Business_Systems_With_Punched_Card_Data_Processing_and_System_3__10_1973 Business Systems With Punched Card Data Processing And System 3 10 1973

Business_Systems_With_Punched_Card_Data_Processing_and_System_3__10_1973 Business_Systems_With_Punched_Card_Data_Processing_and_System_3__10_1973

User Manual: Business_Systems_With_Punched_Card_Data_Processing_and_System_3__10_1973

Open the PDF directly: View PDF .
Page Count: 513

Download
Open PDF In Browser	View PDF

BUSINESS SYSTEMS WITH
PUNCHED CARD
DATA PROCESSING
AND SYSTEM/3 MODEL10

F. R. CRAWFORD

Prentice·Hall, Inc., Englewood Cliffs, N. J.

The author and publisher wish to acknowledge the
International Business Machines Corporation for the
granting of permission to reproduce the following
figures: 4.1,5.1,5.2,5.3,5.8,5.9,5.11,5.13,5.15,
6.6,6.7,6.8,6.10,8.18,8.19,9.1,9.2,10.6,10.15,
10.18,11.1,11.5,11.6,12.5,12.9,12.11,13.1,
13.5,13.6,13.9,13.11,13.16,14.1,15.2,17.1,
17.2,17.5.

To my Dad, who is
always "Just delicious!"

All rights reserved. No part of this book
may be reproduced in any form or by any means
without permission in writing from the publisher.

10 9 8 7 6 5 4 3 2 1
ISBN: 0-13-107698-1
Library of Congress Catalog Card Number: 72-7869
Printed in the United States of America

Prentice-Hall International, Inc., London
Prentice-Hall of Australia, Pty. Ltd., Sydney
Prentice-Hall of Canada, Ltd., Toronto
Prentice-Hall of India Private Limited, New Delhi
Prentice-Hall of Japan, Inc., Tokyo

CONTENTS

PREFACE

BASIC ACCOUNTING PRINCIPLES

Principles of Accounting
Basic Accounting Formula
Posting to Accounts
Income and Expense Accounts
Trial Balance
Closing out Income and Expense Accounts
The Accounting Cycle
The Journal
Ledger Accounts
Chart of Accounts
Journalizing
Posting
Trial Balance
Closing out the Books
Preparing the Balance Sheet
The General Ledger
Glossary of General Accounting Terms
Questions and Exercises

1
4

iii

6
7
8
9

10
10
11
11
14
14
16
17
18
20
21
22

CONTENTS

PUNCHED CARD ACCOUNTI NG

Payroll and Labor Distribution
Source Documents
Payroll Computation
Payroll Writing
Analyze and Distribute Payroll Charges
Accounting Entries
Glossary of Payroll and Labor Distribution Terms
Accounts Payable
Glossary of Accounts Payable Terms
Inventory Control and Material Accounting
Costing.
Accounting Entries
Glossary of Inventory Terms
Questions and Exercises

23
24
27
28
28
28
29
32
33
34
38
40
40
41

PUNCHED CARD ACCOUNTING CONTINUED

Order Writing and Billing
Reasons for Order Writing
Billing
Accounting Entries
Glossary of Order Writing and Billing Terms
Accounts Receivable
Accounts Receivable and the Accounting System
Glossary of Accounts Receivable Terms
Cost Accounting
Relationship to Accounting System
Glossary of Cost Accounting Terms
Manufacturing Control
Management Functions
Glossary of Manufacturing Terms
Questions and Exercises

42
42
44
45
45
46
48
49
50
51
52
53
55
56
57

THE PUNCHED CARD AS A UNIT RECORD

Data Representation for Machines
The 80-Column Card
IBM Card Code
Standard BCD Interchange Code
Card Fields
The 96-Column Card
Questions and Exercises

60
63
66
68
70
72
76

CONTENTS

THE TRANSCRIPTION TO PUNCHED HOLES

80-Column Card Punching
Card Hopper
Punching Station
Stacker
Keyboard
Program Control
Other Features
The Transcription Process
The Machine Verification Process
96-Column Card Punching and Verifying
Punching
Operating Features
Program Control Switch
Program Card
Automatic Duplication
Right Adjust
96-Column Card Verification
Questions and Exercises

78
79
80
80
81
83
84
87
88
89
90
90
92
93
94
94
94
97

PUNCHED CARD DESIGN

Types of Punched Cards
Determining Card Data
Determining Field Size
Determining Data Sequence
Location of Data in Other Cards
Sequence of Data on the Source Document
Machine Characteristics
Final Card Layout
Questions and Exercises

98
102
103
107
107
109
109
115
116

PROCEDURE CONTROL

118

Record Count
Control Total
Unit Count
Self-Checking Number
Modulus 10
Modulus 11
Document Register
Serial Numbers for Batch Control

121
121
123
124
127
129
130
132

CONTENTS

Transmittal and Route Slips
Questions and Exercises

133
134

SORTING AND CLASSIFICATION

135

Machine Operation
Sorting for Printed Reports
Numerical Sorting
Card Handling
Sight Checking
Estimating Sorting Time
Block Sorting
Alphabetic Sorting
Selection Switches
Alphabetic Shortcut Method
Special Sorting Devices
Multiple-Column Selector (IBM 82, 83)
IBM 5486 Card Sorter
Numerical Sorting: 96-Column Cards
Alphabetic Sorting: 96-Column Cards
Counter On (Special Feature)
Digit Select (Special Feature)

137
140
142
144
146
146
147
149
152
153
154
155
157
159
160
163
164

FILE MAINTENANCE

166

IBM 88ColIator
Sequence Checking
Merging Two Files
Merging with Selection
Merging with Equal Selectiort
Other File-Handling Operations
Matching Two Files
Editing
Comparing Two Fields in the Same Card
Selecting Cards by Either of Two Key Numbers
Selecting Zero-Balance Cards
Checking the Filing of Details Behind Master Cards
Questions and Exercises

168
171
175
178
183
184
185
186
187
188
189
190
192

CONTENTS

vii

DATA TRANSFER BY PUNCHED HOLES

193

Gangpunching
Interspersed Gangpunching: X-Punched Masters
Interspersed Gangpunching: X-Punched Details
Checking Interspersed Gangpunching
Offset Gangpunching
Comparing Offset Gangpunching
Selectors
Reproducing
Selective Reproducing
Combined Reproducing and Gangpunching
End Printing
Mark Sensing
Questions and Exercises

194
195
197
199
201
202
203
204
205
206
209
210
214

MACHINE CALCULATING OPERATIONS

218

IBM 521 Card Read Punch
IBM 604 Electronic Calculating Unit
604 Display Panel
604 Operation Planning
Control Panels
Basic Operation
Crossfooting: A - B = ±T
Multiplication: A X B = P
Division: A 7 B = Q
Successive Calculations: (A X B = P) 7 C = R
Sign Control
Approximate Square Root
Gangpunching
Questions and Exercises

218
219
221
223
227
227
227
230
231
233
236
237
239
239

REPORT PREPARATION

245

Methods of Printing
IBM 402 Accounting Machine
Card Reading
Control Panel
Control Panel Wiring
Detail Printing
Accumulating Totals
Program Control
Questions and Exercises

245
246
247
249
252
255
259
263
270

CONTENTS

viii

BASIC CONTROL PANEL WIRING

271

Counter Control
Multiple Selection
Digit Selection
Field Selection
Class Selection
Additional AccotmtingMachine Features
Crossfooting
Progressive Total Printing
Special Program
Summary Punching
Multiple Line Printing
Questions and Exercises

271
277
280
282
284
286
286
287
288
288
292
294

IBM 407 ACCOUNTING MACHINE

298

Multiple Line Read Feature
Storage Units
Tape-Controlled Carriage
Control Tape
Form Design
Questions and Exercises

298
300
302
303
304
308

INTRODUCTION TO COMPUTERS

310

Computer Data Representation
Octal Number System
Integer Conversion
Binary Number Systems
Fraction Conversion
Hexadecimal Number System
Questions and Exercises

313
317
319
315
321
323
326

COMPUTER CODES AND MAIN COMPONENTS

329

Binary Coded Decimal (BCD)
Standard BCD Interchange Code
Extended Binary Coded Decimal Interchange Code (EBCDIC)
American Standard Code for Information Interchange (ASCII)
Storage
Stored Program

329
331
332
333
334
335

CONTENTS

Central Processing Unit
Functional Units
Counters
Machine Cy les
Program Execution
Questions and Exercises

336
337
338
339
340
342

ORGANIZATION OF A UNIT RECORD COMPUTER:
THE IBM SYSTEM/3 MODEL 10

344

Card Input
Card Output
Printed Card Output
Line Printer Output
Processing Unit
Processing Unit Data Flow
Time Sharing
Interrupt
Instruction Format
Printer-Keyboard
Data Entry Keyboard
80-Column Card Compatibility
Questions and Exercises

344
346
347
347
349
351
354
355
356
360
361
361
362

IBM SYSTEM/3 MODEL 10: CARD- AND DISKOR I ENTED SYSTEMS

363

Card-Oriented Programs
Disk-Oriented System
Applications Processing Programs
Questions and Exercises

363
368
375
377

PROGRAM PREPARATION

379

General Considerations
Program Preparation
Program Assembly
Programming Systems
Compilers
COBOL
FORTRAN
Questions and Exercises

380
381
386
388
391
393
394
395

CONTENTS

REPORT PROGRAM GENERATOR, VERSION II

397

File Description
Specification Sheets
Control or Header Card Specifications
File Description Specifications
Input Specifications
Questions and Exercises

398
398
401
404
408
432

RPG II CALCULATION

434

Constants
Use of Result Indicators
Operation Codes
Arithmetic Operations
Compare and Testing Operations
Setting Indicators
Branching Operations
Subroutine Operations
Other Operations
Questions and Exercises

439
444
446
446
456
458
459
462
467
468

RPG 11 OUTPUT-FORMAT SPECIFICATIONS AND
EXAMPLES

469

Output-Format Specifications
Example I: Complete RPG II Program
Example 2: Complete RPG II Program
Questions and Exercises

469
476
483
491

INDEX

492

PREFACE

In recent years, publishers have offered an exceptional number and
variety of books dealing with computers and information processing. Almost
every aspect of these extensive subjects is now in print-including basic data
processing concepts, hardware and software design, application of systems to
problem-solving of all types, and social impact of computers.
However, there seems to have been a tendency to overlook one very important category in this field-the present and prospective users of unit record
equipment. Despite this oversight, thousands of working installations of unit
record equipment businesses, government departments, and military establishments throughout the world. While the total commercial consumption of
punched cards continues to rise every year, it has become somewhat fashionable lately to refer to this versatile data processing medium in the past tense.
Perhaps this is because the punched card was the first universally recognized
symbol of Electronic Data Processing (EDP).
The need to train and educate in this field is urgent. With the introduction of the IBM System/3 Model 10, the possible application of punched
cards, particularly to business problems, has greatly expanded. The new 96column card has provided the greatest increase of speed and flexibility in the
history of unit record data processing.
At the same time, punched cards have been given the versatility of completely automated processing by the stored program computer. The amount
of processing that can be done in a single pass of records through the machine
is no longer restricted by the relatively inflexible control panel or plugboard.

xii

PREFACE

The use of punched cards has continued to expand as an input and out·
put medium for many computers other than the System/3. New subsystems
are available to read and punch cards at speeds more compatible with the processing capabilities of very large systems.
In many businesses, the accepted practices of centralization have given
way to the concept of distributed processing, a policy of placing smaller capacity machines on location where equipment can be utilized by and best serve
the information and accounting needs of that location. Punched card installations are ideal for this purpose. Types of card equipment have been developed that can be connected to a central computer over communication lines.
Thus, the satellite installation takes care of on-the-spot, routine data handling
tasks and can then communicate with the central system to report accumulated activity, maintain central files, or share peak work loads.
No substitute has yet been found for the 80-column punched card as a
"turnaround" record, a record which can be prepared automatically by machine, used by people to record additional information, and can then be processed again by machine with this changed or additional data included. This
concept is almost universally accepted in utility billing and credit card
accounting applications.
This book reviews some basic business practices and shows how these
practices can be carried out with unit record equipment. Although machine
functions and capabilities are described to illustrate these practices, it is not
the primary purpose of this book to prepare the student to actually operate
machines. However, where pertinent, the text will offer illustrations, diagrams, and operating instructions to describe how a machine works. Actual
class exercises can be carried out on machines if such equipment is available.
System/3 functional capability is described to introduce computers and
to show how this system can be adapted to basic data handling and unit
record manipulation procedures. Magnetic disk file function and record organization is described to show how this equipment supplements the machine's. card-handling ability. An introduction to RPG II programming is included to illustrate how the computer is instructed by the programmer to
perform basic tasks.
The author gratefully acknowledges the International Business Machines Corporation for the arrangements which made it possible for him to
draw upon the company's extensive information source·s for material for this
book.
Affectionate acknowledgement is also due to my nice wife, Nancy, who
helped with many hours of proofreading and with the assembly of the manuscript.
Osprey, Florida

F. R. Crawford

BASIC ACCOUNTING

PRINCIPLES

Some knowledge of basic accounting principles is helpful in
understanding the purpose of electronic data processing (EDP) with
unit records. These widely accepted practices are used in most
commercial enterprises. For this reason, the following brief review should provide useful information for those not previously
acquainted with these principles and their many related terms.

Most American businesses use what is called the double-entry system of bookkeeping. As the name implies, two written entries are
made in· accounting records every time a transaction affects a
particular business.
For example, suppose a company buys $1,000 worth of merchandise intended for retail sale and pays cash. To record this
transaction, one entry in the books shows $1,000 as a reduction in
the amount of cash on hand, while a second entry records a $1,000
increase in the value of inventory. This method of record keeping
has become standardized so that when records are kept up to date,
the financial status of a company can be accurately determined at
any time. Also, the results of any company's operation can be
evaluated from available financial data.
Record keeping and reporting have always been necessary
functions associated with the exchange of goods and services
1

Principles
of Accounting

BASIC ACCOUNTING PRINCIPLES

throughout history. As early as 2300 B.C., clay and stone tablets
were used to record payment for services in temples. In the famed
Babylonian empire, tax levy and collection required recorded proof
of individual obligations and payments. The Romans are credited
with introducing the custom of budgeting government expenses.
Bookkeeping and accounting for private business developed
as outgrowths of Italian commerce in the thirteenth century when
Venice and Genoa became great centers of trade. In that era, as
well as in present times, business ventures were financed by loans
and investments of money. Special records and methods of reporting were devised to show the interests of both investors and creditors. At the same time, merchants kept books to record their
financial dealings with both customers and employees. Doubleentry records were used that began the modern accounting practices
of today.
Accounting is a system of keeping the financial records of a
business. The records are always stated in terms of value or money.
If any distinction is to be made between bookkeeping and accounting, bookkeeping may be defined as the routine record-keeping
part of accounting. Much of this routine work can be done automatically by machines, as will be shown in later chapters.
In addition to the making of records, accounting is concerned
with the use of records. For example; various summaries and statements are prepared in all accounting systems. The interpretation
of these statements produces information of vital importance to
the management of a business. The information is also important
to any others who have direct interests in the operation of the
business: bankers and other creditors, stockholders, prospective
investors, and government agencies.
Accountancy is the profession that practices the techniques
of accounting and devises special systems for various types
of businesses. A Certified Public Accountant (CPA) is one who
has fulfilled all legal requirements to hold a public certificate
in· accountancy.
Assets are things that belong to or are possessed by a business.
According to accounting practice, the value of an asset must be
measurable in money. Assets are cash and property, such as land,
buildings, furniture, fixtures, merchandise, and so on. Assets can
also be money owed to a business by other businesses or individuals.
Capital is the original investment in a business, stated as value
in money. Capital may be originated by the owners or by others as
a portion of the assets.

BASIC ACCOUNTING PRINCIPLES

Liabilities. When persons who are not the owners put assets
into a business, they are known as creditors. Contribution by
creditors takes place when the asset cash is loaned to a business or
when merchandise or other assets are sold to it and payment is to
be made at some later date. Creditors who have sold property to
the business expect to get back an equivalent sum of cash at some
future time, in accordance with the terms of sale.
From the time money is loaned until it is paid back, the
creditors are contributors to the assets owned by the enterprise.
Therefore, creditors have an interest in the business and may be
considered temporary investors.
The amounts of money owed to the creditors are liabilities
because a business is liable to the creditors for the sums owed.
Income. Assets are obtained by a business in two ways:
(I) as direct contributions by the owners and creditors and
(2) as a result of operating the business. For example, if an owner
performs a customer service for cash, assets are increased by the
cash amount. The increase in assets through earnings is called
income or revenue.
A business can earn income in various ways. The greater part
of income is usually from its regular activity. For example, a
trading business derives income mainly from the sale of merchandise. A service business such as a garage, a laundry, or a trucking
concern obtains income largely from charges for service. In addition, there may be income in the form of commissions, interest on
amounts owed by customers, rent of premises let to others, and
similar sources.
The increase in assets through the operation of the business
by the owner is a contribution by him and so increases his interest,
the capital. Thus, capital may be increased not only by direct contribution by the owner, but through the earning of income as well.
Expenses. In contrast with the increase in capital by the
earning of income, capital decreases as a result of expenses incurred.
The salaries paid to employees, the rent paid to the landlord, the
interest on indebtedness, and the cost of telephone and other
services are all examples of expenses.
In every business, there are two opposing forces at work to affect the capital: income increases capital, expenses decrease capital.
When expenses are not paid at the time they are incurred but
are to be paid at some future date, the amount of such expenses is
recorded as a liability. Thus, liabilities may increase as a result of
contributions by creditors and by the incurring of business expense.

BASIC ACCOUNTING PRINCIPLES

Net Income or Net Loss. When the total of income, or
revenue, for a specific period of time exceeds the total of expenses
for the same period, the result is net income for the period. Net
income causes an increase in capital.
When the total of expenses for a period exceeds the total of
income, the result is a net loss for the period. A net loss results
in a decrease in capital.
Cash and Accrual. Accounting is practiced on two bases:
cash and accrual.
On a cash basis, income is considered as earned only when
payment is actually received. Expenses are taken into consideration
only when actually paid for.
On an accrual basis, income is considered earned when goods
are sold or services performed, even though payment may not have
been received. Similarly, expenses are taken into consideration
when they are incurred, although they may not be paid.

Basic Accounting
Formula

The basic double-entry accounting formula evolved from the earlier Italian system late in the fifteenth century. It was decided
then that everything owned by a business should be called property. This idea can be expressed as an equation
Property = ownership
If the accounting terms explained in the previous section are substituted, the equation becomes:
Assets

= capital

To illustrate this axiom, assume that a business starts to operate with $10,000 in cash. To record this situation, the first entries
to the accounting records would be made thus:
Assets = Capital
$10,000 = $10,000
This equation must always balance; that is, assets must always
equalcapital.
Now assume that in the course of its operation the business
buys land worth $3,000 and pays in cash. Two entries must

BASIC ACCOUNTING PRINCIPLES

be made to the accounts to record this transaction and to keep the
original equation in balance.
Assets =
Land
Cash

capital

$10,000 = $10,000
+3,000
-3,000
$10,000 = $10,000

Next, the business buys $2,000 worth of inventory for cash:

= capital
$10,000 = $10,000
Assets

Land
Cash
Inventory
Cash

+3,000
-3,000
+2,000
-2,000
$10,000

= $10,000

The business now buys a truck for $4,000. But rather than
pay cash, it is decided to pay at a later date. The business now
assumes a liability, or an account payable. This transaction is
entered as follows:
Assets - liabilities = capital
Land
Cash
Inventory
Cash
Truck

$10,000
+3,000
-3,000
+2,000
-2,000
+4,000
$14,000

$10,000

-$4,000 (accounts payable)
-$4,000 = $10,000

Note that the equation is still balanced. However, a liability
category has been added. Because a liability is something a business owes, this amount must be deducted from the value of the
assets to show the true equity of the ownership account. To do
this, the basic accounting formula is stated as:
Assets -liabilities = capital

BASIC ACCOUNTING PRINCIPLES

or
Assets = liabilities + capital
The above formula shows that a business owns its liabilities
as well as its capital. Capital can also be expressed as net worth.
Therefore, another form of the basic formula is:
Assets

Posting to Accounts

= liabilities + net worth

Accounting transactions could be listed as just shown for an
entire accounting period. However, in this form it would be rather
difficult to see what transactions affected each type of account,
such as cash on hand, accounts payable, accounts receivable, and
so on. It would also be difficult to determine the status of each
account; for example, how much money is owed or due the business at any time.
Much more convenient access to individual accounts can be
maintained by the arrangement shown schematically in Figure 1.1.
Inventory

TI
2,000

Cash

+
10,000

Accounts
payable

Net
worth

-1- 1·-F+ f+

3,000
2,000

Land

3,000

Trucks =

4,0.00.

4,000

10,000

Figure 1.1 Formation of "T" accounts.

Notice that each account is set up in the form of a "T." Asset accounts are considered as being to the left of the equals sign
while liabilities and net worth accounts are to the right. The arrangement makes posting considerably easier, especially when done
manually. Two rules are always followed:
1. For asset accounts, those to the left of the equals sign in
Figure 1.1, posting to the left of the T increases the account. Posting to the right decreases it.
2. For accounts payable and net worth accounts, those to
the right of the equals sign, posting to the left of the T decreases
the account. Posting to the right increases it.

BASIC ACCOUNTING PRINCIPLES

In accounting, posting to the left side of an account is called
a debit entry; posting to the right side is a credit entry. For all
practical purposes, the terms debit and credit mean left and right
postings. The rules may be summarized as follows:
1. For asset accounts, increases are recorded by debits; decreases by credits.
2. For liability and ownership accounts, increases are recorded by credits; decreases by debits.

In a strictly commercial environment, a company engages in
business by selling either service or merchandise for a profit.
Sales increase the net worth account on the ownership side of
the basic equation.
However, in the conduct of every business, there are expenses.
These must be subtracted from net worth. Or, as shown graphically
in Figure 1.2, net worth is increased or decreased by the amount
of the difference (profit or loss) between sales and expenses.

Income and
Expense Accounts

Expense accounts

Figure 1.2 Relationship of net worth to sales and expenses.

Rather than attempt to calculate the effect on net worth of _"
each expense as it occurs, the double-entry system provides for
setting up individual accounts for each category of expense. This
arrangement is shown in Figure 1.3. At the end of an accounting
period, these accounts can be closed out to a profit and loss
account which is then posted to net worth. Thus, for any accounting period, it can be readily determined whether the business is
operating at a profit or loss.
Note that the signs for the expense accounts shown 'in Figure
1.3 are reversed from the normal sign indication on the ownership
side of the equation. That is, the debits are plus while the credits
are minus. Mathematically, when the accounts are closed out to
profit and loss, the signs will be changed by the minus sign in front
of the parenthesis.

BASIC ACCOUNTING PRINCIPLES

To illustrate this process, assume the following transactions:
1. The business makes a $300 sale for which it receives cash.
The cost of goods sold is $150.
2 .. The business receives a bill of $25 for electric power and
pays the bill in cash.
3. A telephone bill is received for $50 but not paid now.
To account for these transactions, the following postings are
made:
1. Cash is increased (debited) $300; sales is increased (credited) $300.
Cost of goods sold is increased (debited) $150; inventory is
reduced (credited) $150.

2. Electricity is increased (debited) $25; cash is decreased
(credited) $25.
3. Telephone is increased (debited) $50; accounts payable
is increased (credited) $50.
Figure 1.3 shows the postings to all accounts as they would
appear in the basic equation.
Accts.

Net

Cost of

31+1o+T+r~1+r~JJr
Figure 1.3 Double entry for expenses.

Trial Balance

Assume now that the transactions shown so far complete the
business activity for an accounting period. The next step is to
prove the accuracy of the double-entry postings; that is, prove that
the debits equal the credits. This is accomplished by taking a
trial balance.
First, individual account balances are calculated by adding the
debit and credit columns of each account and then subtracting the
smaller total from the larger. The trial balance ~s prepared by listing

BASIC ACCOUNTING PRINCIPLES

and totaling the resulting balances, debit or credit, as shown in
Figure 1.4.
Trial balance
Account

Credits

Debits

Cash

Inventory
Land
Trucks
Accounts payable

5,275

1,850
3,000
4,000

Net worth
Sales
Cost of good sold
Electricity
Telephone

150
25
50
$ 14,350

4,000
50
10,000
300

$ 14,350

Figure 1.4 Trial balance.

After the trial balance is taken, adjusting entries can be made
for income earned or for expenses incurred where no entries were
recorded. For example, the accounting period might end in the
middle of a pay period. In this case, an entry is made for pay
earned by the employees but not yet paid. An adjusted trial
balance is then prepared.

The next step is to close out the income and expense accounts Closing Out Income
as shown in Figure 1.5. The sales account is closed out by debiting and
it $300 and crediting the profit and loss account for the corre- Expense Accounts
sponding amount. Each expense account is closed by crediting
Sales -fast of g.s.

3~013~0 \~

Electricity

TelePhone)

~~
Profit & loss account

Cost of g. s.
Elec.
Telephone

(Expense)
150
25
50
225
* 75
300

(Income)
300 Sales

300
300

*The difference between the income and the
expense items is the profit or the loss.

Figure 1.5 Profit and loss account.

BASIC ACCOUNTING PRINCIPLES

with the proper amount and debiting the profit and loss account
with the corresponding amount. The difference between total income and expense is profit or loss. The difference is posted to the
net worth account. Profit credits the account while a loss debits
net worth.

The Accounting
Cycle

The Journal

An example of a typical accounting cycle described in the following section will further explain the concepts described thus far.

As business transactions occur they are normally posted in a
journal, the first entry in the company records. The journal can be

kept in the form of either a book or a sheet. It serves three primary purposes (Figure 1.6):
Journal #1
Date
5/5
5/5
5/5
5/5

....

Description
Inventory
Cash
Accounts receivable
Sales
Cost of goods sold
Inventory
Electricity
Accounts payable

100
100
50
50
20
20
10
10

-~--Figure 1.6 Accounting journal.

1. Pertinent facts about all transactions are shown in the
sequence in which they occur. This is often very important for
reference and auditing purposes.
2. Offsetting debit and credit entries are both shown for
each transaction with a note of explanation. Thus, an identifying
record of all transactions is kept in one location.
3. Possibility of error is reduced by requiring both debit and
credit entries to be shown. If transactions were posted directly to
a ledge~, there would be a chance that a debit or credit entry might
be omitted.

BASIC ACCOUNTING PRINCIPLES

At this point, all transactions have been listed in chronological Ledger Accounts
order. Therefore, it is difficult to isolate postings by any particular
category for reference (for example, entries affecting only inventory
or accounts payable). The next step is to post from the journal to
individual T accounts kept in a ledger. It is then possible to examine all transactions affecting a particular account during any specific
period. This procedure is shown schematically in Figure 1.7.
Journal # 1
Date
5/5

(

V5/5

Description
I nventory

.;
.;

Sales
Cost of goods sold
Inventory
Electricity
Accounts payable

5/5
_

20
20
10

100
50

Accounts 'ec,;v,bl,

5/5

100

Inventory
ledger

Date

Description

REF DR

5/5

Journal #1

100

Cash
ledger

Date

Description

5/5

Journal #1

REF

CR
100

Figure 1.7 Posting from journal to ledger.

A chart of accounts lists all of the accounts to which the com- Chart of Accounts
pany wishes to post transactions. When posting is completed
for a given period, the accounts will be used to prepare financial
statements of the business. Posting to the various accounts is often
referred to as the distribution of amounts. In many businesses, the
number of such transactions runs into the thousands each month.
Processing requires a great deal of effort and expense. A sample
chart of typical accounts follows:
Cash
Capital stock
Land
Equipment
Cost of goods sold
Notes payable

Net worth
Inventory
Profit and Loss
Sales
Accounts payable
Transportation

Accounts receivable
Surplus
Buildings
Equipment
Utility expenses

BASIC ACCOUNTING PRINCIPLES

To classify these accounts, it would be helpful first of all to
assign them to broad categories, then break these categories down
into finer classifications, as follows:
Assets
Liabilities
Net worth
Income
Expense
When the accounts are properly placed under the appropriate
major classification, numbers are assigned to facilitate the recording of entries.
100 Assets
110
111
112
113
114
115
116
120
121
122
123
124
125

Current assets
Cash on hand and in banks
Accounts receivable
Reserve for bad debts
Notes receivable
Marketable securities
Inventory
Fixed assets
Buildings
Depreciation reserve for buildings
Land
Equipment and machinery
Depreciation reserve for
equipment and machinery

200 Lia bilities
210
220

Current liabilities
211 Accounts payable
Long term liabilities
221 Notes payable

300 Net worth
331 Capital stock
332 Surplus
400 -Operating income and expense
Income

410
411 Sales

BASIC ACCOUNTING PRINCIPLES

420
421
422
423
424
430

Expense
Cost of goods sold
Fuel
Electricity
Telephone
Profit and loss

Income and expense accounts are often referred to as temporary accounts because they reflect business activity over a
fairly short period of time and are closed out at the end of the
accounting period. Listed below are examples of income and
expense accounts.
Income

Expenses

Sales
Interest income
Other income

Cost of goods sold
Rent
Insurance
Electricity
Fuel

For purposes of illustration, the following transactions of a
company are shown posted to a journal, then transferred to the
proper ledger accounts. Assume that the company1. purchases $100 worth of inventory and pays for it in cash;
2. makes a sale of $50 (the cost of goods sold is $20);
3. receives an electric bill for $10;
4. receives $50 in cash for a prior sale;
5.

make~

a sale of $150 (the cost of goods sold is $75);

6. buys inventory valued at $5,000;
7. receives a fuel bill for $15;
8. pays a $10 electric bill;
9. pays the $5,000 due for the inventory purchase;
10. makes a sale of $75 (the cost of goods sold is $35);
11. pays a $15 fuel bill;
12. receives a $20 phone bill.

Journalizing

BASIC ACCOUNTING PRINCIPLES

As these transactions occur, they are recorded in the journal.
First, each transaction is examined to determine the way in which
the various asset, liability, and net worth accounts are affected.
The date of the transaction is entered in the date column. Next,
the amount of the transaction is entered to the accounts to be
debited and credited. If required, a note of explanation is included.
A blank line is left after each journal entry.
The posted journal with the above transactions appears in
Figure 1.8. These entries have not yet been posted to the ledger.
Journal
Date

Account debited

Account credited

L.F.

May 5 Finished goods inventory

100
Cash

5 AIR

100
50

Sales
5 Cost of goods sold

50
20

Finished goods inventory
5 Electric

20
10

AlP
5 Cash

10
50

AIR
8 AIR

50
150

Sales
8 ~ost of goods sold

150
75

Finished goods inventory
8 Finished goods inventory

75
5000

AlP
8 Fuel

5000
15

AlP
8 AlP

15
10

Cash
8 AlP

10
5000

Cash
8 AIR

5000
75

Sales
8 Cost of goods sold

75
35

Finished goods inventory
8 AlP

35
15

Cash
8 Telephone

15
20

AlP

Figure 1.8 J oumalizing daily transactions.

Posting

The next step is posting from the journal to the ledger accounts. T accounts are used in the illustration (Figure 1.9), with
debits posted to the left side and credits to the right. Notice that
there are certain balance amounts already in the permanent accounts shown. These are marked with asterisks. The items re-

BASIC ACCOUNTING PRINCIPLES

present money left in the accounts at the close of the previous
accounting period.
Permanent accou nts

Temporary accounts

(Assets & liabilities)

(I ncome & expense)

Cash
*30,000
50

Accounts receivable
100
10
5000
15

Inventory
*6,000
100
5,000

*1000
50
150
75

Accounts payable

20
75
35

10
5000
15

Net worth

600*
10
5000
15
20

Sales

0
150
75

Fuel

COS;;f goods sold

15T

75
35

Electricity

101

Telephone
20

1 36 ,400*

Figure 1.9 Posting from journal to ledger accounts.

1. The first entry debits inventory $100, increasing the value
of the asset account. The offsetting entry credits cash $100, reducing cash by this amount.
2. For the $SO sale, accounts receivable is debited to increase
assets. Sales is credited for the same amount. The merchandise
cost the company $20. Because this is an expense, cost of goods
sold is debited $20. The inventory account is credited with the
same amount.
3. The electricity account is debited $10 to increase expense.
Because the bill is not yet paid, accounts payable is credited $10
to increase liability.
4. For the $SO payment received from a prior sale, the asset
cash is increased by debiting this account $SO. An offsetting entry
of $SO is credited to accounts receivable, reducing this account.
S. For the $ISO sale, accounts receivable is debited to
increase assets; sales is credited to increase income. Since the
cost of goods sold is $7S, a debit is made by that amount.
The offsetting entry is a credit to inventory of $7S, reducing
this asset account.

BASIC ACCOUNTING PRINCIPLES

6. Inventory is debited $5,000 for the value of purchases.
Accounts payable is credited for $5,000 because the company
did not pay cash.
7 . For the $15 fuel bill, the fuel account is debited $15;
accounts payable is also credited for the same amount.
8. As the $10 electricity bill is paid, accounts payable is
debited $10 to reduce liability. The offsetting credit reduces the
asset cash account.
9. When the company pays for the $5,000 inventory purchase, accounts payable is debited $5,000 to reduce liability. The
offsetting entry credits cash for $5,000.
10. For each sale not made for cash, four entries are made:
(a) debit $75 to increase accounts receivable and (b) credit sales
for the same amount to increase income; (c) debit cost of goods
sold for $35 and (d) reduce the value of inventory by crediting the
inventory account for the same amount.
11. When the outstanding $15 fuel bill is paid, accounts payable is debited for $15 to reduce. a liability account and cash is
credited the same amount.
12. The $20 telephone bill is also an expense. The telephone
account is therefore debited and the accounts payable credited
with the offsetting amount of $20.

Trial Balance

To assure the accuracy of all postings to the account ledgers,
a listing is made for each account showing the debit and credit
entries for all transactions posted during the current accounting
period. This listing is called a trial balance. If the period's transactions are properly posted, the total of the debits will equal the
total of credits.
The trial balance shown in Figure 1.10 is made up of the
transactions that were just posted to the T accounts.
For purposes of illustration, the trial balance shows all transactions that have occurred during the period. This is slightly different from common accounting practice which provides for first
determining account balances by adding the debit and credit columns of each account, subtracting the smaller sum from the larger,
and then taking the trial balance by listing and totaling the resultant balances.

BASIC ACCOUNTING PRINCIPLES

Trial balance
Debit

Account

Credit

Sales

Cost of goods sold

Fuel
Electricity
Telephone
Cash

Accounts receivable

Inventory

Accounts payable

20
75
35
15
10
20
50

50
150
75
100
5,000
10
5,000
15
$10,625

50
150
75

100
10
5,000
15
50
20
75
35
10
5,000
15
20
$10,625

Figure 1.10 Final trial balance.

If it is assumed that these are all the transactions for an account- Closing Out
ing period, it is now a major concern of management to deter- the Books
mine what profit, if any, was made. This involves closing out
the books. To arrive at the profit or loss for a corporation during
a given period, the temporary accounts (income and expense) are
used. Again, for purposes of illustration, the statements will be
prepared by working directly from the T accounts. In actual
practice, statements are prepared by working with the account
balances developed on the trial balance (Figure 1.11).
The first step is to add all the figures in the sales account.
The total of $275, when debited to the sales T account, closes it
out. The offsetting entry is a credit to the profit and loss account.
The next area of concern is the cost of goods sold account.
The total is $130. By crediting this amount and debiting to the
profit and loss (P & L) account, the cost of goods sold account is
closed out. The fuel, electricity, and telephone accounts are
handled in the same manner.
The debits in the P & L account represent all the expenses for
the accounting period. The difference between the debits and

BASIC ACCOUNTING PRINCIPLES

credits is therefore the profit or surplus for the accounting period.
In this example, the total is $100. To reflect the transfer of profit
to the net worth section, the P & L account is debited $100 and
the net worth account is credited $100.
Temporary accounts
Cost of goods sold

Sales
50
150
75
275275

20
75
35
130

Fuel

Electricity

~~115 ~
130

Profit & loss
(Cost of goods sold) $130 $275 (Sales)
(Fuel)
15
(Telephone)
20
(Electricity)
__
10-1-_ _
$275
(Profit)

Figure 1.11 Temporary accounts.

The profit and loss account is also called the income statement. The statement shows management the profit made by the
company during the accounting period. In this example, the profit
or surplus represents an increase of $100 in the net worth of
the business.

Preparing the
Balance Sheet

The next step in the closing procedure is to post the asset and
liability accounts to the balance sheet to show the assets, liabilities, and net worth of the company. To do this, individual
accounts are closed out by first adding the debits and credits for
each account. Then the smaller sum is subtracted from the larger
to determine the balance for that particular account.
For example, all the debits in the cash account are added for
a total of $30,050. All the credits are added for a total of $5,125.
The smaller credit balance is subtracted from the debit balance to
produce a remainder of $24,925 debit. Notice that this figure and

BASIC ACCOUNTING PRINCIPLES

all other totals of the various accounts are marked with two asterisks. This is to indicate that these amounts are the closing balances
for the current account period, as well as the opening balances for
the next period.
Permanent accounts
Cash
* 30,000 100
10
50
5,000
15
** 24,925

Accounts receivable

* 1,000
50
150
75
** 1,225

Accounts payable

10
5,000
15

Inventor
* 6,000 20
100 75
5,000 35
** 10,970

Net worth

600 *
10
5,000
15
20
620 **

136 ,400 *

100
36,500 **

Balance sheet
Cash
Accts. receivable
Inventory

$24,925
1,225
10,970
$37,120

Accou nts payable
Net worth

620
36,500

$37,120

Figure 1.12 Permanent accounts and balance sheet.

The debit balance for the accounts receivable account is
$1,225.
For the inventory account, the beginning debit balance of
$6,000 is added to the other entries, $100 and $5,000, to arrive at
total debits of $11,100. The credits are totaled by adding tHe
entries $20, $75, and $35. The difference between the total debits
and total credits is $10,970. This is the closing balance for the
closing period and the opening balance for the next.
The accounts payable credit balance is $620.
The net worth account has a starting balance of $36,400.
When the surplus amount of $100 is added for the current period,
total net worth is increased to $36,500.
Cash, accounts receivable, and inventory (all asset accounts)
are added to arrive at a total of $37,120. The entries are made on
the left side of the balance sheet as shown in Figure 1.12. On the
right side of the balance sheet, accounts payable of $620 (liabilities
to others) and $36,500 (liability to stockholders) are totaled for

BASIC ACCOUNTING PRINCIPLES

$37,120. Thus, assets equal liabilities plus net worth, as stated in
the basic accounting formula.

The General Ledger

When punched card data processing procedures are used, the
general ledger entries are in the form of punched cards rather than
in the form of manual postings on a ledger. Thus, at the end of
the accounting period, either just before or after the completion of
the income statement and balance sheet, a machine-listed general
ledger report is prepared. This report is a summary of all accounts
and their balances. In many cases, it also shows the opening balance of each account and the total debits and credits posted to the
account for the period.
In order to provide a complete historical record in the form
of general ledger accounts, the totals (debit and credit) of the
activity in the temporary income and expense accounts are shown,
even though these accounts were previously closed out to the
P & L account and the profit from it, in turn, was closed out
General ledger

Acct no

110
111
112
116

210
211
300

410
411
420
421
422
423
424
430

Opening
balance

Acct title
Assets
Current assets
Cash
Accounts recvbl
Inventory
Total assets
Liabilities
Current liabilities
Accounts payable
Net worth
Total liabilities
Operating income
& expense
Income
Sales
Total income
Expense
Cost of goods sold
Fuel
Electricity
Telephone
Total expense
Profit & loss

$30,000
1,000
6,000

600
36,400

Debits

Crdts

50
275
5,100

$5,125
50
130

5,025

5,045
100

Closing
balance

$24,925
1,225
10,970
$37,120

620
36,500
37,120

275
275
130
15
10
20

Figure 1.13 General ledger:

175
100

BASIC ACCOUNTING PRINCIPLES

to the net worth account. A sample of the general ledger is shown
in Figure 1.13.
Although the above presentation has been somewhat simplified, the basic steps apply to most accounting procedures. A number of application areas will be summarized in chapters 2 and 3 to
illustrate how the procedures apply.

Asset

Something owned by a company or business, such as cash and ac- Glossary of
counts receivable.
General Accounting

Balance sheet A statement showing in total all the assets owned and all
the liabilities owed by the company.
Chart of accounts
transactions.

A listing of all accounts to which a company will post

Credit An amount posted to an account to decrease it, if an asset, or increase it, if a liability. Credits are posted to the right side of the account.
Debit

An amount posted to an account to increase it, if an asset, or decrease
it, if a liability. Debits are posted to the left side of an account.

Distribution

The allocation of amounts to desired accounts.

Fixed asset Something owned by a company intended for long-term use
and not acquired for resale. The investment in a fixed asset will be recovered through future operations of the business.
Income statement The financial report that shows the profit or loss made
by a company during a particular accounting period.
Journal The book or sheet to which transactions are first posted in chronological sequence.
Ledger The sheet to which entries from the journal are posted. These
sheets are maintained by individual account classification, for example,
cash and inventory.
Liability Money owed by a business.
Net worth The stockholders' equity in the business.
difference between assets and liabilities.

It represents the

T account Another name for ledger account. The T is made by the intersection of the horizontal line under the account name with the vertical
line separating debits from credits.
Trial balance The listing and. totaling of all accounts to insure accuracy.
Debits must equal credits.

Terms

BASIC ACCOUNTING PRINCIPLES

Voucher A business form that shows all the details of a transaction and
authorizes the entry of that transaction into accounting records.

Questions

and
Exercises

1.
2.
3.
4.
5.
6.
7.

What is the main advantage of the double-entry bookkeeping system?
Identify the distinction between accounting and bookkeeping.
Express the basic accounting formula.
Summarize the rules of debit and credit posting to accounts.
What is the purpose of the trial balance? Journal?
List some examples from a typical chart of accounts.
Post the following tra~sactions to a journal in the sequence given, showing
accounts debited and credited.
Sell $50; cost of goods is $24.
Receive cleaning bill of $16.
Sell $200; cost of goods is $145.
Buy inventory worth $2,000.
Receive $29 from previous sale.
Sell $60; cost of goods is $40.
Receive utility bill of $30.
Sell $90; cost of goods is $75.
Pay· $2,000 for inventory purchase.
Receive fuel bill of $120.
Sell $250; cost of goods is $190.
Receive phone bill of $80.
Receive $1,000 from previous sale.

8. Post from the above journal to a ledger.
9. Prepare a trial balance to check posting.
10. Close out books to determine profit or loss for the period. Assume same
previous balances to accounts as shown in Figure 1.12.

PUNCHED CARD

ACCOUNTING

Chapter I outlined general accounting practices more or less
common to commercial enterprises of all types. It can be shown
that the methods of carrying out these practices have also been
gradually standardized throughout the business community, so that
many procedures are similar for such activities as payroll, cost distribution, and inventory control. Individual procedures may vary
considerably in detail, depending upon the business, but the general approach is usually consistent with established practice. In
addition, federal, state, and local regulations for revenue and tax
reporting make it necessary to conform to legal requirements for
providing information to government. Thus, considerable uniformity of systems has been mandated by law.
Chapters 2 and 3 present an overview of basic accounting
procedures which have historically furnished the majority of unit
record applications. A glossary of terms follows each explanation.
Many of the terms will be used later in the text to relate these
accounting procedures to actual data processing steps as they
are described.

The term payroll refers to procedures for recording time worked
by a number of employees and to the calculation and payment of
the resultant wages due. Earnings are reported to each employee,
23

Payroll
and
Labor Distribution

PUNCHED CARD ACCOUNTING

the employer, and to federal and state governments as required by
law. Payroll procedures include the calculation and withholding
of federal income and social security taxes. Local, state, and city
taxes may also be involved, depending upon the place of employment and the residence of the employee.
Payroll is probably the most extensively mechanized procedure of all modern business practices. It is widely accepted as a
standard application of computers and punched card data processing. Payroll is a required procedure wherever people exchange their
time, skills and services for money, not only in business but in
governments, institutions, and military services as well.
All employers must compute employee earnings and maintain records from which wage and tax reports can be prepared.
This is often a voluminous record-handling process involving the
effort and expense of many people. In many enterprises, labor is
the largest and most significant single expense of doing business.
If not properly controlled, this cost can be a major cause of an unprofitable operation or inefficient use of public funds.
Employers must be sensitive to good employer-employee relations. Therefore, it is extremely important for wages to be
computed correctly and paid promptly at the close of the established pay period. At the same time, the system must be designed
to respond readily to changes in individual situations: wage
changes, promotions, location changes, and so on. The operation
of the payroll procedure must reflect the employer's concern for
prompt and accurate remuneration of his employees' services.
Labor distribution is a procedure closely associated with payroll. It concerns the charging, or distribution, of labor costs to the
particular job to which the employee is assigned-by number, account, department, location, et cetera.

Source Documents

Basic steps required to prepare payroll and labor distribution
remain essentially constant, no matter what type or method of
payment is used or what tools are applied to perform the record-keeping function. The first step is to create and maintain
source documents.
1. The personnel or master payroll record includes such typical information as employee name, social security number, and
number of tax exemptions. These basic facts are required for

PUNCHED CARD ACCOUNTING

income tax purposes. In addition, many companies assign an employee serial number for internal identification within their own
organization.
The personnel record usually includes other pertinent statistics
about the employee, including age, marital status, sex, skill or job
code, date employed, assigned location and shift, and so on. These
facts may not only be needed for tax purposes, but for matters of
company policy and union contracts. For example, the employee's
age affects insurance and retirement benefits. Date employed
establishes seniority. Skill or job code may determine work assignment and location.
The record usually identifies the type of payroll involved.
An employee may be salaried, that is, paid regularly by the week,
month, or other period under contract or agreement. Or, an employee may work strictly by the hour at an agreed-upon hourly
rate. Either salary or hourly wages may be affected by overtime,
bonuses, commissions, or production incentives of various kinds.
The master record states the rate of pay for hourly employees or
the base amount of salary per pay period.
The master record can show the employee's work location,
assigned department or division, and even the method of payment
if desired (i.e., by check or in cash).
2. The attendance record, usually a time or clock card, is a
document on which an employee registers time at work. A time
clock is usually provided for this purpose at a location conveniently
near where the employee arrives and leaves at the beginning and end
of the work day. The card clearly identifies the employee,
usually by both man number and name. When the card is inserted
in the clock, the time of day is automatically registered in the
proper position.
Employees are often required to sign the attendance record
at the end of the pay period to signify that the document is a true
record of the time worked. The employee's immediate supervisor
may also endorse the document to testify that the time worked is
also agreed upon by the company management and that the employee is, in fact, properly entitled to compensation.
The time card provides a record of employee absence or lateness, with the reason for any deviation from regular working hours.
The card is a vital record in payroll procedure and may serve as the
basis for settlement of any dispute between employer and employee over wages due.

PUNCHED CARD ACCOUNTING

3. Labor distribution records are prepared. An employee
may spend time on a number of different jobs during the pay
period, or he may be assigned a station on a production or continuous process line. In any case, the time and resultant wages
must be charged to the particular job or jobs at which the employee worked. That is, the labor cost must be distribu ted to the
proper accounts.
Usually, the creation of labor distribution records is the function of a timekeeper, an employee charged by management with
the responsibility of keeping records of how employee time is spent.
The timekeeper often monitors the recording of employee
time on the clock cards and serves as auditor of the attendance
records. Total time distributed for any employee must equal his
clock time for the period.
4. Deduction authorization. In addition to automatically
withholding income taxes, an employer can deduct other amounts
from wages when authorized to do so by an individual employee.
For example, regular payments may be deducted for savings accounts, credit unions, government savings bonds, stock purchase
options, hospitalization and personal insurance, union dues, and
so on. Deductions may be either fixed amounts or a percentage of
gross pay. Or, authorized amounts may be varied in other ways
from one pay period to the next.
An employee may repay an employer by payroll deduction
for advances against wages or for special equipment needed to perform his job but not furnished by the company, such as uniforms,
safety shoes, safety glasses, tools, automobile or truck, et cetera.
A company can be required by law to collect or garnishee a portion of an employee's wages to satisfy creditors.
Records of deduction authorizations must be maintained from
pay period to pay period and inserted in the payroll procedure at
the proper time. Many deduction procedures are supplementary
data processing applications of considerable volume, involving ex..;
tensive record-handling facilities.
A deduction register is prepared at some point in the payroll
run to show an itemized listing of all deductions taken for the
pay period.
5. Personnel records. In many instances there are associated
personnel records that serve as the basis for payroll wage data.
Rate of pay is normally authorized by management through personnel representatives. Pay scales can be established in many ways:

PUNCHED CARD ACCOUNTING

by collective bargaining, by a merit system, civil service scale,
military rating, professional standing, seniority, yearly contract
within a wage structure, and many others. In any case, the rate
of individual compensation is basic data normally received by the
payroll department from some other area within the organization.
Conversely, payroll records are used often in the determination of wage rates which can be affected by such factors
as performance ratings, attendance records, production results,
and job assignment.

When all source records are properly prepared and ready at the end
of the pay period, the payroll computation can begin. This phase
will include at least the following basic steps:

1. Controls are established to assure that all attendance records and other source data are received and properly authorized.
For example, the payroll department accepts a number of time
cards from each department as received from the timekeeper. For
hourly employees, a control total should be immediately set up by
the total number of hours worked by each department. In all subsequent runs, a check back to these totals can verify that all employees have been accounted for at various points during the
procedure. For salaried employees, a man count can be set up
which will serve the same purpose as a control on hours worked.
2. Rates of payor salary amounts are related to each
employee's record of time worked as authorized by the personnel department.
3. Gross earnings are calculated by multiplying hourly rates
times hours worked. Earnings may also include additional pay for
overtime, incentive, bonuses, commissions, and other supplementaryearnings.
4. Federal and local income taxes are calculated for each
employee.
S. Social security taxes are calculated.
6. Miscellaneous deductions are taken, according to the employee's authorization.
7. Net pay is calculated according to the following formula:
Gross - withholding tax - social security tax
- miscellaneous deductions = net pay

Payroll Computation

28
Payroll Writing

PUNCHED CARD ACCOUNTING

A completed payroll includes a number of printing operations:
1. A payroll register is prepared showing all amounts for all
employees on the payroll. At the same time, year-to-date earnings
and taxes are summarized. Totals must balance against preestablished controls.
2. Individual earnings and deduction statements are prepared
for each employee.
3. Checks or cash envelopes are prepared.
4. A deduction register is run and corresponding deduction
accounting is completed.
5. Periodically, federal and local tax reports are printed.
6. Bank account reconciliation is performed.
7. Other management and control reports are made available
as required by the business.

Analyze and
Distribute Payroll
Charges

Accounting Entries

Payroll charges are distributed to orders and accounts covering
manufacturing and operating costs. Expenses can be analyzed by
comparing controllable units with budgets or standards.

The relationship of payroll cost to the general accounts may be
illustrated by the examples shown in Figure 2.1.
a. As payroll expense is incurred, it is charged to the payroll control account. However, until such time as wages are paid,
the company has incurred a liability, so the accrued payroll account
is credited with the $500 payroll cost.
b. Payroll expense must be recovered by charging it to work
in process. In Figure 2.1, it is assumed that $400 of payroll cost
was spent for direct labor. This can be charged directly to a
work-in-process account.
c. It is also assumed that $100 of payroll expense was for
indirect labor, that is, for foremen, timekeepers, and so on. This
labor is not chargeable to a work-iIi-process account. Instead

PUNCHED CARD ACCOUNTING

it is charged to an actual burden account. At the end of the
accounting period, all such costs will be distributed back to the
various production departments and then worked into the cost of
work in process.
d. At the time wages are paid, the liability reflected up to
this point is discharged into the accrued payroll account. This
has the effect of lessening the cash position, as the credit to
cash indicates.
The net result of this series of transactions is to add the cost
of labor to work in process, while lessening the cash position. This
temporary loss should be retrieved when the product is sold.
Payroll expense

Work in process

(A) 500 400 (B)
100 (C)

(B) 4001

Accrued payroll

Actual burden

(D) 5001500 (A)

Cash

1500 (D)

Figure 2.1 Relationship of payroll costs to general accounts.

Accrued payroll The liability for earnings due to employees since the end
of the previous pay period.
Actual cost The purchase or contract price of goods and services used in
manufacturing or other operations involving cost accounting records,
as opposed to estimated or standard costs.
Advance
An amount of money paid to an employee before the customary
time of payment for services.
Attendance card (or clock card) A card showing the time an employee
spends at his place of employment.
Base rate

The hourly rate of pay for an employee.

Bonus Increased earnings or some material reward in excess of regular
earnings granted to an employee for accomplishing results equal to or
in excess of a set standard, or for working less desirable hours or under
less desirable conditions than others.
Burden That part of the cost of producing goods and services which cannot
be directly associated with the items produced.

Glossary of
Payroll and Labor
Distribution Terms

PUNCHED CARD ACCOUNTING

Qock number A number assigned to an employee and placed on his time
card for identification purposes.
Clock station The location of a time recorder where the time cards for
employees are placed.
Continuous job card A card containing time worked and other pertinent
facts concerning the work of one employee on one job order, process,
department, or operation during· successive days within a pay period.
Current earnings card The unit record used to accumulate an employee's
earnings and other pertinent data for a pay period.
Daily time card A card containing attendance time and other pertinent
facts concerning one employee for one day.
Day shift A period of working time for employees which begins in the
morning and ends sometime in the afternoon.
Deduction card A card containing pertinent facts concerning an amount
to be deducted from an employee's pay for such things as contributions, insurance premiums, bond purchases, dues, company purchases
and advances.
Deduction register A list showing the amounts deducted from employees'
earnings, and the reason for each deduction.
Denominating cash payroll Determining the number of bills and coins of
each denomination required to make up each employee's pay with a
minimum number of bills and coins.
Direct labor cost Employees' earnings that are directly applicable to a
job order or process.
Distribution

Assignment of payroll data to the various accounts affected.

Earnings record A record of employee earnings and taxes, both for some
specific pay period and cumulative to date.
Earnings statement A report given an employee, usually at the time he is
paid, showing the factors pertinent to his earnings, taxes, deductions,
and net pay.
Employee or man number A number assigned to an employee for identification purposes. Usually same as clock number.
Gang job card A card containing all pertinent facts about the work of
several employees working as a unit on one job order, process, department, or operation.
Gross earnings card A punched card containing an employee's total earnings before taxes or deductions.

PUNCHED CARD ACCOUNTING

Group bonus plan

A bonus plan whereby the earnings of each employee
are increased when the production by a group of employees with whom
he works is in excess of the set standard.

Idle time

The time during which an employee or machine is available for
work but is not working.

Indirect labor

The work of those employees who contribute to the overall
production of goods and services, but whose work is not directly applied
to the product.

Individual job card

A card containing all pertinent facts about the work
of one employee in one day on one job order, process, department,
or operation.

Job order

A written authorization to perform a specified task.

Master personnel card

For automatic data processing, a punched card containing information about an employee's education, marital status, tax
exemptions, age, sex, company and occupational seniority, salary or
hourly rate, and other factors needed for company records.

Master rate card

A card containing information that sets the rate of pay
for the person or thing identified. An employee's rate is expressed in
terms of money, usually rate per hour. In production or manufacturing,
the rate may be expressed in terms of quantity, for example a standard
number of pieces per hour.

Payroll advice

A memorandum prepared for the payroll department authorizing or giving notice of some change in the status of an employee.

Payroll deduction

An amount authorized to be subtracted from an employee's earnings before payment is made by the employer.

Payroll master card

A card with constant or semiconstant data concerning an employee: name, serial number, social security number, occupation, rate of pay.
Payroll register A detailed list prepared for each pay period of all employees
paid for that period. The register usually contains the same information
shown on the employee's payroll check and earnings statement.

Reconciliation of bank statement

A comparison of the bank statement
with the cash records of the business to see whether they are in agree·
ment. Adjustments must be made for any discrepancies.

Standard cost

An accounting system which uses predetermined costs of
labor and material required to produce goods and services.

Time sheet

A list containing the names or man numbers of a group of
employees, showing the time worked by each employee for a day, a
week, or a pay period.

PUNCHED CARD ACCOUNTING

Unit cost

The purchase price or manufacturing cost of an item.

Variance

The difference between actual and standard measurement.

Work-in-process file A file of punched cards containing labor distribution
and material requisition cards for commodities which are being
manufactured.

Accounts Payable

Accounts payable is the term applied to the procedure of recording and accounting for the amount of money a business owes for
goods and services received from other businesses or individuals.
The procedure is established to record what is owed, to pay indebtedness promptly and thus receive allowed discounts, and to
keep management informed regarding the amount and purpose
of expenditures.
Most businesses follow a general procedure which involves
the following steps:
1. Requisitions for goods or services are approved by designated management authority.

2. Formal purchase orders are forwarded to suppliers as
authorization to ship merchandise or perform services.
3. Goods or services are received.
4. Invoices and purchase orders are audited for agreement
while actual receipt of goods or services is verified.
5. Invoices are placed in a file for payment on due'date.
6. Checks and check register are prepared on due date.
7. Distribution reports are prepared showing expense by
account number for later use in general ledger and cost accounting.
8. Various management reports are prepared, for example,
to show purchases by individual vendor.
Accounts payable has a close relationship to many other
accounting applications. The source data for accounts payable is
usually entered on a voucher which consolidates all the data necessary for payments and distributions. After processing, the data
developed provides input for cost, plant and equipment, inventory
and material, and general ledger accounting. Various accounts are
affected by accounts payable transactions. For example:

PUNCHED CARD ACCOUNTING

a. As goods are received, the invoice is audited and approved
for payment on the due date. The accounts payable account is
credited $50 to increase liability, as shown in the example in
Figure 2.2 Inventory is debited $50 to increase the asset account.
b. Payment is made on the due date. If by the terms of the
invoice, the business is allowed a cash discount of 10%, accounts
payable is debited $50 to decrease the liability account. At the
same time, cash is credited $49 and discount is credited $1. Discount is considered other income because the business did not
have to payout $50 but only $49. The discount entry balances
the double entry, making debits equal credits.
Accou nts payable
(8) 50.00150.00 (A)

Inventory
(A) 50.001

Discount

Cash

149.00 (8)

1,.00 (8)

Figure 2.2 Posting for amount received on account.

Accounts payable distribution reports Distribution of paid expenses by account number according to the chart of accounts.

Glossary of
Accounts Payable

Accounts payable voucher The document used to consolidate all data necessary for payment. Sources of data are purchase orders, vendors' invoices, and material receipts.

Terms

Allowances Concessions or reductions against a~ invoice brought about by
varying conditions and reasons, and agreed to between the purchaser
and the seller.
Anticipation discount An amount which the customer may subtract if he
chooses to pay an invoice before its due date. The amount is usually a
fixed percent of the invoice total, the percent agreed upon between
buyer and seller.
Apron A form attached to vendors' invoices with space for management
approval, vendor code number, voucher number, account distribution
code, and amount.
Cash disbursement register

A listing of paid vendors' invoices.

Credit

An increase in indebtedness to a vendor's account.

Debit

A decrease in indebtedness to a vendor's account.

PUNCHED CARD ACCOUNTING

Due date The date on which, according to the terms and date on the
invoice, payment must be made.
Invoice register

A listing of vendors' invoices approved for payment.

Returns Items of merchandise sent back to a vendor for one reason or
another. Credit is given for its corresponding value or original charge.
Short shipment
billed.

A shipment that proves to contain less than the quantity

Sticker Like an apron, but gummed so that it can be pasted to either the
face or the back of the vendor's invoice.
Trade discount Discount earned by special type of business. For example,
an automobile parts manufacturer will give a discount to a wholesaler
but charge the catalog price to a garage.
Trial balance A listing of outstanding liabilities by vendor.

I nventory Control
and
Material Accounting

Inventory control is a procedure to determine the amount and
type of parts, supplies, and finished goods that will most effectively protect the production, sales, and financial requirements of
a business. Material accounting is concerned with the financial
or money control of inventory.
Much of the capital of a business is likely to be invested in
inventory. Therefore, management should have answers to the
following questions for effective control:
What is in stock?
What is on order?
Will it meet requirements?
When should we reorder?
Is it too much?
How much should we reorder?
What is in Stock?

There are two basic m~thods of keeping inventory:
1. Physical Count.
Each item in inventory is manually
cqunted to determine how much is in stock. However, when inventory is taken periodically, the amount of stock on hand can be
known accurately only after the count has been completed. In

PUNCHED CARD ACCOUNTING

most industries, physical inventory is taken no more than once or
twice a year. With large inventories, it is impractical to count in
shorter periods because of the excessive time and expense involved.
2. Perpetual Inventory. A record is kept of each item in
stock to be controlled. The primary record is the inventory balance card, updated as each tansaction occurs. When items are received, the quantity is added to the record of balance on hand.
When items are withdrawn from stock, the quantity is subtracted
from the record of what is on hand. Thus, the record reflects the
stock balance at all times, as long as the transactions are accurately
posted to the balance card. The formula is:
Old balance + receipt~ - issues

= on hand

With the perpetual book inventory system, a physical inventory is taken at least once a year as a veFification of the accumulated balances. Where very large inventories are maintained, a
continual manual check of selected items may be made, for example, for items with exceptionally high value or volume of transactions. In addition, items may be selected at random for counting as a further check on the functioning of the system.
What is on Order?

Besides maintaining records which show inventory on hand,
it is also necessary to control the material on order. This is true
both for purchased items and items to be manufactured by the
business. This information is required because any decision to obtain additional inventory must be based upon a knowledge of any
open orders. The basic formula now becomes:
Old balance + receipts - issues = on hand
On hand + on order = available
Will It Meet Requirements?

Once the quantity of all items on hand is known, it must be
determined whether this amount is sufficient to meet orders and
thus prevent loss of sales and customers. Also. a minimum allowable balance should be. decided upon for each item of stock so that

PUNCHED CARD ACCOUNTING

the item will not be completely depleted before more can be obtained. The formula can be expanded to include these factors.
Old balance + receipts = issues
On hand - minimum balance = reorder
On hand + on order - planned issues = available
When Should We Reorder?

To provide assurance against out-of-stock conditions, two
reorder methods can be used:
1. Stock clerks can watch the amount of stock on the shelves
to see that it does not get too low. Generally this method is relatively inefficient, especially when there are many different items.
2. Minimums can be established and recorded on stock balance records to signal when items should be reordered. Factors
used to determine minimums area. past usage, such as average sales for a specified period;
b. the total time, usually called lead time, needed to
place the order, process it, and receive delivery;
c. a margin of safety that should be maintained in case
the ordered amount does not arrive on time or in case
usage is greater than anticipated.
The standard formula for establishing minimum order points
is:
(total sales for a specified period X lead time)
Minimum = --'-----------':';...----"'---------'specified period

+ safety factor
For example, assume that usage of a given item for a fourweek period is 10,000 pieces and that lead time to obtain an
additional supply is two weeks. Substituting in the above formula,
the minimum balance is determined to 5,000 pieces, or exactly
two weeks supply. This is exactly the time it takes to have more
items in stock. As a safety factor, therefore, it might be considered
that an extra week's supply, or 2,500 pieces, might be sufficient.
Thus, the total minimumm balance to be kept in stock becomes
5,000 plus 2,500 pieces or 7,500.

PUNCHED CARD ACCOUNTING

Is It Too Much?

Because the yearly cost of maintaining inventory can be 25%
or more of the total dollar value of the inventory, it is important
not to stock too much.
The cost elements that make up the commonly quoted 25%
figure are:
Obsolescence
Interest
Depreciation
Miscellaneous
Total

10%
6%
5%
4%.
25%

These factors are variable, depending upon the type of inventory involved. For example, an inventory of bottled liquors
does not become obsolete nor is it often subject to depreciation.
On the other hand, breakage costs might very well run higher than
4%. An inventory of wearing apparel, however, is clearly subject to
drastic obsolescence from style changes alone. Current economic
conditions affect other costs, such as interest and warehouse
maintenance.
How Much Should We Reorder?

In the retail and wholesale industries, a standard quantity is
usually ordered when the minimum order point is reached. Among
the factors that determine this quantity are:

1. Vendor Discount for Size of Order. It is usually true that
the larger the quantity ordered, the lower the unit purchase price.
However, a vendor may have established his most economical lot
size and offer discounts for orders in increments of this lot quantity. For example, 5,000 bushels of grain may fill one freight car
and is therefore the most economical lot quantity to handle and
ship. An order for 5,250 bushels may require a great deal of
special handling and result in a much higher unit price for the
extra 250 bushels.
2. Storage and Handling Costs. This cost is largely determined by the size and weight of the item relative to cost. For example, the handling and storage cost for major appliances is a sig-

PUNCHED CARD ACCOUNTING

nificant factor in the cost of distributing them to the customer.
Probably the same costs for diamonds is insignificant, relative to
their value. The perishable nature of a product may also be a factor in determining the best order quantity, for example fresh fruits
and produce.
3. Order Processing Costs. This cost may be particularly
significant where elaborate procedures are followed to process
each order. The cost of preparing the order may exceed the cost
of the item to be stocked. This is particularly true for small quantities of hardware items such as nuts, bolts, screws, and washers.
4. Danger of Obsolescence. When obsolescence is a factor in
inventory, considerable judgment must be exercised in the maintenance of stock balances. In some cases obsolescence can be
anticipated by the introduction of new models, techniques, and
products. Changes in style, consumer tastes, and weather conditions cannot always be readily evaluated in terms of their effect
on inventory.

Costing

The preceding discussion was mainly concerned with the quantity
control of inventory. The cost aspect must also be considered.
Costing will provide answers to two additional questions:
What is the value of the inventory on hand?
How much should we cost each item when
it is issued or sold?
In dealing with finished goods inventory, there are three predominant methods of costing:
1. A verage Cost. The total dollar value of an item in stock is
divided by the total quantity. For example, if there are 100 pieces
in inventory and their total value is $100.00, the average unit cost
is $1.00. However, if 100 more pieces were purchased for $120.00,
the average unit cost would be $1.10, the result of dividing $220.00
by 200 pieces.
2. FIFO (First-In, First-Out). The material acquired first is
used first and is casted accordingly. This method demands that
each purchased lot be accounted for separately. For instance,
assume there are three 1~O-piece lots of an item in inventory with

PUNCHED CARD ACCOUNTING

unit costs of $1.00, $1.20, and $1.30, respectively. If 110 pieces
are sold, the costing of the sale is:
100 pieces for $1.00 each = $100.00
10 pieces for $1.20 each = $ 12.00
$112.00
Therefore, remaining in inventory are
90 pieces for $1.20 each
100 pieces for $1.30 each
The FIFO costing method is especially applicable in those
industries where receipts into inventory are in relatively large
quantities and where each lot can be conveniently costed
individually.

3. LIFO (Last-In, First-Out). As the name implies, material
sold is costed according to the lot most recently received. When
that lot is exhausted, the cost of the next most recent lot is used.
This requires that each incoming shipment be maintained as a distinct lot in keeping the records.
Under the LIFO method, if a sale of 110 pieces of the above
item were made, the costing procedure would be:
100 pieces for $1.30 = $130.00
10 pieces for $1.20 = $ 12.00
$142.00
The LIFO method has been used more and more widely in
recent years. In periods of rising costs, the method tends to reduce
the amount of stated profit and resultant taxes.
Finished goods inventory
(Receipt) 200

Finished goods inventory
75 (Issue)

Accounts payable

200

Cost of goods sold
75

Figure 2.3 Finished goods and accounts payable posting.

PUNCHED CARD ACCOUNTING

Accounting Entries

The value of an inventory must be reflected in the accounting
books of a company. When merchandise is received, the inventory
account is increased. Since inventory is an asset, a debit entry will
increase that account, as shown in Figure 2.3.
When a sale is made, the value of inventory is decreased by
the cost of the items withdrawn. Therefore, a sale of merchandise
costing $75 credits the finished goods inventory and debits cost of
goods sold by the same amount.

Glossary of

Available The quantity of stock on hand, plus the quantity on order, minus
the quantity reserved for specific purposes.

Inventory Terms

Average cost The cost of each piece of an item in inventory, arrived at by
dividing the total dollar value of the inventory by the number of pieces
in inventory.
Balance card A punched card on which the quantity of stock on hand is
recorded. The card may also include other quantities, such as maximum and minimum, reservations, issues, receipts, and available stock.
Cycle inventory taking A variation of complete physical inventory, and
often used to augment it, in which checking of actual stock against
stock records is done on a rotating basis, a portion of the inventory
at a time.
EOQ (economical order quantity) Use primarily in manufacturing organizations, EOQ is the result of a formula which incorporates the pertinent
factors required to determine the most economical number of pieces to
be ordered or manufactured when stock reaches the reorder level.
FIFO (first-in, first-out) A method of costing goods issued using the cost of
the goods received first as the cost of the goods issued first.
Finished stock Items or products which have been manufactured or purchased and are ready for sale or use without further manufacturing or
processing.
Issues

The amount of inventory released for production or sale.

Lead time The number of hours or days necessary to place an order,
process it, and receive the material in inventory.
LIFO (last-in, fzrst-out) A method of costing goods issued, using the cost
of goods received last as the cost of goods issued first.
Perpetual book inventory
balances.

Maintaining an up-to-date record of all inventory

PUNCHED CARD ACCOUNTING

Physical count
Receipts
Requisition

An actual count of all pieces of stock in an inventory.

Merchandise or stock that is received in inventory.
Written authorization to release items from stock.

Reservation A means of assuring that certain required quantities of stock
will remain available for some future need.
Standard order quantity A fixed number of pieces ordered when the reorder level for an item is reached.
Stock on hand The quantity of any item or commodity actually located
in stockroom and available for use or issue.
Stock status report A report prepared by various means to show by item
the quantity on hand, quantity on order, quantity reserved or set up as
minimum inventory and thus available for issue, and any other pertinent data.
Transaction register A list of transactions, such as issues, receipts, and adjustments, which affect the balance of material on hand.

1. Define payroll and labor distribution procedures.
2. List five primary payroll and labor distribution records used in nearly
all procedures.
3. Outline some important controls that can be established over payroll
procedures. What purposes are served by these controls?
4. Review the accounting entries showing the relationship of payroll cost
to general accounts.
5. What is the purpose of an accounts payable procedure?
6. Show some possible types of accounting controls for an accounts payable procedure. State the purpose of each.
7. Give reasons why an efficient accounts payable procedure is good business practice.
8. Describe two basic methods of keeping inventory.
9. Determine the minimum requirements for an item in stock if 200 pieces
are used every five weeks. Lead time is three weeks.
10. List factors that determine proper order quantity.

Questions

and
Exercises

PUNCHED CARD ACCOUNTING
CONTINUED

Order Writing
and Billing

No company offering commodities or services for sale can operate
without receiving orders. When an order is obtained, a company's
first objective is to disburse to the buyer the commodities or services ordered in the fastest, most economical manner possible and
then to be paid accordingly.
Following sections explain how an order is handled after it is
received by the seller and the means taken to request payment
from the customer.

Reasons for
Order Writing

An order is often received in handwritten form, written either
by the customer or an authorized sales representative. After an
order is received, the factory or warehouse must be notified to
produce or ship whatever is requested by the customer.
The source document used for this purpose can be the order
itself. However, when orders originate from salesmen or from
many different customers, the document usually does not follow
any standard form. Such orders are apt to be illegible, confusing
in nomenclature, incomplete in description, and irregular in appearance. If this document is processed, it can result in incorrect
shipments, lost time, and a generally unsatisfactory operation.
To avoid this possibility, many companies rewrite orders received from customers or sales representatives. By established pro42

PUNCHED CARD ACCOUNTING CONTINUED

cedure, a new document is prepared which describes the goods or
services ordered by the customer.
A complete customer's order should include ship-to and soldto name and address, date, order number, quantity ordered, description of merchandise, and such special information as shipping,
packing, and marking instructions.
An order writing procedure offers the following advantages:
1. A standard form convenient to use
2. Assured legibility
3. Correct and complete description
4. Correct pricing and packing
In addition, many by-products are obtained as follows:
1. Acknowledgment to customers that their orders have
been received. A company with an order writing procedure will
generally send the customer the original copy of the order when
it is written.
2. Shipping list or packing list. Some companies require a
list before commodities are assembled and packed for shipment.
The list shows the details concerning shipping units, that is, the
number of cans, cases, carloads, etc. for each item.
3. Shipping labels, usually attached to each shipping unit
and showing the name and address of the customer.
4. A bill of lading, to authorize a rail or motor transportation
company to move a commodity from vendor to customer. The
document shows the destination, route, freight class, and gross
weight.
S. Back order control. When a customer's order cannot be
filled, or can only be partially filled, the vendor may prepare a back
order to advise the customer what items have to be sent at some
later date. The material to eventually fill the back order is either
on order or will have to be ordered.
6. Order analysis. This is one of the more important benefits obtained by management from an order writing procedure.
It usually consists of a daily tabulation of all orders received, showing totals by commodity. The report gives management an up-to-

PUNCHED CARD ACCOUNTING CONTINUED

date picture of how sales are moving. In addition, the order
analysis report should give ample warning so that production and/or
purchasing can be geared to handle the volume of orders on hand.
Order analysis can be combined with tabulations of shipments to
report on efficiency in filling orders. Periodic reviews of this
order-shipment report can greatly influence production and purchasing policies, produce a more efficient and economical operation, and help keep back orders to a minimum.

Billing

After goods are shipped or services rendered, the vendor must
notify the customer and state the charges. A document is needed
to show the goods or services purchased by the buyer, the name of
the buyer, the seller's name, and the terms.
A typical invoice for goods sold is composed of essentially
three types of information:
1. Heading information. This includes customer name and
address, customer number, terms, branch and warehouse, salesman, and how shipped.
2. Miscellaneous data. Information can include customer's order number, invoice number, date, and special shipping instructions.
3. Body. Commodity information is listed here-quantity
and pack, description, unit price, and selling price with discount
if applicable.
Much of the data used in a billing operation is repetitive because many companies normally deal with the same customers
over and over again and the products sold remain fairly constant.
Billing can be a by-product of order writing where order
writing cards are reused, or it may be a separate procedure. Some
companies produce an. order and keep the original copy on file.
When items are shipped to a customer they pull the original order;
write in quantity shipped; perform the extension of quantity times
unit price; total the invoice; fill in such information as invoice
number, invoice date, and shipping instructions; and then use the
order as an invoice to be sent to the customer.
With any billing application there is always a method of
handling credit memorandums. When merchandise is returned to
a vendor for any legitimate reason, a credit memo is issued to the
customer along with a check for the credited amount or a state-

PUNCHED CARD ACCOUNTING CONTINUED

ment reducing the indebtedness. The preparation of credit memos
parallels that of invoices and usually takes much the same form.
By the same token, a company will have to issue debit memos if a
customer has been charged less than the correct amount.

After an invoice is prepared, the following accounts are affected:

Accounting Entries

1. Accounts receivable is debited for the amount of an invoice
total; sales is credited for the same amount.
2. Inventory is credited with the total cost of the items withdrawn; cost of goods sold is debited for the same amount. (See
Figure 3.1.)
Accou nts receivable
Invoice total

Inventory

I Cr.

Sales
Tnvoice total

Cost of goods sold
Dr.

Figure 3.1 Accounting entries for billing.

The sale of any commodity is reflected by many accounting
functions. For example, totals appearing on an invoice and recorded in an accounts receivable register are shown on the customer
statement of account and are included in a monthly aged trial
balance of accounts receivable. The trial balance is a report showing a list of customers accounts which are past due and the amount
owed by date.
In addition, the stock record of quantity on hand must be reduced by the quantity of the individual item sold. This transaction
is reflected in the stock status report. Finally, the desired sales
analysis and a number of sales reports (by salesman, by customer,
by product, by geographical location) all originate from individual
sales transactions. Because of this chain reaction, accuracy is very
important.

Acknowledgment Notice to the customer that his order has been received
by the vendor.
Back order An order prepared to cover items which cannot be included in
the original shipment, but which will be sent when available.

Glossary of
Order Writing
and
Billing Terms

PUNCHED CARD ACCOUNTING CONTINUED

Bill of lading Document which the vendor must produce to give a transportation company authority to move a commodity from vendor to customer. The document must show destination, routing, freight class,
and gross weight.
Credit memo Document issued to a customer detailing merchandise he has
returned from a vendor, or other adjustments for which cash is owed
by the vendor to the customer.
Debit memo Document representing a charge to the customer for corrections, additions, and special or other unusual charges.
Description Details required to identify a given merchandise item or
commodity.
Heading information Customer data used in preparing the shipping order
and invoice heading.
Invoice Document that describes the commodity purchased and indicates
name of buyer, name of seller, terms, and amount of sale.
Miscellaneous data Variable invoice details that are peculiar to a given
sale and cannot be predetermined.

Order

Document describing goods or services required by a customer.

Order analysis Classification and summarization of facts about products
or customers from orders not yet filled.
Ship to Heading which calls for the name and address to which merchandise will be delivered.
Shipping label Form usually attached to each container in a shipment,
showing the name and address of the customer who will receive the
shipment.
Shipping list List of commodities packed or assembled for shipment, showing details concerning units to be shipped.
Sold to

Heading which calls for the name and address of the buyer.

Source document

Original record of a transaction.

Terms Condition on which the sale is made; for example, cash, net 30 days,
cash on delivery, etc.
Vendor

Accounts Receivable

An alternate term for seller.

It has been estimated that 90% of American business is conducted

on a credit basis. Only 10% of the customers in the United States
pay for their commodities and services in cash. This widespread

PUNCHED CARD ACCOUNTING CONTINUED

acceptance of credit places an additional burden upon nearly every
type of business to establish and maintain a strict control of customer indebtedness. This record-keeping task is known as accounts
receivable. As an accounting procedure, it is readily adaptable to
unit-record data processing methods.
Accounts receivable appears on a company's statement of
condition as an asset, and may be sold if desired. Many companies
do this to maintain working capital. For example, when a customer buys an appliance from the local dealer, a contract may be
signed arranging to pay for the purchase in regular monthly installments. The dealer commonly sells this contract to a bank or finance company at a discount thereby converting this account
receivable to cash almost immediately. The bank or finance
company also takes over the record-keeping task from the dealer
and collects the indebtedness from the customer.
Customers have widely varied paying habits. If a business is
to realize a satisfactory conversion of accounts receivable to cash,
it must know both the credit status of prospective customers and
the paying habits of current customers.
The general objectives of the accounts receivable procedure
are1. to collect money owed in the shortest possible period
of time;
2. to reduce loss from bad debts;
3. to maintain customer goodwill through prompt and accurate updating of payment records.
Regardless of the accounts receivable system used, the following overall steps must be taken:
1. The customer's account is established. This may involve
a credit check by either the business or a credit agency. The
customer's identity is established with name and address, type of
account, and usually a credit limit.
2. A record is set up within the procedure which will be used
to record transactions to the account as they occur. Transactions
are purchases, payments, returns, and adjustments.
3. The accuracy of entries is controlled so that the business
knows at all times how much the customer owes and when payment is due.

PUNCHED CARD ACCOUNTING CONTINUED

4. Total accounts receivable control is maintained and a collection procedure is established.

Accounts Receivable

It has been established that accounts receivable is an asset that,

and

in most cases, can be readily converted to cash. The following
example shows how typical accounts receivable transactions are
handled on a company's books. The T account postings are shown
in Figure 3.2.

the Accounting
System

Accounts receivable

$500
$350
$850
350

Sales

$500
$350

$500
$500

Cash

$4W1

$850

Discounts allowed

$50
$50

Figure 3.2 Posting for accounts receivable.

1. A sale of $500 is made. Sales is credited for $500, reflecting income in that amount. Accounts receivable is debited to
establish an asset of $500.
2. The customer pays his bill and is entitled to a 10% discount. Since he has discharged his obligation to the business,
accounts receivable is credited $500. The payment increases cash,
therefore the cash account is debited $450. The difference between
cash and accounts receivable is represented by the $50 discount.
This represents an expense of doing business which is reflected on
the books by debiting discounts allowed, an expense account.
3. A sale of $350 is made.
If these three transactions represented an entire month's
business, the effect on the financial statements at the end of the
month would be as follows:
1. Accounts receivable is closed out to the balance sheet as
a current asset.
2. Cash of $450 is closed out to the balance sheet as a current
asset.

PUNCHED CARD ACCOUNTING CONTINUED

3. Sales of $850 are closed out to the profit and loss statement as income.
4. Discount allowed of $50 is closed out to the profit and
loss statement as expense.

Accounts receivable ledger The overall record of customer indebtedness.
The system used might be a ledger book, a me of individual customer
ledger cards, a ftle of IBM punched cards, a 'magnetic tape file, or
other means.
Aged trial balance Same as a trial balance except that open items are listed
in separate groups according to age; for example, all items billed up to
thirty days ago, all items billed 30 and 60 days ago, and all items billed
more than 90 days ago. This report enables management to quickly
analyze problem accounts and take appropriate action.
Allowance An adjustment to a customer's bill, generally authorizing additional credit.
Cashier Generally, the person assigned the responsibility of accounting for
and controlling customer payments, either by cash or check.
Control sheet Document, generally posted daily with summary totals from
other reports, and used to prove that all entries affecting the accounts
receivable ledger have been properly posted and that the accounts
receivable ledger itself is correct.
Credit memo Document authorizing credit to a customer because of damaged merchandise, a billing error, returned goods, etc.
Current balance

Amount owed by a customer at any given time.

Cycle statement writing The practice of subdividing the entire accounts receivable me into alphabetic groups, and rendering statements to each
group on different dates. This eliminates the peak loads that occur
when all statements are rendered at one time.
Debit memo Document increasing the original amount of an invoice because of a billing error, shipment error, etc.
Entry date Date on which an invoice, payment, or adjustment is entered
into the accounts receivable file.
Invoice register Daily listing of invoice totals by customer. It also shows indicative and classification data such as date, invoice number, terms, etc.
Journal voucher Internal document used to make miscellaneous entries to
accounts receivable.

Glossary of
Accounts Receivable
Terms

PUNCHED CARD ACCOUNTING CONTINUED

Open item

A bill that has not been paid.

Partial payment

A customer remittance covering only part of a bill.

Remittance statement Document prepared by the customer, and enclosed
with his check, to describe the invoice(s) being paid. It generally sJ:1ows
invoice number(s), invoice amount(s), and discount(s), and is used by
the vendor to properly credit the customer's account.
Return

Merchandise returned by the. customer to the vendor for credit.

Statement Document periodically sent by the vendor to the customer (fre·
quently at month's end) which shows the total amount owed to the
vendor on unpaid bills.
Trial balance Periodic listing of all open items to prove that the accounts
receivable ledger is in balance with the control sheet.

Cost Accounting

Cost accounting is concerned with the identification and accumulation of the costs of manufacturing and marketing a product. This
section will deal only with the cost of manufacturing, which
consists of:
1. The cost of raw material, parts, hardware, etc., used in
making the article
2. The cost of labor used directly to manufacture the article
3. A fair share of the manufacturing overhead, i.e.,
a. indirect labor (e.g., supervisors, clerks, accountants, and
management in general)
b. indirect material (e.g., paint, wire, wrapping and packing materials)
c. power and heat, gas, compressed air, etc.
d. depreciation of plant and equipment
This can be stated in formula form:
Cost = direct material + direct labor + apportioned overhead
By accumulating unit costs for products, management can
analyze the profitability of these products in comparison with
selling price and can recognize inefficient production methods or
the need to revise the selling price.
When records are kept of costs by operation and job, this
information can be used for pricing new products, for estimating

PUNCHED CARD ACCOUNTING CONTINUED

the cost of future products, and for determining economical ordering quantities. By keeping track of costs by department, it is
possible to use this information to establish budgets for coming
periods of operation.
By comparing current cost figures to preestablished estimates
or standards, inefficient areas can be spotted for management to
take corrective action.
The objectives of cost accounting may be summarized as
follows:
1. Accumulation and identification of costs of pt:oduction for
each job completed
2. Accumulation of costs for each area or department
3. Making the proper accounting entries for the work flow
through the plant
4. Preparation of management reports

As material moves from raw material inventory into work in process 'and ultimately into finished goods inventory, associated accounting entries are made to indicate the work flow. In like
manner, entries must be made to record the expending of labor
and other expenses as they are incurred. The following four transactions illustrate this: _
1. As labor is-expended in the production departments, the
value of work in process is increased by the amount of this labor.
The debit to direct labor work in process accomplishes this increase
as shown in Figure 3.3. An offsetting credit to payroll increases
the accumulation of the amount owed to employees, a liability
account.
2. As material is entered into the production cycle, it also
increases the value of the work in process. A debit'to direct material work in process reflects this increase as &lso shown in Figure
3.3. The offsetting credit to stores indicates a ,reduction in raw
material supply, an asset account.
3. As expenses such as heat and light are incu!fed, they also
contribute to the production effort and increase the value of work
in process. This is called overhead or burden, the third;cost factor.
A debit to manufacturing overhead work in process increases this

Relationship to
Accounting System

PUNCHED CARD ACCOUNTING CONTINUED

account to show the incurred overhead expenses. The offsetting
credit to accounts payable increases the amount owed to suppliers.
Other sources of overhead would be indirect labor and indirect
material.
4. As products are completed and leave the production floor
to be placed in stock and held for sale, corresponding accounting
entries are made. The debit to finished goods inventory reflects
an increase in this account. The offsetting credit to the three work
in process accounts indicates a reduction in the work in process.
Direct labor
Payroll

W/P

100

I 80

100

Direct material
Stores

W/P

50
1

(8) 50

Mfg. overhead

W/P

Accounts payable

I 20

Finished goods
140

Figure 3.3 Posting for cost accounting.

Glossary of Cost
Accounting Terms

Burden

Same as Overhead

Cost accounting The maintenance of detailed records of expenditures incurred in producing goods or rendering services. All expenditures must
ultimately be related to a finished product or service so that the total
cost of each unit can be computed.
Direct labor Labor which directly affects or changes the nature of a product on which it is performed.
Direct material Any material which is used in the manufacture of a product
and which forms an integral part of that product. It can easily be identified and its usage in the product can be readily determined.
Indirect labor Labor which is necessary to the operation of the factory,
but which does not directly affect or change the product. Supervision,
inspection, maintenance, and c1ericallabor are examples.

PUNCHED CARD ACCOUNTING CONTINUED

Indirect material Material used in the manufacture of a product but which
is not an integral part of the product and is used in indefinite quantities.
Job order cost A system of cost accounting used in industries producing
goods in production lots, each of whose specifications differs from
those of the next lot.
Job record The medium for recording information pertaining to a particular job, operation, or process. It is contrasted with attendance record,
which records the time an employee spends in the plant.
Job time Time an employee spends working on a particular job, operation,
or process.
Overhead All expenses of factory operation not directly chargeable to a
particular product. It includes indirect material, indirect labor, and
other manufacturing expense.
Process cost A system of cost accounting used in industries producing
similar type goods in a relatively continuous flow, such as steel strip,
wire, and plate or sheet glass.
Standard cost Costs determined in advance of production and used as a
measure of the efficiency of actual production.

Manufacturing is fundamentally a process through which people
and machines convert goods and materials from one form and value
to another. The manufacturing process usually adds functional
utility of some kind to materials or finished parts.
Nearly all manufacturing can be classified under one of the
following general types of industries:
Basic production
Basic conversion
Fabrication
Standard assembly
Assembly with options
Assembly to specifications

Basic Production. This type, of industry exploits natural resources to obtain raw materials for other manufacturers. A steel
industry mines ore to produce pig iron which it then combines
with other materials and scrap to produce steel ingots. Other examples are producers of aluminum, wood pulp, glass, petroleum,
rubber, and mined or extracted chemicals.

Manufacturing
Control

PUNCHED CARD ACCOUNTING CONTINUED

Basic Conversion. The converter changes the products of the
basic producer into a variety of industrial and consumer products.
For example, the steel mill changes ingots into bars, tubing, rails,
pipe, strip, wire, and structural shapes. Other converter products
include paper, lumber, chemical derivatives, and plastics.
Fabrication. The raw products of the converter are transformed into a larger variety of products. Steel strip is stamped into
bottle caps. Rods become 'nuts, bolts, and twist drills. Wire becomes nails and paper clips. Paper becomes bags, boxes, and newspapers. Glass is formed into bottles, light bulbs, and windshields.
Standard Assembly. In this process the products of the
fabricators are made into finished goods of several or many parts.
Well-known examples include radios, television sets, automobile
engines, and many kinds of appliances.
Assembly with Options. Basic products and devices are
assembled with a given range of options that can be chosen in
various combinations by the customer. Examples include automobiles, airplanes, boats, and data processing equipment.
Assembly to Specifications. The manufacturer produces to
the exact specifications of the customer. This category includes
military hardware of all kinds, space exploration vehicles, ballistic
missiles and defense systems, and atomic reactors.
Manufacturing control is a means of expressing the planned
output of a manufacturing enterprise in terms of the required
quantities and values of materials, facilities, and manpower. Means
must be established to obtain sufficient information to control
these functions.
The following are some of the main objectives of a manufacturing control system:
1. To provide the right quantities of materials and machines
at the proper time. Future demands must be estimated and
material planned for delivery to produce the anticipated amount
of finished goods. Since most companies carry some inventory,
what is on hand must also be related to what will be required. If
it is not enough, additional material or parts must be ordered from
suppliers. This may involve considerable lead time because the
supplier may, in turn, have to produce components for delivery.
To do this, he may have to allow lead time from his supplier, and
so on.

PUNCHED CARD ACCOUNTING CONTINUED

2. To have ample manpower available with proper skills where
needed. This includes both direct and indirect labor according to
the job requirements of the plant.
3. To produce the products on schedule as called for by salesmen, distributors, or government agencies. This involves maximum
utilization of machine time and facilities with manpower skills.
4. To produce all products in the most economical manner
possible. Production management must have some system of record keeping in order to keep track of the progress of plant production and for an adequate system of cost accounting.

An effective manufacturing control system must be exceptionally
well planned. From the planning of facilities to delivery of the
finished product, at least the following management functions must
be performed:
Forecasting
Materials planning
Inventory management
Scheduling
Dispatching
Operations evaluation
Forecasting. The forecast brings into consideration every
known indicator concerning future market acceptance of the product. Essentially, the forecast attempts to predict the maximum
quantity at the most profitable price. Also, some estimate is usually made about the length of time the product will be accepted by
the consumer.
Materials Planning. Based upon the forecast, a plan for the
procurement of materials must be produced. These may include
raw or fabricated goods, or even partially assembled items for inclusion in a larger and more complex product.
Inventory Management. The proper management of inventories of both raw materials and finished goods is the means of
carrying out materials planning.
Scheduling. A plan for facilities, equipment, and manpower
must be integrated in such a way that the manufacturing cycle can
begin and then continue according to design.

Management
Functions

PUNCHED CARD ACCOUNTING CONTINUED

Dispatching. There must be a method of keeping the schedule functioning. Dispatching applies manpower to the schedule.
I t also provides a means for making minor adjustments to compensate for errors in the plan or changing conditions in its execution.
Operations Evaluation. This function evaluates execution of
all plans on a continuing basis. It should signal in advance whenever revisions may be called for to meet the original objectives.

Glossary of
Manufacturing Terms

Assembly Two or more parts or subassemblies put together as a unit to
perform some specific function.
Engineering change A change in the material components or production
methods of a part or assembly. The change may be made to improve
performance or quality, to bring a product up to predetermined specifications, or to reduce the cost.
Machine tool load planning The functions involved in allocating the manufacturing load of a plant to fit the available machine tool capabilities.
Manufacturing or shop order The authority for the manufacturing floor
to produce a given quantity of a part, subassembly, or assembly. The
order also lists the operational steps required to produce the units; for
example, cut, stamp, broach, mill, and polish. When each operation is
completed, the order is posted and usually accompanies the job to the
next operation. The completed order furnishes a record of the progress
of the job through the plant.
Master bill of material List of the components or raw materials needed
to manufacture a product or assembly.
Material or production explosion The quantity ordered times the number
of component parts needed to manufacture one product or assembly.
Parts list

List of all the parts required to make one product or assembly.

Piece or component part
to make one part.

Raw material to which is added labor and burden

Raw material Material in its basic state; for example, iron, wool, aluminum
tubing, stainless steel sheets, various chemicals, cereal grains, etc. In
many cases material used by one company to fabricate into other products is termed raw material. The raw material can be the finished product of another company. For example, bar stock steel is the finished
product of a steel mill but raw material to a tool manufacturer.
Requirements planning Process by which a manufacturer determines the
necessary quantities of product components (either parts or assemblies)

PUNCHED CARD ACCOUNTING CONTINUED

so that they will be available in time for final assembly of the planned
number of finished products.

Requirements, gross Total amount of material required to produce a stated
quantity of finished goods.
Requirements, net Equal to gross requirements minus the amount on hand.
Net requirements represents the additional amount of materials which
must be obtained to produce a stated quantity of finished goods.
Sales forecast An estimate of future sales based on historical, current, and
other pertinent data.
Subassembly Two or more parts or units which are incorporated into a subsequent assembly.
Where-used file A me in part number sequence with a reference to all
products where each part is used.

1. Why is it often necessary to rewrite orders received from customers or
sales representatives? Can you think of any alternate to this requirement?
2. Name some of the accounting byproducts that can be obtained from a
well-organized order writing procedure.
3. List the three most common types of information found on billing
documents.
4. What are credit memorandums? Why are they used?
5. What accounting entries are made after an invoice is prepared?
6. List as many effects of a single sales transaction as you can.
7. Why are accounts receivable sometimes sold? Give exam pIes of this
type of business.
8. List the main steps of an accounts receivable procedure.
9. In a commercial enterprise, which accounting function is more important
to the business, accounts receivable or accounts payable? Why?
10. Name some general types of industries. What function do they all have
in common?
11. Name main management functions in a manufacturing environment. Explain why at least two of these functions are most important.

Questions

and
Exercises

THE PUNCHED CARD
AS A UNIT RECORD

The idea of punching holes in paper to record information
was conceived for the textile industry nearly 250 years ago. In
1725 the Frenchman Bouchon devised a crude automatic control
for draw looms by using a length of paper pierced with holes to
represent a woven pattern. The paper was pressed by hand against
a row of horizontal needles so that the needles opposite the blank
spaces were pushed back while the remaining needles passed
through the holes. This action was mechanically transferred to the
positioning of the vertical or warp threads in the 100m to form a
pattern.
In 1728 an M. Falcon adapted the punched-hole idea to perforated cards, which he used to replace Bouchon's length of paper.
To save space, Falcon arranged the needles or horizontal wires in
multiple rows or ranks. He devised a square prism or "cylinder"
for the cards where they could actuate the loom, one card at a
time. Each card was placed in this "sensing" device by hand.
The next development was by an accomplished mechanic
named Jacques de Vaucanson. He simplified the mechanism of the
draw looms of that period and returned to the use of punched patterns on a continuous band of paper. Vaucanson also built an
automatic feeding cylinder fitted with a rack wheel to pass one
tooth for each change of the threads to automatically form the
pattern. However, the operation of the wheel was quite complex
and this device is not known to have been adopted.
58

THE PUNCHED CARD AS A UNIT RECORD

In about 1790, a working mechanic of Lyons named Joseph
Marie Jacquard invented a fishnet loom that won him an award
from the London Society of Arts. While in Paris to receive the
award, Jacquard was engaged by the Conservatorium of Arts and
Sciences. Here, he later found the opportunity to make improvements in the weaving machine by studying the work of Bouchon,
Falcon, and Vaucanson. Apparently he combined the best qualities of his predecessors to produce the industrial loom that still
bears his name (Figure 4.1). The machine was so successful that
by 1812 there were an estimated eleven thousand looms in France.
The loom was declared public property in 1806; as a reward for his
accomplishment, M. Jacquard was given a pension and a royalty on
every machine. Four years after his death a statue was erected in
his honor in Lyons.

Figure 4.1 Jacquard loom.

THE PUNCHED CARD AS A UNIT RECORD

To set up the machine for weaving, a pattern was first laid out
on ruled graph paper to represent the designer's original weave
draft. The draft showed at least one repeat or complete weave
unit of the fabric to be produced. From the drawing, one card
was punched out to represent each throw of the shuttle carrying
the horizontal or weft thread through the design. Machines were
also invented to mechanically punch and duplicate the cards.
To operate the loom, cards were laced together with loops of
string to form a continuous chain. As the cards moved into place
on the cylinder (actually a quadrangular block), some needles
passed through the punched holes and thereby raised the corresponding warp threads. Where there were no holes, needles were
pushed back by spring action and the corresponding warp threads
were not raised. Any woven pattern could be repeated indefinitely
simply by using the same cards over and over. By adding several
Jacquard attachments to one loom, a weaver could produce not
only intricately figured fabrics but also pictures of considerable
size. One of the most .elaborate patterns woven at that time was
a famous portrait of M. Jacquard himself. It required 24,000
punched cards.
This unique use of information stored as punched holes was
undoubtedly one of the earliest examples of an artificial memory.
That is, information was stored in such a way that it could be
made directly accessible to a machine without human intervention
of any kind. This important concept is a basic function of both
the design and operation of modern data processing equipment.

Data Representation
for Machines

The task of efficient data management has been a perennial business problem since the beginning of organized commerce and industry. In chapter 3 it was shown how very large quantities of
data can originate in almost any modern business environment. It
is hardly necessary to po~nt out that the same problem is also acute
in every government, scientific, and military endeavor. The "paper
blizzard" is everywhere.
The best way to cope with this situation is to process data
with machines. Since the majority of products and services reaching the market today are produced by more or less automated
methods, it would seem to follow that the processing of information should also be carried out more efficiently by machine.
However, in considering this solution to the problem, it is also

THE PUNCHED CARD AS A UNIT RECORD

readily apparent that most data does not originate in a form that
is suitable for mechanical or electronic manipulation. Therefore,
the first step in automated processing is to record data in such a
way as to make machine processing practical. There are two basic
requirements for any system of data representation.
First: The method used must be one of preestablished convention. More simply, there must be some sort of specific agreement about how such representation is to be formed, used, and
understood by machines or people or both. Data and information
are usually represented for people as symbols (Figure 4.2).

IT'

(

EXIT

)

y
Figure 4.2 Data represented by symbols.

THE PUNCHED CARD AS A UNIT RECORD

For example, the printed characters on this page are really
symbols. When understood, they convey meaning to the reader in
a manner established by education, training, and experience. A
symbol may convey one meaning to some people, a different meaning to others, and no meaning at all to those who do not understand its significance. But written symbol representation is not
generally suitable for machine use (although specialized machines
have been developed recently to read both printed and handwritten
symbols directly from paper documents).
Second: The method of representation used must be characteristic of the vehicle of communication. For example, when data
are transmitted over wire, the message is represented as a pattern
of precisely generated electrical impulses. That is, the data are
coded suitably for this type of transmission in a form that can be
sent and received by machines. Data are represented by some
instruments as graphs on ruled paper. Others display data on
gauges, dials, or cathode ray tubes.
We are told that some primitive peoples developed the ability
to talk to each other over long distances by beating on drums or
forming smoke signals. These conventions were devised for rather
unusual vehicles of communication.
It is therefore a general rule that whenever data are to be
processed by a machine, the data must first be reduced to some
exactly specified set of symbols, codes, or patterns which can be
readily sensed and interpreted by a machine. These symbols or
patterns usually differ from those commonly used by people, because convention must conform to the design and operation of the
machine. The choice of data representation and the way the data
will be interpreted is a matter of convention on the part of engineers, designers, and manufacturers.
Dr. Herman Hollerith is reported to have conceived the idea
of recording U.S. census data in cards while traveling by train. He
noticed the conductor punching holes in the railway tickets in a
certain pattern to identify each passenger and destination. The
holes represented specific items of information which the conductor could interpret according to the rules or conventions he had
established.
Dr. Hollerith recognized yet another important aspect of machine data processing. There must be a standard medium in which
to record information, a medium of practical size and form made
from some material of kn9wn characteristics. The early French
inventors of the automatic loom had called for the "best obtain-

THE PUNCHED CARD AS A UNIT RECORD

able paper with a weight of about 16 or 18 cards to the pound."
As machines become more and more advanced in data handling capability, other recording media have gained wide acceptance. Punched paper tape and magnetic plastic tape are
commonly used by computers to read and write data at extremely
high speeds. Many computers use still more advanced recording
media in the form of magnetic surfaces on rapidly revolving metal
disks and drums.

Modern punched-card documents are still often called Hollerith
cards, after their inventor. It is also reported that when Dr. Hollerith
was asked what size card should be used in his new machines, he
casually chose the dimensions of a U.S. Treasury note he happened
to have in his pocket at the time. Whether or not this story is true,
the standard 80-column card is the exact size of the old United
States paper money:

7~ inches long by 3* inches wide.

More frequently, the name IBM card is used, referring to the
International Business Machines Corporation, the largest manufacturer of card-operated equipment. When first used commercially,
IBM cards were punched with round holes. In 1928, the card
capacity was increased from 45 to 80 columns by using smaller,
rectangular holes. IBM had concentrated on the development and
marketing of electromechanical machines. The rectangular shape
provided better electrical contact through the holes as the cards
were fed under metal sensing brushes.
It is always a minor point of semantics whether punch card or
punched card is the proper term. Either may be used. Punch card
is used in the general sense to relate the blank document to the
punching machine which produces the holes. Punched card designates a card after it has been perforated with holes ready for some
sort of machine processing. Card machine operators sometimes
refer to their documents as tabulating cards, after the tabulator, the
first machine to produce printed reports from punched cards.
All card paper stock is manufactured to rigid specifications
to guard against impurities and to insure uniform characteristics
under operating conditions. The stock is 0.007 inch thick and must
be electrically nonconducting.
The actual punching area is divided across the length of the
card into 80 equal spaces or vertical columns, numbered from left

The SO-Column Card

THE PUNCHED CARD AS A UNIT RECORD

to right (Figure 4.3). Each column is subdivided into 12 punching
positions: 0 through 9, 11, and 12. The 11 and 12 positions are at
the top of each column, above the 0 through 9 positions. These
two positions are sometimes referred to as the X and R positions.

rnJ 0 0 00 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 [Q]
1

23456789~~a~Mffiffin~~w~nnM~RnR~~~nmM~~m~~~~aa«~%a~~~~~~M~~D~~~~a~~~~n~~MnnnMftRnn~U

III 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 lIt 1 1 1 1 11 1 1 1 1 11

111 1 1 111 1 1 11 1 1 1 1 1 1 1 1 11 1 1 1 1 1 1 1 11 1 1 11 1 1 1 1 1 1 1 1 1 1 1 1 1 III

12]2222222222222222222222222222222222222222222222222222 2 2222222222222222222222222 [Z]

[~3 J 33333333333333333333333333333333333333333 J J 3 3 3333333333333 J 3 3 3 3 3 3 33333333333@]
~444444444444444444444444444444444444444444444444444444444444444444444444444444~
lID 5 5 55 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 5 @J
&1666666666666666666666666666666666666666666666666666 6 6 6 666666666666666666666666!ID

iz17 7 7 7 7 7 7 1 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 7 iIi
:IDa a a8 a a a a a a a a 8 a a a 8 8 a a a 8 8 8 a a a a a a 88 a 8 8888888 a 8 8 8888 a 8 8 Ba B8 a 8 a 888888 a a 88888888888 8 ~
~999999999999999999999999999999999999999999999999999999999999999999999999999999~
1 2 3 4 5 6 7 8 9 10 ~ 12 13 14 15 16 n 18 192021 22232425262728293031 323334353637 3C 39 40 41 42 4344 45 46 47 48495051 52535455565758596061 62 6364656667686970 71

n 73 74 75 76 77

7879

i·o

Figure 4.3 80-column card.

The 12 punching positions form equally spaced horizontal
rows. The bottom row contains all the 9 positions, the next row
above all the 8 positions, the next all 7 s, and so on. The rectangular holes are always punched out precisely at the intersections of
rows and columns. Therefore, a single card document has as many
as 80 X 12, or 960 possible hole positions. However, in actual use
the card is never completely punched out.
A punching position can be assigned almost any predetermined significance, provided the chosen recording convention can
be recognized by the equipment that is to process the cards.
When a hole is punched, it means that the value, data, or other
significance is recorded. When a punching position is left blank,
it means the information is not recorded. Most important, the
punched hole is permanent. Recorded information cannot be
changed or erased. Of course all recording is restricted to the
960-hole capacity of the card.
Punched cards are unit records because a card is always
punched, read, or otherwise processed as a single unit of information. However, if capacity permits, more than one record can be
punched in a single card.
A group of cards, usually forming a given set of data, is called
a deck. After cards have been placed in some orderly sequence for

THE PUNCHED CARD AS A UNIT RECORD

processing, the deck becomes a file: the payroll file, the accounts
receivable file, the address file, etc.
Because of their exact uniformity, a means must be provided
to keep all cards in a deck or file facing in one direction, rightsideup. This is particularly important for machine handling. For this
purpose, cards are made with specified corner cuts. Any card
out of position can be detected immediately when a corner cut
does not match other properly aligned cards in the file, as shown
in Figure 4.4.

Figure 4.4 Opposing corner cuts in a card file.

Cards are also available from the manufacturers in solid colors
or with distinctive colored stripes. The colors are useful in identifying different types of cards in the same file. For example, in a
name and address file, the name cards may be salmon with an upper
right corner cut; the address cards may be plain manila with an
upper left corner cut. In this case the corner cuts and colors serve
to identify each separate name and address item in a file. Colors
are used only for clerical and manual manipulation of cards;
colors are not recognized by machines.

IBM Card Code

THE PUNCHED CARD AS A UNIT RECORD

The standard IBM card code uses the twelve possible punching
positions in a vertical column to represent the digits 0 through 9,
the letters of the alphabet, and the special characters of punctuation and report printing. The digits 0 through 9 are represented by
single holes in the numeric rows. For example, to represent the
digit 1, a hole is punched in the one row. To represent the digit 5,
a hole is punched in the five row, but not in the same column with
the 1 just punched. To represent 0, a hole is punched in the zero
row, also in a different column. Therefore, to record the quantity
150, three holes are punched in adjacent columns: 1,5, and O.
Refer to Figure 4.5. The choice of columns is a matter of card
design and convenience) a subject that will be discussed later.

00000010000000000000000000000000000000000000000000000000000000000000000000000000
1234S6189n"aaHgffinmmW~nUM~~nu~wmDDM.~D~~~~au«aa~~~~~~~M~~"~~~~~~~~~D~~ronnnMgmnMffi~

111 11111 11111111111111 11111 111 1111111111 111111111 1 1 1 1 11111 1111 11 1111111111111 111
222222222222222222222222222222222222222222222222222222222222222222~2222222222222

33333333333333333333333333333333333333333333333333333333333333333333333333333333

44444444444444444444444444444444444444444444444444444444444444444444444444444444
5555515555555555555555555555555555555555555555555555 5 5 55555555555555555555555555
66666666666566666666666666666666666666666666666666666666666666666666666666666666

17 7 77 71 7 7 77 7 7 7 7 7 7 7 77 7 7 7 7 7 7 7 77 7 7 7 71 7 7 7 77 7 7 7 77 7 7 77 7 7 7 77 7 7 77 7 7 7 77 7 7 7 77 7 71 7 7 77 7 7 7 7 71
88888888888888888888888888888888888888888888888888888888888888888888888888888888
99999999999999999999999999999999999999999999999999999999999999999999999999999999

i23456189n"naHgffinm~W~nn~~~nu~wmmDM.~D~~w~au«aa~~~~~~~M~~"~~~~~~~~~D~~ronnaM.mnnffi~
118M 50Bl!

Figure 4.5 Digits 1-5-0 punched in adjacent card columns.

To represent the letters of the alphabet, the card is thought of
as being divided into two sections-the numeric section just explained, and a zone section. The zone section is made up of the
zero, eleven, and twelve rows. Note that the zero row is common
to both the numeric and zone sections of the card. To represent
the letter A, two holes are punched in a single column: 12 and 1.
To represent B, two holes are also punched: 12 and 2. The letter
C is 12 and 3; D is 12 and 4; and so on through the letter I, which
is 12 and 9.
The eleven row is used in combination with the numeric hole
positions to represent the letters J through R; holes 11 and 1 represent J, holes 11 and 2 represent K, and so on. The zero row in
combination with numeric positions is used to represent the letters

THE PUNCHED CARD AS A UNIT RECORD

S through Z. Since there are 26 letters of the alphabet and 27 possible zone and numeric combinations, the combination of holes 0
and 1 is omitted in the coding scheme. Holes 0 and 2 represent
the letter S; holes 0 and 9 represent Z.
To record alphabetic information, adjacent card columns are
normally chosen, depending upon the format of the record. To
record the name "Jones," for example, five card columns are required, one for each letter in the name. Refer to Figure 4.6.
I
III
00000000000000000000000100000000000000000000000000000000000000000000000000000000
123456789m"aNU~rnRm~wnnn~~~VU~ID~nn~~~n~~~~~~«~~o~~~~~~~~~~~~~~~~~~~D~~ronnnu~~nMnoo

111 1111 1 111 1 111 1 11 111 11 1 1 1 11 1 1 11 1 1 1 1 111 1 11 1 1 1 11 1 1 11 1 1 111 1 1 11 1 1 111 1 1 11 1 1 1 11 1 lit t

222222222222222222222221222222222222222 22222222222222222222222222222222222222222
33333333333333333333333333333333333333333333333333333333333333333333333333333333

44444444444444444444444444444444444444444444444444444444444444444444444444444444
55555555555555555555511555555555555555555555555555555 5 5 5 5 55555555555555555555555
6666666666666666666616666666666666666666666666666666 666666666666666666666 S 6 6 6 6 6 6
17 7 77 7 7 7 777 7 n 7 7 7 77 7 7 77 7 7 77 77 7 7 77 7 7 77 7 7 7 77 7 7 7 77 7 7 77 77 7 7 7 7 7 77 7 7 7 77 7 7 77 7 7 77 7 7 7 7 77 1
88888888888888888888888888888888888888888888888888888888888888888888888888888888
99999999999999999999999999999999999999999999999999999999999999999999999999999999
I 2 3 4 5 6 7 8 9 10 " 11 13 14 15 16 R 18 1910 11 22232' 1526272& 2930 31 323334 35

37 38 39 40 414243« 454647484950 51 52535455565) 58 5960 61 62 63646566676869 )0 n

n 73 74 75 76 n 787980

IIBM 50B11

Figure 4.6 Letters J-O-N-E-S punched in adjacent card columns.

PRINTED
BY THE 26
PRI NTI NG

SPECIAL
LETTERS
CHARACTERS
_ _ _ _~A~---__..
r';

DIGITS

__
:!~l-"- (r'-r.'~"·l'"'"'''''''' 'U~' J~~:!" 0"'" ::;":~::, ;;;:::::~:"'" ~

~=-=~-!.--.L ,)1,",:.=;1'';(7'3')

CARD PUNCH

I
I

"""

221

11121

3333

333333

44444

4444444

444

4444444444444141444444444444444 44 444444
55555 U555 555 55 5 55 55 5 55 515 55155 55 555155

;2~';';

.", ., , , , , "'""",, , , , ,

,g~

CARD COLUMN r /
DESIGNATION;...-f--;;

DIGIT PUNCHES

~=---:-:::-:~::-::-:-::~~:""---;:-;;:;--;:;:-;;:-;;;:::;:;-~~-;-!"";1-;:;::;rifdl

'"""".61 taGS J911 11 iJJ4 IS 16 II Ja

111111111111111111111111111111111111111

1111112212 2 11111 2 2 12 2 11 2111111211111111
333333333333333331333331333333333 333333

555555

55555555

5555

555

66 6 6 6 6 6

666656666

66666

6666

666666666666666666666 61s 6 6 6666666 665666

1 JJ 1 1 1 1 1

1111111111

111111

11111

1 1 1 1 1 1 Jl11 1 1 1 1 1 1 1 1 1111 1 1 1 1 JJ 1 11 11 1 1 1 1 1 1

888888888

88888888888

SSSSSS8

388838

888888888888888888888888Banasc88Bs88888

CARD COlUM N r-L...-. ~ ~ ~ ~ ; ~ ~ ~ ~ ~o 11 9 ~J ~t ~ ~~ ~; ~3 ~~;G;I

9 99999999

9999999999999999991999999999199999199999999999

111
:: 1& ;) 28 n :03: 21 33;~ 35 3ii 31 38 3~ 40 41 11 43 ~t 4i ~i 41 68 ~~ ~9 51515) 54555& 51 5~ S~ 60 61 61 GJ ~J ~5 6!!:N6:F:S~ r~ll.~l~/~ i~ 1) 1$ II 19 1930
DESIGNATIONr--L=-~':"""':':"~'':::,:OB<~ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _--':"'_~~_~--'

A CARD COLUMN

Figure 4.7 80-column card code.

A CARD FIELD

THE PUNCHED CARD AS A UNIT RECORD

Figure 4.7 shows the complete IBM card code as just described, plus two additional zone punches to represent the ampersand (&) and minus or dash (-) characters. The printing at the top
of the card is done by a machine called a printing keypunch, a keyoperated device that prints the information corresponding to the
holes in each column.

Standard BCD
I nterchange Code

The previously described IBM card code is accepted as the standard punched hole representation of alphabetic and numeric data
when used with IBM unit-record equipment. However, punched
cards also serve as an important data recording medium for many
computer systems. To provide compatibility of data for interchange among these various systems, the IBM card code has been
extended to represent a total of 64 different characters. Some of
these characters are useful in the representation of mathematical
formulas and in message transmission over long-distance communication equipment. Other characters mark the end of records,
fields, and blocks of information recorded on magnetic and paper
tape. Some characters, such as the familiar $, #, *, etc., are used
when printing reports, checks, invoices, statements, and the like.
A list of these special characters with their proper names is shown
in Figure 4.8.
The letters BCD stand for binary coded decimal, a method of
coding information. It is related to the binary method of arithmetic which is widely used in the internal computation and data handling units of computers. Figure 4.9 shows the complete BCD code
in collating sequence; that is, the sequence in which computers can
be programmed to sort data. Note that the first column of the
illustration shows the standard characters or graphics that are
coded. Five of the standard BCD combinations print out as either
of two characters, depending upon the type set used in the printing
device. The two variations are called graphic subset 1 and graphic
subset 2 (Figure 4.10).
Graphic subset 1 is used primarily for computer report writing
and most commercial applications. Graphic subset 2 is used when
writing or printing advanced programming languages for computers
as it meets general requirements for mathematical symbolism.
The second column of Figure 4.9 shows the punched hole
combinations for each graphic. The third column shows the bit
code representation for internal computer storage. This column is

THE PUNCHED CARD AS A UNIT RECORD

CHARACTER
Report I Program
low-'

b
0
[

SYMBOL

NAME

Group Mark

:to

Recotd Mark

ill-

Segment Mark

Word Separator

12-3-8
12-4-8

A
A

12-5-8

B
B
B

8
8
8

12-6-8

4
4

12-7-8

B
B
B

8
8
8

B
B
B

8
8

At Sign

Number Sign

;
.:l

11-6-8
11-7-8

Ampersand

Plus

Asterisk

Percent

/
\
0

Slash

Blank

I
i

i
I

%
~

(

Backslash

()

j
I

Substitute Blank

(

Left Parenthesis

)

Right Parenthesis

C
C

3-8
4-8

A
A

8
8

4
4

8
8

4
4

A
A

8
8

5-8

6-8
7-8

4
4

12-0

8
8
8

[

Left Bracket
Right Bracket

Tape Mark

A
B
C

12-3
12-4
12-5
12-6

12-7
12-8
12-9

Greather than

Equal to

U
Z

;

Semicolon

Colon

::>

oUJ

Ul
(!)

z
f=
<{

Minus or Hyphen (Dash)

Delta

-l
-l

o
U

D
E

F
G
H
I
!

J
K

12-1

12-2

B
B
B

A
A
A

11-0
11-1
11-2

11-3
11-4
11-5

11-6

11-7
11-8

11-9

*
S

0-2-8
0-2

B
8
B
B
8

B
B

C
C

0-9

1
2

0
1
2

3
4
5

6
7

8
9

A-8A

%
&
0

(

+
)

Figure 4.10 Graphic subsets 1 and 2.

high -+

C
C

1
1

2
2
2

1
1

2
2

1
2
2

4
4

B
8

4
4

1
1

2
2

8
8

1
2
2

4
4

0-7
0-8

1
2
2

X
Y

1
1

8
8

0-6

4
4

1
2
2

4
4

Print Arrangement H

A
A

Graphic Subset 2

Print Arrangement A

8-2-1

Graphic Subset 1

8-4

A
A

0-3
0-4
0-5

Less than

4
4
4

A
A

>
v'

<
>

2
4

2-8

..fi

11
0-1
0-3-8
0-4-8
0-5-8
0-6-8
0-7-8

ill-

Lozenge

Figure 4.8 Special characters or symbols with names.

B-A-8-4

No Punches

BCD CODE (Core Storage)

CARD CODE

12
11-3-8
11-4-8
11-5-8

Prime or Apostrophe

B-A

&
$

Period or Point

BCD Code

)

A
A
A

1
1

2
2

8
1

8
8

2
1
2
2
4
4
4

8
8

1
1

2
2

1
1

Figure 4.9 Standard BCD code in collating sequence.

THE PUNCHED CARD AS A UNIT RECORD

for reference only and does not pertain to any card representation
of data.
Notice again that the assignment of a particular meaning (in
this case, a character) to a punched hole is a choice that can be
made by the user. Printers can be obtained with typesets specified
for almost any unique application of punched cards. For example,
machines used in the United Kingdom may have little or no occasion to print dollar signs. Instead, equipment is required to print
the signs for pounds, shillings, and pence. Other countries and
locations may also have individual requirements for typesets to
print peculiar language and monetary graphics. The coding system
just described can be used to represent an almost unlimited variety
of data.

Card Fields

All records are made up of a number of related items that may
be called fields. Figure 4.7 also shows a card field at the right
end of the card, made up of columns 75 through 80 and labeled
Date. In this field, columns 75 and 76 are reserved for the recording of month, expressed in numeric fashion. That is, January is
punched as 01, February as 02, and so on. Columns 76 and 77 are
set aside to record the day of the month, also expressed numerically. Columns 79 and 80 record the last two digits of the year.
The design of the field illustrates a number of considerations that
must be taken into account whenever data is punched into cards.
The total amount of information to be put into the card is
determined by the requirements of the reports and other documents that may be prepared from it later. A card field may be from
one to 80 columns in length, depending upon the length of the
particular item of information to be punched. For example, a
field reserved for name would usually be longer than a field reserved for social security number. However, each field must be of
consistent length in all related records making up a file. Also, the
field must be positioned in the same columns in all cards of the
file. And, once a field is assigned to record some specific item of
information, it cannot be used to record any other type of data.
This restriction in card layout and design is true of hand
written documents as well. Although the number of digits or
characters may not be set exactly, the location of most items on
any form is indicated. For example, one of the standardized forms
in common use is the personal check. Each item of information
on this record is located in a designated position.

THE PUNCHED CARD AS A UNIT RECORD

The length of a record field must be known exactly for any
type of machine recording and processing. The length is determined by the maximum number of digits and characters that can
be expected to be recorded in the field. For example, social
security number is always nine digits in length. Therefore, nine
columns are always required to record that item of data in a
card record.
To punch any quantity or value, the number of columns reserved must be sufficient to record the largest quantity expected
in any record. Suppose the amount of sale is one of the items to be
punched in a department store accounts receivable application. If
the selling price of no item in the store is as much as $10,000.00
but might be more than $1,000.00 then, six columns is all that
will ever be required to record this amount. (Dollar signs, and
decimal points are not recorded directly into the amount field.
The postion of these characters is understood in the format of
the record and its component fields.)
From right to left, the positions in a numeric field are known
as the units positions, tens position, hundreds position, thousands position, and so on. Each digit of any number recorded
in a card field must be punched into its corresponding column.
Therefore, if the number to be punched is shorter than the number
of columns reserved for that field, insignificant zeros must be
added at the left of the number to fill the field.
For example, if the number to be punched is 764 and the
card field is five columns long, the number is punched as 00764. If
the number is 64, the field is punched as 00064. The single 4 is
punched as 00004. As previously explained, it follows that fields
of any machine records must be laid out in advance to accommodate the maximum number of characters expected. Data cannot
exceed field size.
To record an alphabetic item such as name, punching starts
properly in the leftmost column of the field and continues on to
the last letter to be punched. If the alphabetic item does not fill
the field, the remaining columns of that field are skipped over,
unpunched. Thus the constant length of the field is always
maintained.
Punched cards are defined as fixed-length records made up of
fixed-length fields. Some types of computers can process and
manipulate records of variable length, particularly where records
are magnetically recorded on tape. In this case, the amount of
data per record is not limited by the physical characteristics of the
recording medium.

THE PUNCHED CARD AS A UNIT RECORD

In July, 1969, IBM announced the most radical change in the
physical characteristics of the punched card since its introduction
by Herman Hollerith in 1890. As the principal recording medium
for the System/3, a small computer, the new card is slightly more

The 96-Column Card

than one-third the size of the old 80-column card: 3t inches long

2~ inches wide. By recording with small round holes, 96 col-

umns of information can be punched in a single card. Thus, the
new card holds 20% more data in about 30% of the old card area.
Figure 4.11 shows the layout of the IBM 96-column card.
[,,/-1-2-3

•

9 10 11

12 13 14 15 16 17 1. 19 20 21 22 23 24 25 26 27 2. 29 : - : - :

'\'J

I
I

33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64

65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96

97 98 99 100 101 102 103104105106107108109 110 111 112 113 114 115 116 117 118 119 12C 121 122 123 124 125 126 127 128

8
4
2
1

8
4

2
1

10 11

12 13 14 15 16

17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

A
8

4
2

33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64

A
8
4

A
8

l=
1

65 66 67 68 69 70 71 72 73 74 75' 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 9995
3::
96 : 2 J .

----IBM

----Figure 4.11 96-column card.

The card is divided into two main recording areas: print and
punch. The print area is subdivided into four horizontal rows of
32 print spaces each. The spaces are numbered from left to right;

THE PUNCHED CARD AS A UNIT RECORD

the top row from I to 32, the second row from 33 to 64, the third
row from 65 to 96. Print spaces correspond to the numbered
punch columns in the lower section of the card. Print line 4 does
not have a corresponding tier but may be used for printing. The
printing can be done at the time the card is originally punched,
either by a key-operated data recorder or during System/3
processing.

•

• • 10 \I

. . .S "

11 •

10 21 U

"""."":t7

H",:IO" "

Tier 1

... -.-m ____

~~_·

........ _..

m_~

Tier 2

II.

1.~

••• M •••••••

nH.".nHH • • • e

..8

;}

Zone
Portion

Digit
Portion

~»M • • n." ••

auMu.n • •

~.~"." • • •

H •••

"~~
A

..
8

2
,

• • n __ • • n n • • • n • • • • • • • • • • • • • • • • • • • '

Tier 3

Figwe 4.12 96-column card punching area divided into tiers.

The punch area is divided into three horizontal tiers, each
with 32 vertical columns (Figure 4.12). However, whereas the
80-column card has 12 punching positions in a single column,
the 96-column card has only six. In these positions, all digits a
through 9, the letters of the alphabet, and 28 special characters
can be represented.
The numeric portion of each column has four punching positions: 1, 2, 4, and 8 (Figure 4.13). The zone portion of each
column has two positions for punching, called A and B. Like the
80-column coding scheme, digits are represented by punched holes
in the numeric portion of a column. But, unlike the 80-column
code, the four numeric positions are assigned a value in the code.
Thus, the digit 1 is represented as a hole in the one position; 2 is a
hole in the two position, while 3 is represented by two holes in the
same column with the value of 2 and 1. The digit 4 is a h~le in the

THE PUNCHED CARD AS A UNIT RECORD

Zone
Portion
Digit
Portion

1 2 3 4

7 8 9

10 11

... 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

S _

f1- -----"--.-.-:~:: ..---------------- t
_9L:

99..JO(UOUOt.!0;VO~.1°1'~ !Q? 1Q8.J...09-l!LlL"L",!, 1~

.illi .1.16--'17..118--'19 J20.J21...!2LI2V2

A A A A A A

8 8 8 8 8 8 8 8 8 8 8 8
4 4 4 4
4 4 4 4
2 2
2 2 2 2
2 2
1
1
1
1
1
1

Figure 4.15 Binary coded decimal representation of 96-column card characters.

Alphabetic characters are represented by a combination of
punches made in the numeric (8,4, 2, 1) and zone (A, B) portions
of a column, as shown in Figure 4.14. Special characters are
represented by a combination of punches in the digit and zone portions of the column. Figure 4.15 shows the combinations used
to represent each character.

Questions

and
Exercises

THE PUNCHED CARD AS A UNIT RECORD

1. Explain two requirements for representing data for machines.
2. Describe the effective punching areas of the 80-column card; the 96column card.
3. List some different types of recording media with which you are familiar.
4. What is probably the most important advantage of punched-hole
recording?
S. Explain the use of punched card corner cuts.
6. What is a collating sequence?
7. Explain why fields of machine records must be explicitly designed.
8. What are some of the advantages and disadvantages of the 96-column
card?
9. What is the BCD code?
10. What is the purpose of printing on punched cards?

THE TRANSCRIPTION TO
PUNCHED HOLES

Data processing is a series of planned operations and actions
upon data to achieve some desired result. The procedures and devices used make up the data processing system. The devices may
vary widely. They may be complex machines or only paper and
pencil. The procedures, however, always remain basically the same.
Of course, there are many types of data processing systems.
These systems vary in size, complexity, speed, cost, and application. But whatever is to be processed or whatever equipment is to
be used, at least three fundamental considerations are involved:

1. The source data, or input entering the system
2. The orderly, planned processing within the system
3. The end result or ou tpu t from the system
In some computer systems, data can be handled on a real time
basis. That is, information in some form enters the system as it
happens. For example, a pressure gauge in an oil refinery transmits
a reading directly to an input device capable of translating that
reading into input data the system can operate upon. After some
processing, an output device directly actuates a control to close a
valve and regulate pressure. Other readings of volume and temperature also may be fed to the computer as changes occur or at
77

THE TRANSCRIPTION TO PUNCHED HOLES

periodic intervals. Data processing is programmed in such a way
that output from the computer directly controls the production
of the refinery.
In a punched-card system, input is always historical. That is,
all data entering the system has originated some time previously.
This chapter describes the various methods of transcribing this
sort of data into punched cards.

80-Column Card
Punching

Data to be recorded on punched cards are very often manually transcribed from some original source document, This process,
known as keypunching, is done by some type of card punching
machine especially designed for this purpose, Keypunches are
equipped with a keyboard similar to the one on a typewriter. An
operator depresses keys to punch holes just as a typist depresses
keys to print letters on paper. When each key is depressed, the
character code corresponding to that key is punched as one or
more holes in a single column of a card. As the operator reads
from an original document, the information is copied exactly in
the form ofpuriched holes, just as a typist might copy from a handwritten document to produce a typewritten page.
Many different models of keypunches can be obtained. Typically, 80-column cards pass through the punch one at a time from a
hopper to a punching station. After punching, each card is automatically released to a reading station. The card then passes
through this station in phase, column by column, with the following card now at the punching station. Whenever a card is completely punched at the punching station, a card at the reading
station moves into the stacker.
Because each card passes the reading station in phase with the
following card at the punching station, it is possible to have automatic duplication from card to card. That is, data from a specific
column at the reading station may be read and transmitted back
by the machine to be punched in the corresponding column of the
card following at the punching station. When this card then
reaches the reading station, the same character is again transmitted
back to the next card. This process continues as long as cards are
fed through the machine or until the duplicating is otherwise controlled by the operator.
The ability to duplicate or to reproduce data automatically
is an important function of card punching in particular and of all

THE TRANSCRIPTION TO PUNCHED HOLES

data processing systems in general. Manual punching and keying
may be considerably reduced, thereby decreasing the chance of
human error in the processing. Data common to any group or set
of records, such as date originated, may be established for the first
card or record and automatically duplicated in all succeeding records in the deck or file.
Figure 5.1 shows an IBM 129 Card Data Recorder. The basic
components of the machine are labeled in the photograph.

Figure 5.1 IBM 129 Card Data Recorder.

Blank cards to be punched are placed in the card hopper, a receptac1e located in the upper right section of the machine. A
sliding pressure plate at the back of the hopper pushes the cards
forward against a set of feeding knives. When the feed key on the
keyboard is pressed, the knives are actuated to move the front
card down into a feed station. The card is held at the feed station
for registration at the punching station. The hopper holds about
500 cards, printed or face side forward, 9-edge down.

Card Hopper

THE TRANSCRIPTION TO PUNCHED HOLES

The first two cards from the deck are fed by pressing the feed
key. All following cards can be fed automatically by setting an
auto feed switch. Operator control switches are just above the
keyboard as shown in Figure 5.2.

fl~

ft~

OFF

O$~' 0·: : :
OFF

OFF

PROGRAM MODE
PIV MODE

AUTO
SKIP/DUP

RECORD
ADVANCE

CHARACTER
MODE
READ

CLEAR

Figure 5.2 129 Card Data Recorder. Operator control switches.

Punching Station

Punching is done at the second station along the card path. As
stated above, when normal punching is begun, two cards are fed
into the card bed at the right of the punching station. As the
second card is fed, the first card is registered for punching. That
is, the left end of the card is accurately positioned under the
punch dies.
While the first card is being punched, the second card is held
at the feed station to the right of the first card. When the first
card is completely punched, it is moved either automatically or
under key control of the operator to the reading station. At the
same time, the second card is registered at the punching station
while the third card is moved from the hopper to the feed station.
The second card is punched. When completed, it also is
moved to the reading station while the first card is ejected to the
stacker in the upper left section of the machine. Punching continues, card by card, in series. As each card is fed from the hopper
to the feed station, a card is also ejected to the stacker. Thus, the
operator controls a continuous flow of cards through the machine
from right to left. Provision is made for single cards to be inserted
by hand if necessary.

Stacker

Cards are ejected by metal fingers that grasp the upper or 12edge of the card. The fingers are located on a roller mechanism

THE TRANSCRIPTION TO PUNCHED HOLES

which stacks the cards horizontally, building the deck by adding
cards, one at a time, to the bottom of the stack. Note that
this arrangement keeps the cards in the sequence in which they are
punched; first card on top, last card on the bottom of the deck,
12-edge to the rear. As will be shown later, it is important in
most applications to preserve the exact punching sequence for a
machine verification of the cards in another keying operation.
A card weight holds the stacked cards, in position until removed by the operator.
The keyboard of the card punch is somewhat like a typewriter
keyboard. Figure 5.3 is a typical key arrangement. The keyboard
can be rotated on the desk or reading table for operator convenience and comfort.

Keyboard

n m n

~ (#J
~ ~ ~ (Ol ~ Q.~
~W~l.:J~~~l=JW~Ol:.::J
(+l r=1
(D-8-;') (II (1l
BLANK
FIELD lQ) ~
E W W ~ lJU ~ lQ) ~ FEED CO~

CD m

WORD
BKSP

n I>l
W
W

m m m

n~ l£)
n RlQ) l!iJ
n CD (5)
(6J 8 0
~ W
U
J

SKIP

~:=~

CTRL

(~____________________sp_a_c_eb_a_r___________________)
Figure 5.3 IBM card punch key arrangement.

Letters of the alphabet are placed in the same locations as on
the standard typewriter. However, whereas the numeric row of
keys is placed at the top of a typewriter keyboard, these keys are
superimposed upon the alphabetic keys of the card punch. For
example, the digit I is on the same key as the letter U, the digit 2
is placed with the letter I, 3 is with 0, and so on. The effect is to
arrange the digits in a convenient grouping while at the same time
maintaining the standard typewriter touch system. This permits
the operator to punch an alphabetic field with both hands and
then, without moving fingers from the home position, to punch

THE TRANSCRIPTION TO PUNCHED HOLES

a numeric field with the right hand only. Use of the right hand
alone frees the left hand for document handling.

"1'

Punched Card

,~--------~.----------~
1 2 3 4 5 6 7 8 9 10

I, I I I I I Iy I I

•

I I I I I I I I I I)
1

2 3 4 5 6

8 9

•

I \J IT] ~Output Storage
f

rn~lnputStorage
79 80

~--Keyboard

~ ~.

-Source Record

Figure 5.4 Data flow, buffered key punch.

The 129 Card Data Recorder contains input and output sections of data storage as shown schematically in Figure 5.4. Each
storage section can hold electronically as many as 80 columns or
characters of information from the source record. All data keyed
for punching is first placed in input storage. A column indicator
shows the operator the storage position where the next character
is to be place~d. After the source record is completely keyed into
input storage, the information is transferred to output storage and
then punched in a card. While one card is being punched from
output storage, the next record can be keyed into input storage.

THE TRANSCRIPTION TO PUNCHED HOLES

If a character is keyed incorrectly, the machine can be backspaced and the correct character keyed. Since a high percentage
of such errors are recognized as they occur, the operator can correct most errors without spoiling a card. Rekeying erases the incorrect character in input storage and stores the correct character.

Control features of the card punch can be made almost entirely
automatic with the program feature. With this feature an operator can preset such functions as field tabulation or skipping, duplication of one field throughout the file, numerical or alphabetic
keyboard shift, etc., before beginning each job.
The 129 is equipped with the capacity to store up to six different programs. Each program is set up by punching a specially
coded card which exactly defines the fields and the operations to
be performed. Using a special procedure, the operator then can
read the program card into one of the six levels of program storage.
The program mode switch is set to the desired level from 1 to 6.
At any time while punching, the operator can select the appropriate program of control for a specific job or card format. This arrangement makes it possible to conveniently handle different record formats within the same file.
11111 111111 1111111 111111

1111

11111 1111 11111 11111 111111111111111111

10000000000000000000000000000000000100000010000010000000000000000000000000000000

;:; ;; ;;;;;';';';';';';';';' ;'~;' ii iiili~7;' ~ ~~i~;' ~ ~ 7;' ;'.;~;;' ~iiliii ~~;' ~~~;' ~~7 ~~;'~~ ~;' ~~;' ~ ~;' ~;~
2222222 2 2222222222222222222222222222222222222222222222222222 2 2 2 2 2 222222222222222
333333333333 3 3333333333333333333333333333333333333333333333333333 3 3 3 3 33333 3 33333
444 4 4 4 44 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 44 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 44 4 4 4 4 4 4 4 44 4 4 4 4 4 44 4 4 4 4 4 4 4 444

55555555555555555555555555B5555555555555555555555555555555555555555555555555555
666 66 66 66 6 66 && & 66 6 & & 6 & 6 & 6 &6 & & & 6 & 6 6 & & & 66 &66 & & 6 6 6 666666 & & 6 6 66 66 66 6 6 6 66 6 6 66 6 6666 66 6
J J JJ J J JJ JJ J J 11 J Jl Jl 11 Jl 11 J Jl J J J J Jl 11 J J J Jl J J J Jl Jl Jl 11 11 11 J JJ 11 JJ J J J JJ J J 11 JJ JJ J 11 J
11118 8 88111188 8 8 8 8 888 88 8 8118 8811181811 BII 8 8 8 88 818811818. 11111 81111 8118 1 8 8 8188 III
999999999999999999999999999999999999999999999999 99993999999999999999999999999 !9 9
1:1 J 45.

IlIllll:r1J141$1I111'II20flZ2nMZ5.1JltJl30l!J2»)4~.nJl

• • 414fq"".'1 • • "'SIV U".M'J.'!ttD'IIUH4rs"IIU_Jiill:'?JJJ415J1J11111.

Figure 5.5 Punched program card.

Figure 5.5 is a punched program card. Codes for four basic
functions are shown in the example, as follows:
Blank

o
12

Indicates the first column of a field to be manually punched
from the keyboard into each card
Shifts keyboard to alphabetic mode for the column where
it is punched
Starts automatic skipping
Defines the length of the field

Program Control

THE TRANSCRIPTION TO PUNCHED HOLES

Columns 14 through 21, 29, 37 through 42, and 56 through
61 are programmed for numerical punching. That is, if a combination key is depressed, a digit will be stored in input storage. Columns 22 through 28 and column 35 are programmed for alphabetic punching.
Columns 1 through 6 and column 36 of the cards being
punched will be duplicated from the first card of the job. Duplicating and skipping are under manual control for the first card
when the information to be duplicated is manually punched. Column 43 is also programmed to be duplicated, but only with alp4abetic information or a blank.
Columns 7 through 13, 30 through 34, 44 through 48, and
column 55 are skipped.
The preestablished machine control of record format greatly
increases the efficiency of keypunching. When freed of the repetitively keyed functions of control, the operator can concentrate
entirely upon the operation of actual data transcription. Figure
5.6 is a tabulation of all program codes for the IBM 129 Card
Data Recorder.

Function'

Punch Code

Field location

Field ()efinition (FD)

All column., except first column

Auto Skip Field

First column

Auto Dup/Auto Ver Field

First column

Alpha Shift (Progrommed)

Each column for alpha shift

Self-check Field

First and last columns

Left-zero Field (Verify Only)-

First and lost columns

Add Accumulator A

First column

Add Accumulotor 8

6
7
4+5
4 +6

First column

Punch and Reset Accumulator 8

4+7
4+5+8
4+6+8

Punch and Reset Accumulator C

4+7+8

First column

Add Accumulator C
Punch Accumulator A
Punch Accumulator 8
Punch Accumulator C
Punch and Reset Accumulator A

First column
First column
First column

First colwnn
First column

Figure 5.6 Program codes, IBM 129 Card Data Recorder.

Other Features

Two models of the IBM 129 can print the character punched in a
column along the top 12-edge of the card. The printed character
corresponds to the punching. The printing feature can be operated

THE TRANSCRIPTION TO PUNCHED HOLES

under either manual or program control. Thus, in one operation
cards can be prepared for both manual and machine processing.
A self-checking number feature provides for verifying precoded numerical information at the same time it is punched. This
type of verification eliminates any need for later verification by
other methods. This feature is described in more detail later in this
chapter. The feature also operates under program control. An 11
punch is punched in column 81 of the data card to signify a valid
self-checking number.
An accumulate feature provides the ability to balance to a
predetermined total, to create a total for a group of cards or a
batch of work, or to crossfoot and punch totals in the same card
or in a following card, under program control.
The machine is equipped with three l4-position, individual
accumulators that add any numerical manually keyed or duplicated fields, in punch or verify mode, into any accumulator. The
maximum size of the input field is 14 positions, thus any accumulation beyond the fourteenth position is lost.
Punchout, or punchout and reset, of individual accumulators
are under program control, and may be done selectively. Group
totals are normally punched out by way of a different program
level than the accumulating program level. Amounts to be accumulated can be entered only by keying or duplicating.
A field will be subtracted in an accumulator if an 11 punchis
multipunched in the units position of the field. Minus totals are
punched out with an 11 punch over the units position of the field.
The production statistics feature provides statistics on machine production for use in measurement of work load, analysis of
errors, and job accounting. A keystroke counter of six positions
(000,000 to 999,999) counts every data keystroke in either punch
or verify mode and functional keystrokes with some exceptions.
A four-position card counter counts every output record in punch
mode and every verify correct or correction punched card in verify
mode. It does not count error cards in verify mode, program cards
read in, and data cards read in. A four-position verify correction
keystroke counter counts each verify correction data keystroke in
verify mode.
All counters may be punched out, or punched out and reset
by the operator. Totals may be obtained by batch, job, day, etc.
All counters are reset to zero when the power is turned on to
the machine.

THE TRANSCRIPTION TO PUNCHED HOLES
DO-IT-YOURSELF, INCORPORATED
Rochester 17, New York
Invoice No.

Z/006

Invo ice Date --.:.l,L0:....:/5::....bc..::7I...::"O,--_
Date Shipped ~r/1°'Terms

='<;":;":'!L.-'!."::-

Single transactions

from a
source document
become UNIT RECORDS ...

~Jel0060715?Q'50

0?3114ROBERT LEE HONEYl·JELL
CUSTOMER
II
II I I
NUMBER II
I CUiOMER ,NiiE
10 0 10 ,

~ INVOICE INVOICE
~ NUMBER

WATT AMP 10B17(018X12fT1329~1329~

CAT NO. ~

NAME OF ITEi I

DATE

SIZE

UNIT TOTAL
PRICE PRICE

uDOC I 0 enD 000 ri 0 II 0 0 n;! 0 0 O~ 0 0 II 0 IuD 0 0 10 10 I 0 II 0 00 0 001 COO II 0010 0 010 0 0 0 0 0 0 DOG

1 I 1 4 5 ; , I

IR II 12 1lll 1:'. 15 I, II 19

11 Ii 13 U Z5

~& ~:;6 ~3

J= JI 11 li 3' Ji

~i jI

35 N

I.' 43 U 45!; 41 "43 SI

~I ~

51 U' 52 U

~4 ~)

iG &J 51 U 1111 12 13 14

. :0

11:~

'it

11111111111111111111111111111111111111111!111111111.111111111111111111111.1111111
22 2 2 2 2 212 2 2Z 21 2 111 22 2 /112 1 22 12122 2 /11 2 11 2 22 2 121 2112122 /12 12222 2 2122 22122 221222 2!
3 313 3 3 3 3 331313 3 lJ 3 33 3 3 ~ 113 3 , 13 3 3 J 3 3 3 3'3 3 3 3 3 3 3 3 3113 3 3 3 3 3 3 33 3 3 3 33 3 3 3 31313 3 3 313 3 3 3 3 3

4444144444 H 44 1 ,j 4 4 4444444444444 4 44444444444 4 4 44 H 414 44444 44 4 4 4 4 4 4 4 4 4 4 4 44 4 4 4 4 4444

5 5 5 5 5 5 5 515 5 , sll5 5 511 j 515 , 5 5" 5 5 5~ 5 5 5 5 5 51 ~ 515 5 50 5 55 5 :; 5 55 5 5 5 5 5 5 55 5 5 S J 5 5 5 5 515 5 5 515 5 5
6 6 6 G6 61 G6 G6 6 6 6 6 Gsl6 5 ~
IIII IIi ; I 1 I 1 1 ) 1 i 1 ;
88888 B Z8 ~ 612 8 8 S a I

n1,18 H

9999919991 9 9 ~ 9 9 9 a 3 9'J
llj~r4

.. !! ~.~ a ~:~ll"

I, f, 6 6 H

ISllll

6 6 6 6 fi 616 n fi fi fi 6C ; 16 6 66 6 6 6 66 GG6 6 6 6 G6 6 6 :.16 6 G6 : 6 6 6 6 G6 GG6

1 1 1 I i I ) I J II ) 1 ~ I I J I' 111 1 1 1 1 7 1 7 1 1
ij

I 7 1 1 1 11 ) 1 1 111 1 J 1 1 1 ) 1 1 I 1 1 1 1 lIT

~ l 8 8 H B ! 88886883 & 8 B a 0 8 8 888888 B 8 9 8 Be 8 818 8 8 n 8 a 8 8 S 8 D!8 8 8 n

g H 9 9 9 9 9 9 9 9 999 • ~ 9 9 9 9 9 9 , 9 9 9 99999 9 9 9999993 9 9999999199 9 9199 9 ~
JlI?l:~;·l~J.:\3~E:

7,;J~!'I'~!1'1~

i5l;:a4~~~~I:;il'~~~~5l!~~:'IiDE

1i2~lH(J6,i'6Inll;I~I:~!liJJi'

Figure 5.7 Transcription process from original record to SO-column cards.

~~'!"

THE TRANSCRIPTION TO PUNCHED HOLES

Figure 5.7 is a schematic of the transcription process from an The Transcription
original record to punched cards. In this exam pIe, the record Process
is an invoice that lists items sold to a customer. The example illustrates the following basic principles that should be considered in
the design of any key punching transcription procedure:
1. The original record always contains historical data. For

example, it is assumed here that the quantity of goods itemized
are being delivered or have already been delivered to the indicated
customer at specified prices.
2. Historical records almost invariably originate in a location
at some distance from the keypunching operation. It is unusual
for the keypunch operator to take any part in the preparation of
a record. In fact, the operator most likely has little or no knowledge of the conduct of the business involved' or of the characteristics of the goods or services to be punched as data in cards.
3. The operation of keypunching must always be reduced
to one of simple transcription. An operator cannot be expected
to decode, rearrange, or translate data from original record to
punched card.
4. Conversion of identification or other data to numerical
codes before the document arrives in the keypunching department
greatly reduces the number of holes to be punched and more efficiently uses the 80-column capacity of the card.
The invoice in Figure 5.7 has been coded in advance for
keypunching. Numeric codes have been assigned to the items of
sales office and salesman, city, state, and customer number. Note
that this coded data is common to all items listed on one invoice. As other unit record processing operations are reviewed, it
will be seen that the handling of numeric data is simpler and more
efficient than the handling of alphabetic data. This is a further
advantage of reducing information to numerical data whenever
practical.
5. The format of the card is arranged to accommodate data
in the same sequence as it is read from the invoice by the operator.
Customer name, invoice number, date, and sales identification are
in the left end of the card (the first columns under the punching
station). This information is punched only once into the first invoice card and is then duplicated into each succeeding item card.

THE TRANSCRIPTION TO PUNCHED HOLES

6. The card record is logically divided into its component
fields, each a separate item of data within the record. The amount
of information to be transcribed should be held to the minimum
required for final reporting as required by the procedure. Field
size is designed with the proper number of columns to contain the
maximum number of characters in the field.
7. The accuracy of the transcription process must be controlled. The first concern is to assure the accuracy of key depressions and punched holes. Aside from the normal handling and
accounting controls, the primary method is one of machine verification. This process is described in the next section.

The Machine
Verification Process

The machine verification process is intended to be an exact repetition of the original transcription operation from written documents to punched cards. The purpose of verification is to assure
the accuracy of transcription. The following considerations should
be noted:
1. The verifying operation, like that of keypunching, is
strictly mechanical. The operator of a card verifier duplicates
exactly the reading, keying, and handling of both documents and
cards. Whenever possible, work should not be punched and verified by the same operator.
2. Normally, all transcribed data need not be verified. In
many applications, descriptive information such as name, size,
weight, color, etc., is not essential to the accounting accuracy of
the procedure. For example, even if the name of an item is misspelled, chances are. the item can still be identified properly on
printed reports. As in all procedures, the value of assuring proper
spelling of names· must be measured against the cost of machine
verification.
On the other hand, if customer names and addresses are involved, it may be well worth the extra cost of verification to avoid
the embarrassment and possible loss of business that could occur
if shipment is made to a wrong address.
3. Mechanical verification does not assure absolute accuracy.
A verifier operator can misread the original document in the same
way the keypunch operator did. Or, the original document may
be illegible or confusing, making transcription accuracy unreliable.
Other methods of assuring accuracy are discussed in a later section.

THE TRANSCRIPTION TO PUNCHED HOLES

4. It must be stressed again that proper design of documen ts
and card forms is essential for accurate, efficient card punching
and verifying. When any good operator is observed in action, it is
immediately apparent that rhythm plays an important part in
achieving efficiency. Just as a good typist falls into an easy, fast
rhythm of word keying, carriage return, and paper handling, the
skilled card machine operator works best with a good rhythm of
card feeding, keying, skipping, duplicating, and document handling. Even the sound of the machine in operation indicates to the
experienced operator that the work is being done accurately; particularly that data fields are in the proper columns.

The 96-column card is the principal data entry medium for the
IBM System/3. As a completely independent unit in this small
computer system, the IBM 5496 Data Recorder does the jobs of

Figure 5.8 IBM 5496 Data Recorder.

96-Column Card
Punching and
Verifying

THE TRANSCRIPTION TO PUNCHED HOLES

both creating this latest form of unit record and verifying data
which has been previously recorded. In addition to combining
the two functions in one machine, the 5496 Data Recorder has
been designed with a number of improvements over the older SOcolumn equipment.
Compactness is made possible by the use of smaller cards, as
shown in Figure 5.S. Also, the unit has the capability to store as
many as four levels of control programs at the same time. This
complements the greater flexibility of the stored program System/3
to process multiple record formats.

Punching

All keyed characters are placed first in entry storage. This operation is very similar to that of the 129 SO-column punch. Keying is carried on without punching, until all data to be placed
in a single card is stored. Then, by pressing a release key, the
stored data is transferred to output storage, from where the entire
card is punched automatically. Punching is done at a speed of 20
columns per tier per second. Three columns are punched at a time,
one in each of the three tiers; for example, columns 1, 33, and 65
or 17,49, and S1. The result is a punching (and printing) speed
equivalent to 60 characters per second. Data for the next card can
be keyed into entry storage while the preceding card is being
punched. When data enters storage, it replaces any information
previously stored there.
With the Data Recorder, provision is made also to erase a keying error from storage before the release key is pressed. Because
no holes have been punched at this time, erasure can be made for a
single column, a field, or for an entire card record. Release of data
in output storage is entirely under the operator's control.

Operating

Some of the more important operating features are reviewed here.
The card hopper, with a capacity of 350 cards, i~ located to
the operator's right. Gards are placed in the hopper printed side
forward, punching positions toward the bottom. Cards are fed one
at a time into a transport mechanism (Figure 5.9) where they are
carried to punch, read, and print stations. Movement of the cards
is from the operator's right toleft.
The stacker at the left holds the completely processed cards
in the same sequence in which they are punched or verified.

Features

THE TRANSCRIPTION TO PUNCHED HOLES

Ribbon cartridge
mounting bracket

Ribbon cartridge

Stacker

Print station

Read station

Punch station

Hopper

Figure 5.9 IBM 5496 Transport.
Program Level
Indicator Lights

Column Indicator

,=1

-=1 (61

PROG 1

PRoo 2}
ERROR

AUTO
AUTO
PROG
SK/DUP REC REL PROG LOAD

RECORD
PRINT ERASE

@ij@@@ij
OFF

OFF

Figure 5.10 5496 operator's paneL

An operator's panel is located just below the transport mechanism. Figure 5.10 is a schematic of available control features on
the panel. Note the column indicator, a lighted digital display that
indicates to the operator the position of the nex t keyed character
entry into storage. The position indicated corresponds to the card
column in which that character will be punched. While operating
the machine, the operator is not concerned with the·actual punching or reading position of the card. The cards are not visible to the
operator as they move through the transport mechanism.
Punching of a record begins when the column indicator advances from 00 to 01. At this point, information for the next card
record can be keyed. The digits 01 through 96 are displayed in
the indicator.
The combination keyboard rests on a flat desk at the front of
the machine. Data keys are arranged in the same positions as the
keys of 80-column machines with the facility for recording either

OFF

THE TRANSCRIPTION TO PUNCHED HOLES

alphabetic or numeric information. Control keys are positioned in
each of the four rows. Figure 5.11 is a schematic of the Data Recorder keyboard.

Space Bar

Shaded - Function keys
Unshaded - Data Keys

Figure 5.11 5496 keyboard.

Program Control

Switch

When the PROG switch is on~ the Data Recorder operates under program control. Information can be stored in four program levels that can be used to control automatic and manual
punching functions.
When the program control switch is turned from off to on,
the information in program level 1 is automatically in controL A
different program level may be put in control by pressing another
program level key on the keyboard: PROG 2, PROG 3, or PROG 4.
While keying data, it is possible to shift between program
levels any number of times if the Data Recorder remains under
program control. If a different program level key is pressed after
the first column of a field has been passed,· the program level in
control automatically changes when the column indicator advances
to the first column of the next field. This protection feature prevents changing inadvertently from one program level to another in
the middle of a card field.
When the program control switch is off, the Data Recorder
operates under manual control. Information is keyed on a column-by-column basis, disregarding any automatically programmed
functions.

THE TRANSCRIPTION TO PUNCHED HOLES

A program card is prepared first by the operator with the Data
Recorder under manual control. Punched hole codes are assigned
to operate the various control functions of the machine. These
codes are shown in Figure 5.12.

93
Program Card

Description

Code

Keyboard Character
to Press

Card Column
Where Pu nched

End of Field

- (minus1*

last column of every field

Automatic Skipping

~ (zero) *

Each column to be skipped

Automatic Duplicating

Each column to be duplicated

lower Shift

Each column necessary

Numeric Shift

Each column necessary

Setf-check Number
Verification

First column of field only

When operating under program control, if a column is not coded for either lower shift or
numeric shift, it is treated as upper shift.
If a column is not coded for either automatic skipping or automatic duplicating, it must be
manually keyed.
* The minus (-) will produce the "B" punch position code.
** The zero (fIJ) will produce the" A" punch position code.

Figure 5.12 5496 control codes.

For a control card to function, it must be read and stored in
a program level storage area. The program load switch is used for
this purpose. Four different storage areas are provided and each
level can store information from one program card at any given
time. As part of the loading procedure, the operator selects which
storage area is to receive the program. The program card is placed
in the hopper, fed through the machine, and read into storage at
the read station. Once placed in storage, the program remains
there until replaced by another program.
When a keypunching job is in process, the operator can select anyone of the four stored programs to control operation.

Automatic
Duplication

Right Adjust

THE TRANSCRIPTION TO PUNCHED HOLES

Information common to a given number of cards can be dupJi.
cated automatically by the Data Recorder. The data is read from
the output storage area under either manual or program control.
This is equivalent to reading data from the preceding card into
the card being punched.
An auxiliary duplicating feature is also provided. Information
is read first into the fourth program level storage from a punched
card in the normal manner. Whenever this information is to be
duplicated, it can be placed in key entry storage by the operator
and from there into the output card.

This function is supplied by the left-zero feature of an 80-column
card machine. It operates only under program control when the
end of the field has been defined by an end-of-field code (hyphen)
in the program.
When the right adjust key is pressed, keyed characters in the
field are shifted to rightmost columns. The insignificant positions
to the left of the field are left blank for alphabetic fields and are
punched zero, without keying, for numerical fields. Right adjusting can thus be done with either alphabetic or numerical fields.
Figure 5.13 shows the source document to card relationship
of the 96-column card. Figure 5.14 is a program card prepared for
this particular punching procedure.

96-Column Card
Verification

The Data Recorder can be used as a verifier by setting the mode
switch on the operator's panel to verify. Cards to be verified
are placed in the hopper in the same sequence as punched. Cards
are read one at a time into the output storage area. This is done
by pressing the read key.
Each card is stopped at a verify wait station after reading.
The operator then reads the same data from the source document
and keys the data a second time. As each character is rekeyed it
is automatically compared with the character in the same column
in the ou tpu t storage area. When the comparison agrees, keying
can continue until the entire card is verified. If the verification is
completed without error, a notch is punched in the trailing edge to
mark the card as correct. The verification notch is shown on a
card in Figure 5.1 5.

THE TRANSCRIPTION TO PUNCHED HOLES

Employee
Name

ST E/ IN tt

NE LS pH

Employee
Numer

I------j
I
I
I
I
I
I
I
I
I
20 I

I
I
I
I
I
I

I
I

5'1

AR MS IT 1< iaNE>

PU 1) I IL
5T E'E:. ILE

Eo 'DN IER

GE RZ. lEV SIcO:t=O:::::>cO:::::>cO
1
1
I
00000 000 0'0
1 0 0 0 0 0 000 0 0 0 0 0 0 0 0 0 0 0 0000 0 00
3' 323331353631 31 39 1tJ., 1213 .. 15 ...1 .. 1150 5' 52 535155 51151 51 Sljaollju 131M 15. Icl:::::>cl=*=l:::::>cl:::::>cl

I
111111111:11111111111111111111111111111
1c2:::::>c2*2:::::>c2:::::>c2
1
1
222222222122222222222222222122222221222

1c3:::::>C3~C3:::::>C3:::::>C3

I
333 3 3 3 3 3 3:3 3 3 3 3 3 33 3 3 3 3 3 3 3 3 3:3 3 3 33 3 3:3 3 3
1c4:::::>c4~4::Jc4::Jc4
4 4 4 4 4 4 4 4 414 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 414 4 4 4 4 4 414 4 4
I
1
1
1c5:::::>c5=i=5:::::>c5:::::>c5
55 55 555 5 515 5 5 5 55 5 55555 55 5 5 515 5 5 5 5 5 515 5 5
I

OESC"""ON

66 66 666 6 6:6 666 666 66666 66 6 6 6:6 6 6 666 6:6 66

FOReMAN'S SIGNATURE

k:6:::::>C6~6:::::>C6:::::>c6

'r:::.7::Jc7-::t:.7:::::>c7:::::>c7

77 77 777 7 717 7 7 7 777 77 77 7 77 7 7 717 7 7 77 7717 77
I
1
'r:::.8:::::>c 8 =x: 8:::::>c8:::::>c 8
1
1
I
88188 888 8 818 8 8 8 888 88 8 8 8 88 8881888 8 8 8 818 88
COST PER MAT'/.. STOCK
~ qoANTITY :y:::::>c9±9:::::>c~bc9
~~ O:~~R:O~ ~ R£(I. VNITj AMOVNT 1~TY. PEL. 1
MtJ.
y
'LASS
IJNIT
NO.
'323331 353631 36 39104' .213 ...516 11.'.9505' 5253 515551515851 60816213818516 61&1691011 12131.1516n1l1911

_ _ _ _~_-_J

12345. 1. 1101112'31.'511111'"20212223242521212121

Figure 6.1 Dual caId designs.

Dual cards are punched from information previously written
on the cards themselves. Figure 6.1 is an example of this type of
card. Such cards serve as both source documents and processing
media. Examples include:

1. Inventory accounting and recording cards. The quantity
of an item in stock can be handwritten on the card when physical inventory count is taken in the warehouse. Identification
or part number may also be handwritten with some description
of the item. Cards are sent to the keypunch department for

86417

100

PUNCHED CARD DESIGN

punching of the written data; they are later used in processing
an inventory report.
2. Payroll change cards may have the change written directly
on the card, for example, a deduction authorization to meet
some obligation which is "payroll deductible." The written information identifies the type of deduction, the amount, when it
is to be paid, employee identification, and also usually includes
the employee's signature of authorization. After processing, the
card becomes a permanent record of authorization to pay the obligation. Payroll change cards may also be authorized by an employer for hourly pay rate or salary changes, work shift changes,
or job assignment changes.
3. In production and inventory control procedures, many
dual cards are originated as requisitions; that is, requests to withdraw an item from stock for production purposes in a factory.
The dual card is the most versatile of all the punched card
record forms. It has the advantage of enabling handwritten
information to be processed by machine methods. Both original and punched hole information are permanently recorded
in one medium.
CENTRAL STOCK EDIT--TRANSACTION CARD
CUSTOMER OR
JOB NUMBER

DOCUMENT NUMBER

ITEM NUMBER

ttANS

QUANTITY

CODE

O:::>CO:;::llcO:::>CO:::>cO
I

l:::>Cl:JiCl:::>Cl:::>Cl
I

Q
DI:
c(

2:::>C2:::>'F=2:::>C2:::>C2

PC3:?~PC3:::>C3

4:::>C4:::>C4:::>C4:::>c4

S:::>cS:::>~::::5:::>C5:::>C5

6:::>C6==\::6:::>c6:::>c6
I

!,~!~S.CTIQN

'A".Al'''CIIO.OU

... Ot,\ul

4UCf"'fC

<;QQti

.PCPC4::J'~fCPC4
5:::>c5:::>c5~5:::>c5:::>c5
I

c(
DI:
t-

6:::>C6:::>C6~C6:::>C6:::>C6

Pc7==\::PcPc7

_ .... e. aeon

S:::>cS::JocS:::>c8:::>c8
I

9:::>c9:::>'C9:::>C9:::>C9

Figure 6.2 Mark sensed card design.

Mark,;.sensed cards are automatically punched from pencil
marks that are made in predetermined positions on the face of the
cards. Figure 6.2 is an example of this type of unit record. The
pencil marks record data at the source of the transaction or event.

PUNCHED CARD DESIGN

101

The marks can be machine read and punched into the same or a
different card for processing.
Examples include weekly attendance records marked by a
timekeeper and meter readings made by employees of a utility
company. The most practical use of the mark-sensed card is to
record quantities of items on hand, in stock, sold, moved, received,
and the like. The card should be marked "on location" at the
time of the transaction. The size of the mark-sensed field is usually kept to 10 positions or less.
Output cards are automatically created as a result of some
machine processing. This type of card is punched as a summary,
balance forward, file update, or exception card by other unit-record equipment. These operations will be discussed in more detail
in later sections.

INI~

R[JfCT[D .....

JI"'''~

, .. "I"UPAJA.\BUi .....

1----'-----

flll.lD

INVFN1rOIRY nCIKlET

•

TO BE RETURNED TO YOUR INVENTORY SUPERVISOR

HUM III OF

CO .. Tt.lfrill_'

000963
Figure 6.3 Stub card design.

Output cards can also be punched by computers and computer peripheral equipment. The cards may be created as intermediate records in a procedure, serving to hold intermediate
results of calculations which will be discarded later. On occasion,

102

PUNCHED CARD DESIGN

source record dual cards are copied or reproduced by machine in
order to preserve the original file intact while processi~g is done
on information in the duplicate file.
Special purpose cards take many forms, some of which ar.e
shown in Figure 6.3. Manufacturers can provide scored cards that
can be pulled apart, leaving a shorter card record of 51, 60, or 66
Some 80-column card machines can be especially
columns.
equipped to handle these shorter card lengths.
One of the most commonly used forms of special 80-column
cards is the punched card check. Checks usually originate as output cards from a payroll procedure. That is, the check may be a
summary card of the period earnings after all calculations of pay
and deductions have been completed for each employee.
Next, the check is interpreted; that is, the information
punched is printed on the card to show payee and amount. The
check is then made valid by affixing the authorized signature of
the payer. When the check is cashed, the payee endorses the card
on the back as would be done with any other type of check. The
card can then be machine processed by the bank and eventually by
the payer to reconcile his bank account.

Determining
Card Data

The type of card record to be used will be determined by the
requirements of the procedure. Sufficient data must be obtained
to produce the desired results. The following should be considered:

1. Source information, to be transcribed from original records or punched in dual cards.
2. Constant data, or such information as price lists, job codes
and wage rates, descriptions, names, sizes, customer names and addresses, and other coded data which must be inserted to complete
the card records. In most procedures the constant data is gangpunched or reproduced from master card files.
3. Calculation results. For example, to work the formula
Quantity X unit cost = total cost the quantity is obtained from
the source document as a transaction, unit cost from the price list
master file, and total cost as a product of a multiplication operation. Space must also be provided to punch any intermediate results that may be obtained in a calculation but which do not
necessarily appear in the end report.

PUNCHED CARD DESIGN

t03

4. Reference data should be adequate to identify each transaction with the source document from which it was created and
include the date on which it occurred. In some cases the data
should designate a person, place, or item responsible for or involved with the transaction. Examples include date, invoice number, batch number, salesman number, etc.
It is nearly always a requirement of the procedure to identify
the different types of cards by punched codes. This makes it possible, for example, for machines to distinguish between master and
detail cards or between transcript and dual cards.
The amount of reference data required will be determined by
the use to which the cards will be put and the reports to be written
from them.
5. Machine produced data. Other than calculation, certain
forms of data can be emitted by the equipment with the use of
special devices: digit emitters, storage, and so on. Such data is
sometimes required to obtain intermediate results.

The card design work sheet shown in Figure 6.4 is one way to
organize information to determine card field size and arrangement.
The number of columns required to record each type of information is listed. For codes, date, invoice number, account
number, etc., this is determined by the largest single number to be
recorded. Thus, four columns are needed for invoice number if
the number series is to be repeated after 9,999; two columns are
needed for office if there are no more than 99 branches.
For quantity and amount fields, the problem is slightly more
difficult. In the first place, the space needed to record the largest
amount may not be known. Second, if the largest amount is
known, its chances of occurring may be rare.
I t is therefore a good plan to provide for all except those rare
cases, and handle them by punching extra cards. For example, if
the seven-digit amount $67,265.80 is a rare case, it may be recorded in a six-column field by punching seven cards for $9,000.00
and one card for $4,265.80 or any convenient combination of
cards totaling $67,265.80. In designing summary and balance
forward fields, however, the amount fields must be made large
enough to accommodate the summary totals.
For punching names, 20 columns are usually sufficient. The
requirements must be checked on each individual job. As a general

Determining
Field Size

104

PUNCHED CARD DESIGN

CARD DESIGN WORK SHEET

"Pa:1:.i..M7.:t Ca.Ju:1

CARD NAME:

DESIGNER:

INFORMATION REQUIRED FOR
PROCESSING AND REPORT
PREPARATION

7~~
~j)ak--

S~P~.1
~

m ..-$,

FIELD SIZE
TRIAL

TRIAL

FINAL

/
5
4

5
4

20
1

LOCATION
IN OTHER
CARDS

SEQUENCE
ON SOURCE
DOCUMENT

1'1-22

20 20
/

4
(0

/
I

/2
10

0:(.0.

...8~~

4
2

"W c.

4
2

2
3
4
2

~P~

14
13

CcrJ.e,

.s~~

'R.croYrL ~

6
4

5
4-

5
4

Co.Adv

..9~

-P.0:vrv 2

,.110. ...., .+V-Q'

::D.

P"r.
C7

~.,-r"

-'J ...

4
/

7-/2

-8~.Ge...A/n,C
:z
'0
'a:
'..'0

!~
:3

99999999999999999999999999999999999999999999999999999999999999999999999999999999
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 19 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 4243 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 14 75 76

n 78 7180

:.~

,Ill
':2;

I:::>
'Z
,0

,a:

!~
:~

'1&1

9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 99 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9
I 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 SO 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 n 78 79 80

!IS
'Ill
:~

·z

:··0
.a:
ltl

99999999999999999999999999999999999999999999999999999999999999999999999999999999
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 4743 44 45 46 47 48 49 SO 51 5253 54 55 56 51 58 S9 60 61 6263 64 65 66 67 68 69 70 71 72 73 74 75 76 n 71 7180

1&1

III
:2;

:::>

z
~

....

(,)

9 9 9 9 9 99 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 99 9 9 9 9 9 9
123456789W"UI3~~~17I9~20~22n~2526n282930~323334~36n383940~~43444546048USO~~S354~56~58~60~~~646566~6869ro71nnu~76n767980

a:
w

III
~

:::>

....a:
~

iLl

9999999999999999999999999999999999999999999999999999 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 999
1234567'~W11UI3~~~17I9~20~22n~2526n282930~323334~36n383940~U4344454604848SO~~5354~56~58~60~~~646566~68nro71nnu~76n797980

Figure 6.5 Multiple-card layout form.

PUNCHED CARD DESIGN

109

After all fields have been laid out and sequenced, a separate form is
available for designing each card in detail. This form will be discussed later.

Keypunching, as a manual operation, is subject to a consider- Sequence of Data
able variation rate. Anything that simplifies it will tend to pro- on the
duce a faster and more accurate operation. Data should be punched Source Document
into a card in the same order in which it is read. If the data sequence on the card is considerably different from that on the
source document, it may be necessary to redesign the source document. There is sometimes greater flexibility in the arrangement of
the data on an original document than on the card record which is
to be machine processed.

Cards can be designed to take full advantage of both machine Machine
characteristics and processing techniques. Although equipment Characteristics
operation has not been discussed yet, the logic of some of the following considerations should be apparent as good practice in all
accounting procedures.
Card Punches

In designing a dual card, it is better to locate the handwritten data towards the left end and the punched data on the right.
This keeps the handwritten information from being hidden under
the card machine pressure lever while the card is read for punching. Figure 6.6 shows what portion of the 80-column card can
be covered when the card is in the bed of the 129 Card Punch.
When other equipment is to be used, the card design should be
checked against the manufacturer's recommendations. Figure 6.7
shows standard and dual 96-column card visibility as these cards
would appear in the hopper of the 5496 Data Recorder. The
shaded areas represent portions of the card that are covered by
machine parts. Figure 6.8 shows dual and standard card forms
at the verify and wait station of the Data Recorder. As shown by
the shading, a larger area of the card is covered here. If data keyed
into a dual card is to be verified, the portion of the card with handwritten source data must be visible to the operator at this station.

110

PUNCHED CARD DESIGN
letters

ABCDEFGHIJKlMNDPQRSTUYWXVZ

UDunn

/'Y'i'''V''AYV/\ll'\J.I''.I''Y'J.I''.I'''fi.l'''lfi~O

Special Characters
~:c.

<<:+ 1- !$it:); ... ,/

nilln

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 a1mB D

0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 aaag

~ ~g
[ll!'i-.''.:I{NNNNl'~NNI'-IJ\.
/v~'vV'YVlrV~-1';yfY'fV1'v('iA'V.,iiI 111819201112231415262118193031323334 J5 36 3138 3940 41 4Z U 4U5 46 4148 4950 5151535455565158 59 60 616163646566 6sTlr~~~~~~~
/ ' 1 ' " . / , / . / , / , / , / , . , . ..

/,-/'"/"/,"/,",1"/

11111111111 B111111111111111111111111111111111111111ll\..I\..I\..N''''''-.NV''''"J'\.

r . .,....r.r..r.r/'·.r.rJ 2 2 212

2 2 22 2 2212 22 2 22 2 B2 2 2 2 22 2 2 2 2 2 2 12 2 2 2 22 122 22 2 2 12 2 2~~l~'f'f'N'\j';'N''f\

3 3 3 313 3 J 3 3 3 3 3 a3 3 3 3 3 3 313 3 3 3 3 3 33 3 33 3 13 3 3 3 3 3 13 3 3 3 3 3 B3 3~~~~~~~

AIa.~~~",,"~fMIUl4

4 4 4 414 4 4 4 4 4 4 4 4 4 4 4 4 4 4 14 4 4 4 4 4 4 4 4 4 4 4 14 4 4 4 4 4 14 4 4 4 4 4 14 LN'\l'\N'.M\.f'o,J""J'\,""'''''J',,",''

/Y.)o~vYll'Y.J,tvv.v'VY.ihl5

55 5 5 515 5 5 5 5 5 5 505 5 5 5 5 5 5'~ 5 5 5 5 5 5 5 5 5 5 5 5 15 5 5 5 5 5 D5 5 5 5 5 5 ~
.. ~~~~'N\

.lVil-~~1lIflI:lI.~MAiAjJ6

6 6 6 6 6 616 6 6 6 6 6 6 6 D6 6 6 6 6 6 6 16 6 6 6 6 6 6 6 G6 6 6 16 6 6 6 6 6 B6 6 6 6 6 6w,;.b'.h'Ii~-fj..!,'\(NNi'oIl'Il\.

/V~'Vl/V1'/!)'I./I.lIl./l/.'V'/1:I'./I...I7 77 77 77 7117 7 7 7 7 77

D1 7 7 7 11 7 &7 1117 7 77 17 1 7 17 17 1 77 e77 711lr-n.I'..I'..I'..I'../"J.J'V'J'J'V'\.

~~~~~m18 888 8 8 8 8 8D8 8 8 8 8 8 8 8 IS 8 8 8 8 8 8 D8 8 8 8 8 8 DB U Ba8 nBDame 8 DO a~"'I~'~~~~~
~~~~~~~9 9 9 9 9 9 9 9 9 919 9 9 9 9 9 9 9 m9 9 9 9 9 9 9 19 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9l~~~~~~
rA"/"/Y"/'.Y'/,..."/,/'V'/'./>/"JI1 111910 1111131415 1"1/819 30 31 3Z 3334 35 36 3138 3i '" 41 4143 4J 4546 H 48 49 50 51 52535455565158596061 6163 6l 6~ 66 61P.

Nonvisual Columns
rzzZZZZZZ4 Model 2

IZlIlZ211liZZZZZZOJ Models 1 and 3 (Punch)
~ Models 1 and 3 (Verify)

Cards with upper left or upper rignt round corner
cuts are acceptable. Cards with lower corner cuts (except
lower left corner cut C·3. 3O-d89ree. O.130·inch base
measurement) cannot be fed.

Figure 6.6 Card visibility at punch station, IBM 129 Card Data Recorder.

Sorters

For sorting operations when cards are being sequenced by
several control codes (e.g., branch, salesman, and product), operator control is simplified if the fields are adjacent to each other.
The minor field should be located to the right; in this case, the
product field is minor. Each progressively higher field (salesman,
then branch) should be located to the left
Automatic Punches

A number of 80-column machines punch serially from left to
right, column by column. When this type of equipment is to be
used in the processing, cards must be designed so that the data
sequence does not conflict with the sequence in which results will
be supplied and punched by the machine. Such a conflict can be
avoided by studying the manufacturer's manuals of operation and
the way processing is to be done.
Computers

Computer characteristics must be investigated to determine
whether the card's design will affect processing speed. With many
computers the ability to move and process blocks of data without

PUNCHED CARD DESIGN

rearrangement makes programming easier and increases processing speed.
ControIPane/~"ing

Many 80-column card machine units are set up and controlled by means of wired control panels. The accuracy and speed
of frequent wiring operations can be improved if the input card's
data sequence is similar to the sequence of reading, punching,
and printing.

" ••

~Q_m*

____ g

•••

a*a.~.M_maw~~~m_

A
8

4
2

~tZJ4S'7 "Dn~DUD."

••

»~UHNH."HH.~U~
A

A
8
4
2

~DM
A
8
4

8
4
2
•• V •••

~UU"

••

o_.~uaUM

••

~ • • M.UUM~
A
8
4

2
2
' • • p.H~~nnn~~"~~ • • • • M • • P • • • flN.N."'

Figure 6.7 Dual aIld standard 96-column card in hopper of 5496 Data
Recorder

111

112

PUNCHED CARD DESIGN

A
8
4

2
1

811345.7

• . • . .0111213 ...

8
4

2
1
B

S3 M 31 M S7 , . HOG ., U U . .

a .. .,

A
8
4
2

Figure 6.8 Dual and standard 96-column card in verify station of 5496 Data Recorder.

Manual Card Handling

If data are to be punched and printed on the 80-column card
punch, information to be checked visually should be placed near
the right or left end of the card. This makes it easier to fan and
check a group of cards. If data are to be read from cards in a file,
the reference data should be positioned so that it can be easily seen.
Positioning of data is somewhat more restricted on the 96column card because of its smaller size.

PUNCHED CARD DESIGN

113

96 - Column Card Layout Form

~_~R_~_f_:~_~_T_-H_~ _-_~~_~_~_-V-_E

---\j'~~~1~~~!

1111/ - - 1

10 11

12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31

PRtN1~~11

'PRINT AREA

-------------]

B
A
8
4

2=-

1 ==
B==
A==

1
B
A
8
4

B~I
A:=
8==
4==

97 98 99 100 101 102103104105 106 107108 109 110 111 112 113 114 115 116 117 118 119 12C 121 122 123124 125 126 127 128

10 11

12 13 14 15 16

17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32

1
B
A
8
4

rt~C;~;u
B

33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 5758 59 60 61 62 63 64

65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91

~~~~IB~M~~~~~~~~~~~~~~~~~~~~~~~~~~~

Figure 6.9 96-column card layout form.

:::::B

8
4

=;;8
:::::4

-=2

B
A
8
4

:::B
:::::A

2:=~
1 ==
1
-

~t~~~~li~

65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96

~;t;r;~~i

~il

1----------------

33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64

~~t~1~li~'
B:=
A==
8:=
4==

---------------

\ MA-C-HTI,fE

B
A
8
4

I~~

I~B
:::::A
:::::8

:::::4

-.j:l.

It ,II "

11l_

"It'.,

llil

1,II111J1U

.:0 /1'.

I.,

l~_a

15/

'~,

1)1

1.' liS, II'

-·14·~-

-1-

n'IGllIlIlJl]~)iJl

J~):'OI~'1

nIl I:

l~' ;~

'\'

:~_o :\,'~ 41'!!

UU"41UUUUS1H5)~U5'''SIU

.:'

,~1.

1,1S

,~,

"41 S

"II.:

5: __~i

'~l_.\~'~

'sSII 5:

S~I.

u ,'".

,',., u

~)'"

I' ,If

.:j

;13145

imiDO 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0

:~45

i!~45

123456789W"nn«ffi~u~~wnnn~~~Du~~~~~~~~n~~~~~~«~~~U~~~~~M ~~~~~~~~~M~"~"~~"nnu~nnn~w

III

111111111111111111111111111111111111111111111111111 11111111111111111111111111111

[iJ

2 2 2 2 22 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2

3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 33 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 33 3 3 3 3 3 3 3 3 3 3

44444444444444444444444444444444444444444444444444444444444444444444444444444444

55555555555555555555555555555555555555555555515555555555555555555555555555555555

6 6 6 6 6 6 6 6 6 66 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6

77777777777777777777777777777777777777777177777777777777771777777777777777777777

ic.

8888 88 88888 88 8888 888 888 8 888 88 888 88888888 888 88 8 888 88 888 888 888 8 88888 88888 88 8888 888

IBl

9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 99 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9 9
123456789W"nn«~~u~~wnnn~~~DH~~~~~~~~n~~~~~~«~~~U~~~~~M ~~~~~~~~~M~"~"~ro"nnu~nnnnw

Figure 6.10 80-column card layout form.

c
00

PUNCHED CARD DESIGN

Once field size and data sequence have been determined, the final
card design can be laid out. Several types of forms are available for this purpose; samples are shown in Figure 6.9. To make
design easier, these forms are considerably larger than standard
size cards. By photography and reduction techniques, the forms
can be reduced to proper size for making printing plates.
When drawing the card layout, there are a number of additional points to be considered. Again, the value of each must be determined by the card being designed and the application in which it
will be used. Figure 6.10 shows the layout for a material accounting
card. Much of the following discussion can be related to this form.
All field and box headings should be explicit and confusing
abbreviations avoided. Headings for written information should
be placed so that they force writing into the proper location. Make
certain that adequate space is provided for this writing.
For dual 80-column cards, the right-hand side of each area or
box containing handwritten data should be at least five card columns to the left of the field in which the information is to be
punched. This prevents the data from being obscured when it is
time to punch it.
The shield in the card punch that extends across the bottom
of the bed between the punch dies and the reading station obscures about one-eighth inch of the card's lower edge. Avoid the
use of this area for data that is to be read and punched into the card.
For easy reference, field headings should be located between
the zero punching positions and the top of the card unless printing
or mark sensing prevents it.
Keep titles and interpreted data between rows of punching
so that they will not be obliterated.
Use pre-printed decimals and commas in boxes where amounts
will be machine printed.
Colored cards and color strips may be used to identify types
of cards. However, color should not interfere with the utility of
the card, particularly the reading and writing of data.
Dotted lines should be drawn between columns (both card
and mark sensing) to indicate the locations of decimal points
and commas.
When designing a dual card, use one area of the card exclusively for handwritten data so that it will not be obliterated by
punching. If there must be an overlap of handwritten and punched
data, use the overlap for recording descriptive information that
that can be easily reconstructed.

115

Final Card
Layout

116

PUNCHED CARD DESIGN

Whenever possible, retain the digits that show the location of
the punching positions. This makes it· easier to read punched information that has not been printed on the card.
Any information, handwritten or printed, that is to be visually
checked should be placed near the right or left end of the card.
In this location, the information can be more easily seen when the
cards are fanned.
Mark sensing should be placed on the right end of the card so
that the card can be held with the left hand and marked more
conveniently.
Corner cuts can often be used to distinguish card types during
clerical as well as machine processing operations. However, lower
left cuts should not be used on some type machines because
they may cause feeding problems. Be sure to check the manufacturer's recommendation.
Special consideration must be given to the design of short
cards. Their position in the hopper of the processing equipment
is predetermined and the card columns should correspond with
the positions of the individual column sensing devices. This will
make machine sorting and control panel wiring easier.

Questions

and
Exercises

This exercise involves the design of an 80-column dual card form to be
used as a requisition for supplies from a small office stock room. Requisitioned items are stationery supplies: typewriter paper, scratch pads, pencils,
paper clips, rubber bands, typewriter ribbons, and the like. Assume that the
activity in the stock room has been surveyed and the following information
established:
a. About 100 different items are carried in stock.
b. No more than 100 pieces of an item are issued on anyone requisition.
c. A separate requisition is ftlled out for each different kind of item.
For example, ribbons and paper cannot be ftlled from a single
requisition.
d. Item description must include at least two fields: name (e.g., bond
paper, rubber bands, paper clips) and unit of measure (e.g., box,
dozen, sheets, pieces, bottle).
e. The stockroom is located in the same building with a number of.offices and departments that obtain supplies there.
f. The stock room is tended by a clerk who services a window. I terns
are picked up by secretaries who carry the supplies back to their
desks for their own use or for distribution to other office workers.

PUNCHED CARD DESIGN

g. Requisitions are originated by the secretaries and are authorized by
managers of individual departments.
h. At the end of convenient periods, filled requisitions are priced and
extended (quantity issued X unit cost) to obtain the total cost of
each item issued.
i. Each department is charged with the cost of supplies used each
month as an expense.
1. List the information fields that are written (or typed) on the requisition
by the secretaries.
2. List the field or fields written by the managers.
3. List any fields written by the stock room clerk.
4. Which fields should be punched in the card?
s. Determine the number of columns needed to record each field by keypunching.
6. DeSign a proposed card to meet the requirements just listed.
7. Outline instructions to the secretary, stock room clerk, and manager for
using the form.
With the instructor's guidance, the class should agree upon a single form
that meets all estimated requirements. The operation of the stockroom can
be simulated as a class activity, with a number of requisitions produced as a
result. If equipment is available, these may actually be keypunched, by class
members or by a data processing department.
If 96-column equipment is available, the same exercise can be carried
out using the smaller card as a requisition.

117

PROCEDURE
CONTROL

Data processing procedures can be divided into three equally
significant parts.
First: the procedure must achieve some planned result and,
for commercial enterprises especially, produce documents that record those results. For example, a payroll procedure is designed to
compensate employees for their work at periodic intervals according to agreed-upon rates of pay. The records produced include paychecks; earnings statements; payroll registers; and the many related
reports for the employer and local, state, and federal governments.
An accounts receivable procedure is designed to keep an account of the money owed a business. Records produced include
monthly statements of customers' accounts, notices of delinquent
payments, credit records, and adjustments to account balances for
payments and purchases.
Second: the results must be accurate and dependable. Requirements for accuracy differ with various types of procedures.
In the case of payroll, where employee wages and tax records are
concerned, there seldom can be any compromise with accuracy.
It is in the employer's best interest to be sure that every employee
on the payroll is paid correctly. In the case of accounts receivable,
it is also in the best interest of a business to see that customer
statements are accurate in every respect.
At the time a procedure is planned, it is necessary to assess
the requirements for accuracy against the cost of obtaining it. For
118

PROCEDURE CONTROL

applications such as cost accounting, labor distribution, and inventory control, a degree of accuracy somewhat less than absolute
can usually be tolerated.
To illustrate this point, there is a story of a munitions plant
during World War II that received an emergency order for a number of specially designed small arms. Because of the critical and
unusual nature of the order, the plant manager assigned his most
dependable expediter to take personal charge of production. True
to his reputation, the expediter was able to effect a smaU miracle.
By largely ignoring established operating procedures, he was able
to have the complete order ready ahead of schedule.
As the plant manager and his man stood on the shipping platform to watch the last of the crated pieces being loaded on a truck,
the manager was extremely. generous with praise for a job well done.
As they turned to leave, the manager paused a moment, then
asked, "By the way Charlie, how much did those things cost?"
Said Charlie, in a classic reply, "About half as much as they
will if we try to find out!"
It should also be noted that a data processing department is
nearly always a service unit within a larger organization. The department accepts information in some form to be processed. Usually, the incoming information is in the form of source records
which must be transcribed to punched cards for further processing.
The processing steps may include sorting, classifying, reproducing,
calculating, report printing, summarizing, and so on. Processing
may also include the preparation of records for further processing
by other types of equipment, including computers and communication equipment.
No matter how complicated the procedure or what the end
result will be, it must be remembered that the machine processing
department does not originate any information. The department
can only handle data as received.
It follows, therefore, that final results can be only as accurate
as the input data. Wrong data can often be processed as easily and
quickly as that which is correct. Systems engineers have a rule
that applies to this situation:
"Garbage in; garbage out!"
In general, the requirements for accuracy in a procedure may
be summarized as follows:
1. The financial status of the business must be accurately reported to management and the stockholders.

119

120

PROCEDURE CONTROL

2. Customer; employee, and vendor records must be maintained with a minimum of error.
3. Accounting for tax purposes, company earnings statements, and government reports must be accurate within the requirements of law.
4. The accuracy of operating reports must conform to the
requirements imposed by internal and external accounting practices. Procedures must be modified to meet these requirements.

Third: the control of a procedure mustbe exercised in such
a way that the results are achieved at reasonable cost and within a
specified time limit. This might also be called the management of
a procedure.
For example, a payroll procedure must be managed so that
employees are not only paid correctly and according to law, but
also so that they are paid on time. In accounts receivable applications, customers must be billed by certain dates and their account
balances must be kept as current aspossible.
Data must be handled in a fashion that makes processing as
efficient as possible. This may involve such steps as machine duplicating of card records so that two different operations can be
carried out at the same time.
Control must provide for the organization of a procedure in
such a way that, in the event of error, the trouble can be found and
fixed with a minimum of lost time. For example, it is possible for
a card record to be jammed or torn by the equipment or by careless handling. If this happens, the corresponding source record
must be available for reference so that the damaged card can be
repunched and put back into the file without costly interruption.
Stringent control must be exercised to prevent theft or fraud
that may affect the financial standing or reputation of a business.
This requirement particularly applies to the production of records
that can be negotiated for cash or other value; for example, checks
of all kinds, bonds, and stock certificates.
The procedure must be designed to leave an easily followed
audit trail, that is, a means of auditing by tracing the flow of data
from input to final output.
This chapter is concerned with the methods used to establish
control and to assure accuracy over the transcription from source
record to punched-card record. Direct machine verification is discussed in some detail with descriptions of the equipment used in
this operation.

PROCEDURE CONTROL

121

As other card processing operations are presented in later
chapters, appropriate methods of control will also be described.

The use of a record count is one of the simplest ways to maintain control over the movement of documents from one stage of
processing to another. An original count is obtained from the tally
of the number of records in a given file or set. The count can be
useful as a quick check that all source documents have been received into a data processing department.
Record count totals should be established at the location
where the documents are originated or first assembled. The total
is obtained by a manual count. The count is then adjusted whenever any re-cords are added or withdrawn from the file. If documents are processed or transferred to a new location, the records
are recounted and the result compared with the original or adjusted total. If the quantity agrees with the control total, it is accepted as proof that all records have been transferred, or processed
in some other manner as called for by the procedure.
Record counts are usually established by batch. This is desirable when all records involved in a given procedure may not be
received into the department at one time. After the batch totals
are set up, they are added together to form the single total for all
records processed.
Although the record count is useful as proof that all records
of a given set have been accounted for, it is difficult to determine
the cause of error if the record count is out of balance. A failure
to balance against a control total does not help to locate a missing
record nor does it indicate which record may have been processed
more than once. Therefore, some provision must be made to check
the error by using a duplicate set or a listing known to contain the
proper number of records.
An error in a record count is merely the first indication that
something is wrong in the operation of the procedure. More elaborate controls must be established to locate the cause of the error.

Record Count

Whenever the term "control total" is used, it is normally taken
to mean a total of some amount or quantity fields in a group
of records. However, as just explained in the section describing

Control Total

122

PROCEDURE CONTROL

record counts, the term can be applied to almost any accumulation
of factors or items that might be used for procedure control.
The control total is accumulated either manually or by machine when a record set is first originated or assembled. Totals
are usually taken by small, easily handled groups of records to
form batch totals which can be added together later to form the
grand total.
For example, assume that a batch of invoices is received into
the machine data processing department. Attached to the batch is
a transmittal slip stating that there are 169 records in the batch.
Now assume that these records are to be keypunched.
The ftrst step in processing is to verify that all records have
been received. This is done by fanning through the bundle and
manually counting the documents. If the count proves to be 169,
then it may be assumed that all records have been received.
Next, assume that it is the job of the receiving Clerk in the
keypunch room to establish a total of the amounts of the invoices
for control purposes. The clerk does this by preparing an adding
machine tape of the totals from each invoice. The tape is attached
to the bundle of invoices and the total is entered on the transmittal slip.
The invoice data is now keypunched into cards in the normal manner, with one card prepared for each item. But the totals
of each invoice are not punched-only amounts for the individual items.
The cards are machine verified for identification data, quantity, and so on. Item totals are not verified. When all cards for
the batch have been completed, the bundle of invoices along with
cards, transmittal slip, and amount total are sent to an accounting
machine. An operator feeds the punched cards into the machine to
add the punched item amounts. At the same time the card data is
listed on a sheet of paper showing all the information punched.
The machine is controlled to print totals by invoice number so
that the listing is a composite record of all information transcribed.
At the end of the listing the machine also prints a grand total.
If all the previous work has been done correctly, the grand
total from the accounting machine should balance to the original
adding machine tape.
If there is an error, each machine total can be checked against
the corresponding original invoice total. When the error is located,
, each detail item is checked until the exact difference is accounted
for. It may be that an error was made in transcribing more than

PROCEDURE CONTROL

123

one item. It is also possible that an error was made in taking the
adding machine total.
When all corrections have been made and the two totals reconciled, a final and correct listing is run on the accounting machine.
This listing can now serve all control purposes that the original
documents served. Each time the corresponding card records are
processed, they can be balanced against the control listing. In case
of error, all source information is readily available for checking.

Closely related to the record count, the unit count differs because no count is established for the actual number of records or
cards. Instead, the total is set up for the number of different units
in a file. In some applications, particularly payroll, units may
mean employees.
For example, a payroll is always made up for a known number of employees. Therefore, if a batch of weekly clock cards are
received in the machine room as part of a payroll procedure, there
should be one card for each employee. This total is established
first by a record count and may be verified by the personnel department or by the various department managers or timekeepers.
As the payroll is processed, employee clock cards may be
combined with other cards, such as master name cards with employee name, serial number, social security number, rate of pay,
and so on. Tax records and previous year-to-date earnings cards
may be included with other deduction cards. When the final payroll is ready for processing, there is an undetermined number of
cards in the file for each employee. However, the total number of
employees must still balance to the original unit count.
When the final payroll records are prepared, the card processing equipment can be controlled to count the number of minor
totals taken on the report-this is one for each employee. Regardless of the number of cards in the file, the machine counts
totals only and consequently, the number of employees. This total
is carried through each step of the payroll procedure. Ultimately,
when checks are written, the unit count control total must equal
the number of employees on the payroll.
The unit count is often combined with other control totals to
provide procedure accuracy. In combination, the two totals can be
used to predetermine amounts or quantities on any reports produced in a given procedure.

Unit Count

124

PROCEDURE CONTROL

For example, when preparing to process a payroll, the total
number of hours worked by all employees can be established from
the original clock or job cards. This figure then becomes the control total of payroll hours for all subsequent reports. Totals may
be broken down by group or department to make it easier to localize any possible errors. The sum of all totals must balance back
to the complete original control total.
Control totals are usually set up by batches of convenient
size, such as department, location, office, account, or division. By
this method, each group of records may be balanced as it is processed. Corrective action, when needed, can be limited to small
easily checked groups rather than one grand total.
To again use a payroll procedure for illustration, it is usually
best to set up control totals by department, location, or office.
As the various reports are run, each group can be balanced to these
basic control totals. Control of earnings, tax amounts, and deductions for a current period, and cumulative year-to-date amounts
will be discussed with the operation of calculating equipment and
accounting machines (Chapters 11 and 14).

Self-Checking
Number

The dependable operation of almost any data processing procedure
depends to a considerable extent upon the accuracy of control
numbers. These are the identification codes assigned to records,
items, persons, locations, accounts, and so on, to make efficient
machine accounting possible.
The entry of an incorrect control number into a procedure or
processing cycle always produces an error of misplaced data. That
is, the information .relating to the control number, usually values
or quantites, is improperly identified. Consequently, that information may be lost. Or, the use of a wrong number can associate
corresponding information with another identification code which
by coincidence happens to match the error number.
The self-checking number is used with numerical identification fields to assure accuracy whenever the field is written or transcribed. The number assures a very high degree of accuracy for
those fields, even on the original source document. When used in
punched card fields, special features must be installed on the equipment by the.manufacturer.

PROCEDURE CONTROL

One typical use of the self-checking number is to positively
identify the recording of items in an inventory. In such an application, handwritten original documents are common. And in nearly all cases, writing the proper part number on a record is just as
important as putting down the correct quantity of parts moving in
or out of a stock room. In this application particularly, an error
in identifying the item or part can multiply into a surprising number of other errors later on in the procedure.
For example, suppose a stock room attendant releases a quantity of item 1234 but inadvertently records the withdrawal of item
2134. It is assumed that 2134 is also a valid part number for
another item in stock.
When the periodic inventory report is run, it may show that
item 2134 is nearly out of stock and should be reordered. Consequently, a surplus of the wrong item eventually gets to stock.
Meanwhile, item 1234 shows a balance on the report of more
pieces on hand than are actually on the shelf. If item 2134 is
stocked at a cost of $100 each and item 1234 at a cost of only
$10 each, then the error has caused an unnecessary and perhaps
serious inflation of inventory costs.
Other costly errors can accumulate as a result of the incorrect recording of the part number. Suppose that part 2134 was
recorded as withdrawn from stock in a maunfacturing plant where
the cost of parts is charged against a shop order for work being
done for a customer. In this case, the cost of the order is increased
by $90 for each part used. Remember, the correct parts were
issued; only the identification number was wrong. Conversely, if
the cost of the recorded parts is less than the cost of the ones
actually used, a serious loss could occur for the plant.
The use of the self-checking number requires that a check
digit be developed for each basic code number to be self-checked.
Once the check digit for the basic code number has been determined, it is added to the units or last position of the number. The
check digit can be developed by a self-checking number generator
on the keypunch, by computer calculations, or by manual arithmetic. Thus the self-checking number is one digit longer than the
original base number.
A number of terms should be defined before describing examples of two different check digit systems.
Transcription error: This is an error of substituting one or
more digits in a number for the correct digits.

125

126

PROCEDURE CONTROL

Examples:

0 for ai, 2, 3, 4, ...
1 for a 0, 2, 3, 4, . . .
2 for a 0,.1, 3, 4, ...

Transposition error:
adjacent columns.
Examples:

457843 for 458743
865745 for 865754

Double transposition:
alternate columns.
Examples:

This is an interchange of two digits in

This is the interchange of digits in

457843 for 454873
865745 for 875645

Random error: A combination of two or more of the above
or any error not covered by the above.
Weights: The amount by which each digit in the control
number is multiplied to arrive at a product.
Example:

457
842
432
765
28 +30 +35 +32 +12 +4

=
=

number
weights
141 = sum

Modulus: The number used to divide the sum of the digits in
the control number to arrive at a remainder.
Examples:

141 divided by 10 = 14
141 divided by 9 = 15
141 divided by 11 = 12

+ 1 remainder
+ 6 remainder
+ 9 remainder

Check digit: The number added to the sum of the products
(weights X control number) to produce a number which when divided by the modulus will give a remainder of zero. Or, it is the
difference between the sum of the products and the next highest
even multiple of the modulus.
Examples: 29 divided by 10 (modulus) = 2 + 9 remainder.
9 + 1= 10 or 30 -29 = 1. Therefore, 1 is the check digit.

PROCEDURE CONTROL

127

141 divided by 11 (modulus) = 12 + 9 remainder.
+ 2 = 11 or 143 -141 = 2. Therefore,2 is the check digit.

Modulus lOis designed to detect either incorrect keying of a
single digit or a single transposition. The following arithmetic
process is used to generate the check digit, regardless of whether it
is generated by a keypunch, a computer, or manually. Refer to
Figure 7.1.
Example:
Basic code number
Units and every alternate position of basic code number
Multiply by 2
Product
Digits not mUltiplied by 2
Cross add
Next higher number ending in zero
Subtract crossfooted total
Check digit
Self-checking number

6
2

X2
6
4

1+2+1+5+4+6

= 19

20
-19
1

Other examples:
Basic code number

45626
30759
73074

Self-checking number
Basic code
Check digit

45626
30759
73074

9
5
7

Figure 7.1 Self-checking number. Modulus 10.

1. The units position and every alternate position of the basic
code number is multiplied by 2.
2. The digits in the product and the digits in the basic code
number that are not multiplied by 2 are cross added.
3. The cross-added total is subtracted from the next higher
number ending in zero.

4. The difference is the check digit.
In this system, the space and zero have the same numeric
value. Therefore, spaces can be substituted for insignificant zeros.
Once the number is generated, no special knowledge is required to write the numbers on an original document or to punch
the numbers into cards. With an 80-column keypunch, the operator keys and punches the number, including the check digit, as it
appears on the source document. The accuracy of the keying and

Modulus 10

128

PROCEDURE CONTROL

the validity of the self-checking number can be automatically
verified by the machine.
When the number on the source document is correct and the
number keyed correctly, the punching operation is not interrupted.
More than one self-checking field may be checked in one card, but
the fields must be separated by at least one card column. If no
errors are detected after punching all the self-checking fields in a
card, an 11 hole is automatically punched after column 80 (in
what would be column 81).
If the number is incorrectly recorded on the source document, or if the number is punched incorrectly, an error is indicated
after the last digit of the self-checking field is punched. A red light
appears on the keyboard, a 12 hole is automatically punched in the
units position (check digit) column, and the keyboard locks. When
an error occurs, the automatic 11 punch in column 81 is suppressed.
Self-check
Number

ARMSTRONG
17 . . . . Z02122Z3Z.. 25H27Z.2t3031:u

275636
425

B... ....
~

••

m~mm~m~m~a~m~n~~~m~M~w~w~mm~~

A •
••
•
8.
•
4
•
•••
2
••••
1 ••
••••
B' 23'S'7"~H~BW~

n~~~~U»~HaVHH~~»B

Self-check
"OK" Punch
Self-check
Number Field

8
4

•

8
4

••
•••
••
•

2
1

•••

2
1

BnM~~»HH~~«U«6_U

••

~~R"~"H~~"~~fl~UB

8
4
2
1

A
8
4
2

•

8
4
2

"MP."~nnnu~~nnn~~NUN~"PH":~N""""l
1.113700

Figure 7.2 Self-check punch location. 96-column card.

The 5496 Data Recorder does not generate the self-checking number; self-checking fields are verified as they are punched.
When no errors are detected, an OK punch is automatically made
in the margin of the card near column 32. Figure 7.2 shows the
96-column card with a self-checking number field that has been
correctly punched.

PROCEDURE CONTROL

129

With the modulus 10 system, the earlier example of the incorrectly recorded part number can be used to prove that a selfchecking number would have revealed the error. First, the check
digits for the two numbers are generated as follows:
Basic code numbers
Units and alternate positions of
basic code numbers
Multiply by 2
Product
Digits not multiplied by 2
Cross-add
Next higher number ending in zero
Subtract cross-added total
Check digit
Complete number

2 3 4

'4
X2
8

4
X2
8

4
3
1
1+4+3+8

3 4

2
2
3
2+2+3+8

16
20
16
4
12344

= 15
20
15

5
21345

Now, the stock room attendant must write a five-digit part
number for each item issued. If the exact same transposition is
made using this system, the attendant would write the part number
12344 incorrect as 21344. Because it has the wrong check digit,
this number would be caught as an error on the keypunch.

Modulus II is designed to detect single digit mispunches, single
transpositions, and double transpositions. The main feature of
this system, distinguishing it from other self-checking number systems, is that it is based on a weighted checking factor for each
digit in the basic number being tested. Regardless of how the
self-checking digit is generated, the following arithmetic process
is used. Refer to Figure 7.3.
Basic number: 943457842
Write digits of basic number: 9

From right to left, write
checki ng factors:

Add the products:

36 + 12 + 6 + 28 + 30 + 35 + 32 + 12 + 4
Total 195

Divide: 195 divided by 11 = 17 plus a remai nder of 8
Subtract: 8 subtracted from 11 = 3 (the check digit)
The new self-checking number is: 9434578423

Figure 7.3 Self-checking number. Modulus 11.

Modulus 11

130

PROCEDURE CONTROL

1. Each digit position of any basic number is assigned a weight
or checking factor. These factors are: 2,3,4,5,6,7,2,3,4,5 ... ,
starting with the units position of the number and working toward
the high-order digit.
2. In the example shown in Figure 7.3, write the number as
illustrated, leaving space between the digits.
3. Below each digit, starting at the right and working to the
left, place the corresponding weighting factor.
4. Multiply each digit by its weighting factor and add the
products.
5. Because this is a modulus 11 operation, divide the sum of
the products by 11, and subtract the remainder from 11. The result is the check digit.
In the modulus 11 system, basic numbers that require a check
digit of 10 cannot be used as self-checking numbers. The coding
system must be adjusted to eliminate such numbers that are to be
self-checked. If an operator is generating check digits on an 80column keypunch, and punches a basic number requiring a check
digit of 10, the machine indicates an error. A 12 punch is placed
over the check digit column, the error light goes on, and the keyboard locks. The operator must release the card and substitute
another basic number.
When the calculations on the basic number result in a check
digit of 11, the digit 0 can be substituted and appended to the
basic number to make a valid self-checking number. Eleven and
zero have the same value in this part of the calculation.

Document Register

A document register can often be used effectively to control the
flow of records in and out of a data processing center. Such a
register is most effective where single records or small batches are
received at frequent intervals by a control clerk. The register is
maintained as proof of receipt, processing, and eventual transfer
of the documents from the center to another area. Further processing of some sort mayor may not be done there. Figure 7.4
shows a sample order register as it might be used in a billing and
accounts receivable procedure.
To begin the system, all the orders are serially numbered at
the point of origin. Each order must have a number, and orders

PROCEDURE CONTROL

131

ORDER REGISTER
MONTH
DATE
RECEIVED

10/14
II

1/
1/

/I
II

"
1/

II
II

10/ 15
II
II

ORDER NUMBER

12831
12832
12833
12834
12835
12836
12837
12838
12839
12840
12841
12842
12843
12844

DATE
AUDITED

DATE
BILLED

DATE
SHIPPED

\0/14

\O/IB

10/10

\0/1 B
10/10

10/15
10/14-

10/17
10/16

10/1 B
10/17

10/17

10/15

10/17

\0/15
10/16
/I

REMARKS

10/15
10/14

10/15
/I

10/16
10/17

10/17

10/19

[0/19

10/18
~~.~

Figure 7.4 Order register.

are transmitted to the data processing center in serial number sequence whenever possible, although there may be some exceptions.
As orders are received by the control clerk, each one is checked off
against a preprinted corresponding serial number on the order
register. Missing orders are immediately identified by this posting
operation. The serial numbers can also be copied by hand if this
is more expedient.
Columns are provided on the register to record the date when
each order is received and to check the completion of other operations as well. Processing points may be either checked off or recorded by entering the date in the proper column. In this way,
the register serves not only to control actual processing, but also
to provide the time each major operation is completed.
Next, it is assumed that the orders are to be transcribed to
punched cards and then verified. Both operations are checked off
in the proper columns. From the punched cards, it is then assumed
that an invoice is prepared. A number of detailed steps can occur
here, all carried out within the machine room. When the invoice is
completed, it is sent to the shipping department for entry of the
date shipped. A column is provided for remarks.
The register can be used to flag delays in processing as shown
in the case of number 12843 (Figure 7.4). After several days, this
order has not been shipped.

132

PROCEDURE CONTROL

It should be noted that a rather elaborate control of this kind
would be maintained only for important and valuable records. In
this case, the register shows the response time of the procedure,
and probably the business, from receipt of any order to its final
shipment.
If the volume of items on the register is large enough, it is also
possible to mechanize this document by preparing a dual punched
card for each line, keypunching the cards, and listing the register
on an accounting machine.

Serial Numbers
for Batch Control

Serial numbers on documents can be used to check that all proper
records are included in a procedure. By arranging the records
in serial number sequence, a check can be made for missing or
duplicate documents at any point in the processing. This system
furnishes effective control of records such as checks, drafts, bonds,
and other valuable papers when all such documents must be accounted for throughout a processing cycle.
If the document is a punched card, the serial number can be
both printed on the card and punched as holes. Cards can be
sorted by serial number and counted automatically. Missing or
duplicate numbers can also be located by machine.
Serial numbers can also be used to identify batches of records.
The number of documents in each batch is recorded on a control
..........

I
BATCH NO.

2
TO:

~ D.ep;;t.

FROM:

-P~

/0/13 /
DATE

j!)~.

NUMBERED

12355

NO. OF DOCUMENTS

12391

FROM

RECEIVED ATTACHED DOCUMENTS SPECIFIED ABOVE

DATE

SIGNATURE
PLEASE SIGN AND FORWARD THE COPY OF THIS BATCH CONTROL
TICKET to SENDING DEPT.'WITHOUT DELAY

Figure 7.5 Batch control ticket.

PROCEDURE CONTROL

133

slip or ticket. In addition, the ftrst and last serial numbers in the
batch are written on the control ticket. The size of a batch should
always be kept to an easily handled number of records.
Figure 7.S is an example of fl batch control ticket. Notice
that space is provided to record the sending and receiving departments, batch number, date, and the ftrst and last serial numbers in
the batch. With this record, both the sending and receiving departments can account for all records.

A letter of transmittal describing a group or batch of documents is
frequently used to set up control and transfer responsibility when
documents move from one department or location to another.
The transmittal is usually a printed form with spaces to indicate
the variable information for the batch. Refer to Figure 7.6.

Transmittal and
Route Slips

CARD SHIPMENT TRANSMITTAL
TO
LOCATION ~~ IDEPT'
INDIVIDUAL

REPORT NAME

individual

rlJtrC.-

I?/,e

~d-,

REPORT CODE OO.3:;.IBOX NO.

441
/

BATCH NO.
DEPT. TO

CONTROL TOTALS
3000

LOCATION n

j?/,e

1'<5; 643,21

8j~.J"';"A
v

a)~
~~

_II.

£>.

FROM
,onIDEPT.

ct:t7U./

AIR
ROUTING SLIP

NUMBERED

/(Jjlt.

173?1

FROM
DATE
DATE FWD. INITIALS

/~~

t;s

173i5
TO

NO. OF DOCUMENTS
REMARKS

peL
v

/()jf

-rLNJ

/o~t

Cl..R

H- 1734'

.JdL ~
7

EXPLAIN ANY DIFFERENCES IN NO. OF DOCUMENTS FORWARDED AND

RETURN TO CONTROL CLERK

Figure 7.6 Transmittal and route slips.

When the volume of work or the number of people who may
perform any given operation is large, it may be desirable to ftx
responsibility and account for documents passed from each operation to the next as well as from one department to another. In

134

PROCEDURE CONTROL

this case, a route slip is used, either in addition to or in combination with the letter of transmittal. The route slip is similar to the·
batch control ticket. However, each department or operational
step which the accompanying documents pass through is identified
together with an indication of the processing time and the operator
or clerk responsible for each job. Responsibility is fixed and the
means to effect a degree of work control as well as document control has been incorporated into the same form.

Questions
and
Exercises

1. Explain how the planned results of procedures must sometimes be modified to meet requirements for accuracy and controL
2. What is an audit trail?
3. To what extent is the record count useful in e~tablishing procedure
control?
4. Show how a reGord or unit count can ,be combined with control totals to
provide effective control.
5. What is the essential difference between the modulus 10 and modulus 11
self-checking numbers?
6. List the purpose of each of the following: document register, serial number, transmittal and route slips.
7. Review your procedure for Use of the stock room requisition as proposed
in the exercise following chapter 6. Decide what controls, if any, should
be established over the origination of this form.
.
8. Present your controls for class discussion.

SORTING AND
CLASSIFICATION

Sorting is the process of ordering and arranging a set of records in some given sequence. In punched card data processing
procedures, the sequence is usually numerical; that is, arranged in
the same order in which we count. For example, in counting from
1 to 100, the count ascends from 1 (the lowest order and first
number in the series) to 100 (the highest order and last number).
Ascending numerical sequence is the most common filing
order for records that have numerical identification or control
fields, such as part number, social security number, and account or
customer number. Numerical sequence has an advantage over
other filing orders because it is most easily handled by punchedcard sorters in the least number of passes of cards through the machine. Numerical sequence also has historical preference with
punched-card equipment since early models of these machines
could process only numerical data; alphabetic or other special
characters could not be handled.
Reverse, or descending numerical sequence is sometimes desirable. In the series 1 through 100, 100 would then be the first
record in the set; 1, the last. Such a sequence might be useful
where records are filed by date. If day, month, and year are numerically represented (for example, July 4, 1970 as 07 04 70),
the first record in the file would be the one with the most recent
date. The last record would be the one with the oldest date. This
135

136

SORTING AND CLASSIFICATION

sequence has obvious advantages where records are filed in order
of their age.
Variations on the numerical sequence can also be used. Although more difficult to manipulate by machine, such sequences
can nevertheless be established where it is necessary to group records by categories or codes that may not correspond to a straight
ascending or descending sequence. One such requirement might
be the arrangement of records in special groups or account numbers where all records of the series 1,000,4,000; and 8,000 might
represent charges to raw material; all records of the series 2,000,
3,000, and 5,000 might represent charges to finished goods; and
all records of the series 6,000, 7,000, and 9,000 would be charges
to inventory expense.
Alphabetic sequences are constructed in the same manner as
numerical sequences. The basic order of a normal ascending series
begins with the letter A as the first or low-order character and ends
with the letter Z as the last or high-order character. Reverse alphabetic arrangements are possible though unusual. Arrangements in
special or variable sequences may be fairly common in name files
of various kinds but not in fields designed for machine processing.
Alphabetic sorts on punched-card sorters require more passes
for a given alphabetic field than for a numerical field with the
same number of characters. Hence the inherent advantage of simplified handling is an incentive to maintain numerical sequences
whenever possible.
Many record fields also contain the special characters of punctuation and report printing: commas, periods, ampersands, parentheses, number signs, and the like. These characters can also be
sorted into a predetermined sequence by both punched-card machines and computers. This is necessary because special characters
often appear in the same fields with both alphabetic and numerical
information. For example, nearly all address fields contain name,
street and number, city, and state: John Doe, 45 Pleasant Lane,
County Center, California. The ascending sequence of special
characters with relation to digits is set as a standard convention by
manufacturers throughout the office equipment industry.
Sets of records can also be arranged in sequence by one or
more fields. For example, a file of insurance records could be arranged by policy number as the first or minor sequence in the set.
The records could then be sorted by issuing office number as the
next or intermediate sequence, then by class of insurance as the

SORTING AND CLASSIFICATION

137

last or major sequence. The resulting order of the file is the grouping of records by class of insurance, by office number within each
class, and by policy number issued by each office.
Punched-card accounting machines can be controlled to recognize at least three orders or sequences within a given set of records.
Subtotals of amount fields can be accumulated and printed within each minor, intermediate, or major group of records as the cards
are fed to the machine.

Figure S.l shows one type of IBM sorting machine designed to
operate with SO-column cards. An explanation of how this machine works will illustrate the basic principles of operation for
many other types of punched-card equipment.
The file or deck of punched cards to be sorted is placed in a
hopper at the right end of the machine, facedown, 9-edge toward
the machine. Two types of hoppers are available with IBM sorters:
the standard vertical hopper and a file-feeding device (Figure S.2).
The standard hopper holds about 1,200 cards, the file-feeding device about 3,600 cards.
Cards are placed in the machine by an operator, usually from
a card file or stacking tray. Depending upon whether or not the

Figure 8.1 IBM 84 Sorter.

Machine Operation

138

SORTING AND CLASSIFICATION

sequence of the cards before sorting is to be maintained, the operator normally fills the sorter from the front of the file or tray.

Figure 8.2 IBM file feeding device.

When the machine is started, cards are mechanically fed into
the transport mechanism, one at a time, from the bottom of the
deck. Cards move horizontally across the machine, from the operator's right to left, into stackers or pockets. Each card passes between a metal roller and a small wire brush or between a small
light bulb and a light-sensitive diode, depending upon the machine
model. The metal bristles of the brush are shaped to fit exactly
through the rectangUlar holes in one column of the card. The
sorter reads only one card column at a time.
The brush (or the light-sensing device) is movable and can be
placed over any of the 80 columns by the operator. When there is
a hole in the column being sorted, the brush drops through the

SORTING AND CLASSIFICATION

139

card and touches the metal roller for an instant (Figure 8.3). (With
the photoelectric sensing method, light shines through the hole
onto the diode.) The metal contact (or the light sensing) creates
an electrical impulse which controls the mechanism of the machine.
Furthermore, the impulse occurs at a time that exactly corresponds
to the position of the hole with relation to the feeding cycle.
Thus, if a hole is sensed in the 4 row, that impulse will occur at
"4 time" in the feeding cycle and is therefore read by the machine

Armature Plate

A B C -

Metal Brush
Contact Roll
80-column Card

Figure S.3 4-punch in SO-column card entering sorter selection device.

as a 4 (Figure 8.4). When a hole is sensed in the I row, the card is
at the 1 position in the cycle. The machine consequently reads
Chute Blades

-Carrier Roll

Roll
Shaft

A
B
C
D

Metal Brush
Contact Roll
80-column Card
Energized Magnet Coils

Figure S.4 4-punch sensing directs card to 4 pocket.

that hole correctly as a 1. The card is then automatically transported to a pocket or stacker by a chute blade that corresponds to
the position of the punched hole in that column.
There are 12 chute blades or bands of metal, varying in
length, that guide the cards from the sort brush to -the pockets

140

SORTING AND CLASSIFICATION

(Figures 8.3 and 8.4). Each chute blade extends from a point
immediately past the sort brush to the opening above one of the
numbered pockets.
Since there are 12 possible punching positions in anyone
column, there are also 12 pockets in the sorter plus one pocket
for cards that have no holes punched in the column being sorted.
For example, if there is a hole in the 3 position of the column, that
particular card is directed to the 3 pocket. If there is a hole in the
1 position, the card goes into the 1 pocket. If there is no hole, the
card is directed to the pocket reserved for blanks, the reject pocket.
. For 80-column cards, IBM offers sorters in three models,
each operating at different speeds:

Sorting for
Printed Reports

Sorter

Cards per
minute

Model 82
Model 83
Model 84

650
1,000
2,000

Printed output is often the end result of a machine accounting
proced ure. Invariably, the information shown is arranged in some
organized form and in a required sequence. In many cases, information is ordered in a certain sequence for one report, then
reordered in another sequence for a second report. The same
card records may be used for a number of reports. I t is only
necessary to sort them properly beforehand.
For example, assume that a business requires reports to assist
management in guiding sales policy. One file of punched-card sales
transactions can furnish all the necessary information when arranged as follows:

1. Product Number. All sales of each product appear together
in product number sequence (Figure 8.5). The report shows quantity, unit price, and total sales value for each item. If other descriptive information is available in the cards, sales trends can also be
shown by color, style, model, size, and so on. To furnish these
analyses, the cards are sorted by the corresponding identification
or descriptive field.
2. Customer Number. On this report (Figure 8.6) all sales
for a period are tabulated by individual customer. Additional in-

SORTING AND CLASSIFICATION

Product sales activity

Selling
price

Prod.
no.

Quan.

Unit
price

1000

125

5.00

625.00

1001

4.00

300.00

1002

130

1003
1004

3.00
2.25

390.00
112.50

65
100

2.50
1.20

162.50

1005
1006

118

2.50

120.00
295.00

1007

5.50

231.00

1008

1.65

1986

430

2.25

85.25

---

......

..-

967.50
125680.25*

Figure 8.S Sorting for printed reports-product number.

formation could include a sales total for some preceding period as
an indication of buying trends. For example, sales this year or this
month could be compared with a corresponding previous period.

Sales by customer

Cust.
no.

Gross
sales

Disc.
amt.

Net
sales

155.50
483.65
529.60

6.22
24.18
21.18
80.63

149.28
459.47

24691
24693
24694
24695
24703
24705
25132
25134

806.30
98.75
102.80
620.00
240.00

26001

95.00
125680.25*

62.00
9.60

508.42
725.67
98.75
102.80
558.00
230.40
95.00

7540.15*

118140.10*

Figure 8.6 Sorting for printed reports-customer number.

141

142

SORTING AND CLASSIFICATION

The credit status of a customer's account could also be shown by
printing out sales, payments, and balance due.

3. Salesman Number. On this report, sales are shown sequenced by salesman number (Figure 8.7). Total sales and resulting commissions are tabulated to show both selling volume and
compensation. The report could be the basis of actual commission
payments to all salesmen in the company.
Salesmen's activity
Comm.
amt.

Sales.
no.

Gross
sales

3175.00
2840.75

287.50

2
3

1690.00

169.00

4
5

2585.90

195.65

550.00

55.00

6
7

5640.50

564.05

4123.00

376.75

780.50

78.05

258.50

2475.00
125680.25*

Figure 8.7 Sorting for printed reports-salesman number.

These examples have been simplified, but in actual practice,
the amount of information that can be printed is limited only by
the data punched in the sales transaction cards.
Notice that because all of these reports were produced from
the same deck of cards, the gross sales amount totals agree. This
is an example of proper procedure control using an amount total
to check the accuracy of each printed report.

Numerical Sorting

To arrange cards in numerical order, the sorter brush (or sensing
head) is set on the right-hand or units column of a numerical field.
As cards feed through the machine, they are separated into the
various pockets corresponding to numerical punching in that column, as previously explained. The pockets or stackers are arranged in ascending sequence from the operator's right to left as
shown schematically in Figure 8.8.

SORTING AND CLASSIFICATION

143

Hopper ....
Card path

-1--Sorter pockets

Figure 8.8 80-column sorter pocket arrangement.

The column of numbers in Figure 8.9a represents a small deck
of cards that are in random order by a two-digit field. Figure 8.9b
02

(c)

(e)

Hopper~

(b)

Hopper~

(d)

66
(a)

Figure 8.9 (a) Random order, two-digit field. (b) First sort on two-digit field. (c)
Sequence after first sort. (d) Second sort on two-digit field. (e) Sequence after sort.

144

SORTING AND CLASSIFICATION

shows the arrangement of the same deck of cards in the respective
soiter pockets after the first pass through the machine. Cards are
directed to the pockets corresponding to the digits punched in
the units column.
In actual sorting, the operator now removes the cards, maintaining the sequence obtained on the first sort (Figure 8.9c). Figure
8.9d shows the cards in the pockets after the second pass through
the· machine. If the cards are removed properly, they will be in
sequence as shown by the column of numbers in Figure 8.ge. For
larger fields, the same procedure is followed for the third and subsequent columns. After each column is sorted, the cards are in
order according to the columns already sorted. Usually, numerical
fields have no unpunched columns. Insignificant zeros are punched
in columns that precede significant digits. In a strictly numerical
sort, no cards should fall into the 11, 12, or reject pockets.

Card Handling

The card sorter, like all other punched-card equipment, is controlled by an operator. And, while the speed of the machine is
constant, an average of 25% of total elapsed sorting time normally
is required for manual card handling. This includes the movements
of removing cards from the pockets or stackers as they are filled,
emptying the machine between sorts, checking the accuracy of the
sort, loading and unloading file trays, and so on. Good operating
procedures will therefore increase efficiency, consequently shortening the elapsed time required to complete any given sort.
By referring to the photograph of the IBM 84 Sorter in Figure
8.1, it can be seen that a flat glass shelf extends over the stackers.
This same arrangement is found on all models of IBM sorters. The
shelf can be used as a convenient surface on which to ready cards
for the file feed or standard card hopper. While a machine is in
operation, pockets or stackers can fill very rapidly. Also, cards are
hardly ever distributed evenly in the various pockets during a sort.
Overflow cards from one or more pockets can be piled on the glass
shelf directly above the pocket from which they are removed.
When the sort is finished and the machine emptied, the operator
knows the exact position in the sequence of each pile of cards on
the shelf. Cards are manually stacked facedown. For each subsequent sort, cards are fed into the hopper starting with the card at
the front or bottom of each stack. This preserves any previously
sorted sequence.

SORTING AND CLASSIFICATION

Special card trays and dollies can be obtained from a number
of manufacturers. The equipment is designed with numbered receptacles or stacking trays to hold cards as they are removed from
the machine. Depending upon the volume involved in anyone
sort-a handful or a number of files-the operator must use some
system that will enable him to accurately return cards to the machine in sequence after each sort.
The hopper of the various model sorters is slightly larger inside than the outside dimensions of an 80-column card. Stacks of
cards must be joggled or aligned carefully so that no protruding
edges will jam the deck in the hopper, thereby preventing free access to the feeding knives. Cards are held down in the hopper with
a removable card weight.
The throat of each machine is designed with tolerances that
will admit one card at a time, but not two. Edges of cards must
not be damaged or else feeding will be troublesome, with the possibility of jamming in the transport mechanism or chute blades.
Folding cards in the center will thicken the edge exactly at the
throat guide with the result that these cards are very difficult to
sort accurately.
The well publicized motto "Do not fold, spindle, or mutilate!" is not exactly a joke to the card machine operator who takes
every precaution to keep punched-card records in the best condition possible.
An operator quickly acquires the simple skill of card joggling
and fanning. To joggle the cards, a convenient size deck is held
loosely in one hand while the ends are struck lightly against some
solid surface with the other hand. Every card machine except the
sorter is equipped with a joggle plate for this purpose. The effect
is to align all four edges of the deck as evenly as possible.
\

",\

Figure 8.10 Sight checking.

145

146

SORTING AND CLASSIFICATION

Fanning or riffling the cards is also good operating practice.
This is done by holding a small deck by the end in one hand and
fanning across the opposite end with the other hand in a brushing
motion. Fanning removes static electricity that may form in dry
atmospheric conditions and cause cards to stick together. Fanning
also removes dust and other foreign particles that sometimes accumulate in a file drawer, particularly when the file is used for
reference. Fanning and joggling also show at a glance that all cards
face in one direction with corner cuts aligned properly.

Sight Checking

The accuracy of every sort should be checked after removing cards
from each pocket. First, the cards from any pocket are joggled
into perfect alignment. Then the operator looks through the hole
corresponding to the pocket from which the cards were removed.
If the cards have been correctly sorted, the holes form a small passage in the deck through which light can be seen (Figure 8.10).
If no light is visible, a missorted card must block the passage
because, in that card, a hole is punched in some other position.
For example, if the cards come from column 3, all cards in that
pocket must have a 3 punched in the column being sorted. If a 2
missorts into the 3 pocket, that card would cut off the sighting of
light through the three hole in the deck.
A missorted card can be located easily by using a thin metal
rod called a sorting needle which is small enough to pass through
the rectangular holes. The needle is pushed through the passage in
the deck (for example, through the 3 hole) until it is blocked by a
missorted card. The deck is opened at that point and the card removed. The error card can be manually filed in its proper place if
it is not damaged in any way. The operator must also file the missorted card in the current sequence resulting from all previous sorts.

Estimating
Sorting Time

Whenever a punched card procedure is designed, consideration
must be given to the total elapsed time required to complete
an entire procedure cycle. The time required for sorting, usually a
significant portion of the estimated time, is always included.
Actual machine time for any given sort can be calculated by
multiplying the number of cards in the file times the number of
columns to be sorted, then dividing the product by the running

SORTING AND CLASSIFICATION

147

speed of the machine (cards per minute). Thus, to sort 20,000
cards on five columns with a 1,000 card-per-minute sorter takes
100 minutes.

20,000 cards X 5 columns
100 mmu
. t es
.
. =
1, 000 cards per mmute
or 1 hour and 40 minutes
A factor of 25% handling time is usually added to the total
machine time. For a novice operator this time may be somewhat
longer; for an experienced operator, considerably shorter. Also
note that the 25% handling factor is. higher for sorting than for
most other card processing operations. This is because the stacks
of cards from each sorter pocket must be manually removed before
the machine can be started on the next pass. Cards must be joggled
and fanned manually between each pass and, after each pass, all
cards should be sight-checked. Adding the handling factor to the
previous example brings the total time to 125 minutes, or two
hours and five minutes.

Block sorting serves two purposes. First, with this technique a
large file can be broken down into smaller groups of cards so that
several blocks can be sorted at the same time on different machines. Thus, overall elapsed time can be reduced even though
the total time required may be slightly more due to increased
handling. Second, the next machine operation in the procedure
can be started without waiting for an entire file to be sorted.
For example, assume that 20,000 sales transaction cards are
to be sorted by customer, a five-digit number. Also assume that
the number is punched in columns 21 through 25. To block sort
the file, proceed as follows:
1. Sort the entire file on column 21. If there is no sequence
to the file that must be preserved, this sort can be done on more
than one machine. If there is approximately equal distribution
of cards in all pockets, there will be 10 blocks of 2,000 cards
each: 2,000 in the 0 pocket, 2,000 in the 1 pocket, 2,000 in the
3 pocket, etc.
2. Take each block of cards and sort in the usual way, from
column 25 back to column 22. Two or more blocks can be sorted
simultaneously on different machines.

Block Sorting

148

SORTING AND CLASSIFICATION

If it takes two hours and five minutes to get the 20,000 cards
into customer number sequence using regular sorting techniques,
this means that the next machine operation cannot be started
until all cards have been sorted as shown in Figure 8.11. If report
preparation requires three hours and 20 minutes, the total elapsed
time from the start of the sorting operation and the completion of
the report would be almost five and one-half hours.

Regular
sorting

Sorting time
2 hours 5 min.

I
il

Sorting time
2 hours 5 min.

Block
sorting

I
o

Report printing time
3 hours 20 minutes

Report printing time
3 hours 20 minutes
I

Hours

Figure 8.11 Serial sorting vs. block sorting.

Using block sorting methods, the amount of time required to
get block a in sequence may be calculated as follows:

First sort
(column 21)
Second sort,
ablock
( columns 25-22)

20,000 cards X 1 column
1, 000 cards per minute
2,000 cards X 4 columns
1,000 cards per minute

20 minutes

8 minutes

Subtotal

28 minutes

25% handling time

7 minutes

Total

35 minutes

Thus, report printing can begin approximately 35 minutes

SORTING AND CLASSIFICATION

149

after the beginning of the sort. Printing can continue while the
balance of the file is being sorted. With this method, the overall
elapsed time is reduced by about one and one-half hours as shown
in Figure 8.11.

Letters of the alphabet are represented in 80-column cards by
both numerical and zone punching in a single column. Therefore,
to sort alphabetically, two sorts are required on each column. The
first sort is numerical, that is, the column is sorted on numerical
punches only. On the second pass, the sort is made on the 0, 11,
and 12 punches only.
In normal sorting procedures, all cards are fed to the sorter
with the 9 edge toward the throat. By referring to Figures 8.3 and
8.4 it can be seen that when a card passes under the sensing brush
(or the diode), the lowest hole in the column is read first. A 9 hole
is read before an 8, a 6 hole before a 0, and so on. The first electrical impluse positions the chute blades to direct the card to the corresponding pocket, regardless of how many other holes may be
punched in the same column. Any higher punching in the column
is ignored.
Therefore, when sorting alphabetically, the first pass sorts the
cards into numerical sequence by the digit punches. The zone
punches are ignored. On the second pass, the sort is made on the
0, 11, and 12 punches in the same column. This is done by disconnecting the digit circuits of the machine. An alphabetic sorting
switch is provided for this purpose.
After the second sort, the 12 pocket contains the letters A
through I, the 11 pocket contains the letters J through R, and
the pocket contains the letters S through Z. In alphabetic sorting, as in numerical sorting, the right hand column of the field is
sorted first. Successive sorts are made from right to left with the
last sort on the leftmost column of the field.
The column of letters in Figure 8.12 represents a small deck
of cards in random sequence. The deck is to be sorted alphabetically by a two-column state abbreviation code. Four sorts are required. Each letter in the column is shown with its corresponding
punched-hole code. Assume that the field is punched in columns
26 and 27.

Alphabetic Sorting

Card
. code

Card
code

Abbr.

0-3

11-5

0-8

11-5

11-1

0-7

0-5

0-3

12-6

F L

11-3

0-2

S C

12-3

11-5

N H

12-8

0-5

12-1

0-6

11-8

0-2

S D

12-4

11-5

N D

12-4

11-5

12-3

12-7

12-1

11-3

12-1

11-4

12-5

11-4

ME
MO

11-9

12-9

11-7

P A

12-1

11-6

Figure B.12 Random order, two-character field.

The first sort is made on the numerical portion of column 27.
Cards fall into the respective pockets as shown in Figure 8.13a.
The second pass is made on the zone portion of column 27. The
same sorting procedure is repeated for column 26. The resulting
distribution of cards is also shown in Figure 8.13b. The final sequence is represented by the column of abbreviations shown in
Figure 8.14.

NY
NH
WY

SD
ND

VT
FL
SC
NC

NJ
VA
GA
LA
PA

Column 27
First pass
(numerical)

12
NH

(a)

NY
WY
TX
VT

FL
NJ

SD
ND
SC
NC
VA
GA
LA
PA

Second pass
(zone)

SORTING AND CLASSIFICATION

GA
PA

WY
FL

NY
VT
NJ
NH
ND
NC
VA

Column 26
First pass
MO
ME

TX
LA

SD
SC

(numerical)

WY
VT
VA
TX
SD
SC

(b)

PA
NY
NJ
NH
ND
NC
MO
ME

Second pass
(zone)

GA
FL

Figure 8.13 First (a) and second (b) sorts on two-character field.

WY
VT

TX
SO
SC
RI
PA
NY
NJ

NO
NC

MO
ME
LA

GA
FL
Figure 8.14 Sequence after second alphabetic sort.

Alphabetic fields usually have blank spaces between words.
These unpunched columns cause the cards to fall into the reject
pocket on the first sort. The pocket need not be sorted again.
Rejected cards are simply placed at the front of the file after the
second sort and before proceeding to the next column.

151

152

SORTING AND CLASSIFICATION

Alphabetic fields can also be block sorted. One method is to
sort the leftmost column of the field on the zone portion only.
This sam divides the file into three sections by the 0, 11, and 12
punches. Each block can then be sorted separately.

Selection Switches

In addition to the alphabetic sorting switch, each type IBM sorter
is equipped with twelve single selection switches. There is one
switch (or button, depending upon the machine model) for each
of the 12 punching rows of an 80-column card. When turned off,
a switch suppresses sorting of the corresponding rows. The sorter
then rejects all cards punched in that position of a column.
More than one switch can be turned off at a time. It is therefore possible to sort cards punched with more than one digit or
zone code into the reject pocket. For example, if the 2 and 3
switches are turned off when sorting numerically on any column,
all cards punched with a 2 or 3 will sort into the reject pocket.
Cards punched with the digits 0, 1, and 4 through 9 will sort
normally. The effect is to select all 2s and 3s from the file into
one pocket.
Conversely, if all selection switches are turned off except any
one single switch, the cards with punching in that row will sort to
the corresponding pocket while all other cards are rejected. For
example, assume that all switches are turned off except the three
switch. When cards are sorted on any column, all cards will reject
except those punched with the digit 3. Those cards will be selected
into the 3 pocket. Note that the sequence of all the cards has not
been disturbed; they have merely been divided into two separate
groups or files.
Selection has many practical uses in punched-card procedures.
It provides a means of high-speed access to records with a method
of readily classifying information into various predetermined categories. Consider the following three examples.
1. Sales transaction cards may be sele~ted from a file for
those customers who have made single purchases of any item of
$1 ,000 or more. If the amount field is in columns 75 through 80
(four digits for dollars and two digits for cents), then selecting Os
in column 75 will remove from the file all cards with an amount
less than $1,000. Cards in either or both files can be printed by
an accounting machine for further analysis.

SORTING AND CLASSIFICATION

153

2 .. Personnel and statistical punched cards are particularly
well adapted to selection techniques, especially when coded by
various categories. Assume that records are coded 1 for salaried
employees, 2 for hourly employees, 3 for commission employees,
4 for salary plus commission, and so on. Separation of employee
records by wage class can be done automatically at speeds up to
2,000 cards per minute. Once selected, reports can be printed
for analysis.
3. As a result of calculation of amount fields, credit or minus
balances often result in accounting procedures. For example, if a
customer overpays a bill, the amount due field in his accounts
receivable record no longer represents an amount due the creditor,
but a refund due the customer. Such minus fields are usually
punched with an 11 or X punch over the units digit of the field.
By selecting X-punched fields, all refund amounts can be separated
from the accounts receivable file for special handling.

Using selection switches, the following shortcut method can save Alphabetic Shortcut
Method
approximately 16% of alphabetic sorting time.
For the first sort on each column, the cards are placed in the
hopper faceup, 12-edge first, with the 9 selection switch off. Because the cards are entering the sorter 12-edge first, a 12 zone
punch is read as a 9, an 11 punch as an 8, and so on. But, because
the 9 selection switch is off, the 12 zone punch is not read. Therefore, all letters with a 12 zone punch sort on numerical punches,
or wliat appear to be numerical punches to the machine. Cards
punched with the letter A (Punched 12-1) fall into the 6 pocket; the
Bs (12-2) fall into the 5 pocket; the Cs (12-3) into the 4 pocket,
and so on up to the Is (12-9) which fall into the 12 pocket. Cards
with 11 and 0 zone punches sort into pockets 8 and 7 respectively.
All cards punched with the letters A through I are completely sorted on this first sort. These cards can now be Temoved from
pockets 6 through 12 and stacked.
The 9 selection switch is now turned on again. The cards remaining in pockets 7 and 8 are sorted separately in the normal
manner, facedown, 9-edge tow.ard the machine. At the conclusion
of this sort, the previously stacked cards punched with the letters
A through I are placed at the front of the deck. The process is repeated for each succeeding column of the sort until the cards are
completely arranged in alphabetical sequence.

IS4

SORTING AND CLASSIFICATION

Sorters can be equipped with a number of special devices to make
the sorting operation more efficient. For example, the IBM 83
Sorter is provided with a five-position sort selection switch. The
setting of the switch determines the sorting pattern to be followed
by the machine, as shown in Figure 8.15. The conventional sorting patterns for either alphabetic or numerical are available with
N or Z settings. In addition, three other variations on alphabetic
sorts can be used by the operator.

Special Sorting
Devices

POCKETS
SORT
SELECTION
SWITCH
SETTING
Numerical
( N)

REJECTS
REGARDLESS
OF EDIT

ERRORS (When
Edit or Edit~Stop
isON)

Blanks

Multiple-punched
cords {incl. letters}

Any cord without
a zone punch

Any cord with more
than one zone punch

0
S-Z

11
J-R

Blanks and cords
with a 12-zone
punch but no digit
punch. Digits 1
to 9.

Any cord with more
than one zone punch
or with more than
one digit punch

Cords with 0 or
II-zone only.
Blanks. Letters A
to I, and 12-zone
spec. char. Digits
1 to 9.

Some as A-I

Zone
(Z)
Alpha-l

R,Z

a,v

(A-I)

Alpha-2

P,X O,W N,V M,U L, T

K,S

J
0-1

(A-2)

AlphaNumericol
( A-N)

0
11
digit J-R

12
A-I

Blanks,
(S-Z)

~zone

Same as A-I

This pattern is based on cards being fed face down, 9 edge first.

Figure8.1S Sorting pattern for IBM 83 Sorter.

Any deviation from a specified pattern can be automatically
detected by setting an edit switch. For example, when sorting numerically, any multiple-punched cards in the column being sorted
are determined to be errors. Also, when sorting by zones, any card
with more than one zone punch is treated as an error. Error cards
are rejected, that is, they fall into the reject pocket of the machine.
By setting an edit-stop switch, whenever error cards reject an
error light comes on and card feeding stops. The operator can
then remove the error card from the file.
The 83 and 84 Sorters can also be equipped with an alphabetic sorting feature. With this feature installed, the machine
operates in one of the three possible sorting patterns shown in
Figure 8.16.

SORTING AND CLASSIFICATION

POCKETS

155

SELECTioN
SWITCH
SETTING

REJECTS
REGARDLESS
Of EDIT

ERRORS (When
Edit or Edit-S.top
isON)

A-I

KN
OT
WZ

BD
FH
JM
PS
VY

Cords punched
with digits only,
zones only, 0-1
combination, or
blank

Any cord with more
than one zone punch
or more than one
digit punch

N
M
L

H
G

F
E

0
C

B
A

Some as A-I

Same as A-I

Blanks.
A,C,E,G,I,
L,O,R,U,X
and the combination 0-1.

Same as A-I

SORT

A- 2

A-N

0
P
0

J
I

KN
OT
WZ
11

BD
FH
JM
PS
VY
12

This pattern is based on cards being fed face down, 9 edge first.

Figure 8.16 Sorting pattern for alphabetic sorting feature.

To sort a column alphabetically, the selection switch is first
set to A-I. Cards punched A, C, E, G, I, L, 0, R, U, and X fall into
pockets 0 through 9. The letters B, D, F, H, J, M, P, S, V, and Y
fall into the 12 pocket. Letters K, N, Q, T, W, and Z fall into the
11 pocket. Blanks and cards without letter coding fall into the
reject pocket.
The selection switch is now set to A-2. It is not necessary to
remove cards from pockets 0 through 9. Cards from pocket 12 are
now sorted, followed by the cards from pocket 11. Cards fall into
their respective pockets and the column is completely sorted.
Occasionally, numerical and alphabetic characters are intermixed in the same field; for example, street and number in address
cards. These fields may be sorted in an alphanumerical pattern.
Digits 0 through 9 fall into their respective pockets. The letters
K, N, Q, T, W, and Z fall into pocket 11. Letters B, D, F, H, J, M,
P, S, V, and Y fall into pocket 12. Cards with only an 11 punch
fall into the 11 pocket, and 12 zone cards fall into the 12 pocket.
All other cards including the letters A, C, E, G, I, L, 0, R, U, X,
and the 0-1 combinations fall into the reject pocket.

The operation of the sorter can be further enhanced by installing
the multiple-column selector feature. Instead of using only one

Multiple-Column
Selector (I BM 82, 83)

156

SORTING AND CLASSIFICATION

brush to read a single card column, the feature uses 10 brushes
that may be set to read any 10 or fewer adjacent card columns.
There is also a small control panel on the sorter, 10 column control keys, and control switches. The function and selection to be
performed by the device is set up by the operator for the particular procedure involved.

5,000 rejects - 5th sort

All cards sort on the first pass

Total number of cards - 20,000

Speed - 1,000 cards/minute

Zero elimination method
Sort
1

Volume
20,000 cards
20,000 cards
18,000 cards
14,000 cards
9,000 cards
4,000 cards

2
3
4
5
6

Time
20 minutes
20 minutes
18 minutes
14 minutes
9 minutes
4 minutes
85minutes

25% handling time

20 minutes

Total

105 minutes

Regular method
20000 cards X 6 columns
1,000 cards per minute

120 minutes

25% handling time

..1Q. minutes

Total time

150 minutes

Figure 8.17 Zero elimination method of numerical sorting.

SORTING AND CLASSIFICATION

157

Multiple-column selection selects into a specified pocket those
cards punched with a predetermined numerical or alphabetic code
in 10 or fewer consecutive columns. For example, all cards with
city names like Rochester, Syracuse, or Albany can be selected
from a given file. The sequence of the remaining cards is not
disturbed.
In a marketing situation for a business, the feature could be
useful in identifying customers in given locations from a punchedcard name and address file. For example, all customers in Rochester could be listed by an accounting machine in planning a sales
campaign in that city.
Other variations of multiple-column selection are possible with
this feature. Common-digit selection sorts out all cards with one or
more common digit punches in ten or fewer consecutive columns.
Zero elimination shortens the time required for sorting numerically by rejecting those cards that require no further sorting
on insignificant zero to the left of the field. For example, while
sorting on a six-digit part number field, number 000065 would reject on the third sort. A card punched 000125 would reject on the
fourth sort. Rejected cards are placed in the front of the file, first
rejects first, second rejects second, and so on. A schematic of this
method is shown in Figure 8.17. In the illustration, it is assumed
that the number of cards given will reject during each of the six
sorts. The estimated time for each sort is also shown and the total
time is compared to the time required to sort the file without the
zero elimination feature. A 25% allowance for handling time is
added for both methods. Use of the feature saves approximately
45 minutes elapsed time on one machine.
A length of field selection can also be made. Cards are distributed by pocket as determined by the last significant column
punched in the field, regardless of the remaining spaces to the left.
When sorting alphabetically, short names like Rye and Olean that
require fewer sorting passes can be separated from longer names
like Kalamazoo and Albuquerque. Selection can be done on one
pass through the machine.

The IBM 5486 Card Sorter is designed to sort 96-column cards.
As a compact, tabletop machine, the 5486 can be operated from a
seated position. The card sorter is an auxiliary unit of the IBM
System/3 small computer.

IBM 5486

Card Sorter

158

SORTING AND CLASSIFICATION

The 5486 operates in much the same manner as the 80-column machines described in the previous section. However, the
sorting patterns and procedures are considerably different. The
basic features of the machine are shown in Figure 8.18.

Figure 8.18 IBM 5486 card sorter.

The card hopper is located at the upper right corner and has a
capacity of approximately 2,000 cards. The cards are placed facedown in the hopper, top edge toward the machine. Cards are fed
from the hopper, one at a time, from the bottom (first card) of
the deck. Two model machines are available from IBM that operate at speeds of 1,000 and 1,500 cards per minute.
The 96-column cards move from the hopper to a read station
where a beam of light shines through the holes punched in the
card and strikes an electronic device. The pattern of holes is sensed

SORTING AND CLASSIFICATION

159

to direct the card to one of six pockets. Pockets are labeled 9/8,
7/6, 5/4,3/2, 1/0, and R (reject). Each pocket holds about 600
cards. The sorter stops automatically when a pocket is full.
A sorting tray is provided at the top of the machine. When a
pocket is full, cards are removed by the operator and stacked in
the trays. There is one tray for each of the pockets.

Numerical sorting begins with the right-hand, or units digit of the
field. With 96-column cards, two selections must be made to locate
the proper column: the tier and the column. A card column selector knob is furnished for this purpose on an operator's panel.

Numerical Sorting:
96-Column Cards

___- UDlunln Select Indicator Window

III

1..-_-- Card Column Selector Knob
(Push knob left and turn it to
select tier; release and turn knob
to select column)

Digit Select Keys
Card Counter

Phase L i a t l t s - - + - - - - - - - - -......

(CNTR)
ON

Phase b a r - - . + - - - - - - J

;;;:,

STK
FULL

READ
CHECK

HOPPER
CHECK

STOP

888BB
NUM

FEED
CHECK 1

NUM
ZONE-

ED
-

ALPHA
-FEAT
-SORT
SUP

Condition Indicators (back-lighted panel)
Mode Switch

Figure 8.19 5486 operator's panel.

160

SORTING AND CLASSIFICATION

A schematic of the panel is shown in Figure 8.19. The number of
the column selected shows in a small indicator window.
Because there are only five numerical pockets in the machine,
numerical sorting is done in two phases. In phase I, all even-numbered digits sort and all odd-numbered digits reject. Thus, Os fall
into the 1/0 pocket, Is into the reject pocket, 2s into the 3/2
pocket, the 3s reject, and so on. Sorted cards are left in the pockets at the end of phase 1.
A lighted phase bar is provided to change the machine to
phase 2 for the second part of the numerical sort. In phase 2, the
cards from the reject pocket are placed in the hopper. These cards
now contain only odd-numbered digits which sort into their respective pockets on top of the even-numbered digits from phase 1.
Digit 1s sort into the 1/0 pocket, 3s into the 3/2 pocket, and so on.
The completed pattern is shown in Figure 8.20.
Phases of Sort

Pockets
Reject

9/8

7/6

5/4

3/2

1/0

After Phase 2 Sort

Blanks and Special Characters

After Phase 1 Sort

All Odd-Numbered Digits, Blanks,
and Special Characters

Figure 8.20 5486 numerical sorting pattern.

At the conclusion of Phase 2, cards are in numerical sequence
by the selected column. The procedure is repeated, two sorts on
each column, until the leftmost column of the field has been
sorted. As a sort proceeds, the operator must be sure that cards
are taken from the pockets or tray in sequence. The bottom card
of the stack is always first into the hopper.
During numerical sorting, the mode switch on the operator's
panel is turned to numerical.

Alphabetic Sorting:
96-Column Cards

To sort 96-column cards alphabetically, the numerical portion of
the punched hole code is sorted first, as just described. After
phase 2 of the numerical sort, the mode switch is set to zone. A
third pass is made on the same column.
Cards from the 1/0 pocket go into the hopper first, cards
from the 3/2 pocket next, then the 5/4 pocket, and so on. Cards
left in the reject pocket after phase 2 are not sorted on this pass

SORTING AND CLASSIFICATION

161

because the column is blank. After the zone sort, cards are distributed in the following order: the letters A through I are in
pocket 3/2, J through R are in pocket 5/4, S through Z in pocket
7/6, and all digits are in pocket 9/8. Rejected cards are placed in
front of the deck before proceeding to the next left column of the
field. Notice that this method sorts the field correctly even though
there may be interspersed alphabetic and numerical data. This alphameric sort pattern is shown in Figure 8.21.
Phases of Sort

Pockets
9/8

7/6

5/4

3/2

A fter Zone Sort

0-9

S-Z

}, J-R

A-I

After Phase 2 Sort

I,R,Z

G,P,X

9
A fter Phase 1 Sort

H,Q,Y

7
F,O,W

E,N,V

C,L,T

5
D,M,U

3
B,K,S
2

1/0

Reject
Blanks and Special Characters

A,J
1

Blanks and Special Characters

}
0

1,3,5,7 ,9,A,J,C,L,T,
E,N,V,G,P,X,I,R,Z,
Blanks and Special Characters

Note: Right Brace (}) is sorted as indicated ,not as a special character.

Figure 8.21 5486 alphabetic sorting pattern.

When only alphabetic information is to be sorted, the 5486
Card Sorter can be equipped with an alphabetic sort feature. With
this feature, cards can be arranged in alphabetic sequence in two
passes through the machine on each column, instead of three. The
Pockets

Phases of Sort
9/8

After Phase 2 Sort

Y
X

116

5/4

3/2

W
V
U
T
S
R

P
0
N
M
L

K
J
I
H
G

V
P
J
0

U
0
I
C

A fter Phase 1 Sort

W
Q

K
E

Z
T
N
H
B

Figure 8.22 5486 alphabetic sorting pattern-special feature.

1/0

Reject

Numerics, Special Characters,
and Blanks

0
C

B
A

Y
S
M

G
A

R
L
F

Numerics, Special Characters,
and Blanks

162

SORTING AND CLASSIFICATION

feature is activated by setting the mode switch on the operator's
panel to the alpha position.
Sorting begins on the rightmost column of the field. The
tier and column to be sorted are selected by the operator with the
column selector knob. The phase bar is pressed to begin the phase
I sort. The resulting pattern of pocket distribution is shown in
Figure 8.22. After phase I, the cards are removed from the pockets and replaced in the hopper: pocket R first, pocket I/O on top
of pocket R, and so on. After phase 2 is completed, cards are in
alphabetic sequence by the selected column. - Note that digits,
special characters, and blanks are in the reject pocket. Numerical
punching is not intended to be sorted with this feature.
The following example shows the actual procedure for sorting
on a two-position alphabetic field. The column of letters in Figure 8.23 is intended to represent a small deck of 96-column cards
to be sorted alphabetically by state abbreviation code. The corresponding card code is shown beside each letter. The abbreviations
Card
code
AB
B

State
abbr.
21

A-

A-8

1.:1

NY
NJ

A-

1.:1

Card
code

A-

BA-

142

A-

SC
NH

BA -

BA-

WY
SD
ND

4 1

A-

Ll2

A-

BA-

A- 8
BA-

BA-

Ll21

BA-

ME
MO

BA-

BA- 8

1
421

1
4

BA- 8

BA-

Figure 8.23 Random order. Two-character field with 96-column card codes.

are repeated from the previous example in Figure 8.12. Assume
the field is punched in columns 26 and 27.

SORTING AND CLASSIFICATION

163

Pockets
Column 27

9/8

7/6

5/4

3/2

1/0

Rej.

VT
NH

NY
VA
WY
GA
LA
PA

TX
FL

SO
NO
SC
NC
VA
GA
LA
PA
1/0

Phase 1

NJ
SO
NO

SC
NC
RI
MO

Phase 2

Column 26

NY
WY
TX

NJ
RI
NH

9/8

7/6

5/4

3/2

NY
TX
NJ
NH
NO
NC

Phase 1

Phase 2

VT
VA
PA

WY
VT
VA
TX
SO
SC
RI

Rej.

MO
ME

SO
SC
GA

FL
RI
LA

PA
NY
NJ
NH
NO
NC
MO
ME

GA
FL

Figure 8.24 Alphabetic sorting. 96-column card and 5486 special feature.

First, the column selector knob is set to tier 1, column 27.
The column is sorted on phase 1. The resulting distribution of
cards is shown in Figure 8.24a. All cards are removed from the
pockets and replaced in the hopper in sequence: rejects first, 1/0
pocket next, etc.
Figure 8.24b shows the distribution of cards after the phase 2
sort on column 27. The entire procedure is repeated on column
26. Figure 8.24c shows the distribution of cards after phase 1;
Figure 8.24d shows the cards after phase 2. Figure 8.25 shows the
final sequence after the sort has been completed.

The 5486 Card Sorter can be equipped with an auxiliary card
counter to count the number of cards being sorted. The counter is
located on the operator's panel.

Counter On
(Special Feature)

164

Digit Select
(Special Feature)

SORTING AND CLASSIFICATION

Five keys on the operator's panel can be used to select cards
punched with specified digits. For example, pressing the 5/4 key
will select all cards punched with a 4 into the 5/4 pocket during
phase I sort. All other cards are sorted into the reject pocket.
During phase 2 sort, all cards punched with a 5 will sort into
the 5/4 pocket. All other cards will be rejected. One or all keys
can be pressed during a sort operation.
Column and tier for the specified digit are selected with the
card column selector knob.
WY
VT

VA
TX
SO

SC
RI

PA
NY
NJ

NH
NO
NL
MO
ME

GA
FL
Figure 8.2S Alphabetic sorting. Final sequence of two-character field.

Questions
and
Exercises

1. Explain why card decks are placed in the sorting machine "facedown,
9-edge toward the machine."
2. Why do sorters normally read only one card column at a time as cards pass
through the reading and selecting mechanism?
3. Outline your concept of good sorter operating practices.
4. Estimate 80-column sorting times for the following fIles:
Model 82 Sorter:
Model 83 Sorter:
Model 84 Sorter:
Model 84 Sorter:

50,000 cards
75 ,000 cards
106,000 cards
95,000 cards

6-digit field
4-character field
8-digit field
3-character field

SORTING AND CLASSIFICATION

5. If each of the above are block-sorted, show estimated elapsed time for
each first block to be in sequence.
6. What characteristics of the files must be known before you can make an
accurate estimate in answer to question 5?
7. How does the multiple-column selector shorten the sorting time?
8. If equipment is available, sort your requisition cards prepared in chapter 6
in preparation for a report showing item, description, unit price, total cost,
and department using. If equipment is not available, cards may be manually sorted to simulate machine action.

165

FILE
MAINTENANCE

It has been shown that many types of punched-card records
can be classified as historical. That is, the card is a coded representation of a transaction that has been completed some time in the
past. Examples of this type of record have been mentioned as sales
transactions, inventory activity records, payroll earnings, census
da ta, and so on.
In some cases, a record may be created in anticipation of a
transaction. For billing operations, cards are often prepunched in
advance of actual payment. When payment is received, the cards
then become historical data to be used in updating the accounts
receivable file.
Utility bills are a common example of such records. The bill
is submitted to the consumer of power, gas, or telephone service
in the form of a prepunched card. The record contains all the pertinent data for the individual account: customer identification,
type and amount of service, and the charge. The customer returns
the card with payment at which time the accounts receivable is updated to reflect the transaction, now historical data.
However, this chapter will discuss the handling and processing of information that might be classified as status records. These
are records which represent the present status or state of some
situation. And in every case, the data processing procedure is concerned mainly with keeping records up to date with changes in
166

FILE MAINTENANCE

the situation that the file represents. This operation is called
file main tenance.
For example, name and address files must be continually updated to reflect the corresponding changes of location by customers, subscribers, members, etc. Insurance files must keep up
with the changing status of policyholders with respect to births,
deaths, marriages, property loss and damage, accident liability,
amount of insurance, and premiums. Files of student records must
be continually updated to show academic standing, subjects completed, credits earned, etc.
Inventory files change as items are taken from the stock room,
ordered, and put back. Accounts receivable files change as customers pay their bills and charge new purchases. Payroll files
change as earnings are calculated and accumulated to new year-todate earnings and tax balances.
Two or more files must be handled in an operation of file
maintenance. One file, the master file, is invariably set up in the
approved collating sequence of the system. In punched-card accounting, this is preferably numerically ascending. The sequence is
established on an identifying or control field of the records. The
field usually contains such identification as customer number,
policy number, item number, or employee number.
A master file contains additional data related to the control
field; for example, value or quantity, balance due, number of items
in stock, policy loan value, hours worked, year-to-date earnings,
and so on.
Transactions that affect the status of the master file make up
the second file in the operation. Records in this detail file must be
identified in the same manner as the master file and must be
sequenced in the same order by the same control field.
There are three basic methods of performing the operations
of file maintenance:
1. Substitution of a completely new and updated master record. This means actually taking the old record out of the file and
replacing it with a new one.
2. Transcription of new or additional information to the master record. That is, the old record is updated with 'the addition
of new data. For punched cards, this means punching an additional
field and assumes that space is available in the card before the
data is added. This method is better adapted to computer filehandling practices.

167

168

FILE MAINTENANCE

3. Summarization of merged master and detail files. Information is calculated or accumulated from both files to update the
master card. This method is most commonly used to adjust amount
or quantity balances in the master record. For example:
Previous balance (master) ± transactions (detail) =
new balance (updated master record)
With 80-column cards, summarizing is done on an accounting
machine. The updated master is produced by a special summary
punch connected to the machine by cable. Reports or other documents are printed as by-products of the process and also as a means
of exercising procedure control over the file maintenance operation.
With the System/3, reading and punching of 96-column cards
is done by a multifunction card unit. Reports are prepared with a
printer that operates under control of the central computer.

IBM 88 Collator

As a review of the various ways of manipulating card files by
machine, the following sections will describe the operation of the
IBM 88 Collator (Figure 9.1). This extremely versatile machine
automatically performs almost every operation of file handling,
including record selection, sequence checking, merging, matching,
and editing.
Because the machine is designed to read and compare two
files of 80-column cards simultaneously, it is equipped with two
feeds, a primary and secondary. Normally, the primary feed is
used to read a master file while the secondary feed reads the detail
file. Each feed operates at a speed of 650 cards per minute. When
both are in use, a maximum of 78,000 cards an hour can be put
through the machine. The volume varies, however, depending upon
the operation being performed. Both feeds can be equipped with
a file-feeding device, each with a capacity of approximately 3,600
cards. Figure 9.2 shows the file feed in place on the primary
card hopper.
Figure 9.3 is a schematic diagram of card paths through the
collator. Five pockets are provided to stack cards as merged,
matched, or selected. Because decks of cards are transported from
opposite directions into one file, they must be placed in the hoppers differently. In the primary hopper, cards are placed facedown, 9-edge toward the machine. In the secondary hopper, they

•

FILE MAINTENANCE

Figure 9.1 IBM 88 Collator.

are placed facedown, 12-edge toward the machine. Figure 9.4
shows the card paths as they would appear looking down from
the top of the machine.
As cards are transported one at a time from the primary feed
hopper, they pass two 80-column reading stations: primary sequence read and then primary read. Each reading station is
equipped with 80 reading brushes. Cards pass between the brushes
and a metal roller, row by row, so that any column in the card can
be read.

169

170

FILE MAINTENANCE

•

____-----Card Weight
Magazine-------.... ,

Figure 9.2 File feed device on primary hopper.

Unlike the sorter, which has only one reading brush, the resulting multiple electrical impulses can be sensed by the machine
as a complete number or name, not just as a single digit or zone
punch. In all operations of card matching and merging,· the machine is set up beforehand to read a particular sequenced control

FILE MAINTENANCE

171

Secondary

Primary

~~~ ~~~~--\---,,----, f~--T-----r-~~~:;6-6~
Secondary
Sequence

Secondary
Read

Primary
Read

Primary
Sequence
R~

Secondaries

Selected
Secondaries

Merged
Cards

Selected
Primaries

Primaries

Figure 9.3 Card paths through the collator.

12-edge

9-edge

Secondary
hopper

Secondary
read stations

Merged
cards

Primary
read stations

Primary
hopper

Figure 9.4 Card feeding through the collator.

field of the file. The field to be read and the type of operation to
be performed are prewired into a control panel. The panel is inserted by the operator. Detailed instructions for wiring the control panel are given in the manufacturer's manuals furnished by
IBM. Some further explanation of control panel wiring is also
presented later in this text.
By proper control panel wiring, the control field of the cards passing through the primary feed can be read into a primary sequence
unit. Cards pass the two reading stations in exactly-timed feeding
cycles so that the control fields may be compared in the sequence
unit (Figure 9.5). That is, the selected card field sensed at the primary read station may be compared with the same field being read
at the primary sequence station. Because a file should normally be
in ascending sequence, the cards are in order if any card is either
higher than or equal to the card ahead. However, if a card is lower
than a card ahead, an error in sequence is indicated. The error is
recognized as a step-down or low-sequence condition.

Sequence Checking

172

FILE MAINTENANCE

Primary
Sequence Unit

/~2

1~-\
I
I

Primary

;:':~;"6'
6/\~
'..Y
----------Control

Primary
Read

Merged
Cards

Selected
Primaries

Primary
Sequence
Read

Primaries

Figure 9.5 Primary sequence unit.

Figure 9.6 Sequence error. Step-down error card.

Figure 9.6 shows a file of cards in ascending numerical sequence with one card (20) out of order. As the cards are fed
through the primary feed of the collator, a step-down occurs before
the error card as shown in Figure 9.7. This error condition can be
wired on the control panel to stop the machine. The cards are removed by the operator and manually arranged to restore the proper
sequence. In this case, the error card should be placed at the front
of the deck.

FILE MAINTENANCE

173

Primary
Sequence Unit

1~2
"

"'6'

Primary

Selection

(.---f---~r
5

Control

23
Secondaries

Selected
Secondaries

Merged
Cards

22
21

11-")
I

Primaries

Figure 9.7 Step-down. Primary sequence unit.

Figure 9.8 Ascending sequence error before step-down.

Figure 9.8 shows an error condition where the step-down
in sequence occurs after the error card (31). Figure 9.9 shows a
sequence where the card after the step-down is in error (23).
In addition, the two cards following the first error card are also
out of sequence (24 and 25). Notice that in this situation, the
machine can indicate only one error, the first step-down. Succeeding cards are in ascending sequence even though they may
be lower in order than the cards which have now passed both
reading stations.

Primary

~~/
Hopper

Primary
Read

Primary
Sequence
Read

174

FILE MAINTENANCE

Figure 9.9 Ascending sequence error-three cards.

The collator can also be set up to insert a blank error indicator
card into the file whenever a step~down condition occurs. Checking is done in one feed while any required blanks are inserted from
the opposite feed. Machine operation is not interrupted and correction of errors must be done manually after the file has been
processed. Indicator cards should be of contrasting color and with
an opposite corner cut from the original file cards.
Because the sequence checking just illustrated is electromechanical, it is possible to check a file only in the order that is valid
to the machine. The sequence must follow the order that is
designed and built into the mechanism by the manufacturer.
Secondary
Sequence Unit

Secondary Secondary
Sequence
Read
Read

Secondaries

Selected
Secondaries

Merged
Cards

Figure 9.10 Secondary sequence unit.

FILE MAINTENANCE

175

The 88 Collator can be used to check the sequence of a file in
either primary or secondary feed. The machine is equipped with a
secondary sequence unit as shown in Figure 9.10. When equipped
with an alphabetic collating feature, the machine can process alphabetic information-selecting, checking, merging, and matching
cards by names or titles. Each letter is treated as a two-digit number; that is, the numerical portion of a letter is read as punched
while the zone portion is read as aI, 2, or 3. Sequence checking,
either alphabetical or numerical, can be done in combination with
other operations of file processing.
Sequence checking of descending orders is also possible by
proper wiring of the control panel.

Each set of 80 reading brushes, primary and secondary, can also
be connected to a comparing unit in the collator with control
panel wiring. Figure 9.11 is a schematic of this arrangement. As
cards pass the two reading stations in sequence, the control field
from the primary read is compared with the control field from the
secondary reading station. One of three possible conditions can
result from the comparison:

Merging Two Files

1. The primary card is lower than the secondary.
2. The primary card is equal to the secondary.
cards match.

The two

3. The secondary card is lower than the primary.
The low primary, low secondary, or equal comparison is available as a timed electrical impulse to actuate the card feeding
and selecting mechanisms. How the comparisons actuate machine
Comparing
Unit

Secondary
Sequence Unit

Primary
Sequence Unit

Primary
Read

Secondary Secondary
Sequence
Read
Read

Secondaries

Figure 9.11 Comparing unit.

Selected
Secondaries

Merged
Cards

Selected
Primaries

Primaries

Primary
Sequence
Read

176

FILE MAINTENANCE

operation is a function of control panel WIfIng. Actually, both
of the comparing units and the sequence units are used to compare four cards simultaneously, one at each of the reading and
sequence stations. The following example has been somewhat
simplified to show only the results of comparisons.
Figure 9.12 is a schematic showing a master file and a detail
file to be merged in ascending numerical sequence. The master file
is placed in the primary feed; details are in the secondary.
Detail
file

Master
file

Figure 9.12 Schematic. File merging.
Detail
file

Comparing
unit

-<$>-

Figure 9.13 File merging. Equal comparison.

Master
file

FILE MAINTENANCE
Detail
file

Master
file

;m'~1

6
1

Figure 9.14 File merging. Low primary.
Detail

Master
file

file

secondar~
read

~
2

Figure 9.15 File merging. Low secondary.

177

178

FILE MAINTENANCE

The first primary card (20) is compared against the first ,secondary card (20). The comparison is equal as shown in Figure 9.13,
and the two cards are fed into the selected pocket. The primary
card is first into the pocket with the matching secondary behind it.
Succeeding cards in both files are read on the next feeding cycle.
The next comparison shows that the following primary card
(21)is lower than the next secondary (22)as shown in Figure 9.14.
Therefore, the primary card is fed into the pocket while the secondary is held. On the next cycle, the following secondary (22) is
lower than the primary (23) as shown in Figure 9.15. A low
secondary comparison results and the secondary card is fed into
the pocket.
Machine operation continues automatically until all the cards
are merged in sequence.

Merging with
Selection

Whenever master and detail files are merged together in prep aration for some further processing, it is usually necessary to be
sure all details are filed behind a corresponding master card. The
details must match the following by the sequenced control field of
the master file: employee number, account number, policy number, part number, etc.
For example, in the preparation of a payroll, it is always
necessary to identify each employee by name. Because such identifying data is constant and does not change from one pay period
to another, a master name file is usually kept as a permanent
record of the identity of each employee. Other information can
include address, number of dependents, sex, age, and so on. Also,
the limited capacity of cards usually does not leave room for name
and address in the same card with earnings, hourly rate, hours
worked, and other earnings information for a particular pay period.
The master file is kept most conveniently in sequence by
employee number. To prepare for merging the current earnings
records with the master name file, it is first necessary to sort all
current pay cards by employee number. The detail transactions
may then be merged with the master records to produce paychecks, earnings statements, pay registers, and other reports of
year-to-date earnings and tax amounts. The two files may be
merged automatically by a collator as previously described.
No employee's current earnings record should reach the final
payroll processing run without a name card. Therefore, during the

FILE MAINTENANCE
Detail
file

Master
file

Comparing
unit

seCOndar}
read

'------.,-------'

-<$>~

Figure 9.16 File merging with selection. Equal comparison~

merging operation, any current pay cards that do not match a
master name card should be selected by the machine for investigation. Cards so selected can mean a new employee on the payroll
with incomplete records or a lost or misplaced name card. The
selection of unmatched cards is also an important element of procedure control. It can be used as a check that no unauthorized
employees are paid.
For the collator run, the master file is placed in the primary
feed; current pay cards go in the secondary. All equal cards are
merged into a single pocket. Unmatched secondary cards are selected into a separate pocket. Figure 9.16 shows the merging
operation. The first cards (20) from the two files match, producing
an equal comparison. These cards are fed into the merge pocket.
On the next comparison (Figure 9 .17), the primary card (21)
is low and unmatched when compared to the next secondary (22).
This card also is fed into the merge pocket. Unmatched primaries
represent employees on the payroll who have no earnings for the
current period. However, they must be included on the payroll
register so they are not selected in this operation. The next comparison (Figure 9.18) shows that the secondary card (22) is low
and unmatched. This card is selected from the file and fed into
another pocket of the machine.

179

180

FILE MAINTENANCE

Detail
file

Master
file

Comparing
unit

<$>

c1i'~1

'
6·
20

Figure 9.17 File merging with selection. Low primary.
Detail
file

Master
file

Comparing
unit

<$>

4 1S~,"d'~1

'
6
20

Figure 9.18 File merging with selection. Low secondary.

FILE MAINTENANCE

The merging and selection process continues automatically.
Figure 9.19 shows that primary card 23 is also low and unmatched
with no corresponding detail card and that one additional detail
(24) has been selected. When merging has been completed, all
unmatched secondaries will have been withdrawn from the file as
exceptions to the regular procedure. Both files can be sequence
checked while merging and selection is carried pu t.
In many cases it is also desirable to select master cards that
have no matching details. For example, assume that a deck of
sales transaction cards must be punched with a unit price before
calculating a gross and net sales amount for each item. The item
cards may be priced by first merging them behind master pricing
cards and then transferring or gangpunching the price from each
master to all following details. The sequenced control field for
both files is item number.
Master
file

Detail
file

Comparing
unit

-<8?>Secondary
read

Primary
read

LJ
Figure 9.19 File merging with selection. Low secondary.

Figure 9.20 shows the pricing procedure in flow diagram form.
I. A batch of sales transaction cards is sorted to item number.
The master file is maintained in this sequence.

181

182

FILE MAINTENANCE

Figure 9.20 Pricing procedure. Flow diagram.

2. The two files are merged on the collator, selecting both
unmatched primaries and secondaries. Unmatched secondaries are
items that have no price. They could be new items in inventory
for which no price has been established. Or, the selected cards
might be punched with the wrong item number and therefore do
not match the current price list. A machine operator or clerk must
refer to the original document for correction or adjustment. When a
price is established or an item number corrected, the matching cards
are manually placed in the merged file in item number sequence.
3. Unmatched primaries are master price cards not used for
this operation. The selected deck is put aside un til gangpunching
is completed.

FILE MAINTENANCE

183

. 4. The transactions are priced by transferring the unit price
field from the master cards to following details.
5. The two files are sorted apart after gangpunching. Transaction cards are ready for further processing.
6. The selected master cards are returned to their file by
merging without selection.
All operations are sequence checked by item number.

The function of merging two files with selection can be utilized
to update a master file by substitution. In this case, equal primaries are selected rather than low primaries or secondaries.

Figure 9.21 Updating master file by substitution.

Merging with
Equal Selection

184

FILE MAINTENANCE

For example, in the previous illustration a master pricing file
was used to price detail transactions. When the price of any particular item changes, it is necessary to pull the old price card from the
master file and substitute a new one. The operation is one of automatic merging with equal selection. A flow diagram is shown in
Figure 9.21.
1. The new master price cards are punched from original
records including item number, description, and unit price.
2. All cards are key-verified for accuracy. An error in unit
price will cause an error in customer billing and consequently in
the accounts receivable file. The error will be repeated every time
the master card is used.
3. The new master price cards are sorted to item number.
4. New master price cards are merged into the master pricing
file by item number. New cards are placed in the secondary; old
masters in the primary feed.
S. As the secondaries are merged, equal primaries are selected.
That is, the new masters replace the old matching masters in
the file.
If required, any new masters without corresponding old cards
can also be selected. However, this selection need not be made if
there are item cards to be placed in the file representing entirely
new items in inventory.
Both files are sequence checked for accuracy.

Other File-Handling
Operations

The previous sections have dealt mainly with the type of file-handling operations that usually are carried out as processing before
the updating of master records. However, there are other collating
operations that can be performed to extract data directly from
cards by various methods of manipulating the record files.
These automatic operations particularly show the unique advantages of the punched card as a "selectable" record. That is,
cards can be automatically withdrawn from files at high speeds to
locate specific classes or individual items of information, as required. Once withdrawn, the same cards can be returned to the
file with equal facility by machine.

FILE MAINTENANCE

185

Some of these file-handling operations are described briefly
in the following sections. The operations can all be performed on
the 88 Collator.

The operation of matching one file against another is somewhat
like comparing one detailed situation against another and making
note of the exceptions. Instead of merging two files together in
one pocket, matched files remain separated in the collator and
are stacked in different pockets. Matching can be done between
primary and secondary feeds, either by single cards or groups of
cards regardless of the number of cards in a group.
Cards in either file that do not match can be selected. Thus,
the two original files may be arranged as four files at the end of the
matching operation: two matched files, selected primaries, and
selected secondaries. All four groups remain in the sequence of the
original files, that is, the ascending order of the control field.
Figure 9.22 is a schematic of a file matching operation. The
master file contains the names and addresses of all current subscribers to a buying service. The file is in ascending numerical
sequence by subscriber number.
Recently, all members were mailed descriptive brochures offering specially reduced prices on specific merchandise. Included
with the brochure was a punched-card order form to be filled out,
signed, and returned by the subscriber as an order.
The detail file in the illustration represents the returned orders, one card per subscriber. The detail file has been sorted to
subscriber account number. The purpose of the file match is to
determine which member subscribers have not yet returned their
order forms. Those who have not responded are to be sent a
second mailing reminding them that the offer expires shortly.
Master cards are placed in the primary feed of the collator,
order cards in the secondary. The machine is set up to match the
two files by subscriber number and to select unmatched primaries. Selected cards represent members who will receive the
second mailing
Since the secondary cards could only have originated from
the subscriber master records, it may not be necessary to wire sec~
ondary selection. However, there might be a special situation where
a time limit for the merchandise offer has been set. Membership in
the service might have expired for some subscribers. Consequently,

Matching Two Files

186

FILE MAINTENANCE

Figure 9.22 File matching.

their cards would have been removed from the master file. Selected
secondaries can identify any former subscribers who are no longer
entitled to the service.
After the second mailing is completed, the selected master
cards are merged back into the file.

Editing

As a check of punching accuracy, cards processed by the collator
can be checked for errors of double punching and blank columns.
Up to 22 columns in each feed, or up to 44 columns in one feed,
can be checked. Because the 88 Collator is essentially a numerical machine, double-punch detection is automatic for every position read into a comparing unit. To detect unpunched columns,
a switch for each position must be wired on the control panel.
When a double-punch or a blank-column error is detected and the

FILE MAINTENANCE

187

switch is wired, card feeding stops, a detect light is turned on,
and another light indicates which feed contains the error card.
The error-detecting circuitry checks punching positions 0
through 9, but does not normally check 11 and 12 punches.
However, complete detection can be specially wired on the control panel.
Double-punch and blank-column detection can be performed
with other operations of merging and matching, or as a separate
operation. When done separately, error cards can be selected into
a separate pocket.

Two fields in the same card can be compared to determine whether
they are equal or, if unequal, which of the two is lower or higher.
Therefore, it is possible to select from a given file all cards in
which field A is less than field B, field A is equal to field B, or
field B is less than field A. The two fields are compared in a sequence unit and a low or equal comparison is selected.
A file need not be in any particular sequence, but both field
A and field B must be punched in the same columns in all cards.
A useful application of this feature is the selection of records from
a given file where two amount or quantity fields are compared.
For example, a customer's total accounts receivable amount
may be compared against a credit limit amount. Any accounts
owing more than the limit can be selected from the file for
appropriate follow-up and any required action.

I STOCK

ON .,. 01/
HAND ORPFR

Comparing Two
Fields in the
Same Card

DATE 11
-O~ 0
a
PATE OF/AS} t
S.O.f;.
ACT/Y.
000000 00000000000000000000000 00000000 00 0000 00000 00000 000000 00000 00000 00000 000000
U VNIT

,yo.

J)£"S C/i/PTION

M COST

RE~E/I- IIYA/~.

MINIM

VEP

~;;;;;; 11111111111111111111111
~

89101'~tl~I~~~~~wnnn~~nnnH

031 ? 3'~ H lS J& J1 38 39 ~o 41 q? 4j 445 46 4~ 48 950515253 4555651585 606161636 656667686 10 n 1'17314 516171B19G
11 II 1 II 1 II 1111 11111 11111 111111 11111 11111 11 11 1 111111

~ 222222 22222222222222222222222 22222222 222222 22222 22222 22222

3 3 3 3 3 3 33333333333333333333333 33333333 3 3 3 3 3 3 33 333 33333 33333

22222 22222 22222 222222

33333 33333 3333 3 333333

444444 444444 4 4 44 4444 4 4 4 4 4 4 44 4 44444444 444444 44444 44444 444444 44444 44444 44444 444444

~ 555555 55555555555555555555555 555555 55 5 5 5 5 5 5 55555 555 5 5 555555 55555 55555 55555 555555

....
::;e 666666 66666666666666666666666 66666666 666666 66666 66666 666666 66666 66666 66666 666666
IU

:::'717777 77777777777777777777777 77777 7 17 7) 77 7 7 71777 77777 77 77 7 7 7177 7 77777 77 77 7 777 77 7

I-ot 888888 8 8 8 8 8.8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 8 88888888 8 88888 88888 88888 88888

88888 88888 8888 8 888888

999999 99999999999999999999999 99gg9999 999999 99999 99999 99999

9 9 9 9 9 999 9 9 99999 999999

1 2 1 4 5 & J8910"UOU~ffi"U~W~nnM~~n~~ OJ131JJJ.H53631 383940414743 fu4541i4148 49 50 ~1 !:253 of

Figure 9.23 Comparing two fields in the same card.

5~ ~s

57 58 5 606 1 62636 656661686 10 1112 n 1. 15 Ji1J781q0

188

FILE MAINTENANCE

Inventory status files often contain quantity fields for the
availability of items. Figure 9 :23 is a sample of how the card fields
in such a record might be arranged. Note that two of the fields
contain the quantities available and minimum balance. By comparing these two fields on the collator, any item with less than a
minimum balance in stock can be quickly extracted from the file.

Selecting Cards by
Either of Two
Key Numbers

Cards punched with either of two key numbers can be pulled
from a file in one run of the cards through the primary feed of the
collator. The selected cards are stacked in two groups.
The two key numbers are punched in a finder card which is
identified from all other cards in the file by a unique control punch.
The finder card is placed at the front of the file by the operator
before the run begins. By control panel wiring, the two key numbers are read into one side of the comparing unit where they are
stored during the entire operation.
As cards feed, they are read at the primary read station and
compared against the stored key numbers. An equal comparison
in either field can be wired to select the corresponding cards. For
this operation, the comparing unit is functionally split into two
sections, one for each key number.
Primary
Sequence
Unit

Comparing
Unit

Primary
Sequence
Read

Finder Card
Fields A & B

/"'- ------{L-__

14
_

t
. 5

14
3

19
2

17
16
15
11
14/19
1

(I---.--JIgr'--------l

_--l

Cards:

14/1y

-11-14-15-16-19-17-14

L///

Finder C;rd

Select:3

Select-2

Select-3

Figure 9.24 Collator planning chart.

The key numbers can be selected only in specified columns;
that is, a search can be made for all cards containing the number

FILE MAINTENANCE

189

2645 in columns 22 through 25 and the number 1006 in columns
32 through 35. If the key numbers are punched in any other columns, those cards will not be selected. Figure 9.24 is a collator
planning chart for this operation.
A number of other variations are possible. For example,
cards may be selected from a file that are equal to one key number
only, instead of two. Or cards may be selected that are either
higher or lower than a given key number. Finally, all cards may be
pulled from a file that are between two numbers representing
maximum and minimum limits.
Key numbers are stored in the comparing unit of the machine
by reading finder cards before a run begins. Each finder card must
be identified by a unique control punch to distinguish it from all
other cards in the file, like an 11 punch in column 20 when all
other cards have only digit punches in that column. The key number is always punched in known fields, both in the finder cards and
in the cards of the file to be searched.
The type of operation to be performed is set up beforehand
by proper con trol panel wiring.

Cards can be selected from a file that are punched with zeros in
any particular field. These zero-balance cards may be selected
from either the primary or secondary feed, or from both feeds
simultaneously. By placing half of the file in the primary feed
and half in the secondary feed, processing time is reduced. When
both feeds are used, a maximum of 33 columns per card can be
checked for a zero balance. If all cards are fed through the primary, up to 66 columns can be checked in one run.
Zero balances are detected by entering the field to be checked
into one side of the comparing unit, and emitted zeros into the
other side. That is, by control panel wiring, impulses timed as
zeros can be generated internally by the machine. These emitted
impulses can also be used to control machine functions during
certain operations in the same way as impulses obtained by reading
cards. In this case, a comparison between emitted zeros and zeros
read from cards can be recognized as equal. The equal comparison
causes automatic selection of the corresponding card.
Blank columns are not recognized as zeros. Therefore, a field
must be fully punched to be selected.
As an example of how this feature might be used.in an actual
data processing application, refer to Figure 9.25, an illustration for

Selecting ZeroBalance Cards

190

FILE MAINTENANCE

an inventory stock status report. It may be assumed that the information listed on the report is also punched in cards, one card
per line.
INVENTORY STOCK STATUS REPORT
Date _ _ _ __
Stock
No.

Descrip.

Old
Bal.

Receipts

Issues

Adjust

On
Hand

On
Order

Reserved

Available

Min.
Bal.

B
e
I

Standard
Order
Quantity

1234
1235
1236
1237

Bolt
Nut
Screw
Nail

150
300
500
400

25
50
100

40
150
300
350

10
25 cr
100 cr
50 cr

145
175
200

30
50
100
200

100

175
225
300
100

100
150
400
200

*
*

1000
8000

Figure 9.25 Inventory stock status report.

In control inventory applications it is often necessary to
quickly identify items with zero balances: items out of stock or
items with zero availability. Any such critical items can be readily
located in a file by searching for zero balances in the on hand and
availability fields. Once located, proper action can be taken to
replenish stock to meet estimated requirements.

Checking the Filing
of Details Behind
Master Cards

A file of punched cards is often made up of master cards with
corresponding detail cards. Such a file can be checked automatically to be sure that each detail follows the proper master. The
check may be particulary necessary when manual filing of cards is
part of the procedure of assembling the file.
When the cards are in some ascending sequence by a control
field, the operation can be done by a normal sequence-checking
operation. However, if the groups are not in any particular order,
they can still be checked if the collator is properly set up by
control panel wiring.
The master cards must be uniquely identified by an 11 or X
punch in a specified column. The control number must be punched
in the same field in both master and detail cards.
Cards are placed in the primary feed of the collator. The
master card control field is entered into one side of the comparing

FILE MAINTENANCE

unit and held until the next master card is fed. That is, when the
distinctive X punch is read by the machine, the comparing unit is
cleared and the new master card control number is read in.
Each detail card number is read into the other side of the
comparing unit. An equal comparison indicates that the detail
follows the matching master card. An unequal impulse (either high
or low) indicates that the detail card is misfiled. The unequal
impulse can be wired to select unmatched cards into a separate
pocket. They can then be refiled properly.
A number of variations on this type of operation are possible. The presence or absence of a specific X-punched card
within a group can cause a special card to be inserted after that
group of cards in a file. "Groups" of cards are considered to be
those which are all punched with the same control number in a
specified field.
For example, in a billing operation, the inserted card can be a
special discount, terms allowed, or description card to explain the
presence of symbols on selected customer invoices. It is assumed
that the X-punched card within the group signals the need for
additional information on the invoice. The special card is inserted
behind a group of cards that all contain the same customer or invoice number. However, the inserted card need not be punched
with the control number of the group.
Conversely, the last card of a group can be pulled from the
file only if it is punched with a specific X punch. This operation will remove those cards previously filed in the preceding
operation.
Note that these operations cannot be done on the sorter
because selection is made both by control number change and
signal card.
The ability of the machine to recognize the first card of a
group by a change in control number can be used to select that
card, leaving all other cards of the group in the file. Thus, master
cards can be pulled even when the file is in no particular order
and when the master cards are not punched with a distinctive
identification.
Single cards can be identified and removed from a file which
supposedly contains only multiple groups. The single cards are selected, leaving the multiple groups intact in a separate pocket of
the collator.
In some accounting machine operations, the last card of each
control group must be an X-punched card. For example, in a

191

192

CARRYING OUT STATISTICAL ANALYSES

payroll operation it may be necessary that the last card for each employee be a summary earnings card punched with a specific X code.
This can be checked on the collator before the cards are run on the
accounting machine.
In this operation, the last card of a group is selected if it is not
punched with the required X code. The selected cards indicate to
the operator which groups must be adjusted before the cards are
processed further.

Questions

and

The following exercises refer to the stock room procedure started in
Chapter 6.

Exercises

1. Consider setting up a master punched-card file for all items carried in stock,
one card for each item. The card may serve as a pricing record and an
inventory control record. What fields should the master card contain?
2. How is the master file to be originated? Maintained? Controlled?
Sequenced?
3. Assume that requisitions will be accumulated for a given period and then
sorted into the same sequence as the master pricing file. Merge the
requisitions as secondaries behind the master file as primaries. Select unmatched primaries and secondaries. Explain the origin of any unmatched
requisitions and unmatched master cards.
4. List the specifications for each master card field and design a card form as
a pricing record. The transfer of price from one record to another will be
explained in the next section.
5. Include controls for these new aspects of the procedure.
6. Present your proposal in class.

DATA TRANSFER BY
PUNCHED HOLES

Punched card records have another unique advantage over
many other types of documents. Information in cards can be
automatically copied from one machine-readable record to another
by the direct duplication or reproduction of punched holes. The
copying process can be varied in a number of ways, from the
direct duplication of a complete file, card by card, to the automatic transfer of selected data from one card to one or more
gangpunched cards.
This chapter describes some of the basic methods of punched
card information transfer. The processing is done by special equipment designed for this purpose, notably the IBM 519 DocumentOriginating Machine.
Information can be transferred in three different ways:
1. Gangpunching-the punching of information from a master card into a following group of detail cards. Masters and details
are identified to the machine by unique X punches.
2. Reproducing-the card-by-card duplication of punched
holes from one document to another. All or any part of the data
from a card can be punched into a duplicate card.

3. Summary Punching-totals accumulated by an accounting
machine can be punched into a summary card. For this operation,
the 519 must be connected by cable to the accounting machine
equipped with a special feature by the manufacturer.
193

194

DATA TRANSFER BY PUNCHED HOLES

Two automatic machine checks can also be used with these
information transfer operations:
1. Comparing. While reproducing or gangpunching, the machine can compare the duplicated hole pattern with that of the
original card. Any errors are indicated automatically for the operator's attention.

2. Double-Punch and Blank-Column Detection. The presence
of two or more holes in a column can be detected. The absence of
any holes can also be detected. The use of this feature will be
more fully explained with the operation of mark sensing.
Additional special operations can be performed by the 519:

1. Mark Sensing. Information recorded by pencil marks on
80-column cards can be punched into the same cards when this
special feature is installed.
2. End Printing. Up to eight digits of information can be
printed on the column 1 end of the card. The printing is in a
specially designed large type positioned to be read when the card
is filed vertically with the column 1 end up. For example, end
printing is used to originate punched clock cards for employee
attendance registration.

Gangpunching

Figure 10.1 is a schematic of the card feeds on the 519 DocumentOriginating Machine. Like the collator, the machine is designed to
feed two files of cards simultaneously. Therefore, the unit is
equipped with two card transport mechanisms with corresponding sets of sensing brushes. The machine may also be considered
to have two separate functional units that can operate independently on a single file, as required. These units are labeled read
and punch in the illustration.
The punch unit will be described first for the operation of
gangpunching.
As cards move from the hopper, they pass under a rotating
mechanism equipped with a set of 80 punches. The punches are
positioned above the card with 80 matching rectangular dies below.
Movement of each card through the machine is timed precisely so
that any of the 960 possible punching positions can be perforated
by extending a punch downward through the card into the die.

DATA TRANSFER BY PUNCHED HOLES
READ UNIT

PUNCH UNIT

Comparing and

Tr~~~~~~!ng

R~~~
Feed
Hopper

Reproducing
Brushes

195

Feed Hopper

.~
Mark
Sensing
Brushes

Gangpunching
and Interpreting
Brushes

Punch
Station

-f '~~i(~~ --==:::::::!.~ ~
Unit

Stacker

Figure 10.1 Schematic of card feeds, IBM 519 Document Originating Machine.

Punches and dies are in exactly the same arrangement as a single
row of hole positions in a card.
Since a card passes the punching station row by row, each card
is punched in the same fashion, that is, all 12 holes are perforated
first, then the 11 holes, then 0, 1, 2, and so on to the 9 row.
However, to actually cut holes, the punching mechanism must
be actuated at the proper time in the feeding cycle. This is done
by sensing the preceding card at a reading station. Here, cards pass
between a set of 80 gangpunching brushes and a metal roller where
any card field can be read. Selection of the columns or fields to be
read and punched is part of the process of setting up the machine
before operation begins.
The card being read moves under the gangpunching brushes
exactly synchronized with the following card at the punching
station. The 12 row of this card is under the brushes at the same
time the 12 row of the following card is over the punching dies.
All the following rows also pass the reading and punching stations
at the same time. Thus, to actuate the punching mechanism for
any particular column, it is only necessary to connect or wire the
punch for that column to the corresponding brush at the reading
station. The electrical impulse obtained by sensing a hole is used
to copy or duplicate that hole from one card to the next.
The 519 is equipped with a control panel. By panel wiring,
any gangpunching brush can be connected to the corresponding
punching die. When the control panel is inserted and the proper
control switches turned on, the machine will perform the operation
of gangpunching in the punched feed.
For example, assume that a procedure requires a file or deck
of cards to be dated for accounting purposes. The date will be
punched as a six-digit number in columns 1 through 6.

~
Stacker

196

DATA TRANSFER BY PUNCHED HOLES

Before operation begins, the control panel of the 519 must
be wired to connect gangpunching brushes 1 through 6 to punching positions 1 through 6. Next, the first card of the deck is manually keypunched with the correct numerical representation of the
date, e.g., 013172 represents January 31,1972.
The deck of cards is then placed in the hopper of the punch
unit. When the machine is started by the operator, the first card is
transported past the punching station in one feeding cycle. Since
no card has yet reached the gangpunching brushes, the card passes
the punching dies already punched.
During the next feeding cycle, the following card (blank in
columns 1 through 6) passes the punching station while the first
card passes the gangpunching brushes. Since brushes 1 through 6
are connected to punches 1 through 6 via the control panel, all
holes are gangpunched from the first card to the second.
On the next feeding cycle, the third card passes the punching
station while the second card passes under the gangpunching
brushes. The punching in columns 1 through 6 is gangpunched
from the second card to the third. The process is repeated until all
cards are punched. Thus, the information from the first card is automatically copied into all succeeding cards of the file.
Accuracy of the entire gangpunching operation can be verified
by sight checking the last card in the file with the first card. The
punching in columns 1 through 6 should be the same (Figure 10.2).

Figure 10.2 Schematic. Gangpunching.

Notice that the machine is set up for this operation in such a
way that any holes in columns 1 through 6 are copied from the card
at the gangpunching brushes back to the corresponding columns of
the card at the punching station. If, by error, there are any holes
in columns 1 through 6 of the card other than the date 013172,
they will be picked up and also copied into all following cards. If

DATA TRANSFER BY PUNCHED HOLES

197

this happens, double punching can occur in one or more of the columns and the file will be damaged.
To prevent this type of error, double-punch detection for the
gangpunched columns can be wired on the control panel. If any
double punching occurs, feeding halts and an indicator light comes
on to signal an error. The operator can disconnect the wiring between the gangpunching brushes and the punching station by removing the control panel. The machine is then cleared of cards.
The operation can be restarted by punching the first card of the
remaining deck with the proper date.

Interspersed gangpunching is the process of punching data from Interspersed
master cards into matching detail cards. The master cards must Gangpunching:
have been previously collated or hand-filed ahead of the details X-Punched Masters
in sequence by an identification field; for example, man number,
part number, or account number.
The purpose of the operation is to transfer some specific field
of information from each master card into the following detail
cards. The gangpunched data can be used in subsequent steps of
processing that include calculating and printing. The data is variable from one master card to another. Common examples include
price, size, description, rate, and name.
The operation is similar to the process of straight gangpunching-with one important difference. The read-back from the
gangpunching brushes to the punching unit must be interrupted
each time a master card is at the punching station. Otherwise, data
from the last detail of a group will be double punched into the
next master.
To interrupt punching for master cards only, the machine
must sense which cards are master cards and which are detail cards.
As previously stated, master cards are normally identified with a
distinctive X punch in a specified column.
A special reading station is located just ahead of the punching
station of the 519 to sense X punches only (Figure 10.3). Movable
Punch X-brushes are mounted on a notched bar above the card
path through the punch unit. A brush may be set on any column
position along the bar and clamped in place with a setscrew. Six
Punch X-brushes are standard equipment with the machine.
Each Punch X-brush has a corresponding exit hub on the control panel. \Vhen an X punch is sensed, the resulting electrical im-

198

DATA TRANSFER BY PUNCHED HOLES

PUNCH UNIT

Punch
Hopper

Punch
Station

Gangpunching
Brushes

Figure 10.3 Punch unit. IBM 519.

pulse can be wired to cause suspension. of all punching during the
following feeding cycle. That is, the cycle when an X-punched
master card is under the punching station.

Figure 10.4 Interspersed gangpunching.

Figure 10.4 is a schematic of interspersed gangpunching.
Masters with varying numbers of matching details are fed through
the punch unit. The card field in columns 25 through 30 is wired
to gangpunch.

DATA TRANSFER BY PUNCHED HOLES

199

Although the above procedure is operationally correct, there is a
distinct advantage in using a unique X punch in all details rather
than an X punch in masters. This is to prevent damage to a file by
double punching.
For example, if an X punch in a master card is not sensed,
punching will not be suspended and double punching will result.
To prevent this situation, the 519 can also be controlled to suspend
punching for all No-X cards.
In this case, the Punch X-brush is set on the column containing the detail X. Whenever an X punch is not sensed (master
cards), punching is suspended. With this system, failure to read
an X punch will result only in a blank detail with no damage to
the file.
All gangpunching can be checked for accuracy by comparing
in the read unit of the machine as follows.

Interspersed
Gangpunching:
X-Punched Details

The read unit of the 519 is equipped with two sets of 80 brushes
as shown in Figure 10.5. By control panel wiring, the punching
in a card passing the comparing brushes can be compared with the
following card as it passes the reproducing brushes.
The field to be compared is wired from each set of brushes to
opposite entries of a comparing unit. These connections can also

Checking
Interspersed
Gangpunching

READ UNIT

Read
Hopper

Figure 10.5 Read unit. IBM 519.

200

DATA TRANSFER BY PUNCHED HOLES
READ UNIT
Gangpunched Card

•

READ UNIT
Gangpunched Card Gangpunched Card

Master Card

COMPARING
MAGNET

COVtPARING
MAGNET

Master card at comparing brushes.

Gangpunched cards at comparing

Gangpunched card at reproducing

and reproducing brushes.

brushes.

READ UNIT

Gangpunched card

Gangpunched card from previous group at comparing brushes.
New master card at reproducing brushes.

Figure 10.6 Gangpunch compare. Master to detail, detail to detail, detail to master.

DATA TRANSFER BY PUNCHED HOLES

201

be made on the panel. Any difference in punching between the
two compared fields is sensed by the unit. The resulting electrical
impulse signals the machine to stop. At the same time, an indicating mechanism displays the positions of the columns that do
not compare. The operator can remove the cards to investigate the
cause of the trouble.
The gangpunch compare operation requires that either master
or detail cards must be punched with a distinguishing X. Like the
punch unit, the read unit has special Read X-brushes for this
purpose. Refer again to Figure 10.5. The brushes are movable
and can be set to read any specified column. When an X punch is
sensed, comparison between the reproducing and comparing brushes
is interrupted for the next feed cycle. This prevents a comparison
between the last detail of a group and the following master. This
situation is shown schematically in Figure 10.6. Comparison may
be suspended for either an X or a No-X reading.
To check interspersed gangpunching, the file is taken from the
punch unit stacker and placed in the hopper of the read unit. Each
feed operates independently and one feed will operate while the
other is empty.

In the previous illustrations of gangpunching, it was assumed that Offset Gangpunching
card fields are always punched into the same columns of all cards
in a file. That is, a field in a master card is punched into the corresponding columns of each following detail of a group.
Occasionally, a gangpunching operation involves punching
from a master card into different columns of the detail. This operation is called interspersed gangpunching. For example, the information in columns 50 through 54 of the master cards can be
punched into columns 3 through 7 of the detail cards. This means
that field selection is necessary. Whenever a master card passes the
gangpunching brushes, columns 50 through 54 must be punched
into columns 3 through 7 of the following detail. Whenever a detail card passes the gangpunching brushes, columns 3 through 7
must be read and punched into the following card-if that card is
a detail. When a master card follows a detail, punching must be
suspended as in normal interspersed gangpunching to prevent double punching the master card. The field selection between columns
50 through 54 and columns 3 through 7 is made at the gangpunching brushes, since all punching is done in columns 3 through 7.

202

DATA TRANSFER BY PUNCHED HOLES

Columns 50 through 54 are only read; they are never punched.
As in normal interspersed gangpunching, master or details
must be identified to the machine by a distinguishing X punch.
In the punch unit, the X is also read by a Punch X-brush. The resulting impulse stops punching action between a detail and master
and, one cycle later, causes selection of columns 50 through 54
as an X field. Figure 10.7 is a schematic of offset interspersed
gangpunching.

Figure 10.7 Interspersed gangpunching.

Comparing Offset
Gangpunching

When offset gangpunching is compared, field selection is used
similar to that required when the cards were punched. The selection must be made to read the proper field from the card at the
comparing brushes. When a master card is at the comparing
brushes, the master field must be read for comparison with the detail card at the reproducing brushes. When a detail card is at the
comparing brushes, the detail field must be read for comparison
with the following detail at the reproducing brushes. The detail
field is always read at the reproducing brushes, and comparison is
suspended when a master card is at the reproducing brushes.
The detail field from the reproducing brushes is entered into
one side of the comparing unit, and the master or detail field from
the comparing brushes is entered into the other side of the
comparing unit.
Selecting the fields read at the comparing brushes requires the
use of a selector, a kind of switch that can be wired from the control panel. The selector is controlled by the identifying X in the

DATA TRANSFER BY PUNCHED HOLES

203

cards. As in the punch unit, the X reading must be delayed one
cycle. At that time the X punched card is at the comparing brushes
and the selector must be controlled.
Whenever gangpunching is compared, it is also good operating
practice to compare the identification fields at the same time. The
comparison checks the filing previously done on the collator or by
hand. If a master is filed ahead of the wrong details, or if a detail is
misfiled, the comparison check indicates the error in the same way
that incorrect gangpunching is indicated.

Selectors can be installed as special features in the 519 and other
card machines. Each selector is a two-way switch that can be
controlled automatically to select a connection over one of two
alternate paths.
Most common switches are operated manually. That is, a
switch is positioned in one direction or the other by moving a
handle, pushing a button, or throwing a lever. The selectors on card
machines are operated automatically by impulsing a pickup hub.

Selectors

Pickup

f(
NORMAL

:
N

(1)

(a)

c
TRANSFERRED
(b)

Figure 10.8 Selector. Normal and transferred.

Figure IO.8a is a diagram of one position of a selector set at
normal. An electrical connection is established between C, or
common, and N, or normal. The switch is held in contact at the
normal position by spring tension.
An electromagnet is positioned in relation to the switch as
shown. Assume now that the magnet is impulsed through the pickup hub (Figure IO.8b). The resulting magnetic attraction lifts the

204

DATA TRANSFER BY PUNCHED HOLES

switch against the retention of the spring and makes contact with
T, or the transferred side of the selector. In this position, a connection is made between C and T. The normal hub is disconnected
when the transferred hub is connected, and vice versa.
Figure 10.9 is a diagram of a selector as it appears on a card
machine control panel. The small circles represent holes in the
panel where plugwires can be inserted. A selector has two positions, all controlled in unison from a single pickup hub.
~-----SELECTORI
0

O~OTO

oNe

()

cCo

Figure 10.9 Selector. Control panel diagram.

If a card field is wired from reading brushes into the normal
side of the selector, the resulting impulses are always available at
the C or common hubs.
If a second card field is wired into the T, or transferred side
of the selector, the resulting impulses will be available at Conly
when the selector is transferred. Transfer is effected by impulsing
the pickup hub with an X reading from a Read X- or Punch Xbrush. This action of the selector makes it possible to read offset
card fields for the operation of interspersed gangpunching.

Reproducing

To duplicate or reproduce a file of cards, the read and punch
units of the 519 Document Originating Machine operate together.
Original source cards are placed in the feed unit and cards to be
punched are placed in the punch unit. Cards are fed from both
hoppers simultaneously, one at a time.
Feeding cycles are timed so that a source card passes the reproducing brushes, row by row, just asa card to be punched passes
the punching station. The entire source card or any selected fields
and columns can be reproduced, hole by hole, into the card at the
punching station.
By control panel wiring, the fields in the reproduced card can
be punched in any desired arrangement, regardless of the way the
data appears in the original cards. For example, columns 1 through
20 can be duplicated into columns 61 through 80, or vice versa.
AlI 80 columns can be reproduced, column for column, as required.

DATA TRANSFER BY PUNCHED HOLES

205

Comparison of the original and duplicate information can be
done one cycle later on the same run of the cards through the machine. When the source card is passing the comparing brushes, the
reproduced card is passing the gangpunching brushes. At that time,
the reproduced card is compared with the source card, using the
comparing unit. Any disagreement stops the machine. Columns
that do not compare equally are indicated to the operator for correction. Figure 10.10 is a schematic of the reproducing operation.

Figure 10.10 Reproducing schematic.

In a data processing procedure, a great deal of flexibility is
provided by the ability to duplicate files automatically in any arrangement. For example, any selected information can be reproduced from a file for processing while the source file remains in its
original sequence available for reference.
Report processing can often be expedited by creating a duplicate file which can be sorted to another sequence while the first
file is being run on an accounting machine.
Two files can be matched, card for card, on the collator. Then,
selected information can be reproduced from one file into another.
At the same time the matching identifying fields can be compared
to check matching.
The 519 is also designed so that gangpunching can be done in
the punch unit while information is being reproduced into the same
cards. Thus, a duplicate deck can be given an identifying punch or
mark to label it as duplicate in all subsequent processing.

It is often desirable to reproduce only certain cards from a file,

without disturbing their arrangement. Cards to be reproduced,

Selective
Reproducing

206

DATA TRANSFER BY PUNCHED HOLES

or those not to be reproduced, are identified by a unique X punch.
The operation is like normal reproducing except that punching is
suppressed for unwanted cards.
During the operation, the two feeds continue to operate in
unison. Cards not reproduced are completely blank as they are fed
into the punch unit stacker. The machine can be equipped with an
offset stacking device that leaves the ends of the blank cards offset from the reproduced cards. If there are too many blanks to be
removed by hand, blanks can be separated from the deck on the
card sorter.
Figure 10.11 is a schematic of selective reproducing.
X punched cards
not reproduced

Blanks

Figure 10.11 Selective reproducing.

Combined
Reproducing
and Gangpunching

For all combined reproducing and gangpunching operations, the
read and punch units operate together. If gangpunching is to be
performed from a single master card, a blank card should precede
the cards in the read unit. This is necessary because the master
gangpunch card in the punch unit should be one card cycle in advance of the first source card in the read unit to avoid reproducing
into the gangpunch master card. The reproduced data can be compared in this same operation. Figure 10.12 is a schematic of this
operation.

DATA TRANSFER BY PUNCHED HOLES
Reproduced
field

Blank
card

Gangpunched
field

Master
gangpu nch card

Figure 10.12 Combined reproducing and gangpunching.

Interspersed gangpunching can also be performed while reproducing. To prepare for this operation, two previous collating
operations must be completed:
1. The source cards and the cards into which information is
to be reproduced are matched card for card. Both decks must
therefore be in sequence by some identifying field.
2. The cards to be reproduced are merged behind X-punched
master cards by the same identifying field. If there are relatively
few masters, they might be hand filed.
The masters and the merged details are placed in the punch
unit. The source cards are placed in the read unit. By control
panel wiring, the 519 can be set up to:
1. Suspend punching whenever the X (or No-X) master card
is at the punching station.
2. Read from the gangpunching brushes to gangpunch the
specified field into the details.
3. Feed the master card in the punch unit one cycle without
feeding a source card. This is necessary in order to keep the two
files exactly matched, one source card for each detail.

207

208

DATA TRANSFER BY PUNCHED HOLES

Reject,

~~<-- -~{ Rejoc~ ~

Figure 10.13 Reproducing and interspersed gangpunching.

4. Reproduce from the source cards to the detail cards.
Both gangpunching and reproducing can be compared for
accuracy. Figure 10.13 is a schematic of this operation.

DATA TRANSFER BY PUNCHED HOLES

The 519 can also be equipped with an end printing unit that
will print numerical digits on the end of an 80-column card. The
unit has eight print wheels, each with the digits 0 through 9 and
one blank space. The location of the print unit in the machine was
shown in FigurelO.l.
Numerical information to be printed can be read from the
punching in the same card or from a card at the comparing brushes
in the read unit. Figure 10.14 shows that information can be
printed on either of two lines. The top of the first line is

t6 inch

from the end of the card; the top of the second line is

inch

from the end of the card. If printing is to be done on two lines
on the same card, two runs of the cards through the machine
are required.

Figure 10.14 End printing.

Figure 10.15 Inverted end printing.

The position of the printing line is selected by latching the
printing unit in one of two notches in a rail on which the unit
slides. Selection is done by the operator before operation begins.
When not in use, the unit is disengaged by sliding it to a third notch
in the rail.

209

End Printing

210

DATA TRANSFER BY PUNCHED HOLES

The unit also can be specified for inverted end printing,
sometimes required in the preparation of merchandise tags and
preprinted stub cards. Samples of this type of printing are shown
in Figure 10.15. For this application, the print wheels are permanently installed in the inverted position.

Mark Sensing

Original information on an 80-column card document can be
transcribed directly into punched holes by a process called mark
sensing. The information is first marked with a pencil in predetermined spaces on one surface of the card. From there it
is sensed by the 519 Document-Originating Machine and automatically punched into the specified columns. The marked information can be punched into the same card or into a duplicate card.
For marking, a card is divided into 27 vertical columns. One
mark-sensing column occupies the same space as three punchedhole columns. Each mark-sensing column is subdivided into 12
vertical spaces corresponding to the 12 punching positions in the
card (Figure 10.16).

Figure 10.16 Column arrangement and eight styles of mark sensing.

The 519 can be equipped with a special device to sense the
marks and punch corresponding holes, as wired on the control
panel. Mark-sensing brushes can be fitted to either the punch
or read units. The position of the brushes in the punch unit was
shown in Figure 10.1. Normally, marked cards are fed through

DATA TRANSFER BY PUNCHED HOLES

the punch unit, where the marks are sensed and then punched into
the same card. However, when the read unit is equipped with
mark-sensing brushes, marked cards may be fed through the read
unit and punched into a duplicate card in the punch unit.
The sensing of the marks is based upon the principle that a
pencil mark on paper can conduct electricity. The mark must be
made ~ith a lead of high graphite content. Best results are obtained
when cards have been clearly and carefully marked. Because the
marks must conduct electricity, only a special electrographic pencil
should be used. Ordinary pencil lead, colored pencils, or crayons
do not contain enough conductive material. A dense, black, narrow
mark from a sharp pencil is always a better conductor than a broad
light mark. The dense mark is much more likely to contain a continuous unbroken deposit of graphite on the surface of the paper.
For best results, the card should be placed on a hard, smooth
surface so that marks are not indented into the paper.

Figure 10.17 Mark-sensing marks.

The left mark in Figure 10.17 is an example of a mark made
by a single stroke of a sharp pencil. Graphite particles are shown
as deposited evenly and closely to conduct electricity.
The right mark in Figure 10.17 is typical of that made by several light pencil strokes. Evidently the pencil was not sharp, nor
was it pressed firmly on the card. The resulting scattered deposit
of graphite could not conduct enough electric current to actuate
the 519 sensing circuits.
Each pencil mark should be long enough to extend all the way
across the mark-sensing column. No other penciled writing can be
done in the area, since these marks may be read by the machine as
if they were intended for mark-sensed data. Marks may be erased
if enough graphite is removed to eliminate conductivity.
Mark sensing has the advantage that data can be recorded at
the site where the transaction occurs. Then, data marked at a remote location can be converted to punched holes in a data processing department without the intervening operation of keypunching.
As previously stated, up to 27 columns of information can be
marked on one side of a card. However, if the cards are to be read
in the read unit of the 519, column 27 is not available. Each of the

211

212

DATA TRANSFER BY PUNCHED HOLES

27 columns of mark sensing is read by a brush. One brush is made
up of three separate brushes. The two outer ones are connected,
and the center brush is wired to an exit hub on the control panel.
A mark on the card conducts a small amount of electricity from
the outer brushes to the center brush.
Because the pencil marks do not conduct enough electricity
to operate the punch unit, the electrical impulse must be amplified.
Each mark sensing entry hub on the control panel is an entry to an
amplifier. The mark sensing brushes must be wired to these entries.
When a brush senses a pencil mark, the small amount of electricity
is amplified and emitted from the corresponding exit hubs as an
impulse suitable to cause punching.
Mark-sensed punching is checked by the double-punch and
blank-column detection device, installed whenever the 519 is
equipped with the mark sensing feature. The device is used to
detect unpunched card columns, columns containing multiple
punches, or both. Consequently, if any column is not marked at
its source location or if the mark fails to convert to a punched hole,
an error is indicated. Similarly, if two marks are made in the same
column, the machine can also be controlled to treat this situation
as an error.
Usually, marked-sensed fields are marked with numerical information no more than 10 digits in length. Alphabetic letters
can be marked, but in this case blank column detection only can
be used in those columns so marked. Alphabetic marking is difficult to do efficiently because it is necessary to memorize the
SO-column card code for each letter.
In a typical payroll application of mark sensing, attendance
cards are prepared for each employee by reproducing and gangpunching from master cards. Figure 10.IS shows a sample form
used for this purpose. Employee name, number, tax class, regular
rate, and overtime rate are reproduced from the master deck. The
date of the pay period is gangpunched into the attendance cards.
During the same operation, the employee number is end-printed.
After preparation, the cards are placed in a card rack with the
printed ends visible at the top.
When an employee arrives at work, he removes his numbered
card from the rack and inserts it in an attendance recorder. The
time of day is automatically stamped on the face of the card in
the space provided. The employee then returns his card to the
rack where it remains during the work period. Time in and out for

!!~

!;(...II

C\I
M

TOTAL
HOURS
WORKED

REGULAR
HOURS

00
00

l!J

a:
i=
0

::J

REGliLA!'l
HOUnS
WORKED

•

«
o-« ....

~;
~o

-,w
0
0

~.. !

....

Il::

w
w
~

W
l!J

)0-

::I~
r~

,.,

ani
n.1

,."

t:::!1
n1

0
0

ani Iftll

ani
M.

"1
....an

0
0

-I

0
0

an
t::Il I
H:J

...

:OJ'"

0:1

,... .... :i::
~ ~O.l:
0

t:!1

....
an

lI:I

.:1

....

.:OJ

."iii

Employee No. - reproduced from a
master file in the read unit

Date-gangpunched from a master date card

' - - - - - - - - Employee No.-end printed after it is reproduced __________ J

!!~

TOTAL
HOURS
WORKED

...
N
M

_L l:::.cl :::,cl-

00
0
Q!

w
U
Z

a:
i=0

«
o-« ....
Z ::J
~,

!;(...II
)0...J

IW~

.!!!.'!ill

__ 2~
~ONl

~:::,c2:::,c2

2:::.c2:::.c2

-3:::,c~:::,c3

IIc

3:::, 3:1C'
C~:::'C4:

4~"",c4:::-

-,c5-

::5:::.c5~Jc5-.

--5-)~
.

Il::

....~
l!J

~Jf

-:--

•

6:::.c6:::.c6-

HOURS
WORKED

a
'" ani

REGULAn

r -,
W

OvERTIME
HOURS
WORKED

•

ani
M.

....an ~

t:::!1
n1

-I

"..

0
0

s:=:!1

I
0
0

ani Iftll

0
0

Ito

n.1

- "1
,."

lin

t:!1

t::Il I
H:J

,...
0

H1 lI:I

....

Figure 10.18 Employee attendance record.

lunch may also be recorded. At the end of the period, the card is
stamped with the time the employee leaves work.
At the end of the week, the total time worked is marked on
the card with an electrographic pencil as shown in the illustration.
Space is provided in another column to mark any days absent.
Space is also provided to mark a reason code, a good example of
how statistical data can be marked when required.
213

~"
0%
~

..~

214

DATA TRANSFER BY PUNCHED HOLES

All employee cards are then returned to the data processing
department. The marks are converted to punched holes by the
519 punch unit. Subsequently, the cards can be used in other
operations to calculate and prepare the payroll.

Questions
and
Exercises

1. What are gangpunching, reproducing, and summary punching?
2. What is the difference between interspersed gangpunching and regular
gangpunching?
3. Why must either master or detail cards in an interspersed gangpunching
operation be marked with special X punching?
4. What is the function of a selector? How is it actuated?
5. In exercise 4, chapter 9, it was assumed that master pricing and requisition cards are matched and merged in preparation for pricing. What field
can now be interspersed gangpunched from master to details?
6. What fields between master and details should be compared? Which comparison is really a machine check on the collating operation?
7. After gangpunching and comparing, what is the next logical step in handling the two files of records? Include these additional steps in your
procedure write-up.
8. As an alternative to keypunching requisitions after they have been filled
by the stock room clerk, it might be possible for the secretaries to mark
sense the item number and quantity requested on the card beforehand.
Design a requisition card for mark sensing with a two-digit department
number, a three-digit item number, and a three-digit quantity field.
9. Devise a way to use one additional mark -sensing column for unit of
measure. (Hint: Code I might represent "each," code 2, dozen; code 3,
box; etc.) Include this field in your card design.
10. In class discussion, compare the two methods of punching the stock room
requisitions: keypunching and mark sensing. Discuss the advantages and
disadvantages of each method.

MACHINE CALCULATING
OPERATIONS

Nearly every complete data processing procedure includes
calculation. In commercial accounting applications, calculation is
usually a matter of simple arithmetic to determine total amount
derived from quantity and unit value. This category of calculation
includes such common applications as billing, accounts payable,
accounts receivable, inventory and production control, ledger accounting, and so on. Payroll and cost accounting require the
multiplication of hours worked times rate of pay, tax rates times
gross earnings, unit cost times quantity produced, and the like.
Practically all commercial calculation can be completed with
the use of the four fundamental operations of addition, subtraction, multiplication, and division. In the majority of procedures,
division is not usually required unless it is necessary to obtain averages of such factors as rates, cost, and sales activity. Specialized
equipment is available to read factors from punched cards, calculate, and punch results in either the same card or in other designated cards, as required. Operation of the IBM 604 Electronic
Calculating Punch is described in this chapter.
The 604 has two basic units: a calculator and a card reading
and punching device, as shown in Figure 11.1. The units are connected by cable, so that information read from cards is transmitted
215

216

MACHINE CALCULATING OPERATIONS

Figure 11.1 IBM 604 Electronic Calculating Punch.

to the electronic calculator where results are produced and then
transmitted back to the punch for recording. Each unit is equipped
with a removable control panel which can be wired to cause the
machine to perform a variety of calculations. All operations are
carried out with 80-column cards.
The 604 can perform a series of operations in one pass of the
cards; that is, two factors can be added, the sum multiplied by a
third factor, the product divided by a fourth factor, and so on. The
order of calculation is predetermined by panel wiring a sequence of
program steps. Throughput is at a constant rate of 50, 100, or 200
cards per minute, depending upon the model. The (calculation
:
capacity is as follows:

Addition and Sub traction. Totals or factors of as many as 13
positions can be accumulated in a counter.
Multiplication. An eight-digit multiplicand can be multiplied
by a five-digit multiplier to produce a thirteen-digit product. Larger products can be produced by one or more additional multiplication recalculations.
Division. A thirteen-digit dividend can be divided by an
eight-digit divisor to develop a five-digit quotient.

MACHINE CALCULATING OPERATIONS

Group multiplication can be performed with either the multiplier or multiplicand as the group factor. This means that one
group factor can be punched in a master card that will then serve
as a constant· multiplier or multiplicand for all succeeding details
grouped with that master. Other group operations can also be performed, such as the accumulation of factors from a group of cards
and the punching of summarized or calculated results into the last
:
card of the group.
Figure 11.2 shows schematically the three types of processing
that can be performed.

Figure 11.2 Three types of punched-card calculation.

a. Each card record is calculated as it passes through the
machine. At least two factors are involved with results punched
in the same card.
b. A group multiplier or multiplicand is punched into a master
card as one factor. The second factor is variable and is punched in
each detail card associated with the master. Details are assumed to
be properly filed behind their corresponding master in a previous
collating operation.
c. Calculations are made for the sum of a group of cards as
well as for individual cards. For example, in a billing job, the unit
price of each item is multiplied by the quantity sold to punch the
billing amount in each card. As the extensions are made they are
added progressively to obtain the total invoice amount.
An X-punched total card is the last card of each group with a
discount rate by which the entire invoice amount is multiplied to
obtain discount amount.

217

218

MACHINE CALCULATING OPERATIONS

Discount amount is subtracted from invoice amount to a btain net amount. All three results, invoice, discount, and net are
punched in the X-punched total card.
Where storage capacity permits, calculations and punching
can be checked in one run. The punched result of the first calculation and the initial input factors are read back into the machine
so that they may be recalculated and subtracted from the punched
result. The input factors are reversed to check the machine operation. The net results of both calculations must be zero, thereby
proving the accuracy of the punched results.
If storage capacity is completely used up for calculation or
accumulation of totals, a second checking run can be made. Results from the second run are over-punched into the same columns
as in the first run. When results are identical, as they should be,
no additional holes are punched. When results differ in any card,
double punching will occur in the result field. If the machine is
equipped with a feature to detect double punching, the error
condition can be signaled to the operator. Blank columns, or
failure to punch in any column of the result field, can also be
detected. By control panel wiring, either checking method can be
used to stop the machine.

IBM 521 Card
Read Punch

The IBM 521 Card Read Punch has three stations as shown in
Figure 11.3. These are
1. A first reading station from where factors to be used in a
calculation are transmitted to the electronic calculator unit.
2. A punching station where results are punched. The station can also be used for gangpunching either with or without
calculation.
3. A second reading station where cards may be read for
gangpunching, recalculation, or a double-punch and blank-column
check.
The feed hopper has a capacity of approximately 800 cards.
Input records are fed one at a time from the bottom of the hopper,
12-edge toward the machine. Cards feed continuously from one
station to the next at the rate of 100 per minute (Modell) or 50
per minute (Mpdel 2). As shown in Figure 11.3, while one card is
being read for calculation another is being punched and a third is

MACHINE CALCULATING OPERATIONS

219

Card Hopper
(800 Card Capac ity )
I

------,

L2 ===lI
pg$Ld
!

First

Rea~ing

Results
are

Lo !
tat,on

f \

Punching
Station

areResults
punched\

Second
Reading
Station

<;;;=====
may
readbe

f~~~~~~

gan~~d~":hing

f;;c~:~ia~~~~i:~,

and controls

occurs

DPBC check

Card Stacker
(1000 Card
Capacity)

III
I

Figure 11.3 Card stations. IBM 521 Card Read Punch.

read for checking. Note that results are produced in a small fraction of a second in the interval (115 milliseconds) between the first
reading station and the punching station. Punched and checked
cards are automatically stacked to the right of the machine. As
with nearly all types of punched-card equipment, the machine
stops when the 1,ODD-card capacity of the stacker has been reached.
A number of keys and lights provide for operator control;
these include start, stop, and reset keys. There is a red indicator
light which turns on whenever the machine stops while power is
on, and turns off as cards are passing through the machine.
Indicator lights also show the status of checking features.
The unfinished program light turns on, the machine stops, and all
punching is automatically suppressed in those instances when a
calculation cannot be completed in 115 milliseconds. The doublepunch and blank-column light is turned on by the detection of
errors in a designated card field. The zero check light turns on and
the machine stops if, in the checking operation, the punched result
subtracted from the recalculated result does not equal zero. The
product overflow light turns on and the machine stops if the result
of a calculation exceeds the number of card columns provided for
punching. All check lights operate only when conditioned by corresponding proper control panel wiring.

The 604 performs all arithmetic calculations. There are five basic
parts to the unit:

IBM 604 Electronic
Calculating Unit

220

MACHINE CALCULATING OPERATIONS

1. Electronic storage units
Factor storage
Multiplier-quotient storage
General storage
2. Common channel
3. Column shift unit
4. Counter
5. Program unit
The factor storage unit (FS) contains 16 positions for the
storage of input factors read from punched cards. Storage is divided
into two 3-position and two 5-position units. Storage may also be
used to retain intermediate results during calculation.
The multiplier-quotient storage unit (MQ) can hold up to a 5position multiplier factor received either from a card, from another
storage unit, or from the counter. During division it is designed to
receive up to a 5-position quotient as it is developed. It may also
be used for the temporary retention of other factors, either read
from a card or developed during the calculating process.
The general storage unit (GS) is intended primarily for the
storage of calculated results until punch time. It may also be used
for input factors from the card. The 16 positions of general storage are divided into two 3-position and two 5-position units.
All of the electronic storage units may be considered as working storage during a series of calculating steps. Once a factor in
one of the storage units has served its purpose, the unit may be
reused for storage of some intermediate result. However, only the
counter and the general storage are output for punching results.
An 8-position common channel serves as a factor path between
all the storage units and the counter. Factors may be moved in
groups of three, five, six, or eight positions at a time under control
of assignment wiring on the control panel. Thus, only one factor
of up to eight positions may be transferred over the common
channel during anyone program step. Detailed description of
channel operation will not be included here.
The column shift unit permits factors that are transferred
from a storage unit to the counter or from one storage unit to
another to be shifted as many as five positions to the left of their
normal entry position. As many as five positions can be dropped
when results are transferred from the counter to a storage unit. An

MACHINE CALCULATING OPERATIONS

221

entry shift (read-units-into) moves all positions of the factor as
many as five positions to the left and prevents positions from the
right from reading in. An exit shift (read-units-out-of) permits the
dropping of as many as five places from the right of a factor.
Shifting is controlled by panel wiring.
Shifting which is inherent in multiplication and division is
accomplished automatically by the column shift unit on multiply
and divide steps.
A l3-position counter is used to perform addition, subtraction, multiplication, and division. Factors cannot be read directly
from a card into the counter, but results developed in the counter
can be punched into a card.
The program unit supplies electronic impulses to control the
arithmetic functions and provide read-in and read-out signals to the
storage units and counter. After data from the card is read into
storage units, calculation is started automatically by an impulse
originating from the exit hubs of program step 1. The machine
then proceeds sequentially through the remaining steps. Twenty
program steps are available with the standard 604. This number
may be increased to 40 or 60 as optional features.
The 604 is provided with switches and keys to control machine operation and with lights to indicate the status of conditions
that affect operations. Of these, only the program test key and
the program test light will be discussed here.
Depressing the program test key makes it possible to operate
the 604 manually one program step at a time. The program test
light turns on to indicate that the machine is in test rather than
run status. By using a test card and a planning chart (explained in
the next section), the operator can take a calculation through the
machine in complete detail. The contents of the storage units are
visible on a display panel located on the calculating unit.

The 604 is equipped with an operation display panel-an arrangement of small neon bulbs. When lighted, selected bulbs show all
factors stored in the machine and the operation being performed
during any given program step. The panel is used as an aid in locating errors in procedure planning or control panel wiring.
To check wiring, a test card is fed through the card read
punch. The punched results are then checked against the calculation planning chart. If there is disagreement, it is usually the best

604 Display Panel

222

MACHINE CALCULATING OPERATIONS

practice to check the wiring first. If no errors are discovered, each
program step in the calculation can be traced as follows:
1. The machine is placed in test status by depressing the
program test key.
2. The test card is placed in the hopper of the card read
punch and fed two cycles. This reads all input factors into their
assigned storage units.
3. The calculation is performed one step at a time by depressing the program advance key. The machine results appear on the
display panel where they can be visually compared with the manual
result on the planning chart. This procedure is repeated for all
steps in the calculation until the error is located.
4. The machine can be returned to normal operation by
depressing the program test key a second time.
Each position of factor storage, general storage, multiplierquotient storage, and the counter is represented by four neons
arranged in a vertical column. The top neon is given a value of 1,
the second a value of 2, the third a value of 4, and the bottom
neon a value of 8. The digit stored in anyone position is mentally
calculated by adding the values of the lighted bulbs in that column.
The arrangement to display the four digits 9637 is shown in Figure
11.4. The machine actually uses binary coded decimal representation for numerical values. This code is explained more fully in a
later chapter dealing with the IBM System/3 computer .

0***2
•o ____
o

*
9

__ 0

o
6

0
3

*4
0
8

Figure 11.4 Schematic. 604 display panel

MACHINE CALCULATING OPERATIONS

223

Other lights on the panel indicate the number of the program
step being executed, the operation being performed, and the position of shift when a factor is transferred from one storage unit
to another.
Digits in the counter are represented as 9s complements of
a positive number; a negative number appears as a true figure with
a minus neon turned on. Both 9s complements and true figures
are represented as binary coded decimals on the display panel.

All calculation performed by the 604 must be accomplished within the predetermined number of program steps available in the
machine. In any given step, the calculating unit can perform one
of a specified list of operations:
Counter Read-in Plus. The contents of a storage unit are
added to the contents of the counter. The particular storage unit
is specified by control panel wiring.
Counter Read-in Minus. The contents of the specified storage
unit are subtracted from the contents of the counter.
Counter Readout. The contents of the counter are transferred
to a storage unit. The counter is not reset or cleared. Storage unit
contents are cleared by the read-in.
Counter Readout and Reset. The counter reads out to a specified storage unit and is cleared at the beginning of the next
program step.
Balance Test for Step Suppression. The plus or minus status
of the counter can be used to activate a corresponding set of suppress on plus or minus balance hubs on the control panel. These
hubs become active on the next program step and continue to be
active until impulsed again or until the end of calculate time.
Suppress on Plus Balance. When the above balance test is
made, the counter is either plus or minus. If it is plus, suppress on
plus balance hubs may be used to suppress succeeding program
steps for that card.
Suppress on Minus Balance. If the above balance test shows
the counter to be minus, suppress on minus balance hubs may be
used to suppress succeeding program steps for that card.
The three program suppression operations just described provide the machine with the ability to condition its operation depending upon the positive or negative balance of results. Thus,

604 Operation
Planning

224

MACHINE CALCULATING OPERATIONS

the machine can follow alternate paths through a procedure as
directed by the actual results encountered in a series of calculations.
Multiply Plus. The contents of a factor storage storage are
multiplied by the contents of the multiplier-qubtient storage. The
product is developed in the counter and is plus. Multiplication in
the 604 is accomplished by repetitive addition. The multiplicand
in factor storage is automatically added and offset internally in the
counter the number of times called for by the multiplier. The maximum number of positions in the product is equal to the sum of
the positions in the multiplier (MQ) and the multiplicand (FS).
Multiply Minus. The product developed in the counter will
be minus. All other conditions of the previous operation, multiply
plus, apply.
Divide. The contents of the counter are divided by the contents of the designated factor storage. The quotient is developed
in the MQ unit. The remainder is left in the counter and can be
cleared by counter readout and reset.
Division on the 604 is the process of reducing the dividend
by subtracting the divisor the number of times that it takes to reduce the dividend to a remainder that is less than the divisor. The
operation is automatic once the divisor has been impulsed to readout and the divide hubs have been impulsed.
Half Adjust. A digit 5 is added to the position of the counter
designated by panel wiring. The operation forces a carry of 1 to
the next most significant position in the counter when the sum of
the 5 and the digit to which it is added is greater than nine. The
effect is one of rounding a result to the nearest penny, nickel,
dime., or dollar.
Zero Check. Corresponding hubs on the control panel can be
used to check the contents of the counter for a zero balance.
Figure 11.5 is the planning chart used to determine the sequence of operations required for a given job. The chart is divided
into (1) a vertical arrangement of the entire calculate cycle of program steps and (2) a horizontal arrangement of the various storage
units into which the factors may be entered.
Program Number. The first column on the chart identifies
the factors to be read in from cards (top line) and provides space
for writing in the program step number. Results to be punched
into cards are identified in the bottom line.
Operation Notes. This column can be used for notes of
explanation about each operation.

MACHINE CALCULATING OPERATIONS

Read-UnUs-Into, Out-Of. Read-units-into shows the counter
or storage position into which the units position of the factor is
read. Read-units-out-of shows the counter or storage position from
which the units position of the factor is read. The column is
normally left blank unless the shift unit is impulsed.
Program Suppress. Program steps may be suppressed depending upon conditions encountered in the sequence of operations; for
example, a positive or a negative balance.
Program Number. This column is the same as the first column.
Factor Storage. The four factor storage units are spaced in
this column. Two units have three positions each; two have five
positions. Transfer of data into or out of these units is indicated
on the line corresponding to the program step.
Note that in the heading of the factor storage column, an
assignment indication is made. For example, factor storage 1 may
be assigned to a 6-4 or an 8-6 position. Depending upon which assignment is· wired on the control panel, this means that the digits
stored in this unit will at all times be transferred as though they
were the leftmost digits of either a six-position or eight-position
number. This feature has the effect of coupling the storage units
together for the purpose of transferring data during the calculation
sequence. Factor storage 3 may also be given an 8-6 assignment.
Assume that factor storage 1 contains the digits 123 and is assigned the 8-6 position. Also assume that factor storage 2 contains
the digits 45678. If FS 1 and FS 2 are both impulsed to read out
into the counter in the same program step, the counter will receive
the number 12345678.
Further assume that FS 1 again contains the digits 123 but is
given the assignment of 6-4. FS 3 contains the digits 456 and is
not assigned. If FS 1 and FS 3 are both impulsed to read out in the
same program step, the data transferred will be the number 123456.
On the planning chart, the factor storage units are shaded on
the bottom punch line to indicate that they cannot be used for
output data to be punched into a card.
Counter. The 13 positions available in the counter are shown
in this column. The spaces may be used to write the results of a
calculation as determined manually. The top or read line of the
column is shaded to indicate that data cannot be entered directly
into the counter from a card.
General Storage. The four general storage units are spaced in
this area. Two units have three positions of storage; two have five

225

226

MACHINE CALCULATING OPERATIONS

/';--r ll)

no.

I~g

READ

IJMP,

OPERATION NOTES

SIZE tJ,c Z3/lS6 If/UMtse.e
lJETERMINEb 011 RENb

Cycle

lY
V
V
V
V

8PSENUMBE2

3
Ii?EMIJlIIDE/Z

iiI

4, IR#'2

7
18

V
V
la'
V

C'ONTlPOl)

Mut.1'IpLy

le;R
1

10 II/Dj) 8f9SE KV.f?.8ER

I II

QuorlElir 70

liZ

/;,

1/3

C'oKrA2OI..

1/5

1/7
1/8
/9

ZO
21

Pi;
,tJj)D

MtA?8ER

Q(}oTJE,.#' "-0
1,3

0$4

123

-coN'r.E'oi

.5' PalM

F 124

125
12'
12/
<

---;;';ulml

{/UOTlelV7

IZ9

,tJ4

'lj)l)

130 PIP

('ot./lv'~cR

eMIT

132 /J?//L7/,DL)I

L... ___ ' ' _

(J.vrRDt'.

--...,.....-

TORAO.

',:

A'SlONM'N

1sSI8t.E

: 1313131313:

.ev

Sf IT

Tl3]l;

5lTT

II:

1111:

: II:

1111:

/I?()

II:

1111:
1i>,r

'jl'l'l'l6:

1/101/1..1.. l'J:

.eo
1:

X?,Z

lTTT

15:
:

liz

9 II: 1111: II: /ISIII'I"
1/0

: Illn:

tTT: iTT 11i Ill;

1// II:

IIII

II:

R'R

1-;

IIII

1 TfT'

1I1I

I II I

l5I5Ol4! TIn

IIII

II:

1111: II

IIII

II:

1111: II

1111:

i I:

1111: II:

1111:

ZI'~If'IGI:

'PIJSSll1U

RO
I~

141

101

1lI.:l

II: 1111 : II:
I/~

II?

X?.r

II i 11I1 ! II:

:R.E

I!):

II:

1111
;;;>0

'~'I"V:

8.3
: /l/181.3l5:

I..,

.po

RO
:

t.:
II: 11I1 ; II: 1111 : 1814141'li TIn
I:

c.v
'::1

liT Till

1111

-,f0

I""

II:

I III : II;

15:
,44

I: I I:

II;

III1 :

II11 :

II'

t2rtf~-

11OI/~

II:

RO
1II1 :

1111: II:

1111:

1111: II

IllTT

I111

:
RO

: II;

lTTT
RO

I'J'k!:II
10/

II:

""SSI8U'

l\'In
n

Ir 151'J':
'I?

12~

II!

II11 : II:

I ill:

: IIII

tf'R

/l?I

TTl I

II,

II:

1111 , II' 121OW1f:

II:

IIII

II i lilT'

II:

IIII

II:

;PI

131 I
12"

III1

1?I

I1I1 : 1I11 : 11I1

I,.,

II:

: 1/1f.VI4IZ

11T18141419111 I:

114

Ie>

I?o

II:

TT ~3"
I IT 1111 : III-I: 111210121814111,1; II: Illl
w,e'
II I I ; 1111 II III I II' 1111
I II I ' II' 1111

II i InT Inn:
h

'P"SSl8J!1f

I I ; IIII
II:

OEN"AL

/11

I ... ,

tJS4-

D"r.I7:~

COU,NTfR

I?L

IZ8 /IuD

131

7iJ ,rS4-

,./u,t'/8EI2

lY
IA
V
1/

/-8

'T"

MULT
QUOT.

:~II: In II: II, ilTn:-

'DIVILJE
4j)/)

/
A
V

I/~

lONM'!':

1/4 kJl;;;>

a§

I~~ ::1

~,,~'

, E,.?/rrlF/)

K,S=/)Z

fACTO' STO.AO'

I",

TITTTIT I:

15:

tf'!7

's()VIlRtF KOOT

I, TnT'

1111' TIn I I

1/~('\5

1III

~ ~<;()UNRERtJOl
Tll

1/'lfVU,{.

Figure 11.5 Calculating planning chart.

positions. The units may be assigned in the same manner as the
factor storage units.
The horizontal space given to a program step for the storage
units and counter is divided into two sections. The upper section
can be used to write in symbols, letters, or words that identify a

MACHINE CALCULATING OPERATIONS

227

factor or result. The lower section is spaced for the actual number
of positions in each unit. An extra space to the right is provided
to note the sign of the factor or result.

The 604 Electronic Calculating Unit and the 521 Card Read Punch
are equipped with separate control panels which can be manually
wired to adapt the machine to specific procedures. Panels are removable so that the units can be readily changed from one jo b to
another by inserting properly wired panels. Figure 11.6 is a schematic of the panels with all exit and entry hubs labeled. Shaded
areas indicate optional features.
Card reading, selection of data for entry into storage, and
card punching are controlled by the 521 or 541 panels. All
operations of calculation are controlled by the 604 panel.

Control Panels

The following sections describe the basic data processing operations
that can be performed by the 604. Some panel wiring diagrams
are shown, but these are not discussed in any detail. Various IBM
machine manuals are available for this purpose.
In studying these sections, the student's main emphasis should
be placed upon the data handling capabilities of the machine, because many of these serve to illustrate methods of factor manipulation used by computer and other automatic calculating devices.

Basic Operation

The 604 executes a series of 20 program steps during each card Crossfooting:
cycle. Program steps are electronic cycles that occur successively. A - B = ±T
All steps are taken whether or not they are actually used to complete a calculation. As many as 40 or 60 steps can be executed by
machines equipped with these optional features.
The number of factors that may be crossfooted is limited
only by the amount of data that can be stored during a read cycle.
In this example, a simple crossfooting operation is performed to
subtract factor B from factor A. The result T may be either plus
or minus; if minus, the true amount will be punched with an X
punch over the units position of the card field. Factor A is
punched in columns 31 through 35 and factor B in columns 25
through 27. The result T will be punched in columns 70 through

----,--2.

1213

'415

1617181920

2122

r--r---PRO~J.~.~.:~:~ACTORSTORREADI~aUS-,--CAlC.SEt~
0 j 0
0
0 I:·:·:ct:·:t:·:·:a-:·:·:·:·:-o:·:::::::::e::;:t 0
Q
0
Q
I 9 Q Q lOT 0 0
0 T 0 I

oeo

I
A

rCALC:r--PCH

91011

151617181920

2122

23242526

SUP----.---8US~X--PILOT
SE.L.!C... ...
TO
...'....S
... '.'.::.'.'.'.'.'.'.'.".'.'.'.".''''-.:'. DIGIT IMPULSES I
I 0----0----0---< I Q 9 (( <;> I( :::~:::::~::::::9.::::::~:;::::~~:~ 0--0----+---0-- 0

0--0 I 0---0---0---0

1 0

'I 0

0, 0

313233

0 2 0

FACTOR
0

3536

~lORAGE
ENTRY
I 0 3 0 0:

0, 0

0 : 0

OeQ

U~T ~U06

EN;,V31
~

aNa

:
,

M I 0---0--0---0 I 0--0--0

0 T.o

':::·ii·····. :···~·. :·:·:·:·..dl::::

r;=~R ~AG~IGNMENT
U

6-4

3.11 .. ~'1

3·1

0 3 0

0 I 0

02 0

o ~pu~cHOsIEl~CT~RS

000000

,1 31 ~I

ro-o·

~OU~TE~ EX~To-o

RlR-,

--PUNCHING---IS
0000000
30
3.5

050

OOlo.o~o~l

0---0--0---0 I

VOOOOOOOOOO

"~'ACTO' STO'AGf 'fAD INl

woooooooooo

"rrt
rrrrT

0 0: oj
o

a
~

~o,
.0

a
~

0-9

0000000000

AGf

,-T

PU~CH oSfIL~CT~'S a

~T1

RfAO.INl

AGf
'fAo-oUTl

0 1 0

, - , - - - - S - - - S E C O N D READING--IS00000000

o
70

- 'a0a

..
a

I.a
a

__________________

~¥~

__________________

604 Calculator Control Panel

0: (5

oNE~At gTIO~~G; E~T:~o 0 0: 0 I

12 & ll--COLUMN SPlITS---------,

a
~

GE~ER~L:s5~~... ~~ EN6R:Y 0 10

0100

OeQ

0 1 0

a
~

0-0

~~L ~RAGE ASSIGNMENT~~ II
'-6 6-4 ).,

OeQ

~~~~~~~~
r;MP

oeQ

:0'00030

oCo

a
a

"n \, \, \,

rIN-PROO OVUFlOW-OUT9
...................... · "ALCULATO' SflfCTO'S-O-I Pui
!-=----=--COUNTER CONTROL-----1

a
~

'\ '\ '\ '\ '\ '\ '\ '\ a
0 \, \, \, \,

000000000

IoZE~O ~~I~~:f::.::~
:.:.'!:.:....::: '-PUNCH SllfCTO'S-I~I

,----OPBC--rs-2NO RfAO
o TOo

0000000

r~T~~t:~?~:~~CO~PL~NG~ l;:'~t ...........

a
~

521-541 Card Read punch Control Panel

Figure 11.6 604 and 521 control panels.

MACHINE CALCULATING OPERATIONS

74. The wIfIng is shown in Figure 11.7, the planning chart in
Figure 11 .8. Typical values of the factors are manually inserted
to plan the calculation.

Figure 11.7 Control panel wiring. Crossfooting.

Figure 11.8 Planning chart. Crossfooting.

Read

Factor B is placed in factor storage 1; factor A in factor
storage 2.
1. On the first program step,A is read out of FS 2 and placed
in the counter (read-in plus). Note that the position of the decimal point is also shown for planning purposes. Whole numbers
may be crossfooted with decimal numbers by shifting the transfer

229

230

MACHINE CALCULATING OPERATIONS

the proper number of positions to the right or left (read-units-into,
out-of).
Punch

All remaining program steps are not used. As the card passes
the punching station in the card read punch, the result is punched
in columns 70 through 74.

Multiplication:

AXB=P

To do multiplication, the multiplicand is placed in any storage
unit; the multiplier in the MQ unit. Factors can be read into the
units directly from a card, or they may be developed as results
from previous calculations. The designated storage unit is impulsed to read out and in the same program step, the machine is
impulsed to multiply. The product is developed in the counter.

Figure 11.9 Control panel wiring. Basic multiplication.

Figure 11.9 is a wiring diagram for basic multiplication. Multiplicand B is assumed to be an hourly rate field read from card
columns 6 through 10 directly into factor storage 4. Multiplier A
is an hours worked field read from columns 2 through 5 into the
multiplier-quotient. Product P (hourly earnings) is to be punched

MACHINE CALCULATING OPERATIONS

231

into columns 76 through 80 from the counter. Note that the last
two counter positions are not punched and that any significant positions to the left of the five-digit field are wired to cause a product
overflow check. If the product exceeds the five positions expected,
the machine stops when the card in error reaches the stacker. The
product overflow light on the Card Read Punch comes on.
Figure 11.10 is the planning chart for the example of multiplying hours X rate = earnings.
PROBlEM

AXS·P

MULT
QUOT

Figure 11.10 Planning chart. Basic multiplication.

Read

The hourly rate (multiplicand B) is read into FS 4. The rate
has three decimal places, for example, $1.255 per hour. The hours
worked (multiplier A) field is read into the MQ. The field has one
decimal place, as shown.
1. FS 1 is read out and the machine is impulsed to multiply.
The product is developed in the counter. The product has four
decimal places.
2. The product is half adjusted in the second position by
wiring read-units-into the second entry hubs.
Punch

Five positions of the product are punched into card columns
76 through 80.

The 604 provides the capacity to divide a 13-digit dividend by an
eight-digit divisor to produce a five-digit quotient in a single program step. If these limits are exceeded, the operation can be done
in parts with several program steps.

Division:
A -7- B = Q

232

MACHINE CALCULATING OPERATIONS

The dividend is placed in the counter; the divisor in any storage unit. The quotient is developed in the MQ. Decimal point
alignment between dividend and divisor is accomplished by shifting the dividend the required number of places as it is placed in the
counter. The divisor may also be shifted as required.

Figure 11.11 Control panel wiring. Basic division.

Figure 11.11 is the wiring diagram for an example of basic
division. The example illustrates the division of total pay (A) divided by hours (B) to produce an average hourly rate. A is punched
in columns 21 through 25 of the card; B is punched in columns 26
APPLICATION

TOTAL frlY.;.-lItJllRS=AVlR./lOURLY RATE

Figure 11.12 Planning chart. Basic division.

PROBLEM

ArB· Q

MACHINE CALCULATING OPERATIONS

233

through 28. The quotient Q is to be punched as output from the
calculation in columns 50 through 52. Figure 11.12 is the planning chart.
Read

The dividend A is read into FS 2 and the divisor B into FS 3.
1. On the first program step, the dividend A in FS 2 is read
out into the counter (read-in plus). The contents of FS 2 are
shifted to the third position of the counter to align the decimal
points to produce a three-decimal quotient.
2. The divisor B is read out of FS 3 and the machine is
impulsed to divide. The quotient is automatically developed in
the MQ.
3. The dividend remainder is cleared out of the counter by
impulsing the counter to read out and reset. The remainder can be
saved, if needed, by impulsing a storage unit to read-in during the
same program step. Note that the counter must be cleared in a
separate step; it does not clear on a read-in as the counters do.
4. The quotient is transferred to the counter in preparation
for half adjustment to the nearest cent.
5. The quotient is rounded to the nearest cent by adding 5 to
the units position of the counter. The remaining program steps are
not used.
Punch

The average hourly rate is punched in columns 50 through 52.

Successive calculations can be made involving any of the arithmetic and data handling operations the machine can perform.
The number of calculations is limited by storage capacity for
input factors, availability of storage for intermediate results, and
storage for accumulating final results to be punched in cards.
Figure 11.13 is the wiring diagram for the calculation of the
formula (A X B = P) 7 C = R or the example of (Pieces X unit
standard time = total standard time) 7 actual time = % over or
under standard.

Successive
Calculations:
(A X B = P) -;- C = R

234

MACHINE CALCULATING OPERATIONS

Figure 11.13 Control panel wiring. Successive calculation.

A card field A (columns 1 through 6) represents the number
of pieces of an item that a worker has produced in an actual period
of time. Field C in columns 21 through 25 is the actual time.
A is a whole number; C has two decimal places for recording to
hundredths of an hour.
The purpose of the formula calculation is to compare actual
production against an established standard rate, represented by

Figure 11.14 Planning chart. Successive calculation.

MACHINE CALCULATING OPERATIONS

field B in columns 9 through 13. The result R shows the percentage over or under standard. In some payroll plans, a bonus is paid
to the worker when standards are met or exceeded.
Figure 11.14 is the planning chart.
Read

Field A is read into FS 1 and FS 2, B into the MQ unit, and
C into GS 2.
Note that FS 1 is assigned to positions 8-6, effectively coupling FS 1 and FS 2 into an eight-position storage unit. This is necessary because the field A is a six-position field and cannot be contained in FS 2 alone. GS 3 is also given an 8-6 assignment to contain an intermediate result which will exceed the capacity of the
five-position GS 4. Decimals are placed in the factors as shown
below.
1. Factor A is multiplied by factor B by reading out FS 1
and FS 2 and by impulsing the multiply plus. Since there are four
decimals in B and none in A, there will be four decimal places in
the product, total standard time.
2. Total standard time (P) is adjusted to the nearest hundred th
of an hour by half adjusting. A digit 5 is added to the second
position of the counter in this program step.
3. The product P must be stored until punching time. Therefore, the counter is read out and reset with two positions dropped,
into GS 3 and GS 4. The positiotls are dropped by impUlsing the
read-units-into the third position.
4. The product P must now be transferred back to the counter
as a dividend in' preparation for the next step in calculating the formula. P is read out of GS 3 and GS 4 into the counter. Two decimals are required in the quotient, plus two in the divisor, plus one
for half adjustment, making it necessary to place the decimal in
the fifth position of the counter. Since P has only two decimal
places, it is offset four positions in the counter to align with the
predetermined decimal point. This is done by impulsing readunits-into the fourth position.
5. P is divided by C by impulsing the divide hubs on the
control panel and reading out C from GS 2 in the same program
step. The quotient is developed automatically in the MQ unit.

235

236

MACHINE CALCULATING OPERATIONS

6. The dividend remainder is not required and is cleared out
of the counter by a readout and reset operation.
7. The quotient R is transferred to the counter for half
adjusting.
8. The quotient is adjusted to the nearest percent.
Punch

The fields P and R are punched as output factors in columns
26 through 31 and columns 78 through 80 in the card.

~ign Control

Multiplication or division of factors identified by plus or minus
signs can be performed automatically on the 604.
Signs of factors are entered, together with the factors, into
any of the storage units that accept information from a card. When
the signs are represented by X punching over the units position of
the field, no special control panel wiring is required for sign control.
When the X is punched in any other card column, the column entry
to storage must be specially wired to segregate the X, which is then
wired to a sign control position for each storage unit. If an X is
punched over the units position of a field for reasons other than
sign control, the X must be eliminated when the field is read into
storage. This is done by wiring a device on the panel called a
column split.
The rightmost position of each storage unit is reserved for sign
control, as shown on the planning chart. This position accepts any
digit or an X to identify the positive or negative value of a factor.
By an internal column split, an X punch over the units position of
a factor is automatically entered in the sign control position of the
unit, with no wiring required. A digit can be used for sign control
by wiring from the card column containing the digit through a
special digit selector on the panel.
The calculation of a positive or negative result as required by
the combination of signs is completely automatic, including the
transfer of factors or results from one storage unit to another.
There is a special channel for the transfer of signs.
When multiplying, the product is positive or negative, depending upon the combination of signs for the multiplier and multipli-

MACHINE CALCULATING OPERATIONS

237

cand, and whether multiply plus or multiply minus is used. The
following combinations are possible.
Multiply Plus

Multiply Minus

(+A) X (+B) = +P
(-A) X (-B) = +P
(+A) X (-B) =-P
(-A) X (+B) =-P

(+A) X (+B) =-P
(-A) X (-B) =-P
(+A) X (-B) = +P
(-A) X (+B) = +P

Sign control for division follows the standard rules of algebra.

The square root of a base number containing eight digits or less
may be produced on the 604 by four iterations of the following
formula:

(1z + Ai)
where

B
A1

X 0.5

= base number for which the square root is to be fO,und

= first approximate square root

A2 = second approximate square root.
A1 is set up in the machine from a digit emitter. Ai may be
3, 30, 300, or 3,000 depending upon the number of significant
digits in the base nUlllber. If the base number has one or two digits,
3 is used. For a three- or four-digit base number, 30 is used. For a
five- or six-digit base number 300 is used; for a seven- or eight-digit
base number, 3,000 is used. Theoretically, any number can be
used for the first approximate square root. The number 3 is chosen
because greater accuracy is obtained when the root is carried beyond the decimal point.
As indicated on the planning chart in Figure 11.1 5, A2 and
A3 are the successive approximates developed by the machine on
each iteration. A4 is the approximate square root.
Some base numbers do not require four iterations, but a maximum of four is required for any base number from one to eight
digits. For example, the square root of 16 is developed in one
iteration:

1; + 3 X 0.5 = 4

Approximate
Square Root

238

MACHINE CALCULATING OPERATIONS

I~ ,4)., tr~LJ_

I~g

i~M?,

OPUATION Nons

5/z£

. . .0

d.c 13/15£ ,vUMU£

l)ErERM/NEb

oN RE.<'b CyclE

QiJontNf

"r",,,V>

&,,,,,,r,,,,,

CONTA?OI.)

,Al/UI.71pLy

let'"
1/0 I/lj)lJ 8195£ A1t,.,8£R
i

V
V
lY

ZJ/no~

,nENr 70

!/.3

/lz,

1/4

C"ulir£R
('OAff'(!Ot..

1/5

A.
/
/
Y
Y

1/7 Pi,

1/8

#UMBER

/JiJi)
'Z/IV/OE

1'1

20 /llJlJ
i21

1/i>i:J

122

'fK

q(/",;?E,I/r' 7"0

123

124
125
12~

127
128
IZ'1
130

".3
~m.;£

.5'

F",(/~ml

l/i:Jv 8.<'sE N'utVllJE.e

Ilj)() quo

-"',.."

1/4

P(;? ('ouNr£R

i31

EMIT

132

;1?uL7/Pt.y

L..-- .....

iA
'/
/

V
V
7iiS4lA
V
1/
IY
IY

C;S4-

cr;;;;';L

G'N",

.3
4

III
13ICr.*:-::
B
Q

TIl:

Ie.

110

III

II: rni: IF Till:

IT ITIT-; II, I I11J

Ll II·

I
NO

'IT

1/1o'II/1l."l'>:

.Q.a

RI'.e

1/1'J1lIt;1"

hn t"nl1 ! Ill: Ilrfr
;po
II i III Ii Iii 1IIIi

I TlT:"

III Tl L _lJ : 1111: II:

II "

;eo

-.!""

:
Q

'P,,!/S//IU

i I:

1515'014: TIn

III

Ii?O

~141
18/~r"

1/3
1/4
1/5
1/

TnTl

13131313131

R"hIlTT: fl: I nTH TF 3~~613:
15,

!ff: iil'

!Tin, "1 n3Itl~l~grTT11
1i'J

7
!

.et:

HlT :

IZ!3"17IOI
R#'R

Pf
:

I!:j:

nTn

II'

21~;71O:

~TT':
120

.Q4

IIIIDI"",V:

/I,)

IT' TTlT"-

ilI/18W

""''

If?O

: "1:"n'

pp~S/!!"!.

1814141'1:

iTT!:

TTTIJIiMI9111T

I."
I-

II,

In:

77"8"33'
12101218141

.('t)

I-I

1111:

11I1

nn1rfl Tn L

11 IT: rrIIS:

II: rill:

t!flM
iTl ,-

1-'
.<'4-

111[' I

1-'

1/1OI/14IZ!O
Ii?1'R

114
II'lYI4I2.

I'"
I"L

127

/P<>S$18U

-OTT"-

.eo
1111 : II,

[I [

191815191

7()14Z
12rl00I1

I?o

f?T

1111 : II:

I"~

1= II i 1IIIi

i I:
R()

111 1liiT

IF "lTIT:
I
, 11: I r 1 I' rho I rn:

/P"SSI(Jt.E

:1~13I3,

nTn

'nil:

' 11919' 919191919' 11 ~ II InTTF til: 'TIIF TT: 1"1 rT':

.eo

TOIAG'

2A .... ~'...:!

COUNTII

:~II'

III Ii II:

.,. / __ 11 ~Jt:,~

MUU.
QUOt.

'~FS4

1/2

c:o"''''.,..~

TOIAG'

;""'.Ml~!

;;::..

~ iTT:"il

:o,)~'

y
V"
V
V

111

I~ RtR

17
18

/
y

'Aero<

'E#NFr'ED

F/RST

I.;

,pI'R

TTTTT

/PI

12l0'00I1:

1?I

/
/

\31

Ip~

.sOUPRE ROO7

I,.~

1111

rnrl

I/~~

SOUN""E KIM,

11 TI IT

I1i'F
I

('{JU{N,

Figure 11.15 Planning chart. Approximate square root.

However, the square root of 5,625 is developed in three
iterations:
S,%gs + 30 X O.S

= 108 (second approximate)

MACHINE CALCULATING OPERATIONS

5,625

108 + 108 X 0.5 =

5,625

----so +

239

80 (third approximate)

80 X 0.5 = 75 (square root of 5,625)

Note that each iteration is a repetition of the previous step,
with the difference that a new approximate is introduced as the
divisor. Regardless of the number of iterations a base number may
require, the machine always takes four.
The calculation requires 32 program steps, eight to each
iteration. The size of the base number is determined on the read
cycle so that on program 1 the digit 3,30,300, or 3,000 as the first
approximate may be entered into FS 4.
The square root of a number having an odd number of digits
to the right of the decimal point, e.g., 12.136, requires the addition
of a zero to the right for accurate development of the root.

Gangpunching may be done as a separate operation involving only
the Card Read Punch or as part of a regular calculating run. Operations include straight or interspersed gangpunching, minor or
major, and offset numerical or alphabeticaL Digits may be emitted
to punch a constant factor as all cards pass through the machine.

Gangpunching

1. Describe the various kinds of calculation that can be performed on the
604 Electronic Calculating Unit.
2. What is meant by "programmed" calculation?
3. How can calculation be machine checked?
4. How can procedure controls also be applied to calculating operations?
5. What are the three types of stora,ge units in the 604?
6. List five of the specific operations that the 604 can perform (e.g., counter
readout) and describe their functions.
7. What is the purpose of the planning chart and how is it used?
8. Plan the multiplication of A X B = P. A is the unit price field from the
stock room requisition prices as discussed in exercise 7, chapter 10. B is
quantity issued from the same card. P is the total cost of the item.
9. Was the previous operation of gangpunching necessary in order to calculate total requisition cost in exercise 8?
10. Revise your stock room procedure to include the calculation of total
requisition cost. Include appropriate controls.

Questions
and
Exercises

240

MACHINE CALCULATING OPERATIONS

= Q. A is the total cost of an inventory
item that has been issued by requisitions over a period of time. B is the
total number of pieces used. Q is the average cost of the item. The field
A is in dollars and cents; B is a whole number; Q is to be calculated to
the nearest cent.

11. Plan the calculation of A 7 B

REPORT
PREPARATION

In punched-card data processing, printed documents or reports
can be obtained as either intermediate or end results of a procedure.
Many types and arrangements of reports can be produced automatically by various kinds of printing equipment known as accounting
machines. This descriptive term is particularly appropriate because
the machines combine a number of common accounting functions
in a single unit. These include the capability to:

1. List or detail print data from each card in a file. Figure
12.l is an example of how an expense distribution report can be
detail printed. The cards from which the report is prepared have
been sorted first by account number and then by department or
branch number. The amount of each invoice, with date originated,
is listed in account number sequence.
Like other punched-card equipment, accounting machines are
equipped with removable control panels. The selection of data to
appear on a given report can be determined by control panel wiring.
2. Tabulate or summarize the numerical data punched in a
file of cards. For this purpose, the accounting machine has counters
to add, subtract, multiply, and crossfoot totals.
Note that the report in Figure 12.1 is printed with totals
after each account. This means that the machine not only accu241

242

REPORT PREPARATION

mulates amounts but recognizes any change in an identifying field,
such as account number, between one card and another. This capability is used to control a total to print after each group has been
listed. For example, in the report shown, two items are listed for
account 431-11 2, six items for account 431-11 3, and five items for
account 431-114. A total is printed after each group.
EXPENSE DISTRIBUTION
BY DEPARTMENT OR BRANCH
DEPT.
OR
BRANCH

ACCOUNT NO.
GEN.
.SUB.
LEDG.
LEDG.

OUR
DATE
INVOICE
MO. DAY
NUMBER

82
82

431
431

112
112

12066
12153

12
12

10
28

82
82
82
82
82
82

431
431
431
431
431
431

113
113
113
113
113
113

12066
12066
12066
12153
12153
12153

12
12
12
12
12
12

10
10
10
28
28
28

82
82
82
82
82
82

431
431
431
431
431
431

114
114
114
114
114
114

12066
12066
12066
12153
12153
12153

12
12
12
12
12
12

10
10
10
28
28
28

431

520

12149

431

700

12082

431

750

12003

431

810

12112

431

850

12043

432

841

12151

AMOUNT

300 00
300 00
600 00*
150 00
150 00
125 00
150 00
150 00
125 00
850 00*
50 00
75 00
50 00
50 00
50 00
75 00
350 00*
360 43
360 43*
225
225*
100 00
100 00*
70 20
70 20*
24 75
24 75*

AMOUNT BY ACCOUNT

AMOUNT
BY DEPT. OR BRANCH

2357 63
1792 86
1792 86*
1792 86

4150 49

Figure 12.1 Detail printing.

The next five accounts have one item each; therefore item and
account totals are the same. All of these totals are printed whenever there is a change in any of the last three or minor digits of an
account number.

REPORT PREPARATION

243

Now notice that a different class of total appears after a
change in the first three, or intermediate, digits of an account number. The intermediate total is the accumulation of all preceding
minor totals. Also, at the end of the report, a major total is taken
by branch or department number. An accounting machine can
recognize up to three classes of control to print totals: minor, intermediate and major.
EXPENSE DISTRIBUTION
BY DEPARTMENT OR BRANCH
DEPT.
OR
BRANCH

ACCOUNT NO.
GEN.
SUB.
LEDG.
LEDG.

OUR
INVOICE
NUMBER

DATE
MO. DAY

AMOUNT

82
82
82
82
82
82
82
82

431
431
431
431
431
431
431
431

112
113
114
520
700
750
810
850

600 00
850 00
350 00
360 43
225
100 00
70 20
24 75

432

841

1792 86

AMOUNT BY ACCOUNT

AMOUNT
BY DEPT. OR BRANCH

2357 63
1792 86
4150 49

Figure 12.2 Group printing.

Figure 12.2 shows that the same cards can be group printed.
In this report, each individual card is not listed. Detail is omitted
that shows how each total is accumulated. However, each total is
identified with its corresponding account number. This form of
control eliminates the effect of printing duplicate totals for
single items.
3. Punch summarized data into cards when connected to a
summary punch, for example, an IBM 519 Document Originating
Machine. Figure 12.3 shows a summary card as it would be
punched for the first line of the group-printed report. The amount
($300.00) is punched in columns 37 through 44 as an eight-position field. Branch and account numbers are punched in columns 13
through 20. Only the identifying fields and totals can be summary punched.

244 REPORT PREPARATION

Figure 12.3

Summary card.

4. Feed and position continuous paper forms automatically
to the line where the data should be printed. Spacing, form positioning, arrangement of the data on the form, total printing, and
form-to-form ejection can all be controlled by a combination of
panel wiring and the setup of an automatic carriage with punched
paper tape.
Because printing can be done at speeds of over 100 lines per
minute, the blank paper forms are usually fastened together as a
continuous belt for machine feeding. After printing, individual
forms are separated along perforated edges by special equipment.
Many documents produced by accounting machines are byproducts of processing within a procedure. Examples are invoices,
statements, payroll checks, address labels, and so on. Figure 12.4
shows a continuous form invoice with matching paper tape for
automatic carriage control.

Methods of Printing

Printing is essentially a mechanical process, aided somewhat by
the chemistry of ink transfer to paper by direct type impression.
For example, a typewriter is a mechanical printer, even
though the mechanism in some models may be driven by electrical
means. The typewriter is also a serial printer. That is, each character is individually chosen by the typist and printed, one at a time,
in a line. After a character is printed, a carriage moves automatically to the left, thus positioning the paper properly to receive the
next type impression. A different typebar is used for each character.

REPORT PREPARATION

Punched card accounting machines are parallel printers, that
is, an entire line of characters is put on paper in a single cycle.
Complete lines are printed in parallel, one after the other, from
the top of the form to the bottom. The printing mechanism is
made up of a number of typebars or type wheels, each with a full
set of characters. These may be numerical only; or they may be
numerical, alphabetic, and special characters combined. The printing cycle is in two parts: a setup portion that positions the required character on each bar, and a very short impression portion
when all the bars or wheels are pressed against the paper simultaneously. A carriage feeds the paper to the next printing line.
Like the typewriter, an inked ribbon is fed between the type
and the paper to make a legible impression. The main advantage of
parallel printing is speed.
Figure 12.4 Continuous form.

GtlJE
r:;=o

Cl)i\i=OODCD ....

C71UO""'toIN-

i I+++++-+++++-I:i

IN ACCOUNT WITH

0
O!

CUST NO.
7756

A B SMITH &CO
1025 E MAIN ST
DAYTON OHIO

.,"

,-~

22" Maximum

M~·I D~Y

I I
YR.

01
1
0 1
01
1
0 1
1
0 11

DATE
MO. DAY
3
4
4

12
2
10

REFERENCE
21046
28522
5096

CODE

CHARGES

CREDIT

206150
134/62
1
1
1
1
1
1
1
1
1
1

1
1
1
206150
1
1
1
1
1
1
1
1
1
1

BALANCE DUE

134.62

0\
1.
2.
3.
4.

10
1
0
0

CODES
1. CASH
2. RETURN
3. ALLOWANCE

ilOl

ill

co.

ENDICOTT, N.Y .

..'"
..

~
~

:,0

STATEMENT
GENERAL MANUFACTURING

Channel 1 - First Printing Line Stop
Channel 2 - First Body Line Stop
Channel 12 - Overflow Start and Page.Total Control
Channels 3 -11 - Normal Stops

0
0
0
10
10
1
10
1
\0
1
0
1

245

246

REPORT PREPARATION

The following sections describe the basic operation of IBM
accounting· machines.

IBM 402 Accounting
Machine

The IBM 402 Accounting Machine has two kinds of typebars in a
single printing mechanism:
1. As many as 43 alphanumeric. Each typebar can print the
26 letters of the alphabet; the 10 numerical digits; and one special
character, the ampersand (&).
2. As many as 45 numerical. Each typebar prints the 10 numerical digits and one symbol -either an asterisk (*) or a credit
symbol (CR). The asterisk marks a plus total when printed on a
report; the credit symbol marks a minus total or credit balance.

When a 9 is punched
in a card, it causes
thetypebar to be
positioned to either
a 9, I, R or Z
depending upon the
previous detection
of a zone punch
( or the lack of.
zone punch ).

n
Figure 12.5 Numerical type bar.

The two sets of bars are separated by a ribbon guide equivalent to one blank space. This space is an important consideration
of form design.

REPORT PREPARATION

247

As an amusing footnote, it is interesting to consider why the
rather unusual number of 43 alphabetic type positions was chosen
in the design of the machine. The first alphabetic printer marketed
by IBM, the 405, also had the same arrangement of 43 bars. The
reason is that it takes exactly 43 letter and space positions to spell
out the name, International Business Machines Corporation. The
original designers made sure their machine could print the full
name of the parent company on a single line!
Figure 12.5 shows the type portion of a numerical and alphabetical typebar. Special characters and symbols can be specified as
required in place of the standard arrangement. A certain number of
removable bars can also be obtained to give the operator some flexibility in selecting characters. For example, extra-large digits of a
special typeface may be desirable for printing the dollar amount
on checks.

Figure 12.6 is a schematic of the card path through a 402 Accounting Machine. The feed hopper is located at the left of the
machine where 80-column cards are placed by the operator, facedown, 9-edge toward the machine. Cards feed one at a time from
the bottom of the hopper. Thus, when a file is in ascending sequence, cards feed from lower- to higher-order records in the normal sequence for report printing. After passing through the machine, all cards are automatically run into a stacker.
The schematic shows a second and third reading station. A
first reading station, not shown here, is available only on another
model (IBM 403) with the capability of printing multiple lines
from a single card. The stations are fitted with 80 individual
brushes, one for each of the 80 columns of the card. Cards pass
between these wire brushes and a metal roller. Electrical contact
is made whenever a brush touches the roller through a hole in a
card. Rows of punched holes are sensed in parallel, from 9s at the
bottom of the card to 12s at the top.
The two sets of brushes are spaced one after the other in the
card path so that when a row of holes passes the third reading station, the same row in the following card is passing under the second
station. For example, if row 8 is under the second reading station
brushes, the following card is also at the 8 row position. This spacing arrangement is designed to permit a comparison to be made
between data in two cards, an important function in sensing selected card fields for control purposes. Or, a control punch sensed

Card Reading

248

REPORT PREPARATION

at the second reading station can be used to prepare the machine
to read, accumulate, or print selected data from that card when it
reaches the third reading station.

2ND

BRUSHES~

NOTE:
FIRST READING
STATION NOT
AVAI LABLE ON

402 MACHINES.

Figure 12.6 Card path, IBM 402 Accounting Machine.

Movement of cards past the reading stations is also timed precisely with the operation of other electromechanical devices in the
machine: typebars, selectors, counters, and so on. Operation of
the entire machine is in cycles, specific periods of time within
which a given series of functions can be completed.
The 402 cycle is divided into 20 equal parts called points.
Each point is further divided into 18 degrees. Thus, there are a
full 360 degrees in every complete cycle. The time required for
any function can be expressed in degrees. For example, card reading occurs every cycle between 0 and 216 degrees. The remainder
of the cycle to 360 degrees represents the space between cards as
they pass through the machine. Cycles are repeated, one after the
other, as long as the machine is running. All functions are timed
to operate in an exact relationship to each other within a given
cycle. Overall machine capability can be adapted to a wide variety
of procedures by control panel wiring.

REPORT PREPARATION

By control panel wiring, input data to be printed is selected from
punched cards as they are fed past the two reading stations. The
data, in the form of timed electrical pulses, is directed to specific
devices within the machine. The devices are also controlled by
wiring to perform different functions, depending upon tlfe requirements and format of the report to be printed.

Figure 12.7 IBM accounting machine control panels.

Control panels are constructed of special insulating material
held in place by metal frames. Figure 12.7 shows several types of
control panel for IBM punched-card equipment, including the 402.
The 402 panel is made up of three sections, each drilled with as
many as 34 rows of holes. Figure 12.8 is a schematic of this
arrangement. Rows are marked A through HH from top to bottom;
holes are numbered I through 66 from left to right. The labeling
system makes it possible to identify any hole by referring to the
intersection of the row and hole locations. For example, card cycle
holes are located at rows J through Q and numbers 49 and 50. The
various machine functions of the holes are printed on the surface
of the control panel.

249

Control Panel

•

<:>

ro ,r "

, ..

u u n n »

Ili

FIIST IfAOING_I S
000000

o
d

001:'00000000
0

20-0000
10--

00000000000

0000

' 1 - - - - - 5 - - - 2 N D LIN! ZON! ENTlY
0

0'0

» u

00000000000
5 - - - 1 S T LINE 'lINT INUY

00000

1 1 - - - - - 5 - - - I S T LIN! ZON! !NTU

20--

00000000000

1....- - - - 5 - - - .-2ND

20-0

1 1 - - - - - 5 - - - 1 0 N ! S!L!CTION IXIT

n n

00000

I I - - - - - S - - - U A N S I U ION! !NUY

o
6S

000

20--0

LINE PlINT INTlY

20--

00000000000

1...- ' - - - - 5 - - 1 1 D LIN! ZON!' '.'NT INTlY
11-:...-::...-::.......:_~_O_o_o_P~IN~ SE~EC~ONoIX~T 0

000

20-0 0
2°0__
_

o 0 0 0 0 0 0 0 0 0 0 0 0 '0 0 0 0 0 .
_ . _ - - - - S - T l A N S F U AlPHAMUICAl PRINT I N U Y - - - 2 0 - . . ._ _ _ ~~NOOI~AL °ZO~E ;NUOy

00000000000

THI.D UADING-IS

00000000

10--

00000000000

000

o·

60
0

20-0000

:0":":"
000

;..~~--'-5----SICOHO lEADING-IS

000'0000000000000000000

I0
'I

~--;-NOOI~Al:lPoHA'6UI~AlO'·I:T ~NT~Y-O-O-O-~-O-O-

!~N~ME~CA~ '~NTOIN~.Y

nrl!Xf,

2~-0-O-

Itel II20I~ I I I"I I 1II 11601I ~ I I.e1III I
I

o , 0

0 lOT 0

o~o

C~'IS~LIC~Ot1 0 0,' 0
0

CNC

oCO

~c:

12AIII"!II I luI ITrrrF'I I III 1...1III 1
12<111 1:11 I I,cl I III I 111 ~I 1,,1 III I cc
20

F
o

THltO IlAOING-IS
0 ~ ~ 0
0 0 ~

20-~ 0 0

0000000000000000000000

75
0

•

'2'

:..

Figure 12.8 Schematic. IBM 402 Accounting Machine control pane1.

J6Jl,.".oer

REPORT PREPARATION

Each hole is fitted with a metal prong in the back which serves
as a retainer for a plugwire. A plugwire is inserted manually from
the front of the control panel. The wires are insulated, except at
each end, where a small jackplug holds the wire in place. Thus,
holes or hubs in the control panel can be connected together with
the pluggable wiring. The 402 control panel shown in Figure 12.7
has plugwires in place.
The control panel is inserted in the machine by means of a
metal slide that clamps firmly in place. Spring contacts on the machine connect with the metal prongs on the back of the control
panel, thus completing an electrical circuit between the internal
wiring of the machine and the external wiring of the control panel.
For example, when contact is made between a brush and the metal
roller at a card reading station, the resulting impulse is directed by
internal wires to the control panel, from where it is redirected to
some machine device. The device is also connected to the panel.
Figure 12.9 illustrates this arrangement schematically.
Each panel, then, must have two kinds of hub: exits and entries. An exit hub is one which emits an impUlse. Some exits are
under the control of the hole in the card; others may result from
some function previously performed, or may be automatic for
every card. An entry hub is one which can accept an impulse wired
to it, for example, a counter entry. A connection must always be
made from an exit to an entry by placing one end of a plugwire in
the exit hub and the other end in the entry hub. Which exits and
entries are used will depend entirely upon what job a machine is
called upon to do.
Whenever two or more hubs are connected together by lines
(Figure 12.10), these hubs are common, that is, two or more exits
and· entries serve the same purpose. An arrow between two hubs
identifies them as a switch that can be turned on by connecting
the two hubs. A curved arrow indicates the switch can be connected through a selector.
Control panel wiring is changed to prepare each type of report,
thereby giving one machine the general purpose flexibility to produce a number of different types of documents or reports for
many different applications.
Once wired, a panel can be stored indefinitely. This means
that whenever a required report is to be printed, the same panel
can be used again and again. A panel can also be wired permanently with a special type of p1ugwire furnished especially for this
purpose. The panel is then fitted with a metal cover for storage

251

252

REPORT PREPARATION

A wire inserted - - - - - - - - -_ __.

;ntoohUb~

which in turn
0....-0

machine

that is connected

its impulse di

ted to control another
internal
device {punch die}

r
Figure 12.10 Control panel common hubs.

Figure 12.9 Schematic. Control panel wiring to internal devices.

(see Figure 12.7). Thus, any number of panels can be kept on
hand to correspond with the variety of report printing requirements called for by a particular procedure.

Control Panel Wiring

Any detailed presentation of control panel wiring is considerably
beyond the intended scope of this text. Complete information on

REPORT PREPARATION

this subject is readily available in reference manuals prepared and
distributed by the manufacturers of punched-card equipment.
However, a few examples are presented in the following sections to
illustrate some basic operational principles of these extremely versatile machines. These principles are more or less common to all
IBM punched-card machines having control panels. The examples
are intended to provide background for use in the design and study
of data processing procedures.
The 402 Accounting Machine is chosen for these examples
because it is about the simplest model of this type of equipment
now available. Later sections include descriptions of some of the
features of the more advanced 407.
As previously stated, the 402 can be equipped with as many
as 88 typebars - 43 alphamerical and 45 numerical. Each alphamerical typebar has all letters of the alphabet, all digits 0 through
9, and one special character position with an ampersand (&). Alphamerical typebars are positioned to the left side of the print area.
Each numerical typebar has 10 numerals and one symbol. In oddnumbered typebars the symbol is an asterisk (*), and in even-numbered typebars it is a credit symbol (CR). From this arrangement,
it follows that alphabetic information can be printed only on the
left side of the report, while numerical information can be placed
in any location. Normally, this restriction is of little consequence
since conventional report design places the name or description to
the left side with quantities and amounts to the right.
As an aid to good form design, it is often helpful to use a
ruled layout sheet. An example is shown in Figure 12.11. All
accounting machine typebar positions are spaced and numbered,
just as they are positioned in the machine. Sheets can be obtained
full siz,e, so that a finished layout will be actual size. The completed sheet is also an aid to control panel wiring, and can be used
for ordering forms of proper size and specifications from the forms
manufacturer.
Figure 12.12 shows a punched-card source document from
which a cash requirements statement is to be run. From the layout
in the preceding illustration, it can be seen that the form has been
designed to print in alphamerical typebars 24 through 43 and numerical typebars 1 through 27. Vendor abbreviation must print
from card columns 7 through 17 into alphamerical typebars 24
through 34. Vendor number must print from columns 18 through
22 into typebars 35 through 39. Invoice amount is printed from

253

254

2 1 •

6 1 •

REPORT PREPARATION

SPACING CHART

') 10111213 14 IS 161718192021222324 2S 26272129 30 11 32 31 34 3S 3637 38 39 40 4142 U 44 I

: ALPHAM[RICAL

TYPE

;~ ~~::

BARS

:::

5 6 7 8 910 II 12 U U 1516 ,. ,. ,. w~nnUnnVUH~n»"U»HnMHW~UQ«.

3 •

::~::~:

NUMERICAL TYPE .AAII

)

GENERAL MANUfACTURING COMPANY

H
CAi REQU IREM ENTS ST
VENDOR

VENDOR
ABBF-SVIATIQN

ATE~.~~~

............................1.

DUE·

I~~
:

2 3 4

7 •

10 II 12 13 14 151611 I' 19 20 2122 23 24 25 2S 2728 29 30 3112

35 36 31 31 J9 40 41 42

43'·~

1 2 ,

7 •

9 10111213 U 15 16 1718192021:l2 23 2425 26272. 2t 3031323334 3S 3631 JI 3t 40 41.20 ..

Figure 12.11 Accounting machine form layout sheet.

K;Z 3IK:~~bA B B~ ~
11m "II "'.

~[I'n

(I'."
un 1111

111110' ...."

.URDU .... lV.

ASSIO 117~1 2 120\3012 11\: h3 1100311 aloo:l 5S!7SI00: 155:6 °1

VIIDO.

It......

.u.

I ... "'. "'. 11I"'IClI.lI"jlm ,.111111l 'lIIl OlSrDU", I ' 'DlCI
IDATI
DISCOUNT
PAID 11j:~[

IIV. II· cu.
I'.
"' llOC

••,.1'

INVOICI
AIiQUMl

••OUlf ,. PI,

AMDUMt
10 'A.

DUAIl

0,000,0000 a8]""0 .. DD[ (JG-O DOD 00 a( 0 [ 0 0000 GO 0 8 0 000000 D0 0 OD,D 0 0,0 OCCO,D 0 0 01~~.0 0 0.0 DIU UD 0 0,0 L0 0 00
51
':' I ~I'_[". """ "" ... n .... '.,11 IJU'If"""''''lr~UU''''''U''''14,.U ..
leu[
' 'III~Utol"n'~1I '''1:'''''''''' III" ...

"'w.

t''''''''":lUi

.1'1' 1'1 Dll' " 110111 DDt I DI DIll II ~ " " " " 1 t I " till tp ,;, t 111011 1 1 ,;DI I:lll"CII;111 1 'I
lI'ea, .:DIY . .
",DlY ",ea, I
I
I
I
,
~ ~D2 DiD' 2 n02 2 202 i D[ 2 2 2222 D2 D2 22 D2 2 2' 2 2 Z2 2 2 2 2 2 Z2 2 2 2 2 ~2 2 bz 2 222:222222:22 N 2 2 2:2 2 2:2 22222

.,.'

•

,I:

: 3:3 [31333333 D3 S 3 3 3 33333333333 D3 3 3 3 3 3 3 3 3 3333333333333103 31113 3 3 III 3 3 3 3 3:3 3 3:3 3 3 N 3 3'33 3 3 3 3

~ 4:444:44 444 4 4 U
0'

4 H 4 4 4 4 4 4 4 4 4 4 4 444 4 4 4 4 4 4 4 4 4 4 4 444 4 4 4 444:4 4 4:4 4 44 4:4 4 4 4 4 4:4 4 4:4 4 4 4;4 4 ':4 4 4 4 44

I·

5:555'55555555555555555555555555555555555555555 IS 5 5 5:5 5 515 5555' 55555 5:S 05'5 5 5 5:5 DI15 5 5 5 5 5
,
I
I
I
,I
I
I

.:' • ':1' II ,
,

n, , , •• , , 1111' , • , 1 , , , &I , '11'1 •• &I &n &••• II:. &&:•• &I ':' I , &• ,:. , .:, && ':1 •,:~& , II &
I

I ' , '

1:111,1111111111 J 1 1 J 7 7 D7ll 7 7 1117177111777777777 J 7111:71 111 1 1 J ':1 J 1 7 7 1:7 1 1:01 1 1:1 7 7:11 J 71 7
, ' .
'
I ' d
1.1
.'I•• ".1
•• 8 ••••• I I •• I ••••••••••••••••• 1 •••••••• I ••••,
'••• :•••••
:•• '.1 I I'I' I•••• ,'. • .'.
I
I
I • • • ••

r, , •,

I:.

':1 • ,:. I ••• , • , • D' 11II11
I 9 9 III , I , 1111 '9 I • 19 " 9 I 9:9 9 1:1 •• I 9:111111:11 ':' , I ,:, 9 9 ,. II
'"~ I,~:,' I • • • n 1111 . . . . ~ ."1t1U ~1I.,nu 111131111 IU4Is.,nU"... 1I4UUU' .... ' bi:illln'S4$l$U'M"ii Iliilll_."... 11 ltll,l/lU..,Sii 111111.

Figure 12.12 Accounts payable source card.

REPORT PREPARATION

255

columns 63 through 69 into numerical typebars 7 through 13, and
so on. Notice that the space between 43 and 1 represents the ribbon guide to divide the alphamerical typebars from the numerical
typebars.

CASH REQUIREMENTS STATEMENT
VENDOR
ABBREVIATION

ABBOT
A· B BOT

BRASS

VENDOR
NUMBER

DUE

.~~

DAI ••

I--=--MO.! DAY

11 79112[3 1

8Bt\SS 117912:31
COAL 118012[31
A8F~AMS
ABHAMS COAL 1180 l2!3 1
J 07612131
8ARH MACH
EL THUST CO ~952112131
KARTAGE INC 4486012131
LEH I G H COAL 48678 12\31

58091 12131
REF
I L T SUP? 60035 121.31
GAS EL 61221 12131
74213 12131
STATE N. Y
81469 121.31
COR TEL
VI
W ICKW IRE B R 8 63 41121.3 1
INC :8 8 2 1 3,1 2: 3 1
Vi ISELO

MA IZr:

N
N

t,i
Y

DISCOUNT

3) 8
1:9 6
8:3 1
10:50

INVOICE

AMOUNT

TO PAY

1 58:78
98[13
277)5
3 00:00
15 0:77 301527
51j25
2 18!7 5
691178
1182.3
214115
675:95
1792186
2 3 712 9
360143
1 95118

TOTAL

BY DATE

155[60
9617
269:04
r;89:50

2864;50
5125
218175
677:94
118~.3

21/1-~5

675PS
1792186
2 3 7~ 9
360143
195~ 8
i

Figure 12.13 Cash requirements statements.

The report is to be detail printed. This means that data from
all cards in the file will be listed on the finished report as shown in
Figure 12.13. The required panel wiring is described next.

When detail printing, the operation of the machine is controlled in
such a way that the typebars are actuated to print during every read
cycle. Two sets of control panel hubs are available to do this (refer
to Figure 12.8).
All Cycles

These hubs, located in HH, 41-50, emit impulses that can be
used to control many functions of the machine. The emitted impulses are available every machine cycle.

Detail Printing

256

REPORT PREPARATION

List

When the list hubs are impulsed from exit hubs, all 88 typebars of the machine rise as each card passes the third reading
station. If there are holes in the card and the control panel is
properly wired, the typebars print every time they rise. Detail
printing thus occurs at a rate of 50 or 150 lines per minute, depending upon the model. Spacing of the paper by the carriage is
automatic before each line prints. If every card is to be printed,
the. list hubs are wired directly from all cycles. The list hubs are
located at Z, 41-42.
Normal Alphamerical Print Entry

Each alphamerical typebar has a corresponding normal print
entry hub on the control panel. These hubs are located in row Q,
1-43. When the hubs are wired from the third reading brushes, the
typebars print the numerical information punched in the corresponding columns of the card passing that station. Therefore, to
print vendor number on the cash requirements statement, card
columns 18 through 22 are wired from third reading to alphamerica1 normal print entry bars 35-39. A diagram of this wiring
is shown in Figure 12.14.

--25--.,
o

--25--0

000

. Figure 12.14 Detail printing. Numerical.

REPORT PREPARATION

In detail printing, the upward movement of all typebars is
synchronized with card reading. The bars move up while a card is
being read from the 9 to the 12 row. As soon as a brush senses a
punched hole, the typebar is latched at that position. For example,
if a 4 is read, the typebar wired to that column stops moving upward when the digit 4 is in the printing position. The typebar is
held in this position during the remainder of the reading cycle.
After the brushes have read all of the punching rows 9 through 12,
all typebars will have been properly positioned for printing. Each
typebar has a hammer which, when fired, pushes the printing type
against the carriage platen. All the hammers are fired at one time
to print an entire line.

Third Reading

The third reading brushes are used for all normal reading operations. Impulses from these exit hubs must be directed either to
normal alphamerical or numerical print entry hubs.
There are two sets of third reading brush hubs on the control
panel. One set is located in rows O-P, 1-40, and the other is located
in rows DD-EE, holes 1-40. The corresponding hubs in both sets
are common and can be wired interchangeably. The two sets are
needed when it is necessary to wire the same card columns into the
typebars for listing and at the same time accumulate the amounts
being printed in counters.

Numerical Print Entry

Each numerical typebar also has a corresponding print entry
hub on the control panel. The hubs are located in row R, holes
1-44. The hub for print entry 45 is located just below 44. When
a numerical print entry hub is impulsed from a third reading brush
hub, the typebar prints the numerical information punched in the
corresponding column of the card. To print the amount to pay on
the cash requirements statement, third reading hubs 70-76 are
wired to numerical print entry hubs 14-20. Wiring for the other
listed items of numerical information is done in the same manner.

257

258

REPORT PREPARATION

Second Reading

The 80 second reading hubs are outlets for another set of 80
brushes located just above the third reading brushes. These hubs
are located in M-N, 1-40. Like the third reading brushes, they read
all 80 columns of the card. However, they are used primarily on
the 402 to prepare the machine to print letters instead of numbers.
They may also be used to perform other specific functions which
will not be covered here.
Normal Zone Entry

To print an alphabetic character, the typebar must receive a
zone impulse on one cycle and the corresponding digit impulse on
the next cycle. Zone punches are read from the second reading
brushes and are accepted by any of the 43 normal zone entry hubs.
A zone punch positions the typebar one cycle in advance so that
the proper character will print on the following cycle when the
digit punch is read.
· , · - - - S - - - - ' I I N T UlICTlON
o 0 0 0 000 0 0
0 0 0

~
I

'XIT
0

20--0

/:--0-0-0-5-TlANS'U AL'"AMltICAL 'tiNT I N U Y - - - 2 0 - - -

000

5-:---~OI:AL010:' ~NT';:

000

COM'AIING 'NTIY
0 0
0 000 0 0 0 0 0 0 0
0 0 0 0 0
0 0
2 ' _ U A N S I U AL'HAMUICAL nlNT I N T " _ « 0 _ « 3 , D I ,
0 0 0
0 0 0 0 0 0 0 000 0 0 000 0 0

~-:-:-

.~_.12S

20---·

-;-;-25

NOIMAL 10NI

~NT~'-:-:-:-~~~ ~Cl

.os~0--':"~5IeOND
so 0
0
0 0 0 0 0

5
.os

000

:-o-NoOl:ALoALPc:tAI~UI~Alo"':T~N 0

S---NUMllICAL 'liNT 'NTlY
---25
NUMIlICAL 'liNT INUY - - - ' 0
0
0 0 0 0 0
0 0 0 ..:.0...:.0....::~:....::....::....::....:.:....:::r"i"===:::::::....::0....::0....::0:....:0:.-0 0 0 0 0 0
0 0 0

00· 0-- 0

0
0

IIADING-15
0

0
0

12 AI IluI II I luI I

SO
0

0 0 0
0 0
0 0 0 0
- - - - - l O - T " ' I D IIADING
0 0
0 0
0 0 0 0

TMIID IIADING~IS
0

~~U~ON! I!AD~NGO

70
000

000

10
COL S'LIf~
0 oxo 010
.0
0'

75
0

oeo

0
0

010
0
0

0
0

010
0
0

0
0

010'0
0 0"0

oeo

TOO

ONO

oeo

~
o

•

0
0

•

0 0 0
,$
000

'MilD IIADING-IS
0 0 0 0 0 0

50
0

000

20--0 0 0

---25
o 0

.0
0

000

•

I, II .. IS

,. "

lO-THIID 1 1 4 D I N G - - - 0 0 0 0 0 000

70
0

"
5

10 21 21

Figure 12.15 Detail printing. Alphabetic.

75
0

tOO

oeo

12AI IluI II I luI IcirrC'i:·1 I III luI III I A: ..I III I I luI I (i~rrrr 1"1 I I
I2eI 1I2DI:1I I 1,,1 I II I I ItDI I ~ I 1,,1 III I :: .01 III I I I·cI I I III I I IIDI I I
o

C
0

rrrrY. I I III I·AIIII 1 s ..1III I I IIAI I I l;rrVI 1..1I I IIS~~T;:
u

010
0 0

10
0

I2CIIIuI:1I 11,,1 I III I I·DI I ~1 I·cI III I .01 I11 I I Ilcl I I I 11 1 I IIDI I I
~
I
I c~-src~o.~ I
~
~
~
~
0
0

rl~~;:71

- - o - - U - - - N O I M A L AL'HAMUICALo'l~NT ~NT~Y-;-~

~-";""-"':""",,!,,----CO-SILICTOlS--~"';"~--~--

OTO
ONO

o· 0

·0

REPORT PREPARATION

259

The cash requirements statement shown in Figure 12.13 has
one column of alphabetical information headed Vendor Abbreviation. The item is printed in typebar positions 24 through 34.
Therefore, card columns 7 through 17 are wired from second reading to normal zone entry hubs 24 through 34; card columns 7
through 17 are also wired from third reading to normal alphamerical print entry hubs 24 through 34. The typebars receive the zone
signals when the cards pass the second reading station; the proper
digit signals when the cards pass the third reading station. The
wiring diagram is shown in Figure 12.15.
This example further illustrates the flexibility of report format that can be produced by the 402 Accounting Machine. Data
read from cards can be printed in almost any arrangement, provided
alphabetic information is kept within the 43 positions to the left
of the printing mechanism. Notice that printing speed remains
constant regardless of the amount of information contained in the
report. However, the time between printing cycles may be extended
slightly by paper form positioning and ejection, depending upon
the length of the form and the position of heading and total lines.

To perform the arithmetic operations of addition and subtraction,
accounting machines are equipped with counters. Each counter is
made up of single-position wheels or accumulators grouped into
units of 2, 4, 6, and 8 positions each. The number of counters
available depends upon the model. A 402 can be equipped with
as many as four 2-position counters, four 4-position, four 6-position, and four 8-position counters for a total of 80 positions.
The result is one of having available a number of individual adding
machines that can be used to accumulate totals.
Each counter in the machine is identified by a number and a
letter. The number indicates how many positions the counter has
and the letters A, B, C, and D identify the specific counter of the
group. Two-position counters are labeled 2A, 2B, 2C, and 2D;
4-position counters are labeled 4A, 4B, 4C, and 4D.
Counters may be coupled together in any desired combination.
An 8-position may be coupled with a 2-position to form a 10position counter. Or, a 6-position and a 4-position may be coupled
to form a 10-position counter. Two 4-position counters may be

Accumulating Totals

260

REPORT PREPARATION

coupled to form one 8-position counter. Thus, the size of the
counter may be adjusted to the expected number of digits in a
report total.
To accumulate totals, four basic wiring steps are required:
1. Determine what card fields are to be accumulated. These
are normally amount or quantity fields read at the third reading
station. Card columns are wired to a counter entry.
2. Select the cards to be added or subtracted. Accumulation
can be restricted to specifically designated cards in a file, provided
these cards are punched with a unique identifying hole. For example, when a file contains both master and detail cards, all cards
pass the reading stations. But the counters can be controlled to accumulate only from details and not from masters. The master cards
are ignored by wiring proper exit hubs to counter control hubs.
Also, by sensing .selected control punches, one class of card
can be controlled to add, while a second class of card in the same
file can be controlled to subtract. This capability is required in
nearly all types of accounting applications. For example, in an
accounts payable file, purchases add while payments to vendors
subtract. In a payroll file, employee earnings add; deductions
subtract.
The selected impulses are wired to a counter control hub.
3. Determine when a total is to be printed. There are four
classes of totals that can be printed on a. report: minor, intermediate, major, and final. These levels suffice for nearly all quantity or amount totals required in commercial accounting practices.
A minor total is printed at the lowest level of accumulation,
e.g., by item, employee, or account number. The intermediate total
is an accumulation of two or more minor totals: all items for a
customer, all employees in a department, all accounts in a ledger.
The major total represents an accumulation of all minor and intermediate totals: all customers in a territory, all departments in a
factory, all ledgers in a closing statement.
By control panel wiring, counters in the accounting machine
can be controlled to print at any of the above levels. The counter
normally resets as printing occurs.
The final or grand total is the end· accumulation of all other
totals, that is, the final accounts receivable total, final complete
payroll, and so on.

REPORT PREPARATION

261

4. Determine where a total is to be printed. Counter exit
hubs are available to cause typebars to print during the total cycle.
Totals may be printed in any typebar positions. However, it is
usually desirable for totals to appear in the same typebar position
as the listed item quantities or amounts being accumulated.
Figure 12.16 shows a cash disbursements report to be listed
from the accounts payable card form also shown in the illustration.
K;2
/

11m

3rK:~ ~ ~ B B~ ~

""I''''

I()'''''ID ,.,.
un

1"[

"'"

8RA

5510117,1 2 , 2 °13 °12 111: IK:3 '10 031 ' 8100: ' 5 a: 7 800: 1 5 5: 6 °

111001 .... "

nlDOI "II".

It.... II

mool
.0

I "'

II'.

II.

1111"1'11. 1I"!11II ""lilt '''[i

o'srO.1f

IQAT! 1(; Out

lUI In II· ClI.
NO.
II' LEDe

PliO

...1

'.0'" " PI,

'''II([ ••••• ,

InolCE

DlSCOUMT

0111

.MaUNT

AMOUIT
10 Pi,

OllAll

0,0 0 0,0 0 000. U0 • 8 0 (][:II 0·0 D0 0 0 a0 ( a ( 0 a a 00 aDOD 0 II 0 000 0 0 0 0 a 0,0 0 0,0 0IU LO,O 0 0 0L~~.O 0 0,0 011J ~O 0 0,0 lOa 0 0
':' 1'1' : I •• ,," """" .111111,.,,, u.nn,"UI ""1 Huuunu"••"ru ....... , ...1H:.UlFlI6.,[I'I " W,SI II 1111 11"'11I11~'" It ":11 1111"11 "'1'111
,1oJ "'I_~ Dill I" 1 0111 ODI 1 DI 0111"
1 I 1 I 1 1111 " ' " I 1 II:!J 1;1 lIIIDlllll:Dll:1 III'C, ';111111

.'DlY I:m
.
",DU ..,au'
I
I
:
I
~D 1lJiD 1 2 Dn2 22 DZ2 D[ 2 ZZZZznznZ zznZ Z2 2 Z2 Z2 2 ZZZ2 Z1 Z2 Z2 ~2 Ztz Z2 2 2:2 2 2 2 2 2:2 2 2:2 2 22:2 2 2:2 2 2 2 2 2
I

I t '

3:3 [ 3'3 3 3 I 3 3 D3 S II 3 3 3 33 3 33 3 3 3 3 D3 3 3 3 3 I 3 II 3 3 I 3 333 3 3 3 3 3 3,[]3 31113 3 3 []3 3 3 3 3 3:3 3 3:3 3 3 N 3 3'3 3 3 3 3 3
I

.:4 •• ;•••• 4 4 • U 4 ... 4 4 •• 4 •• 4 4 •• 4 •• 4 4 • 4 4 4 •• 4 •••••• 4 •• 4 .:••• :. 4 4 •• :4 4 4 4 4 .:4 • 4:. 4 4 4;4 4 4:4 4 4 4 4 4
,

I
1

5:555'55555555555555555555555555555555555555555555 5 5:5 5 5 55555'555555:' OS'S 5 5 5:5 DI15 5 5 5 5 5
,
I
I
I
,I
I
I
I:' • ':111 , ,
1111111111 • 1111' II & , III , I , II , II II .11111:11 &:. III &: •• & & &':1 &.:. & & &:1 &&:1]& III &
,

':' 11t11 1 11 11 11 11 J 1 17 J 11 11 J 11 , 11 ,
I

JJ J , 11 JJ , JJ 77 11 1 11:' , ,,1 1 1 11:'11 , , ,:",:IIJJ ':"

I ' ,

'.III

1:1 11 11 1
I

.',•• III •••••••• I • I II •••••••• I ••• I ••• I • I I III • II •• II ••• ,'•• ',111 II.:.' ,• ... 11'11
U."• H' II"I ... II
'I'

S:.',I t',4,,::.'
I:" I"'"
t n".
9t ~j ""'99"'"
9•••• 9""", 9'9:9' ,:,ItS"9""'",."Ij
,:," I' .:" I:' 9•':19 ':9 9""
,
IJ""
Ii i 9
n.,uII It, iJelll
1 I . n II

.IIJlIII

'I~ In_I.."'''

UHI.II.H .....,.U.., .. ., ••• SIotll

It"'" II' ..",.,,,.,..

CASH DISBURSEMENTS
VENDOR

ABBREVIATION

CHECK NO.
HO.

MO.

10.,

1179 12 31
ABBOT BRASS
1179 12 31
ABBOT BRASS
1180 12 31
ABRAMS COAL
1180 12 31
ABRAMS COAL
3076 ·12 31
BARR MACti
EL TRUST CO 29521 12 31
KART AGE INC 44860 12 31

PHI LA
SAN FRAN

93025
93031

CREDIT

DEBIT
DA T£

YUtDOR

12[1
12

ACCOUNTS PAYABLE

15ai78
98113
277135
300:00
3015127
51125
21 ai7 5

3118
1196
8i31
10150
15 Oj7 7

141:90
178\64
1168 9i6 7

•
·1

DISCOUNT

CASH

15 516 0
96117
269104
28 9i5 0
2864150
51125
218175
141190
17 8i6
4
I

188:56

11501111
j

Figure 12.16 Accumulating a final total.

262

REPORT PREPARATION

Columns 70 through 76 are to be accumulated and printed as a
final total after all cards have been listed. For purposes of illustration, it is assumed that no other classes of totals are required. Figure 12.1 7 is a schematic of the control panel wiring needed to list
and total the amount to pay field.

6.
0

-20--

o
o

60
0

,....

-20--

-I 111 1

(I III I

0:1:::
o

oeo

0·0000000
000
_ _. ._ _
IS 0
70

2$ 26

15 16

31 "

39 40 "

., 4.·

Figure 12.17 Panel wiring. Final total.

1. The amount to pay field is wired fro·m the third reading
brushes (EE, 70-76) to counter entry 8D (UV, 33-40). The lower
set of third reading hubs is used since it is nearest the counter
entry . Note that each counter in the machine has a corresponding
set of counter entry hubs on the control panel.
The size of the counter used must be predetermined so that
enough positions will be available to print the anticipated number
of digits in the total. This can only be done by knowing the appli..,
cation for which the report is to be printed. In this case, any 8position counter or any combination of counters can be used that
total eight positions, such as 6B and 2B or 4A and 4B.

2. Counter 8D is impulsed to add from a card cycles hub.
These hubs emit an impulse as each card passes the third reading
station. In this example, it is assumed that all cards are to be added.
3. The amount to pay (cash) is detail printed by wiring from
the 8D' exit hubs to the numerical print entry, positions 19 through

REPORT PREPARATION

263

26. The same printing could be accomplished by wiring from the
third reading brushes (P, 70-76) to numerical print entry hubs 19-26.
4. The total accumulated in 8D is printed at the end of the
run. The counter is cleared or reset to zero by wiring a final total
exit (BB, 41-44) to 8D total (HH, 66). Each counter in the
machine has a total control hub on the panel (GG, HH, 51-65).
The final total exit hub emits an impulse that is available
when the card hopper is empty, the machine is idling, and both the
final total and start keys are pressed simultaneously by the operator.
Thus, a final total can· only be taken intentionally after all cards
have passed through the machine. In this example, manual spacing
of the carriage would be necessary before printing the final total.
5. All cycles, wired to list, cause the machine to print every
card.

An accounting machine can be set up to print totals automatically
at the minor, intermediate, and major levels by proper control
panel wiring of a comparing unit. These hubs are located on the
402 panel at G-J, 25-44 as shown in Figure 12.18. The comparing
unit is divided into individual positions with each position available to compare one card column. A position has two comparing
entry hubs and two comparing exit hubs. The total number of positions is dependent upon the machine model but at least 10 are
furnished as standard equipment on all models.

Program Control

----------------------COMPARINGENTRY------------------------~

000

COMPARING EXIT

~,'\,'\,,'\,,'\,'\,'\,'\,'\,'\,'\,,'\,~'\,~
~LC PROG. START
o

000

COMPARING ENTRY

000

Figure 12.18 Comparing unit.

From the second and third sets of reading brushes, a field can
be read from two successive cards at the same time. When wired to
the comparing unit, any given field in a card passing the second
brushes can be compared with the same field in a card passing the
third brushes.

000

264

REPORT PREPARATION

Comparison is made by wiring the field from the second reading hubs to one row of comparing entries, and from the third

d
/1
..

DI..r~
OR

ACCOUNT No.

.....

.ua.

UN.

LIaS.

41
41

913
913

660
660

913

700

IIIANe"

OUR
INVOICE
NUM8£1t

EXPENSE DISTRIBUTION
BY DEPARTMENT OR BRANCH

.ATE
MO.
DAY

AMOUNT

12
12

07
14

687150
721;92
140 9i4-2 •

12125 12

675:95
67 Si9 5.

12042
12084

;

915

760 12086 12

2119:50
211 9ls O.

·
i

.. J

913

730

12171

31.

913

740

12164

611:93
611193.

913

750

12089 12

200:00
200:00.

47i40
47:40.

··

AMOUNT BY ACCOUNT
I

!
i

AMOUNT
IT HPJ. OR lRANCH

208 5jJ '/
I

2119150

·
:

;

4204i87

i
··
i

Figure 12.19 Expense distribution report and Accounts Payable distribution card.

REPORT PREPARATION

reading hubs to the other row of comparing entries. Either row of
comparing entries may be wired from either set of brushes and any
positions may be used. However, the field must be wired to the
same positions from each set of brushes. For example, if columns
10 through 15 are wired from second reading to comparing positions G, 30-35, columns 10 through 15 must also be wired from
third reading to comparing positions J, 30-35.
If the fields being compared are identical, no impulses are
available at the comparing exit hubs. When the punching in one
card does not compare with the punching in the preceding card,
impulses are available at these hubs. A comparing exit hub is wired
to a program start hub (FH, 45), either to minor, intermediate, or
major, depending upon which level total is to be printed. Since it
cannot be predetermined which column of the compared field will
cause an exit impulse, the comparing exit hubs are connected together with a single exit wired from either the left or right end
column of the field. The following example explains numerical
program control wiring in greater detail.
Figure 12.19 shows an expense distribution report detail
printed from a payables distribution card. Three levels of totals are
taken on the item amount field punched in columns 63 through 69.
These three levels are:
1. Minor total by subledger. The first minor total, $1,409.42,
is printed from one counter when the 402 comparing unit senses a
change from 660 to 700 in the subledger number.
2. Intermediate total by general ledger number. The first
intermediate total, $2,085.37, is printed from a second counter
when the general ledger number changes from 913 to 915.
3. Major total by department or branch. The first major total,
$4,204.87, is printed from a third counter when the department
number changes from 41 to 43.
Program Start

There are three program start hubs (F-H, 45) labeled minor,
intermediate, and major. These hubs receive comparing exit impulses to stop card feeding and start a series of program total cycles:
one program for minor, two for intermediate, and three for major.
If intermediate program start is wired alone, a minor total cycle is
forced before the intermediate total cycle. If major program start

265

266

REPORT PREPARATION

is wired alone, both a minor and an intermediate total cycle are
forced before the major total cycle.
Total Program

Each program level has seven exit hubs which emit all cycle
impulses whenever the corresponding program start hub is impulsed
(CC-FF, 44-50). Program start must be impulsed before these hubs
become active. Minor program exits emit impulses when the minor
program start is impulsed. Minor and intermediate program exits
emit impulses when the intermediate program start is impulsed.
Minor, intermediate, and major program exits emit impulses only
when the major program start is impulsed.
Counters read out and reset automatically when a total program is wired to a counter total entry. After total printing the machine restarts automatically.
Before the payables distribution file can be fed through the
accounting machine, all cards must be sorted numerically, first on
columns 30 through 35 and then on columns 36 through 38. Figure 12.20 shows the control panel wiring for the expense distribution report.
1. The sub ledger field is wired from columns 33 through 35
of both second and third reading to the comparing entry. Corresponding comparing exits are wired together to make all three positions common. If there is an unequal condition in any of the three
columns of the field, a comparing exit impulse will be available.
This impulse is directed to program start minor. As a result, the machine begins a program cycle for every change in sub ledger number.
2. The general ledger field is wired from columns 30 through
32 of both second and third reading to a comparing entry. The
corresponding comparing exit is wired to program start intermediate to begin a second program cycle for any change in general
ledger number.
3. Department is wired from columns 36 through 38 of both
second and third reading to comparing entry. The comparing exit
is wired to program start major to begin a third program cycle for
every change in department number.
4. Each available counter in the machine has a pair of common
total entry hubs (GG-HH, 51-56). When they are impulsed, the

REPORT PREPARATION

counter will read out to print a total and reset by clearing to zero.
Counter 8A is wired to read out and reset on a minor program,
counter 8B on an intermediate program, and 8C on a major program.
The counter exit and entry wiring is not shown in the diagram.

13 l'

19 30 31 31

30 <0 41

Figure 12.20 Numerical programming.

5. Subledger, general ledger, and department are listed for
each card by wiring the three fields directly to alphamerical
print entry.
6. An asterisk (*) is printed to the right of every minor total
by wiring asterisk symbol hub I to numerical print entry 9. Asterisks may be printed from all odd-numbered numerical typebars.
7. The machine is wired to detail print every card.

267

268

REPORT PREPARATION

Figure 12.21 shows that alphabetic as well as numerical information can be compared and printed. To do this, the second
reading station is used in two ways: to read the zones for printing
and to impulse one side of the comparing entry. Therefore, split
wires are used because there is only one set of second reading hubs
on the panel. As the name implies, a split wire is one with three
jack plugs joined together by lengths of insulated wire. The use of
split wires has the effect of reducing the total number of hubs
needed on the control panel, with a corresponding reduction in
panel size and complexity.

_ _" " ' _ - < 1 . _ , . 0 0 0 0 0 0 0 0 0 0 0 0
000

MEllCAL PlINT I N T l Y _ 4 0 _ . " D I
0

70
75
o 0 0 0 0 0 0
)O-THIIO UAOING

000

•
--25
00

COL $'LlT~

oxo 010
0·'

000

M
H

0
0

000

- - - 2 5 - - - M O l M A L AL'HAMllleAL Pl'NT INTlY-'O
0,

000000000000000

-o-o"'o-o-o·~O"AL 10 INflY --::-:-4:~4; ~Cl
)O-:U~ON! I!AO~N\ 0' 0 0 : I r'~~;~71

---2$
o 0 0

0
0

000

NUMUfCAL,.,
0

INTIY - - - 4 0
0

.-I III I I I·AI I I l~rrl T1..1I I I'$~M9
..I lit I I :"11 I1 I 1luI : :.: ~-;::~:~~

!:: :I: ::

I: :

rc:oI:
! 0 N0
0 0 0 0 0 0 0 oco

oco 0 0

~
~

O•
0,

0'0

Oil

..I II I I I r"l ICl;jrrr luI I I I ~~~~
··1 III I I cl I I III I r 1- 01I I I~fN~LT:TAp :

-:-;-21
o

)O-;'~'.oo II:Df~G

':o~ ~2'~ ~

'!! ~ ~ ~

"IMOA

... 0 4

*OVo",?

"L~t
050

000

Figure 12.21 Alphabetic programming.

" :0

(YeLf

REPORT PREPARATION

The Accounting Machine can be wired for program control for
either detail or group printing. As previously explained, when the
list hub is impulsed from all cycles, the typebars rise as each card
passes through the machine. Consequently, all card information
will be listed that is properly wired to print entry from the reading
hubs. The speed of operation is 50 or 150 lines per minute, depending upon the model.
If the list hub is not impulsed, and program start is not wired,
cards feed through the machine and totals can be accumulated,
but no printing cycles occur. In this case, the machine is essentially an accumulating device or adding machine with final totals
taken manually after all cards have been fed. This type of operation
is often used to balance or establish control totals for an entire
file or deck of cards.
If the list hub is not impulsed, but program start is wired from
comparing entry, machine operation is as follows:
a. The typebars rise for the first card of each control group
to print a group indication. One result of this kind of operation
can be illustrated by referring again to the two reports in Figures
12.1 and 12.2. On the group printed report, note that only the
control fields (department or branch, general ledger, standard ledger) are listed. These numbers are printed during the group indication cycle from the first card of the group. Other detail such as
invoice number and date cannot be listed, except from the first
card. After the group indication cycle, cards continue to feed
without typebar operation until a change in control fields is sensed
by the comparing unit. As cards continue to feed, or tabulate,
totals are accumulated in the various counters as wired on the panel.
b. After the last card of a group passes third reading, card
feeding stops while minor, intermediate, or major totals are printed.
The counters involved are cleared and reset to zero. The form is
spaced or ejected, the first card of the next group is indicated, and
tabulating continues. Note that there is complete flexibility as to
the number of cards that may be contained within any particular
group. Thus, the number of lines shown on the report in Figure
12.2 corresponds to the number of control field changes in the file;
any number of cards could be contained in each group.

269

270

Questions

and
Exercises

REPORT PREPARATION

1. Explain the difference between detail printing and group printing.
2. What is summary punching?
3. What is the relationship between the punched control tape and the form
shown in Figure 12.4?
4. Explain the difference between serial printing and parallel printing.
S. Design a report to show the detailed activity of withdrawals from the
stock room. Use the requisition cards completed in exercise 8, chapter
11. It is assumed the cards are still in item number sequence. Each card
is to be detail printed as follows:
Item number
Quantity used Unit of measure
Description
Unit price
Total costDepartment using

minor total

minor total and final total

6. Can this report be used to check the calculation of Quantity X unit price
= total cost from the previous operation? How?
7. Assume the requisition cards are now sorted by department number.
Design a report to show stationery expense for each department for a
given period as follows:
Department
Item number
Quantity
Unit measure
Description
Unit price
Total cost-

minor total by item, intermediate
total by department, final total

8. Can you demonstrate to the company auditors that this final report includes all items issued to each department for the period? Discuss this
poin t in class.
9. Can additional controls be justified? If so, present your suggestions
in class.
10. Can this report be used to prevent unauthorized use of stationery supplies? Also discuss in class.

BASIC CONTROL PANEL

WIRING

This chapter presents additional information about control
panel wiring for IBM machines. Only elementary principles are
discussed, using the 402 Accounting Machine for examples. The
chapter illustrates typical machine capabilities that can be utilized
in the design and implementation of business data processing systems. It is not intended to provide detailed training in wiring
skills, although the examples presented should be understood in
preparation for technical competence in this area.
If equipment is available for practice and demonstration, at
least some of the examples can be machine run with the card forms
shown and punched with sam pIe data. The wiring diagrams can be
used as a source of information for preparing or planning the stationery expense report designed as an exercise following chapter 12.

Figure 13.1 shows how the machine function of subtraction can be
used for report printing. Negative fields in a sales accounting file
are identified by a distinguishing X punch in column 78; positive
fields are not punched with an X. A salesman commission statement is prepared from the file with invoice number, commodity
number, sales amount, and commission amount detail printed. To
accurately accumulate the total commission amount for each salesman, all commissions for current sales must be added. However,
271

Counter Control

272

BASIC CONTROL PANEL WIRING

(~I I I I'~

SALES ACCOUNTING

DATE

UNIT
COSI

...
f!

con
AIiOUIT

COII'SS'ON DATE
.
AMOUNI 110. DAY

CROSS
PROFIT

'NYO'Cl
HUMin

'S:

'ti=1

=
:1,1

CUSIOln
llAOE ., SAUS
.. IIAN QUANI'TY
NUMIEI SI. CITY CLASS
NO.

SALES ACCOUNTING

ENTRY

~,M!

MO. DAY :!

...

UNIT
COST

con
AMOUNT

CROSS
nOFIT

INVOICE

LOCATION

COI.'SS'ONt...!A,IL INYOICl
AMOUNI MO. DAY MUMlEt

CUSTOIIER I-r-- TRADE .,
HUMin ST. CITY CLASS!

1.1111

•

lUll

QUANTITY

NO.

ITlM

INVO'Cl

AMOUNI~

"" 0 0 0 ODD"D 00D 0 0,1. O~~~ ", '~DD:O 0" O~O 0000010 I I 0011 110' 110" "I~III~~~" I' I .'~';.'I~
D' 10 1~0 11:11111D l:ll1111111111:IIDII~DIDIIIII1DIDDIII110111111111 11111 11'11111 I,ll

Zn 222222 2122 22222,222 222(]2 22 2,22 202 2 202 22 2222222222222222222220222021

13~"

2 20D22 2 2 22122
I

13 33'33'
31' ','0
33 33",
31 ',11"0'"11111111"11
01111[
,
,
I
,

IIIIIIIII~ 11111'33
I
,

44444444414444444,4444
4D,4 4 4 4.414444.44.444444.4
D. 4 D44444444.44.444
,
,

111'11
,

4444444'4'
4444144
,
I

tt.: '. ,"'" ,11,", ':"

5 5 5 5 5 5 5 5 5:55 S 55 S S 5 5 5 5 5 5:5 555 515 5 555 5 555 5 5 ~ 5 5 D5 5 5 S 5555 5 55 5 55 5 555 [ 5 5 5 5 5 5 5 5 S: 5 5 55 S SI 5 5
I

"I' " ,,,,... ,," DO
I '
I

1 7 777 7 77 1~ 7 7 17 7

'.11
, , .. ''II
I " "II
I " " II

"mo , ... II , .. , 0 .,"

I
,
777 7 717D17 7'7 7 77 7 7 77 7 D77 7 ~ 7777 7 77 7 7 77777 [11 77771771 7 7 7t17 1 77 7 1,77

I
'
I
888888.8818888 •• 111.8
I • 8:1
8 8 I .,.
I 1 •• 888.88. I 8. 1.888 ••• D. 88..
,

INVOIft
AMOU

100000'00

COIIODln

S:~~S

.8888888.

11111,11
I

2222 M 2

33333133

44444 144
I

55 5 55:5 5
8 ....:..

17 77 7177
I

""1,11
9999 9199
IU'lI1llllllllo

,
8.:18
•• I I 8,,••

99999999 9'99D999 9U 9 999t9999 9'DD899 9 9 19 19 9ngg 9 999 SlD9 9 99 9 9 9999999999 99919999999199
! "
I. 11t'"11111411I1hl1"'211112IDHZS2lZlIZlIl.I1I2IU411." •• 4041414l~40.4141.101II2!:t~~IIIU.IOIIIIUIUU'
""1011111117411""111"10
IBMII73837.
UClISO , .. UIE U • PAIII' l.m.4n

SA L ES
+NX78

SALESMAN:

MAC Y

-/
/
COMMISSJTATEME~T
6

,
INVOICE NO.

COMMODITY

12176 14202
12176 14702
12176 lql02
12176 63706
99590
99590

RETURNS
+ X78

35106
46106

SALES AMOUNT

DAn

RETURNS AND
ALLOWANCES

COMMISSION AMOUNT

99
365
310
.335

22100
81 00
6885
1
223150
239125
28 '50
52 75

1
12

18 CR
46 CR
1.3*

Figure 13.1 Subtraction for report printing.

amounts must be subtracted for commissions previously paid on
goods which were later returned by customers.
Actual electromechanical operation of counter wheels to do
subtraction need not be explained here, but the use of the pilot
selector is a basic principle of machine control that is common to
all punched-card equipment having control panels.

BASIC CONTROL PANEL WIRING

For example, the 402 is equipped with a number of pilot
selectors, so-named for the guiding function they perform. Pilot
selectors can be used independently or in conjunction with other
selectors on the panel, called coselectors. Pilot selectors are located in rows E-M, holes 51-56, as shown in Figure 13.2.
t I

X~Ujo

~ul 0

o D

PI~OT om~cT~u 0
0

ANt
I PU
0 10 10
0

I010- 50

ONO 10

0 ( 0 10

OTolOIO

1
OTO

ONO

0 ( 0 10 10

X6U~

0 D

COUPLING EXIT
0
0
0
0

10
0

ONO

0(0

OTO

ONO

oeo

I PU
0
0

1S~
oTo

r I
I I
" - - One selector position
Figure 13.2 Pilot selectors.

The action of a selector is the same as that of a switch. Impulses from an exit hub on the control panel can be directed to
either of two paths by opening or closing the switch with a pickup
hub. There are two switches or positions for each selector arranged
vertically on the panel.
The pickup can be actuated by either an X or digit impulse
from a card column sensed at the second reading station. If X
punches are to control the selector, second reading is wired to the
X pickup. If a digit punch in the column is to control the selector,
second reading is wired to the D pickup. The D pickup accepts
any impulse from 9 through 12, while the X pickup accepts only
X or 12 impulses. If a specific digit is to actuate the pickup, and
all other possible punching in the column is to be ignored, then the
digit must also be selected by another device in the machine before
it is wired to the D pickup of the pilot selector.
An immediate pickup is also provided for each selector which
will accept any impulse. This pickup transfers the selector immediately, instead of one cycle later. Also, whenever X or D hubs are
impulsed, the 1 hub emits during the selector transfer for the 101-

273

274

BASIC CONTROL PANEL WIRING

lowing cycle. This impulse can be used to expand the pilot selector
beyond two positions by picking up a coselector. When used in
this manner, the I pickup is called a coupling exit. Thus, any coselector to which it is coupled functions exactly like the pilot
selector.
If the column of the card with the distinguishing X punch is
wired to the X pickup of a pilot selector, one machine cycle later
a card cycles impUlse wired to the C hub of the selector will be
available at the T hub. Or, the card cycles impulse can be wired to
the T hub where it will be available at C after the selector is
transferred.
If there is no X punch in the card, one cycle later a card cycles
impulse wired to the C hub of the selector will be available at the
N side of the selector, or vice versa. Thus, if some cards of a file
contain an X punch and others do not, certain functions of the
machine can be correspondingly actuated or not actuated, whichever is desired. The most common use of the pilot selector is
to control counters. A schematic of this operation is shown in
Figure 13.3.
No-X control card
.--------~::»

Pickup

~=-----........~:.
Normal

N tcounter
plus
C
Card
cycles

...

------~J Pickup

.....- -.....- - - f e )

"""'-_-

T{Co~nter
minus

....I.A->-----aO~-_I@ N

. . .-------~Fe)

c{ Card
cycles

Transferred

Figure 13.3 Schematic. Counter control with pilot selectors.

Assume that cards in a file to be added on a report are No-X
and that cards to be subtracted are punched X in column 78. The
X pickup of a selector is wired to second reading hub 78. Card
cycles is wired to C. When there is no X in column 78, a card
cycles impulse is directed to counter plus. When there is an X in
column 78, card cycles is directed to counter minus. The counter
consequently adds or subtracts as guided by the pilot selector.
Note that the counter can just as easily be controlled to add
X cards and subtract No-X cards. Also, by using two separate
counters, one counter can add No-X cards and the second counter

BASIC CONTROL PANEL WIRING

"'G_'5

24 25

21 ,.. .10

,.c

J9 40

..,

275

43 ....

20_

000

0"0

000

0000

20--

INTlY
000

000

N IIIT-----20-000

0000

20--

INTlY
000

0000

20--

INTlY
000

000

20--

INTlY
000

000

INUY

20--

000

eOUHINGI
0

00000

20--

"NT INTlY

0000000000

20--

NUIT

..:::::~;!...+~...::...

"f"I I I II

JoDI I 1:1 I I·e I II

~ : :1: : ~

Ti:·I

rrlrr r

i H::i~:Y:

11 I I~II I·eI III I
60

NO-1S

20--

('00

00000

IIII

1III luI III I

~--~---C~A."

000

TlANSFER AND S.'.•
0 0
0:0 0
C
0 0
0:0 0

..._1---- TRANSFER AND S.'. I

7 SINVOICE NO.

0000

00000000000000000

0
I

cy~uso_---:-i!I ~1 !I :1 11 :1 ~ ;Y~r

I
8 , 01.

• '~~'O~

35 36

39 40

404

Figure 13.4 Wiring diagram. Subtraction and detail printing.

can add X cards. In this way, a separate total can be accumulated
for commissions on returns or charge-backs as well as on sales.
Figure 13.4 shows the 402 control panel wiring for the report
in Figure 13.5.
1. Commission amount in columns 25 through 29 is wired
from third reading to counter 8A entry.
2. The X in column 78, identifying credit cards, is wired from
second reading station to the X pickup of pilot selector 1. The X
must be read from the second reading station so that the pilot
selector will be transferred by the time the card reaches the third
reading station.
3. A card cycles impulse is wired to the common of the pilot
selector. An X in the card transfers the selector, and the card

50 51

52 53 54

51 .58

.sf

276

BASIC CONTROL PANEL WIRING

5 5 5 5:5 5
6 6 6 6:6 6
1

7 7 7 717 7
I

88 8 8:8

99 9 919 9
j~

ALESMAN:

MACY
R£TUF--o 10'20
..3 3

000000000000000

0'''0-,-0--0-0-1-0-0-~O'5~lI~'OI ~5_0-0-0-0-1-0-0-0-0"'-;0

-;;- 2

-,..

-2.
o
0

1'2

ONO

0
oeo 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 oeo
-14I-------COMPARING I N T R Y - - - - - - - - - - !
0.0

-14

-;-2:.

-2'

COMPARING IXII

~~~~~~~~~~~~~~~~~~~
COMPARING INTRY

_ 2 5 _ _ _ _ __
"

o
65

,)"

:"9

50 51

53 5.

63 64

Figure 13.12 Wiring diagram. Class selection.

Additional
Accounting
Machine Features

A number of additional accounting machine features are available
to increase the flexibility of report printing. The most important
of these are summarized briefly in the following sections.

Crossfooting

As many as three card fields can be crossfooted on a standard 402.
The fields can be read into counters and crossfooted for individual
cards, or totals of the fields can be accumulated from a group of
cards and then crossfooted into one counter. The use of the feature may save one pass of a file through the calculating punch since

BASIC CONTROL PANEL WIRING

287

no field need be punched in the cards for the crossfooted total.
The total only appears on the printed report. The three fields are
entered into three counters as the card passes the third reading
station. The fields may be added or subtracted to obtain the total,
for example A + B - C = T.
Each field is accumulated in a separate counter of appropriate
capacity, like 8A, 8B, or 8e. If the report is detail printed, the
fields are also wired to normal print entry from third reading. All
three counters are connected by wiring their exits together to provide a path for the transfer of totals from one counter to another
during what would normally be a print cycle.
After each card, or at the end of a control group, the machine
is programmed to transfer the minor total into the intermediate
counter and then from the intermediate counter into the major
counter. Printing can be suppressed during the minor and intermediate cycles, if desired.

Progressive totals can be printed by not resetting the counters
during a print cycle. After printing, any total standing in the
counter is retained and additional amounts can be accumulated as
wired from the reading brushes.
On the 402, counter reset can be prevented by impUlsing the
plus and minus hubs of a counter control at the same time that the
total hub is impulsed. This wiring allows the counter wheels to
rotate past zero and to continue rotating to their original position
at the beginning of the print cycle. For example, if a counter contains a digit 6 and is impulsed to clear normally, the counter wheel
turns from 6 to 7 to 8 and on to zero where it stops. However, if
the plus and minus hubs are both impulsed, the stop at zero is suspended and the wheel keeps turning until it reaches 6 again.
Although this feature is designed primarily for printing totals
without clearing the counter, it may also be used for printing any
constant numerical information, such as a date or identification
number. Such information can in effect be stored in the counter
from a master card at the beginning of a run and later read out as
desired. For example, the date or identification number might
be printed at the top or bottom of each new page of a report.
Progressive totals can also be used to consecutively number report pages.

Progressive Total
Printing

288
Special Program

Summary Punching

BASIC CONTROL PANEL WIRING

On the standard 402 there are three types of program starts,
resulting in a maximum of three total program cycles. When the
minor program is impulsed, only one total program cycle is taken.
When the intermediate program start is impulsed, two program
cycles are taken, and when major program start is impulsed, three
program cycles are taken.
When the machine is equipped with a special program device,
any number of total program cycles may be initiated for anyone
program start. For example, four program cycles may be started
by a minor start for the purpose of crossfooting four minor totals
or printing them one underneath the other. As many as five or 10
totals can be crossfooted.

Summary punching is the automatic preparation of one total card
to replace an entire group of cards. The summary card normally
contains the identification of the single group and one or more
accumulated totals for that same group.
The preparation of summary cards reduces card volume for
batch processing, particularly when the ratio of detail cards to
control groups is large. When summarized, any given card file will
be reduced in volume by the average number of details per group.
That is, if there are an average of 10 details per group, the file will be reduced 10 to one. This reduction in card volume produces
corresponding savings in machine processing time thereby also reducing the time between the end of an accounting period and the
production of a finished report.
The preparation of summary cards also is an efficient method
of carrying balances of quantities or amounts forward from one
accounting period to the next. Stock status, year-to-date earnings,
and accounts receivable are common examples of this practice, i.e.,
Old balance ± current transactions = new balance. Previous summary cards from last period's run are always punched with the old
balance. Current transactions are accumulated and their effect on
the old balance is summarized into a balance forward or new balance summary card. A report is printed as a record and to provide
controls over the accuracy of the entire process.
Summary cards are punched by the attachment of a summary punch to the accounting machine by a cable. A number of
units are attachable for this purpose, including the 519 DocumentOriginating Machine described earlier. When the control panels of
both machines are properly wired, the exits of all the counters are

BASIC CONTROL PANEL WIRING

available on the summary punch control panel. Data can be summarized into any fields of the summary card, although at least the
control fields are normally punched in the same location from detail to new summary.
Note that only information standing in counters of the accounting machine can be summary punched. On a standard machine, this means only numerical information since counters are
used to accumulate quantities or amounts. However, machines
can be fitted with an alphabetic summary punching device for the
transfer of group indicated alphabetic data.
Card feeding on the accounting machine is interrupted while
a card is being fed and punched by the summary punch. Provision
is also made for the transfer of credit balances to the summary card
with indicating X punches in selected card columns.
Figure 13.13 is a sample of wiring that shows the relationship
between an accounting machine and a summary punch. The accounting machine wiring has been previously shown for report
printing (Figure 13.5) to produce a salesman commission statement. The following explanation describes the additional wiring required to summarize by salesman number and commission amount.
1. The summary punch switch is turned on. These two hubs
on the accounting machine control panel must be connected to
synchronize the operation of both machines. The hubs may also
be connected through a pilot selector or a coselector to provide
the ability to summary punch only selected groups of cards.
2. A summary card can be punched on any program change
by impulsing the pickup hub of summary punch control. When
a total program exit is wired to the pickup, and the summary
punch switch is on, the summary punch will feed and punch just
before the accounting machine prints the corresponding total.
More than one class of total can be punched in the same run, e.g.,
minor and major. In this case, summary punch control is picked up
from the minor program start. Therefore, a summary card will be
punched for each salesman, because the minor program start exits
an impulse every time a group of cards changes from one salesman
number to another.
3. To identify credit balances in commission amounts, it is
necessary to punch an identifying X in column 80 of the summary
card. To satisfy this requirement, each counter in the accounting
machine has a correspo:qding summary punch X control plus and
X control minus. These hubs are on the accounting machine panel

289

e., 0

::..

••

o 0 !-:-:-:-:-F!15'O Ilo~

0 0 ~~T:AN~nl°7.00
0
0 O· 0 0 0 0 0 0 0

1$ 16

IT.1I 19 .. "

I.J ....

a.D"_.n""u"~9~~.MUM~aM

CO-SIUC1'Oll.U'S

;-;-;-;-;SIt!C'~' ~-O-O-1trl-;-''-ri---

~~ZOO"'~ SI~EC~

_0,"_0_0_O_~_O_"_0_0

:i: ~::O::~I

!I~

o::: !~'?Or'Nl:ZO~~'
, - - - 2 N D UNf '11JrII

00~. i . ~ 03~D ~~:;:~:C~

000000000000

!--;-';AN;n~ "l;""':U~~

_ _ , _ N O I M A L ZaN

O·

~)O

.. 0

~.~~"::N:'~ :~~~I~~~~t~~:!_: ~_:~t~~~~:D~~~_~~~.."_;~_~_~S:~~~~~==~:I~=~-~

I: : : : ~-:-:-:-: ,~.o:. : :' : ': : : ': : ': :O:'AlI~G '.U'HIS.
0

0.00

0.0.,0

oleo

107)0

0.0 0

0 2

~:-;-NoOI~AL oAL'oHA~III~ALo'l':T ~NT~Y-o-o-0-':-0-0, - - - ' - - - N U M I I I C A L nlNT ENTlY

. ".
0

20--

, INT f N T l Y - - ' O - - - - . J "

!~TI{IX~~
II~~"

° ° ° ° °

COUNTIlIXIl

iMCHSTO',,,,

I·cI I1 1:1 I I locI I III I I'_DI-:-1~.:..-I....;..........~
::: ! : : ! : c:",':":oy : :
20

0(0

I: :

I: :--.

_o~o.:..o....,-O_~_....;",,:,_

I'AIlluIII I luI I cnrri:·I I III luI III I ~ ..I III I I
1,clll'Dl~ 11l,el III I I 11 I ~ I
II I : .Dl--25
III I I

'____

THIID IIADING-.S

20--

000000000000000°600000

n
'0
oooooooCoooooo

, "

000000

••• ,. 'fOUIJ.' ...... ",,"JO.a.tI

00000

uunuvH~~n»u~uuvHn~~n~M

a.O"~~.»"UU~~~~.~QM~

Figure 13.13 Wiring diagram. Net baIancesummary punching.

BASIC CONTROL PANEL WIRING

---- ---~

2
3
4
5
6
7
8

" "-

SPLIT COL. CTRL.

9·8~70 6°
I
\

/5°4

3 0 2

/-1-0-0-jl~O
/

T-V.41-44

Figure 13.14 Summary punch X control.

in rows W, X, 51-66. Thus, an amount or quantity field can be
punched with either a plus or minus X, whichever is required.
In the example, the summary punch X control minus of 8A
is wired to summary punch entry 1. When the amount balance
in counter 8A is minus, an impulse is transferred by internal wiring
to the corresponding 11-12 column split hub on the summary
punch, as shown in Figure 13.14. From column split 1 the impulse
is transferred through the common hub to the punch magnets of
the summary punch.
4. Counter exits are provided on the summary punch control
panel for each counter in the accounting machine. To punch commission amount in columns 76 through 80 of the summary card,
counter 8A exits are wired to punch, except the units position.
5. The units position of 8A is wired to column 80 through
the 0-9 position of column split 1. A credit balance summary card
will be identified by an X punch in column 80.
6. Salesman number is wired to counter 4C.
7. The salesman number will be used to identify the summary
card, and hence must add in the counter only once for each salesman. Minor first card is wired to 4C plus, causing the first card of
every group to add.

291

292

BASIC CONTROL PANEL WIRING

8. Counter 4C is cleared on every minor program. Printing
of the group indication is suppressed in the total cycle.
9. Counter 4C exits on the summary punch are wired to
punch columns 54 through 56.

Multiple Line
Printing

It is often a decided advantage while handling many name and address files to be able to print at least three lines from a single card.
Then, one card can contain all the elements of a normal address:
name, street and number, city, state, and zip code as shown in
Figure 13.15. The IBM 403 Accounting Machine is designed for
this purpose with a multiple line printing (MLP) feature.

No.

Mr John Henry Jones 1328 Kenosha Avenue Union City Pennsylvania

FIELD C

FIELD B

FIELD A

a..

.....J
~

7 8

31 32

55 56

791

Figure 13.15 MLP three-card format.

The special MLP card is ideritified with a 9, an 8, and ai, 2,
or 3 punch in a designated card column. The 1, 2, or 3 punches
indicate whether 1, 2, or 3 lines are to be printed from the card.
Normally, 24 card columns are sufficient for each line of an address, but these may be expanded within the 80-column capacity
of the card. One line may be made up of more than 24 columns;
another line may be made up of less.
The 403 performs all the operations of the 402 and in the
same manner. In addition, a control panel wired for the 402 can
be used on the 403 without changing the wiring. To accomplish
MLP, the 403 is equipped with a third set of 80 reading brushes
located at a first reading station. Two other reading stations are
located in the same position as on the 402. A fourth nonreading
station is used for printing the third line of the three-line card. The

BASIC CONTROL PANEL WIRING

293

arrangement of the four stations in the machine is shown in
Figure 13.16.
PATH OF CARD

OPERATION

PRINTING FROM CARD

Zone punching from Field
A zones the type bars.

/
IotR

JOHN

HENRY

Digit punching in Field A
Is combined with zone
punching for Field A, and
Field A Is printed

JONES /

Zone punching from Field
B zones the type bars,

Vl
IotR

JOHN

1328

HENRY

KENOSHA

JONES/
AVENUE

IotR

JOHN

1328
UNION

HENRY

KENOSHA
CITY

JONES /

Digit punching in Field B
is combined with zone
punching for Field B, and
Field B is printed The
normal card following an
M LP 3 card remains sta
tionary at the first station
(between 0 and 1) as the
second line prints
Digit punching for Field C
is stored in counters Type
bars are zoned for Field C

Digit punching and zone
punching for Field Care
combined, and Field C Is
printed The normal card
advances from the first to
the second station

AVENUE
PENNSYLVANIA

Figure 13.16 Schematic, MLP operation.

1. Field A, First Line. The zone punches of field A are wired
on the control panel to read at station 1 and the designated typebars are zoned. The card may be held at this first station between
the 1 and 0 positions before reading the zone punches. This delay
is necessary to prevent overlap in zoning because the same typebars cannot be zoned simultaneously from two different cards one at the first station and one atthe second. The delay in feeding

294

BASIC CONTROL PANEL WIRING

may vary from one to two cycles depending upon the type of card
being read and the relationship of one card to another.
Digit punches are read at station two and field A prints.
2. Field B, Second Line. Zone punches for field B are read
at station 2 and typebars are zoned. Digits punches for that field
are read at station 3 and field B prints. Note that the second line
is printed in the same typebar positions as the first line.
3. Field C, Third Line. Zone punches for field C are read at
station 3 and the typebars are zoned. Digit punches are also read
at station 3 and are stored in 24 counter positions set aside for this
purpose: counters 2A,B, 6A,B, and 4A,B. The digit punching is
combined with the zoning at station 4 when the counters are read
out and field C is printed. Field C also prints in the same typebar
positions as fields A and B.
The use of the 24 counter positions for storage of third line
digits prohibits the use of these counters for totals that are not
cleared before the next MLP card is printed. That is, these counters
cannot overlap the storage of digits for line printing with total
accumulation.
Normal cards may be interspersed with MLP cards, thereby
providing for the use of the feature with such common data processing applications as invoice writing, billing, and the like. The
MLP feature may also be used for crossfooting three fields.
Questions

and
Exercises

1. Explain the operation of the pilot selector and coselector in relation to the
timing of cards as they pass through the accounting machine.
2. What is the purpose of counter coupling?
3. How does multiple selection expand the data recording capacity of punched
cards?
4. Why is digit identification of cards in a file often more convenient than X
identification?
5. How does field selection differ from pilot selection and coselection?
6. What is the difference between field selection and class selection?
7. What is the purpose of the multiple line printing feature?
8. Prepare a 402 Accounting Machine wiring diagram to produce the stationery expense report designed as an exercise following chapter 12.

IBM 407 ACCOUNTING
MACHINE

The IBM 407 Accounting Machine also prepares printed reports from 80-column punched cards. Several advanced features
give this machine considerably more flexibility, speed, and capacity
than the earlier 402 and 403 models. Most important is its unique
method of reading cards. Figure 14.1 is a schematic of the 407
feed unit.
Cards are placed in the hopper with the 9-edge toward the
throat. Cards feed into the machine from the bottom of the deck
in the same sequence as the printed report. Each card is positioned
at the first, then at the second reading station with card grippers
that move horizontally as indicated by the arrows in the diagram.
Cards can be held at the reading stations for any given number of
cycles. After reading, they move around a rotating drum into the
stacker and are held there by a pressure plate. When the stacker is
full, the machine stops.
As a card is positioned at the reading station, it is lined up by
the card aligners so that the 960 possible punching positions are
directly under 960 stationary reading brushes. The brushes are
directly above 960 metal segments,Jabeled 9, 8, 7, ... , 12 in Figure
14.1. Any hole punched in a card allows the corresponding brush
to make contact with a metal segment. The resulting electrical
impulse is transmitted from a commutator as it rotates clockwise
in relation to the brushes. There are 80 commutators at each read295

296

IBM 407 ACCOUNTING MACHINE

Card Lever

SECOND READING STATION

FIRST READING STATION

Figure 14.1 Feed unit, IBM 407 Accounting Machine.

ing station, one for every column of the card. The commutators
rotate together, starting with the 9 position and advancing progressively to the 12 position. Impulses are transmitted to any
brushes in contact with their corresponding segments through holes
punched in a card. Th€ brushes are connected internally to hubs
on the control panel where punched-card data can be wired to
control the various functions of the machine.
Because cards are read while stationary, any card can be read
again and again. This feature permits more than one line to be
printed from a single card. Also, anlOunts and quantities can be
crossfooted from a single card into a counter for printing. Depending upon the model, the 407 speeds range from 150 lines per
minute list and 150 lines tabulate to 50 lines list and 50 lines tabulate. Available counter capacity ranges from 168 positions down
to 21 positions.
The 407 has 120 or 96 print wheels instead of typebars as
shown in Figure 14.2. The wheels are arranged as though they
rotated on a single shaft with their outside rims a fraction of an
inch from the paper. The result is a bank of 120 possible printing

IBM 407 ACCOUNTING MACHINE

positions within a width of 12 inches, or 12 characters to the inch.
Each print wheel contains 47 separate character positions:

10 digits:
26 letters:
11 special characters:

through 9
A through Z
j$D*%@&-#.,

As shown in Figure 14.2, the print wheel is divided into 12
equal parts corresponding to the arrangement of characters shown
in Figure 14.3. Note that special characters are combinations of
8-3,8-4 with the 0, 11, and 12 zone punches or no digits and single
zone punches. In operation, the print wheels remain stationary
until the numerical rows of the card are read. At this time, one of
the corresponding 12 sections of the wheel is selected. A second
selection of one of the four parts of a section is made when the
zone rows are read. If no zone is read, the wheel is positioned to
print the digit only.
8

ROTATION

\
Figure 14.2 407 print wheel.

The print wheel rotates at a high rate of speed until printing
time, when its speed is reduced to 25% of normal. At the actual
time of printing, the wheel is moved against the platen in a straight
line, producing maximum legibility. The rotary motion of the
print wheel at print time is compensated by a special cam.

297

298

IBM 407 ACCOUNTING MACHINE

ZONE

LOWER
PUNCH

8-3
8-4

1. Total Symbol 2. Check Protection

Figure 14.3 407 character arrangement.

Although one line is printed during one print cycle, the
wheels print four different times within that cycle. All the wheels
zoned for zero print first, followed in succession by those of 11,
12, and N.

Multiple Line
Read Feature

The multiple line read (MLR) feature uses the capability of the 407
to read a card more than once. Each time a card is read, a line can
be printed or a field can be entered into a counter. There is no
limitation on the number of times one card can be read.
At least one column of the card must be set aside to identify
the MLR card and then to determine the number of lines to be
printed from that card. This is done by punching the letter A for
one-line printing, the letter B for tw04ine printing, and C for threeline printing. In each case the 12 punch zone for the letters A, B,
and C is used to start MLP operations. The 1-, 2-, and 3-digit
punches of the letters are used to stop MLR operation after one,
two, or three lines have been printed. More than three lines can
be printed by making use of a repeat feature.

IBM 407 ACCOUNTING MACHINE

299

Normally, as many as 28 positions can be printed on a single
line. If required, more than 28 can be printed by coupling
coselectors.
The feature is particularly useful in applications of address
printing. For example, address labels can be prepared on continuous forms. Interspersed three- and four-line addresses can be produced at rates that vary between 1,800 and 3,000 labels per hour.
Specially designed forms, permitting multiple label printing side
by side, increase production accordingly. After printing, the labels
are separated and attached to the material to be mailed.
The 407 can be equipped with an address-writing feature to
print three-line addresses at the rate of 9,000 per hour (or 150 per
minute). Interspersed three- and four-line addresses can be produced at the rate of between 4,500 and 9,000 per hour.
In a high-speed application, addresses are printed on a seveneighth inch tape from either 403 MLP cards or 407 MLR cards as
shown in Figure 14.4. Over 7,000 carbon masters or 8,000 labels
can be produced from one roll of tape. The carbon masters are
used in commercial heat-transfer addressing machines and have a
reverse carbon image printed on the back of the tape. Labels are
processed by automatic mailing machines that cut and affix them
to material for mailing.
MR
CHARL ES GuGGENHE I ME R
C 1492 MINNESOTA eC'-'LEVARO
INTERNATIONAL BUSINESS
BINGHAMTON N Y C MACHI"ES CORPORATION
590 MADISON AVE
NEW YOR~ 11 N Y

MR
AlFqED SCHLESINGER
C \42 FERRUS ST
AKRON
OHIO

MR
PAUL AUST IN
C EMPIRE ~TfEL COMPANY
345 CONKLIN AVENUE
SA" FRANC I seo J2 CAL

Figure 14.4 Address printing on tape.

The address tape is fed by a special attachment that replaces
the standard carriage of the machine. The tape travels across the
platen roller at an angle. With this arrangement, four different
lines of four separate addresses can be printed at the same time.
Continuous form carbon paper passes face forward between the
platen and the tape. Addresses are also printed on the face of the
tape by normal print wheel and inked ribbon operation.
Printing is done from four fixed groups of print wheels. The
right group prints the first line, the second group prints the second ..
line, and the third group the third line. The fourth group at the
extreme left prints only when a fourth line is required. Three-line
addresses are punched in one card. Four-line addresses require two

300

IBM 407 ACCOUNTING MACHINE

cards. A standard carriage tape is used to control the movement of
the address tape across the platen.
Because the information must be printed from four different
sources at the same time, the first line is printed from the first
reading station. The second line is printed from the second reading station, and the third and fourth lines from counters and
storage units.
A perforation is automatically cut in the address tape between
each address as the tape passes from the platen to a rewind reel.
These holes keep the tape in proper registration when it is processed
by an automatic mailing machine.

Storage Units

A 407 can be equipped with four 16-position storage units. The
units store up to 64 positions of numerical information or 32 positions of alphabetic information from a card, from an emitter within
the machine, or from a counter. Data can be read into or out of
the storage units at will, under the control of a digit, an X punch, a
card cycles impulse, a program exit, and certain types of impulses
that are available from the operation of the automatic carriage.
Each storage unit can store letters, digits, and three of the special
characters: minus (-), ampersand (&), and diagonal (/). The remainingeight special characters include 8-3 and 8-4 numerical portions and therefore cannot be stored.
Storage units can be used as follows:
1. For storing information to be printed on sheet headings
after a form overflow so that the name and address is duplicated
on all subsequent forms of that group.
2. To increase the number of usable card columns for detail
information by storing name, city, and state to be printed as
required;
3. To store alphabetic or numerical information to be simultaneously read out and printed line by line with detail information
from a card.
4. To store information to be summary punched.
5. To store information to be group indicated.
6. To store columnar heading information so blank unruled
forms can be headed by reading out of storage for each new sheet.

IBM 407 ACCOUNTING MACHINE

301

Figure 14.5 indicates the use of information placed in storage
for reporting printing.

DATE
NET

CHECK

EARN-

CHECK
NUMBER

EARNINGS
TO-DATE

DATE
NET
PAY

NOTES

4939

12208

61806

6619

12209

76930

5556

12210

69520

4795

12211

40540

8254

12212

110206

MAIL

4837

12213

70215

HOLD

69120

5505

12214

5003

12215

49650

6058

12216

91642

MAIL

REPRESENTATIVE COMPANY
ANY .CITY-ANY STATE

INVOICE TO

--, I

SHIPPED TO

INVOICE
NUMBER
PAGE
26115
2

_E C BROWN & CO

DATE
CODE
SHIPPED VIA

--1

ORDER DATE
ITEM
NUMBER

PLEASE REFER TO
OUR INVOICE NUMBER
WHEN REMITTING
TERMS: 2% TEN DAYS,
F.O.B. FACTORY

ORDER NO.

DESCRIPTION

QUANTITY &
UNIT OF MEAS.

INV TOT
2% DISC
NET AMT

Figure 14.5 Item and total1abeling. Information in storage.

UNIT
PRICE

AMOUNT

$ 1664.61 *
$
33.29CR
$ 1631.32*

COST AND
PROFIT
1239.651664.61
424.96

302

Tape-Controlled
Carriage

IBM 407 ACCOUNTING MACHINE

The tape-controlled carriage feeds and spaces continuous paper
forms at high speed for printing by the 407. The carriage is similar
in operation to the carriages available with the 402 and 403 Accounting Machines previously described. The device can be programmed to control spacing, skipping, overflow; and form ejection
by punching holes along a narrow tape that exactly corresponds
in length to one or more forms.· Once it is punched, the tape can
be inserted in the carriage to control form operation and then removed to be used as often as required.
The carriage will accommodate forms up to a maximum width
of 16 3/4 inches, including punched margins. Minimum width is
4 3/4 inches. The maximum length of the form depends upon the
number of lines to the inch: 33 inches for a carriage that spaces
four lines to the inch, 22 inches for six lines to the inch, and
16 1/2 inches for eight lines. While forms of any size within these
limits can be handled, it is usually advisable to obtain forms of
standard sizes from the manufacturers. Forms can also be designed
to permit printing in practically any desired arrangement. Skipping can be controlled to 10 different sections of the form.
Single-, double-, or quadruple-spacing can vary between lines
as controlled by wiring on the control panel. For example, the
heading section of a form may be single-spaced and the body
double-spaced. Any other spacing required must be controlled
by the tape. Spaces up to two inches between lines can be skipped
at the same rate as normal single-spacing. This skipping is a smooth,
high-speed advance of the form, allowing successive lines to be
printed up to two inches apart at the rate of 150 lines per minute.
The overflow punch in the tape can also be used to start other
operations before ejecting a completely filled form. For example,
a total may be printed at the bottom of each page before the form
is advanced to the next sheet. Several lines of numerical or alphabetic identifying information can be printed on an overflow sheet.
The information can be printed from storage. For example, invoice
and page number of a multipage invoice can be printed at the top
of each successive sheet.
Totals can also be printed on a predetermined total line,
whether the form is completely filled or not. For example, although only two or three items have been printed on a form, the
total of these items may be printed on a previously designated line
of the form, rather than directly beneath the last printed item.
When one sheet or page of a report is completely filled, it can
be ejected and the next form can advance to either the first print-

IBM 407 ACCOUNTING MACHINE

303

ing line or to the first body line. That is, spacing for the heading on
the second sheet of an invoice differs from the spacing on the first
sheet. Usually, printing on the second sheet starts directly on the
miscellaneous data line with invoice and page number.
Overflow skipping is caused by sensing a punch in a specific
position of the tape that starts advancing the paper to the required
line of the next form. The carriage can be programmed so that if
the last card of a group prints on the last available detail line, the
accumulated total for that group prints before the form is skipped
to the next sheet. Thus, the printing of a separate sheet for an unidentified total is prevented. Printing all totals on the last overflow form can be done without reducing the printing space on each
of the preceding forms.

Figure 14.6 shows carriage control tape punching for predetermined printing locations. The tape is ruled for 12 columnar
positions indicated by vertical lines. These positions are called
channels. A maximum of 132 lines can be used for control of a

Control Tape

GldJE
;;

=O~CD

....

(II'"

(,IIN_

...
..'"...
...

0 I1
I
01

'tI

1'1

;;;

~....

..•

1'1

iii

1'1

::;

0
....

iii

'tI

1'1
:I:
1'1

N
N

1'1

tl
~

~
~
~

;=OCDCD'"

CftUl"''''N-

III
!:l

01
1
01
1
0 1
1
01
I

01
1
01
1
01

STATEMENT

\0
1
10

GENERAL MANUFACTURING CO .
ENDICOTT, N.Y.
IN ACCOUNT WITH
A. B. SMITH & CO.
1025 E. MAIN ST.
DA YTON, OHIO

ICUSTOMER Nol ! MO.! DAY ! YR.!
7756
I 5
01

CODES
1. CASH
2. RETURN
3. ALLOWANCE

10
I

DATE
MO. DAY
3
4
4

12
2
10

REFERENCE
21046
28522
5096

CODE

CHARGES

CREDIT

10
I

206.50
134.62

0
206.50

0
0

'--'I

BALANCE DUE

134.62

Figure 14.6 Control punching for predetermined printing locations.

304

IBM 407 ACCOUNTING MACHINE

form, although for convenience the tapes are slightly longer. Horizontallines are spaced six to the inch for the entire length of the
tape. Round holes in the center are prepunched for a pinfeed drive
in a tape sensing mechanism that controls the carriage. The tape
advances through the mechanism in synchronism with the movement of the printed form through the carriage. The effect is exactly the same as if the control holes were punched along the edge
of each form.
Twelve brushes, one for each channel, are positioned over the
tape for sensing the punched holes. Brush 1 rests on channell,
brush 2 on channel 2, and so on. A hole in the channel position
allows the brush to make contact with a metal roller and set up the
necessary circuits that can be used to control form operation.
Carriage control hubs are available on the control panel. A special
punch is available for punching the holes in the required positions
in the channels. After the tape is punched it is cut and looped into
a belt. The two ends are glued together. The completed tape is
inserted in the carriage reading mechanism by the operator as part
of the setup operation in preparation for a report run.

Form Design

One of the basic tools used in designing forms is the spacing chart,
such as shown in Figure 14.7. The numbers across the top from
o to 11 represent the tens position of the print wheel number.
The numbers directly beneath represent the units position of the
print wheel number. For example, print wheel 42 is located by
referring first to the 4 column and then to the digit 2 within the 4
column. Print wheel 9 can be located by referring to the 0 column
and then to the digit 9 within that column. The form alignment
symbol (black triangle) locates print wheel 1, 60, or 120 and
should be embodied in the form design as an aid to the operator
in aligning the form when setting up the machine.
In this example of the spacing chart, a facsimile of the carriage control tape is shown at the left for marking the control punching for a specific form. Notations have been included concerning
standard form widths and depths, lateral movement of the carriage,
and instructions to forms manufacturers.
The following are some of the considerations that should be
given to good form design on the 407:
1. Unlike the 402 and 403, every print wheel on the 407
contains all the characters that can be printed by the machine.

IBM 407 ACCOUNTING MACHINE

6 iI~' 7

305

123456189012345678801234561890123456189012345618901234561890 12 34561890 12345161890 12 34561890 12 34 561890112 34 561890 123456189

TTTTT

!!I!IIIII

2
3
4

'f-

SHIPPING SCHEDULE

•
7

jscHEDULE~

•

PItODUCT

t-:'. r;:; "•.

2
3
4

,
•
•

ORDER

CUSTOMER

QUANTITY

NO.

12
13
14

II
17

II
II

20
21
22
23
24

20
21
22

24
21
21
27
21
21
30
31
32
33
34
35
38
37
38
31
40
41
42
43
44

15
21
27
21
2.
30
31
32
33
34

31
37
31
31
40
41
42
43
44

01213,4561

9 123 56189012

1 3

,
,

~Hj

III

I I

I Ii

41
47
4'
4'
10
11
12
53
54
51
H
57

+H++ +

-H..+

'1,

,.'

,T,
I
;

,
II

-i-

,T
I:

12
13
14
I!
II

3~o
I

,
I -

67 901234 5 618 9 0 1 23456189 1234 5 618 9 01234 516118

f..l+

l'
'

'r+

l~
I,

" rr't- ,
'W- HJlit nr I II
"
'1'mwf~f+
4+M+h-r-rI
l+" t tIT
,I,

50
11
52
53
54

I
I

-!+Hiii

f'
:

tl"~ ,+

Figure 14.7 407 spacing chart.

Therefore, there is no restriction to position alphabetic information only to the left of the print area. Form space can be more efficiently utilized in some applications by taking advantage of this
additional flexibility. For example, sold to and ship to names and
addresses can be printed on the same line, one on the left side of
the form and the other on the right. This is a particular advantage
in the preparation of invoices, bills, orders, and the like. In this
case, careful consideration must also be given to the header card
design.

81
62
63

i i ,

~ 1+,

l '

57
51

19
10

t+H

ITHh- t

tt -t-tI'-

7]8 9

i:'
I

t+,

~! ~

9form~~_o 1~ ~~l1faIstolj1213rJ51rl!ts 'h14)5 6f~8~~J3hlsOO7~_91!.1.Jfu~~ ~J!~I~J~L~
I

48
47
41

++t+ 1

-! '

-T

I
I

i f ' i

85
II

306

IBM 407 ACCOUNTING MACIDNE

In some billing operations, the bill is made up in two parts:
one an itemized statement and the other a remittance stub to be
returned with payment. Customer name can be printed on both
parts of the form by split wiring from the reading station to the
print wheels.
2. The preprinting of vertical lines between two adjacent
printing positions can be avoided because there is a maximum
tolerance of only 0.010 inch between adjacent printed characters.
The machine can be wired to print commas, decimals, oblique
lines, dashes, etc., which can replace the vertical lines.
3. The number of legible multiple copies obtained depends
upon the weight of the paper used for each form, the carbon coating, and the hardness of the platen. The striking force of the print
wheels against the paper is not adjustable. Therefore, paper and
carbon should be tested with a platen of the recommended hardness to be sure of best results.
4. On the 407, the credit symbol (CR) prints from two print
wheels and the minus sign prints from one. For this reason, the
minus sign is recommended as a credit symbol.
5. The dollar sign does not need to be preprinted on check
forms. This symbol can be printed directly to the left of significant
digits in the amount field. This check protection can be wired on
the control panel to print amounts as $.50, $1.00, or $12.00,
rather than $ .50, $ 1.00, or $ 12.00. The printing of the
floating dollar sign by the machine prevents the printed amount
from being fraudulently altered by filling in blank spaces in front
of significant digits.
6. The number of alphabetic positions that can be assigned
in a forms layout to print from storage is 32. More than 32 positions can be stored by combining the use of counters and co""
selectors as required.
7. A character emitter can be wired through selectors or
storage control to print report headings on unruled forms and to
identify the first and last line of a report with such information
as brought forward on the first line and carried forward on the last
line.
8. A variable length overflow feature makes it unnecessary
to reserve space for multiple total lines.

IBM 407 ACCOUNTING MACHINE

9. When spacing is six lines to the inch, skips of up to two
inches are possible without interrupting the normal printing speed
of the machine. Therefore, it is an important means of increasing
machine production to hold forms to a size that does not require
skipping in excess of these limits between print cycles.
10. In the 402 and 403 print units, zero suppression to the
right of a numerical field is mechanical. That is, unless certain
levers are hand-set, zeros print automatically to the right of an
amount field (Figure 14.8). Therefore, to suppress unwanted zeros,
it is necessary to align all numerical fields to print in a columnar
arrangement, including the total, thus:
Example:

(Customer number)

603918
123
1234
12345
1
12
Total 13715*

Zero suppression on the 407 is electrical and completely flexible. It is not necessary to vertically align the right-hand digit of
an amount field with the right-hand digit of a heading, for example,
to print the customer number above an amount column. Also,
total identification permits the flexible positioning of totals, thus:
Example:

(Customer number)

603918
123
1234
12345
1
12
Total 13715*

Complete descriptions of all the operating features of tapecontrolled carriages can be found in the manuals of operation
supplied by the manufacturer. Complete forms specification for

307

308

IBM 407 ACCOUNTING MACHINE

use with the carriage are also furnished by the forms manufacturers. These topics will not be covered in this text .
.311800158:780015516800
11960009811.30009011700

BROS

I .

Hammersplit levers not raised.
ABBOT
A B BOT

BROS
B R 0 S

11791121311
1 1 7 9 1 21.3 1

3!181
1: 9 6

158!78
9 811 3

1
1

155!68
9011 7

Hammersplit levers 34 (alphameric) and 6, 13 and 20 (numerical) raised.

HAMMEJlSPUT

LEVmS

Figure 14.8 402, 403 zero suppression.

Questions

and
Exercises

1. How does the 407 Accounting Machine differ' from the 402 in card
reading and feeding? Printing?
2. Describe the function of the 407 multiple line read feature.
3. What are the special design considerations for MLP cards?
4. What are storage units used for? How do they differ from counters?
5. What are the functions of the tape-controlled carriage?
6. Why is form design considerably more flexible on the 407 than on
the 402?
7. What features of form design can be originated by the machineas reports
are printed?
8. What is zero suppression?

IBM 407 ACCOUNTING MACHINE

9. What is a floating dollar sign?
10. Design the stationery expense reports from chapter 12 for printing on
the 407.

309

INTRODUCTION TO
COMPUTERS

Preceding chapters have described a number of data processing operations that can be effec,tively mechanized with specially
designed machine equipment. For this type of automatic processing, the punched-card record has been presented as the standard
medium of data handling and communication.
Operation

Type of Equipment

Transcribing original data into
machine-readable records
(punched cards)

Keypunch (80 column)
Data recorder (96-column)
Mark sensing punch (519)

Checking data transcription

Key verifier

Sorting and classification

Sorter

Reproducing and transferring
data

Document-origina ting
machine (519)

File handling

Collator

Calculating

Calculating punch

Report preparation

Accounting machine

Output card preparation

Summary punch

In all unit-record processing, the above operations are arranged in some predetermined sequence to form a planned procedure or system. Typically, machine records are punched, verified,
310

INTRODUCTION TO COMPUTERS

sorted to sequence, associated with previous records, and then
printed to report results. Each operation within the procedure is
carried out on a separate machine unit specifically designed to perform that function. While this type of system is very effective
when compared with many manual methods, some inherent disadvantages can quite readily be shown.
First: . The sequence of operations is always under manual
control. That is, to follow a given procedure, card records must be
transported by hand from machine to machine. As a result, the
system is entirely dependent upon the skill and ability of human
operators to run the various machine units efficiently and to follow
instructions correctly.
Second: Punched-card equipment is mainly electromechanical. Each operation requires physical movement of a paper record
past reading or punching mechanisms. Therefore, processing has
obvious limitations of speed. Production is still very fast by manual standards but nowhere near approaches the rates possible with
electronic components.
To overcome these limitations, more advanced computing
equipment has been designed and produced with the capability to
perform all essential data processing operations in a single integrated
unit. In such a computer system, electronic components are utilized
in combination with electromechanical devices. Processing is almost entirely electronic while the operations of "reading and writing" records are performed with improved input and output devices
interconnected through a central processing unit. The organization
of a simple system is shown schematically in Figure 15.1.
As shown in the illustration, the input devices read directly
from a record medium, in this case punched-card files. All information from each card is transmitted over a channel to a storage
unit where the data is retained for processing. More than one input device can be provided, so that records from at least two different files can be read into storage at the same time; for example,
a master record and associated details.
Before entering storage, input data is converted to a representation suitable for internal electronic manipulation. The entire
system is controlled by a central processor made up of two sections:
control and arithmetic/logical. The action of the processor is
directed by a stored program also retained in the storage unit. Input data can be handled according to the procedure established by
the program and then returned to storage where it is available for
output. The completed output may take the form of either

311

312

INTRODUCTION TO COMPUTERS

G ----. '--__

Processor
and
e

---,mr-a_in_st_o_ra-rg_ _ _..... - - -

ljJU'1 J

Figure 15.1 Schematic. Basic computer system.

punched cards, printed reports, or both. As data is "written"
from storage, it is reconverted to punched-card code or to the
graphic characters of printed information.
Limited input and output is also provided by a printerkeyboard. Operator control of the system is by means of a console
where power is turned on or off, where the contents of storage can
be displayed, and where system operation can be observed and
monitored as required.

Figure 15.2 IBM System/3.

INTRODUCTION TO COMPUTERS

313

The following sections and Chapter 16 discuss each of these
components in general terms as basic units common to all digital
computers. Chapters 17 and 18 deal specifically with the IBM
System/3, a small unit-record computer designed to operate with
96-column cards (Figure 15.2).

It has been shown that data can be represented in many ways

Computer Data

depending upon the medium of communication, recording, or
storage, and the established conventions of coding for the medium.
For example, in punched cards data is represented as patterns of
holes. When transmitted to machines, the data is then represented
by timed electrical pulses, the positions of counter wheels, and the
relationships of various mechanisms within the device.
In a computer, data is represented by many electronic components: transistors, magnetic cores, electrical pulses, wires, and
so on. The storage and flow of data through these devices is
represented as electronic signals or indications. The presence or
absence of these signals in specific circuitry is the method of
representing data, much as the presence or absence of holes in a
card represents data.

Representation

"0" State

"1" State
MAGNETIC CORE

ESc:>]

~CJ~
RELAY OR SWITCH

~ •

•

ELECTRICAL PULSES

.--.
Figure 15.3 Binary state of computer components.

Computers function in binary states; that is, computer components can indicate only two possible states or conditions. For
example, the ordinary light bulb operates in a binary mode - it is

314

INTRODUCTION TO COMPUTERS

either on or off. Likewise within the computer - transistors are
maintained either conducting or nonconducting, magnetic materials
are magnetized in one direction or in the opposite direction, and
specific voltage potentials are present or absent. These conditions
are illustrated in Figure 15.3. The binary state of the various components is a signal to the computer, just as the presence or absence
of light from an electric bulb can be a signal to a person.
For example, a device to represent decimal values can be designed with four electric bulbs and with switches to turn each bulb
on or off as shown in Figure 15.4. The light bulbs are given decimal values of 8, 4, 2, and 1. When a bulb is lighted, it represents the
decimal value associated with it. When a bulb is not lighted, no
value is represented. With such an arrangement, the total decimal
value will be the numeric sum indicated by the lighted bulbs.

Off

©)

Off

Figure 15.4 Binary coded decimal representation.

All decimal values 0 through 15 can be represented. The value
The
value 9 is represented by having the 8 and 1 lights on and the 4 and
2 lights off. The value 5 is represented by having the 4 and 1 lights
on and the 8 and 2 lights off, and so on.
Note that the bulbs may be given almost any significance, depending upon what mode of operation is to be established. For
example, the 16 possible on/off combinations can represent letters
of the alphabet. When the right-hand bulb is on, assume that it
represents the letter A. When the second bulb from the right is on,

o is represented by all lights off; the value 15 by all lights on.

INTRODUCTION TO COMPUTERS

315

assume it represents the letter B. When both bulbs are on, the
letter C is represented, and so on. Because there are 26 letters in
the alphabet, one more bulb must be added to the device to represent all letters.
Data is represented within a computer by assigning a specific
value or significance to a single binary indication or group of binary
indications. The significance assigned to the specific indications
becomes the code or language for representing data. Calculations
can also be performed using the binary number system and binary
arithmetic.
Many data processing operations are logical in nature. Since
logic is fundamentally a yes/no system, these operations can be
performed very efficiently in binary. For example, when one value
is compared against another, the result of the comparison can be
indicated as high or low, equal or not equal. When a calculation is
performed, the result is nearly always indicated as plus or minus,
positive or negative, debit or credit.
Many control operations within the machine system can also
be expressed in the binary state. Is the card reader ready to feed
the next card? Yes or no? Is an error indicated in the transmission
of input data to storage? Yes or no? Does the code structure of
the data received conform to the accepted convention of the
system? Yes or no?
In computer parlance, the binary state is often referred to as
the presence or absence of a "bit." Probably the term originated
to describe the recording on magnetic tape where combinations of
magnetized "bits" and "no-bits" in the surface of the recording material are used to represent data. As will be shown in the next section, these terms have carried over into references commonly made
to the ones and zeros of binary notation and to computer codes.

The design and programming of an application for computers does Binary Number
not normally involve any direct use of an internal machine lan- Systems
guage. It will be shown later that programming systems have
been developed which make it possible to write instructions for a
computer in a language much more convenient to use than direct
binary representation. The computer itself can convert from this
programming language to its own internal system of coding.

316

INTRODUCfION TO COMPUTERS

However, to understand the actual capability of the machine,
including storage capacity, operating characteristics, and so on, it
is helpful to have some general knowledge of the various machine
number systems.
The familiar decimal (base 10) number system uses 10 symbols or digits to represent numeric values. The symbols are 0
through 9. The place value of the digits also signifies units, tens,
hundreds, thousands, etc.
The binary (base 2) number system uses only two symbols
or digits: 0 and 1. The place value of the symbols is based on the
progression of the powers of 2. That is, the units position of a
binary number has a value of 1; the next position, a value of 2; the
next, 4; the next, 8; the next, 16; and so on. These and additional
place values are shown in Figure 15.5.

Figure 15.5 Place values of binary numbers.

In pure binary notation, the digits indicate whether the corresponding power of 2 is present or absent in each position of the
number. A 1 represents the presence of the number; the 0 represents
the absence of the value. The place value does not signify units,
tens, hundreds, or thousands as in the. decimal system. Instead,
the place value signifies units, twos, fours, eights, sixteens, and
so forth. In this system, the quantity 12 is expressed with the
symbols 1100, meaning (1 X 22) + (1 X 22) + (0 X 21) + (0 X 2°)
or (1 X 8) + (1 X 4) + (0 X 2) + (0 X 1). Figure l5.6shows the
binary representations of the decimal values 0 through 16.
Computers using the binary system of data representation are
typified by the IBM System/3, System/360, and System/370. For
all of these systems, the basic unit of information is a byte, made
up of 8 binary positions. Four bytes make a computer word with
32 consecutive bit positions of information which are interpreted
as a unit, much as a character of digit in other systems.
The bit sections within a word have a place significance related to the binary number system. That is, the place position of
a bit in the word determines the value of the bit. In the binary
number system, the decimal values of the places (from right to
left) are 1, 2, 4, 8, 16, 32, 64, and so on.
Although the place value of the bits of a word are always those
of the binary number system, they can be interpreted or processed

INTRODUCTION TO COMPUTERS

317

in such a way as to represent a significance other than quantity.
For example, a 32-bit word can be interpreted as a 32-place binary
number, as an 8-digit hexadecimal number, as four alphabetic or
numeric characters, or as any predetermined representation.
Cij

Place
Value

.:l
(.)-

0>
Ql

Figure 15.6 Binary equivalent of decimal values for 1 through 16.

It is apparent that binary numbers require several times as many

positions as decimal numbers to display the equivalent number. In
talking and writing, these binary numbers are awkward. A long
string of ones and zeros cannot be effectively transmitted from
one individual to another. The octal and hexadecimal number systems fill this need. Because of their simple relation to the bina.ry

Octal Number System

318

INTRODUCTION TO COMPUTERS

number system, numbers can be converted from one system to
another by inspection.
The base or radix of the octal system is 8. This means that
there are eight symbols to represent all values. The symbols are 0
through 7; there are no symbols 8 and 9. The important relationship is that three binary positions are equivalent to one octal
position. The sample chart in Figure 15.7 is used to convert between binary, octal, and decimal systems.
Decimal

Binary

Octal

000 000

000 001
2

000 010

000 all

000 100

000 101

000 110

000 111

001 000

001 001

001 010

001 011

001 100

001 101

001 110

001 111

010 000

Figure 15.7 Binary and octal equivalent of decimal values for 1 tluough 16.

The octal system is used mainly to provide a shorthand. The
digits of the number represent the coefficients of the ascending
powers or8. Consider the octal number:
173 = (1 X 82 )
=
64

+ (7 X 81 ) + (3
+
56 +

= 123 (decimal)

X 80 )
3

INTRODUCTION TO COMPUTERS

By remembering what a number represents in the binary or octal
system, the number can be converted to its decimal equivalent by
the method shown above. However, with larger numbers, this
method becomes more difficult to use. More convenient methods
are described in the following sections.
Decimal to Octal

Rule: Continuously divide the decimal number by 8 and
develop the octal number from the remainders of each step in
the division.
Example: Convert the decimal number 149 to its octal
equivalent.
8 ~ remainder 5
8 ~ remainder 2
11 remainder 2
8

The octal equivalent of decimal 149 is 225.
The original number to be converted is first divided by 8 as
shown above. The remainder from this division becomes the loworder digit of the octal number, or 5. The quotient received from
the first division is then divided by 8. The remainder becomes the
next higher-order position of the octal number, or 2. The steps
are repeated until the quotient is smaller than the divisor. After
this step, the final quotient is considered to be the high-order digit
of the conversion; in this example, 2.
Octal to Decimal

Rule: Continuously multiply by 8 and then add the next
octal digit.
Example: Convert the octal number 225 to its decimal
equivalent.
2

16
~~
+2
18

144

149

319

Integer Conversion

320

INTRODUCTION TO COMPUTERS

The high-order digit of the octal number is multiplied by 8
and the next lower-order digit is added to the product. The result
is then multiplied by 8, and the next lower-order octal digit is
added to that amount. When the lowest-order octal digit has been
added, the process ends. The decimal equivalent of octal 225 is
shown to be 149.
Octal to Binary and Binary to Octal

This can be done visually with no calculation required.
Rule: Express the number in binary groups of three ~
Examples:

Octal to Binary
2
010

2
---010

101

Binary to Octal
010 010 101
2

Decimal to Binary

Rule: Continuously divide the decimal number by 2 and
develop the binary number from the remainders.
Example: Convert decimal 149 to its binary equivalent.
21149 remainder 1
2lH. remainder 0
2ln remainder 1
21.l§. remainder

2l2. remainder 1
21..1 remainder a
21.1 remainder 0
2U remainder 1

remainder a

Arrange the binary number as a1a
bottom of the column to the top.

a1a

101, reading from the

INTRODUCTION TO COMPUTERS

321

Binary to Decimal

Rule: Continuously multiply the binary digits by 2 and add
the next binary digit.
Example: Convert 010 010 101 to its decimal equivalent.
10

010

101

X
+0
2
X2
4
+0----1
4
X 2
8

+ 1-----..1
9

X2
18
+0-----1
18
X 2

+ 1--------...J
37
X 2
74

+0--------1
74
X2
148
+ 1--------..1
149
The decimal equivalent of 010 010 101 is shown to be 149.

Fractions can also be converted from one number system to another
as shown in the following examples.

Fraction Conversion

322

INTRODUCTION to COMPUTERS

Decimal to Octal

Rule: Continuously multiply by 8 and develop the octal
number from the carry.
.149
X 8
1.192

Example:

1.536
X 8
4.288
X 8
2.304
The octal equivalent of the decimal fraction .149 is shown
to be 0.1142+.
Octal to Decimal

Rule: Express as powers of 8, and add.
Example:

0.1142

1(8-1 ) + 1(8-2 ) + 4(8- 3 ) + 2(8- 4 )

"8 +

114
2
64 + 512 + 4096

= 0.1489 or 0.149

The decimal equivalent of the octal fraction 0.1142 is shown
to be 0.149.
Octal to Binary and Binary to Octal

Rule: The same rule applies for fractions as for whole
numbers.
Example:

.001 001 100 010
.001 001 100 010
.1

INTRODUCTION TO COMPUTERS

323

Binary to Decimal

Rule: The same rule applies as for whole numbers .
.001 001 100 010

Example:

= 1(2 -3) +
1

1(2 -6) + 1(2 -7) + 1(-11 )
1

= "8 + 64 + 128 + 2048
=

305/2048

= .1489 or .149

While many computers use the binary number system to indicate
the power of 2 up to the size of a full word or designated field,
any direct man-machine communication is in the hexadecimal
numbering system. Large binary numbers are made up of long
strings of ones and zeros, frequently awkward to interpret and
handle. The hexadecimal (base 16) numbering system is used as a
convenient way to represent such large binary numbers. Each
hexadecimal digit stands for four binary digits.
Hexadecimal notation requires 16 symbols to represent 16
number values. Since the decimal system provides only 10 number symbols (0 through 9), six additional symbols are needed to
represent the remaining values. The letters A, B, C, D, E, and F
have been adopted for this purpose, though any other six marks
could have served equally welL Thus, the entire list of hexadecimal symbols consists of 0, 1,2,3,4,5,6,7,8, 9, A, B, C, D, E,
and F, in ascending sequence of value. The following table shows
equivalent decimal, binary, and hexadecimal numbers.
Decimal

Hexadecimal

2
3

7
8
9

Binary

0000
0001
0010
0011
0100
0101
0110
0111
1000
1001

Hexadecimal Number
System

324

INTRODUCTION TO COMPUTERS

Decimal

Hexadecimal

10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26

Binary

1010
1011
1100
1101
1110
1111
10000
10001
10010
10011
10100
10101
10110
10111
11000
11001
11010
11011
11100
11101
11110
11111

A
B
C
D
E

F
10
11
12
13
14
15
16
17
18
19
1A
IB

27
28
29

1C
ID
IE
IF

30
31

Note that upon reaching decimal16, the hexadecimal symbols
are exhausted, and a "carry" is placed in front of each hexadecimal
symbol. This carry symbol (1) has a place value of decimal 16.
Binary to Hexadecimal

Rule: Divide the binary number into groups of four binary
digits, starting from the tight, and replace each group with the corresponding hexadecimal symbol. If the left-hand group is incomplete, fill in zeros as required.
Example:
111110011011010011 =0011 1110 0110 1101 0011
=

or 3E6D3 (hexadecimal)
Hexadecimal to Binary

Rule: Substitute the corresponding group of four binary digits
for each hexadecimal symbol and drop off any unnecessary zeros.

INTRODUCTION TO COMPUTERS

325

6C4F2E (hexadecimal)

Example:

=0110
=

1100

0100 11110010 1110

a110 11000 100 111100 1a111 a

(binary)

Integer Conversion, Hexadecimal to Decimal

Converting large hexadecimal numbers by either direct expansion or a multiplication-addition method becomes quite tedious
and difficult. The use of conversion tables makes possible rapid
conversion of hexadecimal integers into equivalent decimals, so
that necessary arithmetic can be performed in the decimal system.
H

E
X

E
X DEC

Hexadecimal
Positions

1
2
3
4
5
6
7
8
9
A
B
C
D
E
F

DEC

° °

268,435,456
536,870,912
805,306,368
1,073,741,824
1,342,177,280
1,610,612,736
1,879,048,192
2,147,483,648
2,415,919,104
2,684,354,560
2,952,790,016
3,221,225,472
3,489,660,928
3,758,096,384
4,026,531,840
8

1
2
3
4
5
6
7
8
9
A
B
C
D
E
F

DEC

° °

16,777,216
33,554,432
50,331,648
67,108,864
83,886,080
100,663,296
117,440,512
134,217,728
150,994,944
167,772,160
184,549,376
201,326,592
218,103,808
234,881,024
251,658,240
7

1
2
3
4
5
6
7
8
9
A
B
C
D
E
F

DEC

° °

1,048,576
2,097,152
3,145,728
4,194,304
5,242,880
6,291,456
7,340,032
8,388,608
9,437,184
10,485,760
11,534,336
12,582,912
13,631,488
14,680,064
15,728,640
6

1
2
3
4
5
6
7
8
9
A
B
C
D
E
F

DEC

° °

65,536
131,072
196,608
262,144
327,680
393,216
458,752
524,288
589,824
655,360
720,896
786,432
851,968
917,504
983,040
5

1
2
3
4
5
6
7
8
9
A
B
C
D
E
F

DEC

° ° ° ° ° ° °

4,096
8,192
12,288
16,384
20,480
24,576
28,672
32,768
36,864
40,960
45,056
49,152
53,248
57,344
61,440
4

1
2
3
4
5
6
7
8
9
A
B
C
D
E
F

256
512
768
1,024
1,280
1,536
1,792
2,048
2,304
2,560
2,816
3,072
3,328
3,584
3,840
3

Figure 15.8 Hexadecimal-decimal integer conversion.

Figure 15.8 is a hexadecimal/decimal in teger conversion table.
Each column of the table is labeled at the bottom with the number
of the hexadecimal place position from right to left. Each column
also has all the hexadecimal symbols arranged in ascending order
of values from top to bottom with the corresponding decimal
equivalent. Thus, the hexadecimal symbol D in column 1 has an
equivalent decimal value of 13; in column 2 the equivalent value is
208; in column 4 the equivalent is 53,248; and so on.
Rule: Add the decimal equivalents of the hexadecimal digits,
starting with the units position of the hexadecimal number. Work
from right to left, using the table for the decimal equivalents of
the higher-order positions.

16
1
32
2
48
3
4
64
5
80
96
6
7 112
8 128
9 144
A 160
B 176
C 192
D 208
E 224
F 240
2

1
1
2
2
3
3
4
4
5
5
6
6
7
7
8
8
9
9
A 10
B 11
C 12
D 13
E 14
F 15
1

326

INTRODUCTION TO COMPUTERS

Example: Convert hexadecimal A4B5 to decimal.
5 in hex position 1 =
5
B in hex position 2 =
176
1,024
4 in hex position 3 =
A in hex position 4 =
40,960
The sum = decimal value 42,165
Integer Conversion, Decimal to Hexadecimal

While not quite so convenient, it is possible to convert from
decimal to hexadecimal using the conversion table as Figure 15.7.
Rule: Look up in the table the next smaller number than the
decimal number to be converted. Note the hex equivalent and position number. Subtract the decimal value of that hex number
from the decimal number and then look up the remainder in the
table. Repeat the process.
Example: Convert the decimal number 42,1 65 to hexadecimal.
Find the hex equivalent of decimal = 42,165
A in position 4
= 40,960
Remainder
= 1,205
4 in position 3
= 1,024
Remainder
= 181
B in position 2
176
Remainder
5
=
5 in position 1
5
=
The hexadecimal equivalent of decimal 42,165 is shown to
be A4B5.

Questions
and
Exercises

1. List the normal sequence of steps in a punched-card data processing
procedure.
2. What are the essential elements of a computer system?
3. How can values, characters, or other data be represented by binary
devices?
4. What other types of logical operations can be conveniently represented
by binary devices?
5. What isa byte? What can it represent?
6. Write the binary and octal equivalents of the following decimal numbers:

3,5, 7, 9, 12, 15,21.

INTRODUCTION TO COMPUTERS

7. Convertthe following decimal numbers to octal: 125, 19,288, 966, 1346.
8. Convert the following octal numbers to decimal and binary: 437, 16,
114,525,50.
9. Convert from binary to decimal: 101111101, 111101010.
10~ Using the table in Figure 15.8, convert the following hexadecimal numbers to decimal: 4A5B, lCD, 7EF, E6B.
11. Using the table in Figure 15.8, convert the following decimal numbers to
hexadecimal: 24681, 4832, 1234.

327

16 COMPUTER CODES AND
MAIN COMPONENTS

Data can be symbolized visually in many ways, including the
most common method of using printed characters and digits to
record information. In the computer, as in punched-card machines,
data cannot be seen as printed information but must be represented
or coded with a fixed number of binary indications. For example,
a code can be constructed to represent all numeric and alphabetic
characters with eight positions of binary indications made up of
yes/no, on/off, or 0 and I bit combinations in some electronic form.
Most computer codes are self-checking; that is, they are constructed with some method of checking the validity of the coded
representation. The code checking is designed to occur automatically within the machine as normal data processing operations are
carried out. The method of validity checking is a part of the
code design.
In some codes, each unit or character of data is represented
by a specific number of bit positions that must always contain an
even number of I bits. Different characters are made up of different combinations of 0 and 1 bits, but the number of I bits in any
valid character is always even. With this code system, a character
with an odd number of I bits is detected as an error. Likewise, a
code may be used in which all characters must have an odd number
of I bits. In this code, an error is indicated if a character is detected with an even number of I bits.
328

COMPUTER CODES AND MAIN COMPONENTS

329

This type of checking is known as a parity check. Codes
that require an even number of 1 bits are said to have even parity. Codes that require an odd number of 1 bits are said to have
odd parity.
In other codes, the number of 1 bits present in each unit of
data is fixed. For example, a code may use eight bit positions to
code all characters, but exactly four 1 bits will be present in each
character. Characters having more or fewer than four 1 bits cause
an error indication. This system of checking is known as a fixedcount check and is often used for data transmission in teleprocessing networks where computers may receive and send data over
communication lines.

In this code, all characters, including numeric, alphabetic, and
special, are represented with six positions of binary notation plus
a parity bit position. These positions are divided into three groups:
one check position, two zone positions, and four numeric positions. The code structure is shown in Figure 16.1.
Check

Zone

{

N"m";~1 ~
{

0123456789
0110100110
0000000000
000 0 0 0 0 0 0 0
1000000011
00001 1 1 100
1011001100
0101010101

Binary Coded
Decimal (BCD)

ABCDEFGHI J KLMNOPQRSTUVWXYZ

&.D-$*/,%#@

11010011000101100101011001
11111111111111111100000000
11111111100000000011111111
00000001100000001100000011
00011110000011110000111100
o 1 100 1 1 000 1 100 1 100 1 100 1 100
10101010110101010101010101

01010100110
11111100000
11100011100
o1 10 1 1 0 1 1 1 1
00100100101
o 1 001 001 010
01001011010

Figure 16.1 Binary coded decimal.

The four numeric positions are assigned the decimal values
8, 4, 2, and 1, and can represent the numeric digits 0 through 9.
Note that the digit 0 is arbitrarily represented as 1010, actually the
binary number for 10. The B and A zone bits are 0 when only
the numeric digits are represented.
Combinations of zone and numeric bits represent alphabetic
and special characters. The B and A bits provide for four possible
bit combinations: 01, 10, 11, and 00, or the binary notation for
decimal 1, 2, 3, and O. The code structure follows that of 80column cards, that is, three zones plus numeric values represent
letters. No zones with numeric values represent digits.
The C position is the check bit used for code checking only.
Because the six-bit alphameric code is an even parity code, the
total number of bits used to represent a character must include

330

COMPUTER CODES AND MAIN COMPONENTS
CHARACTER
Report I Program
Low~

BCD CODE (Core Storage)

CARD CODE

No Punches

12-3-8
12-4-8

)

8
8

[

12-5-8

< I

12-6-8

'*'
&

$
*
;
~

12-7-8

B
B

8
8
8

4
4

1
1

11-7-8
11
0-1
0-3-8
0-4-8
0-5-8
0-6-8
0-7-8
2-8
3-8
4-8
5-8
6-8
7-8

12-0

I
I
(

%
~

....
1)

#
@

C
C

C
C
C
C

A
A

8
8

4
4

8
C

8
8

4
4

12-1

12-2

o=>w

12-3

12-4

1
2

A
A

11-1

B
B

11-2
11-3

B
B

11-4

12-5
12-6

F
G

12-7

...J
...J

12-8

12-9
11-0

!
J
K

C
C

2
1
2
2

11-5

11-6

11-7

11-8

11-9
0-2-8
0-2

0-3

0-4

8
A
A

1
2
2

0-5

W
X
Y
Z

0-6

£1

0-7

0-8
0-9

A
A

5
6
7
8
9

5
6
7

1
1

8
8

1
1

1
2
2

4
C
C

8
9

A
C

B
C

1
1

B
B

z
i=

High~

1
2
2
2

o(,)

=
,
:

.
I

NAME
Group Mark
Record Mark
Segment Mark
Word Separator
At Sign
Number Sign
Ampersand
Plus
Asterisk
Per Cent
Slash
Backslash
Lozenge
Blank
Substitute Blank
Left Parenthesis
Right Parenthesis
Left Bracket
Right Bracket
Tape Mark
Less than
Greater than
Equal to
Semicolon
Colon
Period or Point
Prime or Apostrophe
Minus or Hyphen (Dash)
Delta

Figure 16.4 Standard terminology for BCD special characters.

Extended Binary
Coded Decimal
Interchange Code
(EBCDIC)

This code uses eight binary positions for each character, plus one
position for parity checking. With eight bit positions, 256 different characters can be coded, including uppercase and lowercase
alphabetic characters, a much wider range of special characters,
and many control characters that are meaningful to input and output devices in a computer system. At present, many bit combinations have no assigned function and are reserved for future use.
EBCDIC is the internal coding system of the IBM System/3.
332

COMPUTER CODES AND MAIN COMPONENTS

333

This is a seven-bit code developed through the cooperation of the
data processing industry communications equipment users. The
code was standardized in an attempt to simplify machine-to-machine and system-to-system communication.
The IBM System/3, System/360, and System/370 all have an
eight-bit character capacity to correspond with the standard eightbit byte. Therefore, for use in these systems, the code is expanded
to an eight-bit representation, referred to by IBM as ASCII-8.
The code may be used for internal processing and input/output
purposes in those media for which ASCII has been standardized.
A complete table of these codes is shown in Figure 16.5.

American Standard
Code for Information
Interchange (ASC II)

Computer storage accepts data from input devices, exchanges data
with the processor, and retains the completed information to be
written by the output devices. Thus, all data to be processed by
the system must pass through storage.
Storage is always divided up into locations, positions, or sections, each of which is given a number or address so that stored
data can be readily located as needed. This arrangement can be
thought of as resembling small mailboxes in a post office, each box
identified and located by an assigned number. Like a mailbox, a
storage location has capacity for a specific amount of information
that can be inserted or withdrawn by the processor. To insert or
withdraw information in any location, the address must always
be known.
Unlike the mailbox, however, whenever data enters computer
storage, the previous contents of that location are replaced with
the new data. When information is taken from a location, the
contents remain unaltered. Once put in storage, the same data
can be used many times. In effect, an exact reproduction of the
information is made available for processing or writing by some
output device.
An almost unlimited amount of processing can be carried out
before data is returned to storage to be made available as output.
In many procedures, all required processing can be completed in
one pass through the system. Mechanical movement of the records
is therefore greatly reduced in relation to processing. As data is
"written" from storage, it is reconverted to punched-card code or
to the graphic characters of printed information.
Computer storage normally contains at least three kinds of
data:

Storage

334

EBCDIC

NUL
SOH

STX
ETX

LC
DEL

RLF
SMM
VT

CR
SO
51

OLE
OCl
OC2
TM

RES
NL

BS
IL

CAN
EM
CC
CUl
IF5
IGS
IRS

IUS
OS
50S
FS

BYP

LF
ETB
ESC

CU2
ENQ

ACK
BEL
SYN
PN
RS

UC
EOT

CU3
DC4

NAK

SUB
SP

Bit
Configuration

00000000
00000001
00000010
00000011
00000100
00000101
00000110
00000111
0000 1000
0000 1001
0000 1010
0000 1011
0000 1100
0000 1101
0000 1110
0000 1111
00010000
00010001
00010010
00010011
00010100
00010101
00010110
00010111
00011000
00011001
00011010
00011011
00011100
00011101
00011110
00011111
00100000
00100001
00100010
00100011
00100100
00100101
00100110
00100111
00101000
00101001
00101010
00101011
00101100
00101101
00101110
00101111
00110000
00110001
00110010
00110011
00110100
00110101
00110110
00110111
00111000
00111001
00111010
00111011
00111100
00111101
00111110
00111111
01000000
0100 0001
0100 0010
0100 0011
01000100

COMPUTER CODES AND MAIN COMPONENTS

EBCDIC

<:
(

$
)

7/12

::;
7

610

a
b

c
d

e
f
g
h
i

Bit
Configuration

01000101
01000110
01000111
0100 1000
0100 1001
0100 1010
0100 1011
0100 1100
0100 1101
01001110
0100 1111
01010000
01010001
01010010
01010011
01010100
01010101
01010110
01010111
01011000
01011001
01011010
01011011
01011100
01011101
01011110
01011111
01100000
01100001
01100010
01100011
01100100
01100101
01100110
01100111
01101000
01101001
01101010
01101011
01101100
01101101
01101110
01101111
0111 0000
01110001
01110010
01110011
01110100
01110101
01110110
01110111
01111000
01111001
01111010
01111011
01111100
01111101
01111110
01111111
10000000
10000001
1000 0010
10000011
1000 0100
10000101
10000110
10000111
1000 1000
1000 1001

EBCDIC

j
k
I
m
n

o
p
q

w
y

PZ 7/11
A

B
C
D
E
F
G
H
I

Bit
Configuration

1000 1010
1000 1011
1000 1100
10001101
10001110
10001111
10010000
10010001
10010010
10010011
10010100
10010101
10010110
10010111
10011000
10011001
10011010
10011011
10011100
10011101
10011110
10011111
10100000
10100001
10100010
10100011
10100100
10100101
10100110
10100111
10101000
10101001
10101010
10101011
10101100
10101101
10101110
10101111
10110000
10110001
10110010
10110011
10110100
10110101
10110110
101101.11
10111000
10111001
10111010
10111011
10111100
10111101
10111110
10111111
1100 0000
1100 0001
11000010
11000011
11000100
11000101
11000110
11000111
1100 1000
11001001
1100 1010
1100 1011
1100 1100
1100 1101
1100 1110

MZ 7/13
J
K
L
M
N

o
p
a

RM 5/12
S
T
U
V
W

Y
Z

o
1

4
5
6
7

8
9

Bit
Configuration

1100 1111
11010000
11010001
11010010
11010011
11010100
11010101
11010110
11010111
11011000
11011001
11011010
HOll011
1101 1100
11011101
11011110
11011111
11100000
11100001
11100010
11100011
1.1100100
11100101
11100no
11100111
11101000
11101001
11101010
11101011
11101100
11101101
11101110
11101111
11110000
11'''0001
11110010
11110011
11110100
11110101
11110110
11110111
11111000
11111001
11111010
11111011
11111100
11111101
11111110
1111 1111

Figure 16.5 American Standard Code for Information Interchange (ASCII-8).

COMPUTER CODES AND MAIN COMPONENTS

335

1. Data to be Processed. These may be elements currently
being operated upon, such as arithmetic factors to be used in calculation, record fields, complete records, blocks of records, and
so on. As far as storage is concerned, this kind of stored data is
transient; that is, as soon as one element or piece of data is processed by the central processing unit, the next is moved into storage. Each new data element replaces that which was previously
stored. Data is always received from some input device. After
manipulation by the processor, the results are transmitted to an
output device. The intervening processing may also involve the
fetching and storing of intermediate results. This may occur any
number of times before the final results are ready to be made
available as output.
2. Constant factors. These items of data are usually required
throughout the carrying out of any given task or procedure: arrays
of tables of information, such as pay rates, tax rates, miscellaneous
values, reference factors, and so on. These items are stored at
some predetermined location. Space may also be allocated for
temporary storage of intermediate results. Since the processor
usually has a limited number of registers, adders, and accumulators,
temporary storage is required as a kind of scratch pad where these
results can be referred to as needed.
3. Instructions. The instructions to the computer for processing are always placed in storage - at least those instructions under
whose control the machine is operating.

All reading, processing, and writing must be done according to an
established procedure. Control over the entire system is maintained by placing instructions to the machine in storage in exactly
the same form as data. Each detailed step of the procedure is
coded as an instruction and then placed in known storage locations. These locations must be apart from the transient data or
constant factors.
A complete procedure can be developed from individual instructions if they are grouped in proper sequence as a program.
Instructions in the program direct the computer to produce desired results. This stored program controls all operations of internal processing as well as those of the input and output devices.
Once stored, the program can be retained indefinitely during normal system operation. Thus, storage is a type of artificial memory

Stored Program

336

COMPUTER CODES AND MAIN COMPONENTS

that can "remember" instructions and recall them as required.
The number of instructions that can be stored, and the consequent
length of the program, depends upon the capacity of storage.
When a particular task is finished, new instructions can be
loaded into storage, replacing the program previously retained.
Any of the standard input devices can be used for this purpose
because instructions take the same form as data.
The possible variations of the stored program provide a computer with almost unlimited flexibility. A single machine can
execute a great many different procedures by simply loading the
appropriate program into memory. Most important, because instructions are stored in the same form as data, the machine can
operate upon its own instructions. This basic concept is a most
remarkable departure from the operating principles of all other
types of previously developed calculating or information-handling
equipment. It enables the machine to alter its own program in
response to conditions encountered while following a procedure.
Consequently, a computer can be programmed to exercise a limited degree of judgment, but always within the framework of the
operations it can perform and as directed by its human operators.

Central Processing
Unit

The central processing unit (CPU) controls and supervises the entire machine system. It performs the actual arithmetic and logical
operations, and in a real sense it is "the computer." From a functional viewpoint, the CPU is made up of two sections: control
and arithmetic/logical.
The control section has the various facilities required to direct and coordinate all operations called for by instructions in.the
stored program. These operations include addressing main storage
to locate, fetch, and store data or instructions, and initiating communications between storage and the external devices. By means
of the control section, automatic, integrated operation of the entire computer system is achieved.
In many ways, the control section can be compared to a
telephone ex.change. All possible data paths already exist, just
as there are connecting lines between all telephones serviced by
the exchange.
The telephone exchange has the equipment to carry sound
pulses from one phone to another, ring the phones, connect and
disconnect circuits, and the like. The path of communication be-

COMPUTER CODES AND MAIN COMPONENTS

337

tween one telephone and another is set up by appropriate controls
and mechanisms in the central exchange. In a computer system,
the control section can respond to an interruption from some
source, start or stop communication with an input or output device, turn a signal device on or off, fetch or store data and instructions, sequence instructions in the desired order, and so on.
The arithmetic/logical sections contain the necessary circuitry
to perform arithmetic and logical operations. The arithmetic section calculates, shifts numbers, sets the algebraic sign of results,
rounds, compares, and the like. The logical section carries out the
decision-making operations to change the sequence of instructions
as may be required. These include comparison, translation, editing,
testing indicators, and so forth.

A register is another type of storage device capable of receiving
some specific unit of information, holding it, and then transferring
it as directed by controlling circuitry. Various electronic components may be used, such as transistors, depending upon what
function the register is designed to perform. Registers are normally divided into positions or places, each of which can indicate
a binary one or zero.
Some types of registers and their functions are:
1. An accumulator holds the results of a calculation. When
sequences of arithmetic operations are performed, the results in an
accumulator are often used from one step in the procedure to the
next. For example, two fields may be added to produce a sum in
the accumulator. The sum may then be multiplied by another
field and the product added to still a fourth field. The final result
can remain in the accumulator until transferred to storage for
output.
2. A storage register contains data taken from, or being
sent to, storage. This type of register serves as a place to store
data temporarily as it passes between the processing unit and
main storage.
3. An address register holds the designation of an input or
output device or the address of a storage location.
4. An instruction register contains the instruction being interpreted or executed.

Functional Units

338

COMPUTER CODES AND MAIN COMPONENTS

5. General registers may hold several types of data. Register
A may contain one factor to be used in a calculation, and Register
B the second factor. The two factors can be combined and placed
in Register C.
The contents of some registers can be shifted to right or left
within the register and in some cases between registers. As the
contents are shifted, places vacated can be automatically filled
with zeros. The contents of a register can be masked in various
ways so that individual binary positions can be either selected
or ignored.
Registers may temporarily hold information while associated
circuits analyze the data. For example, an instruction can be
placed in a register and associated circuits can determine what
operation is to be performed and locate the data to be used. Data,
such as single characters, bytes, or other data segments, may be
placed temporarily in registers to check validity of coding or data
representation.
In the computer, all data is actually represented with binary
indications of ones and zeros.
Usually, the more important registers in a computer system,
particularly those· involved in normal data flow and storage addressing, have small incandescent or neon lights associated with
them. These lights are mounted systematically on an operating or
maintenance console where they show register contents and various
program conditions. With these lights, an operator or maintenance
man can "see" into the machine and determine, if necessary, just
how instructions are being performed. During this kind of operation, the machine is manually cycled or otherwise "slowed down"
so that detailed action of the various registers can be followed.
When the system runs at normal speed, the lights usually have no
significance but display the blinking effect often shown on news
or entertainment media to dramatize the complexity of computer
operation.
Console lights are also used in diagnostic and maintenance
procedures.

Counters

An adder receives data from two or more sources, performs addition in electronic circuitry and sends the results back to some re.,
ceiving register. Binary carries to higher order positions are made

COMPUTER CODES AND MAIN COMPONENTS

339

automatically. The final sum goes to corresponding positions of
the receiving register.

All operations of the entire computer system take place in fixed
intervals of time. These intervals are measured precisely by regular
pulses emitted from an electronic clock at extremely high frequencies, perhaps several million pulses per second. Each machine cycle
lasts for a fixed number of pulses.
Time references are stated in milliseconds, microseconds, and
nanoseconds. The following table lists these terms with their abbreviations and decimal and fractional equivalents.
Decimal
Equivalent

Fractional
Equivalent

Term

0.1

1
= -second
10

= 100 milliseconds

0.001

1
= 1,000 second

= 1 millisecond (msec)

0.000001

1
1,000 ,000 second

= 1 microsecond (Jlsec)

1
0.000000001 = 1,000 ,000000
, second = 1 nanosecond (nsec)

It is sometimes difficult to understand such ~xtremely short
intervals of time unless they are compared with some known refence. For example, "quick as a wink" is actually about one-tenth
of a second, or 100 milliseconds. A jet plane traveling at 600 miles
per hour covers 10 miles in one minute, 880 feet in one second,
0.88 of one foot in one millisecond and approximately 0.01 inch
in one microsecond. Light travels at a speed of 186,282 miles per
second, about 982 feet per microsecond, and slightly less than one
foot per nanosecond. Or, a nanosecond is about the time it takes
the light from this page to reach the reader's eye.
Within the interval of a timed machine cycle, the computer
can be programmed to perform a specific operation. The number
of individual operations required to execute a single instruction
depends upon the instruction. Thus, various machine operations
may be combined to execute a single instruction.
Instructions usually consist of at least two parts: an operation
and an operand. The operation part tells the machine which basic

Machine Cycles

340

COMPUTER CODES AND MAIN COMPONENTS

function to perform: read, write, add, subtract, and so on. The
operand can be the address of data in storage. Therefore, a com.;.
plete add instruction includes both the function of adding and the
specification of what data is involved. If the operation is "write,"
the operand will be the location in storage of the information to
be written and the amount of data to be transmitted to the previously selected output device.
Instructions also cause the machine to perform a control
function such as shifting a quantity in a register to prepare for
division or testing an indicator to determine· if a coding error has
been detected.
In order to receive, interpret, and execute instructions, the
processing unit must operate in a prescribed sequence as determined by the specific instruction and carried out during a fixed
interval made up of timed pUlses.

Program Execution

Program instructions are always placed in known locations of storage. Each addressable location may contain just one instruction
or, depending upon the computer, an instruction may occupy
several locations. If instructions are variable in length, a part of
the instruction must indicate exactly how many storage positions
are occupied. In any case, instructions are stored in exactly the
same manner as data.
Instructions are placed in storage in the sequence in which
they are normally executed and in ascending locations. That is,
the first instruction of the program may be in location 000 1, the
second in location 0002, the third in 0003, and so on. Or, the first
instruction can just as well be placed in location 5421, the second
in 5422, and so on, in any convenient space. But in order to be
executed, the program must always be in storage.
When machine operation begins, an instruction counter is set
to the address of the first instruction. To begin processing, this
instruction is fetched from storage and decoded in a register. Then,
while it is being executed, the counter automatically advances or
steps to the address corresponding to the location of the next
stored instruction. In this way the instruction counter acts as a
pointer, directing the processor to each instruction as it is required.
If an instruction occupies one position of storage, the counter
steps one. If an instruction occupies five positions, the counter
steps five. In either case, by the time an instruction is executed,

COMPUTER CODES AND MAIN COMPONENTS

the counter has located the next instruction in sequence. The
stepping action of the counter is automatic. The result is that
when the computer is directed to any series of instructions in any
storage locations, it proceeds to execute them one after another
until instructed to do otherwise.
For example, assume that a program is stored beginning in
location 0142. To start operation, the instruction counter is set
to address 0142. The first instruction is fetched from storage and
brought to the processor. The operation part of the instruction
is decoded. In this way the machine is told what operations to
perform: add, subtract, compare, etc. Refer to Figure 16.6.
The operand part is placed in an address register. This tells
the machine where in storage to locate the factors to be operated
upon. The instruction operand may also indicate where the addition is to be performed, since at least two factors are always involved to produce a sum. The operation of addition is always designed to occur in a specific register or accumulator. The location
of the second factor in the addition is therefore predetermined
and need not be specified.

Storage

<...

ADD 0002 (complete instruction)

l
(operation part) ADD ~

Storage Register

.
r

0002 (address part)

Instruction Register

I nstruction Counter
Figure 16.6 Instruction decoding.

The next instruction is located. This instruction may return
the sum just developed to storage or fetch and add another factor
to the sum. When arithmetic operations are complete, instructions
may start an input or output device to send or receive data.

341

342

COMPUTER CODES AND MAIN COMPONENTS

Execution of instructions does not have to proceed sequentially. That is, a control instruction can act directly upon the instruction counter. Such an instruction only directs the machine
system; no data is processed. Since the counter is capable of addition, it can be incremented to point to any location of storage
by simply adding a constant or variable factor to the address contained by the counter.
The counter can also be reset to the address of the beginning of the program so that all· instructions can be repeated for
another group of incoming data. Thus, the execution of a program
can be repeated any number of times as records are read serially
into storage.
The counter can be reset to zero or to any storage address by
a branch instruction. The program can be arranged so that branching is conditional or unconditional.
For instance, the branch or transfer back to the beginning of
the program should occur only if there are more records to process.
However, if the input device has fed all records and indicates an
end-of-file condition, then the program should branch to a control instruction to terminate or halt operation.
The computer can also be directed to examine some indicating
device and then branch if the indicator is either on or off. Such an
instruction has the effect of saying, "Look at the sign of the quantity in the accumulator. If the sign is minus, take the next instruction from location 5000. If the sign is plus, continue with the next
instruction in sequence." The instruction is set in one of two possible location addresses, 5000 or the address of the next instruction.
At location 5000, a special series of instructions, or subroutine,
will direct the computer to carry out special procedures for the
handling of minus or credit balances. When that subroutine
is completed, a transfer is effected back to the main program.
Thus, the logical path through a program followed by the
machine may be controlled either by unconditional branching or
by a series of conditional tests applied at various points. However, the actual arrangement of instructions in storage is not
normally altered.

Questions

and
Exercises

1. Explain the structure of the BCD code.
2. Explain the concept of the storage address.
3. Name three kinds of data normally contained in main storage.

ORGANIZATION OF
A UNIT RECORD COMPUTER:
IBM SVSTEM/3 MODEL 10

The elements of any computer system, including the System/
3, are each designed to perform specific functions pertaining to
data input, processing, data output, and control.

Card Input

The 96-column card serves as the principal input medium for the
System/3 Model 10. Two card feeds, primary and secondary, are
provided in a multifunction card unit (MFCU). Cards are placed
in the hoppers facedown, left end toward the machine. Each feed
can be controlled to operate independently of the other. Cards are
directed to the read, wait, punch, and print stations by transport
mechanisms as shown schematically in Figure 17.1.
When a card is fed from one of the hoppers, it passes first
through the read station where light shining through the punched
holes in each column is converted to electrical energy by an array
of solar cells. There is one solar cell for each punch position.
Reading is done serially, column by column, three tiers at a time,
from the left end of the card. That is, columns 1, 33, and 65 are
read; then columns 2, 34, and 66, and so on. The data in all 96
columns is transmitted to internal storage where it is held for
processing. Input data is converted automatically from card code
representation to the internal electronic form usable by the processing unit.
344

COMPUTER CODES AND MAIN COMPONENTS

4. Can information in storage be used more than once? Why?
5. What are the two main functional units contained in the central processor?
6. Describe the functions of the following: accumulator, storage register,
counter, adder.
7. Name the two parts of a computer instruction.
8. Name the following time intervals in seconds: 0.2; 0.004; 0.000001;
0.000000002.
9. Explain the general concept of the stored program.
10. How is a program executed?

343

ORGANIZATION OF A UNIT RECORD COMPUTER

Secondary
___ Hopper

Figure 17.1 Schematic. IBM Multifunction Card Unit.

Input data entering the system is checked to insure reading
accuracy. Each set of three tiers is read twice and both readings
must agree. Disagreement between the two readings produces a
read check. The hopper that fed the error card can be determined
from indicator lights on the MFCU control panel. The incorrect
card can be located at the primary or secondary wait station.
The speed of card feeding and reading is determined by the
operations performed. The rated speeds of 250 cards per minute
(Model AI) or 500 cards per minute (Model A2) are for reading
operations only. If punching or printing is to be performed, the
reading rate will be reduced to correspond with the slower of these
two operations.
If required, cards may be fed through the reading station
without being read. The reading sensors are not actuated and
the cards pass through the station without transmitting any data
to storage.

345

346

Card Output

ORGANIZATION OF A UNIT RECORD COMPUTER

After a card passes the read station of the MFCU, it enters a wait
station. Note that there is an upper wait station for the primary
feed and a lower wait station for the secondary feed. The card
then passes to a punch station where it is punched column by column, three tiers at a time. All punched output data is transmitted
from storage and is converted automatically from the internal system representation to punched card code.
Punching each hole in a column generates a signal that represents that hole. After a column has been punched, all the signals
for holes in the column are compared with the character code
specified for that column. If the two do not match, a punch check
occurs. Data to be punched from storage is also checked to insure validity within the 64-character set allowed in the card coding
structure. An error causes a punch invalid check. The card that
contains the incorrectly punched data is in the stacker previously
selected at the start of the punching operation. Card punching is .
done at the rate of 60 cards per minute (Model AI) or 120 cards
per minute (Model A2). The unit can also be controlled to feed
cards through the punch station without punching.
Output cards can be stacked in any of the four stackers .. If
there is no stacker selection, cards from the primary hopper will
be placed in stacker I; cards from the secondary hopper will be
placed in stacker 4.
When collating operations are performed, one file of sequenced cards is placed in the primary hopper and the other in
the secondary. Cards can he merged, matched, and selected to
simulate the operation of the 80-column collator described earlier.
To either reproduce or gangpunch, blanks are placed in one
hopper of the MFCU, original cards in the other. Under control of
the processing unit, cards can be reproduced and gangpunched as
required by the procedure.
In file maintenance operations, master cards can be fed from
one hopper while detail cards are fed from the other. Operations
of matching and merging can be combined with calculation, gangpunching or reproducing, and total accumulation.
If summary punching of new balances is required, masters
and detailS are merged first in one operation using the two feed
hoppers of the MFCU. In a second operation, the merged files
are placed in one hopper and blanks in the other. The MFCU can
then be controlled to summary punch into blank cards. In this
case, the System/3 can combine the operations of calculating,

ORGANIZATION OF A UNIT RECORD COMPUTER

347

total accumulation, summary punching, and report printing in
one operation.

From the punch station, cards pass through a cornering station to
a print station equipped with a series of printing wheels. As many
as four lines of 32 characters each can be printed on one card.
The MFCU prints all of the possible 64 characters in the 96-column
card code. Data to be printed is transmitted from storage in the
processing unit and mayor may not be punched iIi the card. Cards
not printed are ejected through the station to the stackers.
Card throughput, for up to three rows of printing, is 60 or
120 cards per minute. There is a reduction in the rated speed if
four lines per card are printed.
Two checks are made on printing accuracy. An error in the
synchronization between the print wheels and the MFCU attachment circuitry causes a print data check. The check insures that
the wheel has been positioned to the proper character as transmitted from storage. An error in synchronization between the
MFCU circuitry and a printer stepper clutch causes a print clutch
check. This check insures the proper positioning of the print lines.
The error card is fed to the stacker designated before the start of
the printing operation.

Printed Card Output

A line printer is provided with the System/3 to produce reports at
high speed as output from the computer. All information to be
printed is transmitted by the processing unit from internal storage
where each line of data is arranged in a predetermined format.
The basic printer can place as many as 96 characters on one line,
spaced 10 characters to the inch. The user can obtain an optional
feature from IBM which extends the 96-character capacity to either
120 or 132 positions in a single line.
Forty-eight different characters can be printed by the standard machine: 10 numerical, 26 alphabetic, and the 12 special
characters. + & $ * -/ ,% # @ '. No type wheels or bars are used.
Instead, multiple arrays of type slugs are placed end to end to form
a loop or chain. There are two characters on each slug. When the
printer is operating, the chain moves in a horizontal plane across
the blank paper form as shown in Figure 17.2. The chain is driven

Line Printer Output

348

ORGANIZATION OF A UNIT RECORD COMPUTER

Idler Pulley
Type Slug

Figure 17.2 Print chain and paper form.

continuously over two sprockets, one on each side of the printing
unit. Each chain has a total of 120 slugs or 240 character positions
and the standard set of 48 characters is repeated five times. The
first array of slugs is depicted below as they appear on the chain.

Printing production is either 100 lines per minute (Modell) or
200 lines per minute (Model 2).
The usable character set can be extended from 48 to as many
as 120 different characters by means of a universal character set
feature. For example, the printer can be controlled to print both
uppercase and lowercase alphabetic characters. However, when
this feature is used, the line production speed is reduced somewhat, depending upon the text being printed.
Printing is accomplished by striking the paper from the back
with small hammers, thus forcing the paper forward against the
type. An inked ribbon between the paper and the type chain
, transfers the character impressions to form the printed line. The
rotation of the chain and the action of the hammer mechanism is
synchronized in such a way that type slugs are struck in the proper

ORGANIZATION OF A UNIT RECORD COMPUTER

349

positions to exactly duplicate the image of the line in storage.
Thus, the arrangement of the line is completely flexible. Unlike
punched-card accounting machines, mechanical zero suppression
is not required. Printed output does not need to be within a set
columnar or tabular format. The chain printer can produce a
free-form printed page in exactly the same manner as a typewriter.
The entire print chain cartridge can be removed from the machine by the operator and replaced with another cartridge. This
interchangeable type makes it possible to use a number of different
character sets or fonts on one machine.
The movement of the paper forms is also controlled from the
processing unit. Maximum length of a single sheet is 14 inches.
This is 112 lines at eight lines per inch, and 84 lines at six lines per
inch. The 6- or 8-line vertical spacing can be selected by the operator. Spacing between lines and skipping to a predetermined line
on a page are controlled from the processing unit.
As another optional feature, a dual-feed carriage enables the
printer to produce two side-by-side documents at the same time.
Each document can be independently controlled.
A number of checks are made automatically to insure printing accuracy, including those on the action and synchronization of
the printing hammers and' chain. A print check error is indicated
when the hammer circuitry does not respond properly to a print
signal. A thermal check is made to prevent the hammer unit from
overheating. Other indicators show a forms jam, carriage slipping
or spacing too far, or various interlock conditions in the chain and
forms chute.
An indicator also shows if a character that is to be printed is
not available on the chain. This checks that the proper type cartridge is on the machine.

The processing unit of the System/3 is divided into three main
parts:
Storage
Control
Arithmetic/logical

Processing Unit

The basic storage size for the system is 8,192 eight-bit words
or bytes. Additional storage capacity can be obtained in incre-

Storage

350

ORGANIZATION OF A UNIT RECORD COMPUTER

ments of 4,096 bytes up to a total of 32,768 bytes. Storage
capacity for computers is often referred to in increments of "K"
meaning 2 10 or 1,024. Therefore, basic System/3 storage is 8K
(8 X 1024 = 8,192) bytes with additional sizes available as 12K,
16K, 24K, and 32K.
In addition to the eight data bits, each byte contains one
parity bit to maintain an odd number of one bits per byte. Each
time a byte of data is used, it is checked to insure that it contains an odd number of one bits, including the parity bit. If an
even number of ones is detected, the processing unit indicates a
process check.
Although the parity bit is really an inherent part of the internal coding system, the bit is "invisible" to the programmer. It
represents no data and cannot be manipulated in any way by instructions. Capacities are stated and thought of in terms of eightbit data representation, but it is understood that the byte always
includes the ninth parity bit when used within the machine system.
Data is stored in extended binary coded decimal interchange
code (EBCDIC). Numeric data can also be handled as zoned decimal format where each byte is considered to be divided into two
groups of four bits each. Bits through 3 make up the zone portion; bits 4 through 7 lllake up the digit portion as shown in Figure 17.3. Whenever data is handled in this format, the zone bits
do not participate in any arithmetic operations.

Byte

Ipl I I I I I I I )pl I I I I I I I)
~~~~
Zone

Digit

Zone

Digit

Figure 17.3 Zoned decimal format of data storage.

A group of bytes may form a field in adjacent locations of
storage just as adjacent card columns may form a punched card
field. When the field is numeric and in zoned decimal format, the
zone bits of the low-order byte indicate the sign of the field: 1111
for plus, and 1101 for minus.
Data can also be handled in binary format as an eight-bit
binary integer as shown in Figure 17.4. Note that all data in
storage is in bytes of eight binary bits. The format of instructions
executed by the processing unit determines whether the data is
treated as zoned decimal, graphic characters, or binary integers.

ORGANIZATION OF A UNIT RECORD COMPUTER

351

Each of the 96 columns of input card records can contain one
six-bit character. During input operations, card data is converted
automatically to EBCDIC. During output operations, EBCDIC is
converted back to card code or to printed graphics.
Byte

Ip[ , , ,
,0

I , , ',pi' , , f, , , I

f
3

4
Y

Binary integer

Figure 17.4 Binary integer format of data storage.

Addresses of storage locations are consecutively numbered in
hexadecimal from 0000 to the upper limits of the largest size storage. Any field or group of bytes can be located by addressing
either the leftmost or rightmost (lowest or highest address) byte,
depending upon the instruction.

Figure 17.5 is a simplified schematic of the System/3 processing
unit data flow. Reference numbers on the illustration correspond
to the item numbers below, as follows:
1. Information enters the processing unit by way of the channel and is translated automatically to EBCDIC. A parity check is
made, byte by byte, on the code translation.
2. Incoming data is routed to storage by the control section
and placed in the locations designated by an I/O instruction.
3. The storage data register stores data temporarily as it
passes between processing and storage.
4. The A and B registers serve as temporary storage for data
to be processed by the arithmetic/logical unit. Each of these registers holds one byte of data and each can be entered from several
other units in the data flow.
5. The storage address register holds the address to be accessed in storage.
6. The condition register stores one byte. Each bit indicates
the result of specific processing operations. For example, bits are
placed in the condition register to indicate a high, low, or equal

Processing Unit
Data Flow

352

ORGANIZATION OF A UNIT RECORD COMPUTER

AAR
XR1
XR2
~CR-DRR

~M~~:R

1/0 Channel

XRDAR
LPDAR
LPIAR
MPDAR
MRDAR
PSR

® - Parity Check
PG

- Parity Generate

_ _ - ALU Data Buss Out
~- LSR Data Buss In

====:>-Misc Data Buss
- - . - Control Buss

Figure 17.5 Schematic. System/3 Model 10 Data Flow.

comparison. After an arithmetic operation, bits may indicate that
a binary or decimal overflow has occurred. The assignment of each
register position to indicate a condition is as follows:
Bit

Name

7
6
5
4
3
2
I

Equal
Low
High
Decimal overflow
Test false
Binary overflow
Unassigned
Unassigned

ORGANIZATION OF A UNIT RECORD COMPUTER

Instructions can test the register for these conditions and
direct the machine to execute corresponding program routines
accordingly.
7. The op code register holds the operation code byte of each
instruction taken from storage. When an operation is decoded, the
control cireuitry is actuated to perform that specific function.
For example, if the operation code is hexadecimal 06 (0000 0110),
the arithmetic/logical unit is directed to add two fields and place
the sum in storage. If the operation code is hexadecimal 07 (0000
DIll), the unit is directed to subtract one field from another and
place the remainder in storage. The addresses of the fields are
designated in the operand part of the instructions.
8. The Q register holds a special Q byte from the instruction
that is to be executed. The byte is read into circuitry to control
the operations performed by instructions. For example, in the
. operations of add and subtract zoned decimal, the Q byte designates the length of the fields that are to be operated upon.
9. There are 16 local storage registers in the basic card system
that also contain data required for the execution of various instructions. The function of these registers will not be explained here.
10. The arithmetic/logical unit performs all arithmetic and
logical functions on the processing unit. It is capable of decimal
add and subtract, binary add and subtract, compare, and other
data manipulating functions.
Note that at various points in the data flow, parity checks
are taken as the data moves through the processing unit. Also, as
new data is generated, parity is generated at the same time. That
is, when a sum or remainder is returned to storage, the parity portion of the byte is generated and returned also.
Each operation performed is done in two phases: instruction
and execution. Some instructions combine the phases so that there
is no distinct execution phase.
The time interval in which the processing unit reads one byte
from storage is known as a cycle. At least three cycles must be
performed for every instruction. Cycles can also be taken by the
input and output units in sending and receiving data.
The reservation of storage areas for either data or instructions
is originally done by the programmer. Actually, the processing
unit cannot distinguish between data and instructions. If by mis-

353

354

ORGANIZATION OF A UNIT RECORD COMPUTER

take the instruction counter is pointed to data, the machine will
erroneously attempt to operate according to whatever information
is placed in the operation code register. Improper functioning
of the computer would undoubtedly result and quickly halt
the system.
Conversely, however, instructions can legitimately be treated
as data. For example, in item 7 above, the operation code for add
zoned decimal was given as hexadecimal 06, while the code for
subtract was given as hexadecimal 07. Therefore, by simply adding
a binary I to the code for addition, the operation is changed to
subtract. Or, by taking I from the subtract code 07, the remainder
becomes the operation code for add. In a program, instructions
can be given to test the condition register for the sign of a result.
If the sign is plus, the machine may be instructed to subtract a
given factor to obtain some final result. Or, if the sign is minus,
the machine may be instructed to add the factor to reach the final
result. That is, the program can instruct the computer to alter
other instructions to either add or subtract a factor, depending
upon the sign of the previous result.
The processing unit also acts as a controller for all I/O devices.
The unit communicates with the devices by way of an interface
called the input/output channel. Each device attached to this
channel has an address in the same manner that each storage location in the processor has an address. The device can thus be identified and activated by instructions in the stored program. The
I/O channel consists of:
1. A set of signal lines that carry information to and from
the processing unit;
2. Logic to translate card code into EBCDIC and vice versa;
3. Logic to time share I/O operations with processing and to
interrupt the main program for I/O servicing, as required.

Time Sharing

In this mode, internal data handling can be carried out while input
or output information is being routed between an I/O device and
main storage. The I/O device is directed to start operations by a
specific instruction and is then left to perform its operation until
storage is required. The device then steals one or more cycles from

ORGANIZATION OF A UNIT RECORD COMPUTER

355

other processing unit operations in order to store or retrieve data.
The processing unit then continues to perform other operations
until the I/O device again requires storage cycles. This interleaving
of functions is possible because the processor can move data into
storage in a small fraction of the time required to complete an I/O
cycle, such as feed a card or print a line.
An optional dual programming feature provides the system
with the capability to execute two independent programs on a
time-sharing basis. The feature allows the processing unit to transfer to a different program when the current program must wait for
completion of an I/O operation. This uses the high performance
capabilities of the processing unit rather than forcing it to wait for
completion of the execution by active I/O devices. The feature
allows two independent programs to be placed in storage at the
same time.

Normally, the interrupt capability of computers implies that the
processing unit can interrupt a currently operating sequence of
instructions to perform some other instruction sequence required
by an I/O device, and then return to the main or interrupted program. For example, certain I/O units require special program subroutines to handle data to be entered within a limited period of
time. Such a situation can occur when a communication device is
ready to send messages to the computer and may require servicing
before a current sequence of instructions has been completed. The
System/3 processor is equipped with an interrupt capability to
service an input or output unit whenever the device has information to transmit or is ready to receive.
When interrupted, the system is so designed that:
1. Information is saved to enable the machine to restore itself to the same status it had before the interruption occurred.
This may include 'the status of the condition register, storage registers, and so on.
2. Information is saved to enable the processor to return to
the program at the last instruction to be executed from the point
where the interruption occurred. This means that after completing
the subroutine, the instruction counter must be reset to the address of the next instruction in the main program.

Interrupt

356

Instruction Format

ORGANIZATION OF A UNIT RECORD COMPUTER

The System/3 is designed to perform 28 different operations including decimal addition and subtraction, binary addition and subtraction, moving and editing data, testing and changing the contents
of registers, controlling the I/O devices, comparing, branching, and
others. The functions of these· operations are complementary in
such a way that the system can meet nearly all data handling requirements of a general-purpose commercial machine. Each possible operation that can be performed is coded as an eight-bit byte.
It is interesting to note that the arithmetic operations of
division and multiplication are not included in the operation set.
However, by programmed subroutines, the machine can divide and
multiply just as though this calculating ability were built into the
circuitry. It is a common characteristic of computers that the operation set is generally designed to include only basic functions
while more specific or specialized functions must be sim ulated by
programming.
In the System/3, the operation code is combined with a Q
byte and an operand to form an instruction. The operation code
of the instruction always tells the processor what to do; for example, add zoned decimal, test, branch, etc. The operand of the
instruction normally tells the processor where in storage to find
the data upon which to operate. Therefore, an operand is usually
one or more storage addresses. The Q byte serves to further define
the instruction by various means, such as indicating the number of
bytes in the operand, the address of an I/O device, registers to be
used in the operation, and so on.
Figure 17.6a shows the format of an instruction that will
clear a designated area of storage (Operand 1) and replace it with
a specified zoned decimal field obtained from another area of
storage (Operand 2). The execution of the instruction has the
same effect as a clear and add operation in a desk calculator or
adding machine. In the System/3, the operation is called Zero and
Add Zoned. The instruction can be used as the first step in preparation for accumulating quantities in some record field in storage.
Since two storage areas are involved, the instruction contains
two operand addresses and each address is two bytes in length. The
address of Operand 1 is shown as hexadecimal 10, the location of
the units position of the numeric field. The address of Operand 2
is shown as hex 20, also the location of the units position. The two
operands as they appear in storage are shown in Figure 17.6b, c.

ORGANIZATION OF A UNIT RECORD COMPUTER

The Q byte in this instruction format designates the length of
the two operands involved. Positions 4 through 7 of the byte
designate the number of bytes in Operand 2, minus one. Since
Operand 2 is actually three bytes long, positions 4 through 7 contain a hex 2. Positions 0 through 3 of the Q byte designate the
number of bytes by which the length of Operand 1 exceeds the
length of Operand 2. Since Operand 2 is five b'ytes in length, positions 0 through 3 also contain the quantity 2. For this operation
in the System/3, the maximum length of Operand 1 is 31 bytes
and the maxim urn length of Operand 2 is 16 bytes. These restrictions are imposed by the design of the system.
Instruction
(a)

04
Op code

22
Q

Operand 2

Operand 1

Byte

Operand 1 before instruction execution
(b)

Data

oooe

0000

OOOE

OOOF

0010

Address

Operand 2 before instruction execution

(c)

Data

DOlE

001F

0020

Address

Operand 1 after instruction execution
(d)

Data

oooe

0000

OOOE

OOOF

0010

Address

Figure 17.6 System/3 instruction format. Zero and add zoned.

The operation code for Zero and Add Zoned is coded as 04
in hexadecimal, indicating that the eight-bit byte in storage would
be indicated as 0000 0100.

357

358

ORGANIZATION OF A UNIT RECORD COMPUTER

When the instruction is executed by the processor, Operand 2
is placed byte by byte into the storage locations of Operand 1.
Extra high-order zeros are inserted into those positions by which
Operand 1 exceeds Operand 2 in length. The zone bits in each
byte of the result are all set to ones, except the units position.
This byte carries the sign of the result field: 1111 if positive, 1101
if negative. A result of zero is considered positive.
The result as it would appear in storage is shown in Figure
17.6d. Note that all zoned decimal digits appear with binary ones
in the zone portion of the byte. Therefore the hex symbols for the
result field are FO, FO, F4, F2, and F5.
Operand 2 remains in storage unchanged (unless the two fields
overlap), and can be used in any other data handling operations
that may be executed later on in the program. The condition
register is set as follows:
Bit

Name

Setting

7
6
5

Equal
Low
High

Result is zero
Result is negative
Result is positive

Bits 4 through 0 are not affected. In this example, the fivebit position of the register is set to 1. When a result is zero both
the 5- and the 7-bit positions are set to 1.
To continue accumulating in the field designated as Operand
1, an operation called Add Zoned Decimal can be performed. The
instruction is the same as that of the previous example, as shown
in Figure 17.7 a. Operand 1 remains in the same storage position
with an address of hex 0010. The address of the new Operand 2
is hex 0020. Refer to Figure 17.7b, c.
The operation code for Add Zoned Decimal is hex 06. The
Q byte remains the same since Operand 1 is five bytes in length
and Operand 2 is three bytes. The maximum number of bytes,
and consequently the size of the fields that can be added together,
also remains the same.
When this instruction is executed, Operand 2 is added algebraically to Operand 1, byte by byte, and the result is stored in
the location of Operand 1. The operand fields are addressed by
the units position bytes. The zone bits in each byte of the result
are all set to ones, except the units position. This byte carries the
sign of the result field: 1111 if positive, 1101 if negative. A zero
result is considered positive. The result of the addition is shown in

ORGANIZATION OF A UNIT RECORD COMPUTER

Figure l7.7d as it would appear in storage. Ope:rand 2 remains in
storage unchanged (unless the fields overlap). The condition register is set as follows:
Bit

Name

Setting

7
6
5
4

Equal
Low
High
Decimal overflow

Result is zero
Result is negative
Result is positive
Carry occurred from the highorder position of Operand 1

Bits 3 through 0 are not affected. In this example, the 5bit position of the register is set to 1.
Instruction

(a)

06
Op code

Operand 2

Operand 1

Byte

Operand 1 before instruction execution
(b)

. F2

Data

oooe

0000

OOOE

OOOF

0010

Address

Data

Operand 2 before instruction execution
(c)

Data

002E

002F

0030

Address

Operand 1 after instruction execution

(d)

oooe

0000

OOOE

OOOF

0010

Address

Figure 17.7 System/3 instruction format. Add zoned decimal.

To carry out these arithmetic operations, any area of storage
can be used as an accumulator for decimal quantities. An operation of Subtract Zoned Decimal (hex code 07) can also be performed to arrive at a net balance in Operand 1.

359

360

ORGANIZATION OF A UNIT RECORD COMPUTER

The following formula is given to calculate the time required
by the Sytems/3 to execute the two instructions described above:
Time in j.JSec = 1.52 (N

where

+ Ll + L2) + 1.52 (R)

N = instruction length in bytes
L1 = length of destination field in bytes (destination
field is that field addressed by Operand 1)
L2 = length of source field in bytes (source field is that
field addressed by Operand 2)
R = length of Operand 1 in bytes when recomplementing is necessary (not required in above
examples)

Substituting in the formula:
Time

Printer-Keyboard

= 1.52 (6 + 5 + 3) + 1.52 (0)
= 21.28 flsec or, for the two instructions
= 42.56 flsec

The System/3 Modell 0 can be equipped with a printer-keyboard
which is intended to serve primarily as an inquiry station and secondary printer. However, the device can also be used for direct
data entry into the system and for communication between operator and machine. The unit is conveniently mounted on the console
work table with a forms stand on the floor behind it. Although
similar in appearance to an electric typewriter, the keyboard and
printer operate independently of each other under stored program control.
The keyboard will accept data whenever a "proceed" light is
on. Key depressions cause the corresponding character to be transmitted to processor storage on an interrupt basis. However, because the keyboard and printer are not physically linked, key
depressions do not automatically cause a character to be printed.
There are 44 character keys as well as shift, space, and carriage return keys, all arranged in typewriter style (Figure 17.8).
Functional keys and indicator lights are located to the right of the
keyboard for the operator's use.
The printer operates at a rated speed of 1.5 characters per
second. Printing is serial, character by character, across a

l2~ -inch

ORGANIZATION OF A UNIT RECORD COMPUTER

361

writing line at 10 characters per inch. All printing is transmitted
from processor storage.

Figure 17.8 System/3 keyboard character arrangement.

This unit is used with the processing unit and the MFCU for on-line
data recording and data verification. Mounted on the console work
table, the keyboard operates under stored program control.
The keyboard has the touch and format of the 5496 Data
Recorder, making operator transition from one device to the other
quite convenient. The 64-character set is equivalent to that of the
5496. Card punching, reading, and column-for-column printing is
provided by the MFCU. Completed cards are placed in the fourth
stacker of the MFCU where they are accessible to the operator
from a seated position at the keyboard.
The 96-column cards can be punched and verified at the keyboard in exactly the same manner as carried out with a 5496.

Data Entry
Keyboard

New users of the System/3 often change from existing 80-column
punched-card equipment. In these cases there is a requirement to
convert existing 80-column card files to 96-column cards, particularly master files and other permanent records. To meet this need,
IBM makes available an 80-column card read punch for attachment
to the System/3. Not standard equipment, the device is normally
used on an IBM system that may be rented on an hourly basis in a
Basic System Center. Once the files are converted, the user usually
does not require regular 80-column input in his application. The
one-time conversion of files can be done in advance of actual installation of the new system.
The 80-column files are read by a special card reader and are
punched into 96-column cards by the MFCU. The operation is
under control of a stored program available from IBM.

SO-Column Card
Compatibility

362

Questions

and
Exercises

ORGANIZATION OF A UNIT RECORD COMPUTER

1. Name the four card stations of the MFCU. What function is performed
at each station.
2. How is the MFCU similar to the 88 Collator in card handling?
3. How is the MFCU similar in function to the 519 Document-Originating
Machine?
4. How are files arranged in the MFCU for summary punching?
5. What takes the place of a control panel in the MFCU and the printer?
6. What purpose does the printing function of the MFCU serve?
7. List some advantages of the printing capability of the System/3 line
printer compared to a punched-card accounting machine.
8. How are paper forms positioned and controlled on the printer?
9. How are the printer and carriage controlled?
10. How does the processor distinguish between instructions and data?
11. What is an 110 channel?
12. Why is the "interrupt" capability important?
13. What are the three main parts of the instructions presented in this chapter and what is the function of each?

IBM SVSTEM/3 MODEL 10:

CARD- AND DISC-ORIENTED
SYSTEMS

Using the configuration of equipment described thus far, the
System/3 Model 10 can perform the data processing operations
previously done by single-purpose unit-record machines. A number of these operations can often be combined in one run of the
cards through the machine.

The System/3 is controlled for each type of procedure by a stored
program of instructions in processor storage. The programs are of
two general types: utility-oriented and application-oriented.
Utility programs are offered as products by IBM and other
vendors from whom they may be obtained by computer users.
The following utility programs are typical:
Reproduce!1nterpret Program

This program has a number of options. Information from one
set of 96-column cards is reproduced into a duplicate deck. Reproducing is on a column-for-column basis. Punched cards can
also be printed by the MFCU.
Information from one card deck can be punched into different field positions of blank cards. The user punches format cards
363

Card-Oriented
Programs

364

CARD- AND DISK-ORIENTED SYSTEMS

to load with the program to specify the arrangement of fields in
the duplicate deck. The printing location corresponds to the location of punching in the reproduced deck.
96-Column List Program

This program lists cards on the printer without reformatting.
The user may select single-, double-, or triple-spacing; automatic
overflow is provided on the standard II-inch form. After listing
has been completed, a card count is printed. The user can also
select an option to bypass the listing and obtain only a card count
as output.
Sort/Collate Program

This program is designed to perform a wide variety of sorter
and collator functions. A specification sheet is used to describe
the job to be performed. Cards are punched from the sheet and
are used by the program to perform the desired functions. The
following are available:
sort, merge, match, select, and sequence check.
Selection may be accomplished based upon a variety of logical tests including comparison of field contents to a constant or to
other field contents. Several tests may be combined for easy
separation of cards which meet certain tests. Tests for zones,
digits, or characters may be used during a run to select only those
cards to be processed, leaving the remainder of the deck in its
original sequence. Specific cards may be directed to a pocket of
the MFCU during a match or select run.
Control fields for the various functions may be in ascending
or descending sequence and in combinations with numeric fields.
Multiple and split control fields may be described, provided they
do not exceed an accumulated total length of 100 characters.
In collating operations, the user may substitute a value table
which changes the normal sequence values of the sort field contents. It is also possible in the sort program to assign new values
to existing punches in control fields for purposes of sequencing,
sorting, and matching. Thus, an A may be recoded to a B so that
As and Bs would be considered equal. Similarly, the values of A
and B can be reversed to provide a new sort pattern.
The control field for sorting or collating need not be in the
same location in all cards in the deck. For example, the man num-

CARD- AND DISK-ORIENTED SYSTEMS

ber might be in two different locations in two different types of
cards being sorted together.
Note that on-line computer sorting differs considerably from
sorting practice off-line with special purpose equipment. With the
System/3, the sort is begun by placing approximately half of the
card file in the MFCU primary feed and half in the secondary feed.
The machine attempts to merge the two groups of cards according
to a given control field. In-sequence groups or strings of records
are built up in each of the output stackers. After the initial
merge, the machine is programmed to stop and print out for the
operator the number of additional passes that will be required.
Cards from the output stackers are returned to the feed hoppers:
stackers 1 and 2 to the primary feed, and stackers 3 and 4 to the
secondary feed. The process is repeated until the entire file is
merged into one long string and is therefore in sequence.
Figure 18.1 is an example of this type of sorting. Assume
there are 10 records in random order that are to be sorted in ascending sequence by a two-digit number. The deck is divided into
two groups and merged in the MFCU. The first two cards (62 and
97) are placed in ascending order in stacker 1. At this point, there
is a step-down in sequence because 06 and 18 are both lower in
sequence than the last card placed in stacker 1. Therefore, a new
string is started in stacker 3. Merging continues until the next
step-down occurs. When it does, the next string is started in
stacker 2. A fourth string is required and the last two cards are
placed in stacker 4. When additional strings are produced when
sorting larger decks, they also fall into stacker 1, 3, 2, and 4, in
that order.
For the second pass, cards from stackers 1 and 2 are placed
in the primary feed; cards from stackers 3 and 4 are placed in the
secondary feed. Two strings result, and each is placed in stackers
1 and 3 respectively. When the last two strings are merged in the
last pass, all cards are in sequence in stacker 1.
Unlike off-line sorting where records are sequenced one column at a time, the computer reads the entire control field at once.
Therefore, the total time required to sort depends upon the number of cards being sorted, the degree of presequencing, and the
speed of the machine. The shortest possible sorting time results
when the records are already in sequence and only one merge pass
is required. The longest time for sorting occurs when the deck is
in complete reverse order. Normally, when cards are in random
order the number of strings may be assumed to equal one-half the

365

366

CARD- AND DISK-ORIENTED SYSTEMS

number of cards. For a file of 5,000 cards the sorting time can be
estimated as follows:
1. Let S equal one-half the number of cards in the file. Example: S = 2,500~
2. Find the highest power of 2 that is less than S. This number is the probable number of merge passes required. Example:
211 = 2,048 or 11 passes will probably be required.
3. Time in minutes for sorting is T= NC/R
where N = number of passes
C = number of cards in file
R = read speed of MFCU.
11 X 5,000
11 X 5,000
Example: T =
250
or
500
T = 220 minutes with MFCU Model Al or 110 rninutes
with Model A2.
Data Recording Program

A Systemj3 equipped with the optional data entry keyboard
can be used for on-line data recording with 96-column cards as
output. The MFCU is used as a punching device; the keyboard is
used for entering the data from an original document. The program controls the system to perform most of the functions of
punching and interpreting of the 5496 Data Recorder. Key depressions cause corresponding characters to be entered into processor storage. When the stored record is complete, a card is
punched out by the MFCU. There are two advantages:
1. In a low-volume application, the computer can be used to
perform all the operations of data recording without the additional
cost of providing a separate machine unit.
2. In those installations that have one or more data recorders,
the computer can be used to extend card punching capability to
take care of peak loads of short duration.
Data Verifying Program

The Systemj3 may be used to verify the accuracy of recorded
data in 96-column cards. The cards may have been previously

CARD- AND DISK-ORIENTED SYSTEMS

18
Stackers

62
58

First pass

21
34

97
Original
sequence

f---+ 18

2
Feed
hoppers

Stackers

Second pass

Feed
hoppers

Stackers

Third pass

Figure lS.l Forming sequences or "strings" by sorting on MFCU.

punched by the MFCU or by the 5496 Data Recorder. The MFCU
is used in this operation to read the previously punched cards into
storage. Blank cards may also be fed from the second hopper for
correction of errors.
As each character is keyed from the original document, it
enters processor storage where it is compared against the corresponding character in the card record read by the MFCU. If a comparison is not identical, an error light is turned on and the operator
may key the original character again. If there was a keying error,
the comparison should now be correct and verification can continue. If the card character is incorrect in storage, the error light
remains on.

367

368

CARD- AND DISK-ORIENTED SYSTEMS

After the third re-try, the keyed character is assumed to be
correct and is transmitted to storage. The balance of the record is
then keyed. When verification is completed, the correct record in
storage may be punched into a blank card and automatically inserted into the card file to replace the error card. The error card
will be stacker selected.
A verify OK notation is printed on the fourth print line of
each correct card.
80- to 96-Column Conversion Program

This program controls a System/3 equipped with a special
IBM Card Read Punch and associated attachments to read 80column cards, translate the punched data, and punch out 96-column cards in the MFCU. A number of options are provided by
the program, including standard character-for-character translation,
reformatting the data in the 96-column card, special translation
for special characters, and ability to handle invalid and unwanted
characters.
The program does not handle conversion from 96 to 80 columns. If this operation is to be done, the user must prepare his
own program which can be used only with a System/3 equipped
with the special card read punch.

Disk-Oriented
System

The System/3 Model 10 can be equipped with disk storage to
provide the computer with the advantages of direct access to large
amounts of data on line. Two IBM units are available: the 5444
and the 5445 Disk Storage Drives. As many as 7,370,000 eightbit bytes can be made accessible with two 5444 drives. Nearly
41,000,000 eight-bit bytes can be made accessible by attaching
two 5445 drives to the system.
The following section discusses basic operating principles that
apply to many direct-access storage devices. Operation of the IBM
units is described for purposes of illustration.
Disk Storage Operation

Flat metal disks, similar in appearance to phonograph records,
serve as storage media when mounted on disk storage drives. The
disks are coated on both sides with a magnetic recording material.

CARD- AND DISK-ORIENTED SYSTEMS

Depending upon the type and model, one or more disks are rotated at a constant speed on a common spindle. For example, disks
on the 5444 rotate at the rate of 1,500 revolutions per minute.
One or more disks can be removed from the drive units in a
disk cartridge or disk pack. These removable disks are interchangeable among specified drives in much the same way that phonograph records are interchangeable among record players operating
at a given speed and equipped with a proper pickup arm. A 5440
Disk Cartridge, containing one disk, can be used with any 5444
drive and consequently is accessible to any System/3 with a 5444
attached. The 5445 drives are equipped with a 2316 Disk Pack,
each containing 11 disks. By following System/3 programming
conventions on an IBM System/360 or System/370, data on a
2316 pack can also be accessed by any of these systems. In this
way, disks provide virtually unlimited storage capacity that can be
accessed by one or more types of computer systems.
Unlike phonograph records, however, information can be recorded as well as written on the disks by any controlling system.
Thus, a file produced by one computer can be changed- and updated by another.
All data is written in concentric tracks on the disk surface by
movable read/write heads. These heads are attached by support
arms to a carriage that slides back and forth across the disk surfaces. There is one head for each surface, as shown schematically
in Figure 18.2. When the drive is running, the heads float on a
Read/write head

Removable disk
(in disk cartridge)

-,
I
I

____ J

Carriage

Drive spindle

Figure 18.2 Schematic. Disk storage drive operation.

Fixed disk
(5444 models 1
and 2 only)

369

370

CARD- AND DISK-ORIENTED SYSTEMS

thin film of air just clear of the surface of the rotating disks. The
heads do not actually touch the surfaces.
Disk drives are provided with contamination control devices
to avoid interference between the read/write heads and the disk
surfaces. Dust and dirt particles must be excluded from the drive
and from the disk cartridge or pack. For example, the 5444 is
equipped with a brush that sweeps dust particles from each disk
surface during the starting sequence when the drive is first used.
The disks run in a closed air circulation system, so that no air enters or leaves. The cartridge is specially designed to prevent offline contamination when the disk is not on the drive.
The 5444 contains safety devices to control both the start
and stop procedures and the various machine operations. These
devices include interlocks to prevent the operator from gaining access to the drive while it is operating and also to prevent the drive
from starting while the operator has access to it.
The 5444 also has circuits to protect recorded data. These
circuits insure that read and write operations take place only when
it is intended for them to do so, and that recorded information is
not lost accidentally.
Disk Storage Organization

All data on disks is magnetically recorded along concentric
circular paths or tracks. Each track is placed in a specific position
where it can be located by the access mechanism and the attached
read/write heads. Each disk on the 5444 can contain either 100 or
200 tracks, depending upon the model. A schematic of the track
arrangement is shown in Figure 18.3.

Figure 18.3 Disk storage track arrangement.

Because the read/write heads are all attached to the same
carriage, they move across the disk surfaces together. At each access position, two or more tracks are available for reading and
writing, those on the top surfaces and those on the bottom sur-

CARD- AND DISK-ORIENTED SYSTEMS

faces. The tracks thus vertically aligned are considered to form a
cylinder. Information can be accessed from any track in the cylinder without moving the read/write heads. An index pulse indicates
the start of each track and aligns the start of tracks on a cylinder
on anyone disk.
Cylinders are numbered 000 through 102 or 202. Cylinder
000 is nearest the periphery of the disk, cylinder 202 nearest the
spindle. Cylinders 001, 102, and 003 contain alternate tracks for
storage of data should any other tracks be defective. The address
of a track is defined by cylinder and head number.
Each track is divided into 24 sectors of equal length (Figure
18.4). Each sector contains a data field and an identifier field.
One track

Note: The sector address is:
IclclclclclclclclhlslSlslSlslolol

"II(

2 address bytes

AM - Address mark (2 bytes)
S - Sync. byte (1 byte)
F - Flag (1 byte)

Figure 18.4 Disk sector storage arrangement.

[

c --'- cylinder no.
h - head no.
s - sector no.

Add - Sector address (2 bytes)
Check - 3 check bytes
Id field - Identifier field

371

372

CARD- AND DISK-ORIENTED SYSTEMS

Sectors, data fields, and identifier fields are separated by gaps.
The gaps contain no data; their length varies according to their
location within the track. Any given data record can be located
by addressing cylinder, head, and sector number.
Data Field. The data field of a sector contains one synchronizing byte, 256 data bytes, and three check bytes. The synchronizing byte has a fixed bit pattern to synchronize the read/write
circuits with a file control unit. The 256 data byte positions are
reserved for record storage. The record may be arranged in any
format to meet the requirements of the data processing procedure.
If record lengths are less than 250 bytes, the unused portion of the
field should be filled with any valid bit pattern. For operations of
reading and writing, records are therefore considered to be of
fixed length.
The check bytes verify the information stored in the fields to
which they are appended. The bytes are computed by the file
control unit when data is transferred to disk storage. Subsequently, when data is read from the disk, the file control unit recomputes the check information and compares it with the check bytes
read from the corresponding data field. Any difference in the
comparison indicates an error.
Identifier Field. The identifier field contains the physical
address of the sector. The start of each sector is denoted by an
address mark with two bytes of specially recorded information
that is uniquely identifiable. A synchronizer byte is recorded
after the address mark. The identifier field contains a flag byte,
two address bytes, and three check bytes. The address defines
the cylinder, head, and sector numbers.
Flag By te. Occasionally, because of damage to the disk or a
defect in the recording material, a particular track will not be
usable. The condition of the flag byte in the identifier field indicates that either (I) the track is a normal data track, (2) the track
cannot be used because of a defect, or (3) the track is an alternate
for a defective track. Normally, all eight bytes of the flag byte are
set to zero when the sector address is first written. Setting certain
positions to ones will flag the machine that one of the above three
conditions exists. Significance of the flag byte coding for the
5444 is shown in Figure 18.5.
When a track is found to be defective, an alternate track is
assigned. On the defective track, the address of the alternate track
is written into the identifier field of every sector. On the alternate
track, the address of the defective track is written into the identifier field for cross-reference.

CARD- AND DISK-ORIENTED SYSTEMS
Flag byte
bit

Setting

Function

Not used

Operative track

Defective track

Primary track

Alternate track

o through

Figure 18.5 Flag byte significance.

When an access operation is performed to a defective track,
the flag byte indicates that the track is defective. The address of
an alternate track is given in the identifier field. The read/write
heads then move to the new track address where the address of
the original or defective track is given in the identifier field. For
operating purposes, the read/write heads are now positioned at
the correct location as defined in the address instruction.
Alternate tracks are assigned one at a time. Facilities in the
system automatically insure that the proper exit procedure and
return procedure to the original location are made if a data operation overlaps.
Disk File Organization

Data sets in disk storage may be organized in a number of
ways. Three common methods are:
Sequential
Direct
Indexed sequential

Sequential. The records in a sequential data set are consecutive according to a given sequence. Sequential organization is the
only one possible for magnetic tape, printer, and punched-card
records. Normally, no record in the file can be read or written
until all preceding records in the sequence have been read or written, but special facilities are available to p.ermit arbitrary positioning. On tape, records cannot be lengthened, shortened, or deleted
without rewriting the subsequent portion of the file.
This same sequential file organization can be used for disk
storage. Records may be written sequentially beginning at a given
address and continuing through successively higher addresses until the entire file is written. A record control field or key may

373

374

CARD- AND DISK-ORIENTED SYSTEMS

be used to determine the relative location of a record within the
file.
Sequential disk storage is useful in those procedures where
direct access to particular records is not required; for example, the
storage of programs, subprograms, routines, and entire programming systems. Also, sequential storage organization may be used
to form a queue of records or data sets waiting to be printed,
punched, or transmitted over communication lines as output.
Direct. Records in a file are assigned directly to specific storage locations. In this type of file organization, the addresses in
storage must bear a direct relationship to control fields or keys in
each record. For example, the assigned part numbers of inventory
records might correspond with the storage addressing scheme.
Thus, the record for part number 1 is stored in location one, part
number 2 in location 2, and so on.
Regardless of the order in which records go into the file, each
record always occupies a specific disk location. Therefore, missing
or unavailable records produce blank storage locations. A schematic of direct file organization is shown in Figure IB.6.
Indexed Sequential. Neither the sequential nor the direct file
organization allows convenient access to any particular record
without first reading the preceding records. The indexed sequential file organization overcomes this difficulty by associating the
key of each stored record with its corresponding storage address.
The keys, with their locations, are then stored separately as a table
or index. The index contains the key of every record in the file
and its address in storage. The data file mayor may not be in sequence when it is stored. If it is not, the index can be sorted by
key later.
For example, assume records 075, 024, 253 ... ,046 are placed
in storage in locations Dl, D2, D3, ... , D99 as shown in Figure
1B.7. As they are stored, the computer is programmed to compile
a table including each record key with its corresponding address.
The table appears as follows:
Key

Disk Address

075
024
253

Dl
D2
D3

046

D99

CARD- AND DISK-ORIENTED SYSTEMS

375

Card record
file

1
. . f. - - - - - - - - - D i , k ,"o"g. ac.,

-I

Figure 18.6 Direct file organization.

The table is now sorted by key so that the index appears:
Key

Dis k Address

024
046
075

D2
D99
Dl

253

To retrieve any record from the file, the key must be matched
against the index to obtain the corresponding address. This organization has the advantage that storage can be filled to capacity as
records are received, and that new records can be inserted without
rewriting the entire data set. Also, if records of known length are
stored, new records can update or replace old ones as transactions
occur, provided the index is modified accordingly.

Complete programs are also available from IBM and other vendors
to process specific applications on the Systemj3. Programs may be
tailored to the needs of a specific industry, a general accounting

Applications
Processing Programs

376

CARD- AND DISK-ORIENTED SYSTEMS

l~/I
/

/
253

Index storage

/
/

099

Record storage

I-

Sort by
record
key

1~~r-

_____

I_n_de_x_st_o_ra_ge______~~~1

1~~r-

__________Re_c_or_d_st_o_ra~ge

__________~'1

Figure 18.7 Indexed sequential file organization.

function, or to a management system. Users of such programs are
generally required to· construct their files and other records in some
standard manner in accordance with the formats handled by the
program.
Typical products offered include such programs as hospital
accounts receivable, hospital patient billing, apparel business control, utility billing, and so on. The IBM Apparel Business Control (ABC) system is an excellent example of ready-made programs
available to System/3 users. The following features are offered:
Order Edit. This function is performed by two separate programs, order edit and pricing. The order edit program checks new
orders for valid product codes, crossfoots the order detail cards,
and balances the totals from the order detail cards against the totals

CARD- AND DISK-ORIENTED SYSTEMS

377

in other miscellaneous data cards. The pricing program validates
the style and color numbers against those actually carried, and
prices the order detail cards.
Order Writing. This program prepares printed orders for use
as internal documents and as order acknowledgments to the customer. An order register program controls the machine to print
out a register of all orders received.
Bookings Reporting. This program prints a combined bookings and over and under report by style, color, dimension, :lnd size.
The report includes a summary of sales received since the last report, a detailed listing of key orders received since the last report, a
summary of remaining orders received, a summary of all outstanding or completed cuts, and an over-under summary.
Fabric Requirements Reporting. Two programs produce output that shows seasonal fabric requirements based on either actual
or projected sales.
Finished Goods Requirements Reporting. This program produces a report showing finished goods requirements by time period,
based on the start shipping and complete shipping dates specified
on the orders.
Stock Allocation. This program matches outstanding orders
against completed cuts. Cards representing allocated orders details are selected into separate stackers on the MFCU for review
and shipment.
Invoicing. This feature is made up of three programs to
crossfoot order detail cards, extend and total invoices, print an
invoice register, and punch invoice summary cards.
All of these programs can be tailored to the user's specific
requirements by filling out a questionnaire. Responses to questions are keypunched into 96-column cards that are used as input
to a special modifier program. The modifier program then punches
out modified program instructions for the system application programs. That is, the System/3 is used to generate the system programs properly modified to suit the individual user.
A specific configuration of equipment is also required to
operate the ABC system.

1. What are the advantages of the utility programs to the computer user"!
2. Trace the card paths during a System/3 sort/merge operation.
3. Compute the approximate sorting times for files of 4,200, 6,850, 2,225
and 12,900 cards. (Use Model A2.)

Questions

and
Exercises

378

CARD- AND DISK-ORIENTED SYSTEMS

4. Under what circumstances is it justifiable to use the computer system for
keypunching and verifying?
5. Under what circumstances can the System/3 convert from 80-column to
96-column cards?
6. What is a disk cartridge and how is it used?
7. Describe the physical arrangement of the disk for data storage.
8. Explain the three main types of disk file organization: sequential, direct,
and indexed sequential.
9. How do application processing programs differ from utility programs?

PROGRAM
PREPARATION

The preceding sections described various components of the
System/3 with particular regard for function and system organization. It was shown that the computer operates only under control
of a stored program contained in memory. The program completely details a procedure, step by step, in terms of the 28 basic
operations the machine can perform. Each step is called an instruction. By executing instructions in the proper sequence, the
computer can produce results from available data input. Two
arithmetic instructions were described as examples of how the
processor manipulates data.
The program is developed for each different procedure by a
programmer. Programs are read into memory like any other input
data and are retained there for the duration of a particular task.
When a task is completed, a new program can be stored in preparation for the next procedure to be followed. Transition from job
to job can be made automatically if supervisory programs are
available.
This chapter will deal more completely with the development
of the stored program and will lead into the use of a programming
system called RPG II which is especially designed for the System/3.
379

380

General
Considerations

PROGRAM PREPARATION

It is always surprising to realize how difficult it can be to describe

even the simplest task to someone who has little or no preknowledge of what is required. It is especially difficult if the task is to
be detailed in writing. As an example, try to write down step by
step instructions for tying a shoelace or a necktie.
When giving instructions verbally, a number of assumptions
are usually made. For example, it is assumed that even a child
knows what a shoelace is, why it must be tied, and where it is used.
But, depending upon the person receiving the instructions, none of
these assumptions may be valid. Imagine telling-not showing-a
barefoot aborigine how to use shoelaces!
Such an illustration may be a ridiculous exaggeration, but the
same principles apply to the requirements for describing to a computer exactly how a procedure is to be carried out. Nothing can
be assumed. In the case of the computer, however, the capabilities
of the machine and the extent of its resources are completely
known-such facts are not always available when instructing people.
First, the programmer or systems engineer must take into account the overall aspects of the application which is to be mechanized. This "big picture" of the task can be viewed in three
main parts:
Problem definition
Problem analysis
Procedure design
Problem Definition

The application must be understood completely, particularly
in terms of what results are required and when they will be needed.
Then it must be determined if proper input is available and whether
or not this information can be used to produce the required result
on schedule.
Problem Analysis

In most commercial applications, this is primarily the development of an economic justification. There may be many tradeoffs
among requirements, schedules, equipment, and implementation
costs. For those jobs that must be done, like payroll, accounts receivable, and billing, the objective is almost entirely to produce
results as economically as possible. For applications like sales
analysis, inventory control, and cost analysis, the return on the data

PROGRAM PREPARATION

381

processing investment is the prime consideration. For example:
What savings are possible if management can know the inventory
position by the fifth of the month instead of the fifteenth? Do
these savings offset the additional costs of accelerating the schedule? What savings can be made if the inventory position is known
weekly instead of monthly?

Procedure Design

The preparation of a diagram or flowchart is a basic approach.
The flowchart may show the overall aspects first to assist in the
problem analysis. Later, the proposed procedure is diagrammed
in complete detail.
Figure 19.1 is a system flowchart of an inventory control
procedure. The chart uses specific symbols for both processing
and input/output functions. The procedure involves a multiple
warehouse system. Items are stocked in a central warehouse for
distribution to remote warehouses. All customer orders are received in the remote locations and transmitted by teletype (communication link symbol) to the central data processing installation.
The system provides four major groups of operations:
a. Updating stock status, based on actual transactions. Shown
on the flowchart as Run S 1.
b. Response to inquiries from auxiliary warehouses and the
central warehous~. Shown as Run 01.
c. Reorder analysis, including purchase order preparation.
Shown as Runs PI, P2, and P3.
d. Weekly analysis reports to show slow-moving items, major
changes in usage rates, behind-schedule deliveries, economic lot
sizes, etc.
All of the above planning and analysis is usually carried out
independently of the computer and the programming system.

To develop a final program, the programmer must be concerned .Program Preparation
with at least some of the following in direct relation to the machine and the programming system with which he is to work.

Prepare
Inventory

Receipts

Prepare
Stock
Withdrawals

Withdrawals

Misc.
Transactions
RUN 01
Determine &
Update Stock
Status and
Prepare Ship.
Instructions

Master
Stock File

Update
Stock Status

Prepare Misc.
Transactions

At End
of Day's
Transactions
I

RUN PI
Select Items
Below Protection
Level Determine
Order Qty.
and Date
RUNP2
Group Orders By
Commodity Class

RUNP3
Select Vendor &
Purchase
Orders

~-r---------IPrepare

Figure 19.1 Inventory control procedure.

382

At End
of Week's
Transactions
I
I

..J

RUN S2

Analyze Usage
~-~ and Status

PROGRAM PREPARATION

1. Conversion of original data to the input medium. With
the System/3, the preparation of 96-column cards is involved. A
small amount of data can also be entered with the printer-keyboard.
2. Allocation of the computer storage areas to data, instructions, and related information. Storage capacity directly affects
the method of job execution; that is, the number of records stored
at anyone time, the length of the program, and consequently the
extent of the processing that can be completed in anyone pass of
the data through the machine.
3. The availability of reference data: tables and constant
factors. Such information may be required throughout a job run
but cannot be conveniently read in with each input record. Such
information must be given space in memory.
4. Requirements for accuracy. These vary with each application. For example, a payroll must be 100% accurate. On the
other hand, some tolerance for error may be permissible in a sales
analysis where trends rather than actual figures are to be shown.
5. Ability to restart the system when an unscheduled interruption occurs, such as a power failure. Usually, work is performed in manageable batches with controls and totals established
by each batch. If an unexpected work stoppage occurs only the
last batch must be redone, not the entire job.
6. Format of output data in the form of punched cards and
printed reports.
7. Availability of program routines that have been used and
tested in other procedures. Some may be used to advantage in the
current procedure being developed. Many programming systems
make provision for a library of subroutines or programs that are
available. Means are provided to properly insert a library item in a
new program automatically. The programmer calls out an item by
writing down proper identification in his program.
8. Resource management. Consideration should be given to
recording the utilization of the system, that is, logging the running
time for each job, recording exceptions to procedures that cannot
be processed, planning for system maintenance, and allowing for
unsched uled downtime and diagnosis.
9. Housekeeping routines. Although these routines are usually an inherent part of the programming system, there must be pro-

383

384

PROGRAM PREPARATION

vision when starting a job to be sure all housekeeping has been
done first. That is, indicators may be preset, storage areas cleared,
and constant data moved or arranged according to instructions.
In case of restart, additional housekeeping may be required to restore the system to a ready state to reprocess the last batch of
work not completed.

Problem
or
application

Programmer

Machine
language
program

Program
deck

Problem
data

Computer
execution
run

Output
cards
Report

Figure 19.2 Relationship between the computer and programmer. Actual machine
language coding.

PROGRAM PREPARATION

There may also be instructions to the operator printed out on
the printer-keyboard telling which form to insert in the printer or
what cards to put in the MFCU. Additional messages may be
printed as entries in the log of system utilization.
Figure 19.2 shows the relationship between computer and
programmer if it were practical to write a program in actual machine language. The problem is first analyzed, defined, and diagrammed as described earlier. Then the programmer writes the
program and supplies all tables, formulas, codes, or other reference
materials needed for the specific application.
When the program is prepared, the problem then becomes
input data and the computer produces the planned output. A
number of difficulties make this approach entirely impractical:
1. All instructions must be coded in machine language. Even
though the System/3 is a relatively simple configuration of units,
the production of instructions of varied format in hexadecimal
code would be very difficult indeed.
2. All instructions must be arranged in the exact sequence in
which they are to be executed by the computer. If one or more
are omitted by error, all succeeding instructions must be relocated
in storage to make room for insertions. Otherwise, some branching technique must be used to include new instructions. This
clerical accounting for all storage space must be carried on entirely
by the programmer.
3. The full burden of logic and program organization is
placed directly on the programmer.
4. Previous experience (i.e., tested programs that might be
used for part of the procedure) is difficult to work into the new
program. Such programs must be linked to the new program by
additional handwritten instructions.
5. The programmer must understand the computer in detail.
He must know each indicator and register and he must program
their functions entirely.
6. To test the program for error, all checking must be done
in machine code. If corrections are required, the clerical work of
modifying and rechecking becomes exceedingly complicated and
time-consuming.

385

386
Program Assembly

PROGRAM PREPARATION

Early in the history of computer development it became obvious
that methods must be worked out to free the programmer from
the burden of machine language coding. As engineering sophistication of machines or hardware advanced, there was a corresponding need for the techniques of programming or software to keep
pace. Otherwise, full utilization of future machines could not
be realized.
Because a large proportion of the task of program writing
proved to be clerical, it seemed logical to attempt the assignment
of such work to the computer itself. The machine might thus do a
large share of the work in preparing its own program. Consequently, in the evolution of computer application, a number of software
systems have been developed. The simplest concept centers around
the principle of symbolic programming, or assembly.
In this approach all data, records, fields, and related information to be used in the procedure are labeled with symbolic identifiers when the program is originally written. No specific storage
locations are assigned. That is, each item of data to be addressed
by an instruction operand is given a name to easily identify that
item for the programmer.
The amount of storage required is specified by indicating the
length of each item on the written program sheet; This will later
tell the machine exactly how much memory space will be required,
but not where the space will be located for the actual program.
For example, an input pay record might be described as follows:
Label
PAY PER
DEPT

Length

LaC

2
3
2

SHIFT
MANNO
RATE
REGHRS
OTHRS

5
3
3
3

BLANK

The sum of field lengths within a record must equal the total
length of the record. Forms for writing the program in longhand
are provided to standardize the format and to enforce the estab-

PROGRAM PREPARATION

lished conventions of the system. For example, labels may be restricted to 10 spaces; length of field to three spaces. Means are also
provided to address an entire record as well as individual fields
and items.
Next, a standard abbreviation is given each operation the
machine can perform. The abbreviations take the form of mnemonics: ADD, SUB, DIV, ZAZ (Zero and Add Zoned), SZ (Subtract Zoned Decimal), etc. The use of mnemonics is intended to
make the programming language as much like everyday English as
possible. As a result, the language is designed to be easy to learn
and relatively simple to use. The programmer writes these standard mnemonics in instructions just as if they were actual machine
coded binary notations.
Instruction operands address the previously named record
fields and reference items. Thus, a small program routine to calculate gross pay might appear as follows:
Label

Operation

Operand

Comments

START

SEL
RD
CLA
ADD

READER
PAY REC
REGHRS
OTHRS

MPY
STO

RATE
GROSS

Select input device
Read one record into storage
Put regular hours in accumulator
Regular hours + overtime hours
= total hours
Total hours X rate = gross
Put gross pay in record

SEL

PRINTER
PAY REC
START

Select output device
Write one record
Branch to repeat routine

WR
BR

This example is simplified for purposes of illustration and is
not an actual sample of any particular language. Several basic concepts are shown.
First, the programming sheet is laid out to include at least
four items of information: label, instruction operation part, instruction operand, and comments. The comments have no effect
on the procedure but are used as required for clarification purposes
only. The comments also make it easier for others to follow the
procedure in case changes or alterations must be made later.
The first step of the program is labeled START as a reference
point that may be addressed by other instructions. The last in-

387

388

PROGRAM PREPARATION

struction operand addresses the first instruction label indicating a
branch back to the beginning of the routine. Thus, the portion of
the program shown forms a complete loop that is repeated for
each record of the input file. Only an instruction to be addressed
is given a label.
After the program is completely written and all items of data
properly identified, a single entry is made indicating the actual
storage location of the first instruction. The completed program
with all related data is now punched, line by line, into cards. The
resulting deck is held for input to a computer assembly run as a
source program.

The computer is loaded with the assembly program which has
been previously prepared as part of the system. In this run, the
source deck is processed as data. All symbolic information is converted to actual machine coding. Essentially, two passes of the
source data are required, although the system is often designed to
produce a complete object program in one run.
1. From the beginning memory location in the source program, the machine can determine the actual locations of all succeeding instructions by progressive addition. The length of all
record fields and constant factors is also given in the source program. Again, by addition of the lengths, the computer can assign
an actual memory location to each item.
All symbolic labels with corresponding actual locations are
stored asa table in memory.
2. After all actual locations have been determined, instruction operands can be converted by an operation of table look-up.
The assembly program also stores as a reference table all possible
mnemonic operations with their machine code equivalent. During
the processing, these symbols in the source program instructions
are also translated into machine codes.
The converted or assembled object program can be punched
out in cards as an object deck. A printed list of the source program including the machine coded translation can be obtained as a
by-product of the assembly.

Programming

Systems

Figure 19.3 shows the relationship between computer and programmer when a programming system is used. The system includes
at least two parts. First, there is the system language with a stan-

PROGRAM PREPARATION

Problem
or
application

Programmer

---1-1"'......

Source
program

Source
deck

Computer
processor
run

storage

Object
deck

Processor
listing

Problem data

Read into

Computer
execute
run

Output
cards

Figure 19.3 Relationship between computer and programmer. Programming system.

389

390

PROGRAM PREPARATION

dardized convention of grammar and syntax. This is the language
in which the programmer writes the program. Second, there is an
assembly program the computer uses to translate from the system
language to another machine language program.
The input to the· assembly, as written by the programmer, is
called the source program. It states the requirements of a given
task and the method of solution. As previously stated, before the
programmer writes his source program, he must have completely
analyzed and defined the problem.
The output from the assembly is the object program. This is
the translation of the source program to the machine language of
the computer system on which the object program will be used.
The assembly program is available in cards and thus can be loaded
into the computer for the assembly run.
The source program is keypunched directly into cards directly
from the programmer's original written sheets. These cards become input data to the assembly. When all the source information
has been translated, the resulting object program is available as any
normal output from the computer. Usually, the object program is
listed to record the assembly process. Exceptions or errors preventing assembly are printed as the source program is being translated. Unless severe trouble is encountered, the program is punched
into output cards as an object deck.
During the assembly process a number of checks can be made
on the programmer's work. For example, the programmer may
have referred to nonexistent parts of the program or may have forgotten to make reference information available. Any illegal mnemonics for operations can be listed, grammatical violations can be
noted, and out-of-sequence instructions can be shown. Sometimes
a controlled tolerance for error is built into the assembly program
in such a way that source information is used even though misspelled or otherwise in error.
Once the object program is properly assembled and punched,
it can be loaded into the computer storage for an execution run.
Normally, this first run is to test the object program with sample
input data. The test data should be arranged to cause the machine
to execute all possible operations and to follow each exception to
the procedure completely. If the program is fairly complex, it will
be rather unusual for this test run to be executed perfectly.
The process of removing errors, or debugging the program, is
carried out next. Many programming systems provide special aids

PROGRAM PREPARATION

391

to locating and fixing errors with printouts, tracing routines, and
other diagnostic methods. One method of error detection halts
machine operation at the point of error and prints out the entire
contents of memory. The programmer may then examine this
"snapshot" of storage for some clue to the malfunction. Meanwhile, the machine can be performing other work.
When all errors are finally located, correction cards are
punched and inserted in sequence in the source deck. The deck is
reassembled and the test is rerun on the new object program. This
program is debugged if necessary, reassembled and tested again
until a final, error-free object program is obtained. The proven and
tested object program can be used again and again with varying
sets of task data.

The preceding discussion has described one method of overcoming
the language barrier between computer and programmer. With this
method, programs can be written in terms more easily understood
(and learned) by the programmer. At the same time, the assembly
does much of the clerical work of program preparation. The assembly process also performs some auditing functions, thereby reducing the occurrence of many common manual errors and making
easier the job of testing and correcting the operation of the object program.
When an assembly system is used, the programmer writes all
the detailed computer steps. And even though there are many advantages to the method, he must furnish all instructions on a onefor-one basis. For one instruction written in the symbolic language,
the assembly will produce only one instruction in machine language coding.
The next step in programming system development is to increase the effectiveness of the assembly by enlarging its functions.
For example, many phases of procedures are duplicated or repetitious. All programs require routines to read and write records, to
check for errors, and to perform many standard calculations of
arithmetic. Many procedures also use standard methods of converting data from one form to another.
Common routines can be included in a table, array, file, or
library that is available to the assembler. The programmer is then
given the option to instruct the assembly program to select appropriate routines from the library and insert them in the object pro-

Compilers

392

PROGRAM PREPARATION

gram as specified. The operation part of library instructions is
standard, but the addresses must be modified to reference the location of data as defined by the source program.
To make these enlarged functions of the assembler available,
additional mnemonic operations can be added to the language.
For example, assume that a tested routine for extracting square
root is stored in the library. This routine may be given the name
SQRT. If the programmer needs the square root routine as a part
of his procedure, the mnemonic SQRT is written on the source
program sheet as the operation part of a symbolic instruction.
The operand designates the storage location of the number and
the location where the root is to be stored.
When the assembly program encounters SQRT, reference is
made to the library rather than to the table of machine operation
codes. The SQRT subroutine is then assembled into the object
program with addresses obtained by the regular assembly process.
In this case, a number of machine instructions are produced in the
object program for one instruction in the source program.
To make more and more functions available, the size of the
library can be increased. As the resulting programming language becomes more flexible, it also becomes more independent of actual
machine operations. Consequently, there is the effect that the
programmer now thinks in terms of the programming system
rather than in terms of actual computer functions related to a
particular machine.
In fact, as the language becomes more and more sophisticated,
the programmer is required to know less and less about the computer system. Or, as the language is developed to a higher and
higher level, it tends to become problem oriented rather than machine oriented. Thus, the language is a more effective aid in expressing the problem while the computer does a greater portion of
the conversion from procedure to machine operations.
Functions can be added to the system to link entire program
routines together and to automatically include housekeeping,
checkpoints, restart, and control routines. The program assembler
now becomes a compiler with the ability to search the library and
compile an object program from given specifications and available routines.
Finally, the language can become standardized among a large
number of users so that there is general agreement as to the rules
of grammar, syntax, format, limitations, and so on. At the highest
level, the language is completely independent of any particular

PROGRAM PREPARATION

393

computer system. Compilers can be developed to translate an
original source program for use by any computer, even those machines which are produced by a variety of manufacturers.
Therefore, programs written in the standardized language can
be compiled on a number of different computer systems. Of
course, the necessary configuration of equipment must be available with sufficient storage capacity and proper input/output
devices.
Several such high-level languages are widely used throughout
the computer industry.

This COmmon Business Oriented Language is based on a system
first developed in 1959 by a committee of government users and
computer manufacturers. The language itself, as well as the techniques for using it, are concerned mostly with commercial problems and results.
The COBOL language bears little resemblance to machine
language. And the problem programmer has little to do with the
method by which his COBOL program is translated into machine
language.
The following example illustrates the basic principles of the
problem oriented type of programming system. Assume that the
value of an item called INCOME is to be increased by the value of
an item called DIVIDENDS. COBOL can specify this addition by
the following sentence in the source program:
ADD DIVIDENDS TO INCOME.
Before the COBOL processor can interpret this sentence, however, it must be given certain additional information. For example,
the programmer must write DIVIDENDS and INCOME in a special part of the program called the data division. The facts about
the data represented by those names are stated here, such as maximum size, how the data is expressed, etc.
When the compiler encounters the sentence, it also has access
to certain information that will aid it in translating the sentence.
In addition, it can obtain certain information built into the processor itself. The exact procedure varies from machine to machine.
In any case, the problem programmer is not directly concerned
with the details.

COBOL

394

PROGRAM PREPARATION

First, the compiler examines the word ADD. It then consults
a special list of words that have clearly defined meanings in COBOL.
This list is part of the compiler. If ADD is one of those words,
the compiler interprets it to mean that it must insert into the object program the machine instruction or instructions necessary to
perform addition.
The compiler then examines the word DIVIDENDS. Since
it can obtain certain information about DIVIDENDS, it will know
where and how this information is to be stored in the computer,
and it will insert into the object program the instructions needed
to locate and obtain the data.
When the processor encounters the word TO it again consults
the special word list. In this case, ;it finds TO directs it to the value
of INCOME which is to be increased as a result of the addition.
The compiler must now examine the word INCOME. Again
it has access to certain information about this word; as a result, it
can place in the object program the necessary instructions for locating and using INCOME data.
Notice that the programmer placed a period (.) after the word
INCOME, just as he would in terminating an English-language sentence. The effect of the period in the COBOL compiler is quite
similar. It tells the compiler that it has reached the last word to
which the verb ADD applies.
The steps just described are performed by the compiler in
creating an object program. They may not be performed in exactly
this way or in the same sequence, because machines vary and because each compiler is adapted to a particular machine. Regardless
of the machine, the same COBOL sentence produces machine instructions that will cause the 0 bj ect program to add together the
values DIVIDEND and INCOME.

FORTRAN

FORTRAN is an acronym for FORmula TRANslation system.
The language is made up of individual commands or statements
of a program, operators (such as + or -), and expressions (such
as C = A + B). The compiler is a program for the computer which
tells it how to translate the program written in FORTRAN into a
program in machine language.
The FORTRAN language is intended to be capable of expressing any problem of numerical computation. In particular, it
deals easily with problems containing large sets of formulas and

PROGRAM PREPARATION

395

many variables, and it permits any variable to have up to three
independent subscripts.
Each FORTRAN source program is composed of a number
of statements. Each statement deals with one aspect of the problem or procedure. For example, a statement may cause data to be
fed into the computer, calculations to be performed, decisions to
be made, results to be printed, and so on. Also, some statements
written by the programmer do not cause any specific computer
action, but rather provide information to the processor. For example, the following two statements are listed with their effects:
READ 1, A This causes the computer to read a card and
handle the data in such a way that if the card has a value or quantity 97.0 punched in it, then A will have the value of 97.0.
C = 3.* A The asterisk indicates multiplication. Thus, this
statement means multiply A by 3.0 and set C equal to the result.
Using the data of the previous example, C would be given the
value 291.0, the product of 3.0 X 97.0.
Notice that when the source program is executed, the computer is not instructed merely to find the value of C. It is given
the data (in this case 97.0), instructed how to make the computation, and told what to do with the result.
An important aspect of FORTRAN, as of many other programming systems, is that the language can be largely independent
of the computer on which the object program is to be executed.
But the FORTRAN compiler is dependent upon the object machine because it must produce an object program in that machine's
language. For this reason, each object machine must have its own
processor. This is another way of saying that an original FORTRAN source program can be translated for use on several types
of computers. But the source program must go through the translating process for each type of machine on which it is to be used.

1. What are the three main aspects of a procedure or application that must
be considered for mechanization? Explain each.
2. In what detail would a system engineer need to understand machine
functions to mechanize the procedure in Figure 19.1? Explain.
3. How does the data capacity of storage affect the method of doing a job?

Questions

and
Exercises

396

PROGRAM PREPARATION

4. What are tables and reference data and where would these be obtained
in a procedure?
S. What is resource management?
6. What is housekeeping?
7. Give some important reasons why programming in machine language is
impractical.
8. What is the purpose of a program sheet?
9. What are some important advantages of a programming system?
10. What is the difference between a problem-oriented and a machine-oriented programming system?
11. What are some of the main characteristics of COBOL and FORTRAN?

REPORT PROGRAM GENERATOR,
VERSION II

The Report Program Generator, Version II, compiles or generates a machine language object program for the Systemj3 from
written specifications of a problem or application. Input data files,
the planned output, and all operations of data handling are expressed in predetermined format by the programmer. Originally
developed for the IBM Systemj360 as RPG I, Version II has been
expanded with additional features for the Systemj3 that can be
used in a wide range of commercial data processing jobs. The
RPG system has both a language and a processor.
The language is primarily a set convention that has been
established for filling out the various programming forms. Different specification sheets are used to code each aspect of the
source program. For example, one form is used only for input
file specifications. Another is used only to describe calculation
and data handling operations. Headings on each sheet describe the
particular use of the forms.
After specifications have been written, cards are punched
from the data. Each line on a form is a card in the source deck.
This deck and the RPG processor deck are read into the computer
by the multifunction card unit during an assembly run. One
MFCU hopper contains the source program, tables (if any), and
blank cards for the object deck. The other hopper contains the
processor, approximately 3,000 cards containing about 40,000 in397

398

REPORT PROGRAM GENERATOR, VERSION II

structions. The assembly is done in six phases requiring about
2,000 bytes of storage per phase. The object program can be produced with two options: (1) as a printed listing only or (2) both
in cards and as a listing with all machine instructions to run a
planned task. After the assembly run, the object program can be
loaded into the computer for a test or execution run with actual
problem data.
During assembly, the processor checks many clerical aspects
of the source program and produces printed messages to the programmer to identify any errors encountered during production of
the object program.
Debug statements may be placed in the source program deck
for listing and to display fields and indicators, information that
can be helpful for locating programming errors. The contents of
any field and any program indicator can be printed. The object
program can be generated with or without the debug option.

File Description

In RPG language, the term file refers to any logical collection of
related records. The records may be on either cards or paper.
Therefore, there is the rather unusual connotation that a printed
report is also a "file." Three kinds of files are considered in
the system:
1. Input files froin which information is read.
2. Output files on which information is printed, punched,
or both printed and punched.
3. Combined files; that is, card decks in which some or all of
the cards are to be used for both input and output. Example: two
fields in the input deck are to be added and the sum punched in a
third field in the same cards.
The RPG programming system is designed mainly to process
files of cards.

Specification Sheets

Certain information is common to all of the specification sheets
used in the RPG system. Refer to the Control Card and File
Specification form shown in Figure 20.1.

International Business MachInes Corporation

IBM

Form X21·9092-Q
Printed in U.S.A.

RPG

CONTROL CARD AND FILE DESCRIPTION SPECIFICATIONS
1

IrU~chi~~--'-i;~~i~rJ-_[I~ l_=t-~

Date _ _ _ __

L::

r
:::_

Page

Punch ~~~1_1--=r~_ .=cJ

75 76 77 78 79 80

Program
Identification

IIIIIII

Control Card Specifications
Sterling

Core
Line

(l)

Sizet~
Compde

~ .~

00
~

::i

Core
Size to
Execute

Number

Of Print
Positions

Etn

Refer to the specific System Reference Library manual for actual entries.

8
;;'

12 13 14115116 17 18 1920 21 22 2324 25 2612} 28 29 30 31 32333435 3637 38 39 4041 424344454647 4849 50 51 52 53545556 5758596061 626364 6566 676869 7071 72 73 74

I I

I I r I L I ILI_LLLllltW_W Lill LLI III I l LI I I r 11 ITT In
File Description Specifications
Mode of Processing

File Type

File Addition

Length of Key Field
File Designation

End of File

Record

Add((~ss

Type

Number of Extents

Type of File

Sequence

Filename

Line

Number of Tracks
for Cylinder Overflow

of Record Address Field

Symbolic
Device

Device

, File Format
0-

er:
u

~
51617

~1wI~1 ~

Block
Length

Tape

for DAM

Record
Length

Rewind

Extent Exit

Name of
Label Exit

~I~

::>

.--t-.'-rlltffifh1tlilti-rtt-i-rullnm

Location

101112 1314115116117118119120212223124252627128129 30131132133 34135 36 37 38139140414243444546147484950515215315455565758 59160616263646516616716869170171727374

I I

iJ
-t
LJ--f
--r
-+-+-+-+-+-+. ++-+--+- ~L-t±
---r---t-1+r Ll-l
\!
i!
II

'-0
'-0

f-- -

- _.- -

1-+-+ -- -.-

I I I I I I I I:

111111

.[--1+ - - -

1.1.

~_il

-- --r_ -

J---1
- -- t

ii-ttt--Lt±H-Tft-r-f+f-1-tttH-ttlllllllllllll'
Figure 20.1 Control card and file description.

I I I I I

-I-I--I--+--+-+--H--

t-+++++-+--

I I I I I I I I I I I I I I I I I I I I I I I I I

400

REPORT PROGRAM GENERATOR, VERSION II

Each sheet is divided into 80 vertical columns for convenient
punching into cards, either 80- or 96-column. Each horizontal line
is divided into various fields for coded entries which will describe
the application. The programmer writes one character, digit, or
symbol in a column.
Columns 1 and 2 are located at the top right corner of the
form and are used to number the pages. Columns 75 through 80
are just to the right of the page identification and serve to contain
the program identification. Columns 3 through 5 are used to number the lines of coding. The first two digits are preprinted on the
sheets; the third may be used as an insert digit when items are
omitted from the normal sequence. When punched in cards, the
line numbers keep the records in order within each page.
Column 6 indicates the type of form so that this information
will also be punched in the source deck. The following codes
are used:
H
F
I
C

o
E
L

Control or "header" card (because it is the first
card in the source deck)
File description (cards or printed output)
Input specifications
Calculation specifications
Output specifications
File extension specifications
Line counter specifications .

All type codes are preprinted on the corresponding form.
Only Hand F are used on the form shown in Figure 20.1. All of
the above items are aids to the clerical organization of the coding
and have no effect upon the object program other than to insure
proper sequencing and identification of the source data.
Figure 20.2 shows the proper card arrangement in the source
deck after punching. Cards outlined in dotted lines are optional.
For example, a program may be assembled for a card-to-print operation where no calculation is required.
Column 7 may contain an asterisk (*) to identify written
comments. The asterisk indicates that all of the following information on that line is not to be assembled as part of the object program. The programmer may use the comment line to document
notes explaining the procedure. The notes may be particularly
helpful later if the source program is to be revised or changed

REPORT PROGRAM GENERATOR, VERSION II

401

and must therefore be understood by someone other than the
original programmer.
/*

---- - "

I Tables Loaded at
I Compilation Time

_...1.______
/

* *

'I
I

,- _1- _ _ _ _ "
I

r--'

I
I

r-" I

,....L ______ ,
( Alternate
I Collating Sequence
I Specifications

....

I
I

I- ~)

(-Fiie - - ---, :
1

Translation
Specifications

1- ./
I

-~-~=:==-=~~
~11'1
I!11",

Output
Specifications

- ==:- -=-==-=~

~= -=--=-=-~~~il
....I\n
III1II

( Calculation
1 Specifications

I!~

~:~:;:t;o~s~ 1. /~i
(~--=-=~i'~ II
-

11iil

Extension

1 _Specifications

11111

'
I

----~

RPG Control
Card

"~"

:11

File
Description
Specifications

III,.!

II
"

II:

III

**
/*

Separator cards
End of file
------ Optional decks

'1
~-/

f--I

Figure 20.2 Arrangement of RPG II source deck.

One control card is required for each source deck. The card
specifies to the processor a number of general items of information which apply to both the assembly process and the entire application.

Control or Header
Card Specifications

402

REPORT PROGRAM GENERATOR, VERSION II

Columns 7 through 9 indicate the maximum number of bytes
that can be stored by the system used to compile the object program as follows:
Entry

008
012
016
024

032

Explanation
8,192 byte capacity
12,288
16,384 "
24,576
32,768

Column 10 indicates whether or not an object deck is to be
punched during assembly. An object program can be generated
with only a listing. It is sometimes desirable to check for programming errors so that corrections can be made before the final object
deck is punched. The following entries are used:
Entry

Explanation

Blank

Object program is punched into cards and a listing
with diagnostics is given.
Object program is not punched into cards, but the
listing with diagnostics is given.

Column 11 controls system halts and the printing of the program listing during assembly:
Entry

Explanation

Blank

Compilation stops if any programming errors are detected. A program listing with diagnostics is given.
Compilation stops if serious programming errors are
detected. No program listing is given.

Columns 12 through 14 indicate the maximum number of
bytes that can be stored in the system which will execute the object program.
Entry
Blank

Explanation

The size of the system that will execute the object
program is the same size as the system that compiled
the source program.
(Other entries are the same as for columns 7 through 9.)

REPORT PROGRAM GENERATOR, VERSION II

Column 15 is used to indicate whether or not the DEBUG
operation is to be performed during assembly.

Entry

Explanation

Blank
I

DEBUG operation is not performed.
DEBUG operation is performed.

Column 16 is not used.
Columns 17 through 20 are used to describe the format of
sterling fields used in sterling currency. Otherwise, the fields are
left blank. The formats will not be explained here.
Column 21 describes the type of punctuation used in numeric
constants; that is, numeric information retained for reference by
the object program at execution time. The entry also controls format of an UDATE field and the punctuation provided by certain
edit codes. This entry will not be explained here.
Columns 22 through 25 are not used.
Column 26 provides for altering the collating sequence of
characters. Normally every alphabetic, numerical, and special character holds an assigned position in relation to all other characters.
This special order is known as the collating sequence and is the
basis for all sorting (Figure 20.3). The System/3 uses a collating
sequence based on the way in which characters are represented in
the machine.
With this entry in the control card specification, the programmer may elect to change the collating sequence to one other than
that used by the machine. Or, two or more characters may be assigned the same position in a sequence. This sequence will then
apply to the compare operation using alphameric fields and also
to match fields. The alternate collating sequence will have no effect upon numerical compares, control fields, or table look-up operations. A Translation Table and Alternate Collating Sequence
Coding sheet is used with this entry. Every character that is to be
out of the normal sequence must be specified. Cards are punched
from the sheet and are placed in the source deck as shown in
Figure 20.2.
Columns 27 through 74 are not used.

403

Blank

(

;(

32
33

)

;

}

15
16

• (minus)

17
18

(underscore)

Figure 20.3 System/3 Model10 collating sequence.

File Description
Specifications

Each file to be used in an application must be described in the
source program. As previously stated, a printed report is also considered as an output file in RPG language. The programmer enters
the proper items on a File Description Specification sheet shown
in Figure 20.1. The description serves to:
1. Assign each file of the job a positive identification. Whenever possible, this should be either the true name of the file or an
easily recognizable symbol (such as PAYREC, ACCTSREC, INVREC). The assigned name will be used later on other specification
sheets to identify the files for data handling operations.
2. Assign an input or output device to every file.
3. Indicate whether the file will provide input data, serve for
output data, or both ..
4. Specify if an input or combined file is to be sequence
checked.

REPORT PROGRAM GENERATOR, VERSION II

5. Show which files must be completely processed before
the job is finished.
6. Show that a rile mayor may not be used for a specific job.
7. Specify whether additional I/O areas are required.
The specification sheet is designed so that only one line is required to describe one file.
Columns 1 through 6 and columns 75 through 80 were described in the previous section.
Columns 7 through 14 are used for the file name made up of
from one to eight characters. The name can be any combination
of digits and letters, but the first character must be a letter. The
name must begin in column 7 and cannot contain any special
characters or blanks. Every file must be given a name, but no two
files can be given the same name. The file must be referred to by
its assigned name throughout the program.
Column 15 indicates what kind of file is named as follows:
Entry
I

o
C

Explanation
Input file
Output file
Combined file

All files must be further described on input or output specification sheets. If either punching or printing is to be done on the
input deck, it must be specified as a combined file. If only one
input or combined file is used, it must be placed in the primary
hopper of the MFCU.
Column 16 is used to further describe an input or combined
file:
Entry

Explanation

Blank

Must be left blank for all output files.
Primary file, the main file. in the job, is used to control
the order in which cards from all files are processed.
Secondary file, any file not considered to be the primary file.
Table file, a file containing only related data items of
the same type, length, and decimal positions.
Demand file, an input file from which records can be
read by a special FORCE operation only. (The FORCE
operation can be used· to bypass normal RPG record
selection.)

T
D

405

406

REPORT PROGRAM GENERATOR, VERSION II

A primary file is required for every job but there can be several secondary files. Certain other restrictions and considerations
for this entry will not be discussed here.
Column 17 applies only to combined files, primary input
files, and secondary input files. It is used to indicate that the input
or combined file named in columns 7 through 14 is to be checked
for an end-of-file condition.
End of file means that there are no more cards to process because the end-of-file card (/* punched in columns I and 2) has
been read. Whenever an end of file occurs, a last record (LR) indicator turns on. The job terminates after all operations conditioned
by the LR indicator have been completed. The entry is especially
significant when there are two input or combined files which do
not contain the same number of cards. The entry in column 17
indicates which file causes an end of job.
Entry

Explanation

Blank

1. The job execution can end whether or not all of
the records in this file have been processed.
2. If column 17 is blank for all of the input files, all
records from every file must be processed before the
program can end. The LR indicator turns on only
after all files have reached end of file.
Check the file for end-of-file condition (input or combined file). If more than one file is to be checked,
the LR indicator turns on only after all files to be
checked reach the end of file.

Column 18 is used to describe the sequence of an input or
combined file. In conjunction with other entries on the Input
Specification sheet,· the entry indicates which record fields are to
be sequence checked, which fields are sequenced control fields,
and which records are to be matched for processing. The following entries are made here:
Entry

Explanation

Blank

No sequencing or matching. Always blank for output, table, and demand files.
Ascending sequence.
Descending sequence.

Columns 19 through 23 may be left blank.
Columns 24 through 27 are used to indicate the length in

REPORT PROGRAM GENERATOR, VERSION II

characters of both input and output file records. The length must
be within the size of the input or output device being used. For
example, the record length for cards is always 96. The maximum
length specified for the printer depends upon the number of printing positions available on the model being used: 96, 120, or 132.
The entry must be right-justified.
Columns 28 through 31 are not used.
Column 32 may be used to assign two input or output areas
to I/O files. When one I/O area is used, the cOlllPuter reads, pro..,
cesses, and punches one card at a time. When two I/O areas are
assigned, the computer reads or punches one card while processing
another card record. The use of two areas can make the job execution faster. Since more storage space is required for the two records, the programmer must determine if this additional space will
be available at execution time.
Entry

Explanation

Blank
1-9

One I/O area per file
Two I/O areas per file

Columns 33 and 34 are used to assign an overflow indicator
to the printer. This indicator signals that the last line on a page
has been printed. As a result, the correct action can be taken to
eject to the next page, print headings, and so on. The functions of
the indicator will not be discussed here.
Entry

Explanation

OA-OG or OV

Overflow indicator (anyone symbol)

Columns 35 through 38 are not used.
Column 39 applies only to tables that are to be loaded at execution time and to output files that are to be printed on the line
printer. The column is used in conjunction with the Extension
Specification sheet to further describe the structure of the tables.
The column is also used in conjunction with the Line Counter
Specifications to further describe printed output.
Entry

Explanation

Blank
E

No line counter or extension specifications.
Extension specifications are used.
Line counter specifications are used.

407

408

REPORT PROGRAM GENERATOR, VERSION II

Columns 40 through 46 are used to assign a specific input or
output device to the file names in columns 7 through 14. Each
device normally has only one file associated with it. The device
code must start in column 40.
Entry

Explanation

MFCVI

Primary hopper of the MFCV.
Secondary hopper of the MFCV.
Printer. If the printer has a dual carriage feature, this
entry refers to the left carriage.
Right carriage of the printer, if the printer is equipped
with the dual carriage feature.

MFCV2

PRINTER
PRINTER2

Note: An output card file must first enter the System/3 by
way of the MFCU, even though it~!llay be blank cards before processing. The file is assigned to an MFCU stacker on the Output
Format Specification sheet.
Columns 47 through 70 are not used.
Columns 71 and 72 are used to assign an external indicator
to condition the file described on this line. The indicators control
whether or not a file is to be used for a job at execution time.
For example, for one run it may be required to produce both
output cards and a printed report. For the next run, only the report may be needed. To avoid using two separate programs for
the different runs, the external indicator can be set on or off according to the output required. If the indicator is on, the corresponding file is produced; if it is off, the file is not produced. Once
the indicator has been set, it cannot be changed during an execution run. When the execution program is being loaded, provision
is made to preset the indicators at that time.
External indicators may be used for all files: input, output,
and combined files but not for table input files.

Input Specifications

Entry

Explanation

VI - V8

Indicators to condition a file

Figure 20.4 is the RPG II Input Specifications form. It is used to
further describe and identify the input files to a procedure. On
this sheet the programmer may:

IB~

Form X21-9094·· 1
Printed in U.S.A.

I nternational Business Machines Corporation

RPG

INPUT SPECIFICATIONS
1

Date _ _ _ _ __
Page

75 76 77 78 79 80

[I]

Program
Identification

Program _ _ _ _ _ _ _ _ _ _ _ _ _ _ __

IIIIIII

Programmer

Record Identification Codes

Field Location
>

1 - 1- - - , - - - - - ;

iii

Filename

Line
0.

I~Zllro
~

g .Q §C "'C
5- E·g 0
~ ~ 8 £

(l)

Position

~Iel~
o

~ ~

Position

~I§I

Position

I~I~

~I§I i ~

Field Name
From

~
~

]

Field
Indicators

, ,I I-I I,.,"~
5

~~
~~ ~

Sign

~.:
C)

C7)

:rl

*~ u.~
8 ::'u

'.ro "''''00

Plus Minus or
Blank

1 !1 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I

IoGTTrl
013
014

015
016

l--l----l----l--i-l----l-l----I------!---+-H

1 1 1 1 I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I

I
~

--+--+---I-+-+-+--U-l-----W- 1 1 1 1 1--1-1 1 1 1 1 1 1 1 1-++ 1 1 1 1 1--+--1--+-1 1 1 1 1 1 I 1 1 1 1 I 1 1 1 1 1 I 1 I 1 1 1 1 1 1 1 1 1 I I I I

1.0

III

1--,-- --- -U-t=1:t1!t:ttt.-I+++-+-l
I I II I
~±:l-H-t±HTl i 1111111 UTI 111111111 , 1111 ! I
HHH1l ~~ 1+ - I I I I I I I I

I . - - - - --

Figure 20.4 Input specifications.

I I

I I I I

410

REPORT PROGRAM GENERATOR, VERSION II

1. Identify the different types of card records within a single
file, for example, master name, address, transactions, and so on.
2. Specify the sequence, if any, of the various types of cards
in a file.
3. Define card fields and identify the data by name in each
field.
4. Set indicators based on individual fields.

Describe control fields.

6. Specify fields to be used for matching records or for sequence checking of data.
7. Direct input or combined file cards to specific stackers in
the MFCU after processing .
.The input specifications may be divided into two categories:
1. File and card type identification describe each input card
and its relationship to other cards in the file.
2. Field description describes the individual fields in cards.
These entries on the sheet must be written one line lower than file
and card type identification entries.

Columns 1 through 7 and 75 through 80. The uses of these
columns on fhe sheet have been described in the previous section.
Columns 7 through 14 are used to identify the input file by
name, the same file name previously assigned in the File Description Specifications.
Columns 15 and 16 are used to assign sequence numbers to
different types of cards within an input file. For example, in a
billing or invoice-writing procedure, the sequence of information
on the printed form is predetermined. Normally, the customer
name appears first, followed by street address, city and state, miscellaneous data, and transactions or items to be billed or invoiced.
Transactions .are listed in the body of the form after the heading
information has been printed. The transactions may be in no
particular sequence.
For this application, each type of card is assigned a sequence
number beginning with the name as 01, street address as 02, and
so on, with the transaction cards assigned the highest number in

REPORT PROGRAM GENERATOR, VERSION II

the sequence. Numbers from 01 to 99 may be assigned, but the
sequence must begin with 01.
Figure 20.5 shows how the various types of input cards are
arranged in the file for this job. A card out of sequence will cause
the computer to halt at execution time for corrective action by
the operator.
Entry

Explanation

01 - 99

Check for card sequence. No check is made on the
sequence of data.
Sequence of cards is not checked.

Any two alphabetic characters

Transaction

(07)

(04)

(03)

(O~)

Customer name

(01)

(07)

(04)

(03)

(02)

(01)

Figure 20.5 Card sequence. Invoice writing procedure.

411

412

REPORT PROGRAM GENERATOR, VERSION II

Column 17 is used to indicate whether one card, or more than
one card, of a particular type can be in the card sequence. An entry
is made only if the corresponding numerical sequence entry is
made in columns 16 and 17. Thus, in a billing operation the program can insure that no more than one customer name, address,
or city/state card is in the file for each bill being printed. However, the program can be written to permit more than one transaction card per bill.
Entry

Explanation

Blank

Cards are not being checked for sequence. Columns
15 and 16 have alphabetic entries.
Only one card of this type may be present in the sequenced group.
One or more cards of a particular type may be present
in the sequenced group.

If sequence checking entries are made in columns 15, 16, and
17, the types of cards checked must be in the file.
Column 18 makes this rule optional. An entry indicates that
the card type being sequenced checked need not be in the file.
For example, some invoiced customers may not have a street address. Therefore, a street address card does not have to be present
for every customer.
Figure 20.6 shows RPG coding in columns 16 through 18 for
the invoice example just discussed. The various types of cards are
numbered 01, 02, 03, and 07. (Any two-digit ascending sequence
may be used, starting with 01.) The file cards are sequence checked
to be sure that each card is in its proper order for the invoice.
Street address cards are considered optional.
In some instances, it might be required to run a report showing all customers, even though some customers have not made
purchases during the period. For this procedure, item cards are
also considered to be optional.
It is standard practice to construct unit-record files so that
different types of cards are used for specific kinds of data. Each
type of card is distinguished by unique hole punching in specified
columns. For example, items purchased by a customer may be
coded P in column 10 while items returned for credit may be coded
C in the same column. The RPG system provides for recognition
of card types.
Columns 19 and 20 are used to assign arecord identifying
indicator to each input record type.

REPORT PROGRAM GENERATOR, VERSION II

413

41
, 6/, X X
I

Date

MONJ HL'< BI LLt NG

Program

A . COD E R

Programmer

S«I

e:*

Filename

Line
(1)

(1)

0 2

0 4

(1)

..0

::T

(1)

u..
3

(/)

~
(1)

~*~
';:; 0

(5 (1)
:!:!
c: 'C

0
E ';:; u:I C.
(1)
2 0 a:

10 11 12 13 14 15 16 17 18 19 20 21

---

8 \ L L R E C D f/J \ I
I¢ 2 \0

0 31 ¢ 1 NO -

Name Card
Street Address Card

City/State Card

Transaction Card

Figure 20.6 RPG coding. Invoice example.

It should be noted here that the indicators available to the
program are not machine features. Program indicators are turned
on or off by setting binary bits in predetermined locations of storage. Thus, program indicators may be used by the programmer
in virtually the same manner as if they were designed into the
computer.
When a specific indicator is assigned to a type of card record,
the indicator is turned on as the card is selected for processing at
execution time. The indicator specifies throughout the rest of the
program cycle which card type has just been selected. As will be
shown later, the indicators are used to condition calculation.
Columns 19 and 20 are also used to indicate that fields named
in columns 53 through 58 of the following specifications are lookahead fields. That is, a field in the next card to be processed may
be examined to determine what operations should be done next.
The operation is similar to using the first station reading brushes in
an accounting machine to control the calculations to be done when
a card reaches the next reading station.

414

REPORT PROGRAM GENERATOR, VERSION II

The number of previous RPG entries have identified the various input files and records to the program. This identification
must also be related to actual card punching of corresponding
codes as it appears in the records.
Columns 21 through 41 are used for this purpose. No entries
are required if all records are to be processed in the same manner.
A maximum of three identifying characters may be shown on one
specification line. Seven columns are provided for the description
of one character or combination of punches in the code.
Columns 21 through 24, 28 through 31, and 36 through 38
specify the position of the character or code in the record.
Entry

Explanation

Blank
01 - 96

No record identification is needed.
Card column number of one character in the code.

Columns 25,32, and 39 indicate whether the specified character is or is not in the record. The entry is equivalent to an X or
No-X designation for punched-card accounting machines.
Entry

Explanation

Blank
N

Character is present in the code.
Character is not present in the code.

Columns 26, 33, and 40 indicate whether the entire character,
only the zone portion, or only the digit portion is used in the
identifying code.
Entry

c
Z

Explanation
Entire character
Zone portion of the character
Digit portion of the character

When setting up record identifying codes, it is important to
consider that many characters have the same zone or the same
digit portion.
Columns 27, 34, and 41 identify the actual character. Any
alphabetic character, special character, or digit can be used to
serve as the code or part of the code.
Figure 20.7 shows four types of input records identified by
the characters A, B, and C and the digit 7 in column 10. Figure

REPORT PROGRAM GENERATOR, VERSION II

20.8 shows a payfile input record identified by the digit 6 in column 1, the digit 8 in column 90, and the zone portion of the
character B in column 91.

4/16/XX
r
I

Date
Program

MONTHLY

[

BILL' NG

A. C Q 0 E B

Programmer

....

Record IdE

:0
.!:

c:*

~*,-

Filename

Line

eQ)

::l

::r
Q)

u..
3

(J')

'';; 0

... 0c:
Q)

c:
Q)

E '';;

::l

Position

0 4

----uu~
N

Positi

co
~

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 272829::

1St LL R E C D ¢I l
I rp
I
,r; 2 10 21¢
I

.......u

0 31

2 rjJ

,¢ 7N 0 3 ¢

cA
¢ CB

I¢
I

t¢
I¢

CC
C7

Figure 20.7 Four types of input records.

A maximum of three identifying characters can be described
on one specification line. If the identification code consists of
more than three characters, an AND line must be used. This
means that the first three identifying characters are described on
one line. The additional identifying characters are described in
as many of the following lines as are needed. The word AND is
written in columns 14 through 16.
Figure 20.9 shows a payfile identified by four characters,
using the AND line: digit 6 in column 10, digit 8 in column 90,
zone portion of the character B in column 91, and the digit 5 in
column 92.

415

416

REPORT PROGRAM GENERATOR, VERSION II

IB~
Date

International Business Machines Corpora~ion

RPG

4- /2()t~

7
II() () R. L Y Pt9YRo-LL
JONeS
Programmer /II. FIE
I

Punching

Program

Instruction

INPUT SPECIFICATI
Graphic
Punch

I I
I I

I I \
1 I 1\

Record Identification Codes

FielJ

l!'.

Filename

line

~
I-

!
3

0 2

0 3

I--

=0
]"§ ~
E "

~ ilg

0 1

iii

~·o

U III

Position

Ii;

u u

Ii;

Position

~e ~

(5 ~ ~

-ll ~

~o~~
ei~

Position

~~cS

Zuu

9 1011 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 J6 J7 J8 J9 40 41 42 4J

AY FI L I" JI

1'¢ N/) 8

L--:.-I-I-...L.......
L--

.....

L-L-I-...

....... ..............

L--L--I..--"-"'

'"--

...... L.......

]

--=

Figure 20.8 Payfile input record.

IB~
Date

International Business Machines corporal

RPG

/3/XX

Punching

Program

S.qi..,q,e r P;Il YRoil.

Programmer

~/1....:....::i:.:..F-------.:IC=--_.Tt_O_N_~_~_ _ _ __
~

~"u

Instruction

1 Graphic

Punch

I I

I \

I ·1

-_i*a

=§....
~~ ~g~
~ g~

Record Identification Codes

]

Filename

INPUT SPECIFIC1

t; ::,

Position

~~ucS
~~

Position

~~u6
~~

Position

~~~

~zC30U5
~ ~ ~ II~

9 10 11 12 lJ 14 15 16 17 18 19 20 21 22 2J 24 25 26 27 28 29 JO Jl J2 JJ J4 J5 J6 J7 J8 J9 40 41 42 4

/ t.

ll~

tit

9 I

1\
1\

Figure 20.9 Use of AND for record identification.

A particular card type may be identified by two different
codes. In this case, an OR line can be used to indicate that either
one of the codes may be present to identify the record. The word

REPORT PROGRAM GENERATOR, VERSION II

OR is written in columns 14 and 15. Figure 20.10 shows a record
identified with the digit 5 in column 1 and the digit 8 in column 2
OR the digit 2 in column 1.
IB~

International Business Machin8$ C

INt/,ENTOR Y

Program

~
~

Filename

o.g

c.
>

I-

~
5

I /

......... "" ........ 1

(J)

Punch

I I
I I

)
/

Record Identification Codes

]
'"
';'.
Line

Graphic

Punching
Instruction

SrocK

.I.N.

Programmer

INPUT SPECIFI

RPG

t(jUX

Date

:>!

Position

o
~

_0 E
_N ..c:"

-~.e

Position

~
~

Position

zuu

~Q\
o~

o~~

zuu

9 10 11 12 13 14 16 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41

/>IV EN TR.

r8e.

I S

oR.

........1-......1......

.........................

vvl.--

"---

. . . . . Ll

Figure 20.10 Use of OR for record identification.

Column 42 is used to specify an MFCU stacker (1-4) for the
input cards after processing. If no entry is made, primary cards
are automatically placed in stacker 1, the primary hopper. Secondary cards are placed in the secondary hopper, stacker 4. Combined files may be stacker-selected either on this form or on the
Output Format Specification form.
If the same stacker is selected for both an input and an output
file, a card from the output file is put into the stacker before a
card from the input or combined file.
Entry

Explanation

Blank

Cards automatically fall into a primary or secondary
stacker, 1 or 4.
Stacker where card type will be stacked.

1-4

Column 43 is not used.
Columns 44 through 48 describe the record fields of the input

files. Columns 44 through 47 contain the number of the card column in which the field begins. Columns 48 through 51 contain
the number of the card column in which the field ends. Both

417

418

REPORT PROGRAM GENERATOR, VERSION II

numbers (from 1-96) must be right-justified on the form and the
leading zeros may be omitted~ A single column· field is defined
by writing the same number in both the From and To entries.
Numerical fields may have a maximum length of 15 characters.Alphameric fields may be up to 256 characters long. However, fields read in from cards are restricted to the length of one
punched card.
The number of positions to the right of the decimal is indicated for numerical fields in column 52. If there are no decimal
positions, enter a zero.
Entry

Explanation

Blank
0-9

Alphameric field.
Number of decimal positions in a numerical field.

Input card fields are named in columns 53 through 58. The
same name is used throughout the program to refer to the field.
Only fields to be used in the processing need to be named. Rules:
1. Must begin with an alphabetic character.
2. May not contain special characters or blanks.
3. May be any combination of from one to six numbers
and letters.
Figure 20.11 is an example of field description for a master
salesman name and commission card.
If two or more card types contain identical fields, each field
must be described. However, to avoid duplicate coding, an OR
entry may be used. The OR relationship means that the fields
named may be found in either one of the card types. OR lines
can be used when 1. Two or more card types have the same fields in the same
positions.
2. Two or more card types have some fields which are the
same and some fields which differ in location, length, or type
of data.
The word OR is written in columns 14 and 15 to indicate an
OR line as shown inFigure 20.12.

IBM
Date

/1)

International Business Machines Corporation

RPG

/211xx

GX21·9094·1 U/M050
Printed in U.S.A.

INPUT SPECIFICATIONS
1

Page

SI1,L£5 RC'PORT

Program
Programmer

75 76 77 78 79 80

~;~~;~f:ation

/1. C" .f) c R.
Record Identification Codes
Field Location

iI-I---,..-----\

Iii

Filename

Line
c.
>

Position

~
011

~ §~

151AIL 1£15

\418

~ 1°1 ~'

011

Position

j§

1 1 I Position

jl§llj*I~f

Field Name
From

016

311

017

81¢

81Slzls~IL IEI.s

III
21~

018

019
110
111
112
113
1 14

.j:>.
~

IMlnlNIMo

91 IrlEIRlelY
2161 INIIlJMIE
31¢1 IPIE'IPlr
3151 Ie Ivls 111#10

015

115

Figure 20.11 Field description.

Field
Indicators

~
c;

g ~

/I¢I Ie 16'

013

I.C

012

014

IIIIIII

,~ I I. I'"" I ~,;,."
c::

~
a;
u:

Sign
&";00

Plus MinUS or

Blank

IBM

International Business Machines Corporation

DateW~

RPG

GX21·9094·1 U/M050
Printed in U.S.A.

INPUT SPECIFICATIONS
1

Page

S.f1LFS REPoRT
~1 FIE JaNES

Program

75 76 77 78 79 80

~;~~~~f:ation

[L I I I I I

Programmer

Record Identification Codes

ol::o

Filename

Line
0-

,~r'
i~

~.~ ~

~I~I!

I Position

Position

il§1

~I~
il§l\j ~

Field Name

From

15 Ai £5

o
o

]

A. 8.

liS

]
e
]

~:~

':;

g'2'

:;;;0

I I I I··';~
Sign

Zero

Plus Minus or
Blank

Position

u::

4&
6&
Ac

Field
Indicators

Field Location

DE. Pi
E"'tI PLIA10

5¢ f/J 1 iE~1
7¢ 2C o.si

t ol£pr
/14- IClhI p LINIO
5; fJl rcM

{,b

17_i¢ zc:osr

5
9

11
I
- , - f---1
1 ?

10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 5960 61 62 63 64 65 66 6768 69 70 71 72 73 74

14-

AL £5

AS
O~

I¢

4'
66

DIE PT

£1,4J p'l. No

5'1> o I TeM
7!¢ 2C osi

Figure 20.12 Field description. Two or more cards.

REPORT PROGRAM GENERATOR, VERSION II

Field description entries can also be used for the page numbering of reports. These specifications will not be described here.
Columns 59 and 60 are used to assign control level indicators
to specified fields of input records.
Card files normally enter the computer system in a predetermined sequence produced by mechanical sorting of control fields.
When two or more types of cards are combined in the sa'me sequence, groups of cards frequently contain like control numbers.
The groups may represent such conditions as several purchases
made by one customer, several job cards for the same employee,
a number of inventory transactions for the same part number,
and so on.
When the sequenced files are processed, amounts and quantities are accumulated and punched or printed by groups. With an
accounting machine, this kind of operation is called group printing. As previously described, totals may be printed in minor, intermediate, or major groups depending upon the sequence of the files.
The RPG language provides for programming the System/3
to operate upon control fields in a manner similar to that of a
punched-card accounting machine but with considerably more flexibility. For example, up to nine control levels may be programmed
by assigning control level indicators to specified fields. Whenever
a card with a control field is read, the field is compared with the
same field in the previous card. The accounting machine compares cards at two different reading stations. The computer compares data in the control fields from two card records in storage.
When a control break occurs; that is, when a comparison between two fields is unequal, the corresponding indicator turns on.
When the indicator is on, operations can be performed to cause
total printing or punching as required. Normally control level
indica tors are used to 1. Condition calculations to be performed whenever information in a control field changes.
2. Condition summary punching or total printing to be done
after totals have been found for one control group.
3. Condition operations to be done on the card record that
causes a change in a control field, for example, group indicate the
next group.

421

422

REPORT PROGRAM GENERATOR, VERSION II

Entry

Explanation

LI-L9

Any control level indicator .

Figure 20.13 shows entries to assign three control level indicators to the fields Employee Number, Department, and Location.
L3 is the highest indicator and is equivalent to a major control in
an accounting machine. When L3 is turned on, all lower level indicators are also turned on.
Additional rules, considerations, and restrictions for assigning
control levels can be found in the IBM reference manuals for
RPG II.
Columns 61 and 62 are used to compare records from two or
more input and combined files. Records can be matched by one
field, by as many as nine fields, or even by entire records. When
match fields from a primary card are the same as match fields from
a secondary card, the matching record (MR) indicator turns on.
Matched or unmatched records from two files are processed
as follows:
1. When a card from the primary file matches a card from
the secondary file, the primary card is processed first.

2. When cards from ascending files do not match, the card
with the lower match field is processed first regardless of the file.
Refer to Figure 19.14. When cards from descending files do not
match, the card with the higher match field is processed first.
3. A card type with no matching field specification is processed immediately after the card it follows. The MR indicator
is off. If this card type is first in the file, it is processed first even
if it is not in the primary file as shown in Figure 20.14.
Entry

Explanation

MI-M9

Any matching level.

Rules:
1. Sequence checking is automatically done for all card types
with matching field specifications. An error in sequence stops the
program. All files containing records to be matched must be in
the same sequence - either ascending or descending.

IBM
Date

International Business Machines Corporation

RPG

/5"!)(~

Page

L/ .5 T /N ~

IlLIC/(

Programmer

Printed in US_A.

p)l Y,

Program

GX21-9094-1 U/M050

INPUT SPECIFICATIONS
2

75 76 77 78 79 80

[I]

~~~~;~f:atiOn

JON€..5
Record Identification Codes

Field
Indicators

Field Location

Filename

Line

~~:o
~; !
IE .g _
...:..
2 _

a.
>-

f0-

(.f)

~f-I- - - . - - - - - - l

Position

~Iel~

Position

E
z" ~
u u

Position

::z21°
o NI '"
uu
~

.<:

~I~

jl§llj ~

Field Name

From

Ip
I

a 13

/ f

3 rJ

Nil

014

8¢

/)1 VS olN

015

016

33
2
78

8¢
33

I I I I I I I

1c.IOI I 121¢

-r-+-

-I-

~
u::

MinUS

or
Blank

M£

51-! IF,
DE PT

bt*

c 1/ TNL3

L 0

I--+~

-t--+!

~t·r

-t

j-~. +.
t-

;

t- t-t-

;t-

1 tt

Sign

Plus

E.M PLINlolLl1

+---+--+-+-+-+-+- -1-+-+-+--+ +--+-++++-t----I-l--+-+_+_

,a: I I. I,.. I'·~,;,;""
"""

..LjP~PIL
INIOlll1z.
/141 IoT£ rl

-- + - - +I' - ~ I

f INIc.ln

II-t,I IIII1 1111,-+-i-++-t--+-+--++++

1 i3 !

If>

- -I

1: 2 •

A A

.5 _r~~
~ ~~
8 :2u

a 12

lJy ell RoO 5

tii

all

017

IIIIIII

I
j

l
Figure 20.13 Control level indicators.

-.~~+~L;-~ ---~:_'-l'f
-1) t r·Lft' ILt~ t-t- t--t-tt-++ . -l+- - -_t 1 t +-t+,
+_1

I I

L -; .
I 1

~
i

tt-

424

REPORT PROGRAM GENERATOR, VERSION II

*Match
field

The card with no match field
is processed immediately after
the card it follows, regardless
of file.

--} ®
- - - --}

When cards have no match,
the card with the lowest
sequence number is processed
first, regardless of file.

CD Primary file cards are processed
- - - -

before all matched secondary
file cards.

Processing sequence

Figure 20.14 Processing order according to matching records.

REPORT PROGRAM GENERATOR, VERSION II

2. At least one card type from each of the primary and secondary files must have matching fields.
3. The same number of matching fields must be specified for
all card types used in matching. The same MR indicators must be
used for all types.
4. All match fields given the same MR indicator must be the
same length.
5. If more than one matching field is specified for a card
type, all the fields are treated as one continuous matching field.
They are combined according to ascending sequence of matching
record indicators. The low field is on the right, the high field on
the left, as shown below.
Division

Department

Employee Number

6. Matching fields may be either alphamerical or numerical.
However, all matching fields assigned a given level indicator (Ml
through M9) are considered as numerical fields if anyone of those
fields is described as numerical.
7. Only the digit portions of numerical fields are compared.
Even though a field is negative, it is considered to be positive for
comparing purposes. Thus, +123 will match -123.
8. Whenever more than one matching indicator is used, all
match fields must match before the MR indicator turns on.
9. Field names are ignored in, matching record operations.
Therefore, fields from different card types which have the same
match level may have different names.
Figure 20.15 is an example of three card types which are to
be used in a matching operation. All card types have three matching fields specified and all use the same indicators (Ml, M2, and
M3) to indicate which fields must match. The MR indicator turns
on only if all three match fields in either of the card types from
the MASTER file are the same as all three fields from the card in
the weekly file.
Figure 20.16 shows sequence checking for a single master
card file.

425

IB~

I nternational Business Machines Corporation

RPG

2/2'l/xx
~I

Date

I I I I

uJair/(LY P,q Y

Program

Programmer

/"',f€iJ

GX21-9094-1 U/M050

Printed in U.s_A_

INPUT SPECIFICATIONS

~k!f)
Record Identification Codes

.;..

iii

Filename

~§

I-

~
4

75 76 77 78 79 80

Program

III,,,,

;:l c£ PRoGRA MMEf!
Record Identification Codes

line

;!
Field Name
From

-lS

Field
Indicators

::;;u

,I~ I I I"""""
a:

~
Q;
u::

Sign

'.0 """."

Plus M,nus or
Blank

3 4 51617,8,9 ,lO,II,12,13,14P5 ,16P7118P9 ,2°121 ,22 ,23,24125126127128,29, 30,31P2133134135,36,37,38P9140141142143144,45,46,47148,49,5°,5'152153,54,55,56,57,58159,6°16',62163,64165,66167,68169,7°17,,72,73,74
rl.¢\N\O\4
91,¢ IlVIo\S
91¢\N\O Ibll
IIMI,llI.slrl£IR

011

121111111111111111111111111111111111111111111;1~111;1:II~I~I:I:U~II
~'1--;-4
!I
0: 5 i
- t·_-t
a i6

rl 1111111111111

11'11 I Izl.¢1 I/lNl/lrlAll.
21/1 I 131b\ \NI,qIMI£

.3171 I 13181 101/1'1
311\ 1 141¢1 ILIOI£:
41/1 1141/1 ~1f{l/lflr

.j:;.
t-.)

....,J

Figure 20.16 Sequence checking, master file.

IMII
\MIl.

428

REPORT PROGRAM GENERATOR, VERSION II

Columns 63 and 64 are used for field record relation under
three different situations:
1. More than one card in an input file can be exactly related

by entering an OR in columns 14 and 15. As shown in Figure
20.17, each record is assigned a record identifying indicator and
entries are made for the unique card code in each card type. All
card columns are commonly used in the specified OR relationship.
That is, at execution time, cards with 5 punched in column 1 will
be considered equaUy with cards punched with 6 in column 1.
However, it may be necessary to provide for exceptions to
this rule, particularly when all fields in the two types of cards are
not the same. Therefore, there must be some way to relate specific fields to particular OR cards as required. To do this, columns
63 and 64 should contain the same record identifying indicator
shown in columns 19 and 20 of the card type where that field is
located. In Figure 20.17, Field C is in columns 40 through 50 of
the digit 5 card and Field D is in columns 60 through 70 of the
digit 6 card. Field C is related to record indicator 14 and Field D
is related to record indicator 16.
2. Data from an input card field may only be required when
a control break or matching record 'occurs. In this case, columns
63 and 64 contain only the entry for the proper level indicators
(LI-L9) or the MR indicator. An example is shown in Figure
20.18. The manager's name will be printed on the output report
only when a control break in department occurs.

3. Columns 63 and 64 may also be used to condition aspecification by an external indicator (UI-U8). These indicators may
be set prior to actual processing by a special control card used at
system initialization time. When the indicator is on, the specification applies; when it is off, the specification is ignored.
Columns 65 through 70 are used to determine if the data content of an input record field is plus, minus, zero, or blank. From
one to 99 program indicators are available. By entry in one of the
field indicator spaces on the form, a specified indicator is associated with the field named in columns 53 through 58. At execution
time, the indicator is turned on if a given condition is true and is
turned off if the condition is not true. The indicators are used to
condition either calculation or output operations.

IB~

I nternational Business Machines Corporation

RPG

12 / :l7 /J(J{

Date

GX21-9094-1 U/M050
Printed in U.S.A_

INPUT SPECIFICATIONS
1 .2

r I

Program

Page

w£€ K'L 'r £x PE"N.st;- LI6 T

Programmer

75 76 77 78 79 80

[I]

Program

Identification

FRe-P Folf!O
Record Identification Codes
Field Location

Filename

Line

c.
to-

~Io~ro
~ 1~'3 ~

Position

fI§II

Position

Ifl§ I'

Position

1rl1~

~1§ljli ~

Field Name

From

g
§
U

zl9S

31~1

o
o

~I¢

'Id

islf6lz Ifi/l€lL.1 Ole
7f!2S1 IF/I/!"/ L/g1t:

~
1 3

1 4

oj:.

ZUUIIlCl.

1 5

Figure 20.17 Field record relationship.

1 - 1- - - , - - - - - ,

75 76 77 78 79 80
Program

L /s 1111/4

page[]]

IJ,)

Line

GX21-9094-1 U/M050
Printed in U.s.A.

INPUT SPECIFICATIONS

~I{'

$ I~

E'~

~f-I---r--I

o
.,

0
0
0
0
0
0
0
0

R
B
R
B
R
R
B
B

R
R
R
R
R
R
B
R

R
R
R
R
R
R
R
R

0
0
0

B
B
B

R
R
R

0
0
0
0
0
0
0
0
0
0
0

ADD
Z-ADD
SUB
Z-SUB
MULT
DIV
MVR
X FOOT

Move Factor 2 into Result Field, right justified.
Move Factor 2 into Result Field, left justified.
Move zone from low·order position of Factor 2 to
low-order position of Result Field.
Move zone from high-order position of alphameric
Factor 2 to high-order of alphameric Result Field.
Move zone from low-order position of Factor 2 to highorder position of alphameric Result Field_
Move zone from high-order position of alphameric
Factor 2 to low-order position of Result Field.

MOVE
MOVEL
MLLZO

0
0
0
0
0
0
0
0
0
0
0

MHHZO

MLHZO

MHLZO

Compare Factor 1 to Factor 2.
Identify the zone in the leftmost position of an
alphameric Result Field.

CaMP
TESTZ

0
0

R
B

Setting
Indicators

Set one, two, or three specific indicators on.
Set one, two, or three specific indicators off.

SETON
SETOF

Branch to another RPG 11 calculation specification line.
I dentify the name in Factor 1 as a destination label
to which GOTO may branch.

GOTO
TAG

0
0
0

B
B

Branching
within
RPG 11

0
0
0
0

Lookup
Operations

Table Lookup.
Array Lookup.

LOKUP
LOKUP

0
0

Subroutine

Beginning of the subroutine.
End of the subroutine.

Compare
and zone
testing

Cl~
0

"t;;,

Add Factor 2 to Factor 1.
Clear Result Field and add Factor 2.
Subtract Factor 2 from Factor 1.
Clear Result Field and subtract Factor 2.
Multiply Factor 1 by Factor 2.
Divide Factor 1 by Factor 2.
Move remainder of preceding division to a Result Field.
Sum elements of the array and put sum in Result Field.

Move
Operation

~.g

5
~
u.

t::

.,,,

J:~

0
0
0
0
0
0
0
0

B
B
B

B
R

B
B

R
R

B
B

R
R

B
R

R
B

B
B

0
0

R
R

0
0

B
B

R
R

B
B

0
0
0
0

B
B
R

B
B
B

B
B

R
B

B
B

operations

Call to execute the subroutine.

BEGSR
ENDSR
EXSR

Program
Control

Forcing record to be read next.
Forcing outputjlrinting.

FORCE
EXCPT

Debug
Function

Aid in finding programming errors.

DEBUG

0
0

0- Optional
R - Required
B - Blank

Figure 21.2 RPG II operation summary.

2. All operations are performed in the order specified on the
calculation form.
3. All operations conditioned by control level indicators in
columns 7 and 8, except those which are part of a subroutine, must
follow those that are not conditioned by control level indicators.

Columns 43 through 48 are used to name the field which will
contain the result of the operation specified in columns 28 through
32. A new field may be named, either alphameric or numerical.
The following rules apply:
1. The result field must begin with an alphabetic character
and contain no blanks or special characters.

RPG II CALCULATION

2. If the field has not been described elsewhere, columns 49
through 52 should contain an entry to define the length of the
new field.
3. The field may also be described elsewhere, but it must be
described in exactly the same manner.

Columns 49 through 51 are used to designate the length of a
result field. Numerical fields may have a maximum length of 15
digits. Alphameric fields may be up to 256 characters long. If the
designated length is not sufficient to hold the result, digits or
characters will be dropped.
Column 52 indicates the number of positions to the right of
the decimal point in a numerical field. If the result is alphameric,
the column is left blank. If the numerical field has no decimal positions, enter a zero.
The number of decimal positions can never be greater than
the length of the field; the maximum number of positions is 9.
However, the number of positions may be more than the
number of decimal positions that actually result from an operation. In this case, zeros are filled in to the right. If the number of
decimal positions is less than the actual number that results from
an operation, the extra positions to the right are dropped.
Column 53 is used to indicate that the contents of the result
field are to be half adjusted or rounded.
Entry

Explanation

Blank

Do not half adjust.
Half adjust.

The adjustment is made by adding 5 to the first digit to the
right of the last decimal position specified in column 52. If the
field is negative, a minus 5 is added. All decimal positions to the
right of the number specified in column 52 are dropped.

Example:

123.4679
5

Product of a multiply operation.
Add 5 to the digit to the right of
the last decimal position specified.
123.5XXX Drop all decimal positions beyond
the number specified in column 52.
123.5
Result after half adjusting.

441

442

RPG II CALCULATION

Column 53 should be left blank for all nonarithmetic
operations.
Columns 54 through 59 are used to specify indicators that
are turned on by the result of the operation specified in columns
28 through 32. Following operations can then be conditioned by
testing these indicators.
Three fields are provided for result indicators: columns 54
and 55, 56 and 57, and 58 and 59. Each field is used for different
conditions: plus or high, minus or low, and zero or equal.
Indicators may also be turned on or off by the special operations SETON and SETOF. Thus, the condition of an indicator can
be under control of the programmer. Normally, only indicators
01-99 and H1-H9 are used in these fields, but other types of indicators can be used if required.
Entry

Explanation

01-99
H1-H9
L1-L9

Any
Any
Any
Last
Any

LR
OA-OG,OV

two-digit number.
halt indicator.
control level indicator.
record indicator.
overflow indicator.

Columns 54 and 55 (plus or high): An indicator specified in
this field is turned on when:
1. The result field in an arithmetic operation is plus.
2. FACTOR 1 is higher than FACTOR 2 in a compare
operation.
3. FACTOR 2 is higher than FACTOR 1 in a table look-up
operation.
4. The zone tested in a TESTZ operation is a plus zone.

Columns 56 and 57 (minus or low). An indicator specified in
this field turns on when:
1. The result field in an arithmetic field is minus.
2. FACTOR I is lower than FACTOR 2 in a compare
operation.
3. FACTOR 2 is lower than FACTOR 1 in a table look-up
operation.

RPG II CALCULATION

4. The zone tested in a TESTZ operation is a minus zone.
Columns 58 and 59 (zero or equal). An indicator specified in
this field turns on when:
1. The result field in an arithmetic operation is zero.
2. FACTOR I is equal to FACTOR 2 in a compare operation.
3. FACTOR 2 is equal to FACTOR I in a table look-up
operation.
4. The zone tested in a TESTZ operation is neither plus or
minus.
The following rules also apply to result indicators:
1. When the same indicator is used to test the result of more
than one operation, the operation last performed determines the
setting of the indicator.
2. Any indicators to be turned on or off by the SETON or
SETOF operations can be entered in any of the three fields (columns 54 through 59, as shown in Figure 21.3.
3. Halt indicators are normally used only to check for error
conditions.
Resulting
Indicators
Arithmetic
Zero

Plus IMinus

Operation

Factor 2

Result Field

Field
:~ !.
b 1;;;
Length 0.. .:l

~ ~
.~

Compare
High Low Equal
1>2 1< 2 1 = 2

~
I

Lookup
Table (Factor 2) is
High

Low Equal

27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 ,

i¢ I

Sf TON

(Og

S ETON
S er OF

HI
-

Figure 21.3 Result indicators.

443

444

RPG II CALCULATION

Columns 60 through 74 on the Calculation Specifications
sheet are used for comments.

Use of Result
Indicators

Figure 21.4 is a program calculation routine that illustrates the
use of the result indicators just described. The routine calculates
gross pay by employee in a payroll application. The routine is also
flowcharted in the illustration to show the logic of the procedure.
The item numbers below refer to line numbers on the Calculation Specifications sheet.
01. Compare the employee's hours worked to the constant
40. (Forty is the standard number of hours in the work week.)
Set indicator lOon if hours worked are greater than 40 and set
indicator 20 on if hours worked are less than 40. If hours worked
equal 40, no indicator is set. As a result of this comparison and
the setting of the indicators, the machine will follow one of three
possible paths shown as A, B, and C on the flowchart.
Path A: Hours worked less than 40.
02. Indicator lOis off; this calculation is not executed.
03. Indicator 20 is on. This calculation is executed only
when indicator 20 is not on.
04,05. Indicator 10 is off; these steps are ignored.
06. Indicator 20 is on; pay rate is multiplied by actual hours
worked to obtain gross pay.
Path B: Hours worked greater than 40.
02. Indicator lOis on; the constant 40 is subtracted from
the actual hours worked to obtain overtime hours.
03. Indicator 20 is not on; the pay rate is multiplied by 40
to obtain regular pay.
04. Indicator lOis on; overtime rate is multiplied by overtime hours to obtain overtime pay.
05. Indicator lOis on; overtime pay is added to regular pay
to obtain total pay.
06. Indicator 20 is not on; this step is ignored.
Path C: Hours worked equal 40.
02. Indicator lOis not on; this step is ignored.
03. Indicator 20 is not on; pay rate is multiplied by 40 to
obtain gross pay.
04, OS, 06. Both indicators are off; these steps are ignored.

IBM

International Business Machines Corporation

RPG
Date

_!?Lx)C

CALCULATION SPECIFICATIONS

@s~~:~~;on I :~:::iC

II. _fj 0

1-1

V Cj

Programmer _ _

I I_I I I I I I
Resulting

Indicators

Indicators
Arithmetic

-.;Ul

Line

~3S

I
And

Plus Minus

Factor 1

Operation

Factor 2

Result Field

.... 0: en·
E

Z...J

Field
Length

;

B Hrswkd >40

Hrswkd
-40=
Ovtme
hrs

Low

~ 1 >2 1< 2

Compare
High

Lookup

High

9 10 1112 t3 "

/Ial

zls6

/1<:6
/I~
Zl~

1 0

1 3

1 4

Low
15 t6 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57

Ko
H~
HR 5W Kt>
PA rR Ar£
£R T/I1
Oy I2 1< 2 1 = 2

..."'"'

c
c

1 3

1 4

1 5

Comments

Lookup
Table (Factor 2) is
High

Zero

Compare

Low Equal

9 10 1112 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74

2 N¢/
I 2 I'Ir¢ /

MvR

FI £L Oy

cpu or INlr
teE' NO FR

C
C
C
C
C

Figure 21.5 Move remainder.

452

RPG II CALCULATION

The remainder is 15 positions in length. The number of decimal positions equals the number of positions in factor 2 of the
divide· operation. If the remainder is less than 15 positions, the
leftmost zeros may be dropped. Resulting indicators may also be
used for MVR.

Crossfoot (XFOOT)

Elements of an array are added together. The sum is placed
in the field named as the result. Factor 2 contains the name of the
array. Factor 1 is not used.

Move Operations

Move operations place all or a part of factor 2 in the result
field. Factor 2 is not changed. Factor 1 is not used and must be left
blank on the specification sheet. Resulting indicators are not used.

Move (MOVE)

This operation causes characters or digits from the field named
as factor 2 to be placed in the field named as the result. Moving is
done, character by character, beginning with the rightmost position of factor 2 and proceeding from right to left until all positions
are placed in the result.
If factor 2 is longer than the result, excess positions in factor
2 are not moved. If the result field is longer than factor 2, excess
positions to the left of the result field are not changed.
An alphameric field or constant is changed to a numerical
field by placing it in a numerical result (0 in column 52). The digit
portion of each character is converted to its corresponding numerical character and then moved to the result field. Blanks are moved
as zeros. The zone portion of the rightmost character becomes the
sign of the r~sult.
Conversely, a numerical field or constant may be changed to
an alphameric field by moviI!g it to an alphameric result (blank in
column 52). Examples of the MOVE operation are shown in
Figure 21.6.

RPG II CALCULATION

453

Result Field

Factor 2
Before MOVE Operation

Numeric or +
Alphameric (4 = letter D)

Alphameric
After MOVE Operation

I 2 I 31 41 7 181 41 2 I 51

11 12 1 31

4 1 p I H 14 I SiN

Numeric Result Field

Alphameric Result Field

Result Field Larger than Factor 2
Before MOVE Operation
Alphameric

Numeric or +
Alphameric (4 = letter D)

After MOVE Operation
Numeric Result Field

IplHI41sINI
Result Field Smaller than Factor 2

Alphameric Result Field

Before MOVE Operation
Alphameric

Numeric or +
Alphameric (4 = letter D)

After MOVE Operation
Numeric Result Field

Alphameric Result Field
Before MOVE Operation
Numeric
Alphameric
After MOVE Operation
Numeric

Alphameric Result Field
Result Field and Factor 2 Same Length

Figure 21.6 Move operations.

Move Left (MOVEL)

This operation is the same as the MOVE operation, except
that factor 2 is left justified in the result field. If factor 2 is longer

454

RPG II CALCULATION

Factor 2
a. Numeric

Result Field

17\8.412151

Before MOVEL Operation

\51617.8141

/7 18.4\2/51

After MOVEL Operation

17/8/4.2151

/71 8.41 2 15/

Before MOVE L

IAIKITI41DI

1718.412151

After MOVEL

17181412/NI

b. Numeric

Numeric

Alphameric

----·-j;I~ISjNI---~eforeMOVEL-------151617!81~--c. Alphameric

_ Numeric
IPIHI 4 1S INI

After MOVEL

171 8 14 12 15 1

jPIHI 4 1SINI

Before MOVEL

IAIKIT 141DI

IpIHI4.\s\N\

After MOVEL

IpIHI41sINI

d. Alphameric

Alphameric

Factor 2and Result Field Same Length
10101010101 8 1412151

Before MOVE L Operation

+
15.6171 8 14 1

10101010101 8 \412151 '

After MOVEL Operation

10.010 10101

1910131117181412151
_
191013\11718141215\

Before MOVEL

IAIK\TI4IDI
.
Alphameric
19 101 3 11 171

I BI Rl w lelxIHI41 SINI

Before MOVEL

1BI Rlwle Ix1H\41 sl NI

After MOVEL

\5161718.4/
Numeric
121916\3.7\

I BI Rlwlelx\Hl4lsl NI

Before MOVE L

IAIKITI41DI

IBIRlwlelxIHI41SINI

After MOVEL

IB IRlwlelxl

a. Numeric

b. Numeric

Numeric

After MOVEL

---- --- - - - --------------:r--c. Alphameric

d. Alphameric

Alphameric

Factor 2 Longer Than Result Field

Numeric

After MOVEL Operation

leIP\T\5INI

Before MOVEL

le\PIT\5\N I

After MOVEL

/113\0/ 9 14 13 \2\11 0 1
_ Numeric
1317131515131211101

17 18 14\2151
171814\2151

Before MOVEL

IB\Rlwlelx\H\4IsIAI

After MOVEL

171814121NIHI41 S\AI

Before MOVEL Operation

+
11.310\9\4\3\2\1\ 01
+ Numeric
17.814\2151312111 0 1

- - - - - - - - - - - - - - - .-----------=---Alphameric

.1 Numeric

17181412\51
_
17\8\4\2151

Alphameric

------------------------

Alphameric

lelPIT 151N I

Before MOVEL

\B\RlwlelxIH\4\sIAI

\e\pIT\5IN I

After MOVEL

lei P\T\5\NIHI4IsIAI

Alphameric
Factor 2 Shorter Than Result Field

Figure 21.7 Move left.

than the result, the excess rightmost positions of factor 2 are not
moved. If the result is longer, excess positions to the right of the
result are not affected.

RPG II CALCULATION

Alphameric and numerical fields are changed in the same
manner as in the MOVE operation. Examples are shown in Figure 21.7.
Move Zone Operations

These operations affect either the zone or numerical portions
of a character. They provide detailed flexibility to operate within
the character structure, for example, to change the sign of a numerical field.
Four operations are available in RPG II:
Move
Move
Move
Move

High to High Zone (MHHZO)
High to Low Zone (MHLZO)
Low to Low Zone (MLLZO)
Low to High Zone (MLHZO)

In these operations, the word high refers to an alphameric
field; the word low to a numerical field.
Figure 21.8 is a tabulation of the functions of these four
opera tions.
MHHZO

Factor 2 Alpha

MHLZO

Factor 2 Alpha

Result

Alpha

Numeric

Factor 2 Alpha
Result

MUZO

Alpha

Factor 2 Alpha
Result

Alpha

Factor 2 Alpha
Result

Numeric

Factor 2 Numeric
Result

Alpha

Factor 2 Numeric
Result

MLHZO

Numeric

Factor 2 Alpha
Result

Alpha

Factor 2 Numeric
Result

Alpha

Bits 0 - 3 of leftmost byte of
factor 2 are moved to bits 0 - 3
of leftmost byte of result.
Bits 0 - 3 of leftmost byte of
factor 2 are moved to bits 4 - 7
of rightmost byte of result.
Bits 0 - 3 of leftmost byte of
factor 2 are moved to bits 0 - 3
of rightmost byte of result.
Bits 0 - 3 of rightmost byte of
factor 2 are moved to bits 0 - 3
of rightmost byte of resu It.
Bits 0 - 3 of rightmost byte of
factor 2 are moved to bits 4 - 7
of rightmost byte of result.
Bits 4 - 7 of rightmost byte of
factor 2 are moved to bits 0 - 3
of rightmost byte of result.
Bits 4 - 7 of rightmost byte of
factor 2 are moved to bits 4 - 7
of rightmost byte of result.
Bits 0 - 3 of rightmost byte of
factor 2 are moved to bits 0 - 3
of leftmost byte of result.
Bits 4 - 7 of rightmost byte of
factor 2 are moved to bits 0 - 3
of leftmost byte of result.

Figure 21.8 Move zone operations.

455

456

RPG II CALCULATION

Compare and Testing By performing compare and testing operations, the computer can
Operations be programmed to follow one of several possible paths through a

procedure. The selection of the proper path is made by comparing
one item of stored data with another. Single characters, selected
fields, or entire records 'can be compared. One of three possible
conditions invariably results: a given field is always higher, lower,
or equal to another.
The condition resulting from the comparison is registered by
designated indicators. When the indicator is on, the machine can
be instructed to branch to a high, low, or equal program routine.
Such branch instructions are termed conditional branches and have
the effect of telling the machine to GO TO another instruction
provided the designated indicator is on. If the indicator is off,
continue with the next instruction in the sequence.
As previously explained, the determination of whether one
field or character is higher, lower, or equal to another is made according to the established collating sequence of the system. In
RPG II, that sequence can be varied by using a special coding form.
Compare (COMP)

Factor I is compared with factor 2. As a result of the comparison, indicators are turned on as follows:
High:
Low:
Equal:

Factor 1 is greater than factor 2
Factor 1 is less than factor 2
Factor 1 equals factor 2

The record fields named as factors 1 and 2 are automatically
aligned before they are compared. If the fields are alphameric,
they are left-justified. If one field is shorter, the unused right positions are filled with blanks. Fields of unequal length to be compared may contain as many as 40 characters. Fields of eq uallength
may contain a maximum of 256 characters.
Numerical fields to be compared are aligned by the decimal
point. Any missing digits are filled in with zeros. Maximum field
length is 15 digits.
Figure 21.9 shows a number of compare operations with
n.amed fields and constants. The item numbers below correspond
with line numbers in the illustration.

GX21·9093·'

IBM

International Busmess Machines Corporation
Pnnted In U.S.A

RPG

CALCULATION SPECIFICATIONS
1

Oate _ _ __
Page

Program
Identification

77 78 79 80

IIIIIII

Resulting

Indicators

Arithmetic

T T

~
S·

Line

And

Plus Minus

And

Factor 1

Operation

Factor 2

Result Field

E
0

..J

Field
Length

~ High Low Equal
~ 1>21<21=2

Table (Factor 21 is
High

fll('$ Iu,KD

1 3

1 4

-....J

1 5

Low Equal

9 10 1112 13 14 1516 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74

o
o

Comments

Lookup

Zero

Compare

1(8

iN£ rp /I

ilL. 70 CO MP

sr oc; K8 AL 7 I

CO MP

21<21=2
Lookup
Table (Factor 21 is

High Low Equal
9 10 11 12 13 14 15 16 17 t8 t9 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74

PI £L PA
I

IF I

£L

R.1 Nil

SUB
PI EL DB
GO TO Rr NJ
MU LT 1+
GO TO Rr NJ

FI GL DB

S A '-IE

II~

111('

(
15

R.T Nt

TflG
(

,
I)

It; 0 TO R.T Nt

N/~

T£ S T 1.
ENlJ
Cio
Mil I. Z blf I £ L P

c
c

Operation

Field
Length

END

IE l

If /

EL DO

ll¢

21¢

1191G

c
c
c

Figure 21.10 GO TO and TAG coding.

' - - ......&........

462

RPG II CALCULATION

Routine

BX4

Indicator 20
Off

Off

Indicator
200n
MHLZO
Field C

Routine

END

Figure 21.11 GO TO and TAG flowchart.

Line 05. If indicator 20 is off, go back to DOOVER. If indicator 20 is on, the loop has been executed 5 times.
Line 06. Subtract 5 from count to set the field back to zero
in preparation for the next execution of the loop.
The routine has the effect of multiplying Field A by 5.

Subroutine
Operations

Subroutines are written in the same manner as other calculation
specifications in the RPG source program. The subroutines are
then compiled with the source program and become part of the

GX21-9Q93-1

IBM

I nternational Business Machines Corporation
Printed in U.S.A.

RPG

CALCULATION SPECIFICATIONS
1

Date _ _ _ _ __

page[D

75 76 77 78 79 80

~;~~:~::ation I

IIIIII

Programmer _ _ _ _ _ _ _ _ _ _ _ _ _ _ __

Resulting

Indicators

Arithmetic

And

Line

Plus Minus Zero

And

Factor 1

Factor 2

Operation

Result Field

:x:

Field
Length

:;]"

Compare

~ High Low

Equal

~ 1>2 1< 2

1= 2

Lookup

9 10 1112 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74

Table (Factor 2) is
High

1 5

Comments

PO Ov EI<.
{."I 1£ L f)A

,114
ADP

1
Co

AD/)
IJ IV

-r

N 21¢
CO UNIT

CO fliP
CiO TO

SuB

PI £L OA
UN,
5
Do oV IE R
5

(.", EL DA
co UN1

Low Equal

bt/J
I

Z~
I

fJN1

0'1
~

C
C

C
C
C
~

Figure 21.12 Branching operations.

464

RPG II CALCULATION

resulting object program. One subroutine can be compiled with
any number of different source programs by including the cards
punched for the subroutine as part of the various source decks.
Each subroutine must be given an identifying name and must
begin with the operation BEGSR and end with the operation
ENDSR. These two operation codes mark the limits of the subroutine and separate it from the other elements of the calculation specifications.
All subroutines must be written in specification lines following all operations conditioned by control level indicators specified
in columns 7 and 8. Subroutine lines are always identified by the
letters SR in columns 7 and 8.
Execute Subroutine (EXSR)

This operation causes the designated subroutine to be executed. EXSR may be written anywhere in the source program.
After the subroutine has been executed, the operation in the line
following the EXSR operation is executed.
The EXSR may be conditioned by any indicators. Then, the
subroutine can only be executed if all conditions are satisfied.
Factor 2 contains the name of the subroutine to be executed.
This same name must appear on a BEGSR line.
Begin Subroutine (BEGSR)

This operation serves as the beginning point of the subroutine.
Factor 1 must contain the name.
End Subroutine (ENDSR)

This operation must be the last statement of the subroutine.
It defines the end of the routine. The ENDSR automatically causes

a branch back to the next operation line after the EXSR line.
Example: Use of Subroutines

Figure 21.13 is a flowchart showing logic of subroutine use in
calculating net pay in a payroll application. Three aspects of the
calculation procedure are shown.

RPG II CALCULATION

Calculate
gross

EXSR
FICA
FICA
BEGSR
ENDSR
UPDATE
PAY REC

EXSR
WHTAX
WHTAX
BEGSR
ENDSR
UPDATE
PAY REC

YES

SVBOND
BEGSR
ENDSR

EXSR
SVBOND

NO
NO

FIXNET
BEGSR
ENDSR

WRITE
PAY
RECS

Figure 21.13 Subroutine logic flowchart.

1. F.I. C.A. After gross pay amount is calculated, a branch is
effected to the F.I.C.A. subroutine. The branch is unconditional.

465

466

RPG II CALCULATION

.. The subroutine can include all instructions to calculate the
social security tax, including the year-to-date accumulation and
t~e test for limit of payment according to current federal tax
regulations.
When F.I.C.A. is calculated, the machine is directed to the
next instruction in the main program. The amounts obtained by
the subroutine are placed in the employee pay record. If no tax is
calculated, the current pay period tax field is filled with zeros.
The year-to-date tax field is unaffected.

2. Withholding Tax. An EXSR operation causes a branch to
the WHTAX subroutine where federal income tax for the employee
is calculated. The branch is unconditional. When the subroutine
is completed, the main program is entered at the next instruction.
The pay record is again updated to include the result of the tax
calculation.
3. Savings Bond. It is assumed that in this application, employees may have authorized amounts deducted from their pay
toward the purchase of a U.S. savings bond. The amount of each
authorized deduction is included in the master records for the
employee.
If the employee has authorized a deduction, an indicator is
on, and the branch is effected to the SVBOND subroutine. The
branch is conditional upon the designated indicator. The subroutine should contain all instructions needed to do the accounting
for bond deductions.
The following formula is calculated at the end of the
subroutine:
GROSS PAY -c-(FICA + WHTAX + SAVINGS BOND) = NET PAY
The net pay field is then tested to determine if the amount is less
than or equal to zero. Net pay could be a minus or zero quantity
if an employee had not worked long enough in the current period
to earn the deduction amount. In this case, no paycheck can be
written and a corresponding error would appear on the pay register.
If net pay is less than or equal to zero, a FIXNET routine is
entered to add the bond deduction back to the employee's pay
and to reverse the bond accounting entries. At the end of the routine, the main program is reentered to prepare the output documents for printing and punching ..

RPG II CALCULATION

467

The use of subroutines in the above procedure provides a
distinct advantage of making the program modular. That is, certain parts of the program can be changed or eliminated without
affecting the main program. If federal tax laws are modified, the
tax subroutines can be readily altered without affecting other
parts of the application. Also, should bond deductions be discontinued, the routine can be eliminated by using one of the external
indicators available in the RPG II system. Branching can be made
conditional on the specified indicator.
Note also that the fix routine corrects a potential error automatically. The routine could also include provision for printing
out a list of those employees who receive no pay for a period or
who are over-deducted.

The RPB II language includes a number of additional operation
codes that will not be completely explained in this text. For the
most part, these are supplementary features not always required
for programming an application.
Look-up (LOKUP).

The operation executes a search for an item in a table. Factor 1 is the search word; factor 2 is the table name.
Exception (EXCPT).

The operation allows records to be written out at the time
calculations are being performed. A number of multiple records
can be produced from a single output record, for example, a given
quantity of duplicate shipping labels for one customer.
Force (FORCE).

A record can be read from any file thereby overriding the
standard sequence of input operations.
Debug (DEBUG).

Placed at any or several points in the program, this operation
causes a printout of all indicators that are on at that point in the

Other Operations

468

RPG II CALCULATION

program. The contents of a named field can also be printed. The
operation is an aid in tracing programming errors during the testing
of the object program from the RPG assembly.

Questions
and
Exercises

1. List the four parameters of all calculations in RPG II in the order in which
they are placed on one line of the specification sheet.
2. How are RPG II indicators similar to pilot selectors and coselectors in an
accounting machine?
3. What is a constant? How is it used?
4. What is an operation code? List some RPG II examples.
5. What are result indicators?
6. Given the data supplied, what result is obtained from the following
operations.

Field A

Oper.

Field B

16.24+

ADD
ZADD
ADD

0.56923+
123
1532.00
7431.0+
14.632+
246+
0.55
241.234

12.346283+
12.782
24624682468+
463.82

SUB
SUB
MULT
MULT

DIV
DIV

Field
Length

Dec.

5
3
6
5
5
6
5
3
4

3
0
3

Pas.

Half
Adj.

3
1
0
2

H
H

7. Prepare the calculation specification sheet to continue the case study
problem following chapter 20.

RPG II OUTPUT-FORMAT

SPECIFICATIONS AND
EXAMPLES

Output-format specifications describe all output files. Entries
are of two types:
1. File identification entries describe the files and the records
to be produced within the files.
2. Data description entries locate the data within the records.

Like other RPG fonTIs, the Output-Format Specification sheet is
divided into 80 columns as shown in Figure 22.1. This section
describes the entries which can be made on the form.
Columns 1 through 7 and 75 through 80. Entries for these
columns are the same as described in chapter 21.
Columns 7 through 14. This entry identifies the output file
by name. Construction of the name must adhere to the same rules
as those established for the file description form. The name must
be written on the first line that identifies the file.
Column 15. This entry identifies the type of record to be
written: H for header, D for detail, and T for total. The record
may be either a printed line or a 96-column punched card. An E
entry indicates a record to be printed or punched during calculation time.
469

Output-Format
Specifications

IBr.,

Form X21.9090-1

International Business Machines Corporation

Printed in U.S.A.

RPG

OUTPUT - FORMAT SPECIFICATIONS

1 2

Date _ _ _ _ __
Page
Program _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __

75 76 77 78 79 80

[I]

~;:::f:ation I

IIIIII

Programmer _ _ _ _ _ _ _ _ _ _ _ _ _ _ __

.j:o.

Filename

Line

8.
>

I-

~
3 4 5

o
o
o
o

617

~
t:: t

1.1

~ ~

And

-t!

~ ~ co «

J9
~

~
OJ

O"'P'" "dI~t<'"

s."

Field Name

:;, iii

And

8 ~
:5 j

~ ~

10.

~
«

End
Positon

Edit Codes
Commas

Zero Balances
to Print

Yes
Yes
No
No

Yes
No
Yes
No

Record

Outpu~

J>.

n
~

o
o

1 0

1 1

1 2

1 3

1 4

1 5

I CR
A

L
M

x = Remove
Plus Sign
Y = Date
Field Edit
Z = Zero
Suppress

Sterling
Sign
Position

Constant or Edit Word

8 9 10 11 12 13 14115 16 17 18 1920 21 22 23 24 2526 27 2829 30 31 323334 35 36 37 38 39 4041 4243 4414546 47 4S 49 5051 52 5354 55 56 57 58596061 626364 65666768 69 70171 72 73 74

No Sign

0
0
0
0
0

Figure 22.1 Output-format specifications.

RPG II OUTPUT FORMAT

Record type must be entered in the order in which the records are to be produced: heading, detail, total, and EXCPT. E
may not be specified for a combined file.
Column 16. An entry in this column is used:
1. to select the stacker for specified cards;

2. to indicate that the page overflow routine can be used at
this point.
Entry

Explanation

Blank

Primary cards automatically fall in stacker 1, Secondary cards in stacker 4.
Indicates selected stacker.
Fetch overflow.

1-4
F

The F entry in the column causes the overflow routines to be
performed in advance of their normal execution. In some instances, it may be desirable to print total and detai1lines at the top
of a new page rather than after the heading lines conditioned by
overflow have been printed.
Columns 17 and 18. Entries in these columns indicate the
number of line spaces to be taken before or after printing. No
spacing, single-spacing, double-spacing, or triple-spacing is indicated by 0, 1, 2, or 3 respectively. An entry in column 17 indicates spacing before the line is printed; entry in column 18 indicates spacing after printing.
Columns 19 through 22. Skipping is used when more than
three lines of space are needed between printed lines. Spacing or
skipping should normally occur after printing to avoid delay waiting for the form to be positioned. Entries in these columns indicate the line number for the form to which the skip is to be made.
On the System/3, overflow skipping to the next page takes
place automatically after line 60 is printed. Therefore, when using
standard 66-line paper, six-line margins are placed at the top and
bottom of each page. Procedures that need different form specifications must be described with Line Counter Specifications, a
special form for this purpose.
Entry

Explanation

0-99

Lines 0-99
Lines 100-109
Lines 11 0-112

AO-A9

BO-B2

471

472

RPG II OUTPUT FORMAT

Entries in columns 19 and 20 indicate skipping before a line
is printed; entries in columns 21 and 22 indicate skipping after the
line is printed.
Columns 23 through 31. Output indicators are entered in
these columns to specify the conditions under which a record is to
be punched or printed. As many as three indicators can be specified. The indicators are written on the same line as the name of
the record to be produced as output.
Entry

Explanation

01-99

Any resulting indicators, field. indicators, or record
identifying indicators previously identified. Thus a
record may be punched or printed when the result of
a calculation is plus, minus, or zero; when the result

of a comparison is high, low, or equal; and so on.
LI-L9
HI-H9

VI-V8

OA-OG,OV

LR
IP

Any control level indicators previously specified.
Halt indicators. When a condition occurs that stops
the program, a record can be produced as a result of
the halt; for example, a message to the operator.
An external indicator set prior to program execution.
For example, different report formats can be produced in separate runs from the same program and
input data. External indicators condition printing at
the option of the operator or programmer.
Any overflow indicator previously assigned to this
file. Totals may be printed at the end of a page, for
example, even though no control breaks have occurred in the input files.
Matching record indicator. V sed to print or punch a
record when records match as previously specified.
Last record indicator. Can be used to print or punch
the total at the end of a job.
First page indicator. Can be used to print headings or
other identifying information on the first page of
a report.

As many indicators as may be needed can be specified by
using more than one line of the form. AND, written in columns
14 through 16, establishes the relationship between the two lines
of indicators. OR relationships can also be specified by writing OR
in columns 14 and 15 of a subsequent line and entering the alternate indicators on that line. Figure 22.2 is an example of using
indicators to condition output records. Item numbers below refer
to the corresponding lines of the specification form.

RPG II OUTPUT FORMAT

OU~

IBM
RPG

IZ/;S/xX
r I
SflLe.s
Program

Date

Programmer

rRE'/J

R&-NJer

Punching
Instruction

Ft:J~j)

)
IL

~ Space

Output Indicators

Skip

!
.t:

Filename

Line

e.¥ "
!.
to
>.J<

~
3

~!!;

~~ .2 ~

I-

00 or pur

o
o

or 01 AL ~E C7

1 2

L-J.-J.--L--

(

ANO

o
o

H
oR.
~

10 11 12 13 I. 15 16 17 18 1920 21 22 232. 2526 27 28 29 30 3132\

o
o
o

1 0

or cl,4

RIJ

Tl.

Z
J

IP
NI I¢I!/ ZI¢
ZS-

3[¢
I

zl¢
3J¢

NMR

L f

IAi R

I. Z

1\
~

I-

....................

.......j...-I--I--

.......... U

Figure 22.2 Indicators to condition output records.

01. The printer carriage skips to line 6 before printing. The
heading is printed only when an overflow or first page condition occurs.
03. A detail1ine is printed when indicators 10 and 20 are off
and indicators 25 and 30 are on. The indicators could have been
previously specified as field indicators from input specifications or
as resulting indicators from calculation specifications.
05. A detail line is printed if indicator 20 or 30 is on.
07. Two lines are spaced before printing a total if level 2 indicator is on and the matching record and halt 2 indicators are off.
Thus, a total is taken after a control break, if no error has occurred.
(It is assumed an error would turn the halt indicator on.)
08. A total is printed if control level indicator L2 is on and
the matching record indicator is also on.

473

474

RPG II OUTPUT FORMAT

09. A summary card is punched· and placed in stacker 2 if
control level indicator L2 is on.
Co·lumns 32 through 37 are used to enter the name of every
field that is to be written out. Names must have been previously
assigned on the preceding forms.
Column 38 provides an entry to edit numerical fields produced on printed reports. That is, the field may be punctuated by
automatically adding commas and decimal points in the proper
positions. Insignificant zeros to the left of the field may be suppressed as required. The various codes are printed on the form for
ready reference. Figure 22.3 is a table which further explains the
available editing codes.
For example, if code I is specified in column 38, the field
will be printed with commas and a decimal point but with no sign,
Commas

Decimal
Point

No
Sign

.00

Blanks

.00

Blanks

Code

Signs for Negative Balance
Minus (.)
CR

Print Out on
Zero Balance

Zero
Suppress

.00

Blanks

.00

Blanks

.00

Blanks

.00

Blanks

X*
y **

Note:

* Code X moves a positive sign from the units position.
** Code Y puts slashes in a numeric field according to the following pattern:

nnln
nnlnn
nnlnnln
nnlnnlnn
Code Y zero-suppresses only the left-most portion.

Figure 22.3 Edit codes.

RPG II OUTPUT FORMAT

475

even if the field is minus. In case the field is all zeros, two zeros
will be printed after the decimal point as shown. The code A will
produce the same editing as code 1 with the exception that the
credit symbol (CR) will be printed after minus fields. If code J is
used the minus symbol (-) rather than the credit symbol will be
printed after minus quantities.
Columns 45 through 70 may be used for more elaborate editing of fields by inserting the desired symbol in the proper position
of the field. The use of edit words is shown in Figure 22.4. Constants may also be entered in these columns. These are usually
words used for report headings, column headings, and card identification. Structure of the constant must conform to the same
rules as previously stated, and the word must be enclosed in apostrophes as shown.
International Business Machines Corporation

OUTPUT - FORMAT SPECIFICATIONS

I 2
pageDJ

:ators

Field Name

II §
~ ~
8 ~

t-r-r0

~ ~

Edit Codes

....

Iii:

>
~

End

iii

Positon

"
~

in
Output
Record

~rf

Commas

Zero Balances
to Print

Ves
Ves
No
No

Ves
No
Ves
No

No Sign

I
2
3

A
B

J
K

C
D

l
M

Constant or Edit Word

29 30 31 32 33 34 35 36 37 38139 40 41 42 43

!+I~

NIT

YR. 1"0 Dr

Sr/J

41¢

055

61~

ITo TAL..

6191

PR I cl£

~~ /. II NC
WE 114 IIr

U P:l( £1<

21
22
23
24

is !III"

I.
'I

19
20

I!
i

! :

15
16
17
18

~If

XI)(;·.'1.'"/..)(

.(tl

ws J

'X~I·!~Jt')(.:'ll:X

f-l-t TlliMx:x

_ _ _ _ r-8++++-HH -+-+-+
9
10
11

X ii'
I,'

Ix. A' v

I ' , !

- - - t - + - l - + I+-f"-f-t+f~!J
----+-6::..1-+-t-+
I "

'I
1)\.,;( ,I~

)(iJ\XX • XXlx.x

'sa<:

V<
~ CDS

, :> iJ

d.-

. I

r: rl'

FJ I'

ifl rv r~

i
I
I

Figure 22.6 Sales report. Printer spacing chart.

Figure 22.7 is the Control Card Specification form showing
that the program is to be assembled on an 8K Systemj3. No other
entries are required.
Figure 22.8 describes the two files involved in the processing:
the input card file ITEMS and the output file REPORT.
Figure 22.9 is the input form with each field of the input
record named and located. The record identifying indicator is 01.
Control level indicator Ll is assigned to the item field. The indicator will be turned Oil each time a control break occurs. Accumulation of the quantities, cost, and selling price occurs for every
record. Totals are taken for each control break as conditioned by
indicator Ll (Figure 22.10).
Figure 22.11 is the Output-Format Specification form. The
detail line is' printed as specified on lines 01 through 04. Note that
edit code 3 is used to print the amount fields with commas and
decimal points. The end position of each field corresponds with
the location of the field shown on the printer spacing chart.

477

"'-I
00

I nternational Business Machines Corporation

IB,.,
Date

RPG

Izf~/x)(

Form X21·9092·1
Printed in U.S.A.

CONTROL CARD AND FILE DESCRIPTION SPECIFICATIONS
1 2

Page

Program~t9' as REPtJRT - - - - Programmer f1L.Tc~
£"t;o

ltIZ]

75 76 77 78 79 80

~~~;i~:ation IslA IL IciS I

Control Card Specifications
~

Sterling

~I ~re

Core

I~
f--

o Sizeto
Execute

1'' "'0 0, 8Compile

ojlL11lJ

e~ g

~
Number

.§

Of Print 2!
Positions

Refer to the specific System Reference Library manual for actual entries.

al .~c:

:.J

I I I lilllUJlillll I I I I I II I I I I I I I I I I I I I I TTITllTIJ

Figure 22.7 Sales report. Control card specifications.

File Description Specifications
Mode of Processing

File Type

End of File

lille

!
4

File format

...J
W

Device

Number of Extents

Tape
Rewind

Name of
Label Exit

Symbolic
Device

..---File

lil.lm

Type 01 rile
Organization
or Additional Area

Number of Tracks
for Cylinder Overflow

Extent Exit
for DAM

Record Address Type

Sequence

Filename

File Addition/Unordered

Length of Key Field or
of Record Address Field

File Designation

...:.. Overflow Indicator

Block
Length

Record
Length

a:
...J

""I~
« ::::
~

Condition
Ul·U8

Core Index

.::

=>'

r--

10 '1 12 13 14

0121

IFI/[T[E[MIFII [L[EIII?

0131

'Fklc-IPloIR[T[ 1 10

916
'117

IMIF[C[oll
PI~I/INITIC-IR

014

8;t1
o

++-+-+++++ 1 1 I I I I 1++++ 1 1 I I 1 I I I 1 I I 1 I I I 1 I I 1 1 1 I 1 1-+ 1 1 I 1 1 1 1 1 I 1+1 1 1 1 I I I 1 I I 1 I 1 I

I I I I I I I I I I
017

++++-+ I

I I I I I I I I 1 I I I I I I 1 I 1 1 1 I I 1 1 1 1 1-+-1 1 1 1 1 1 I I 1 1 I 1 1 I 1 1--++-+ I I I 1 1 I 1 1 1 I

Figure 22.8 Sales report. File description specifications.

~
~

1.0

IB~

International Business Machines Corporation

RPG

IZ/t.b

Date

Program

GX21-9094-1 U{M050
Printed in U_S.A.

INPUT SPECIFICATIONS
1

'S,QLE s R€?O'(T

Programmer

8LrER

75 76 77 78 79 80

page~

Program

Identification

Record Identification Codes

Field
Indicators

Field Location

a:;
Filename

Line

,..~

Position

~
3

o
o

~
~

~ :~
"'.c

:;';u

, I I. I,., I'·""
~.''''"
Sign

:rl

-.;

Plus Minus or

Blank

10 11 121314115161171181192012122 23 24125126127128 29 30 31132133134135363738139140141142143144454647148 495051152153545556575815960161621636416566167 68169 7017172 73 74

bl IJ 1T1£IM
7
11¢1~19IVIA 1/1/
/III I I/ISI21clolsir

21t I I Izlb,Zlsl£I,PI,RIC

1~IAILklsl

£((0

1 0
11
1 2
1 3
1 4
1 5

Figure 22.9 Sales report. Input specifications.

Lil

GX21·9093·t

IBM

International Business Machines Corporation
PrInted m U.S.A

RPG

Date~_'~
Program

J4'

iF S

CALCULATION SPECIFICATIONS
1

pa ge [ t 0

Reiber

75 76 77 78 79 80
Program
Identification

Is I~ IL lEis I I

Programmer~£~ _ _ _ _ _ _ _ __
Resulting

Indicators

Indicators
Arithmetic

Line

I
And

Plus Minus Zero

Factor 1

Factor 2

Operation

Result Field

Field
Length

]

"- z...J

'0
z

Compare

~ High Low Equal
~ 1>2 1< 2 1=2

Lookup

~\F~~_
High

00
~

C '-

C /..

I 3

I 4

I 5

Low Equal

9 to 1112 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 5t 5253 54 55 56 57 58 59 60 6t 62 63 64 65 66 67 68 69 70 71 72 73 74

¢I tt'IJ A.N
Co Sf
I¢ I 5£ PRe
5(/ 8e oS
SIJ 8f I?.'-

I¢ I

Comments

ADl>

ADD
ADD

1\ p
A

stJ 89

Sf.}

rY
Be 0,5
SU 8P ~L
lYe

1'1 I~

8 Z.

".s

7Z.

p lee.

C
C
C

C
C

Figure 22.10 Sales report. Calculation specifications.

IB~

International Business Machines Corporation

GX21-9090-1 U/M 050·

RPG

Date~

Printed in U.S.A.

OUTPUT - FORMAT SPECIFICATIONS
1 2

.s~L ES Rlif'O,e T

Program

75 76 77 78 79' 80

~~~~~f:ationl

PageDJ

1.1 IW

/lLr€R E"CjO

Programmer

.j:;o.

Space

Output Indicators

Skip

~
.I:

Filename

Line

t~
o u
___..!i!

I-

:!l.

]

W £

Ii;

>"'"
I- i;l

e
~~

::(

Field Name

~ ~

;3 ::(

i;
W

cl5
3

oR. E/' ORT
0

o
o

5£ p~

su Be oS

o
o

SlJ

1 2

1 3

1 4

1 5

in
Output
Record

CD
"0

Yes
Yes
No
No

Yes
No
Yes
No

e
0

X = Remove
Plus Sign
Y = Date
Field Edit
Z = Zero
Suppress

Sterling
Sign
Position

<.J

Constant or Edit Word

Ilfll

17 £~

I~·LJ AN

/ 6
3¢
41¢

LI
I 8

3rP

4¢

8P Rc.

FI NC o.s

FI liP ReI

4i¢

r 21

'1-9

S VB 70

'If IN f4

TA L

0
0
0
0

No Sign

A
B

I¢ /

1
2

Positon

Zero Balances
to Print

I~iii

Edit Codes
Commas

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 323334353637 38 39 4041 42 43
4445464748495051525354 55 56575859606162'6364'65666768 6970 71 72 73 74

1 0

IJ~

"'i"".,.

0
0

Figure 22.11 Sales report. Output-format specifications.

RPG II OUTPUT FORMAT

483

Subtotals are printed after each control break with the constants to identify the amounts as either subtotals or a final total.
The fields are blanked after printing so that they can correctly
accumulate quantities for the next control group.
The final totals are printed by conditioning with the last
record indicator. Setting these fields back to zero is optional.

A planned stock status report shown in Figure 22.12 is to be prepared from a file of inventory transaction cards. The report will
include item number, item description, previous balance of the
quantity in stock, withdrawals and receipts for the current period,
and new balance in stock. As the report is printed, summary cards
are punched with item number, description, and new stock balance.
Input records are the file of previous balance cards sorted to
item number with current withdrawal and receipt cards. Figure
22.13 is a layout of the various types of cards. Figure 22.14 is the
system flowchart of the overall procedure. The Control Card and
File Description form is shown in Figure 22.15. All of these entries should now be familiar.
Each file is named and associated with an input or output device. Note that the old balance file contains three types of cards:
previous balance, receipts, and withdrawals. These are described
in the input specifications shown in Figure 22.16. Only one previous balance card is allowed in the file; the record is assigned
indicator 01. More than one receipt or withdrawal is allowed because there may have been any number of transactions for an item
number during the current period. Receipts are assigned indicator
02; withdrawals, indicator 03. Each type of card is identified by a
unique punch in column 10: the digits 5, 6, and 8.
All three types of cards contain fields in the same card columns and are therefore associated with an OR in columns 14 and
15 of the form. The fields are located by their card columns, corresponding to the previous card layout form. Control level 01 is
assigned to item because a total is to be printed and punched at the
end of each group of cards with the same item number. The quantity field is assigned the halt indicator HI. If the input file contains a minus previous balance, an error is indicated and the program is stopped.

Example 2:
Complete R PG II
Program

PROG. ID.,

'AN, AT 10 CHARACT.

PAGE
DATE

, PER INCH, 6 LINES PER VERTICAL INCH)

~'TAr().s
R.'tDt:. Ie

-.Sroc:.K

UMENTALlST:

[

:~ Fold back ut dotted line.

7IiJU./)..~

[

I 1 111 11 llJl ~~2 2)21?[Xrz 22b 333131313333441)41441441445555 JJJJJ516J6 616!6i6 16 66777 171771717i71de elelelelel81~lelei9 991919191919
1121*1*/71890 ;1234 51617 ~1~lo I ~1~14151617 8910 1 2jl~1516 7e 9 on21il4 ;;16 ;Ie 90 1 23 ~1~IWI~I~lo 1/2 34i5.1617 8 90 1 21;14 ;16l718!910 l[2l314151617j8~9Io I 2131415,617
I
•
I!
I I
r I ~ ill!
:II I !
' ! i ' I
I
1 ';;,
I [
,: 1 I !J-b::; i_'J
I ! ,I 1 ;
i
,~
~1Jj ~XI)j !t~x')lJJCI~lIltl) ltiJllxiJl t~~ ; i • iX!Jlixlx.iJC[)c I l~iJ(i,Qc!Il:J
1 ~~:)cX)(IJI ! I bc.xixix~:H~ Il~~J
I i 1.1 !
[[
I!
[:
I!~: !
!! J! I [;
I~: I I ! !, I 1. . "
I iii
! I I ' ! '11!
I

~UllIIIIIII
~1'r+ttlJ+lW;>~41-H-+-+-+-++
Hit I I i,I ,
It I [ . I. '"

·~~--."t+
+~r:=j~~
~ ~~+;;-rr~~t-+~-r-+~~~~~~~r+~~++~~~~r+~~r+~-r~~~rr+

-.!
~

-+I

• !

'T--'TT~-"f:;-;m[TTT:nm~i~i~'

ITTTTTT'Ti I : I:' I [ ! \ :;n rm iTI : : : Ii! I ! I\!

I:ffii
'~!!I'\I'T

~ I--+-I--l--l--I I i I I I i I I i , +t---;--+--i--i--..-+-'---;+-H-.+-+-+-1-+-++-+-+-+-+---+-+-+-i--t--+-+-+-if-+-t-+--+-+--+~
I'
':"
I
~
I t , , !-+-~I+-+.+-~.+-T+-+-i-+-~.-+-+-+-i--t-+-+-+-~~-+-+-+-+-+-+-+-+-'--+-+

~-l--l-++--t--.',
t++--+
I i
~I
'~L.:TH i
,! :-U -\"+-++-'-+-"'-+--+-4-+'f--!-~-+-f-+-++-H~++-+-+-+-l-+++-H-+-+-+-i-+-

!§.

!2
!!

: I!'

ill

.!!

~
, -t+'

[

~
~

~
~
~
~

!!
~
~

!1
~
36

•
I

•

~~
H
11
Ii
' !

I '+++++----+--+-++++++-+--t--l--I--i----J-++-!--l-+-I

I I 11 !: i I I i I I i 111TTiTTT

~
~.l

" ,lTTi

.~>---t-I+-+-+--~-+-+-+--H'-I-.t-+-+-+-+--+-+++-~I--++-+-+-

iii

: i

. i

'I!

~IIIIII!

Ii i

~
I

[ I

I I I II I I I I i

111111111111111111

,
I 111111 ITTW rt±U " ! I ! I ' 11111 ! II
II 11II i I: 1111 : i I! : I i III Hfffm 111111111111/1
I-I
i

i I: I!

'Ii

i Tl TITl TTTrTTTnTTI

ITIllTrm

Figure 22.12 Stock status report. Printer spacing chart.

[ : 1111

IIi III

IB~
96 COLUMN CARD MULTIPLE LAYOUT FORM

5 TOC K
Print
Print Lin
Tier 3

Punch

DESCRIP liON

qU/~N

~;;::'71 ' 1'1'1' I" I° I ' I° I' ""'" I" 1''1'' I" 1'°," l'OI'"I'VI" Illl'"'I"I,ollOll'ltCI"I)UIJ'IU'JJ IJ'lIJ~IJbIJI 'JCIJ~lqVlqllqllqJ Iqqlq~lqblql Iqc Iq~IOUI~II~ll~J l~ql~~I~bl~ll~el~"ltiU Ib1lbllbJ IMlb~ Ibbl61lbelb"1 IV III IIII 101 "I I~ I1"1 (( IIe I'"IOVI
Card

Card Name

I=: (

197198/991100110111021103/104110511061107110811091110111111121

E I PIS

I 12 1314151617\8\9 1011\12\13\14\15\16 17118119 120121 122123124125126127128129130j3lj32 33\34\35\36\37 8\39\40\41\42\43\44\45\46\47\48\49\50\51\52\53\54\55\56\57\58\59\60\61 \62\63\64l65\66\67\68\69\70\71 \72\73\74\75\ 76\77\78\79\80
Print
Print Lin 1
Tier 1

:
:

Print Line 2
Tier 2

Print Lin
Tier 3

((loRD

(OD£..

Punch

---?-I

Program
Control
Card

I 1213141516171819 10 11\12113\14\15\16 17118/19120121122123124125126121128129130131132 33134135136137 38139140141142143144145146147148149150 151 152153154155156157/58159160 161162/63164 65166/671681691701711721131 74115/761171781191801

1111/1 I

11111

111111111111

I III

II I I I 1-' II

I I I I

I I I I I I

I I I I I

111111111111111

Print Line 1
Tier 1

Print Line 2
Tier 2

Print Lin
Tier 3

CAR.D
Punch

COD£

8-+

6~;;:o7 1'1'1'1'1" I ° 1'1 ° l'f"I" I" I "1" I'" 1'° I" 1'° 1"1'"1"1"1"1"1'"1" I" 1£°1£'1'"1" 1'£I""I"I"I'OI"j'OIOOI'"I" 1"1"1'"1"1'°1" 1'°1"1 00 1" 1'£1"1"1'"1'°1" 1"°1"1°" 1°' 1°'10>1°'1 00 1""1°' 1""1°'1'"1" 1"1"1 "1'"\'"1" I'·I'·I~I
Card

Figure 22.13 Stock status. Card layout.

486

RPG II OUTPUT FORMAT

(Blanks)

Old bal

Recpts Wthdrl =

Stock

Figure 22.14 Stock status. System flowchart.

Figure 22.17 shows the calculation specifications. The following item numbers refer to the line numbers on the form.
01. Quantity from the previous balance card (indicator 01)
is placed in the old balance field in computer storage. The field
has been cleared after printing totals from the previous control
group.
02. Quantity from the previous balance card is also placed in
the new stock balance field.
03 and 04. Quantity from the receipt card (indicator 02) is
placed in the receipts field in storage and is added to the stock
field.
05 and 06. Quantity from the withdrawal card (indicator 03)
is added to the withdrawal field and subtracted from the stock
field. Indicator 09 is turned on if the field becomes minus.

IB~

International Business Machines Corporation

Date

RPG

//77&>1

Form X21·9092·'
Printed in U.S.A.

CONTROL CARD AND FILE DESCRIPTION SPECIFICATIONS
1

.5 r~ c K :s r;:; rUS
----Programmer E.
R. Y£) e ,e

Page

76 77

~;;~;~:ation !s ITlole I~

Program

Control Card Specifications

i _~r~

....

Core
Size to
~ Compile

Line

(L)

0111

Core

Size to
:i 0. Execute

~ .~
"B't:
o :J

~
3

~ .~

"'::;;
~ u

.g "o ::;;

Number ~
Of Print ~
Positions

Bil !"~ ~ ~ """~ """""~ . ""

10 11 12 13 14 15 16 17 18 1920 21 22 23 24 25

Refer to the specific System Reference Library manual for actual entnes.

« .... " .... ;'" " "

j j

j j j

~~~ ~~
"

00 "
I

"
I

.. "

00 ..
I

ro" " " ,.

File Description Specifications
Mode of Processing

File Type

End of File

Type ot t"lIe

Organization
or Additional Area

File Format
a>
.,.!.

]
3

0121
.j::a.

00
.....:J

1wI~1~
7

Block
Length

Record
Length

a:
...J

Number of Tracks
for Cylinder Overflow

Extent Exit
for DAM

Record Address Type

Sequence

Filename

Line

File Addition/Unordered

Length of Key Field or
of Record Address Field

File Designation

Device

Symbolic
Device

Number of Extents
Tape
Rewind

Name of
Label Exit

z·

Overflow Indicator

!!i

::::I~

~ 0
«
:::,

Condition

Core

Index

Ul.~

z=>

10 11 12 13 14

IFIOILIDI~4ILI I

IIlpl£1AI II 1 1 I 119161 1 I I 1 1 I 1 I I I 1 IMIFIclul1

:: :~~ ~;I~III:I~jJ.J±lliH/I~~i±l±illJ 1IIIIm~;j€~1ttllll LJllD1DJJjJJJll11111

1015+_IF

raTs

.---.....-

.... -

1111111-+-+--+
017

-{ I I I I I 11--+ l----j--t I 1 1 I 1 1 1 I 1 I I 1 1 I 1 I I I I
-I--+-++--+-I I I 1 I I

I I I I 1 I I I 1 I I

+--+-+-+--

I I I I 1 I 1 I 1--+-++-+-1

1 1 I I 1 1 1·+++-+-+--+-+--+-++-1 1 1 1 1 1 1 1

Figure 22.15 Stock status. Control card and file description.

IBM,

International Business Machines Corporation

Date

12/0'
E. R Y

S~cK

GX21-9094-1 U/M050
Printed in U.s.A.

INPUT SPECIFICATIONS

RPG

5Tnrv.s

75 76 77 78 79 80

Program
Identification

pageD]

f) E R

Program

IIIIIII

Programmer

Record Identification Codes

Filename

~{,
1.8;
a. E·g
_

Jl
7

~
iii

;

I Position

~I.eli
o 2
Z

Position

il§1

---r----,

1-1

~1~ljlll~

CI>

Field
Indicators

Field Location

Field Name
From

~
-.;

.tJ ~
0>

g ~:~

ZUU(/)o..

I I~~,';~. "O"
~, ~8' i~ I I.'~
0

Sign

1Il

Plus Monus or
Blank

1011121314115161171181192012122 23 24125126127128 29 30 3113213313413536 37 38139140141142143144 45 46 47148 49 50 5115215354 55 56 57 58159 60161621636416566167681697017172 73 74

olLIDIBIAlll

I I I I I INII

Figure 22.16 Stock status. Input specifications.

GX21·9093·1

IBM

International Business Machines Corporation
Printed in U.S.A.

RPG

E#7 Ix)<

Date

CALCULATION SPECIFICATIONS
1

S~. SI /1T 1/.5

Program

E:.

Programmer

page[D

Program

Identification

76 77 78 79 80

IIIIIII

R ~...::t£~"t:.-=---- ______
Resulting

Indicators

Indicators
Arithmetic

Plus Minus Zero

~
5-

Line
Cl.

:'i

--'

Factor 1

Factor 2

Operation

Result Field

Field
Length

15
z

Compare

'§ High Low Equal
~1>21<21=2
Lookup

Table (Factor 2) is
High

¢I

plJ AN

liDO

OL D 8

I¢ I

plJ AN

A(J[)

51'

o elKI

I¢ Z

<;JIJ

R£

£:

I!!

116 9

i i ! i i
.,. oirieiAiL. : 11 q,
1'10 1R E,e. I 1q,

rio Irs rl(
I
I

i
I

:-f61¢

.t--t-+

10 TIW rll

-lb4

oS "io!e:K:
I

AD!)

oLOBAL
5

P7~ i I
R.£!£:!p~
-+-t--+oc!~!
1 s'qOC.K
-r

.j::o.

00
\C

Low Equal

9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 5253 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74

Comments

i
I

I 1t

OUTPUT - FORMAT SPECIFICATIONS

1 2

Page

srock ST/?TUS

Program

75 76 77 78 79 80

[I]

~;::~f:ation I

IIIIII

E. ~ YPcR

Programmer

...
~

I.C

Space

Output Indicators

Skip

I Jd

.J:;

Filename

Line

o
o
o

o
o

2'5
o u
~

And
Q.I

~~«

~ Iii co
>-"

j
3

W ~

~~
7

Field Name

(~
I~
. •. iii

; $

;3 ~

~ ~

End

Positon
in
Output
Record

Yes
No
Yes
No

No Sign

CR
A

B
C
0

J
K
L
M

X = Remove
Plus Sign
Y = Date
Field Edit
Z = Zero
Suppress

Sterling
Sign
Position

Constant or Edit Word

I. I
IT EfIII

DE SC Rl-r
OL D8 tiL
~£ CF TS
IWIT fll) RL
S1 OCk

I, Z

«.~Q~~~~~~~~~~~W~~~~~~wu~ooro

Is(! 9
Lf(

zlsD
38
S¢8

6f/J 8
7(J8

81pS

88
TO "13 AI-

S"I~ 8
61> B

1 0

Iro il!. €C

rw 7/1

71$6 B

1 2

To 1s riA::

Blp 8

1 3

1 4

ON EW 8AL

88
L

'E IJ(~ OR'

'~ IN II L '

I 7 EI1

16
1 7

DE :5C Ri

I fa
32

¢u AN

317

to Print

0..

rl4

Yes
Yes
No
No

Zero 8alances

9 10 11 12 13 14 15 16 17 18 1920 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38139 40 41 42 43

rK RPT

Edit Codes
Commas

0
0
0

Figure 22.18 Stock status. Output-format specifications.

71 72 73 74

RPG II OUTPUT FORMAT

491

(In planning any inventory application, it is necessary to anticipate the possibility of a minus balance in stock. The condition
can occur when incorrect quantities of receipts and withdrawals
are recorded.)
09, 10, 11, 12. When a control break in item number occurs,
the balance just accumulated will be printed and punched as specified in the Output-Format Specifications. At the same time, the
totals are added to fields when final totals are accumulated: total
balance, total receipts, total withdrawals, and total stock.
Figure 22.18 shows the Output-Format Specifications. The
two files are named the printed stock report and the punched
stock card. The summary card produced in one run will be the
previous balance card for the succeeding run.
One line is printed on the stock report whenever a control
break occurs in item number. The fields named in lines 02 through
07 are printed. If indicator 09 is on, the minus balance is printed
with the word ERROR. An output summary card is punched as
specified in lines 14 through 17.
After the last record is processed from the input file, the
final totals are printed as specified in lines 9 through 13. Totals
may be crossfooted and balanced to control totals to check the
accuracy of the run. With additional programming, the summary
cards can also be totaled to prove that balances are carried forward
to the next report accurately.

1. Complete the Output-Format Specifications for all output files of the

Questions

case study problem presented following chapter 20.
2. Prepare a review of your solution for class discussion.

and
Exercises

INDEX

A and B registers, System/3, 351
Accountancy, 2
Accounting, 2
Accounting cycle, 10
Accounting entries, 40
Accounting formula, basic, 4, 6
Accounting machine, 241, 246-269, 298-308
Accounting period, 7
Accounting transactions, posted, 6, 7
Accounts payable, 5, 32, 33
distribution reports, 33
voucher, 33
Accounts receivable, 46-48, 118, 187
Accrual basis, accounting, 4
Accrued payroll, 29
Accumulate feature, IBM 129,85
Accumulating totals, IBM 402,259-263
Accumulator, 337
Accuracy:
requirements for, 383
transcription, 88
Acknowledge, 45
Actual cost, 29
Add, RPG II, 447
Add zoned decimal, RPG II, 358, 359
Adder, 338,339
Addition, 341
Addition and Subtraction, IBM 604, 216
Address:
disk track, 371
storage, 333, 340
Address label writing, IBM 407, 299, 300
Address register, 337, 341
Advance, payroll, 29
All cycles, IBM 402, 255
Allocation of storage, 383
Allowances, 33
Alphabetic sorting switch, 150
Alternate traCk, disk storage, 373

American Standard Code for Information Interchange
(ASCII), 333
Amount fields, 105
AND, RPG II, 415
Anticipation discount, 33
APplications processing programs, System/3,
376,377
Apron, 33
Arithmetic/logical unit, 31, 336, 337, 353
Arithmetic operations, RPG II, 446-455
Assembly, 386-388
with optional manual control, 54
phases of RPG II, 398
program, 390
with specifications, manual control, 54
Assembly run, 397
Asset, 2, 21
Asterisk, RPB II, 400
Attendance record, 25, 29, 101
Audit trail, 120
Automatic carriage, 245
Available, inventory, 40
Average cost, inventory, 40
Back order, 45
Back order control, 43
Balance card, 40
Balance sheet, 18, 21
Balance test for step suppression, IBM 604, 223
Base rate, 29
Basic conversion, 54
Basic operation, IBM 604,227-239
Basic production, manual control, 53
Batch control, 132-134
Begin subroutine, RPG II, 464
Bill of lading, 43, 46
Billing, 42,45,46,166,191,411,412
accounting entries, 45
Binary coded decimal (BCD), 329, 331

493

494
Binary coded decimal interchange code (EBCDIC),
350
Binary to decimal, fraction conversion, 323
Binary to hexadecimal, integer conversion, 324, 325
Binary notation, 316
Binary number systems, 315-317
Binary to octal:
fraction conversion, 322
integer conversion, 320
Binary State, 313, 314
Bonus, 29
Bookings reporting, System/3, 377
Bookkeeping, 1,2
Bouchon, 58,59
Branch instructions, 342
Branching operations, RPG II, 459-462
Brush, sorter, 138
Brushes, mark sensing, 210
Burden, 29
Burden account, 29
Buying service, file maintenance, 185, 186
Byte, 316
Byte capacity, RPG II, 402
Calculation, IBM 604, 215-218
Calculation capacity, IBM 604, 216, 217
Calculation results, 102
Calculation specifications, RPG II:
columns 7,8, 436
columns 9-17, 436, 438
columns 18-27,438
columns 33-42, 438-440
columns 43-48, 440, 441
columns 49-51, 441
column 52, 441
column 53, 441
columns 54-59,442
Capital,2
Card arrangement, RPG II source deck, 400
Card capability, System/3, 80-column, 361
Card capacity, 70
Card check, 102
Card counter:
IBM 129,85
sorter, 163
Card data recorder:
IBM 129, 79-86
IBM 5496, 90-97
Card design:
automatic punches, 110
card punches, 109
computers, 110, 111
control panel wiring, 111
machine characteristics, 109
sorters, 110
Card feeds, System/3, 344
Card fields, 70, 71
Card handling:
manual,112
sorter, 144-146
Card input, System/3, 344, 345
Card-oriented programs, System/3, 363, 364
Card output, System/3, 346-349
Card read punch, IBM 521,218,219
Card reading, IBM 402, 247, 248
Card selection, sorter, 152, 153
Card speeds, System/3, 345

INDEX
Cash basis, accounting, 4
Cash disbursement register, 33
Central processing unit (CPU), 311, 336, 337
Chain printer, System/3, 348, 349·
Channel, System/3, 351
Character emitter, IBM 407, 306
Character printing, System/3, 347,348
Characters, special, 136
Chart of accounts, 11, 21
Check bytes, disk storage, 372
Check digit, 125, 126
Checking filing' of detail behind master cards, 190
Checking interspersed gangpunching, 199, 201
Chute blade, sorter, 139
Class selection, 284, 285
Clock number, 29
Clock station, 30
Closing out books, 17
Closing out income and expense, 9
COBOL, 393,394
Code checking, 328, 329
Codes, one digit, 105
Collate, System/3, 364-366
Collating sequence, alteration of, 403
Collator:
card paths, 168, 171
IBM 88, 168-192
Column indicator, IBM 5496,91
Columns:
mark sensing, 210-212
punched card, 63, 70, 71
Column shift, IBM 604,220,221
Column split, IBM 604,236
Combined files, RPG II, 398
Combined reproducing and gangpunching, 206-208
Comment line, RPG 11,400,401
Comments, program, 387
Common channel, IBM 604, 220
Compare, RPG II, 456-458
Compare and testing operations, 456-458
Comparing:
IBM 402, 264, 265, 268
IBM 519, 194,205
two fields in same card, 187, 188
Comparing unit:
collator, 175
IBM 519,202,203
Compilers, 391-393
Computer-programmer relationship, 385, 388-391
Computer system, 311
Condition register, System/3, 351-353, 358, 360
Console, 312, 338
Constant factors, 335
Constants, RPG II, 439
Contamination control, disk storage, 370
Continuous job card, 30
Control break, 421
Control field, 105,167,171,178, 181, 1~0, 191,
374,421
Control or Header Card Specifications, RPG II:
columns 7-9,402
column 10, 402
column 11, 402
columns 12-14,402
column 1 S, 403
columns 17-20, 403
column 21, 403

INDEX
Control or Header Card Specifications, RPG II (cont.):
column 26, 403
Control level indicators, RPG II, 421, 422,436,438
Control panel, 171, 195,204,216,218
IBM 402, 249-252
IBM 521,227
Control panel wiring, IBM 402, 252-269
Control section, 311, 336
Control tape, IBM 407, 303, 304
Control total, 121, 123
Conversion program, System/3, 80- to 96-column
cards, 368
Corner cuts, 65, 116
Co-selectors, 273, 282, 284
Cost accounting, 50-52
Counter:
counter control, IBM 407, 271-277
counter read-in minus, IBM 604, 223
counter read-in plus, IBM 604,223
counter read-out, IBM 604,223
counter read-out and reset, 223
IBM 402,259-263
IBM 604,221,225
Coupling exit, IBM 407, 274
Credit, 21, 33
Credit entry, 7, 15, 18
Credit memoranda, 45, 46
Crossfoot, RPG II, 452
Crossfooting:
accounting machines, 286, 287
IBM 604,227-230
Current earnings card, 30
Cycle, System/3, 353
Cycle inventory taking, 40
Cycle stealing, System/3, 354, 355
Cylinder, disk storage, 371, 372
Daily time card, 30
Data:
constant, 102
determining card, 102, 103
historical, 78, 87
input, 77
location in other cards, 107
machine produced, 103
real-time, 77
reference, 103
representation for machines, 60-63, 313-315
symbolic representation, 61, 62
Data checking, System/3, 344
Data conversion, System/3, 102
Data entry keyboard, System/3, 361,366
Data fields, disk storage, 372
Data format, System/3, 350, 351
Data output, 77
Data processing, 77
department, 119
procedure, 118
system, 77, 310, 311
Data protection, disk storage, 370
Data recorder IBM 5496, 90-97
Data recording program, System/3, 366
Data sequence:
determining, 107
source document, 109
Data verification error, System/3, 367, 368
Data verifying program, System/3, 366-368
Day shift, 30

495
Debit entry, 7,15,18,21,33
Debit memo, 46
Debug, RPG II, 398, 467, 468
Debug operation, RPG II, 403
Debugging, 390, 391
Decimal to binary, integer conversion, 320
Decimal to hexadecimal, integer conversion, 326
Decimal to octal:
fraction conversion, 322
in teger conversion, 319
Decimal positions, RPG II calculation, 441
Deduction authorization, 26
Deduction card, 30
Deduction register, 30
Defective track, disk storage, 372, 373
Denominating cash payroll, 30
Description, item, 46
Detail file, 167
Detail print, 241, 255-259
Digit selector:
IBM 407, 280, 281
IBM 604,236
sorter, 164
Direct access, disk storage, 374
Direct Labor, 28
cost, 30
Disk cartridge, 369
Disk file organization, 373-375
Disk pack, 369
Disk recording, 369-370
Disk storage:
operation, 369, 370
organization, 370-373
System/3,368-375
Display panel, IBM 604, 221-223
Distribution of costs, 21, 30
Divide, RPG II, 449, 450
Division, IBM 604,216,224
Document register, 130-132
Double-punch and blank-column detection, 194, 197
Double transposition error, 126
Dual cards, 99, 100, 115
Dual programming feature, System/3, 355
Due date, 34
Duplication, IBM 129,78
Earnings record, 30
Edit entries, RPG II, 474, 475
Edit stop switch, sorter, 154
Edit switch, sorter, 154
Editing, 186, 187
Electrographic pencil, 211
Electronic calculating punch, IBM 604,220,221
Employee number, 30
End-of-file condition, RPG II, 406
End printing, 80-column card, 194,209,210
End subroutine, RPG II, 464
Erasure, card records, 83, 90
Exception, RPG II, 467
Execute subroutine, RPG 11,464
Execution run, 398
Expense, 3, 7, 13
Extended binary coded decimal code (EBCDIC), 332
External indicator, RPG II, 408, 428, 438
Fabric requirements reporting, System/3, 377
Fabrication, 54
Factor storage, IBM 604, 220, 222, 225

496
Falcon, M, 58, 59
Fanning cards, 146
Feed station, IBM 129, 79
Field, card, 70, 88, 136, 142, 151, 152
Field, determining size, 103-106
Field, fixed length, 71
Field description, RPG II, 410
Field indicators, 431, 437
Field length, 71
'
RPG II calculation, 446, 447
Field name, 440
Field positions, 71
Field record relation, 428
Field reference, 105
Field selection, IBM 407, 282
FIFO (first-in, first-out), 38-40
File, RPG II, 398
File and card type identification, RPG II, 410
File description, RPG II, 398
File description specifications, RPG II:
columns 7-14, 405
column 15,405,406
column 17, 406
column 18, 406
columns 24-27, 406
column 32, 407
columns 33, 34, 407
column 39, 407
columns 40-46, 408
columns 71,72,408
File feeding device, 137
File handling operations, miscellaneous, 184-192
File identification, RPG II, 404
File maintenance, 167-192
summarization, 168
transcription, 167
File name, RPG II, 405
Final total, IBM 402, 260
Finder card, 188
Finished goods requirements reporting, System/3, 377
Finished stock, 40
Fixed asset, 21
Fixed count check, 329
Fixed length records, disk storage, 372
Flag byte, 372
Floating dollar sign, 306
Flowchart, 381
Force~ RPG II, 467
Forecasting, manual control,S 5
Form design, IBM 407, 304-308
Form spacing, IBM 407, 302, 303
FORTRAN, 394, 395
Functional units, computer, 337, 338
Gang job card, 30
Gangpunched extensions, 30
Gangpunching, 181, 193-203,205,239
Garnishee, wage, 20
General ledger, 20
General register, 338
General storage, IBM 604,220,222,225-227
Glossary:
accounts payable terms, 33, 34
accounts receivable, 49, 50
cost accounting, 52, 53
general accounting, 21, 22
inventory, 40
manufacturing, 56, 57

INDEX
Glossary. (cont.):
order writing and billing, 45, 46
payroll,29
Go to, RPG II, 459
Grand total, IBM 402, 260
Graphic subsets, BCD code, 68,69,331
Gross earnings, 27, 30
Group bonus plan, 31
Group indication, 269
Group multiplication, 217
Group printing, 243, 421
Groups, card, 191
Half-adjust:
IBM 604,224
RPG 11,441
Halt indicators, RPG II, 431, 437, 438, 443
Halt operation, 342
Hardware, 386
Heading information, 46
Hexadecimal to decimal, integer conversion, 325, 326
Hexadecimal number system, 317,326
High-order character, 135, 1 36
Hollerith, Dr. Herman, 62, 63
Hopper, 116,79, 90,137,158,168,174, 175
Housekeeping routines, 383,384
Hubs, control panel, 251
IBM,63
IBM card code, 64, 66, 67, 68
Identifier field, disk storage, 371, 372
Identifying characters, RPG 11,414-417
Idle time, 31
Income, 3, 13
Income statement, 21
Index, disk storage, 374
Indexed sequential access, disk storage, 374, 375
Indirect labor, 28, 31
Individual job card, 31
Input of combined file description, RPG II, 405
Input device, 77
Input files, RPG II, 398
input file name, RPG II, 410
input file record fields, RPG II, 417-421
Input/output channel, System/3, 354
Input specifications, RPG II:
columns 15, 16,410,411
column 17,412
column 18,412
columns 19,20,412-414
columns 21-41, 414
column 42, 417
columns 44-48, 417-421
columns 59,60, 421
columns 61, 62, 422
columns 63, 64,428
columns 65-70, 428-431
Insignificant zeros, 71
Instruction counter, 340-342
Instruction execution, System/3, 353, 354
Instruction format, System/3, 356-360
Instruction register, 337
Instruction timing, 360
Instructions, 335, 336
Instructions as data 336, 354
Integer conversion, 319-323
Interleaving functions, System/3, 355
Intermed.iate total, 243, 260, 265, 287

INDEX
Interpreting, punched cards, 102
Interrupt, 354, 355
Interspersed gangpunching, 197, 354, 355
Inventory, 34, 125
average cost, 38
costing finished goods, 38
maintenance cost, 37
obsolescence, 38
order points, 36, 37
perpetual, 35
physical count, 34
processing cost, 38
Inventory cards, 99
Inventory control, 34
Inventory formula, 35, 36
Inventory management, 38
Inventory requirements, 36
Inventory status file, 188
Invoice, 46
Invoice register, 34
Issues, 40
Jacquard, Joseph Marie, 59,60
Job order, 31
Joggling cards, 145
Journal, 10, 21
Journalizing, 14
K bytes, 350
Key, disk record, 373, 374
Key numbers, 188, 189
Keyboard:
IBM 129,78,81
Keypunching:
80-column card, 78-84
96-column card, 89
Labor distribution, 24, 26
Language:
data representation, 315
programming, 315, 397
Last record indicator, RPG II, 406, 436, 438
Layout, card forms, 107, 113, 115
Lead time, 40
Ledger, 21
Ledger accounts, 11
Level, language, 392
Liability, 3, 5, 21
Library, program, 383, 391, 392
LIFO (last-in, first-out), 38, 40
Line Counter Specifications, RPG 11,407,471,472
List, IBM 402,256
List program, System/3, 96-column card, 364
Local storage registers, System/3, 353
Logical nature of computers, 315
Logical path, program, 342
Lookahead fields, RPG II, 413
Lookup, RPG II, 413
Low-order character, 135, 136
Machine cycle, 248, 339, 340
Machine oriented, 392
Manual cycle, 388
Manufacturing control, 53-56
Mark-sensed cards, 100, 116
Mark sensing, 194, 210-214
Master cards, 98

497
Master personnel cards, 31
Master rate card, 31
Matching fields, 422, 425
Matching record indicator, RPG II, 422
Ma tching two files, 1 85, 1 86
Material accounting, 34
Ma terials planning, 55
Memory:
artificial, 335, 336
machine, 60
Merging:
with equal selection, 183-184
merging two files, 175-178
with selection, 178-183
System/3, 365, 366
Microsecond, 339
Millisecond, 339
Minor total, 241, 242,260, 265,287
Miscellaneous data, 46
Mnemonics, 387
Modifier program, System/3, 377
Modulus, self-checking number, 126
Modulus 10, 127-129
Modulus 11, 129, 130
Move, RPG 11,452
Move left, RPG II, 453-455
Move remainder, RPG II, 450, 452
Move zone operations, RPG II, 455
MR Indicator, 438
Multifunction card unit, System/3, 344, 345
Multiple column selector, sorter, 155-157
Multiple line printing, IBM 407, 292-294
Multiple line read, IBM 407, 298-300
Multiple selection, IBM 407,277-280
Multiplication, IBM 604,216,230-233
Multiplier-quotient storage, IBM 604, 220,222
Multiply, 449
Multiply minus, IBM 604, 224
Multiply plus, IBM 604, 224
Nanosecond, 339
Net income, 4
Net loss, 4
Net pay, 27
Net worth, 6, 14, 18,21
Normal alphamerical print entry, IBM 402,256,257
Normal zone entry, IBM 402, 258, 259
Numeric bits (BCD), 329, 331
Numerical print entry, IBM 402, 257
Object deck, 388
RPG II, 402
Object program, 388, 390
Octal to binary:
fraction conversion, 322
integer conversion 320
Octal to decimal, integer conversion, 319, 320
Octal number system, 317, 318
Odd parity, 329,331
Offset gangpunching, 201, 202
Offset gangpunching comparing, 202
Op code register, System/3, 353
Operands:
instruction, 339, 340, 341,387
RPG II, 446
Operation codes:
System/3, 353
RPG II, 439, 446

498
Operation mnemonics, 387
Operation notes, IBM 604, 224
Operation part, instruction, 339,340,341
Operation planning, IBM 604, 223-227
Operations evaluation, manual control,S 6
Operator controls, IBM 129, 80
OR, RPG II, 417, 418,428
Order acknowledgment, 43
Order analysis, 43, 46
Order edit, System/3, 376, 377
Order writing, 42, 377
Output cards, 101
Output data, format, 383
Output device, 77
Output device assignment, RPG II, 408
Output files, RPG II, 398
Output-format specifications:
columns 7-14, 469
column 15,469,471
column 16,471
columns 17, 18,471
columns 19-22,471
columns 23-31, 472
columns 32-37, 474
column 38,474,475
column 39, 475, 476
columns 40-43, 476
Output indicator, 472
Output stacker, RPG II, 471
Overflow indicators, RPG II, 438
Overflow skipping:
IBM 407, 303
System/3, 471
Paper forms, 349
positioning, 245
Paper tape, 63
Parity bit, Systeml3, 350
Parity check, 329
System/3, 351, 353
Payroll, 22,118,123,124,212,214,284,285,444,
465,467
file maintenance, 178-181
Payroll accounting entries, 28
Payroll advice, 31
Payroll change cards, 100
Payroll computation, 27
Payroll deduction, 31
Payroll master card, 31
Payroll record, master, 24, 25
Payroll register, 31
Payroll source documents, 24
Payroll writing, 28
Perpetual book inventory, 40
Personnel record, 25, 26
Physical count, 41
Pilot selector, IBM 407, 272, 274
Planning chart, IBM 604, 224
Plugwire, 251
Posting to accounts, 6, 7
Posting from journal, 14
Pricing file maintenance, 184
Primary sequence unit, collator, 171-174
Print chain, 349
Print checks, Systeml3, 349
Print wheels, IBM 407, 296, 297
Printer, System/3, 348, 349

INDEX
Printer-keyboard, System/3, 360, 361, 385
Printing methods, 245
Problem analysis, 380,381
Problem-oriented program, 392
Procedure design, 381
Procedure management, 121
Processing unit:
System/3, 349
System/3 data flow, 351-354
Processor check, RPG 11,397
Production statistics feature, IBM 129,85
Profit and Loss (P & L), 17
Program assembly, 386-388
Program assembly checks, 390
Program control:
IBM 129,83,84
IBM 402, 263-265
Program execution, 340-342
Program generator, 397
Program indicators, RPG II, 428, 431
Program number, IBM 604, 224,225
Program preparation, 381
Program routines, availability, 383
Program start, IBM 402, 265, 266
Program steps, 216
Program suppress, IBM 604, 225
Program unit, IBM 604,221
Programmer, 379, 380, 391,403,404
Programming sheet, 387
Programming systems, 315, 380
Progressive total printing, IBM 407, 287
Property, Business, 4
Punch unit, IBM 519, 194, 195, 196
Punch X brushes, IBM 519, 197-202
Punched card:
80-column specifications, 63
96-column specifications, 72-75
format, 87, 88
types of, 98-102
Punching, 96-column card, 91
Punching position, 64
Punching station, IBM 129, 78, 80
Q register, System/3, 353

Random error, 126
Read check, System/3, 345
Read unit, IBM 519, 194
Read-units-into, out-of, IBM 604, 225
Read-X brushes, IBM 519,204
Reading, IBM 407, 295
Reading station:
IBM 129,78
IBM 402, 248, 249
IBM 407, 295, 296
Receipts, 41
Reconciliation of bank statement, 31
Record count, 121
Record identifying indicator, RPG II, 412, 414, 437
Record keeping, 1
Record length, 406, 407
Records, historical, 166
Reference data availability, 383
Register, 337
Register masking, 338
Register shifting, 338
Reproduce/interpret program, Systeml3, 363, 364

INDEX
Reproducing:
card, 193,204-208
selective, 205, 206
Reproduction of data, 78, 79
Requisition, 41
Reservation, stock, 41
Resource management, 383
Restart, system, 383
Result field, RPG II, 441
Result indicators, RPG II, 434, 436, 437, 443, 444
Returns, 34
Revenue, 3
Right adjust, IBM 5496, 94
RPG II program:
example 1, 476, 477
example 2, 483-491
ROW, card, 64, 169
Safety devices, disk storage, 370
Sales transaction, file maintenance, 181-183
Scheduling, manual control, 55
Scored cards, 102
Second reading, IBM 402, 258
Sectors, disk storage, 372
Selecting cards by key numbers, 188, 189
Selecting zero-balance cards, 189, 190
Selection switches, sorter, 152
Selector, 202-204
Self-checking number, 85, 124-130
Sequence:
alphab~tic, 136
ascending, 135
descending, 135
intermediate, 136
major, 136
minor, 136
numerical, 135, 136
variable, 136
Sequence checking:
input files, 406
Sequencing, types of cards in file, 410, 412
Sequential access, disk storage, 373, 374
Set off, RPG II, 459
Set on, RPG II, 459
Ship to (invoice heading), 46
Shipping label, 43, 46
Shipping list, 42, 46
Short shipment, 34
Sight checking, 146
Sign control, IBM 604, 236, 237
Software, 386
Sold to (invoice heading), 46
Sort/collate program, System/3, 364-366
Sorter:
IBM 5486, 157-164
machine operation, 137-140
special devices, 154
Sorting:
96-column alphabetic, 160-163
96-column numerical, 159, 160
alphabetic, 149-152
alphabetic short cut, 153
block, 147-149
numerical, 142-144
printed reports, 140-142
Sorting needle, 146
Sorting time:
estimating, 146, 147

499
System/3, 366
Sorting tray, 159
Source deck, preparation of, 397
Source document, 46
Source program, 388, 390, 397
Special program, IBM 407, 288
Specification sheets, RPG II, 398-401
Square root, IBM 604, 237-239
SQRT,392
Stacker:
collator, 168
IBM 129,78,80,81
sorter, 138, 144
Standard assembly, manual control, 54
Standard BCD interchange code, 65, 69, 331
Standard cost, 31
Standard order quantity, 41
Status records, 166
Step-down, sequence, 171-174
Sterling currence, RPG II format, 403
Sticker, 34
Stock allocation, System/3, 377
Stock on hand, 41
Stock status, 45
Stock status report, 41
Storage, 333-335
computer, 336
IBM 129,82,83
System/3, 349-351
Storage address register, System/3, 351
Storage amount, program assembly, 386
Storage data register, System/3, 351
Storage register, 337
Storage units, IBM 407, 300, 301
Stored program, 335, 336, 379
Strings, records, 365
Subroutine operations, RPG II, 462-467
Subroutines, 342
Subtract, RPG II, 448
Subtract zoned decimal, 359, 360
Successive calculations, IBM 604, 233-236
Summarize, 241, 243
Summary punch, 168
Summary punching:
IBM 407, 288-292
IBM 519, 193, 194
Supervisory programs, 379
Suppress on minus balance, IBM 604, 223
Suppress on plus balance, IBM 604,223
Synchronizer byte, disk storage, 372
System' halts, RPG II, 402
System language, 388, 390
T account, 6, 7
Tables, RPG II, 407
Tabulate, 241
Tag, RPG II, 459-462
Tape-controlled carriage, IBM 407, 302, 303
Terms, 46
Test zone, RPG II, 458
Third reading, IBM 402, 257
Tiers, punched card, 73
Time card, 25
Time sharing, 354, 355
Time sheet, 31
Timekeeper, 26
Total program, IBM 402, 266, 267
Trade discount, 34
.

500
Transcript cards, 98
Transcription error, 125, 126
Transcription process, 87, 88
Transcription register, 41
Transmittal letter, 133
Transportation error, 126
Trial balance, 8, 16,21,34,45
Type bar:
alphanumeric, 246, 253
numerical, 246, 253
Unit cost, 32
Unit count, 123, 124
Unit record, punched card, 64

Variable length overflow, IBM 407, 306
Variance, 32

INDEX
Vaucanson, Jacques de, 58, 59
Vendor, 46
Verification, 96-column card, 94, 95, 96
Verification process, 88, 89
Voucher, 21
Warehouse system flowchart, 381
Work-in-process file, 32
X-punched cards, 191, 192, 217
Zero and add, RPG II, 447, 448
Zero balance, 189, 190
Zero check, IBM 604, 224
Zero elimination, sorter, 157
Zero and subtract, RPG II, 448, 449
Zero suppression, 307
Zone positions, BCD 329,331
Zoned decimal format, 350

Source Exif Data:

File Type                       : PDF
File Type Extension             : pdf
MIME Type                       : application/pdf
PDF Version                     : 1.3
Linearized                      : No
XMP Toolkit                     : Adobe XMP Core 4.2.1-c041 52.342996, 2008/05/07-21:37:19
Create Date                     : 2015:08:18 21:04:24-08:00
Modify Date                     : 2015:08:18 20:47:42-07:00
Metadata Date                   : 2015:08:18 20:47:42-07:00
Producer                        : Adobe Acrobat 9.0 Paper Capture Plug-in
Format                          : application/pdf
Document ID                     : uuid:ab194851-af27-d246-9fa0-c3a844748423
Instance ID                     : uuid:fa173d5b-e99a-454c-850b-3cce2c8544ce
Page Layout                     : SinglePage
Page Mode                       : UseNone
Page Count                      : 513

EXIF Metadata provided by EXIF.tools

Business_Systems_With_Punched_Card_Data_Processing_and_System_3__10_1973 Business Systems With Punched Card Data Processing And System 3 10 1973

Business_Systems_With_Punched_Card_Data_Processing_and_System_3__10_1973 Business_Systems_With_Punched_Card_Data_Processing_and_System_3__10_1973

Navigation menu

Versions of this User Manual:

Views

Navigation