G_1616 G 1616
User Manual: G_1616
Open the PDF directly: View PDF .
Page Count: 84
Download | |
Open PDF In Browser | View PDF |
GenRad GR 161 6 Precision Capacitance Bridge GR 1621 CapacitanceMeasurement System Form 1616-0100-00 InstructionManual WARNING Dangerous voltages may be present at the terminals of this instrument. Observe all warnings contained in this manual. Refer all servicing to qualified personnel. G R 161 6 Precision Capacitance Bridge G R 1621 CapacitanceMeasurement System Form 1616-0100-00 @GenRad, Inc., 1988 Concord, Massachusetts, U.S.A. 01742 January, 1988 Tableof Contents Page Section CONDENSED OPERATING INSTRUCTIONS ix 3.4 SPECIFICATIONS . . . . . . . . . . . . . . . . . . . . . .. x Section 1 INTRODUCTION 1.1 Purpose 1.2 Description 1.2. 1 General................. 1.2.2 Bridge Circuit 1.2.3 Standards 1.2.4 Oscillator 1.2.5 Detector 1.3 Controls, Indicators, and Connections . . . . . . . . . . . . . . . . . . . 1.4 Accessories. . . . . . . . . . . . . . . . . . . 1-1 1-1 1-1 1-1 1-2 1-2 1-2 3.5 . 1-3 . 1-3 3.6 2.5 2.6 2.7 2.8 General 2-2 Dimensions 2-3 Environment 2-2 Bench Models 2-2 2.4.1 Cabinet Removal 2-2 2.4.2 Conversion For Rack Mounting 2-2 Rack Models 2-2 2.5.1 General 2-3 2.5.2 Installation 2-3 2.5.3 Conversion To Bench Use .. 2-3 Power-Line Connection 2-4 Line-Voltage Regulation 2-4 System Connections 2-4 2.8.1 Oscillator. Bridge, and Detector 2-4 2.8.2 BCD-Capac itanee-Output Connector 2-4 2.8.3 BCD-Conductance-Output Connector 2-5 2.8.4 Analog Outputs 2-5 3.7 3.8 3.9 Preliminary Checks Functional Self-Checks Phase Adjustment Balance and Readout 3.5.1 Readout Multiplier 3.5.2 Initial Settings of Lever Switches 3-3 3-3 3-3 3-4 3-4 . 3-5 3.5.3 Balance Procedures 3.5.4 Final Balance 3.5.5 Readout Correction 3.5.6 Units of Measurement Parameters of The Unknown Capacitor . . . . . . . . . . . . . . . . . . . . .. Series Equivalent Parameters . . . . . . . . . . . . . . 3.6.2 Dissipation Factor Frequency 3.7.1 Setting The Frequency 3.7.2 Monitoring Frequency 3.7.3 Locking To a Frequency Reference . . . . . . . . . . . . . .. 3-5 3-6 3-7 3-7 3-7 3-7 3-8 3-8 3-8 3-8 3-1 0 Voltage Level 3-10 Accuracy 3-10 3.9.1 Accuracy Versus Frequency and Cx 3-10 3.9.2 Accuracy Versus Temperature 3-11 3.9.3 Range and Dissipation Factor Limitations 3-12 3.9.4 Shunt Capacitance To Ground 3-1 2 3.9.5 Fringing Capacitance 3-13 3.9.6 Conductance Accuracy 3-14 3. 10 Precision 3-15 3.11 External Standards 3-15 3.11.1 Range Extension To 111 Microfarads . . . . . . . . . . . . .. 3-15 3.11 .2 Extension of C Resolution .. 3-16 3.11.3 Externally Determined Accuracy/ Comparisons 3-16 3 OPERATION 3.1 3.2 3.3 Connection of Unknown Capacitor 3.4.1 Three-Terminal Capacitors 3.4.2 Two-Terminal Coaxial Capacitors 3.6.1 2 INSTALLATION 2.1 2.2 2.3 2.4 Page 3-1 3-1 3-2 iii Tableof Contents(Cont'd) Page Section 3.12 3.13 3.14 3.15 Section 3.11.4 Test-Fixture Compensation. 3-17 3.11. 5 Range Extensions To 11 Microsiemans 3-1 7 Precise Comparisons 3-17 3.12.1 Balance Comparisons 3-17 3.12.2 Direct Substitution 3-17 Non-coaxial 2- Terminal Capacitors . 3-17 3.13.1 Unshielded 2- Terminal Capacitors 3-18 3.13.2 Shielded 2- Terminal Capacitors 3-18 Dc Bias 3-20 3.14.1 Normal Bridge Configuration/ Parallel Bias 3-21 3.14.2 Reversed Bridge Configuration/ Series Bias 3-21 3.14.3 Dc In The Ratio Transformer/ Demagnetization 3-22 Reversed Configuration 3-22 3.15.1 Explanation 3-22 3.15.2 Procedure 3-22 4 THEORY 4.1 4.2 4.3 4.4 4.5 Introduction 4-1 Properties of Capacitors 4-1 4.2.1 Basic Components of Capacitance 4-1 4.2.2 Inductive and Lossy Components 4-2 4.2.3 Frequency Characteristics .. 4-4 Basic Bridge Circuitry 4-5 4.3.1 Elementary Capacitance Bridges 4-5 4.3.2 Transformer-Ratio Bridges .4-5 Circuitry of The 1616 Bridge 4-7 4.4.1 Excitation 4-7 4.4.2 Circuit For The Unknown 4-7 4.4.3 Capacitance Standards 4-7 4.4.4 Conductance Standards 4-8 4.4.5 External Standards 4-9 4.4.6 Zero Adjust 4-9 C-Standards Accuracy 4-9 4.5.1 Calibration 4-9 4.5.2 Sealing 4-10 4.5.3 Thermal Lag 4-10 iv 4.6 4.7 Page G-Standards Accuracy Ratio Accuracy 4.7.1 Residual Impedances . . . . 4.7.2 Example With 1: 1 Ratio . . 4.7.3 The 10:1 Ratio 4.7.4 The 1:100 Ratio 4.7.5 C Offset Due To Induction 4.7.6 Ground Circuit Impedance .. .. . . 4-9 4-11 4-11 4-1 2 4-13 4-13 4-14 4-14 5 SERVICE and DIAGRAMS 5. 1 5.2 GR Field Service Minimum Performance Standards .. 5.2.1 General 5.2.2 Zero Setting, Offsets, and Sensitivity 5.2.3 Capacitance Accuracy 5.2.4 Capacitance Ratios 5.2.5 Conductance Accuracy 5.2.6 Conductance Multipliers 5.3 Dissassembly 5.3.1 Knobs 5.3.2 Cabinet Removal 5.3.3 C-Box Removal 5.3.4 The G Box 5.4 . Recalibration and Adjustments 5.4.1 Internal Capacitance Standards 5.5 5.6 '" " 5.4.2 5.4.3 5.4.4 5.4.5 Conductance C301/Setting Lever-Switch Maintenance Switches Multipliers Zero C Stiffness Note On Trouble 5.5.1 5.5.2 5.5.3 Analysis Mechanical Damage BCD Circuits Non-Repairable Subassemblies . . . . . . . . . .. 5.5.4 Typical Parameters Parts Lists and Diagrams 5-1 5-1 5-1 5-3 5-3 5-3 5-3 5-4 5-5 5-5 5-5 5-5 5-7 5-7 5-8 5-8 5-8 5-8 5-8 5-9 5-9 5-9 5-9 5-9 5-9 APPENDIX A AN INTERPRETATION of MILLEA'S METHOD TO ELIMINATE FRINGING FROM TWOTERMINAL MEASUREMENT A-1 ! ' Listof Illustrations Title Figure 1-1 1-2 1-3 2-1 2-2 2-3 2-4 2-5 2-6 1621 Precision Capacitance-Measurement System 1616 Precision Capacitance Bridge . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Elementary Diagram, 1621 Precision Capacitance-Measurement System-GR1616 Bridge with 1316 Oscillator and 1238 Detector Front-Panel Controls, Indicators, and Connectors of the 1621 Precision CapacitanceMeasurement System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. Rear-Panel Controls, Connectors, and Cables ii x Dimensions of the Bench Models Dimensions of the Rack Models Bench-Cabinet Installation Rack-Cabinet Installation Terminal Identification at BCD CAPACITANCE OUTPUT Connector (A-J13, Rear Panel of Bridge) Terminal Identification at BCD CONDUCTANCE OUTPUT Connector (A-J 12, Rear Panel of Bridge) DC Meter Outputs Socket on 1238 Detector, Exterior (Rear) View 2-1 2-1 2-2 2-3 Simplified Bridge Diagram 3- Terminal Connection of Unknown Capacitor Simplified Bridge Diagram 2- Terminal Connection of Unknown Capacitor .... The 1621 Capacitance-Measurement System Simplified Diagram ..... Equivalent Circuits of the Unknown Capacitor ... Dissipation Factor Versus Directly Measured Parameters Specified Capacitance Accuracy of the 1616 Bridge Worst-Case Measurement Error Due to Temperature (Based on Maximum Temp Coef of C Stds) Non-Coaxial 2- Terminal Capacitors and Some Measurement Configurations ... Circuit For Applying Bias To The Unknown Capacitor-Normal Configuration .. Reversed Configuration - Bridge Circuit With Oscillator and Detector Interchanged 3-3 3-4 3-5 3-8 3-9 3-11 0 2-7 3-1 3-2 3-3 3-4 3-5 3-6 3-7 4-1 4-2 4-3 ••• 0 0 ••• 0 0 •• 0 0 0 0 0 0 ••••••••••••• 0 0 • 0 0 •••• 0 0 0 0 ••••••• 0 ••••••••• 0 0 ••• 00 ••• 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 v • 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 3-12 3-19 3-21 3-22 ••••••• ••••••• •••••••••• 0 0 0 0 •••••••• 0 0 ••••• ••••••• 0 0 0 0 Schematic Diagram of a Capacitor .. Structure of a 3- Terminal Capacitor With 2 Coaxial Connectors A Coaxial 2- Terminal Capacitor, It's Structure, Component Capacitance, and Connection To a Bridge Capacitor Lumped-Parameter Equivalent Circuits and Vector Diagrams Capacitance Versus Frequency Characteristic of a Mica Capacitor Dissipation Factor Versus Frequency Characteristics of a Mica Capacitor An Elementary Capacitance Measuring Bridge An Elementary Capacitance Bridge With Transformer Ratio Arms Circuitry For 3 Methods of Balancing a Transformer-Ratio Capacitance Bridge .... Simplified Bridge Schematic Diagram 0 • 0 •••• 0 0 •••• 0 •• 0 00 1-3 1-7 2-5 2-6 ••••• 0 ••••• 0 • 0 •••• 0 0 • ••••••••••••••• 0 0 0 1-2 2-5 •••••••••••• ••••••••• 0 o. 4-4 4-5 4-6 4-7 4-8 4-9 4-10 0 0 00000.0.0 3-8, 3-9 3-10 3-11 Page 0 •• • 0 0 • 0 0 0 • 0 • • • • • 4-1 4-1 4-2 4-3 4-4 4-5 4-5 4-6 4-6 4-11 Listof Illustrations (Cant'd) Figure Page Title 5-1 5-2 5-3 5-4 5-5 5-6 5-7 5-8 5-9 Three-Terminal Conductance Standards Interior View, Upper Rear, With the C-Box Removed Interior Bottom View . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Detent Mechanism of a Typical Lever Switch, A and B Constructions Front View; Mechanical Replaceable Parts Identified Rear View; Mechanical Replaceable Parts Wiring Diagram of BCD Output Circuits Conductance Multiplier Board Assembly 1616-4720 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Schematic Diagram of the 1616 Precision Capacitance Bridge . . . . . . . . . . . . . . . . . . . . . . . 5-5 5-6 5-7 5-9 5-11 5-11 5-11 5-12 5-13 A-1 Test Setup For 2- Terminal Measurement, With Fringing Eliminated A-1 Listof Tables Table Page Title 1-1 1-2 1-3 Front-Panel Controls, Indicators, and Connectors 1-4 Rear-Panel Controls and Connectors 1-6 Accessories Supplied With the 1621 System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. 1-9 2-1 2-2 2-3 2-4 Rack-Mounting Cabinets and Brackets for 1621 System Key to Screw Sizes Binary-Coded Decimals (BCD) Readout-Multiplier Code . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3-1 3-2 3-3 Senses of Phase-Sensitive Meters Characteristics At Bridge Detector Output Connector Bias Current Limits For Transformer Saturation 4-1 4-2 Approximate Magnitudes of Residual Parameters Error Example, Ratio 1:1 5-1 5-2 5-3 Test Equipment 5-2 Capacitance Ratio Checks 5-2 Conductance Accuracy and Multiplier Checks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5-4 vi 2-2 2-3 2-5 2-5 . . . . . . . . . . . . . 3-2 3-21 3-22 4-11 4-13 Condensed Operating Instructions WARNING Dangerous voltages may be present at the terminals of this instrument. To reduce the risk of electric shock, turn the voltage source to "0" before connecting or disconnecting device under test. CAUTION Be sure the line-voltage switches on Oscillator and Detector rear panels are properly set for the available power. b Set TERM. SELECTOR to upper X1 position unless C; is larger than 1 n F (1000 pf ). then use X 10 If appropriate, raise C-MAX lever. c. Set C and G levers to approximate values of C; and G x (parallel components of "unknown"). d. With GAIN control keep MAGNITUDE indication on scale Turn SENSITIVITY and TIME CONSTANT cw, if required to achieve final balance (below) For best resolution, increase Eo RANGE to 150 V TUNING AND PHASE ADJUSTMENT. a. Set the controls as follows. POWER : ON (up) oscillator and detector FREQUENCY SELECTORS: as desired or 1.01 kHz Oscillator Eo RANGE: 15V Eo ADJUST: MAX Detector TIME CONSTANT: 0.1 s PUSH BUTTONS: both out IN-PHASE FINE ADJUST: midrange QUAD FINE ADJUST: midrange PHASE SHIFT (inner knob): 180 deg SENSITIVITY full on (cw) Bridge TERM SELECTOR: CAL EXT MULT: OFF CAUTION Do not exceed either 350 V rms or 0.16 f volts (example: 16 V at 100 Hz) in normal configuration. e. Refine the balance, left-to-right with C levers, bringing IN PHASE meter toward zero until the G error predominates. Then continue with G levers, bringing QUAD meter toward zero until the C error predominates. Repeat step e until balance reaches'the resolution you need. 6 G levers X G Mull: 0/15 000.0 nS X 10C levers: 05 pF 000 fF 000 0 aF (First three digits masked) GAIN: 30db (or keep MAGNITUDE on scale) 2-TERMINAL MEASUREMENT. b. Fine tune oscillator to detector frequency (peak on MAGNITUDE meter). (Note 3rd osc dial is continuous). c. Reset C levers to 00 pF 000 fF 500 0 aF Reset GAIN to 100db Read just C levers for MAGNITUDE of 20-40 d. Using PHASE SHIFT (large knob) bring QUADRATURE meter to zero. Reset C levers for MAGNITUDE of 80-100, reset zero using PHASE SHIFT first and then QUAD FINE ADJUST. NOTE Keep M/\GNITUDE meter on scale for correct phase indication; other meters may be off scale. e. Reset C levers to all zeros. Adjust G levers for MAGNITUDE of 80-100. Using only IN-PHASE FINE ADJ, set IN PHASE meter to zero. 3-TERMINAL MEASUREMENT. a. Connect unknown capacitor between inner conductors (HIGH, LOW) of 3-term C, port, shield to outer conductor of either or both connectors. Cables are optional; shield at least the LOW one. (Cable capacitance is excluded from the measurement \ vii a. Set TERMINAL SELECTOR to right X1 position unless C, is larger than 1 nF, then use X 100. Raise C-MAX lever, if appropriate. b. Set readout to value of fringing capacitance of 2-terminal port, i.e., ** nF *00 pF 115 fF 000.0 aF, 0 j-tS 000.0 nS X 10-6 , if you selected Xl in step a. If XlO or Xl00, set C proportionally smaller (11.5 fF or 1.15 fF, respectively). c. Balance the bridge with ZERO ADJUST. d. Connect unknown capacitor to the 2-TERM C; port outer shell - ungrounded to outer conductor of connector, inner (shielded) terminal to inner conductor. e. Proceed as before - steps, c. d. e of the 3-terminal instructions. FURTHER INFORMA TION. Refer to the Table of Contents. Specifications 1616 PRECISION CAPACITANCE Capacitance measurement, 3-terminal: DECADES: 12. RANGE 01 aF to 1 IlF 00 'to 10 6 F). ACCURACY:' =10 ppm, when most-significant decade is 1, 10, or 100 pF per step; otherwise, and at other frequencies, accuracy is =[50 ppm (05 + 20 CIl , ) (f.,,)' ppm + (f H . ) aF]. Capacitance, 2-terminal: Same as above, except as follows. RANGE: One additional decade, to 10 IlF (10 'to 10 ' F). Conductance measurement, 3-terminal: DECADES: 5 (virtually extended to 11 by G multiplier). RANGE 100 as to 100 p.s (10 'to 10'u) ACCURACY' (01% 1 step in least significant decade). There is a small reduction in conductance accuracy at frequencies other than 1 kHz. RESIDUAL C « 0.03 pF). (across conductance standards): Conductance, 2-terminal: Same as above, except as follows: RANGE One additional decade, to 1000 p-S (10 'to 10 S). xi. and DETECTOR Connect to rear BNC con- Required: OSCILLATOR: GR 1316 recommended. DETECTOR: GR 1238 recommended. The 1616 Bridge is available with this oscillator and detector as the 1621 Capacitance-Measuring Assembly. Available: 1316 OSCILLATOR, 1238 DETECTOR and a broad line of capacitance standards. Mechanical: Bench or rack model. DIMENSIONS (wxhxd). Bench, 19.75x1381x12.88 in. (502x351x327 rnrn), rack, 19x 1222xlO.56 in. (483x310x268 mrn). WEIGHT: Bench, 57 Ib (26 kg) net, 69 Ib (32 kg) shipping; rack, 49 Ib (23 kg) net, 61 Ib (28 kg) shipping. xi. Multipliers: FOR 3-TERM: X10; FOR 2-TERM: X10, X100; affect both C and G. FOR CONDUCTANCE ONLY: Xl, Effects of these multipliers X10 , ... XlO 6 (7 positions). are included in the specified ranges. Frequency: OSCILLATOR nectors. BRIDGE for these conditions: 23() ± 1 C: humidity, <50% Frequ81lc/ RH 1 kHZ, except as noted: temperature See manual for detailed accuracy analvsis. t Registered trademark of the Carpenter Steel Co G900 and G874 - Gilbert Engineering. Glendale. Arizona 85301 10 Hz to 100 kHz. Standards: CAPACITANCE Air dielectric with TC < +40 ppm/C and D 10 ppm for 7 lowest decades; lnvar t , air dietectric with TC of 3 1 ppm/C and D <10 ppm for 3 middle decades; mica dielectric with TC of 20 =10 ppm! 'C and D <200 ppm for 2 highest decades. ADJUSTMENTS for all capacitance standards available through key-locked door on panel. THERMAL LAG: C standards for first 8 decades mounted in an insulated compartment with a thermal time constant of 6 h (time required for compartment interior to reach 63% of ambient change). CONDUCTANCE: Metal-film resistors in T networks With small phase angles. Comparison: Terminals provided to connect external standard for comparison measurements; 13-position panel switch multiplies standard by-0.1, 0 . 11. input: The smaller of 160 f.". or 350 V rrns can be applied to bridge transformer at the GENERATOR terminal without waveform distortion; 500 V rrns max, depending on conductance range, when GENERATOR and DETECTOR connections interchanged. interface: G900 locking coaxial connector on panel to connect 2coaxra I connectors on unknowns. 'G8 74 locking panel to connect 3-terminal unknowns and 2 to connect external ccta,ndcird.DATA OUTPUT: 50- pin and 36- pin type 57 connectors on rear provide connection to 8-4- 2 -1 weighted BCD contacts (reated at 28 V. 1 AI on each SWitch for capacitance and conductance values CAPACITANCE Catalog Number Description 1616 Precision Capacitance 1621 PRECISION - MEASUREMENT I nternal Temperature: C standards in bridge, about 1°C above ambient, for ultimate accuracy. allow 24 hrs to stabilize with ac power on. Frequency: 10 Hz to 100 kHz. Supplied: 1616 Precision Capacitance Bridge, 1316 Oscillator, 1238 Detector, all necessary interconnection cables, and power cord. Available: 1404 REFERENCE STANDARD CAPACITORS (10 pF, 100 pF. and 1000 pF) for calibration. Power: 100 to 125 and 200 to 250 V, 50 to 60 Hz, 51 W. Bridge Bench Model Rack Mudel 1616-9700 1616-9701 SYSTEM Mechanical: Bench or rack models. DIMENSIONS (wxhxd): Bench, 19.75x24.25x15 in. (502x616x381 rnrn): rack, 19x 20.91x11.44 in. (483x531x291 mrn). WEIGHT: Bench, 105 Ib (48 kg) net, 140 Ib (64 kg) shipping; rack, 90 Ib (41 kg) net. 125 Ib (57 kg) shipping. Description 1621 Precision Capacitance·Measurement Bench Model, 60-Hz Rack Model, 60-Hz Bench Model, 50-Hz Rack Model, 50-Hz viii Catalog Number System 1621-9701 1621-9702 1621-9703 1621·9704 FREQUENCY SELECTORS PHASE SHIFT ~ QUAD 1238 DETECTOR "'- \..: GAIN c, MAGNITUDE EXT STD ZERO 1616 BRIDGE G LEVERS 1621 Precision Capacitance-Measurement k System GenRad WARRANTY We warrant that this product Is tree from defects In material and workmanship and, when properly used. wm perform In accordance with GenRad's applicable published specttlcations , If WIthin one tt j year after orlglnalshlpment It Is found not to meet this standard, It will be repaired or at the option of GenRad. replaced at no charge when returned to a GenRad service facUlty. CHANGES iN THE PRODUCT NOT APPROVED BY GENRAD THIS WARRANTY SHALL VOID GENRAD SHAll NOT BE liABLE FOR ANY INDIRECT ,SPECIAL, OR CONSEQUENTIAL DAMAGES. EVEN IF NOTICE HAS BEEN GIVEN OF THE POSSIB1UTY OF SUCH DAMAGES THIS WARRANTY IS IN LIEU OF ALL OTHER WARRANTIES, EXPRESS OR IMPLIED, INCLUDING, BUT NOT LIMITED TO ANY IMPLIED WARRANTY OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE, SERVICE POLICY Your local GenRad office or representative will assist you In all matters relating to product rnamtenance. such as calibration, repair, replacement parts and service contracts GenRad policy IS to maintain product repair capability for a period of five 151years after the unit goes out of production and to make this capability available at the then prevailing schedule of charges. ix Introduction-Section 1 1.1 1.2 1.3 1.4 1-1 1-1 1-3 1-3 PURPOSE. DESCRIPTION CONTROLS, INDICATORS, AND CONNECTORS ACCESSORIES. 1.1 PURPOSE. The 1621 Capacitance-Measurement Svstern is designed for the precise measurement of capacitors and capacitance standards. In the standards laboratory, its high resolution for capacitance and conductance make th is system well suited for capacitance standards measurements. Its phaseindicating error meters facilitate rapid balancing. Convenient in-line readout maximizes accuracy of manual data recording and BCD outputs are provided for automatic data processing. The 1621 system measures either 3-terminal or 2-terminal capacitors. The transformer-ratio-arm circu itry of the bridge assures that 3-terminal measurements can be made accurately, even in the presence of large capacitances to ground. For instance, a ground capacitance of 1 pF produces an error of only 0.03% in the measurement of 1000-pF capacitor. This feature makes the assembly very useful for in situ measurements of ungrounded circuit capacitances. The 1616 Precision Capacitance Bridge, one of the 3 instruments in that system, may be obtained separately. The bridge will perform as described herein, if used with an oscilIator and detector equivalent to the G.R 1316 and 1238. NOTE This manual describes the 1621 system generally and provides its operating instructions. This manual also describes in detail the 1616 bridge only. A wide range of capacitances can be measured, extending from the resolution limit of 0.1 aF (10- 7 pF) to a maximum of 10 pf, with internal standards, or farther with external standards. For 3-terminal unknown capacitors, a pair of coaxial terminals is provided; for 2-terminal coaxial "unknowns," a single precision connector facilitates exacting control of fringing effects. Since an important use of this bridge is the comparison of capacitance standards, another pair of coaxial terminals is provided on the bridge to which a 3-terminal reference standard can be connected and designated EXTERNAL STANDARD. The other standard is then connected to the selected UN KNOWN terminals, and the internal standards are used to complete the balance. If the ratio between the two standards is close to 0.1, 1., or 10, the accuracy of the measurement is equal to the accuracy of calibration of the reference standard, and the precision of comparison is 1 part in 108 (0.01 ppm) of a 10-pF capacitor (or even better for larger ones). 1.2 DESCRIPTION. Figure 1-1. 1.2.1 General. The 1621 Precision Capacitance-Measurement System consists of the 1616 Precision Capacitance Bridge with the 1316 Oscillator and the 1238 Detector, a complete system for the precise measurement of capacitance. Oscillator and detector are mounted above the bridge, in a pedestal cabinet, as pictured in the front pages (and Figure 1-2); or the three instruments may be rack mounted. Connecting cables, supplied, go neatly behind the assembly. An elementary system diagram is given in Figure 1-1. 1.2.2 Bridge Circuit. The ratio arms of the bridge are transformer windings, tapped on the standard side in deci mal steps (-1, 0, 1, 2 ... 9, X) and on the unknown side in decade steps (X 100, Xl 0, Separate, fixed-capacitance standards are used, whose values range in decade steps from 1 aF to 100 nF. This combination of internal standards and transformer xn INTRODUCTION 1-1 ratios makes possible the wide measurement range of 1 to 10 14 Loss in the measured capacitor is expressed as parallel conductance from the resolution Ii mit of 0.1 f S to a maximum of 1 mS a measurement range of 1 to 10 1 3 The values of the set of 5 conductance standards are effectively extended by series resistance standards, in 6 decade steps (X 1 ... X 10-6 ) READOUT MULTIPLI ER UNKNOWN CAPACITOR o~LOW GR 1316 OSCILLATOR CAPACITANCE short-circuit loading. Auxiliary outputs are provided, both in-phase and quadrature, for detector references. These signals are comparable in quality with the main output. The level of each aux iIiary output signal is about 1.3 V rrns, driving the minimum recommended load impedance of 47 kSl. The phase separation between them is typically 87 to 90°, except below 50 Hz it may be a few degrees less. (F INE ADJUST controls on the 1238 Detector panel enable you to establ ish the desired quadrature phase in the detector.i Xl HIGH you are sure of frequency within ±1%. After warmup the frequency stability is typically within ±0.001% for a few minutes. Set the level as desired, up to 1.6 W into a wide range of load impedances (0.25 Sl to 2.5 kSl) with the help of the front-panel meter (reading 0.1 to 125 V) and you are sure of a pure and constant output signal. Its level varies less then ±2% with tuning; its distortion remains less than 0.4% over 3 decades of frequency and from open to With the synchronizing circuit you can conveniently lock this oscillator to a more stable source or provide sync to a scope, counter, or another oscillator. LEVER SWITCHES Figure 1-1. Elementary diagram, 1621 Precision Capacitance-Measurement System - GR 1616 bridge with 1316 Oscillator and 1238 Detector. Conductance circuitry is omitted. NOTE For more details about the 1316 Oscillator, refer to its instruction manual. 1.2.5 Detector. 1.2.3 Standards. Internal capacitance standards are of 4 types. The 3 most stable capacitors, made with Invar* steel, have temperature coefficients about 3 ppm per °c, and the next 3 smaller capacitors (also Invar), about 20 ppm per 0c. Their nitrogen dielectric is hermetically enclosed to assure independence from effects of changing atmospheric pressure and humidity. The 2 largest standards, of sealed, low-loss mica construction, and the 4 smallest, being open air-dielectric capacitors, have temperature coefficients about 20 ppm per "C. The set of 6 Invar and 2 mica standard capacitors are well insulated from the environment, the thermal time constant being 6 hours. So the bridge is remarkably insensitive to fluctuations of environmental temperature caused, for example, by a cycling air conditioner or the movement of personnel. The 1238 Detector, also developed for the 1621 system, complements the osci lIator and bridge with ccnven ience and sensitivity. Set the frequency dials to match those of the oscillator, or select the flat response characteristic. Set the gain as required; you can have full-scale readout for any bridge error from 70 nV (tuned response) to 400 mV (FLAT) - a range of 135 dB. Yet the instrument is immune from damage by signals as large as 200 V, at any gain setting Watch the in-phase and quadrature meters as you balance the bridge, they indicate conveniently whether to adjust C or G next, and whether to increase or decrease the weighting. (Convenient phase adjustments enable you to compensate for any phase shift (0-360°) through bridge, cables, and filter and to set the 2 phase-detector references exactly in quadrature.) 1.2.4 Oscillator. In addition to filter tuning, gain, and phase, front-panel controls select linear vs 20-dB-compressed response, rejection of power-line frequency components, and meter time constants from 0.1 to las. The 1316 Oscillator, developed for the 1621 system, isa conven ient, stable, powerfu I source. Set the 5 in-line decade frequency dials anywhere between 10 Hz and 100 kHz, and • Registered 1-2 INTRODUCTION trademark of the Carpenter Steel Co., Reading, Pa. 1.4 ACCESSORIES. NOTE For more detai Is about the 1238 Detector. refer to its instruction manual. 1.3 CONTROLS, INDICATORS, Table 1-3 lists the accessories supplied with the 1621 Precision Capacitance Measurement System. Power cords are supplied for the 2 instruments that use them. Rackmounting hardware is also supplied with the system, if it is the "rack" version. For mounting refer to paragraph 2.5. AND CONNECTORS. Figures 1-2 and 1-3 illustrate the instrument system, front and rear. Tables 1-1 and 1-2 further describe the individual controls, indicators, and connectors. 7 8 34 9 33-_ 10 32 31 30 29 28 27 16 2G 17 18 19 20 21 22 23 Figure 1-2. Front-panel controls, indicators, and connectors 1621 Precision Capacitance-Measurement System. of the INTRODUCTION 1-3 Table 1-1 -----------------FRONT-PANEL CONTROLS, INDICATORS, AND CONNECTORS Fig. 1·2 Item Name Description Function Oscillator POWER switch Toggle switch, down position OFF. Turns oscillator on and off. 2 FREQUENCY selector Set of 2 rotary switches with decimal steps, 0 ... 9, and 1 stepless pot with similar calibration and detent at O. Illuminated decimal points Selects and indicates frequency, a 3-digit number. Decimal-point illumination serves as pilot light. 3 FREQUENCY range switch Rotary switch with 4 positions Hz, Hz, kHz, kHz Selects a frequency range, indicates units, and controls illuminated decimal point in item 2. 4 OUTPUT VOLTAGE RANGE switch Rotary switch with 5 positions 15,50, 150,500, 150. Selects output-voltage range and indicates fullscale meter range (item 5) Simultaneously switches the output impedance from 0.25 S1 to 25 kS1 in decade steps. 5 Voltmeter Ac meter with 0-50 and 0-15 scales; has mechanical zeroadjustment screw. Indicates output terminal voltage, in ranges selected by itern 4. 6 OUTPUT ADJUST Step less rotary pot with limits labeled 0 and MAX Controls output level in the range selected by item 4. Detector, Right Side 7 TIME CONSTANT Rotary switch with 5 positions 0.1,0.3, 1, 3, 10 SECONDS. Controls the smoothing (integration) of detected signals and hence, effectively, the meter damping. 8 IN-PHASE meter Zero-center meter graduated 50-0-50; has mechanical zeroadjustment screw. Ind ication of one component of bridge error signal, such as the C component. 9 FINE ADJUST (IN-PHASE) Step less rotary pot. Adjusts phase of item-8 reference so the quadrature component is rejected (a fine adj ustrnent). 10 PHASE SHIFT (smaller knob) Rotary switch with 4 oositions: 0°, 90°, 180°, 270° Selects phase shift in coarse steps, supp lemented by items 9, 11, 12. 11 PHASE SHIFT (larger knob) Step lessrotary control, calibrated -50° to +50°. Shifts phase-detector references; set so that nearby meters (items 8, 13) respond independently to C and G error. 12 FINE ADJUST (QUADRATURE) Step less rotary pot. Adjusts phase of item -13 reference so the inphase component is rejected (a fine adjustment). 13 QUADRATURE Zero-center meter graduated 50-0-50; has mechanical zeroadjustment screw. Indication of one component of bridge error signal, such es the G component. 14 SENSITIVITY control Step less rotary pot. Fine gain control; use it to keep IN-PHASE and QUADRATURE meters reading on scale (does not affect MAGNITUDE meter) 1-4 INTRODUCTION Table 1-1 (Cont) ----------------FRONT-PANEL CONTROLS, INDICATORS, AND CONNECTORS Fig. 1-2 Item Name Description Function Bridge 15 Ground Socket for banana plug. Direct connection to master ground (electrical midpoint of ratio transformer) and chassis. 16 2-TERMINAL UNKNOWN port G900 precision coaxial connector. Connection for 2-terrninal "unknown" capacitor. Note Neither terminal may be connected directly to gnd (itern 15); outer = high; inner = low (F ig. 1-1). 17 3-TERMINAL UNKNOWf\i port Pair of G874 coaxial connectors, LOW and HIGH. Connection for 3-terminal "unknown" capacitor Note outer shields are tied to master gnd (item 15) 18 EXTERNAL STANDARD port Pair of G874 coaxial connectors, identified as HIGH and LOW. Connection for 3-terrninal external standard capacitor for special measurements, comparisons, or range extension. 19 ZERO ADJUST Step less lO-turn pot. Capacitance offset adjustment Range a few aF in CAL or 3-TERM positions of item 20; 3,3, and 50 pF (respectively) in 2-TE RM Xl, X 10, and X 100 positions 20 TERMINAL SELECTOR (READOUT MUL TIPLlER) Rotary switch with 6 positions 3 TERMINAL (XlO, Xl); CAL; 2 TERMINAL (X 100, X 10, X 1) Selects which UNKNOWN port (items 16, 17) connects to the bridge, or neither (CAL position); grounds the terminals of each port not so connected. Selects the READOUT MULTIPLIER apply it to both C and G. 21 EXT MUL TIPLIER switch Rotary switch with 13 positionsOFF,-Ol,O,O.l 10 Gives any external standard (at item 18) one of 12 weights (including zero) or disconnects that port from the bridge and grounds both terminals. 22 CONDUCT,L"NCE rnultiplier lever Lever switch with 7 positions identified as Xl .. X 10-6 Gives the conductance standards an add itional set of multipliers. 23 CONDUCTANCE standards levers Set of 5 lever switches, each 12-position with readout indicator -1,0,1 ... 9, X 1st digit, JLS next 3 digits, nS, Determines effective value of internal conductance standards, along with items 20, 22. Note at G balance, unknown (readout X CONDUCTANCE multiplier + ext std G X EXT MUL T) X READOUT MULTIPLIER. 24 CAPACITANCE standards levers Set of 12 lever switches, each 12 position with readout Determines effective value of internal capacitance standards, along with item 20 Note at C balance, unknown = (readout + ext std C X EXT MULT) X READOUT MULTIPLIER. 25 C MAX switch Lever switch with 4 positions Up, no effect, down, shutters over 1st 3 capacitance digits, cumulatively in 3 steps. Allows insertion or removal of first 1,2, or 3 (largest) standard capacitors from bridge Each is removed when a shutter covers its indicator. Note removal of large standards not used serves to reduce capacitive loading across detector (for best sensitivity), setting item 24 to /ero does not. 26 C-standards tri mrners Set of 12 screwdriver adjustments hidden behind a small panel; labels 100 nF .. 1 aF. To trim or adjust internal C standards if necessary. Labels indicate nominal (X 1) weight of corresponding standard when its switch is up (readout = X) indicator 1,0,1 .9, X 1st 2 digits, nF; then in blocks of 3pF,fF,aF. INTRODUCTION 1-5 Table ,-, (Cont] FRONT-PANEL CONTROLS, INDICATORS, AND CONNECTORS Fig. 1-2 Item 27 Name C-standards tri rn rners lock Function Description Secures the small panel over item 26 to preserve its Lock with keys. adjustments. Detector, Left Side 28 MAGNITUDE meter Meter, calibrated 0 to 100; has mechanical zero-adjust- Ind ication of bridge error signal level. ment screw GAIN, dB 29 COMPRESSION 30 Step attenuator; 12 positions, Coarse gain control; use it to keep MAGNITUDE 20 .. 130 dB meter reading on scale Push-button switch (push to engage, push again to release). Out linear response, full gain In compressed response, 20-dB-larger signal can be handled with meters on scale 31 LINE REJECTION Push-button switch (push to engage, push again to reo lease) Out normal. In 40-dB attenuation of line-frequency component in bridge error signal (Circuit can be adapted for 60 or 50 Hz) 32 Frequencyrange switcn Rotary switch with 5 positions FLAT, Hz, Hz, kHz, kHz. Selects broad-band characteristic or frequency range of tuned response, indicating the units and controlling the decimal point in item 34. 33 POWER switch Toggle switch, up ON; down OFF. Turns detector on and off 34 FREOUENCY selector Set of 3 rotary switches with decimal steps, O. 10. IIlumi nated decimal points Selects and indicates frequency to which detector is tuned (unless item 32 says F LA T). Decimal"[Joint illumination serves as pilot light Table '-2 REAR-PANEL CONTROLS AND CONNECTORS Fig. 1-3 Item Oscillator Name Description Function 1R QUADRATURE REFERENCE OUTPUT BNC jack * Provides a reference output, 90° leading the "in-phase" reference, at 1.3 V open circuit (connect to item 21R) 2R IN·PHASE REFERENCE OUTPUT BNC jack * Provides the other reference output, at the same level. Approx in phase with item 4R (connect '-6 INTRODUCTION to item 20R) Table 1-2 (contl----------------- ----------------- REAR PANEL CONTROLS AND CONNECTORS Fig. 1·3 Item Name Description Function 3R EXT SYNC BNC jack * Use for synchronization, if desired As an input, lock range is ±2%/V r ms, up to 10 V As an output. 0.3 V behind 27 kSL 4R POWER OUTPUT BNC jack * Main output, up to 16 W max, rnay be 125 V open circuit or 5A short circuit (connect to item 13R) 5R 8/10 AMP fuse Fuse in extractor-post holder Protection against damage from short circuit. lR 2H 3H 411 Gil lH 8H 10H 11 H 21R 2011 121l 19R nil 18R 141l 17R Figure 1-3. Hear-panel controls, connectors, and cables. Dashed lines indicate cable connections. INTRODUCTION 1-7 Table 1-2 (Contl----------------REAR-PANEL CONTROLS AND CONNECTORS ----------------Fig. 1-3 Item Description Name Function 6R Line-voltage switch Slide switch (labeled 50-60 Hz); 2 positions 100-125 V, 200-250 V. Accomodates power supply to either range of line voltages. 7R Power plug 3-pin power plug. t Connects from power line and earth ground. Detector, Left Side 8R 1/2 AMP fuse Fuse in extractor-post holder Protection against damage from short circuit. 9R Line-voltage switch Slide switch (labeled 50-60 Hz) 2 positions 100-125 V; 200-250 V. Accomodates power supply to either range of line voltages. lOR Power plug 3-pin power plug. t Connects from power line and earth ground. 11R DC-METER OUTPUTS 5-pin socket. Outputs for remote metering; all 3 meter circuits included. 12R INPUT SIGNAL BNC jack.* Main input to be detected (connect from item 14R). Bridge and Cables 13R GENERATOR INPUT BNC jack.* Input port for audio-frequency power to bridge circuitry. Connect from item 4R. 14R DETECTOR OUTPUT BNC jack." Output port for bridge error signal (unbalance). Connect to item 12R. 15R 0776-2020 BNC patch cord Interconnect items 4R, 13R.6. 16R 0776-2040 8161-5200 BNC patch cord BNC patch cord (red band) Connect items 1R to 21 R, 2R to 20R 6. Connect items 12R to 14R 6. 17R BCD CONDUCT ANCE OUTPUT 36-pin socket Indicates in BCD code to external instruments the CONDUCTANCE-readout and CONDUCT ANCEmultiplier values (items 22, 23) and the position of the TERMINAL SELECTOR (READOUT MUL TIPLlER) switch (item 20).6 18R BCD CAPACITANCE OUTPUT 50-pin socket Indicates similarly the CAPACITANCE readout, i.e.. the positions of the levers of items 24 and 25.6 Detector, Right Side 19R AMPLIFIER OUTPUT BNC jack* Output for remote instrumentation; ac voltage. 20R IN-PHASE REF INPUT BNC jack* Input reference for phase-sensitivedetectors; required level> I' V rms. 21R QUADRATURE REFERENCE INPUT BNC jack* Input like 18R except leading that by 900. ·BNC jack accepts Amphenol 6Refer to para. 2.8. t Refer to note, pg. 1-12. 1-8 INTRODUCTION "BNC" plug or military connector no. UG-88/U. -------------------Table 1·3 -----------------ACCESSORIES SUPPLIED WITH THE 1621 SYSTEM Name Power cord Description or Function 3-wire AWG number 18 type SVT cable, rated at 7A, 230 V. The connectors, designed for 125-V operation, conform to the Standard for Grounding Type Attachment Plug Caps and Receptacles, ANSI C73.11-1963. Length. 7 ft. 2 required (1 each for oscillator and detector). GR Catalog No. (Type) 4220-0220 Patch cords Shielded cable with BNC plugs; see para. 2.8; length 15 in. (2 req'd) length 24 in. (1 req'd) length 15 in, double shielded, red banded (1 req'd) 0776-2040 0776-2020 8161-5200 Plug To fit DC METER OUTPUTS socket; pins 5 (Amphenol 126-217) 4220-5401 Cap Plastic dust cover for G900 0900-7190 connector INTRODUCTION 1·9 Installation -Section2 2.1 2.2 2.3 2.4 2.5 2.6 2.7 2.8 GENERAL DIMENSIONS ENVI RONMENT BENCH MODELS RACK MODELS POWER-LINE CONNECTION LINE-VOLTAGE REGULATION SYSTEM CONNECTIONS 2-2 2-2 2-2 2-2 2-2 2-4 2-4 2-4 1616 1621 DIM A B C o 19.8 FRONT ., C 1 I 348.0 381.0 596.9 617.2 mm mm mm mm (13.7 (15.0 (23.5 (24.3 in.) in.) in.) in.] 246.6 327.7 330.2 350.5 mm mm mm mm (11.6 (12.9 (13.0 (13.8 in.I in.) in.l in.l BENCH MODEL Tn PANEL BRIDGE SYSTEM END VIEW D J Figure 2-1. Dimensions of the bench models. ~ CLEARANCE FOR REAR CABLES 1621 P~ 1I TOP VIEW ---------- U DIM P Q R 1 U 19 FRONT S T SYSTEM 368.3 292.1 332.7 40.6 576.6 mm mm mm mm mm (14.5 in.] (11.5 in.l 113.1 in.) (1.6 in.] (22.7 in.l 1616 BRIDGE 368.3 269.2 309.9 40.6 309.9 mm mm mm mm mm (14.5 in.) (10.6 ln.) (12.2 in.) (1.6 in.) (12.2 in.) RACK MODEL PANEL ~ END VIEW Figure 2-2. Dimensions of the rack models. INSTALLATION 2-1 2.1 GENERAL The 1621 Precision Capacitance Measurement System, or the 1616 Bridge alone, is available for either bench use or for installation in an EIA Standard RS-31O 19-in. relay rack with universal hole spacing. Appropriate cabinet and hardware sets are available for conversion of a bench model for rack installation or vice versa. Locate the instrument for convenience of operation and in a suitable environment. Avoid blocking the flow of air through the vents. Some open bench area, to the right or in front of the bridge, should be provided. NOTE If you assemble a 1621 System from separate instruments, either install them in the 4177-2621 cabinet or consult GenRad about the need for a magnetic shield between bridge and detector. If you convert a 1621 from bench to rack, transfer the magnetic shield. 2.2 DIMENSIONS. The dimensions of bench and relay-rack models of the system and of the bridge are given in Figures 2-1 and 2-2. 2.3 ENVIRONMENT. The system is designed to operate in standards laboratories, in which the environment is typically very well controlled. All specifications are valid over a temperature range of 22-24°C. Storage range is -20 to +70°C. 2.4 BENCH MODELS. b. Withdraw cabinet. each instrument out of the 2.4.2 Conversionfor Rack Mounting. To convert a bench instrument for rack mounting, exchange the cabinet and install appropriate hardware, as follows a. Obtain the appropriate Rack-Mounting Cabinets, as described in Table2-1, from GenRad. b. Obtain, optionally, a Bracket Set (Table 2-1) for each cabinet. Brackets are especially recommended for heavy instruments, which need support from the rear rail of the rack. --------Table 2-1--------- RACK-MOUNTING CABINETS AND BRACKETS FOR 1621 SYSTEM Quantity 1 1 1 3 Description Rack-mounting cabinet (for oscillator). Rack-mounting cabinet (for detector). Rack-mounting cabinet (for bridge). Sets of rear-support brackets and screws (1 for each instru ment). Part No. 4174-3240 4174-3624 4174-3627 4174-2007 c. Remove the cabinet, as in paragraph 2.4.1. d. Remove the rear cover from the bench cabinet, with screws (B, Figure 2-3), for later installation on the rack cabinet. e. Proceed with the rack installation, skip to paragraph 2.5.2, step b. Figure 2-3. 2.4.1 Cabinet Removal. 2.5 RACK MODELS. To remove the bench-model cabinet, first stand the system (or instrument) in the normal, horizontal position, free of all cables, and proceed as follows a. Remove the 4 dress-panel screws (A) accessible through holes in the handles of each instru ment. 2.5.1 General. Each rack model comes completely assembled in a suitable metal cabinet, which is designed to stay semi-permanently in a rack. Each instrument can be drawn forward on extending tracks for accesswith support, or (with a lift) Figure 2-3. Bench-cabinet installation. 2-2 INSTALLATION forward, o ~~~ TRACK / ~INSTRUMENT B A .17-5 Figure 2-4. Rack-cabinet installation. withdrawn completely. The cabinets listed in Table 2-1 are all included with a rack-model 1621 system, together with screws. Table 2-2 lists the screw sizes for reference. 2-2-------- ---------'Table KEY TO SCREW SIZES Ref Figs. Description 2-3; 2-4 A 8 C E No. Length, thds/in. inch Dress-panel screws with washers 10-32 Thread-cutting screws Thread-cutting screws Thread-forming screws 10-32 Figure 2-4. 2.5.2 Installation. Directions follow 10-32 8-32 .56 .50 .50 .19 for mounting the cabinet in a rack and installing the instrument on its tracks a, Remove 4 dress-panel screws (Al and slide the instrument out of the cabinet until the tracks are fully extended Continue pulling the instrument forward until motion along the tracks is stopped. At this juncture, tilt the front of the instrument up slightly and continue withdrawal, past the stoos. unti I it is free. b. Insert the rack cabinet wherever desired in the rack be sure it's level - and fasten it with 4 screws (C) to the front rails c. If the rack contains a rear support rail, use brackets (0) to support the cabi net with the rear rai Is; open-slotted screw holes allow positioning. d. Use the set of slots in the sides of the cabinet that allow alignment of the open-slotted holes in the brackets with threaded holes in the rail. The long flange should extend to the rear. e. Insert screws (E) from inside the cabinet, through the slotted holes and drive them into the holes in the long flange of the bracket. Each side takes 2. NOTE Start the screws in the appropriate holes off the rack, to make the threading easier. f. Pass screws (8) through brackets and screw 2 into each rear rail. (Details may be varied to suit particular situations. ) g. To install the instrument, first set its rear edge in the cabinet front open Ing. SIide the instrument back, mak ing sure that the rear and the upper front slide blocks engage the tracks. (Stops prevent further insertion.1 h. Pull the instrument forward with the tracks, keeping a hand on each side (fingers underneath). Slide the instrument back about Y, in. along both tracks, past the stops, by pressing down on the tracks (with thumbs) while tilting the front of the instrument up slightly. i. Push the instrument back into the rack, checking for smooth operation of the tracks and slide blocks. NOTE The instrument is now readi Iy accessible for behind-the-panel adjustments. It slides in and out freely on extendi ng tracks. 2.5.3 Conversion to Bench Use. To convert exchange the a.Obtain a 1621 system. from GenRad. a rack-mounting instrument for bench use. cabinet. as follows: for the Bench Cabinet. part no. 4177-2621 or 4172-4106 for the 1616 bridge alone. b. Remove the instrument from the rack cabinet, after removing the panel screws (A, Figure 2-4). (When free motion along the tracks is stopped, tilt the front of the instrument up slightly to clear the stops.) c. Slide the instrument into the bench cabinet. d. Fasten instrument to cabinet using dress-panel screws (A, Figure 2-3). e. Transfer the rear cover, with screws (8), from rack cabinet to bench cabinet. INSTALLATION 2-3 2.6 POWER-LINE CONNECTION. Power requirement for the 1621 system is 51 W at 100-to- 125 or 200- to- 250 V, 50- to-60 Hz Make connection as follows a. Set the line-voltage switches on the rear panels of oscillator and detector (Figure 1-3) to correspond with the available power-line voltage. Use a small screwdriver to slide the switch. b. Connect the external power line to each power plug using the power cords supplied or equivalent, 3-conductor cords (para. 1-4) The fuses should have the current ratings shown on the rear panels (Figure 1-3) regardless of which line-voltage range is chosen in step a. 2.7 LINE-VOLTAGE REGULATION. The accuracy of measurements accomplished with precision electronic test equipment operated from ac line sources can often be seriously degraded by fluctuations in primary input power l.inevoltaqe variations of ±15% are commonly encountered, even in laboratory environments. Although most modern electronic instruments incorporate some degree of regulation, possible power-source problems should be considered for every instrumentation setup. The use of line-voltage regulators between power lines and the test equipment is recommended as the only sure way to rule out the effects on measurernent data of variations in Iine voltage. 2.8 SYSTEM CONNECTIONS. Figure 1-3. 2.8.1 Oscillator, Bridge, and Detector. Make the 4 essential connections among the instruments, using the 4 BNC patch cords supplied with the 1621 system (refer to Table 1-3) as follows a. Test power to the bridge 1316 POWER OUTPUT to 1616 GENERATOR INPUT. b. Unbalance signal to the detector: 1616 DETECTOR OUTPUT to 1238 INPUT SIGNAL. (Red-banded cable). c. Reference siganls to the detector: 1316 REFERENCE OUTPUTS to 1238 REFERENCE INPUTS (one cable for IN PHASE, one for QUADRATURE). punch coupler, or comparator Use a 50-pin plug, Amphenol PIN 57-30500 (or equivalent) and cable such as Alpha No. 1181/50, which has AWG No. 22 stranded wires.* For pin identification, refer to Figure 2-5. Notice that the output data is provided by SWitch closures only. The switches in the bridge are rated for up to 0.5 A at 110 V (ac) with resistance loads. If, however. you want for example a logical "1" (the on state of each data bit) to be represented by +5 V with respect to the system ground (pin 30) then use an external +5-V power supply and connect it from ground to VREF (pin 25). Logical "0" is an open circuit. The capacitance readout is available in binary-coded decimal form, the code being "1-2-4-8" as detailed in Table 2-3. The body of the table contains only binary numbers, composed of bits 0 and 1. Notice the extension of the usual BCD table, to include X (ten) and negative 1. As an example, suppose the bridge readout is 396 pF. Figure 2-5 shows us that pins 20, 19,43, 17, 16 and 40 are "1", while pins 45, 44, 18, 42, 41, and 15 are "0". This kind of data is commonly accepted by printers. However, suppose the bridge readout is 4(-1) 6 pF the same capacitance I Figure 2-5 and Table 2-3 tell us that pins 44,43,18,17,16 and 40 are "1" while pins 20,19, 45,42,41 and 15 are "0". A sufficiently sophisticated system will make a computation and print 396 pF (for example). A simple system utilizes a 12-character printer (including -1 and X as well as the usual 0 . 9) and will print out just what the bridge readout shows. If the system must use a printer that does not recognize -1 and X a logic circuit may be fabricated, to detect the occurrence of either -1 or X anywhere in the readout, and trigger an alarm such as a buzzer or a change in color of the printout (Table 2-3 shows that when both 2-weight and 8-weight binary signals of anyone decimal digit are" 1", there is such an occurrence. In our example, the pins involved are numbered 18 and 43.) Either provide the system complexity needed to handle data containing -1 and X without ambiguity, or make sure that the operator removes them from his final balance adjustment. Operating instructions in this manual are written for the latter case. NOTE The capacitance data must be scaled up by 1, 10, or 100 (the READOUT MULTIPLIER) code for which is given below. 2.8.2 BCD-Capacitance-Output Connector. Figures 1-3,2-5. Make connections from the BCD CAPACIT ANCE OUT PUT socket at the rear of the bridge, if you want to record or process the C-measurement data with a printer, card- 2-4 INSTALLATION 'Alpha Wire Corp., Elizabeth, N.J. pF 20 IF 2 200 The READOUT MULTIPLIER is one of 3 values, Xl, X 10, X 100. However, the BCD data has 6 possible values, corresponding to the 6 positions of the TE RM INA L SELECTOR switch, as shown in Table 2-4. For example: a binary 101 (decimal 5) means the READOUT MUL TlPLIER is X10 and (incidentally) the measurement is being made via the 2- TERMINAL port. OF 20 40 80 8 800 ~~~~ pF IF of 161618 Figure 2-5. Terminal identification at BCD CAPACITANCE OUTPUT connector (A.J13, rear panel of bridge). Arrow toward connector = input; away = output. These weights apply when the TERMINAL SELECTOR switch is set to READOUT MULTIpLER = Xl (either of 2 settings). This is a rear view, exterior of socket. The BCD output is zero for any internal standard that has been removed from the bridge by the C MAX switch, regardless of the position of the lever switch normally controlling that standard. ------Table If your printer has decimal points, drive them from the READOUT MULTIPLIER as follows (for printout in pF) pins 33, 34, 35 drive the points following columns 5,6,7, respectively Printing of a 2nd point indicates 2 terminal Xl or X 10; all 3 points indicate CALIBRATION -------- Table 2-4-------READOUT-MULTIPLIER 4 0 0 1 1 1 1 Decimal (Equiv.) Weight 2 0 1 1 1 0 1 1 2 7 0 0 1 0 1 0 4 5 6 CODE TERMINAL-SELECTOR Switch Position XlO Xl 3-term ina I 3-term inal CAL 2-terrninal X 100 2-terminal XlO 2-terminal Xl 2-3----- BINARY-CODED DECIMALS (BCD) Signal Weight (Binary) 8 4 2 1 Digit (Decimal) 1 0 0 0 0 0 1 0 0 1 0 1 -1 0 1 0 0 0 0 0 1 1 1 0 0 1 0 2 3 4 0 0 0 1 1 1 0 1 1 1 0 1 5 6 7 1 1 1 0 0 0 0 0 1 0 1 0 9 8 X, 2.8.3 BCD-Conductance-Output Connector.Figures 1-3,2-6. Make connection, similarly, from the BCD CONDUCTANCE OUTPUT socket, using a 36-pin plug, Amphenol PIN 57-30360 (or equivalent) For pin identification refer to Figure 2-6. In addition to the basic conductance-readout data, here is multiplier data. The conductance multiplier is one of 7 values, 1 10-6 Its exponent (magnitude only, expressed in BCD code) appears at pins 13, 14, and 30. For example: these 3 pins at "0" state means zero exponent, i.e. the G multiplier is 1. Figure 2-6. Terminal identification at BCD CONDUCTANCE OUTPUT connector (A-J12, rear panel of bridge). Arrow toward connector = input; away = output. The indicated conductance weights apply directly when the effective multiplier is 1, for example: CONDUCTANCE multiplier = 10'2 and READOUT MUL TlPLlER = Xl00. This is a rear view, exterior of socket. 2.8.4 Analog Outputs. Figures 1-3,2-7. AMPLIFIER OUTPUT. Use a BNC patch cord to connect this signal to remote monitoring or recording equipment if desired. This ac signal is proportional to the MAGNITUDE meter deflection, and is 4 V rms at full scale. DC METER OUTPUTS Use the 5-pin plug supplied (see Table 1-3) and cable suited to your system, if you wish to have remote indication of the 1238 front-panel meter INSTALLATION 2-5 deflections. A suitable cable is Alpha No. 1175, which contains AWG No. 22 wires. Pins are designated as shown by Figure 2-7 Only pin H of these circuits may be grounded. A is +, B is - for the MAGNITUDE meter circuit (6 V corresponds to full-scale deflection) D is +, H is - for the IN PHASE meter circu it; E is +, H is - for the QUADRATURE meter circuit. For each of these 2 circuits, the level is 1 V for full-scale deflection of the corresponding front-panel meter, when the SENSITIVITY control is ccw (minimum) However, 2-6 INSTALLATION because that control affects the front-panel meters, not the DC METER OUTPUT voltage, the voltage is relatively lower when that control is cv«. .m _ ~-~ ::1 0-- e• - ._- A +} MAGNITUDE B- D + IN PHASE QUADRATURE Figure 2-7. DC METER OUTPUTS Detector, exterior (rear) view. socket on 1238 Operation - SectionJ 3.1 PRELIMINARY CHECKS 3.2 FUNCTIONAL SELF-CHECK 3.3 PHASE ADJUSTMENT . . 3.4 CONNECTION OF UNKNOWN CAPACITOR. 3.5 BALANCE AND READOUT . . . . . . 3.6 PARAMETERS OF THE UNKNOWN CAPACITOR 3.7 FREQUENCY . . 3.8 VOLTAGE LEVEL. 3.9 ACCURACY. . . 3.10 PRECISION 3.11 EXTERNAL STANDARDS 3.12 PRECISE COMPARISONS. 3.13 NON-COAXIAL 2-TERMINAL CAPACITORS 3.14DCBIAS. . . . . . . . . 3.15 REVERSED CONFIGURATION. . . 3-1 3-1 3-2 3-3 3-4 3-7 3-8 3-10 3-10 3-14 3-15 3-17 3-17 3-20 3-22 & WARNING Dangerousvoltagesmay be presentat the terminals of this instrument. Refer to specificwarningscontained in this section. NOTE The following instructions apply literally to the 1621 system. If your 1616 bridge is connected to an oscillator and/or detector other than the GR 1316 and 1238, interpret the corresponding parts of this section appropriately. & CAUTION Do not connect a power cord until each linevoltageswitch has been set properly. 3.1 PRELIMINARY CHECKS. Refer to paragraph 1.3 for figures illustrating the controls, indicators, and connectors and for tabulation of their functional descriptions. The recommended initial operating procedure follows: a. Check that the line-voltage switches on the rear panels of oscillator and detector are positioned according to the available power-line voltage (either 100-125 Vor 200-250 V). To slide these switches, use the tip of a small screwdriver b. The LINE REJECTION filter in the 1238 Detector has been set for 50 or 60 Hz by GR (it can be purchased either way) or by a user. If you need to verify that it matches your power-line frequency (or reset it), refer to the 1238 Instruction Manual. c. Check that the 4 BNC patch cords are in place at the rear, connecting oscillator, detector, and bridge as described in paragraph 2.8. d. Check that 4 meters read zero (2 of them at midscale). If necessary, adjust each with a small screwdriver, at the recessedscrew just below the meter. e. Set the OUTPUT VOLTAGE RANGE to 15 V. (Higher voltages should be used only when they are needed, usually to facilitate high precision in measurements.) & WARNING Beware of hazardous voltages: ~ 350 V at 3-TERM UNKNOWN HI or EXT STD HI; ~ 35 V at 2-TERM UNKNOWN HI (outer shell), while oscillator is set to a high level. f. Connect the power plugs (rear panel) to a suitable power line, using the power cords supplied. Flip the POWER switch UP (front panel) on each, oscillator and detector. Verify that a decimal point is illuminated in each FREQUENCY selector. 3.2 FUNCTIONAL SELF-CHECK. Adjust oscillator and detector to the same frequency and make a balance to check for system operation, as follows: OPERATION 3-1 a. Set the front-panel controls as listed (from upper left to lower right, Figure 1-2) FREQUENCY 101 kHz (Oscillator) OUTPUT VOLTAGE RANGE - 150 OUTPUT ADJUST - MAX FREQUENCY - 101 kHz (Detector) TIME CONSTANT 1s LINE REJEr:TION - push button out COMPRESSION - push button out GAIN - 30 dB SENSITIVITY ccw (minimum) PHASE SHI FT - 180° a FINE ADJUST - rnidrange (both controls) EXT MULTIPLIER - OFF (Bridge) TERMINAL SE LEC, UR - CAL ZERO ADJUST -- ccw C MAX - down (Asterisks below represent 3 closed shutters) CAPACITANCE * * nF *05 pF 000 fF 0000 aF CONDUCTANCF a fLS 0000 nS X 1(f6 b Reset the GAIN If necessary to make the MAGNITUDE meter read near midscale (20 to 80) Tune the detector to match the oscillator, as indicated by a peak reading of the MAGNITUDE meter, to the nearest step Fine tune with the oscillator 3rd dial d Reset the CAPACIT ANCE to ** nF HOO pF 000 fF 0000 aF Verify that MAGNITUDE drops to zero e. Reset the GAIN to 100 dB. Observe that a noticeable unbalance occurs when the 1 fF lever is moved to 1 or -1 QUADRATURE meter reads 25 or more (in either direc tion) and both read on-scale c. With the PHASE SH I FT controls, bring the QUADRATURE meter approximately to zero. If the range of the larger knob is not sufficient, turn the smaller knob to an adjacent position. d. Now the I N PHASE meter should read upscale, to the right (because C readout is greater than C unknown) Reverse the direction of the meter deflection, if necessary, by turning the smaller PHASE SHIFT knob 2 clicks e See that both FIN E ADJUST knobs are set to amid position (dot L) Increase the C unbalance until the MAG· NITUDE meter reads 8099. Set SEf'>JSITIVITY to max, cwo Set the PHASE SHIFT for zero on the QUADRATURE meter. If zero cannot be obtained wi th the PHASE SH 1FT, use the QUADRATURE FINE ADJUST.* t Reestablish the null reading and refine it, If necessary, uSing the C and G levers The MAGNITUDE reading should not exceed 2, or % division g Unbalance the bridge by raisinq CONDUCTANCE lever switches, achievinq about the same magnitude of unbalance as before 80-99 (Try usrnq the 1- fLS lever) Now the QUADRATURE meter should read upscale, to the right (G readout is greater than G unknown) Trim the IN PHASE FINE ADJUST onlv to make the adjacent meter read zero, or a uiinu num (Do not expect zero at all possib Ie G unbalances at high !req ) h. If, In step g, the meter reads down scale, interchange the IN PHASE and QUADRATURE connections at the REFERENCE INPUTS behind the 1238 Detector Then repeat steps a through g, above. 3.3 PHASE ADJUSTMENT. In order to enjov the convenience of the phase-sensitive indicators, adjust the phase shifters as described below. The adjustment is then useful as long as the frequency is unchanged, the detector FREQUENCY remains tuned for maximum response, and the CONDUCTANCE multiplier is unchanged Readj ustrnent of phase shitt may be made while the bridge IS berng balanced, refer to paragraph 35 NOTE The goal IS to minimize the response of one phase sensitive meter to changes in C, the other to changes in G The following procedure, though not ideal for every measurement condi tion, ISadequate in general a Reestablish the null reading Oil the iv'lAGNITUDE meter, as in para 32 However, set the CONDUCTANCE multiplier so its exponent IS 5, --4, -3, 2, for a fre quency of a 1, 1, 10, or 100 k Hz . respectively Then switch GA IN to 60 dB and unbalance the bridge by raising CA PACITANCE lever switches only, making the MAGNITUDE meter read about 20 (For the first attempt, use the 10 fF lever) b. With the SENSITIVITY control, set the signal level in the phase-sensitive detectors so that either I N PHASE or 3-2 OPERATION --------- Table 3-1 SENSES OF PHASE-SENSITIVE REFERENCE connections METERS PHASE-SHI FT controls set to Meter 0° 90° 180 0 270 0 Normal IN PHASE C+- G+- C-+ G-+ Normal QUADRATURE G+- C-+ G-+ C+- Crossed IN PHASE G*'- C+- G-+ C-+ Crossed QUADFiATUFiF C+- (;-+ C-+ G+- We have assumed that it is preferable for the IN PHASE meter to respond upscale (to the right) for C, > C x' and the QUADRATURE meter, similarly, for Gs G x However, if you prefer to associate "I N PHASE" with "conductance", interchange the REF ERE NCE connections at the rear panel (cross the patch cords) If you prefer the opposite sense of response, set PHASE SHIFT near 0° instead of 180° Table 3- 1 shows the 8 possible combinations. > • Phase adjustment is now f and 9 are refinements. adequate for most measurements. Steps 3.4 CONNECTION 3.4.1 OF UNKNOWN Three-terminal Capacitors. CAPACITOR 3-TERMINAL, SHIELDED UNKNOWN CAPACITOR Figure 3-1. As the simplified daigram shows, the 3-terminal measurement evaluates the direct capacitance C (and conductance) between HIGH and LOW bridge ter?ninals. Shielded cables or patch cords (G874-422A or 874-R22LA) ma be used to connect these terminals to a capacitances to ground (C ' C ' including cable capacitances) are excluded . Is hs f rom the measurement. but If large enough they may affect accuracy. Enclosure. Check that the capacitor to be measured is enclosed in a shield, for connection to bridge ground. Not only is the shield necessary for precision measurements, but it must be mechanically fixed (permanently or repeatably) for Cx to be defined. Physically, as the shield gets larger and farther from the plates of Cx, the value of Cx becomes larger and less dependent on spacing of the shield. However, that dependency never vanishes. Use the 3-terminal component mount G874-X, or (if that is too small) a metal box, if an enclosure must be supplied -, Connectors. Check that the "unknown" capacitor has two coaxial connectors for convenient connection to the bridge. If the capacitor is relatively small and light weight and is equipped with G874 connectors spaced 1.25 in. apart (center-to-center), patch cords are not necessary. The capacitor can then be plugged directly into the bridge. However, shielded patch cords are convenient. their use allows the capacitor being measured to sit on a bench surface or at a remote location. The parameters of such cables are negligible in a 3-terminal measurement, except for noise introduced into the detector when they are moved. Provide connectors if necessary, such as those listed in Table 1-5. Connect the capacitor to the 3-TERMINAL UNKNOWN ports. Reversal. The measured direct capacitance is not changed in most three-terminal measurements when the connections to the capacitor are reversed, i.e.. bridge HIGH to capacitor L instead of bridge HIGH to capacitor H Shielding. Keep both HIGH and LOW connections shielded as a general rule. At least one must be shielded or the capacitance between connecting terminals and wires becomes part of the measurement. The HIGH connection may be made without shielding, If you prefer. This terminal, being at the high-voltage but low-i mpedance output of the transformer, is not sensitive to pickup from external sources and seldom needs to be shielded. There is voltage from the HIGH terminal to ground. Much capacitance or conductance connected from this terminal to ground reduces the transformer output voltage and introduces possible errors into the measured direct capacitance and conductance. In general, keep such shunt C below 200 pF (see para. 3.9) Keep the LOW connection shielded. This terminal, being at the low-voltage but high-impedance input to the detector, is very sensitive to noise and signal pickup from ex- LOW .... OSCILLATOR LOW Figure 3-1. Simplified bridge diagram - 3-terminal connection of unknown capacitor. ternal sources and must be completely shielded for lowcapacitance or any precision measurements. There is no voltage from LOW to ground when the bridge is balanced. Much capacitance or conductance from the LOW terminal to ground shunts the detector, and by reducing sensitivity limits the precision of your measurements. In general, keep such shunt C below 200 pF. Capacitance of the recommended R22A (or 874-R22LA) is 90 pF. patch cord, 3.4.2 Two-terminal Coaxial Capacitors. G874- Figure 3-2. As the diagram shows, a 2-terminal measurement evaluates the capacitance Cx (and conductance) between the inner conductor and the shell of a coax ial structure. The inner conductor connects to LOW bridge terminal (which must be shielded for precision measurements), the shell to HIGH. Although the shell serves as a shield, do not connect it to ground. Capacitance to ground Ch g is excluded from the measurement. & WARNING Beware of possibly hazardous voltage on shell of "unknown" capacitor whenever generator is set to a high level. If necessary. use of a coaxial adaptor (or special test fixture) between the single G900 connector on the bridge panel and the capacitor being tested. The capacitance of any such adaptor is included in the measurement. However, the ZERO ADJUST control can be used to compensate the bridge - up to 3 pF - making the CAPACITANCE READOUT directreading for any unknown capacitor connected to the adaptor, as described below. OPERATION 3-3 Gx ex 2-TERMINAL COAXIAL UNKNOWNCAPACITOR LOW L-----11I OSCILLATOR LOW Figure 3-2. Simplified bridge diagram - 2-terminal connection of unknown capacitor. NOTE Non-coaxial 2-terminal capacitors can also be measured; refer to para. 3.13. Enclosure. Check that the capacitor to be measured is so constructed that the shell shields the inner terminal. and that both are isolated from ground. Use the 2-terminal component mount G874-ML if an enclosure must be used. Connector. Check that the capacitor has G900 or G890 connector for convenient connection to the bridge. If not, provide the latter type. or whatever fits your special test fixture (see below]. Zero Adjustment. Compensate the bridge for capacitance of its 2-terminal connector and - if one is used - the attached short coaxial line, as follows: a. Check that the front-panel controls are set as of the completion of functional checks and phase adjustment, paragraphs 3.2 and 3.3. (Frequency is optional; see para. 3.7.) b. Set the TERMINAL SELECTOR to 2 TERMINAL Xl, X 10, or X 100 for measurements up to 1 nF, 10 nF, or 10 pF, respectively. c. Set CAPACITANCE lever switches to the value of the fringing capacitance of the connector that will receive the unknown capacitor. In the simple case that the READOUT M U LTIP LIE R is Xl, no adaptor is needed, and accuracy of ±8 fF is sufficient, select CAPACITANCE and CONDUCTANCE readouts as follows ** nF *00 pF 115 fF 000.0 aF and 0 fJ-S000.0 nS X 1(J6 If the READOUT MULTIPLIER was set to XlO or X 100, use proportionally smaller settings (11 fF 500 aF, or 1 fF 150 aF, respectively). For greater accuracy, refer to para 3.9.5. d. Adjust the ZERO ADJUST control for minimum reading on the IN PHASE meter. Turn the GAIN control to 3-4 OPERATION 90 dB, or as far as is appropriate to get a good indication for this adjustment. The adjustment is now valid for a particular terminal capacitance and READOUT MULTIPLIER. If you change either of these (and leave the TERMINAL SELECTOR on one of the three 2-TERMINAL positions) repeat the zero adjustment. e. Connect the capacitor to the 2-TERMINAL UNKNOWN port. Special Test Fixtures. The details of special fixtures are beyond the scope of this manual. However, these comments apply: 1. The ZERO ADJUST range is sufficient to compensate up to about 3 pF with READOUT MULTIPLIER set at Xl or Xl0, 56 pF at Xl00. Therefore, bridge readings will generally need correction (by subtraction) unless you provide an external capacitor for test-fixture compensation. Refer to para. 3.11. 2. Highly precise measurements are possible; refer to "direct substitution", para. 3.12. 3. Mount the test fixture directly on the 2-TERMINAL UN KNOWN port or use rigid coaxial lines, elbows, etc. (Flexible cables change capacitance with position. If you use them, your precision is only about ±0.1 pF.) 4. Support the fixture, if it's on a long, rigid line, to avoid damage to the panel-mounted connector. Lock each connector so it will not shift position. 5. Consider also the possibility of connecting your fixture to the 3-TERMINAL UNKNOWN port; thereby tak ing advantage of the fact that various cable and ground capacitances are excluded from the measureruent, NOTE Refer also to NON-COAXIAL CAPACITORS, para. 3.13. 3.5 BALANCE AND READOUT. 2-TERMINAL Figure3-3. This is the actual measurement process. Preceding paragraphs explain how to tune oscillator and detector, make phase adjustments, and connect the capacitor to be measured. Balance the bridge, as follows, and you have the measurement in both visual and digital forms. The standards for measurement are contained in the bridge. 3.5.1 Readout Multiplier. Set the TERMINAL SELECTOR to the appropriate Xl position (2 or 3 terminal) unless the capacitor being measured is larger than 1 nF (1000 pF). Then set the selector to the highest READOUT MULTIPLIER (Xl0for3-terminal, Xl00 for 2-terminal measurements). The purpose is to utilize one of the three "pf " lever switches for the most significant digit, if possible, thus maximizing the accuracy of your measurement. Cx ~ EXT MULTIPLIER Gx -z- .2 UNKNOWN TERMINAL SELECTOR READOUT MULTIPLIER -zr 0-+- .40-+ BRIDGE, GR 1616 Figure 3-3. The 1621 Capacitance-Measurement System - simplified diagram. *Note: Effective values are shown for all G standards and for C standards smaller than 1 pF. 3.5.2 Initial Settingsof lever Switches. If the capacitor value is known, even approximately, set that value into the CAPACITANCE readout, and set the doubtful digits at the right to 5's. For example, if the capacitor is about 12 pF, set the bridge to read 12 pF 555 fF 555.5 aF. C MAX. Move the C MAX lever down, covering as many as 3 of the unused digits to the left of the desired readout. Covering as many as possible of the zeroes at the left of the final readout helps to maximize the precision of your measurement, by removing stray capacitance which otherwise loads the ratio transformers slightly. CONDUCTANCE. For a typical measurement, set the CONDUCTANCE levers initially to the zero readout J.LS 0000 nS X 10-6 . However, if conductance is known approximately, set the levers to that approximation. Unknown, A Black Box. If the capacitance or conductance settings cannot be set initially to the right order of magnitude, the MAGNITUDE meter will read off-scale. Temporarily, use the following aids to obtain a preliminary a balance a. Push the COMPRESSION button in. b. Turn OUTPUT VOLTAGE RANGE and OUTPUT ADJUST ccw as may be required to bring the MAGNITUDE meter on-scale. c. Explore large ranges of C and G lever-switch settings, leaving most of the levers at zero so that the ones being moved are always most-significant digits. d. Watch the phase-sensitive meters (even though they may be pinned and not properly phase related). A reversal of sense (pointer crossing midscale) is an indication of passingnear a rough balance. 3.5.3 BalanceProcedures. Figures1-2, 3-3. GAIN. Set the GAIN control initially so the MAGNITUDE meter reads in the upper half scale. Later, as balance is improved, turn the control cw to maintain useful readings on the other 2 meters. SENSITIVITY. Set the SENSITIVITY control initially at midrange. Later, as balance is improved, adjust for a large on-scale reading on one or both of the IN PHASE and QUADRATURE meters. OPERATION 3-5 NOTE Keep the MAGNITUDE meter on scale, otherwise the phase sensitive detectors may be over loaded. However, the IN-PHASE and QUADRA TU RE meters may be pinned without loss of phase sense indication. TIME CONSTANT. Set the TIME CONSTANT initially to 0.1 s. for immediate response to your balancing operation. Later, as balance is improved and the GA IN control has to be set quite high, turn to larger time constants. A noisy reading (meter pointer jumping) can be expected at high gain if the time constant is too low. So increase it to calm jittery IN-PHASE and QUADRA TU RE meters, allow time enough for each reading, and you can achieve a highly precise balance. Lever Switches. Decide which levers to adjust first by looking at the PHASE-SENSITIVE DETECTOR meters If the IN PHASE meter reads farther from zero, adjust CAPACITANCE first if QUADRATURE, adjust CONDUCTANCE. CAPACITANCE balance. Refine the balance as follows, left-to-right a Start with the most significant digit that is in doubt. Preferably, every CAPACITANCE digit to the right of it should be set to 5. b. Push the lever down if the IN PHASE meter reads up scale; pull up if it reads down scale. Continue until the meter pointer crosses zero. c. Leave that lever in the position closest to balance, i.e.. making the IN PHASE meter point closest to zero, on either side. d. Adjust the next lever to the right, similarly, if there is sufficient meter deflection to be a guide. (I f the suggestion in step a is used, you will probably not have to readjust the lever you set in step c.) e. Increase the GAIN as much as necessary to provide that deflection, but not enough to deflect the QUADRATURE meter off scale. f. Continue refining the capacitive balance until the QUADRATURE indication becomes a limitation, then refine the CON DUCT ANCE balance. CONDUCTANCE balance. If an approximately correct conductance setting has been initiated, refine the conductance balance from left to right, similarly to the capacitance balance. Minimize the reading of the QUADRATURE meter. However, if the CONDUCT ANCE levers have been set to zero, as described above, proceed as follows a. Raise the CONDUCTANCE levers from 0 to 5, starting at the left of the multiplier and continuing left until either the QUADRATURE meter pointer crosses zero (if so, skip to step c) or all these levers have been raised. 3-6 OPERATION If the meter points to the right, drop the significant lever to -1, so the readout is -11J.S 555.5 1(f6 , and skip to step c. If the meter points to the left or zero, drop the significant lever to + 1, so the readout is 1 1J.S555.5 mostnS X mostnS X 1(f6. b. Raise the conductance multiplier lever until the meter crosses zero and stop at the setting nearest balance. l f this lever is set to X1 or x io '. PHASE SHIFT will have to be reset, as described below. c. Refine the conductance balance as outlined above (for capacitance) starting with the left-most (1J.Sj lever and proceeding to the right. 3.5.4 Final Balance. Alternate between CONDUCTANCE and CAPACIT ANCE balances, as described above, improving whichever is worse until the other becomes a limitation. Conclusion. Stop when the desired precision has been achieved. The bridge is probably capable of greater resolution than you need. Voltage. If there seems to be a gain or noise limitation, even though GAIN is 130 dB and TIME CONSTANT is 10 s, raise the oscillator OUTPUT VOLTAGE However, do not exceed the level of 350 V rms. If the frequency is less than 2200 Hz, do not exceed 0.16 f volts, where f is frequency in Hz. & WARNING While generator is set to a high level, beware of hazardous voltages: ~ 350 V across 3-TERM UNKNOWN or EXT STD; ~ 35 V across 2-TERM UNKNOWN (outer shell high). Phase Shift / Fine Adjust. While making a precise measurement of a low-loss capacitor, check as follows a. As the 4th significant digit of C is being selected, verify that the QUADRATUR E meter responds negligibly (compared to the IN-PHASE meter) for Clever changes b. Correct the phase if necessary with the QUADRA TURE FINE ADJUST control c. When balancing conductance, make the corresponding check (and adjustment) of the other phase. NOTE A slight change in tuning of either oscillator or detector will affect phase, even though the change has negligible effect on magnitude of response. Phase Shift / Reset. If the capacitor is lossy, the phaseshift settings of paragraph 3.3 are unsuitable. Reset as follows a. Make an initial balance, watching the MAGNITUDE meter, until the 4th significant digit of C or G has been chosen (whichever predominates). b. Temporarily turn SENSITIVITY and GAIN controls ccw enough to bring both phase-sensitive meters within % division of zero (scale value of 2.5). Set both FINE ADJUST controls to mid range (pointer up) c. Unbalance the bridge by raising C levers (or G if it predominates) enough to make large deflections on these meters. d. If C predominates, adjust PHASE SHIFT for zero QUADRA TU RE and upscale IN-PHASE meter readings. Raise the GAIN and SENSITIVITY settings as is appropriate; refine the phase with the QUADRATURE FINE ADJUST control (as described under Phase Shift / Fine Adjust). If G predominates, adjust similarly, for zero IN-PHASE and upscale QUADRA TU RE meter readings. e. Return to the initial-balance condition, then unbalance G (or C) The meter that read upscale in step d should not deflect. Trim the nearest FINE ADJUST control for this condition. NOTE Be sure the EXT MULTIPLIER switch is OFF (not simply zero) whenever the EXTERNAL STANDARD terminals are unused. Even though those terminals are open, their stray capacitance may introduce error into the measurement of capacitors below 1 fF and the unshielded LOW terminal may pick up enough "noise" to have a detrimental effect on precision. NOTE Keep the MAGNITUDE meter reading on scale. If the MAGNITUDE meter is pinned, distortion in the detector may cause a spurious phase shift. However, the IN PHASE and QUADRATURE meters may point off scale without loss of sense information. output data (rear panel) that data is now in standard form. Refer to para. 2.8. NOTE There may be occasions when it is desirable or necessary to leave X in the readout. If, for example, the bridge balances at ** nF X23 pF 456 fF 785.5 aF (the last 2 or 3 digits may be insignificant! you have this choice For greatest accuracy and a record as to which internal standard is significant, leave the X in the readout Take the responsibility for making sure the recorded data is unambiguous (particularly if a printer is being driven from the BCD CAPACITANCE OUTPUT) For a simpler readout, move the C MAX lever and correct the readout to *1 nF 023 pF 456 fF 785.5 aF. But then the bridge is almost certainly unbalanced; if you now refine the balance you are almost certainly reducing accuracy by depending on a larger, lessaccurate internal standard 3.5.6 Units of Measurement. The units of each, Cx and Gx, appear on the front panel; but be sure to apply the multipliers. For example, (Figure 1-2) if the readout is X98 nF and the READOUT MULTI PLiER indicates XlO, C; is 1098 nF or 1098 fJ.F The conductance readout has 2 multipliers. For example, if the readout is 5.43 p..SX 10-6 and the READOUT MULTIPLIER indicates Xl00, c, is 5.43 X 10-4 us or 5.43 X 10- 1 0 S NOTE The symbol "S'" represents the Siemens unit, the unit of conductance (or admittance). Siemens is equivalent to mho (the reciprocal of ohm). represented by the symbol U. 3.6 PARAMETERS OF THE UNKNOWN CAPACITOR. 3.5.5 Readout Correction. Figure 3-3. If there is a -1 or X in the CAPACITANCE or CONDUCT ANCE numbers displayed above the lever switches, it is generally recommended to correct the readout as follows a. Proceed from left to right unless X is adjacent to -1, in either order. (Correct the right-hand digit first in any such pair.) b. Change each -1 to 9 and decrease the preceding digit byone. c. Change each X to 0 and increase the preceding digit by one. d. Verify that the final balance is still valid. e. The readout is now correct for manual recording. Also, if your measurement system makes use of the BCD 3.6.1 SeriesEquivalent Parameters. Figure 3-4. This bridge always measures directly the parallel equivalent parameters Cx and G x ' even in case the loss component of the unknown capacitor is in fact entirely in series (as resistance of the connecting leads). Whether a series R-C, a parallel R-C, or a more elaborate equivalent circuit best describes the capacitor you are measuring cannot be decided on the basis of one measurement. At the very least, measure at several frequencies; there is further discussion in Section 4. Here in this paragraph is a group of formulas and diagrams relating the parallel equivalent parameters to the series equivalent parameters. Notice that D, Q, and ware equally valid members in either set of parameters. OPERATION 3-7 Gx PARALLEL SERIES Figure 3-4. Equivalent circuits of the unknown capacitor. CSE~ C,[1,( WGC',)} C,11'0'1 R G2+W2C2 x where C G R 3.7 FREQUENCY. 3.7.1 Setting the Frequency. Gx =----S x To make measurements at any frequency in the range 10 Hz-lOG kHz, set the oscillator FREQUENCY selector dials and range switch so that they read as desired. Their indicators together make a convenient, in-line readout. The 3rd-d igit selector is a continuous control (although the readout indicates only 10 positions) From the detent, at the zero position, this control can be rotated cw, through the indicated decade range, or ccw for a small negative extension of that range. Observe the limitations on high voltage at high frequency as described below. In brief, use the following procedure to set the frequency of measurements by the 1621 Precision CapacitanceMeasurement System a. Set the oscillator FREQUENCY controls. b. Set the detector FREQUENCY controls to match. Without balancing the bridge, peak the response of the MAGNITUDE meter by fine tuning of the oscillator. c. Make the phase adjustment as in paragraph 3.3. G x(Q2+1) capacitance in farads in Siemens = resistance in ohms w= 21Tf f = frequency in hertz. = conductance 3.6.2 Dissipation Factor. NOTE Unless your measurement must be made at exactly 1.00 kHz (for example) set the oscillator a little above such a round number. Doing so will simplify the tuning of detector and oscillator to the same frequency. Figure 3-5. Dissipation factor or loss tangent 0 (the reciprocal of storage factor Q) is defined above. Refer also to para. 4.2. o is presented in convenient, graphical form in Figure 3-5, which is used as follows a. Find C x on the appropriate scale, if the measurement frequency is one of those given. Otherwise, notice that the horizontal scale is really the product fC x The higher your frequency, the farther right your capacitance scale shifts. (For example, at 3 kHz, imagine the l-kHz scale shifted to the right nearly half a decade; if your C; = 8 pF, locate 24 pF on the l-kHz scale.) Imagine a vertical line through that point, representing fC x b. Find G x on the slanted scale. I magine a slanted line to represent G x . c. Find 0 on the vertical scale, directly left of the intersection of your fC x and G x lines. d. For greater precision, use the expanded partial chart (but determine order of magnitude from step c) For still f C; higher precision, use the formula 0 = G x / 21T If the range of bridge measurement is extended by the 3-8 OPERATION use of an external standard, extrapolate the corresponding scale. (For example, an external standard of conductance may be advantageous for measuring large, lossy capacitances at high frequency. Add more slanted lines in the upper right portion of chart to represent such an extension.) The shaded square in the main chart and that of the expanded partial chart represent the same range of data. A visual impression of the magnitude of this expansion helps you use the partial chart. 3.7.2 Monitoring Frequency. a. Connect a counter to the EXT SYNC connector (rear). b. Select PERIOD measurement (1 IJ.sTIME BASE) and 10 PERIODS AVERAGED on the counter. For a frequency near 10 Hz, most or all of the counter readout is now utilized. The most significant digits will spillover to the left as you proceed; remember them as they disappear. c. Increase the PERIODS AVERAGED, turning the control cw as far as is necessary to obtain the desired resolution of measurement. d. Calculate frequency, the reciprocal of the period just measured. 10. 6 10- 5 10- 4 10- 3 o 2 10 10. 2 0:: o 10. 10 f- u « u, ' Q z o ~ 10. o 10. 2 en en 10 ' 10 10 2 3 4 10 10 10 100 of I I 10 100fF 10 100fF 10 100pF 100aF I (f = 120)_ 0.1 10 100 of 10 = 10 ) -.. O.I 10 100aF I CAPACITANCE CONDUCTANCE G 2 o -, ~ I a:: 8 o f- ~ 6 u. 5 z o 4 f- Ci 3 (f) (f) Ci 2 ~~ ~~ '", "" -, I _ .'-' " <, ,,"-. -, -, ,"-. <, -, -, (1=1000) (1= 120) ~ -, 1 1 2 t CAPACITANCE 3 2 c, I 100nF I I 10fLF I " 14 5 6 3 5 6 4 FOR CERTAIN -, " <," -,-, "-:\~ -, "<, -, "-. ~ ~~ -, -, ~ -, ,,1'\ 6' I 10 -, -: -, -, -, <, "- i\.'\ "-. ~ I I 100nF <, '" "'" <, -, t\.' I"-.'", ",,,,~ -, "-" , """-, "-" <, I I 100pF 10 -, "', ~ -, .:'\ . i 10 ~ ," ".'" -, " f« ,,"-. ."- i 100pF SCALES) ~ -,"0-~~ "-,'\ " ~9 (SLANTED "-\:'" 106 10 fLF 10 100fF I 10 100pF I 10 100nF C x • AT 4 COMMONLY USED FREQUENCIES '9"S'6',\:'~ ,f 10 100fF I j (f 100nF i i 10 10 5 8 ,,~ Q 3 4 5 6 8 12 1 8 -, 2 1 I 2 FREQUENCIES 161628 Figure 3-5. Dissipation factor vs directly measured parameters. Above: entire range of bridge with internal standards. Left: expanded partial chart, used for greater resolution. For any arbitrary frequency, enter C x scale at correct product fC. (Example: if 1=60, C x=6 pF, use the scale for "1=120" but read "C x=3 pF"1. OPERATION 3-9 3.7.3 Locking to a Frequency Reference. If the reference frequency is known within ±1%, and the reference source can drive the 27-kQ EXT SYNC circuit rms, proceed as follows with 1-10 a. Connect the reference source to the EXT SYNC jack. b. Tune the capacitance-measuring system to the reference frequency, setting the phase shift as described before. However, if the reference frequency is not known to ±1%, or its signal level is < 1 V rms, follow this procedure a. Using a tee, connect both the frequency reference and a scope to the EXT SYNC jack. b. Vary the 1316 Oscillator frequency over a range of 30% or so, observing the scope. It should be possible to see both varying-frequency and fixed-frequency components in the waveform. (Synchronize the scope, preferably, to the reference. ) v c. Tune the oscillator through lock, a condition in which the waveform "stands still". Determine the range over which lock can be maintained, and set the FREQUENCY dials to the center of that range. d. Tune the detector and set the phase shift. 3.8 VOLTAGE LEVEL. Selection. The source voltage is selected by the 2 controls, OUTPUT VOLTAGE RANGE and OUTPUT ADJUST, on the 1316 Oscillator. Read the front-panel meter, full scale being the number indicated by the OUTPUT VOLTAGE RANGE switch. Another source can be used instead; but always select a level within the limits described below, and never more than 350 V rrns, max. * Frequency Dependence. Select a source voltage ~ O.16 f volts rms, to avoid saturation in the bridge ratio transformer. For example, at frequencies of 0.1, 1.0, 10, & 100 kHz, maximum levels are 16, 160,350, & 350 V rms. NOTE Voltages at the unknown capacitor, greater than this limit, can be obtained by using the reversed configuration of para 3.15 Large capacitors. You can resolve large C at high frequency using relatively low voltage. Appropriately, levels available in the 1621 system are then limited by oscillator loading. Normal operation is below these limits level < 4 X 10-3 / fC volts rms for OUTPUT VOLTAGE RANGE = 150 V; level < lef 2 / fC for lower ranges Example for 01 flF at 100 kHz, select 10 V max, on the 15-V range 2-TERMINAL Xl. Use 20 V rrns, max, at this one setting of the TERMINAL SELECTOR Otherwise, clipping circu its designed to protect you from a shock hazard will cause an error in your measurement. Resolution. The chief reason for using a high voltage level is to facilitate balancing the bridge to a high resolution. Generally, make preliminary balances with 15 Vor so. *500V Then switch to 50, and finally 150 V, as you refine the measurement beyond the 6th significant digit. Do not ex- ceed the voltage limits described above. '200 400 600 200 600 400 1000 ppm 800 10 IfF +>-l..I-J.-t.Ll..I-J.-t.Ll..I-J.-t-'-'L.w..-t-'----+-Wu...L+-Wu...Lf-Wu...Lj-l-'-'-'-j-l-'-'-L.i o 10 20 30 40 STEADY-STATE TEMPERATURE - 50 DEGREES CELSIUS Figure 3-7. Worst-case measurement error due to temperature Ibased on maximum temp coaf of C stdsl. Basic specifications apply in shaded band 123°C ±1°CI. In remainder of chart the error may add lin the worst easel to the tolerance given in Figure 3-6. the readout were ** nF 103 pF 456 fF 789 aF, then the measurement would be 103.4583 pF, +.0011, -.0013pF NOTE If the bridge is to be used regularly at a temperature other than 23°C, it can be calibrated for such use. Refer to para. 5.4. For temperature rise in 1621 System, refer to Specifications, in front of manual. 3.9.3 Rangeand Dissipation-Factor Limitations. Figure 3-5. G-Range Limits. The accuracy of a Cx measurement may be limited by an extreme of Gx (or vice versa) For example, if both Cx and frequency are very large (near the limits specified for the bridge) and the capacitor is very lossy, the largest internal G standard may not be large enough for balance. Similarly. a very small Cx with very low losses measured at a low frequency, may require smaller G-standard steps for the desired resolution than the smallest in the bridge. The simplest way to extend the range is by the use of external standards. The range limits are represented in the figure by the ends of C; and Gx scales. Dissipation Factor. Figure 3-5 covers somewhat more range of D than can easily be evaluated by this bridge. For moderate values. all the parameters discussed in paragraph 3.6 can be determined readily. However. at extremes of D (or QJ.only the predominate admittance can be measured accurately. Very low D. At about D ~ l(fs the unknown capacitor has lossesas low or lower than those of the internal stand- 3-12 OPERATION ard capacitor. If. for example, D = 10-7 , you cannot measure G; directly by the CONDUCTANCE readout. and so cannot measure D. However, this situation does not limit the Cx-measurement accuracy, because any decade of the CONDUCTANCE standards can be set to 1. C-Accuracy with High D. BecauseG standards have some stray C, C accuracy is sacrificed when much G has to be balanced. The error can be + or -, depending on certain compensations. Calculate accuracy as follows: where Do is the ratio of conductance to l-k Hz susceptance of the G standards. and is 103 or more for this bridge. Comparison accuracy depends on the G standards that are changed between measurements. Substituting b.G for G in the formu la. comp acc = ± b.G / (21T103 CD o). For example. if C 1 = 12.3456 pF, G 1 = 1.239 X l(fl f.lS and f = 500 Hz, then D = 3.2 and accuracy is ±1600 ppm. If a similar capacitor has C2 = 12.3123 pF and G 2 = 1.241 X l(fl f.lS, then comparison accuracy is ±13 ppm. Seealso para. 4.7. 3.9.4 Shunt Capacitancesto Ground. Figures3-1, 3-2. These capacitances, Cis' Chs' and Chg are excluded from the measurement of Cx. However, if large, they can reduce accuracy and precision (particularly at high frequency). If you can keep these capacitances below 200 pF each (unknown capacitor directly connected to bridge terminals) or 100 pF plus cable capacitance (unknown capacitor connected by a pair of G874-R22LA patch cords) measurement accuracy is unimpaired, even at 100 kHz. Three-terminal Measurements. If you must measure in the presence of larger shunt capacitances to ground, calculate a bound on the probable error as follows C readout = Cx - w Greadout = G 2 ChsCls Lsg 2 w ChsClsRsg ground, the error from ground capacitance should be negligible. Coaxial 2-terminal measurements differ in this aspect because the LOW bridge terminal is cornpletelv shielded by the HIGH Capacitance from HIGH to qround does not affect accuracy except for loading of the ratio transformer That effect is negligible for C hg < 100 pF or 0.1 C x (whichever is greater) However, a distinction must be made between C hg and fringing capacitance; see below. 3.9.5 Notice that the error terms are both negative. Cis is low-side-to-shield capacitance; Chs' high-side-to-shield. Include cable capacitances, if patch cords are used. RSg and LSg are, respectively, the resistance and inductance of the path from a virtual common point in the shield surrounding the unknown capacitor to the ground point in the heart of the bridge. Include the Rand L of the patch-cord outer conductors (2 in parallel, if 2 patch cords are used) in R Sg and LSg' As Figure 3-1 shows, our error calculation is based on a lumped-parameter situation. If cables are used, C, L, and R are really distributed along the cables, and the corresponding error terms are even less significant than those calculated above For example, assume shunt capacitance of 45 pF at each end of the unknown capacitor, in addition to 90 pF for each G87 4- R22 LA patch cord. Assume 1 /LH and 10 mD for for the series-impedance parameters of each cable and its associated connectors Assume Frequency is 100 kHz. Then: Lsg = 5 X Hr7 henry Rsg = 5 X W- 3 ohm G error = w 2 Chs Cis RSg = 3.6 X 10-11 Siemens. If C x = 1 pF, compare this C error with the normal error, 5 X 1(JIS F. (Refer TO Figure 3-6; measuring 1 pF, we expect 500o-ppm accuracy.) Notice that most of that error is due to series inductance in the bridge (see para 4.7) and is accentuated by any inductance of the cable inner conductors. At lower frequencies, with very long cables, the series inductance thus usually contributes more error than the ground-return inductance. Two-terminal Measurements. Non-coaxial 2-terminal measurements can be considered a special case of 3-terminal measurements in this aspect. If neither terminal has more than 200 pF to nearby instrument or bench "ground", and if that ground is connected by low impedance to the bridge Fringing Capacitance. Fringing capacitance is significant in coaxial 2-terminal measurements. (The HIGH and LOW connectors for 3terminal measurements are far enough apart so the fringing between their inner terminals is only 125 aF.) As explained in paragraph 4.2, the recommended setting of ZERO ADJUST makes the balanced-bridge C readout express just the capacitance C due to fields beyond the reference plane of the x G900 connectors, entirely within the "unknown" capacitor and its connector. As described in para. 3.4, you may assume the fringing is 0.11 5 pF and be correct within ± .008 pF, i e., 8 fF. Two methods are suggested for greater accuracy, as follows. Standard Capacitor,Use a Coaxial Capacitance Standard (2 pF) or (1 pF) with calibration specified within ±5 fF, or a G900-W04 Precision Open-Circuit Terminal (2.670 ± .0067 pF). You may have one calibrated to greater accuracy by the National Bureau of Standards. a. Set front-panel controls as of the completion of functional checks and phase adjustment, para. 32 and 3.3, at the desired frequency, para. 3.7. b. Install the coaxial capacitance standard and set the CAPACITANCE lever switches to its value. c. Set the TERMINAL SELECTOR to 2 TERMINAL Xl, XlO, or X100 depending on whether you wish to measure capacitors up to 1 n F, 10 n F, or 10 J.1.F,respectively. Keep the C MAX switch down, during this procedure (and during measurement of your unknown capacitors except when their values require large C standards with a READOUT MULTIPLIER of Xl00) d. Balance the bridge, using the ZERO ADJUST control for C and the CONDUCTANCE lever switches for G Turn the GAIN and SENSITIVITY controls as required for suitable indications of balance. NOTE The resolution of ZERO ADJUST settings is about 0.5 fF for X 1 and X 10, about 5 fF for Xl00 READOUT MULTIPLIER. e. If you want to obtain a more exact value for the fringing of the G900 Connector on your bridge (for use in para. 3.4) remove the capacitance standard and rebalance the bridge using the CAPACITANCE lever switches. The OPERATION 3·13 repeatability is about ±1 fF, verify above procedure a few times. this by repeating the Direct Measurement. Calibrate a coaxial 2-terr'ninal capacitor against the internal bridge standards. If you prefer. by the method of Millea'* (Refer to Appendix A.) This method requires the capacitor to be measured twice. while attached to a fixture made with a tee. suitable connectors. and 2 lengths of flexible cable. The fixture is not readily available; details are left to your ingenuity. 3.9.6 Conductance Accuracy. The accuracy specification is ±O.1% ± 1 step in the least significant digit, particularly for G :::=10- 1 ] S However, that last step need not be a limitation at larger conduc tance. If you avoid making the first digit of the CONDUC TANCE readout a zero G readout = Gx ±O.1% Gx, for Gx;;;;'l cr S p.S Offset. Measure Go and use the correction, as follows. a. With EXT MULTIPLIER OFF, TERMINAL SELEC TOR at CAL, and readouts initially at zero, balance the bridge b. The G readout is G,,; it rnay be + or -, with magnilcr 8 p.S This offset is unaffected by tude (typically) ZERO ADJUST c. For any measurement (typically if the G multiplier from the readout, thus G; in which Go is significant is lcr s or lcr 6 ) subtract it = (G readout ± 1 in 5th window o': coefficients . are -700, -300, ±50, ±50, and Of course, you can calibrate the bridge at a temperature other than 23°C, making use of the "±15 ppm" internal standards or any suitable external standards for references Refer to para. 5.4. The internal conductance standards respond to environmental temperature change with a srnall time constant (a few minutes) in contrast to the large capacitance standards. C-Range Limits. As discussed above (conversely), G x accuracy may be lirnited by the available ranqe of internal C 'Millea, Aurel, "Connector Pair Techniques for the Accurate Meas· Journal of urement of Two-Terminal Low-Value Capacitances," Research, 3, of the National Bureau of Standards, Vol 74C, Nos 3&4, .Julv-Dec , 1970. 3-14 OPERATION Capacitance to Ground. As explained in the Cxaccuracy discussion, G, accuracy may be affected by the ground capacitance and series impedance from the shield around the unknown device through the cables to the bridge ground. I f you must use long cables, and particularly if you need to test at high frequency, use the formula given there to esti mate the consequent error Go) ± 01% Temperature. The temperature effects on conductance accuracy are nearly negligible. The basic specification of ±1000 ppm (see above) applies over the temperature range of 23 ±l°C For further temperature changes, the worst case obtains when the CONDUCTANCE multiplier is set to 1cr 4 , 1 or 1cr 6 , then the readout at balance varies inversely with temperature at the rate of 700 ppmtC For settings of 10-4 , 10-], 1cr 2 , 1cr 1 , and 1, respectively, the temperature ±15 ppmtC standards. For example, if G; 10- 1 ] S ?"d D = 10 at f = 105, the smallest internal C standard is not small enough for a precise balance. (Stray C in the G standards has to be cancelled, so we are not able to measure Cx anvwav.) If Gx = lcr 4 Sand D lcr 2 at f = 120, the largest internal C standard is not large enough for balance. Use an external standard, if you want to ex te.id either limitation. Dissipation Factor. At very large D, the unknown con ductance may have less capacitance than the internal stand ard conductance. Then you cannot measure C; directly and so cannot measure D But (in contrast to small-D limits) this situation is generally no limitation on C;x measurement accuracy. C standards can be set negative if necessary for balance. At very low D, the accuracy of G, measurements is limited because a significant part of the conductance in the standard arm of the bridge consists of losses in the internal capacitance standards. It is therefore impractical to measure G; in the region near or below the bottom edge of the chart Figure 3-5. If the low-loss Cvstandards are used (C; < 1 nF) this G, accuracy limitation is ~'Yr) at D io '. 10e;;" at D = 10-4 , and about 100 e;;,) at D 10- 5 . 3.10 PRECISION. The comparison precision of the 1621 system (or the 1616 bridge if set up in an equivalent system) is specified to be ±01 ppm (one part in 108 ) for low-loss capacitors between 10 pF and 10 pF This statement means that the system has the resolution, sensitivity, stability, repeatability, and operating convenience necessary for you to balance capacitance to ±01 ppm. Such precision is significant in measuring capacitance changes, comparing capacitors, adjusting and evaluating capacitance standards, etc. The precision is available for measurements described in preceding paragraphs, but far exceeds the absolute accuracy to which we can guarantee the many possible combinations of internal standards. Resolution. The smallest internal standard step is l 9 lcr F, so resolution of ±.01 ppm extends from C x 1cr l 1 F (10 pF) to 1cr 5 F (10 pF) without external standards. An extra decade, at one end or the other of the range, can be provided by a suitable external standard. Sensitivity. You can balance the 1621 system to ±.01 ppm because (among other things) you can see aOl ppm unbalance on the detector. This remarkable sensitivity depends on instrument capabilities, but also on your selection of a high source voltage, proper tuning of the detector, and setting of the phase shift, as described before. Because source voltage must be kept below a limit proportional to frequency, sensitivity is generally adequate for the above-mentioned resolution only at frequencies above about 900 Hz. These somewhat arbitrary limits obtain: at 10, 102 , 103 , 104 , and 105 Hz you can resolve to ±.01 ppm when c, 10-9 , 10-1 0 , 10-1 1 , 10-1 2 , and 10-1 2 F, respectively. (To realize the 1 Stability. You can enjoy the above-mentioned precision because (also) the internal standards which determine the most-significant digits of any such precise measurement are sufficiently stable. They are stable in terms of mechanicai shock, aging, atmospheric changes, ambient temperature, and other factors. Consider temperature in more detail. These standards are thermally isolated, with a time constant of 6 hours. Assuming you can make a comparison (2 measurements) in 6 oC minutes, the effect of a l_ ambient temperature change on that comparison cannot exceed 1/60 of the specified temperature coefficient. That is 3 ppmtC for C; in the range 1-1000 pF, so the effect is .05 ppm. Therefore, for ±.Ol-ppm precision under these conditions, be sure the ambient temperature is regulated within ±0.2°C. NOTE Because of the long thermal time delay in the bridge, it must be held at the desired temperature a long time before precise measurements can be made Refer to para 4.5 for a quantitative discussion. For example, a 12...?Cchange, 24 hours before, will disturb a comparison taking 6 minutes as much as a 0.2°-C change during the comparison. Take reasonable care to avoid mechanical shock, regulate temperature as described above, and you can rely on the stability of the internal standards, for C; about 1-1000 pF However, the internal 10 and 100-nF standards, used for C; in the range of 2-100 nF, may have as much as an order of magnitude larger temperature coefficient. To obtain .01 ppm precision using them, make more rapid comparisons, provide better temperature regulation, or use an unusually-stable external standard as described below under the heading of Externally Determined Accuracy. 3.11 EXTERNAL STANDARDS. The 1616 Precision Capacitance Bridge has considerable versatility to make special measurements and comparisons in addition to those already described. These paragraphs deal with the use of external standards to extend range, 3.11.1 RangeExtensionto 111 [.1F. Accessories. A 1-JlF. highly stable. external standard. i e.. GR 1409- Y Standard Capacitor and two G874- R33 patch cords. Connections. Connect the capacitor to the EXTERNAL STANDARD port of the bridge, thus capacitor H to bridge HIGH, capacitor L to bridge LOW, and capacitor G to the shield of the LOW cable. Connect the unknown capacitor to its port as usual. Calibration. Verify the calibration of the external standard with adaptors and cable at the frequency you plan to use, by the method of para. 3.4 and 3.5, or as follows set EXT MULTIPLIER to -0.1, TERMINAL SELECTOR to CAL; balance the bridge; and multiply the CAPACITANCE readout by 10. Balance. Measure the unknown capacitor as usual except that you now have a more-significant C digit than before, controlled by the EXT MULTIPLIER switch. Readout. If the external standard is 1 [.1F,with sufficient accuracy for your measurement, and if the series-inductance error described below is tolerable, interpret the EXT MULTIPLIER setting multiplied by 10 as an extra digit at the left of the usual C readout. (Example ext std value 1.00011 [.1F;EXT MUL T 0.6; readout 54 nF 321 pF . READOUT MU LT X 100; desired accuracy ±0.1 %. Final value is 65.4 [.1F.) But if v.ou want greater accuracy, the measurement is [Lexternal standard C value) (EXT MULTIPLIER) + CAPACITANCE readout] (READOUT MULTIPLIER) (series-inductance correction). In the same example (1.00011 X 0.6 + .054321) X 100 = 65.4387 [.1F,without the correction (see below). In this example, that might be .05 /IF. Unless you can refine that, only 2 decimal places are valid. Then the final measurement is 65.44-.05 = 65.39 [.1F Similarly, the final G readout is [(external standard G + (CONDUCTANCE readout) value) (EXT MULTIPLIER) (CON DUCT ANCE multiplier)] (READOUT MUL TIPLlER). Series-L Correction. The accuracy of measurement of large capacitance is usually Ii mited by the bridge and lead inductance in series with the capacitance. The bridge reading of capacitance is greater than the unknown capacitance C; by a capacitance error tJ,C = ",} C2 Q. If bridge induc- OPERATION 3-15 tance Q is about 0.3 pH, in series with the UNKNOWN terminals, you have a minimum error of the order of +0.002% C}.If (f k H z ) 2 Hence, at 1000 Hz and 100 pf, the bridge reading might be high by about 0.2%. Refer also to para. 4.7. 3.11.2 Extension of C Resolution. Although the resolution of the bridge is exceptionally fine, you can make it still finer with a sufficiently small external capacitance standard. Thoroughly shield anything connected to EXTERNAL STANDARD port, particularly on the LOW side. greater accuracy than the internal standards permit, and if you have suitable external standards, use this procedure. (For example: the bridge may not yet be stabilized at room temperature.) Alternatively, you may wish to measure against a standard deliberately set a few ppm different from the internal standards. Comparisons between external standards of the same nominal value, or values differing by convenient ratios, like 1, 2, or 10, can be made to great precision (para. 3.10). Use one as "standard", one as "unknown." As a check on possible zero offset, interchange them and compare again. Accessories. A Accessories. A well-shielded O.l-aF (i.e., 10-\ 9 F) 3-terminal capacitor; well-shielded cables, such as 874-R22LA. NOTE If you make such a capacitor, use button-sized plates spaced several cm either side of a shield with a small pinhole in the center. Zero Adjust. If the measurement is to be made at the 3-TERMINAL port, shield the 3-T LOW connector using an open-circuit termination; set TERMINAL SELECTOR to 3-T Xl, C MAX lever down, and the entire C readout to zero. Balance the bridge with ZERO ADJUST and the G lever switches. [If the measurement is to be 2-terminal, omit the shield, install a coax ial C standard, set TE RMINAL SELECTOR accordingly, and balance the bridge with the C readout fixed at the value of the C standard.) Calibration. Measure the external standard at the EXTERNAL STANDARD port. A balance should obtain with C readout set to 0.1 aF, EXT MULTIPLIER to 1.0. [If the measurement is to be 2-terminal, set the C readout 0.1 aF above the value of the C standard.) Use a high frequency (10-100 kHz) and high oscillator level. Measurement precision of ±1O% is probably adequate. (Setting the EXT MULTIPLIER to 0.9 should noticeably unbalance the bridge) If possible, adjust the external standard to OlaF. Balance. Connect the unk nown capacitor to its port. Measure as usual except that you now have a less-significant C digit than before, controlled by the EXT MULTIPLIER switch. Readout. Interpret the EXT MUL TIPLI ER indication multiplied by 10 as an extra digit at the right of the usual C readout. (Exarnple ext std value 0.1 aF; EXT MU LT 0.6; readout 234 aF 5; READOUT MUL T X 10. Final value is 2.3456 fF) 3.11.3 Externally Determined Accuracy I Comparisons. Applications. If you want to make a series of measurements with the convenience of direct readout but with 3-16 OPERATION suitable reference, i.e.. GR 1404-A, and 2 of G87 4- R22 LA patch cords. Connections. External standard capacitor to EXTERNAL STANDARD port; H to HIGH, L to LOW. Unknown capacitor to its port. Balance. Measure as usual, except be sure to control the most significant digit by the EXT MULTIPLIER switch. Notice which C-Iever switch (call it S) corresponds in magnitude to the EXT MU LT. Set "S" and all at its left to zero. For good external determination of accuracy, EXT MUL T should be set high, and one or more levers right of "S", low (preferably zero). (Example Ext std 100 pF; "S" is the lO-pF lever, regardless whether READOUT MU LT is 1X or higher. Note "S" set to X would represent 100 pf . and so would EXT MUL T set to 1.0.) Readout. Read out as usual except that EXT MUL TIPLiER reading multiplied by 10 must be substituted for the zero of lever switch "S." (Example, as above, also EXT MUL T 0.8; READOUT MUL T lOX; C readout 123 fF 456 aF. Final value is 801.23456 pF.) Accuracy. Estimate the. C and G accuracies by adding these error contributions for each. 1. Error in calibration of the external standard capacitor, ppm. 2. Error of internal standards as they apply (refer to para. 3.9) express temporarily as C or G (not ppm) and multiply by the indicated READOUT MULTIPLIER. Express result as ppm of the "final value". 3. Error of ratio transformation. Typically error is < for ratio 3 10 10 100, ppm where the ratio is the product of EXT MU LTI PLI ER and READOUT MULTIPLIER settings. Ratio of 1 must be (1.0) (Xl) not (0.1) (XlO) 4. Error of zero offset. Bridge reading at balance with both ports open but shielded (use a G874-W) Termina- tion on each LOW connector), EXT MULTIPLIER at 1.0, and READOUT MULTIPLIER at Xl. This error should be zero. 3.11.4 Test-Fixture Compensation. Applications. You may have a test fixture, leads, terminals, etc. which have parameters included in C x and Gx but which you want to balance separately so that the bridge readout is only the additional C and G of capacitors connected to the fixture. Notice that ZERO ADJUST performs just this function for 2-TERMINAL measurements only, for C only, and for a very limited range, but with great precision. (Its function, specifically, is to balance the capacitance of the G900 connector on the bridge.) mulae for C and G are both given there. The "external standard G value" is 100 JLS. NOTE External standards of C and G can be used simultaneously by connecting them in parallel. 3.12 PRECISE COMPARISONS. 3.12.1 BalanceComparisons. Very precise comparisons between 2 capacitors nominally related by convenient ratios, such as 1,2, or 10, can be made by connecting one as EXTERNAL STANDARD, the other as UN KNOWN capacitor. The internal standards are used only to evaluate the difference. Refer to para. 3.11. 3.12.2 Direct Substitution. Accessories. Stable, adjustable, 3-terminal shielded, precision capacitor and (optionally) a conductance, i.e.. GR1422-CB capacitor, and two G874-R33 patch cords. (Adjustable conductance standards, though not usually needed, may be connected in parallel.] Connections. Connect the test fixture you plan to use to the appropriate UNKNOWN port; the compensating capacitor (and conductance, if any) to the EXTERNAL STANDARD port. Set the READOUT MUL TIPLIER as you intend to keep it for subsequent measurements. Adjustment. With the C and G lever switches set to zero and EXT MULTIPLIER on 1.0 (or a lower setting) balance the bridge by adjustment of the external standard. If that is only capacitive, manipulate the CONDUCTANCE lever switches as usual. Notice that is is quite possible to make satisfactory testfixture compensation without reaching a perfect balance with the external standard. For example, suppose you need to compensate about 20 pF: READOUT MULT: Xl: available capacitor range: 10 to 110 pF. Set EXT MULT to 0.2. This capacitor has a resolution better than 4 fF, i e., < 1 fF, referred to the test fixture. If that resolution is sufficient simply verify that you have achieved it. (Refine the balance using C lever switches and verify that appreciably less than ±0.5 fF is required of them.) If, in that example. resolution must be lOaF, you need another external variable standard capacitor in parallel with a suitable fixed capacitor. 3.11.5 Range Extension to 11 mS. Accessories. A standard 10-k!1 resistor and two G874R33 patch cords (or one cord and G874- MB adaptor). General Procedure. Refer to the analogous paragraph 3.11.1, which concerns C rather than G. The readout for- Even more precise comparisons between 2 nominally equal capacitors can be made by connecting them sequentially as the UNKNOWN capacitor. External standards may be used. Balance Technique. Make use of (-1) and X settings of lever switches if doing so will permit more of the most-significant digits to remain unchanged. Example: if capacitor A measures 12345.0012+ pF, and B something more like 12344.9998+ pF, then make the readouts 12 nF 345 pF 001 fF 255 aF and 12 nF 345 pF 00(-1) fF 855 aF respectively. Repeat measurements A, B, A, B, etc. to eliminate any affects of temperature changes, connector reliability, etc. In this example, then, B is 0014 pF, i.e., 0.12 ppm, lessthan A. Precision. The only limit on how precisely a pair of capacitors can be matched is the resolution of the bridge. The errors listed in para. 3.11.3 do not apply; absolute "farad" value is of secondary importance. If A and B are so similar that each can be balanced repeatedly with the same 9 most-significant digits, and if the 9th digit is significant (changing it upsets the balance) then you may be sure that A and B are equal (for the moment) within ±.001 ppm Generally, comparison precision is specified at ±.01 ppm for the 1621 system. Refer to para. 3.10. Terminal Capacitance. Terminal, fringing, and test-fixture capacitances cancel in direct-substitution measurements, a fact that makes this method convenient for many kinds of 2-terminal measurements, even when great precision is not required. Refer to para. 3.13. 3.13 NON-COAXIAL 2-TERM. CAPACITORS. Figure 3- 8. Measurement of non-coaxial 2-terminal capacitors can be precise and repeatable only to the extent that the various stray admittances can be brought under control. Some can be excluded from the measurement, some need to be included. Some are unavoidably influenced by the method of OPERATION 3-17 connection to the bridge, and so the measurement is valid only if such details are specified. The capacitor terminals are represented in a general way. If they are a pair of binding posts at standard spacing or a coaxial connector and if the bridge UN KNOWN port can be adapted to mate with these terminals, the method of connection can be specified reasonably well. 3.13.1 Unshielded, 2-Terminal Capacitors. Figure 3-8a. Diagrammed is a very simple capacitor, with an internal capacitance C 10. C4 0 represents capacitance between the terminals; C2 0 and C 3 0 capacitances to the nearest conductor in the environment (we assume for simplicity, there is only one such conductor of importance) Intrinsic Value. As it stands alone, the capacitance of diagram a is C20 C30 C 10 + C40 + C C' 20 + 30 Connections. If the capacitor has a G874 connector or 1.75-inch-spaced binding posts. use a G900-0874. or G900- 09 adaptor. respectively. A length of cable or rigid coaxial line can be used to locate the point of attachement away from the bridge. Other capacitors (in general) will require special adaptors or test fixtures. Measured Values. When you connect that capacitor to the 2-TERMINAL port of the bridge, all the strays change. Diagram b shows a wire used to make connection at one terminal. Even if you properly adapt the bridge port so the capacitor mates without such a wire, C4 0 is now different, i.e.. C4 1 . C2 1 is certainly larger than C2 0 · C3 0 may be practically unchanged. C5 0 is capacitance from that "nearby conductor" to bridge ground. Because C2 1 , C 3 1 , and C5 0 form a Y network with a midpoint that we don't care about we can substitute the equivalent 6. (Use a Y-6 or T-1Ttransformation.) Two parts of the 6 are excluded from the measurement.* The third part, across the bridge LOW and HIGH terminals, is C2 1 C31 C2 35 =. C . C21 + C31 + 50 Certainly, if we make C5 0 large enough with respect to C2 1 and C3 1 , C2 3 5 will approach zero. If C5 0 is small enough, C2 3 5 approaches Cj , C 3 1 /(C 2 1 +C 3 d · As represented by diagram b. the measured value is It is usually preferable to make the third term zero rather than try to make it approximate the third term in the "intrinsic value." If you do the former, by grounding any nearby conductors (diagram c) the measured value is: C 1 0+C "Because the 1616 ground as guard. circuitry 3-18 OPERATION 4 1· is basically a 3-termlnal bridge with Correction for Cable, Connector, etc. There are 2 distinct methods of removing adaptor capacitance from your measurement (results differ). For each method, there are 2 techniques (results are identical). Open-Connector Method. Be sure there is nothing but air at the connector or fixture to which the capacitor will be connected. The 2 equivalent techniques are: 1. Use an external standard and/or ZERO ADJUST to balance the bridge with zero CAPACITANCE readout (ref. para. 3.11). Now when you measure an unknown capacitor, the readout is the corrected measurement. Alternatively. 2. Balance the bridge and record the readout Co F ' the capacitance of adaptor, cable, and fringing. Now when you measure an unknown capacitor, correct your measurement as follows: "readout" minus Co F ' (Refer to para. 3.9.5 for more about fringing.) Substitution Method. Connect a standard capacitor Cs with the same terminal configuration as the unknown to the test point. The 2 techniques are 1. Use an external standard and/or ZERO ADJUST to balance the bridge with CAPACITANCE readout set to Cs' Now when you measure an unknown capacitor, the readout is the corrected measurement. Alternatively 2. Balance the bridge and record the readout (Cs+C o ) Refer to para. 3.12. Now when you measure an unknown capacitor, correct your measurement as follows "readout" plus Cs minus (Cs+Co )' NOTE If the standard used in the substitution method was calibrated to eliminate fringing and C4 0 , the corrected measurement is C 10. Use the 3-Terminal Port (Figure 3-8. d). If you wish to measure. as directly as possible. the internal capacitance C . 10 make a separate. shielded connection to each terminal. Make connections (if the capacitor has a pair of binding posts) with a pair of G777-03 adaptors and a pair of G874-R22LA cables. Connect only the shielded plug of each G77 7- 03 to the capacitor. Ground any nearby conductor as shown. If the shields extend far enough. C42 will certainly be less than C40' The measured value is: NOTE As a rule, this bridge readout is valid without correction for cable and adaptor capacitances, although either method described above may be used for greatest accuracy. 3.13.2 Shielded 2-terminal Capacitors. Figure 3-9, e. These capacitors are enclosed in a conducting shield: examples are GR 1409-Y. or 1422-0. Some may alternatively be measured as 3-terminal capacitors (refer to CASE leI lal : } i / I I I I / I I ::;:<.. 7\ c" I J ,....--'"""ilt----..,- I ~C21 I I I ~C50 I I I I I I I I '-'1E-/-"-::: ~----if~~: --/ I I 60 " I --L I .,C C70 -h-,E--- -'- WIRE C81 "- CAS lOW I I ! H l BRIDGE 2-TERMINAl I H H l H 2-TERMINAl PORT PORT OF BRIDGE 1 fI [b] C 70 ---iE-- " I/ I\, C60 C~/1\-:-- t--/ !'--j:---- " ".- r- t----,HIGH lOW 'i(---- r-::V "82 CA SE WIRE) H l L -=l H BRIDGE 2-TERMINAl PORT H BRIDGE 3- TERMINAL PORT (gl Icl C 70 r- " ;x./--iE-----~ ........ "- :...--' 42 "- II ---1E---- C6 0 "- HIGH LOW CASE :-- "'if ---r-::-:.....-- C83 10 C93 l- H T-GAP H Idl BRIDGE 3-TERMINAl Figure 3-8. PORT Ihl BRIDGE 3- TERMINAL PORT Figure 3-9. Non-coaxial 2-terminal capacitors and some measurement configurations. OPERATION 3-19 para. 3-4). Others may have one terminal already connected to the shield. Intrinsic Value. As it stands alone, the capacitance of diagram e is However, we are specifically interested in 2-terminal value, with one terminal connected to the case. Assuming the right-hand terminal is the one, then the intrinsic capacitance reduces to. Connections. Connect the LOW terminal to the GND or shield terminal using a link. 938- L, if it fits. Connect the 2 capacitor terminals to the 2- TERMINAL UNKNOWN port of the bridge as follows: capacitor LOW to bridge outer (HIGH): capacitor HIGH to bridge inner (LOW). If the capacitor terminals are 1. 75-inch-spaced binding posts. use G777-03 and G900- 08 74 adaptors and, if desired, a shielded cable (G874- R22 LA). Alternatively, use only the link and a G90009 adaptor. Measured Value. When you connect that capacitor to the bridge at least 2 of the strays change. C8 0 is probably reduced (by shielding) to C8 1 ; C9 0 changes completely (now that LOW and GND are connected) to C9 1 . As shown by Figure 3-9, f, the measured value is NOTE Correct the bridge readout for cable and adaptor capacitance, as described before. Use the 3- Terminal Port, Simply (F igure 3-9, g) If you wish to make virtually the same measurement without the need to correct for cable and adaptor capacitances, proceed this way a Connect LOW to shield or GND of the capacitor with atink as before. b. Install a G874- MB adaptor at the bridge 3- TERMINAL HIGH connector. Tie that to the capacitor LOW with a wire or plain patch cord. c. Connect the G77 7- 03 adaptor shielded plug only to the capacitor HIGH terminal. Tie that to the bridge 3TERMINAL LOW connector using a shielded cable. G874R22LA. Be sure the shield of the G777-03 adaptor does not contact the shield surrounding the capacitor. Make a washer of paper or plastic(.003 to .015 in. thick) for this purpose. d. Your measured value (diagram g) is 3-20 OPERATION (where C8 2 is comparable to C8 1 + C9 1 of diagram f.) Use the 3-Term. Port, with Correction(Figure 3-9, h). Some capacitors are specified in terms of the capacitance added to a given bridge port (Adapted, if necessary, to suit the capacitors). For example the 2-terminal capacitances of certain GR 1422 capacitors are so defined. A typical procedure follows: a. Install a 938-L link on the capacitor-case (GND) binding post, connecting also to the adjacent LOW post, if any. b. Install a G777-03 adaptor on the capacitor - insulated plug at capacitor HIGH. shell connected plug at the adjacent bind post with the link. c. Install a G874-MB Coupling Probe as an adaptor on top of the G777-03JJnserew the binding post just installed, far enough to determine the gap as follows d. Install anotherG777-03(shielded plug only) on top of the stack. With it firmly seated, verify that a 1/16-in. gap (2 mm or less) exists between the shields. NOTE This gap is the demarkation between part of the adaptor stack included in C9 3 and part that is not. To stabilize the gap, wrap assembled adaptors tightly with electrical tape e. Connect the top adaptor to the bridge 3-TERMINAL UNKNOWN, LOW using a shielded cable, G874- R22 LA. f. Install another G87 4- MB adaptor at the 3- TERMINAL UNKNOWN, HIGH connector. Tie this, using a patch cord or wire, to the link in step a. (If the capacitor has only 2 binding posts, fasten to the link with an aligator clip; otherwise, use the 3rd binding post.) g. Temporarily unplug the first G777-03 adaptor from the capacitor binding posts and support it nearby with the inner, insulated plug exposed. Tie the shell-connected plug (uninsulated) to the capacitor-case binding post (bridge HIGH) as before. h. Compensate the bridge by the open-connector method (technique 1 or 2) described above. This step eliminates C9 3 (diagram h) from your measurement. i. Reinstall the first G777-03 adaptor as in step b. Measure the unknown capacitor as usual. J. Your measured value (diagram h) is 3.14 DC BIAS. Dc bias voltage may be applied in either of two ways to a capacitor that is being measured on the bridge. l~ & CAUTION Do not apply voltage at the bridge DETECTOR OUTPUT connector in excess of E M A X in Table 3-2, or the G standards may be damaged e. Connect the unit holding R B to the tee. and that to the DFTECTOR OUTPUT connector of the bridqe iiSlflq an adaptor (GB740BPA) A recommended power supply is the G R 1265-A. & WARNING • To minimize electrical shock hazard, limit bias to 60 V. • Bias voltage is present at connectors, test fixtures and on capacitors under test. • Capacitors remain charged after measurement. • Do not leave instrument unattended with bias applied. • To remove the risk of electric shock, turn the voltage source to "0" before connecting or disconnecting the device under test. ...L + L l~---------+--__ -, Figure 3·10. Circuit for applying bias to the unknown capacitor - normal configuration. Notes: • Resistor in G874-X Insertion Unit; G874-02 Adaptor provides binding posts for power supply. .. G874- T tee with G874-QBJA and G874-QBPA Adaptors to BNC connectors. It should be noted that the instrument is rated to accept bias voltages up to 500 V rms or de. However, to minimize the risk of electrical shock hazards, the use of biasing voltages of less than 60 V is highly recommended. 3.14.1 Normal Bridge Configuration / Parallel Bias.Fig. 3·10. To apply bias In parallel with the detector a Connect the bridge as usual, but add the de bias supply In parallel with the deter-tor. as shown in the circuit of F iqure 310. b. Shield all leads connected to the high side of the detector. Use a G87 4- T tee connector for convenient parallel connections and a G87 4- OBJA Adaptor If your detector is not the GR 1238, connect a series capacitor between detector input and bias supply to block bias voltage train the input stage. (This capacitor IS built into the detector input stage of the 1238) d. Connect a series resistor R B between the high lead in the tee and the bias supply to prevent the low impedance of the bias supply from shorting the detector input A resis ranee of about 100 kn is recommendod Lower resistance reduces the detector sensitivity; higher resistance reduces the de voltage across the unknown capacitor since 1. Either polari tv of de bias can be appl ied. Choose the polarity required by the capacitor being tested. Observe the r.urr ent limitations described below. g Calculate the bias from a known E B , using the formula in step d. 3.14.2 Reversed Bridge Configuration / Series Bias. Fig. 3-11. Compared to the normal configuration, this method allows higher oscillator voltage to be applied at low frequencies (unless the CONDUCTANCE multiplier must be set very high), has lower sensitivity; and permits direct bias measurement In the reversed configuration, you apply bias in series with the oscillator; use the following procedure a Connect the br idge "reversed"; refer to para 3.15 b. Connect the de bias supply or battery in series with the asci Ilator. Observe the current limitations described below. ----------Table Refer to Table 3-2 for the effective bridge resistance R at BO the DETECTOR OUTPUT connector. Install the resistor In a shield such as a G874-X Insertion Unit 3-2--------CHARACTERISTICS AT BRIDGE DETECTOR-OUTPUT CONNECTOR CONDUCTANCE Multiplier 1 NOTE Use a choke in place of R B for high ac impedance, when low dcvoltage drop is needed. Shield the choke from both rnagnetic and electric fields. R BO 70 V 9kn 10-1 210 89 1(J2 500 810 10- 3 1Cf4, 1O'5 , or 1Cf6 ~)OO 500 8Mn 80 "Vof ts rms or dc. OPERATION 3-21 d. Calculate the bias from a known E B, using the formula given above, except that R'B depends on the settings of FREQUENCY range and OUTPUT VOLTAGE RANGE on the 1316 Oscillator. (R' B can be as large as 2.4 kQ.) Alternatively, measure Ex' (F igure 3·11 ) "GENERATOR INPUT" CONNECTOR GR 1238 OETECTOR c. Turn the oscillator OUTPUT and then slowly ccw to zero again. ADJUST 3.15 REVERSED CONFIGURATION. cw to MAX Figure 3-11. 3.15.1 Explanation. etJ I OPTIONAL FLOATING 81AS SUPPLY To apply higher ac test voltage (up to 500 V) to the unknown capacitor than that normally permitted), use this configuration, in which the oscillator is connected to the bridge DETECTOR OUTPUT connector and the detector is connected to the bridge GENERATOR INPUT. Use this reversed configuration when lower sensitivity can be tolerated. H In normal operation, the maximum voltage across the unknown capacitor would be limited by transformer-core saturation to E M A X 0.16 f (para. 3.8) However, in reversed operation, the maximum voltage is lirnited by power dissipation in bridge resistors, or by insulation breakGR !316 OSCILLATOR GR 1616 8RIDGE Figure 3-11. Reversed configuration - bridge circuit with oscillator and detector interchanged, showing optional application of bias to unknown capacitor. down. The former depends on the CONDUCTANCE plier setting refer to EM A X in Table 3·2. When the configuration multi- is reversed, the ac test voltage across the unknown capacitor is equal to the generator voltage at balance, or between zero and twice that voltage for any unbalance 3.15.2 Procedure. 3.14.3 Dc in the Ratio Transformer / Demagnetization. When the capacitor that is measured with bias (by either of the preceding configurations) passes leakage current, de will flow through the bridge ratio transformer. This current will magnetize the core and affect the accuracy of the a. Interchange 2 connections at the rear of the bridge (Figure 1·3) so you have oscillator POWER OUTPUT tied to bridge DETECTOR OUTPUT and bridge GENERATOR INPUT tied to detector INPUT SIGNAL. b. Balance the bridge as usual (para. 3.5) but observe the limits of Table 3·2 instead of para. 3.8. ratios, and such use generally is not recommended. It is, however, possible to operate the bridge with some de in the transformer, if you keep th is cu rrent below the I M A X values shown in Table 3-3 to avoid saturation of the core. Table 3-3 --------BIAS CURRENT LIMITS FOR TRANSFORMER SATURATION ---------- Operation within these limits will not damage the bridge, but demagnetize the transformer after such use to insure accuracy in normal operation. Demagnetize as follows a. Connect oscillator to bridge in the normal confiqura- READOUT MULTIPLIER tion. b. Set oscillator FREQUENCY to 100 Hz, OUTPUT VOLTAGE RANGE to 50, OUTPUT ADJUST to zero. Set bridge TERMINAL SELECTOR to CAL. XlO 3-22 OPERATION X 100 X 1 200 mA 20 2 • An elastic limit ~ for best resolution these values are too high; for some measurements, currents of 5 'MAX may be tolerable. Theory-Section 4 4.1 4.2 4.3 4.4 4.5 4.6 4.7 INTRODUCTION ..... PROPERTIES OF CAPACITORS BASIC BRIDGE CIRCUITRY . CIRCUITRY OF THE 1616 BRIDGE C·STANDARDS ACCURACY G·STANDARDS ACCURACY RATIO ACCURACY 4-1 4·1 4-5 4·7 4·9 .4·10 . 4-11 4.1 INTRODUCTION. 4.2 PROPERTIES OF CAPACITORS. The 1616 Precision Capacitance Bridge is a standardslaboratory instrument of exceptionally high precision Together with the other components of the 1621 Precision Capacitance-Measurement System. it is designed for accurate measurements of capacitance, conductance, and therefore the properties of dielectrics, as well as high-resolution comparisons among capacitance standards This section deals with the 1616 bridge For theoretical discussions of the 1316 Oscillator and 1238 Detector, please refer to their individual instruction manuals 4.2.1 BasicComponents of Capacitance. Figure 4·1. Three Terminals. Most physical capacitors can be precisely represented by the three capacitances shown in Figure 4-1. the direct capacitance, C H L' between the terminals Hand L (capacitance between the plates of the capacitor), and the two terminal capacitances, C H G and C L G , from the corresponding terminals and plates to the case, surrounding objects and ground (to which the case is connected to either conductively or by its relatively high capacitance to ground) A 3-terminal capacitor (Figure 4-2) has connected to the G terrninal a shield that cornpletely surrounds at least one of the terminals (H), its connecting wires, and its plates except for the field that produces the desired direct capacitance to the other terminal (L) Changes in the environment and the connections can vary the terminal capacitances, usually C H G and C L G , but the direct capacitance C H L referred to simply as the capacitance of the three-terminal capacitor - is determined only by the internal structure. Figure 4·1. Schematic diagram of a capacitor, showing the direct capacitance and the associated terminal capacitances. f----f .. IL Figure 4·2. Structure of a 3·terminal capacitor with 2 coaxial connectors. This direct capacitance can be calibrated by 3-terminal measurement methods, utilizing guard circuits or transformer-ratio-arm bridges, which exclude the terminal capacitances. The direct capacitance can be made as small as desired, since the shield between terminals can be complete except for a suitably small aperture The losses in the direct capacitance can also be made very low because dielectric losses in the insulating materials can be made a part of the terminal impedances. When the 3-terminal capacitor is reconnected as 2-terminal, the 2-terrninal capacitance will exceed the calibrated 3-terminal value, C H L, by the terminal capacitance C H G . Two Terminals. In the common 2-terminal connection, the capacitor has Land G terminals connected together, ie, L terminal connected to case The terminal capaci- THEORY 4·1 m oklJ tance C L G is thus shorted, and the total capacitance is the sum of C H Land C HG. In general, since one component of the terminal capaci- tance C H G is the capacitance between the terminal and surrounding objects, the total capacitance is changed by changes in the environment of the capacitor and particularly by the introduction of the wires required to make connection to the capacitor. (o) -x STRUCTURE, IT MUST BE UNGROUNDED Lb 1 CAPACITANCE COMPONENTS NOTE C" There is a possible confusion between Hand L. Usually the 2-terminal capacitor is labeled H for the more completely shielded connection, L for the one connected to shield (if any) However, the termi nals of the 1616 bridge are labeled the opposite way L for the inner, H for the outer part of the 2-terminal port Bridge H connects to capacitor Land G, but G must not be grounded. The uncertainties in the calibrated value of this 2-terminal capacitor can be of the order of tenths of a picofarad, if the geometry, not only of the capacitor proper but also of the environment and of the connections, is not carefully defined and specified. For capacitors of 100 pF and more, the capacitance can usually be adequately defined for an accuracy of a few hundredths percent, if the terminals and method of connection used for cal ibration are specified For smaller capacitances or for higher accuracy, the noncoaxial 2-terminal capacitor is seldom practical. The 3-terminal arrangement is generally preferred. Nevertheless, there are very accurate capacitors (in all but the smallest sizes) with 2, coax ial, ungrounded terminals. Two Terminals, Coax Connection (Figure 4-3) In the coaxial 2-terminal structure, capacitor terminal L is again connected to the case, but not ground The case is a complete shield around the field that determines the direct capacitance C H L. Fields outside the case contribute only to C LG, which can be excluded from the measurement So most environmental changes have little effect on accuracv. If the bridge is designed to measure such a capacitor and each incorporates a suitable precision coaxial connector. the uncertainties of measurement can be as low as 100 attofarads. The realization of such precision depends largely on exacting definition of the boundary between capacitor and bridge. This boundary is represented in Figure 4-3 by line XB. Physically. the boundary is the plane of the mating face of the outer conductor of the G900 connector. i.e.. the reference plane of the connector. For precise measurements. the electric field must be entirely parallel to that reference plane so the portion represented by C is unequivocally part x2 of C. whereas C and C are not. x 4-2 THEORY Fx 0 C" --IE--~ 8 C, Co -1E- -IE- ~ tel BRIDGE (e1) CAPACITOR BRIDGE II~ ON Figure 4-3. A coaxial 2-terminal capacitor, its structure, component capacitances, and connection to a bridge. The fringing capacitance C F x makes the freestanding capacitance of the capacitor slightly greater than C, How ever, C; is the value we always measure, calibrate, and refer to as the capacitance of the device. The fringing capacitance C makes the bridge terminal capacitance slightly greater ~~en open- circuited than the value C which obtains when a suitable capacitor is attached. ADJUST cannot be so accurately set with the The UNKNOWN port open as with a standard coaxial 2-terminal capacitor installed. Limitations on the accuracy with which GenRad can specify C involve uncertainty as to the position . of spring-loaded partsFB.In the G90 0 connector an d reduce d resolution because of "noise" picked up by the unshielded LOW terminal. However the repeatability is much better than the tolerance given C in para. 3.4. i.e.. ±8 fF. zmo FB 4.2.2 Inductive and LossyComponents. Figure 4-4. No physical capacitor is ideal, ie., free of inductance and dissipation, although some are excellent So also, the equivalent networks that are used to represent a nonideal capacitor do so imperfectly However, they are satisfactory in many instances, particularly if some of the parameters are understood to be quasi-constants (They may vary somewhat with temperature, humidity, pressure. frequency, acceleration, aging. illumination, etc.) Generalized Circuit. Figure 4-4 (a) represents the nonideal direct capacitance (C H L of Figure 4-1) with 5 lumped constants R represents the metallic resistance in the leads, supports and plates; L, the series inductance of the leads and plates; C I , the capacitance between the plates; C K' the capacitance of the supporting structure Conductance G represents the dielectric losses in the supporting insulators, the losses in the air or solid dielectric between capacitor plates, and the doc leakage conductance. For most purposes, C, and C K are added as C (Notice that C K is zero if support-structure capacitance is entirely with in C H G and/or CL G ) 1, HIGH r more lag in IT By itself, G has no effect on apparent capacitance because the bridge resolves them separately Equivalent Parallel Circuit. Figure 4-.4 (c) is the equivalent circuit based on the 2 components measured directly by the bridge Terminal properties (E T, IT, (J) are identical with diagrams a and b Diagram d shows the vector relationships The algebraic relationships between these components and those of diagram a are given below, followed by simplifying approximations that are convenient for studying the effects of one parameter at a time. It is the capacitance C; (not CSE ) that we generally mean by the term "apparent capacitance". C (1 - w 2 LC) C =------------2 x ET (1 I G LOW G2 L w LC + RG)2 + (wGL + WRC)2 G (1 + RG) + w 2RC 2 w 2 LC + RG)2 + (wGL + WRC)2 = -------------- x (1 (e) There is a resonance between Land C at a frequency f o . We are concerned only with the condition f f o (and C; is positive) At sufficiently low frequency (f« f o ), we observe that w 2 LC « 1, but various losses may still make C; =I=-C, thus < I, Ie;. i:-fl' I C, REFERENCE DIRECTION L OF E,. (c) (f) '---"'-'-'1 ERs·RSI,. I I --REFERENCE DIRECTlQN I G "':: I I I x I nu __ I I I I x C - G2 L C - G2 L "'::---- (1 + RG)2 + (wGL + WRC)2 OF IT I __ C "':: G(1+RG)+w (1 + RG)2 2RC 2 (1+RG)2+(wGL+wRC)2 "'::--- G 1 + RG E. Figure 4-4. Capacitor lumped-parameter circu its and vector diagrams. equivalent The corresponding vector diagram, Figure 4-4 (b) shows how the capacitive and conductive current components add to make IT, and how the capacitive, inductive, and resistive voltage components add to form ET. IT leads ET by the phase angle (J, between 0° and 90°, as is characteristic of a physical capacitor. (We are not interested in conditions at such high frequency that IT lags ET) The complement of (J is 8, the dielectric loss angle Notice that, regardless of the relative magnitudes of the other vectors, a small inductive component (IELI < IEel) always makes lET I < IEeI In other words, the presence of series inductance makes a capacitor apparently increase in capacitance as frequency (or the inductance) increases. The effect of series resistance on apparent capacitance is opposite, because ER tends to make IETI > IEel But the effect of shunt conductance G on apparent capacitance is mostly to accentuate the effect of R, if any, by causing where the fractions at the right are valid at sufficiently values of Rand L. low < fo If, however, we relax the frequency restriction to f but consider the case of negligible series loss (wRC « 1) and reasonably small shunt loss (G < WC), then C; increaseswith a term proportional to f2 as follows With this information, the capacitance at, for example, a frequency of 2 MHz can be con puted with high accuracy from the calibrated value at 1 kHz. For fifo up to 0.3 or so, the accuracy may be greater than that of a measurement at 2 MHz because of the difficulties in determining the measurement errors produced by residuals in the connecting leads outside the capacitor (unless, of course, one uses a bridge particularly designed for measurement at 2 MHz) You may use a grid-dip meter for measuring f o, with the capacitor terminals shorted. THEORY 4-3 There are secondary effects on the apparent value of capacitance in a 3-terminal measurement that will not be described here except to say that they involve resonances among Land C components in all 3 basic parts of the network of Figure 4-1 (The series inductor in Figure 4-4 (a) may represent, at least in part, a wire carrying current not only to CH L but also to CH G or part of it) These effects are generally negligible if the following approximation is valid within the desired measurement accuracy Q tan e= wC x 1 5 x 3. Power factor is the ratio of the real power (watts of dissipation) to the product of rrns voltage and current (voltarnperes) P. F = cos ERs e E = sino T where f' a is the lowest resonant frequency of the network Estimate f'o as the free ringing frequency of the largest C component, such as C H G' (given some initial energy) with the capacitor-lead and cable inductances; assume the bridge HIGH terminal is virtually grounded and its LOW terminal is an open circuit. Equivalent Series Circuit. Figures 4-4 (e) and 4-4 (f) present another 2-component equivalent circuit which may be appropriate if Rand C are the only significant components of diagram a However, the 1616 bridge does not I n terms of the parallel equivalent measure Rs directly circu it Loss Factors. An important characteristic of a dielectric material, and hence of the capacitor made from it, is the ratio of energy dissipated to energy stored, per cycle of ae There are 3 commonly used "factors" to express the same characteristic. 1. Dissipation factor or loss tangent 0 is the ratio as Notice that several other convenient expressions can be taken from Figure 4-4, such as Q or 0 in terms of current components in diagrams c and d or voltage components in diagrams e and f Other convenient expressions are given in para 36, for use of 0 or Q in converting data from series-equivalent to [)<.rallel-equivalent circuit forms and vice versa. NOTE For low-loss capacitors, the difference between o and PF is very small For 0 1CfI , 1Cf2, and 1Cf3 , (0 PF )/0 is 05%,005%, and 5X 1Cfs%, respectively 4.2.3 Frequency Characteristics. The lumped-parameter equivalent circuit such as Figure 4-4, a, for a real, physical capacitor is most appropriate under the conditions in which the parameters (R, L, C, G) are constants I n the following examples are some situations in which they are, others in which we have to treat them as quasi-constants, or develop special "constants". Capacitance. At high frequencies, the inductive effect predominates (C; increases with f2) At low frequencies, C; is essentially constant if Q is very high and the dielectric is air. But if the capacitor has a solid dielectric, such as mica, it is responsible for a slight capacitance change with frequency; C 1 is then a quasi-constant 10 defined above fZ w li It 1'0 I Expressed in terms of the components of Figure 4-4 a + w 2RC 2 0=--'----·--wC (1 - w 2LCl - wG 2L G (l+RG) - ---r-- 0.0 I 10Hz 100 1kHz 10 / J --. 100 / IMHZ "----C 10 .--- FREQUENCY Figure 4-5. Capacitance vs frequency characteristic of I f frequency is well below resonance and if both series and shunt losses are small, a good approximation is G 0:=::::-+ wC RwC . 2 Storage factor or quality factor Q is the ratio of energy stored to energy dissipated per cycle of ac 4-4 THEORY a mica capacitor. The change in capacitance C; with frequency, of a 1000-pF capacitor with mica dielectric, is shown in Figure 4-5. The dashed line slanting downward to the right represents the change in the dielectric constant of mica resulting from interfacial polarization: that slanting upward to the right shows the change in effective capacitance C x resulting from series inductance. Though the former phenomenon may be true at much higher frequency, it can be neglected above 300 kHz in this example, because there the latter effect is so predominant. The magnitude of the change at low frequencies depends upon the dielectric material and is, for example, much smaller for polystyrene than for mica Dissipation Factor. At high frequencies, "series" loss predominates because of skin effect in the leads and plates (as explained below) At low frequencies, "shunt" loss predominates, but G may be far from constant. If goodquality, low-loss, solid dielectrics are used, G is nearly proportional to w; then it may be useful to treat G/wC as the quasi-constant DZ The skin effect can be represented by considering R a variable, R = Rz J(;J, in the second term, RwC, of the formu la for 0, then effective conductance at w capacitor 1. Thus, for the variable air At very low frequencies, a "shunt" loss conveniently represented by G may be significant. Leakage conductance G is usually negligible at frequencies above a few Hz and is important only when the capacitor is used at dc for charge storage. The dominant components of 0 at audio frequencies are the dielectric losses in the insulating structure and in the dielectric material between the plates 4.3 BASIC BRIDGE CIRCUITRY. * 4.3.1 Elementary Capacitance Bridges. Measurements of capacitance, particularly those of high accuracy, are made by a null method that uses some form of the basic ratio bridge. shown in Figure 4-7 The capaci- 0= Dz + RZCw 2 /3 The dissipation factor as a function of frequency for a mica capacitor is shown in Figure 4-6 There 0 is the sum -2 10 I 0:: ~ ~163 z o ~164 o, f M 1N -- ~ - - -- - -- (f) (f) -- -- -- - - -- 100 I kHz 10 .- / - °10. 5 10 Hz ! /, _1 __ , / 100 / Figure 4-7. An elementary -- -- -- -- - - -- I MHz - -- 100 FREQUENCY Figure 4-6. Dissipation factor vs frequency characteristic of a mica capacitor. of three principal components a constant dissipation factor caused by residual polarizations and shown by the horizontal dashed line; a loss produced by interfacial polarizations, which contributes the 0 shown by the dashed line slanting downward to the right; and an ohmic loss with sk in effect in the leads and plates. which results in a 0 proportional to the 3/2 power of frequency and is shown as the dashed line slanting upward to the right The total dissipation factor has a minimum value at a frequency that depends on the size of the capacitor; f M I N varies inversely with capacitance and ranges from 1 k Hz to 1 M Hz for capacitance values from 1 pF to 100 pF. In an air capacitor. the losses in the air dielectric and on the plate surfaces are negl igible under cond itions of moderate humidity and temperature The loss is, therefore, largely in the insulating supports, and the above formula applies. However, if the capacitor is variable. and the varying part is lossless, D Z varies inversely with C Then it is preferable to treat G/wC as G I w/wC, where G I is the capacitance measuring bridge. tance of the unknown, Cx,s. balanced by a calibrated, variable. standard capacitor, C s' or by a fixed standard capacitor and a variable ratio arm, such as RA . Such bridges with resistive ratio arms and with calibrated variable capacitors or resistors can be used over a wide range of both capacitance and frequency and with a direct-reading accuracy which seldom exceeds 0 1% For higher accuracy. resolution. and stability in capacitance measurements at audio frequencies, a bridge with inductively-coupled or transformer ratio arms has many advantages, and increasing use of transtorrner-ratio-arrn bridges is being made in the measurement of many types and sizes of capacitors 4.3.2 Transformer-Ratio Bridges. The advantages of transformer ratio arms in a bridge are that accuracies within a few parts per million are not difficult to obtain over a wide range of integral values, even for ratios as high as 1000 to 1, and that these ratios are almost unaffected by age. temperature, or voltage The low impedance of the transformer ratio arm also makes it easy to measure direct impedances and to exclude the ground 'Thomas, Instruments Cliffs, H. E., and Clarke, and Measurement N. J. (1967) C. A .. Handbook of Electronic Hall, Englewood Techniques, Prentice p 36·57. THEORY 4-5 impedances in a three-terminal measurement without the use of guard circuits and auxiliary balances. To illustrate these characteristics, a simple capacitance bridge with transformer ratio arms is shown in Figure 4-8. On the toroidal core, a primary winding, connected to the generator, serves only to excite the core; the number of primary turns, N p , determines the load on the generator but does not influence the bridge network. If all the magnetic flux is confined to the core - as it is to a high degree in a symmetrically wound toroid with a high-permeability core - the ratio of the open-circu it voltages induced in the two secondary windings must be exactly equal to the ratio of the number of turns The ratio can be changed by the use of taps along the two secondaries, but, when the ratio is fixed, the voltage is highly invariant Changes in the core permeabil ity with time and temperature have only very small effects on the effective ratio. It depends not only on the turns ratio (a perfect integer) but also on leakage flux, wh ich is not confined to the core in a practical transformer. The ratio is, therefore, both highly accurate and highly stable. \ I, the terminal G connected to the junction of the ratio arms. The capacitances between Land G shunt the detector, so that they affect only the bridge sensitivity The capacitances between Hand G are across the transformer windings. To the extent that the transformer can be assumed ideal, i.e., with no resistance in the secondary windings and with no flux that does not link equally both secondaries, the current drawn by the H-G capacitances does not change the voltages V 5 and V x or the balance cond itions In practice, the transformer resistances and leakage inductances can be kept so small that quite low impedances or large capacitances can be connected from H to G before there is appreciable error in the bridge. The junction of the ratio arms, G, is therefore a guard point, or guard potential, in the bridge All capacitances to G from the H or L corners of the bridge are excluded from the measurement In the three-terminal capacitors represented by the H, L, G terminals in Figure 4-8, the bridge measures only the direct capacitance. C x' of the unknown in terms of the direct capacitance, Cs ' of a standard without additional guard circuits or balances. One can take advantage of the accurate and stable ratios of the transformer by designing a bridge with an "unknown" arm that is fixed and a ratio that can be varied to balance the bridge Far greater measurement accuracy is feasible with such a design approach (making C, a fixed capacitor rather than a variable one) For example, consider the following alternatives. Figure 4-9 shows three of the possible ways of balancing a simple transformer-ratio capacitance bridge. For simplicity, the generator and primary are not shown, but it is assumed that the two secondaries have 100 tu rns each and are excited so that there is 1 volt per turn. The capacitor in the unknown arm is assumed to be 72 pF IOOV Figure 4-8. An elementary capacitance bridge with transformer ratio arms. -100 V This balance condition is not affected by the capacitances shown from the Hand L terminals of Cs and C x to 4-6 THEORY -70V (0) (b) (e) FIXED RATIO MULTIPLE DIVlDERS SINGLE FIXED CAPACITOR SINGLE DIVIDER MULTIPLE FIXED CAPACITORS VARIABLE CAPACITOR In Figure 4-8, the two transformer secondary windings are used as the ratio arms of the capacitance bridge with the standard capacitor, Cs, and the unknown, Cx' as the other two arms in a conventional four-arm bridge network. The condition for balance, or zero detector current, is easily shown to be that IOCV IOOV Figure 4-9. Circuitry for 3 methods of balancing a transformer-ratio capacitance bridge. In Figure 4-9a, the two ratio arms are equal and the bridge is balanced in the conventional way with a variable standard capacitor which is adjusted to 72 pF The detector current can equally well be adjusted by a variation in the voltage applied to a fixed standard capacitor. In Figure 4-9b, the standard capacitor is fixed at 100 pF This is balanced against the 72pF "unknown", connected to the lOa-Vend of the transformer, by connect ion of the standard to 72 V of the opposite phase, obtained from suitable taps on the transformer windings The inductive divider shown has a winding of 100 turns with taps every 10 turns and, on the same core, another winding of 10 turns tapped at every turn. If, as shown, the second winding is connected to the 70 V tap on the first winding and the capacitor to the 2-V tap on the second winding, the required 72 V is appl ied to the capacitor. Six or more decades for high precision can be obtained in a similar fashion with more windings on one core and the use of additional transformers driven from the first. Such inductive dividers have very accurate and stable ratios, but the errors increase with the nu mber of decades because of loading effects Another method of balance by voltage variation is shown in Figure 4-9c, where a single decade divider is used in combination with multiple fixed capacitors. The lOO-turn secondary is tapped every 10 turns to provide lO-V increments If, then, a 100-pF capacitor is connected to the 70 V tap and a lo-pF capacitor to the 20 V tap, the resulting detector current balances that of the 72-pF "un known", connected to 100 volts. This bridge can be given 6-figure resolution, for example, through the use of 6 fixed capacitors in decade steps from 100 pF to 0001 pF, each of which can be connected to anyone of the taps on the transformer. In any of these bridges, the bridge ratio can also be altered by use of taps on the "unknown" side of the transformer to select the voltage applied to the capacitor being measured. For example, a 720-pF capacitor would be balanced in any of the circu its of Figure 4-9 if connected to a lO-turn (10 V) tap on the upper wind ing The range of the bridge can thus be extended to measure capacitors which are much larger than the standards in the bridge These advantages of transformer ratio arms and dividers make possible a bridge of wide capacitance range, and high accuracy, one that is usefu lover a wide frequency range It is economically reasonable to construct the relatively few, fixed capacitance standards to have the necessary stability and accuracy for such a bridge, one that will measure with .01% accuracy over 6 decades of capacitance and 3 decades of frequency. At low frequencies, a limit is imposed on sensitivity by the maximum voltage obtainable from the transformer, since, for a given core, the voltage at saturation is proportional to frequency. At high frequencies there is a decrease in accuracy resulting from the decrease in core permeability with frequency, from the increased loading of the transformer by its self-capacitance as well as the bridge capacitances and, of course, from the usual residual capacitances and inductances in the bridge wiring and components. 4.4 CIRCUITRY OF THE 1616 BRIDGE. Figure 3-3. In general, this capacitance bridge is based on the 3rd method, Figure 4-9, C Each of the 12 internal capacitance standards is calibrated at a cardinal value and each is switched to an appropriate tap on the ratio transformer as you bring the bridge to balance. Conductance (or loss in the capacitor being measured) is balanced by the same method. Conductance standards of fixed, cardinal values are switched to appropriate taps on the same ratio transformer The circuitry is described in more detail in the following paragraphs. 4.4.1 Excitation. The application of sufficient voltage (power is mostly reactive) from the oscillator to the ratio transformer provides a test signal so the detector will respond to an unbalance The level of excitation is limited, always below the saturation level of the transformer core. The primary winding of 200 turns is driven directly by the oscillator POWER OUTPUT port. Several design features eliminate stray coupling to other parts of the bridge and particularly to the detector. 1. The GENERATOR INPUT port is connected to the transformer by a shielded wire. 2. This shield is floating in the bridge (grounded only at the oscillator) 3 In the ratio transformer, the primary winding is surrounded by an electric shield that is grounded In the bridge 4.4.2 Circuit for the Unknown. Figure 3-1, 3-2. The 2 terminals of C x , HIGH and LOW, must be connected, respectively, to the ratio-transformer secondary (upper half, in the simplified diagrams) and the summing point (which drives the detector) Because the midpoint of the ratio-transformer secondary winding is grounded, both terminals of C x must be floating However, the 3rd terminal or shield of the capacitor under test is grounded at the 3-terminal port The facility for measuring capacitors larger than the largest standard, as mentioned above, obtains when the TERMINAL SELECTOR switch points to XlO or X100 Then, the HIGH side of C; is connected to the 20- or 2-turn tap, respectively, of the secondary winding When the selector indicates Xl, the connection is to the 200th turn from ground. Then the voltage across the unknown capacitor at balance equals the oscillator voltage Each of the UNKNOWN terminals is disconnected from the bridge and grounded whenever the TERMINAL SELECTOR selects CAL. When it selects 3-TERMINAL, the 2-terminal port is likewise grounded, and vice versa 4.4.3 CapacitanceStandards. Figure 3-3. There are 12 capacitance standards, one for each mu It ipie of 10 from 100 nF down to an effective value of 1 aF The "high" side of each is connected to the front-panel selected tap on standards side of the transformer secondary THEORY 4-7 (generally the lower half in the simplified diagram) The "low" side of each standard is permanently connected to the same summing point mentioned above, which drives the detector. However, the 2 largest C standards can be disconnected from the summing point by dropping the C MAX lever switch. Thus, stray capacitance shunting the detector can be reduced while you measure extremely small values of increments of C, Each of the first 6 CAPACITANCE lever switches indicates values in tenths of the value of its associated standard. Thus, the lever with a 1, 0, 1,2, ... 9, XpF readout actually switches the lO-pF standard capacitor The max imu m current through each of these standard capacitors (when the corresponding readout is X, i.e.. ten) obtains when its HI G H side is connected to the transformer 0 secondary at the 200th turn, 180 out of phase from the unknown-capacitor-X 1 connection. That is the bottom of the transformer as. shown in simplified diagrams The intermediate taps are 20 turns apart The zero position is a ground connection. The 1 position is a connection to a 2G-turn tap on the "unknown" side of the transformer secondary. In order to provide greater stabil ity in the set of capacitance standards it is preferable to use a moderate valued capacitor connected to fewer turns than an extremely small valued one connected to the usual number of turns. Therefore a set of taps is brought out from the ratio transformer at intervals of 2 turns (to a maximum of 20 turns) Each capacitor that connects to this set of taps is 10 times as large as it would have to be if it were connected to the usual (20-turn-per-step) taps The last 6 "C" lever switches span 2 turns per step. The 6 associated standard capacitors are each 10 times as large as the effective values shown in Figure 3-3. Therefore, each of the last 6 CAPACITANCE lever switches indicates values in hundredths of the value of its associated standard Thus the lever with a - 1, 0, 1, 2, 9, X -aF readout actually switches the 100-aF standard capacitor For these, of course, the - 1 position is a connection to a 2-turn tap on the "unknown" side of the transformer Physically, the transformer is manufactured with a separate winding of 22 turns, tapped every 2 turns for the small C standards. A set of 20 windings of 20 turns each serves the large C standards and the "unknown", by a series combination that acts as a center-grounded 400-turn secondary Except for the primary (200 turns) all these windings are very intimately coupled by multifilar con struction Each C standard is cal ibrated to the desired value by means of a trimmer. All 12 are located behind the locked door in the front panel (Figure 1-2), where they are labeled with the effective values of the corresponding standards, as shown in Figure 3-3 Calibration procedures are given in Section 5. 4-8 THEORY 4.4.4 ConductanceStandards. Figure 3-3. There are 5 conductance standards and a multiplier circuit for reducing their effective values in steps, so that you have virtually 11 conductance standards, one for each multiple of 10 from 10 p.S down to 1 fS Like the capacitance standards, the conductance standards connect as follows high side, via lever switches, to taps on the standards side of the transformer; low side, via multiplier circuitry to the summing point that drives the detector. Because this bridge is designed for the greatest precision in C rather than G measurements, it is preferable to have the main (20-turn-per-tap) secondary winding carry only C-standards currents Therefore, the G standards are connected to the other (2-turn-per-step) winding described before. Consequently, each of the 5 conductance standards is 10 times as large as the effective values shown in Figure 3-3. It follows that each of the 5 CONDUCTANCE lever switches indicates values in hundredths of the value of its associated standard. Thus the lever with a - 1, 0, 1, 2, ... 9, X . p.S readout actually switches a 100 p.S standard of conductance, i.e.. a precision 10-kSl resistor When the G multiplier is set to Xl, the connection from the set of 5 conductance standards to the detector is direct When that setting is X 10- 1 , a resistor network passesonly 10% of the conductance standards' current to the detector summing point (The other 90% returns directly to ground) Similarly, at XlO- 2 , l(Yo passes to the summing point The "box" (with these G standards and networks) also contains pots for setting the multiplier ratios exactly to the appropriate powers of 10 R 18 for 10- 1 , R 19 for 10-2 etc. These pots are intended for calibration purposes as described in section 5. Ideally, the current through each conductance standard should be perfectly in phase with the ratio- transformer voltage. However, the resistor is bound to have some endto-end capacitance. For example. R 1 has some stray C in parallel. Without compensation, the current through this stray C affects the bridge balance condition like increasing the value of the C standard. The compensation is represented schematically by C23, connected between the midpoint of R 1 and ground Refer to Figure 5-9 Physically, C23 is formed by a threaded, spring-loaded sleeve surrounding the main part of R 1 and adjustable by turning with a special wrench. The current component from the transformer through C23 leads by a large angle (nearly 90°) causing the voltage at the midpoint of R 1 to lag slightly with respect to the ratio-transformer voltage If the compensation is adjusted correctly, the resulting lag in the current frorn the middle of R 1 to the detector is just sufficient to cancel the leading current through the stray C Not only is each of the 5 G standards compensated in this way, but also 1 of the 4 series resistors in the G multiplier network Thus, R9 passesthrough a similar sleeve with capacitance represented as C22. There is, of course, more stray capacitance to ground in the G·standard circuit when the G multiplier is set to some multiple other than X 1. Consequently, in some of the positions of that switch, the necessary compensation to achieve zero phase shift is the addition of capacitance shunting the series resistor. So, for example, for the io ' multiplier, R6 has the adjustable C18 in parallel The desired result is complete independence of C and G balances, and hence measurements A measure of the quality of the compensation is Do, the ratio of conductance to lk Hz parallel susceptance of Cc of the G standard G s (We refer to the effective conductance, including the multiplier network if it is involved, and the effective susceptance, including the compensation described above i Refer to para 3.9.3 Do Gs ~--- 21Tl03 c, Because of the compensation described above, Cc is very small and can be negative Thus, the factory adjustments assure that the rnaqnitude of Do is very large for any G·standard and G·mu Iti plier combination you chose Because it is extremely difficult to obtain external standards of conductance with zero or known shunt capacitance, we do not recommend any readjustment of the "C" compensation in the G box, i.e.. do not turn any of the sleeves, do not replace any resistor passing through a sprinq-loaded threaded sleeve in the G box. C18 and C20. C27 are "factory adjustments" 4.4.5 External Standards. An external standard, or whatever you connect to that port, is connected like a capacitance standard (a large one) The LOW terminal is grounded, when EXT MUL TIPLIER is set to OFF, and otherwise connected to the detector summing point. The HIGH terminal also connects to ground in the OFF position, as well as in the position HIGH connects to the ratio- transformer in steps of 20 turns Unlike the CAPACITANCE lever switches, the EXT MU LTI PLIE R indicates directly the fraction you select of a the external standard (in farads and Siemens). Thus. the settings of .1, 0, 1,2, .9, 1.0 indicate (respectively) that - 0.1, 0, 01, 0.2, 09, 1 times the external standard C and G are added to the CAPACITANCE and CONDUCTANCE indicated by the lever switches. 4.4.6 Zero Adjust. Figures3-2, 6-3. This front panel adjustment is primarily intended to compensate the bridge for a few pF of terminal capacitance when you chose a 2·terminal position of the TERMINAL SELECTOR switch The adjustable component is the pot R24, connected across 2 turns of the ratio transformer (on the standards side), with its wiper arm capar.itively coupled through C17 to the detector summing point. Therefore, a capacitive balance can be made without moving the Clever switches and a "zero offset" can be chosen so that the readout is zero for some convenient reference condition in 2·terminal measurements. (See para 3.4.2.) A second function of the ZERO ADJUST control is to compensate the bridge for a few aF of variation in effective zero offset among the CAL, 3·TERMINAL Xl, and XlO positions of the TERMINAL SELECTOR switch. (In the latter 2 positions, the 3·TERMINAL UNKNOWN L.OW connector must be shielded with a Type 874WN OpenCircuit Termination to remove the 125 aF or so of termi nal capacitance from the bridge) The circuit from R24 through C17, although it is now disconnected from the detector and grounded by the TERMI NAL SE LECTOR switch, neverthe less couples enough signal to the summing point for the purpose described before (Notice that the range of adjust ment is now reduced by a factor of 10 6 or so.) A coarser adjustment, C301, very lightly couples some signal Of the opposite phase to the detector (via the G box) and is provided so you can make sure that the ZERO ADJUST control, sornewhere in its range, reaches the desired zero condition for each of the first 3 positions of TERMINAL SELECTOR. (Each position will require a different ZERO ADJUST setting) The obvious utility of this second function is to make the CAPACITANCE readout true, down to the 12th digit, so you need not add a "zero correction" to each measurement. However, accuracy in that digit is insignificant for C, above about 100 fF (01 pF) Therefore, only if you must measure such small capacitors will you need this function regularly. It does have utility in recalibration of the bridge, particularly in trimming the smallest internal C standards with the larger ones as references. 4.5 C-STANDARDS ACCURACY. 4.5.1 Calibration. The set of 12 internal capacitance standards can be cal ibrated qu ick Iy and accurately by a series of comparison balances starting with a single external standard capacitor of almost any size within the range of the bridge. Since the 8 figure resolution of the bridge permits comparison with a precision of 1 ppm down to 1 pF, the accuracy of calibration is usually determined by the accuracy of the standard. a Only one external standard is required, most conven iently a 3 terminal 100·pF standard, such as the GR Type 1404·B. With a test frequencv of 1 kHI, the accurate, Internal 0 1 transformer ratio can be used to ensure accurate decade ratios of the internal capacitance standards The ~ 1 position of any capacitance balance switch connects the corresponding internal capacitor to a 20·turn tap on the "unknown" side of the ratio trans THEORY 4·9 former. This capacitor can be compared with the next lower decade capacitor, which is connected to the 200-turn winding on the standard side when the corresponding lever is set to the X position. Any adjustments required can then be made with one of the trimmers accessible beneath a hinged cover on the front panel Refer to para 54 Such checks or recal ibr ations of the bridge need not be made often The capacitors are constructed to be so stable that after cal ibration they may be expected to change less than 10 ppm per year in normal use. The temperature coefficients of the internal standards are stated in the Specifications and the consequent effect of temperature on accuracy is summarized in para. 3.9. 4.5,2 Sealing. The "air-dielectric" standards, except the 4 smallest, are filled with dry nitrogen and sealed. If they were open to the atmosphere, their capacitance would change 20 ppm for each 10% change in humidity and about as much again for each 300 meters of elevation or 3% change in barometric pressure because of the weather. Sealing practically eliminates these infl uences. The 2 largest capacitors, with selected mica dielectric, are also sealed to keep them dry and stable with respect to the effects of humidity cient a is therefore 6C RM Example 1 what is the fractional change of the 3rd-decade internal capacitance standard during a 5-minute measuring process, if the environmental temperature is known to be maintained in the range 23° ±l°C) The worst ° possible case is 0 = 22°C and 60 e = 2°C; From the specifications, a representative value for a is 3 ppmtC; the time constant for the lag box in the bridge is 360 minutes (6 hours) So 6C C = (3 ppm) ( 5 min I (2° C) EO °c 360 min = 083 ppm Example 2 What is the effect of a change, 18 hours ago, from 10° to 23°C) 4.5.3 Thermal Lag. The 8 largest internal capacitance standards are housed in a "lag box," a thermally insulated container that comprises most of the "C box," within the bridge For the purpose of estimating transient temperature effects on precision of measurements, we represent the thermal properties of the lag box simply as follows 1 The mass M inside the box is all at temperature j 2 Heat is transferred through massless insulation of thermal resistance R Then the lag box will respond to a step change 60 e in to 0) at time zero environmental temperature (from 0 with a simple, exponential change of internal temperature 0, as follows (A close analogy exists to a series R-C circuit. capacitance being analogous to M, applied vokaqs to 0, and the capacitor voltage to OJ For the electrical circuit. the time constant IS RC, for this thermal model, RM ) ° ° where E IS 2718 and t IS time (in the same units as RMI Consider rate of change of inside temperature at any t I me t d 60 e dt OJ = RM 4·10 THEORY During any time interval 6t, that is very small compared to the time constant, the internal temperature change is that rate of change multiplied by 6t The corresponding fractional C change for a capaci tor with a temneratu re coeffi- RM E 4.6 G·STANDARDS ACCURACY. The basic conductance accuracy of 01% depends on precision resistors and multiplier networks that can be calibrated against external standards The very high degree of independence between G and C balances depends on factory-set compensation in the G box, measured in terms of 0o, and the minute losses in the C standards, which are specified As indicated in the parts list, the greatest accuracy is provided in the standards for the first 2 positions of the G readout, R 1 and R2 The nex t, R3, has a tolerance of ±O 1%; the following, R4 and R5, of ±1% Intercomparisons can easily be made using the bridge ratios, in a manner simi lar to the cal ibration of C standards You can measure the error in R5, for example, (against R 1 or R2) but no calibrating adjustment is provided Adjustments for the multiplier network are described in para 5.4. ,Essentially, one makes the bridge balance with the correct G readout while a standard external conductance (or network) is connected to the UNKNOWN port The factory adjustments make 00 at least 103 for any G readout This means. for example, if the unknown capacitor has a 0 of 104 , the 1kHI error in C measurement due to stray capacitance in the G standards will not exceed 0 1 2 ppm However, if the unknown conductor has a 0 of 10 at 1 kHI, the error in C measurement may be as much as 10% Refer to para 39 and 44 As a rule of thumb, at 1 kHz DC C 1 00 0 o Do that approximation is invalid for small numbers of turns CG represents stray ground capacitance that is not associ ated with the transformer, but appears in parallel with CG T at any given setting of the switches Typical parameters are given in Table 4-1 These residual impedances make the voltage V applied to the capacitor C, or C; differ from the voltage E induced in the transformer winding For small residuals, the relationship between these voltages is 4.7 RATIO ACCURACY. To measure accurately with this bridge, you must properly balance a well defined unknown capacitor against accurate standards via accurately known ratios Section 3 deals with system operation; 42, definition of the device being measured, and 4.5 accuracy of internal standards The following paragraph deals with the accuracy of significant ratios in the bridge and their effect on your measurement The transformers in the Type 1616 Precision Caoacitance Bridge are not qu ite the ideal transformers shown, for example, in Figures 1-1 and 3-3 The resistance, leakage inductance, and capacitances of the ratio-transformer windings, which are assumed to be zero in the simplified bridge theory, have been kept sufficiently small in the instrument so that errors from these residual impedances are less than 10 ppm for capacitances up to 0 1 IlF at a frequency of 1 k HI However, the residual impedances make the voltages at the transformer terminals differ slightly from the voltages induced in the windings and produce bridge errors that increase with frequency and with the rnagnitude of the measured capacitance The accuracy of the ratios when the transformer is lightly loaded is better than 1 ppm for the unity ratio and is better than 3 ppm for the 0 1 ratio at 1000 HI or lower frequencies The winding self-capacitances act as a more significant load as frequency increases, so that the error in a 0.1 ratio increases to about 30 ppm at 10 kHz and to 05% at 100 kHz The phase errors are, in general, somewhat larger than the magnitude errors of the ratios At 1000 Hz, the phase error is probably within ± 10 11radians, but the error increases in approxirnate proportion to ratio and to the square of frequency 4.7.1 Residual Impedances. where CT = C + CG T + CG, and we use add itional subscripts s or x to designate the standard or "unknown" arm of the bridge For convenience we postulate "effective" values C' of the main capacitor C in each arm (Figure 4-8) C', and C'x are related by the simple balance condition of para. 4.3, ie, they are inversely proportional to the turns ratio. From the circuit of Figure 4-10, for small residuals, and subscripts x or s can be used as before Residuals on the C, and your readout rather low; standard side make C's C x and your residuals on the "unknown" side make C' x readout tends to be high. There is Iikely to be some cancellation of errors, particularly when the turns ratio is 1 1 > > t I E I E ~ GE"I[RATOR Figure 4-10. Simplified bridge schematic diagram showing residual impedances of the ratio transformer. Figure 4-10. The simplified bridge diagram of Figure 4-10 is similar to that of Figure 3-2 but shows, in the equivalent circuit of the transformer, winding resistances (r), leakage inductances (Q), and winding capacitances (C G T ) in both the standard and unknown ratio arms The magnitudes of these residuals vary on the standard side as the lever switches are moved, and on the unknown side as the TERMINAL SELECTOR is reset. The resistance r and inductance Q are approximately proportional to turns in the ratio transformer itself However, because we include the series impedances of intimately associated wiring and switches, ---------- Table 4-1 ---------- APPROXIMATE MAGNITUDES OF RESIDUAL PARAMETERS Transformer turns r 2 06£1 31'H 08 4 10 nF 500 5 1 nF 150 7 10 pF 75 10 100 fF 100 1 fF 35 6 20 15 60 35 200 105 Note Q CGX';:; 10 pF CGT Standard capacitor 100 nF 300 pF CGS 850 pF THEORY 4-11 The difficulty in determining the residual parameters for the many switch combinations makes impractical the detailed calculation of errors for correction of most measurements However, the uncertainty or worst-case errors are presented in para. 3.9. Also, if the capacitance being measured is invariant with frequency, you can determine the error due to residuals for any given measurement with a few ancillary measurements and calculations (Notice that the error term is proportional to frequency squared.l More convenient formulas are given below. An effective conductance G' can be defined, analogous to C'. G' is larger than G by a term that is a hypothetical conductance adding a component of current to the detector summing point (under the ideal conditions of zero rand zero Q) equal to the true current component that flows through C in phase with E (because of r ) We assume the "unknown" is a capacitor: a slightly different analysis would be appropriate if it were a resistor. Subscripts x or s can be used, as before By definition of the primed symbols, they are related by the turns ratio. Usually, only the most significant standard is considered: C' x C' s residual errors from small residuals C =------ x N s C's ~ +w C error 2 Assuming NsC"s N x(l + w2QxCTxl C"s = C s The fractional both arms of the bridge (Qs C T s Nx QXCTxj is [the increment differs from C"sl/C"s' for simplicity, C, in the denominator by which Cs we replace C", with Notice that the coefficient of w 2 can be determined experimentally, for any given C; and switch settings for multipliers and most significant digits, by mak ing 2 measurements at different frequencies, assume C"s constant The effects of residual impedances on accuracy are summarized in 3 observations 1. The effective value of the internal C standards is mu Itipl ied by (1 + k f2) where k is constant as long as the ratio-transformer switches are not changed. 2 In general, k has either sign, and its magnitude is within the bounds indicated graphically in para 3.9 3 Because of the factor f2, the residuals cause the predominant error at high frequency, above 10 kHz Other errors predominate below 2 kHz However, if more than one standard of similar significance is involved, the relationship is generalized as 4,7.2 Examplewith 1:1 Ratio. An alalogous formula can be used for G However, if we assume for simplicity that only one G standard is being used, we can express G and the G error, x (readout - G) in Siemens, thus: x The errors in the bridge readings can be determined by the measurement of a calibrated capacitor. A convenient standard of capacitance and loss is a three-terminal air capacitor that can be connected directlyto the bridge terminalsto add a minimum of series inductance and resistance. For such a capacitor it can be assumed with good accuracy that the capacitance and loss have negligible changeswith frequencyup to 100 Hz and that any changesin bridge readings with frequency indicate bridge errors. Ahhough the dissipationfactor is not generallya simple function of frequency, in a cleanair capacitorthe magnitudeshould be sufficientlylow, i.e., in the tens of ppm, that it can also be neglected. 6G Assuming, similarly, only one C standard is being used, we write the following expression for C, and define a new effective capacitance standard C", which incorporates the 4·12 THEORY The tabulated results were obtained when such a 1000-pF Capacitor was measured on a 1621 CapacitanceMeasurement System; see Table 4-2. Note that the apparent capacitance change at 100 k Hz is only 200 ppm, as compared to the estimated standards-arm ---------- Frequency 10 Hz 100 1 kHz Table 4·2 --------ERROR EXAMPLE. RATIO 1.1 X00500 G readout Relative error C readout 0035S +7 ppm pF 0.136 +2 XOO.495 XOO.493 0 5 15X 102 10- (--1)240 10- 5 10 XOO.488 -5 (-1)884 10- 2 50 XOO.385 -108 1-1)608 io" 100 XOO.164 -329 1--117XO 1 error (W 2 £sCTs) of 5650 ppm The smaller measured error results from a partial cancellation of transformer-impedance errors in the 2 ratio arms of the bridge. When. as in this example. the ratio is 1 1. the residual rand £ are equal on both the standard and unk nown sides of the transformer in Figure 4-10 If the total load, CT' is also the same on both sides. the coefficient (£sCT s - £xCTx) is zero and C, = C", = C, at any frequency In this example. and in general, the errors in the two bridge arms do not quite cancel because the total C T on one side does not equal that on the other The internal and external stray capacitances to ground. CG 5 and CG x' are seldom equal, since the bridge shields and wiring make CGs Cs and the relatively high. Therefore. commonly. C", bridge readout is low For example, other things being > equal. if CGs - C Gx = 100 pF, the bridge would read low by 1600 ppm at 100 kHz This kind of error. being dependent on the internal ground capacitances. is altered by any rearrangement of decade switches to select alternative internal standards. In the following example, the same 1000-pF capacitor is measured with 2 of the several possible decade settings 1 kc X00498 9X0498 100 kc pF XOO 163 pF 998686 For a bridge ratio of 1 1, therefore. the err ors In C and G are typically less than. say, 1000 ppm at 100 Hz Both errors are proportional to frequency squared I t is not usually practical to apply corrections for these errors to the bridge readings, chiefly because the magnitudes of the stray capacitances to ground and their variations with switch settings are not easily determined 4.7.3 The 10:1 Ratio. Bridge errors from transforrner residuals can, in theory, be reduced or eliminated by symmetry in the bridge rirr.uit for any transformer ratio, Just as in the example of the 1 1 £sCT s or £x/£s ratio All that is required is that £xCT x CT/C T x to make C, = Cs ' hence, the residuals should be proportional to the ratio In practice. there are several reasons why the errors. 2 2 W £CT or w rCT Cs' cannot be kept the same on both Sides of the bridge as the ratio is increased The residual impedances rand £ (Table 4-1) differ slightly from proportionality to ratio, even when that is small For high ratios, the rand £ on the low Side can never be less than the minimum values set by the wiring, switch, and terrninal impedances Any wires used to connect the unk nown capacitor to the bridge also increase these residual impedances The capacitance residuals also are seldom proportional to ratio Although the bridge ratio is determined essentially by the ratio of the unknown and standard capacitances. Cx/C s' the error depends upon total capacitance in the bridge arm, ego C T = C + CG T + CG Neither the transformer wind Ing capaci ranees. CG T' nor the ground capacitances of the capacitors, CG , are proportional to ratio. Generally, for a ratio of 10 1. the bridge errors In C may be as high as 03% at 100 k Hz The errors are dependent upon decade settings and stray capacitances in the bridge and also outside, so that corrections are not easiIy made 4.7.4 The 1: 100 Ratio. At 1 kHz. a setting of 9X was equivalent to XO. as it should be At 100 k Hz , not only was the measurement in error because of residual impedances. but the change from XO to 9X resulted in a new balance requirement so the final readout was 99 In fact the two 100-kHz measurements differed from each other by considerably more than the original error due to frequency The read ing was lower for the second 100-k Hz measurement than for the first. because the capacitance C G s was larger With the XOO setting. the internal 1000-pF standard and its ground and wiring capacitance loaded the transformer predominantly; with the 998 setting. both the 100-pF and 10-pF standards were also connected and the ground capacitance was thereby increased about 170 pF. Also, several other capacitors were switched to higher taps in the final balance. This example is nearly an extreme case. A typical situation in which you would use a 100 1 ratio is measuring a capacitor larger than 1 pF The large ratio generally means a large bridge error resulting from the residual impedances, but also simplicity in their effect and, therefore, a chance to correct your measurement by calculation Capacitance to ground from the UNKNOWN HIGH terminal, due to the capacitor attached there, can be measured (with another bridge) or estimated That and CG x (Table 41) are usually negligible in determining the dominant factor C T x in the error THEORY 4-13 Notice that this error is positive (readout is too high), that it is fractional (a dimensionless percentage, not a number of picofarads), and that it is proportional to frequency squared, the value of capacitor measured, and the total series inductance (including transformer leakage and those components of \\ in cables outside the bridge). Whether Qx includes also the series inductance within the structure of the unknown capacitor depends on definition. If not, then C, is an effective "terminal" capacitance and is liable to be frequency dependent. If yes, then Cx is the series capacitance within that structure, and may be essentially independent of frequency. The G error is more conveniently treated as an additive term (mhos) rather than a fractional error Because of the sense of the large ratio and the limitations of resistance in switches and connections, the term containing r x predominates, thus described, another source of error at high frequencies is the voltage induced in the internal bridge wiring connected to the detector circuit by currents flowing in the bridge circuits connected to the transformer The bridge does not have the short, coaxial current paths required for radiofrequency accuracy, and mutual inductances of the order of 1 IlH between bridge arms may be expected Since the currents drawn from the transformer by the bridge capaci tors increase with frequency (i = wCE) and the voltages induced in the detector circuit are proportional to wMi, these error voltages, e = w 2 MCE, increase with the square of frequency The errors depend also, in a complicated way, upon the internal capacitors used, the transformer voltages applied to them, and therefore the front-panel switch settings. Experiments confirm, however, that these errors appear not as a percentage of the measured capacitance but as a more-or-less constant picofarad error or C offset at any one frequency The order of magnitude of the error is 100 aF at 100 kHz a 4.7.6 Ground-Circuit Impedance. Again, this error is positive (readout is high) However, it is a number of Siemens, not a ratio. It is proportional to frequency squared, C; 2, and to the total series resistance including those components of r x in cables outside the bridge. Whether r x includes any lossy component inside the capacitor under test depends on your definition. At any one frequency, the conductance can be expressed as an equivalent series component Rs' as shown in para 3.6. Usually that is not included in r x However a known series resistance (often called "ohmic" resistance or d-e series resistance) may be treated as part of r x and thereby excluded from your measurement As a rule of thumb, calculated corrections of the types given above (for C and G, for 1·100 ratio) are accurate to 5 or 10% Use them to reduce the error of high-frequency measurement an order of magnitude; but if frequency is 30 k Hz (for example) do not ex pect corrected measurements to be as accurate as l-kHz measurements 4.7.5 C Offset Due to Induction. In addition 4-14 THEORY to the residual impedances previously Another source of error, which may be significant when you measure a 3-terminal capacitor remotely at high frequency, is the finite impedance of the path between the shield of the capacitor and ground in the bridge This path, usually through one or both shields of the cables connecting the capacitor to the 3-TERMINAL UNKNOWN port, has series Rand L components ZSg = RSg + jwL Sg (Figure 3-1) Currents through ground capacitances Cis and return through ZSg The effects on the measurement are negative errors in both C and G; thus the. correction is additive c., The error is most commonly noticed when Greadout is negative, both because it is often the larger of the two components and because negative loss in a capacitor is an obvious indication of the presence of error. An example is given in para 3.9. ServiceandDiagrams -Section5 5.1 5.2 5.3 5.4 5.5 5.6 GR FIELD SERVICE . MINIMUM PERFORMANCE STANDARDS DISASSEMBLY . RECALIBRATION AND ADJUSTMENTS TROUBLE ANALYSIS PARTS LISTS AND DIAGRAMS Exterior Views (Mech parts) Federal Manufacturing Code Figure 5-9. Schematic Diagram Appendix A . 5-1 5-1 5-1 5-7 5-8 5-9 . 5-11 .5-13 .5-15 . A-' 11 5.' WARNING Dangerous voltages may be present inside this case. These instructions are intended for the use of qualified service personnel only, to avoid electric shock, do not perform any servicing other than that contained in the operating instructions unless you are qualified to do so. GR FIELD SERVICE. dures. except as noted, set the front-panel controls initially Our warranty (at the front of this manual) attests the quality of materialsand workmanship in our products. When difficultiesdo occur, our service are available for technical telephone assistance. If the difficulty cannot be eliminated by use of the following service Instructions. contact the GenRad Service Department in Concord. MA giVing full information of the trouble and of steps taken to remedy it. Describe the instrument by type serial, and 10 numbers. (Refer to front and rear panels) Instrument Return. When returning an instrument to GenRad for service. please identify the failure mode as accurately as possible and include this information with the instrument. For instruments not covered by the warranty. a purchase order should be forwarded to avoid unnecessary delay. For return shipment, please use packaging that is adequate to protect the instrument from damage. i.e.. equivalent to the original packaging. Advice may be obtained from the GenRad Service Department in Concord, MA 5.2 MINIMUM PERFORMANCE STANDARDS. 5.2.' General. The equipment, methods, and criteria for verifying the specified performance of the 1616 Precision Capacitance Bridge in the 1621 Precision Capacitance-Measurement System are given below. Recalibration is described in para 5.4. If performance is grossly inadequate, erratic, or cannot be corrected by the adjustments, refer to trouble analysis, para. 5.5. Equipment needed for the measurements and procedures of this section is listed in Table 5-1. For all these oroce- to the standard positions, as follows Oscillator (1316) POWER - on (after preliminaries, para. 3.1). FREQUENCY - 101 kHz OUTPUT VOLTAGE RANGE - 15 V OUTPUT ADJ UST - 10 on lower scale of meter Detector (1238) POWER _ ON (after preliminaries, para. 31) FREQUENCY _ 101 kHz TIME CONSTANT - 0.3 S FIN E ADJUST - each control at midrange PHASE SHI FT - 180° (Large dial at O~) SENSITIVITY maximum (cw) GAIN 40 dB COMPRESSION - disabled (button out) LINE REJECTION disabled (button out) Bridge (1616) C MAX - down (3 windows closed) CAPACIT ANCE - zero CONDUCTANCE - zero Conductance Multipl ier - )( 10-6 EXT MULTIPLIER - OFF TERMINAL SELECTOR - CAL ZeRO ADJUST - any position 5.2.2 Zero Setting, Offsets, and Sensitivity. Verify that the capacitance offset Co can be set to zero and measure the conductance offset Go' as follows. Sensitivity and precision are also verified. a Perform the preliminary and functional checks, tuning, and phase adjustments of para 3.1-3.3. b. Install an open-circuit shield (Table 5-1) on the SERVICE & DIAGRAMS 5-' --------- Table 5-1 -------TEST EQUIPMENT Item Requirements Stability: 20 ppm/year; temperature coefficient: 2 ppmrC; terminals: 3; value: 100 pF ±10 ppm. Reference standard capacitor, (GR1404A) Standard resistor, 100 kn Accuracy, : ±.01% at 23°C. Precision metal-film resistors 10 Mn ±0.1% 1 Mn ±0.1 % (4 required) 10.2 kn ±01%. 100 n ±0.1% DC Resistance Basic accuracy: ±O.01 %. direct reading: range 1p.S to 1 6- digit resolution. bridge rn (GR1666) Binding-post adaptor Fits 2- TERMINAL UNKNOWN port (G900 to BANANA JACK 3/4" spacing Patch cords Doubly shielded coaxial cable with locking G874 connectors: length: 3 ft. (2 required) Open circuit Shielded mount for conductance network: G874 connectors: (2 required). Shielded mount for conductance connectors; (2 required). Mount connector). network, G874 Table 5-2 CAPACITANCE RATIO CHECKS TERMINAL SELECTOR 3-ter x1 CAL CAL CAL CAL CAL CAL CAL CAL CAL CAL CAL 5-2 SERVICE & DIAGRAMS nF pF ·. XOO NXO 'NX 'ON '00 '00 '00 '00 '00 '00 XOO 000 •• ·. 'N NX fF see 5.2.3c 000 000 XOO NXO ONX 000 000 000 000 000 000 aF 000 000.0 000.0 000.0 000.0 000.0 XOO.O NXO.O ONX.O OOn.X 000.0 000.0 Tolerance see 5.2.3d 500aF 50aF 10aF 5aF 5aF 50aF 5aF 5aF 2aF 5fF 50fF Approx GAIN Trimmer Label 70db 90 110 120 130 130 130 130 130 130 50 30 1nf 100pF 10pF 1pF 100fF 10aF 1aF 100aF 10aF 1aF 10nF 100nF 3-TERMINAL LOW connector. Set the TERMINAL SELECTOR to 3-TERMINAL X 10. c. Balance the bridge precisely, with increased GAl Nand (if necessary) voltage, as described in para. 35, using the lever switches for G but only the ZE RO ADJUST control for C. Leave the C readout at zero, except for temporary changes of a step or two in the last digit to test whether you have sufficient sensitivity for 9-figure resolution Record the G readout as Go' the conductance offset It should not exceed 02 X 10-6 p.S in magnitude, and may be either + or d. Sensitivity is normally sufficient so that an oscillator voltage of 100 V, GAIN setting of 130 dB, and SENSITIVITY set near midrange result in a deflection of 25 (5 divisions) for 5 steps of the last C lever switch, near balance e Repeat step c with the Terminal SELECTOR set to each of its 6 positions, the final one being 3 TERMINAL X 1. The bridge is now set for Co = 0 ±O 1 aF, on 3TERMINAL Xl, and Go = a known offset Note that the sensitivity of ZERO ADJUST is variable Its setting is significant in 2-terminal measurements, but 3-terminal measurements can be made to specified accuracy without regard for how this control is set. A rear-panel, screwdriver adjustment is required if the range of ZERO ADJUST is inadequate for balance in step e. 5.2.3 Capacitance Accuracy. Verify that the absolute van.e of one of the prime internal standard capacitors has been trimmed within specification, as follows NOTE If the required external standard is unavailable, it is still possible to make a calibration verification with high confidence by making only the Capacitance Ration Checks of paragraph 5,2.4. The internal standards are all of high stability so that checking all their ratios will almost always detect any excessive errors. The chance of all the standards changing by the same amount (in %) is very small. a. Stabilize the temperature of the bridge at 23° ± 1°C r 24 hours previous to and during the following steps of ra. 5.2.3 (36 hours is required after the bridge has been bove 45° or below O°C; refer to para 46) b. Remove any shields from the 3- TERMINAL UNKNOWN erminals and connect instead an external reference standard ith a nominal value of 1000pF known to an accuracy of Oppm or better. (Note: the following instructions pertain in etail to the use of a 1000pF standard but a 10 or 100pF tandard may be used if the calibration principle is nderstood, c. Set the readout to the exact value of the 1000pF standard, starting with an X. For example, if the value is 1000.016pF, set XOO.016pF. If the value is 999.989. set Set first row of Table 5-2. XOO.N89 where N is -1. d. Verify that the final C readout is not more than 10 ppm different from the absolute 3-terminal value of the C standard I f that is known to ± 12 ppm, for example, the acceptable tolerance on this measurement 5.2.4 Capacitance Ratios. The following procedure checks each internal standard against the next higher. or lower. standard starting with the 100pF standard being checked against the Just-calibrated 1000pF standard as indicated in the 2nd row of Table 5-2. This and the next eight checks work down to the lowest standard. The last two checks work up from 1000pF to .1 uF. (If a 100pF or a 10pF standard was used as the external standard. then this ratio procedure would start further down on this table and checks should be added that go up from the value of the standard used to values skipped.) The procedure for checking all other standards against the internal 1000pF standard is as follows. a. Set the TERMINAL SELECTOR to CAL. Be sure EXT MUL TIPLIER is set to OFF. Set C readout to all Zeros with first three digits masked using the C-MAX switch. b. Balance the bridge with ZERO ADJUST and the G levers. Use the lowest G multiplier (10-6). Get negative G values using a -1 setting of the smallest digit possible. Leave the ZERO ADJUST as set You now have Co < 0.1 aF for bridge configuration set c. Set up each of the C-readout combinations starting with the second row of Table 5-2. For each. balance the bridge using the smaller C lever switches and the G controls as required. Verify that each balance is achieved within the tolerance limits indicated. Here * indicates a zero digit covered by a mask as set with the C- MAX lever and N indicates a setting of -1, If the detector phase adjustments are made as suggested by the Condensed Operating Instructions. the IN-PHASE meter may be zeroed with the C levers and a precise G balance is not necessary. This meter can also be used to extrapolate between C lever steps to estimate the error. The GAIN settings in the table are suggested as initial settings. more gain may be required. 5.2.5 Conductance Accuracy. The following procedure makes use of 2 external stand- ards to check the 5 basic internal G standards (An equally valid check could be made by measuring each with a d-e bridge, a procedure which would require some disassembly and unplugging of connections NOTE The accuracy of a suitable standard depends on the certified accuracy when calibrated by NBS or a standards laboratory, its stability and the time elapsed since calibration, it temperature coefficient and the temperature, etc is ±22 ppm, If ±3, then ±13 e Remove the external standard to the "G box 'OJ & CAUTION Restrict the oscillator level to 20 V rms or less when the TERMINAL SELECTOR is set to 2-TERMINAL Xl; otherwise safety diodes will distort the test signal and invalidate the balance condition. SERVICE & DIAGRAMS 5-3 At high settings of GAIN, movement of your hands or other objects near the unshielded 2-TERMINAL UNKNOWN port will upset the balance. Keep hands away and nearby objects still. a. Install an adaptor on the 2-TERMI NAL UNKNOWN port and plug in a 100-kQ standard resistor (Table 5-1) as the first G standard. b. Reset the frequency to 102 Hz, the oscillator level to 10 V, and tune the detector and oscillator to match as before. However, reset the PHASE SHI FT in the process of balancing the bridge, as described in para. 3.5 under Final Balance Phase Shift/Reset. NOTE Reset the phase shift, as required, as you approach balance in any step of the procedure: para. 5.2.5 and 5.2.6. The need to do so is recognizable when both phasesensitive meters respond to a change of either C or G lever switches. (The inconvenience of having to reset PHASE SH IFT is to be expected only for precise measurements of very lossy capacitors, not for most measurernents.) c. Set up, in turn, the combination of G standard, TERMINAL SELECTOR setting, and G readout for each of the first 5 rows in Table 5-3. Complete steps d through f for each setup before proceding to the next. d. Start with a convenient GAl N setting. Balance the bridge using the C lever switches, increasing the GAIN to the "dfs" number tabulated. (Reset PHASE SHIFT as required.) e. Adjust the SENSITIVITY so that a change of 5 divisions on the QUADRATURE meter corresponds to the percentage change tabulated as "Tolerance". Notice that a External Standard 100 k 100 100 n 10 M 10 100 k 100 CONDUCTANCE Terminal Selector o 1% change is the step from XOO to X01, or XO( -1) Similarly, a 1-% change is from XO to Xl, or X(-1) A 10-% change is from X to 9. Change the GAl N if necessary. f. Verify that the internal G standard is within th specified accuracy. In other words, the QUADRATUR meter should read between the points labeled 25 when the readout is set as tabulated. (The pertinent standards are R 1 through R5, in that order.) 5.2.6 Conductance Multipliers. A procedure very similar to the preceding one is used to check the G multiplier ratios (steps of 10-1). In this paragraph, the maximum accuracy (0.1%) is required for every step. Unlike the internal G standards, the multipliers can be trimmed by you, the customer, if you find they fail to meet the following specifications (para. 5.4) a. The setup is much the same as before. Initially, reinstall the lO().kQ standard resistor at the 2-TERMINAL (see below) to the port and connect network "A" 3-TERMINAL UNKNOWN port, using shielded patch cords. b. Networks"t>:'and "B" are 3-terminal conductance standards that you can assemble in a few minutes. Each is a tee of 3 resistors,Figure5-1, enclosedin an electric shield. Use a G874-X as an enclosurefor each. Measurethe resistanceof each resistor, after soldering, at room temperature, to an accuracy of ±.03%. Use a precisionresistancebridge, such as the GRType 1666. For convenience in making 4-terminal "Kelvin" connections to each resistorat the test fixture of such a bridge, leavethe resistorpigtails full length. (After the measurementthey can coiled up for future use or clipped off.) c. Calculate the 3-terminal conductance of each network you have made, thus: Table 5-3 ACCURACY AND MULTIPLIER CHECKS CONDUCTANCE Approx us ns Mul GAIN Tolerance 000.0 Trimmer Resistor 2- TERM X1 2- TERM X10 2-TERM X100 X 0 0 XOO,O OXO.O 50 dB 70 90 ±0.1% 0.1 0.1 none none none n 2- TERM X10 2-TERM X100 0 0 OOX,O OOO.X 80 90 2 10 none none n X X X 000.0 000.0 10-1 10-2 CDE.F 10-3 60 dB 70 80 ±0.1% "A" 2- TERM X10 2- TERM X100 3- TERM X1 R18 R19 R20 "A" "B" "B" 3- TERM X10 3- TERM X1 3- TERM X10 X X X CDE.F GHI.J GHI.J 10-4 10-5 10-6 The letters C thru J represent described in paragraph 5.2.6c. 5-4 SERVICE & DIAGRAMS 90 R21 R22 R23 110 120 digits calculated as NOTE To separate the bushing from the knob, if for any reason they should be combined off of the shaft, drive a machine tap a turn or two into the bushing to provide sufficient grip for easy separation. Here, C, D J represent numerical digits, of which F and Notice that X (ten) is the preferred J are insignificant most significant digit in following procedure For example, if GA IS 99765 n ,express it as X (1 )765 nS d. Set up, in turn, the combinations of G standard, TERMINAL SELECTOR setting, and G readout for each of the last 6 rows in Table 5-3. Complete steps e through g for eachsetup, before proceding to the next e. Balance the bridge, as before f. Adjust the SENSITIVITY as before, for 0 1% change equals5 divisions g Verify that the most-significant G standard multiplied y the selected G multiplier is within ±O 1% of its nominal alue In other words, the QUADRATURE meter should ad between the points labeled 25 when the readout is set tabulated. Ra 5.3.2 Cabinet Removal. ~ i "8" & NOTE Keep POWER OF F during disassembly or reassembly. Rb ",.::t NETWORK To replace a knob a. Slip bushing on shaft and rotate to correct position as observed in disassembly of knob. b. Keep bushing away from panel by at least the thickness of a filing card Pull it out farther if necessary to prevent tip of shaft from protruding. c. Tighten the setscrew in the bushing. d. Place knob on bushing with retention spring opposite setscrew. e. Push knob on until it bottoms and pull it lightly, to check that the retention spring is seated in groove in bushing. Rf Figure 5-1. Three-terminal conductance standards. Ra = Rb = Rd = Re = 1 Mn. Rc = 10.2 kn. Rf = 100 n. .3 DISASSEMBLY. .3.1 Knobs. & CAUTION Do not use a screwdriver or other instrument to pry off the knob if it is tight. To remove the knob from a front-panel control, either replace. one that has been damaged or to replace the ociated control, proceed as follows a. Grasp the k nob firmly with dry fingers close to the nel and pu II the knob straight away from panel b. Observe the position of setscrew in bushing when the antral is fully ccw. a e. Release the setscrew with an Allen wrench; pu II the ushing off the shaft. Figure 2·3. To remove the bench-model cabinet from the bridge, first set the instrument in the horizontal position, free of unnecessary cables, and proceed as follows a. Remove the 4 dress-panel screws (A) accessible through holes in the handles b. With caution not to let the instrument drop out of its cabinet, turn it face down to rest on the handles. c. Pull the cabinet up and off. Carefully return the instrument to a horizontal position. To remove each instrument from the 1621 system cabinet or each rack-mounted instrument from its cabinet, apply step a, above. Then withdraw the instrument forward carefully as described in para 24 and 25 . 5.3.3 C·Box Removal. Figure 5·2. The subassembly containing the capacitance standards comprises a substantial part of the instruments' weight and volume Although it is not to be repaired, it can be removed as follows, either for replacement or for access to miscellaneous wiring. a With the bridge right-side-up, outside its cabinet, remove the upper 3 screws (of a total of 5) that are about % in. from the rear edge of the right side panel b. Remove the corresponding 3 from the left Also on the left side, remove the 3 top screws, 3 more 6 in. below those, and finally 2 more for a total of 11 that form a rectangle. c. Slide the C box back only '% in, so it has support, and SERVICE & DIAGRAMS 5·5 disconnect the 3 BNC connectors behind the EXTERNAL STANDARD port d. Slide the box back about 3'12 in. and support it by hand while disconnecting the edge connector from the etched board and the 2 wires that are separately clipped near the middle of the bottom edge of that board e. Lift the C box out toward the rear. f. In reassembly, be sure to bring the C MA X gears into mesh as nearly right as possible. That is, preset the gear on the C box to the ccw position (of 4 detent stops) as seen from the left; preset the C MAX lever down (all 3 windows blanked) g Connect the white wire (left) and black (right) first; then the etched-board connector; then the 3 BNC connectors in this sequence violet (bottom), (top) , red h. Reposition the idler gear associated with C MAX, i necessary, by loosening, moving, and retightening the small recessed screw in the left side panel. Moving it toward the rear decreases the backlash but may cause binding, toward the front makes the action more free. i Slip the gear engagement (if the attempt in step f was unsuccessful) as follows Tip the instrument onto its right side, remove the 11 screws from the left side, and tilt the C box by hand enough to disengage the gears while you move the C MAX lever slightly for better alignment C TRIMMER PANel J7,,8,,9 gray (middle), GEAR FOR S3 ~,,~ ~~ "': ~CBOX C TRIMMER ACCESS DOOR .......J~--Jl0 WHITE (WT5) RATIO TRANSFORMER <, T1 <; BLACK (WT6) C301 ACCESS Figure 5-2. Interior view, upper rear, with the C box removed and shown above. 5-6 SERVICE & DIAGRAMS • G-MUlT TRIMMERS: R23, R2l, R20, R19, R1B, R22 ~ _f I i~ I FOR S4 Figure 5-3. Interior bottom view. The G box cover is shown fastened 5.3.4 The G Box. Figure 5-3. This "box," containing the conductance standards, is the most conspicuous subassembly in the bottom part of the bridge behind the bank of lever switches Although the G box is not to be repaired, it can be uncovered for access to the G-multiplier adjustments and it can be removed if necessary Use this procedure. a Place the bridge upside down b. For access to adjustments, remove the 4 cover screws Move the cover to the position shown and replace 2 of the screws through the 2 spare holes (nearest the edge) of the cover. Adjust only with the cover attached c. For removal of the box, make a note of the positions of the 10 wires to be disconnected, something like this left end, front-to-rear red tracer, green tracer, blue tracer, brown tracer, black tracer; rear panel, left-to-right red, violet, black, grey, brown. d. Disconnect those 10 wires and the yellow-banded BNC connector e. Remove 2 screws in the rear panel (near, but not holding, the BCD CAPACITANCE OUTPUT connector) and 4 screws in the right panel.Iarnonq. but not holding, the slide blocks) while holding the G box so it does not fall. Lift it free f. In reassembly, mesh gears properly by presetting the switch on the G box ccw (as seen from the right) and the G temporarily multiplier in position for trimming lever "down," the G-multiplier indicating network. 1()6. If necessary, to obtain proper mesh, try again g. Adjust the gears for a reasonable amount of backlash by moving the G box slightly, while you have the 4 rightside screws temporarily loose NOTE Do not reposition or replace any parts or turn any adjustments in the G box except as instructed in para. 5.4. Otherwise Do will have to be reset, a factory operation. 5.4 RECALIBRATION AND ADJUSTMENTS. Recalibration is principally the trimming of the internal C standards for absolute accuracy and exact decade ratios The G multipliers (but not the 5 standards) can be trimmed also. Adjustments include C301 which affects the ZERO ADJUST ranges and mechanical adjustments on the lever switches. The temperature of the bridge should be maintained well within the allowable tolerance, just as the accuracy of each adjustment should be greater than the acceptable accuracy of the bridge For example, in addition to the temperature requirements described in para 52, the bridge should be held at 23° ±05°C for 6 hours previous to and during recali brat ion SERVICE & DIAGRAMS 5-7 For special uses, it is possible to recalibrate the bridge at temperatures other than 23°C. The range of trimming provided for the internal C standards sets the upper and lower limits at about 21° and 26°C, respectively If you do recalibrate at a nonstandard temperature, be sure to tag the bridge with that information 5.4.1 Adjustment of Internal Capacitance Standards. To recalibrate the one or all of the internal capacitance standards, proceed as in paragraphs 5_2.3 and 5.2.4 but make adjustments of the trimmers (listed in Table 5-.2) to minimize the errors. Additional procedure is as follows. a. Unlock and open the panel that covers the trimmer screws (para. 13) b. Obtain a small. sturdy screwdriver with an insulated handle and shank (or enclose the shank in a plastic sleeve). .&. WARNING Full oscillator voltage appears on the trimmer screws. c. Notice that each screw is labeld in terms of the effective value obtained when the corresponding lever switch is set to maximum. Thus, the 3rd screw is labeled 1nF; it trims C100 which adds 1000pF to the standard arm when 100pF per step lever is set to X. To trim C100 against the external 1000pF standard (para. 5.2.3) use the nF (or 1000pF) trimmer as indicated in the first line of Table 5-2. d. To trim a smaller standard against a larger (as in the next 9 rows of Table 5-2) use the trimmer whose labeled value is equal to the capacitance setting. See right- hand column of Table 5- 2. e. To trim a larger standard against a smaller (as in the last 2 rows of Table 5-2). use the trimmer whose labeled value is 10 times the capacitance setting. See right-hand column of fable 5-2 .• f. If you are interested in preserving the accuracy as long as possible, guard your bridge against any violent motion. Even the impact of a dangling patch cord swung against the panel will cause measurable changes in calibration (but insignificant compared to the specifications) Close the access door over the C trimmers with care, otherwise it will snap in closing. 5.4.2 Conductance Multipliers. The G-multiplier calibration is similar to that above. Proceed as in para. 5.2.6 and trim so as to minimize any errors you find. The additional procedure is as follows. (Notice that any possible errors in the 5 G standards, para 5.2.5. must be corrected only by GenRad, which has means for checking and adjusting Do at the same time. a. Gain access to the set of 6 screwdriver adjustments on the G box, as in para 5.3.4. b. In turn, make the connections and settings Indicated in the last 6 rows of Table 5-3, and as described in paragraph 5.2.6, but instead of determining the error, adjust the trimming resistor listed in the last column for the best null possible. Always adjust these trimmers in the order given (R18 first) because the last three adjustments have some interdependence. 5-8 SERVICE & DIAGRAMS 5.4.3 C301 / Setting Zero C. The ZERO ADJUST control varies Co' the capacitance offset, in any position of the TERMINAL SELECTOR / READOUT MULTIPLIER switch. So does the rear panel screwdriver adjustment C301 (Figure 5-2) Normally, C301 is set so that the lO-turn range of ZERO ADJUST spans conveniently the 3 conditions referred to in para 522, which are conditions of Co = 0 on CAL and both 3 TERMINAL-UNKNOWN positions (The 3TERM LOW connector must be shielded I The adjustment shou Id be made at 1 kHz. 5.4.4 Lever-Switch Stiffness. The front-panel lever switches that control C and G digits, ie, all but the first and last levers, have detent mechanisms that can be adjusted as required The factory setting is approximately 12 07 (350 grams weight), the force required at the normal finger position to push each switch from step to step (between 0 and g) Other settings are possible, to satisfy ind ividual preferences. About 9 07 will provide a light "feel" and enable you to slide each switch up and down with your thumb on the front of the knob, but you need more skill to avoid stopping between detents. About 1607 will provide a positive detent action that may be preferable if the bridge has many operators, each mak ing relatively few measurements at a time. Perhaps more important than the absolute stiffness of the detent action is consistency If one or more of your switches responds differently from the majority it will be an annoyance. Adjust the detent stiffness as follows. a. Remove the bridge from its cabinet and position the instrument upside down. b. Determine which construction you have, by reference to Figure 5-4. "A" has 3 screws in a row along the bottom under each detent wheel; the screw farthest from the panel is an adjustment (not locked) "B" has 2 such screws, both locked. c. If you have "A" construction, turn the adjustment screw, cw to stiffen the detent action, ccw to loosen it d. If "B," loosen screws Band C just enough to allow you to move the block that holds the spring Tilt the block as is required to obtain the desired detent action and tighten the same screws 5.4.5 Maintenance Note On Switches Both rotary and lever switches may become noisy or erratic from lack of lubrication. For rotary switches such as the TERMINAL SELECTOR switch and MULTIPLIER switch, use CRAMOLIN R5 oil. For lever switches, such as C and G decades, use LUBRIKO H101 grease. Clean contact area of lever switches with solvent (i.e.. FREON TF) before applying a thin film of lubricant. 5.5 TROUBLE ANALYSIS. If the bridge requires service beyond the realm of recalibration and adjustment described above, or the categories of repair described below, please return the instrument to GenRad. Refer to para. 5.1. 5.5.2 BCD circuits. If the BCD code availableat the rear panel does not indicatethe measurementas described in para. 2.8, trace the circuit with the help of the schematic diagram, Figure 5- 7. If the trouble is in the wiring, repair it. If in the switching, return the instrument to GenRad. NOTE The lever switches are not replaceable, except by GR. r: . - ,~,,~""" - ~ . FRONT PANEL 5.5.3 Non-Repairable Subassemblies. ._c-, SCREWSB,C ....-. BRACKET I I I I ~ DETENT WHEEL / I The following components are not to be disassembled for repair. the ratio transformer T1, the coaxial choke T2, the G box. and the C box. In the unusual event that one of them apparently needs replacement. please consult with GenRad or return the instrument as described before. x 5.5.4 Typical Parameters. The following parameters are typical They are listed for reference in trouble shooting. Unless otherwise noted, set the controls as follows TERMINAL SELECTOR - CAL EXT MULTIPLIER C and G readout - zeros C MAX - shutters closed G multiplier - 10-6 a (B) Figure 5-4. Detent mechanism of a typical lever switch, A and B constructions. The instrument is upside down Ishowing the bottom edge of the front panel). The type of wrench shown in part 10.25 inch, 80°, open end) is recommended for instruments having the B construction. 5.5.1 Mechanical Damage. Parts, such as connectors at front or rear, that are damaged or worn, can be replaced A wire broken loose can be resoldered or replaced. However, the replacement of any lever switches is not recommended. Only the simplest repairs to the mechanisms of the C-MAX or the G-multiplier levers is recommended. At the GENERATOR INPUT Jack 3.7 H at 100 Hz, dc, ungrounded Demagnetize after dc measurement; n para 314. Measured series inductance increases with test voltage; may be 5 H at 1 V. At the EXTERNAL STANDARD port, with EXT MULTIPLI ER at 10, 37 H at 100 Hz, 1 n dc, qrounded Demagnetize as above From DETECTOR OUTPUT to EXTERNAL STANDARD LOW, with the EXTERNAL MULTIPLIER set to 10, 14 mH (below l-V test voltage) at 1 kHz; R < 1 n dc 5.6 PARTS LISTS AND DIAGRAMS. The following pages contain mechanical and electrical parts Iists. schematic diagrams, and supplementary information. SERVICE & DIAGRAMS 5-9 MECHANICAL PARTS LIST Fig Ref Qnt FRONT 1 I. 2 2. 1 3. I 4. 2 5. 2 6. 7. 8. 9. 2 10. 19 II. 12. Description GR Part No. Fed Mfg Code Mfg Part No. Lock with 2 keys. Handle. Cabinet gasket. Bench cabinet assembly; detailed breakdown listed separately. * Panel locking connector asm., J I,EXTERNAL STANDARD HIGH; J4,3-TERMINAL UNKNOWN HIGH. Panel locking connector asm., J2, EXTERNAL STANDARD LOW; B,3-TERMINAL UNKNOWN LOW. Banana jack, J5, ground connection. Precision connector, J6, 2-TERMINAL UNKNOWN. Knob asm., ZERO ADJUST, including: retainer. Knob asm., TERMINAL SELECTOR; EXT MULTIPLIER, including: retainer. Knob for CAPACIT ANCE or CONDUCTANCE lever switch. Door assembly. 5605-0100 5360-2032 5331-2220 4172-4016 24655 24655 24655 24655 5605-0100 5360-2032 5331-2220 4172-4016 0874-4006 24655 0874-4006 0874-4005 24655 0874-4005 4150-0900 0900-4410 24655 24655 4150-0900 0900-4410 5520-5333 5220-5402 5500-5321 24655 24655 24655 5520-5333 5220-5402 5500-5321 5220-5402 5500-5120 24655 24655 5220-5402 5500-5120 1616-1041 24655 1616-1041 Multiple socket, J 12, BCD CONDUCTANCE OUTPUT. Multiple socket, J13, BCD CAPACIT ANCE OUTPUT. BNC connector, JII, GENERATOR OUTPUT; J14, DETECTOR OUTPUT Foot (resilient strip). 4230-4036 93916 57-40360 4230-4049 93916 57-40500 4230-1200 91836 KC-19-161 4171-7010 24655 4171-7010 Fed Stock No. REAR I. 2. 2 3. 2 4. NOTE Electricel parts information in this section is presented in such a WIY that all tha dlta for I pert-numberld sub· assembly Ira visible in I single opening of tha mlnult. Thus. thl perts list Ippeerl on Ilft·hend plges. whila the part-location diagram Ion tha Ipron) Ind the schemltic dillGram hip out] Ira on right-hand Plges. 'Cabinet kits and hardward are listed on page 5-12. REFERENCE DESIGNATOR ABBREVIATONS B BT C CR OS F J K KL KS L M MK 5-10 SERVICE & DIAGRAMS = • • • • • • • • Motor Battery Capacitor Diode Lamp Fuse Jack Reley Reley Coil Relay Switch Inductor Meter Microphona P Plug T"nsistor Q R S T U VR X Y Z .. • • • • • • • Resistor Switch T,andormer Integreted Circuit Diode. Zenar Socket for Plug-In Crystal Network ReferellC8l ASA Y32.16end MIL-STD-16C 5935-062-1776 2 9 10 11 Figure 5-5. Front view; mechanical replaceable parts identified. 3 2 4 Figure 5-6. Rear view; mechanical replac;eable parts !T 404 i;~~ZI~~~~~SI~~~~~~A~;, ~o~, 16';; ,l~: 10-1 I VOL TAGES EXPLAINED fN INSTRUCTION BOQI'( SERVICE NOTES c:==::J' PANEL CONTROL REAR CONTROL 0" SCREWDRIVER CONTROL WT ~ WIRE TIE TP ~ TEST POfNT COMPLETE REFERENCE DESfGNA TlON INCLUDES SUBASSEMBl Y LE TTER, C·RI, B·R1, ETC (:::::::1 Ie F ROTARY CONNECTfONS SWITCH NUMBERING L,W.,,, S"",ACE 'RON', -..........+ REAR - DUTPUTLEAVE~SUElASSEMBLY ~ INPUT FROM DIFFERENT SUBASSEMB -' --CONTACTS FIRSTCONTACTCW FROM STRUT SCREW ABOVE I'(EY IS 01 SECTION SECTlON NEAREST PANEL IS '1 ----t> OUTPUT REMAfNS ON SUBASSEMBLY fNPUT F ROM SAME SUBASSEMBLY [)---- ROTORS SHOWN CCW MAX-! 53 '-----t--+-----------, I WT51 +WT5'3 WT5e- J 10 3 i 55 i 5<0 x 57 0 9 8 7 5 43 2 I 0 -I '" WT2. WT4- YE-VT-BK ~J Ii,:no 0 YE-GY-BU YEGY-BK YE-GY-GN 24- WTN 23 1 WT4 WH-C,NI J13"47 ~ -000- WH-BU j'"' , WH-BR '1-c;, 22- I, I~ f-~---i WHoO WTI ,WH-SK t t 4-8 ~: 0 WTB Y JI3""4:'l I~ , c -------.---'-.... CAPACITANCE =0 THRU CONDUCTANCE SIG 517 THRU 52.1 o WT5 C , J g:;1 ~I , J : 2.. ¥'- \4 '" i") oJ f\) m ~ J r-'-_ , tw r '70 '------i---------. VV- 8 - \iI ::; (\) 1'1 r<'> '" '-'! t -s I> J WT 170 - WT 210 VI I OWTI R , RD-BK l:l J 12."'18 J I 2. -l~NC?.U~Tf\ N C~-.!?U TPU~ : co~ GY-BK, I GY-BR GY-BU G"!-GN 17 '1(3 VT-BK VT'BR VT-BU VT-GN 14 OR-13K OR-BR OR-BU OP,-(;N IS 4 8 31 4 i'.8 29 1(; 2 3 4 s I 2 4 e , -- I 2 2G 27 VT-GY-8K VT-GY-BR VT-GY'BU VT-C,Y'GN ,J 12 tt -." I 2 i'. 4 19 S 5 G 30 I WT tv" I 15 " 4 B 20 3 q 21 22 OR-GY,BK OR-GY-BR OR-GY-BU OR'GY-GN 13 d YL -f\D-BK YE'RD -SR YE-P,D-BU YE-RD'G N 2.0 2- 8 4 25 2<;, 2.\ 4 27 B 28 I 5 2. IQ 23 2'1- 8 I B I 2 7 9 10 WIRE COLO"" YE-RD'BK YE'RD'BR WH-RD-BK, WH'RD'BR YE'BU ,'lWT24 ~ -z: CHA:,SIS JIZ'*3G, OR-GY'BK OR-GY-BR WH-BK WH-BR VT-GY-BK VT-GY- BR YT-GY-GN YE-GY-BK YT'RD-BK VT-RD-BR YE-GY-BR YE-RD-BU GY-RD-BK GY-f\D-BR YE-RD-0N OR-GY-BU Figure 5-7. Wiring diagram of BCD output circuits. SERVICE & DIAGRAMS 5·11 ELECTRICAL PARTS LIST CHASSIS MOuNTED PARTS DESCRI PTlON REFDES C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C C ** ,.,. J J ** ** J J J J J ,.,. J J J J J J J J * R R R R ,.,.,. R R R R R R R S S S S S S S S S S S * ** *** 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 10 17 18 20 21 22 23 24 25 20 27 28 100 101 102 103 300 301 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 CAPACITOR .0<;993 UF CAPACITOR 9990 PF FACTCRY SELECT CAPACITOR AIR 3-32PF FACTCRY SELECT CAPACITCR AIR 1.7-8.7PF TRIMMER CAPACITOR ASM TRIMMER CAPACITOR ASM TRIMMER CAPACITOR ASM TRIMMER CAPACITOR ASM TRIMMER CAPACITOR ASM TRIMMER CAPACITOR ASM TRIMMER CAPACITOR ASM TRIMMER CAPACITOR ASM TRIMMER CAPACITOR ASH TRIMMER CAPACITOR ASM CAP PCLYPRCPYL 500PF 10PCT 200V TRIMMER CAPACI TOR ASM TRIMMER CAPACITOR ASM TRIMMER CAPACITOR ASM TRIMMER TRIMMER TRI MMER TRIMMER TRIMMER TRI MMER PF 5PCT 500V CAP MICA 02 CAPACITOR ASM CAPACITOR ASM CAPACITOR ASM CAPACITOR ASM FACTORY ADJUST TRIMMER CAPACITOR ASM PANEL CONNECTOR PANEL C(~~ECTOR PANEL CC~~ECTOR PANEL CONNECTOR BUSHING ThREADED BANANATURRET PANEL CCMEC TOR RECPT BNC RECPT BNC RECPT BNC CONNECTCR PC 15 POS SR .150 SP RECPT BNC BULKHEAD.200 CABLE RECPT MICRO RIB 30 CONT RECPT MICRC RIB 50 CONl RECPT BNC BULKHEAD.206 CABLE RECPT BNC ORES 1 RES 2 RES 3 RES 4 RES 5 RES 6 RES 7 RES 8 RES 9 RES 24 POT 1 2 3 4 5 6 7 8 9 10 11 CCMP 100 K 5PCT 1/4W FLM 10K 51l0l>PCTl5PPM1/4W FLM lOOK 5/100PCT15PPM1/4W 1/10PC T 15PPM1I4W FLM 1M FLM 10 M IPCT 100PPM 1/4W FLM 100 M IPCT 100PPM 11410 FLM 80K 1/10PCT 15PPM1/4W FLM 800K 1/10PCT 15PPMl/4W FLM 8 M IPCT 100PPM 1/4W FLM 80 M IPCT 100PPM 1/4W loW Kf\CB lK OHM 5 PCT lOT SWITCH SWITCH SWITCH SWITCH SWITCH SWITCH SwiTCH SWITCH SWITCH SWITCH SWITCH RCTARY ASM RCTARY ASM SHIELD RCTARY ASM AS" ASM ASM ASM ASH ASM ASM PART OF SWITCH ASM 1616-2500 SUPPLY AS LOOSE PARTS ORIENTATION OF PART IS CRITICAL 5·12 SERVICE & DIAGRAMS PIN 1016-3000 NO. FMC 11505-4029 0505- ..02 8 24055 24655 PART .*••-•••• .*.*-.* •• MFGR PART NUM8ER il505-4029 0505-4028 4380-3725 74970 160-130 438v-361>u 1615-2120 1015-2120 1615-2110 1615-2110 1615-2110 1615-2110 1615-2110 1015-2130 1015-213" 1615-2180 4803-1569 1615-22VO 1615-2120 1615-2110 1616-6600 1610-0600 1610-6600 1616-660J 1616-6600 1010-0600 0+700-0364 1010-4810 1616-4810 1616-4810 1610-4810 .f- •• -++f-. 1615-2130 80583 24655 24655 24655 24655 24655 24655 2.. 655 24655 24655 24655 19390 24055 24655 24655 24655 24655 24055 24655 24655 24655 81349 24655 24055 24655 24655 MAC-Iv 1615-2120 1615-2120 1615-211i) 1615-2110 1615- 211 0 1015-21h.l 1615-2110 1015- 2130 1015-2130 1615-2180 PP481 56ilPF 1015-220v 1615-2120 1615-2110 1616-060U 1616-6600 1016-0600 1 bi6-6600 1016-06vO 1616-0600 CMu5ED620JN 1616-4810 1616-4810 1616-4810 1616-4810 24655 1615-2130 0874-4006 0874-4005 J814-4005 1>874-4006 4150-0900 0900-4410 4230-2300 4230-2300 4230-2300 4230-2715 423 .. -120" 4230-4036 4230-4049 4230-1200 4230-2300 24655 H655 24655 24655 24655 24055 24655 24655 24055 26601 91836 02000 02660 91836 24655 v874-4006 0874-4005 0874-4005 0874-4000 4150-0900 0900-4410 4230-2300 4230-2300 4230-23vu 143-015-08 KC-19- 161 57-40360-9 57-40500-4 KC-19-l01 4230-2300 oU99-4105 6019-3450 6019-3455 0619-3400 0188- 5100 6188-6100 661 <;-3465 0619-3470 0188-4800 6188-5800 0000-0142 81349 24655 24655 24655 24655 24055 2.. 655 24655 24055 24655 80294 RCR07GI04J 6619-3450 6619-3455 6619-3460 0188-5100 6188-6100 6619-3465 6619-3470 6188-4800 618B-5800 3540S-1-102 7890-5338 7890-5339 789D-825U 7890- 533 7 1616-2500 1010-2500 1616-2500 1610-2500 1610-2500 1616-2500 1616-250v 24655 24655 24055 24655 24655 24655 24655 24655 24655 24655 24655 7890-5338 7890-5339 7890-8250 7890-5337 1616- 2 500 1616-2500 1616-2500 1616- 2 500 1010-2500 l610-2500 1616-2500 10PCl ELECTRICAL PARTS LIST (cont) CHASSIS SWITCH SWITCH SWITCH SWlT CH SWI TCH SWITCH SWITCH SWITCH S\oIIT CH S\oIITCI- ASM ASI" A SI" ASM ASM ASM ASM ASM ASM AS'" S S S S S S S S S S 14 15 16 17 18 19 20 21 T T 1 2 TOROID TRANSfORMER CHOKE A 51" VR VR 1 2 ZENER lMZ-68 A ZENER lMZ-68A 13 AS,"I 68V 6av 5PCT 5POT PC BOARD R R R R R R R R R R R R R IJ 11 12 13 14 15 16 17 18 19 20 21 22 1.5\01 1.5\01 PIN RES RES RES RES RES RES RES RES POT POT POT POT POT 1.02K 1110 PCT 50PPM1I2" 10J OHMlII0PCT 50PPM 1/8\01 1 PCT 1I4W fL'" t49K t4.9K 1 PCT 1I4W flM 64.9K 1 PCT 1/4101 flM 1 PCT 1/4" FlM t4.9K 1/8W 1 PCT FlM 6.04K FLM 715 CHM 1 PCT 1/81>1 CERM TRM 200K OHM 10 PCT 15 T CERM TRM 20K OHM I.> PCT 15T CERM TRM 20K OHM 10 PCT 15T CERM TRM 20K OHM 10 PC T 1ST CERM TRM 2K OHM 10 PCT 15T flM flM 1616-31)0U NO. fMC MfGR PART 1616-2500 1616-2500 1616-2500 1616-2500 Ib16-25011 1616-25.:10 1616-2500 161b-250J 1616-250J 1616-2500 24655 24655 24655 24655 24655 24655 24655 24655 24655 24655 1616-2500 1616-'25UO 1616-2500 1616-2500 1616-2500 1616-2500 161b-2500 1616- 2 501i 1616-2500 1616-2500 1616-2310 1616-245v 24655 24655 1616-2310 1616-2450 6083-1064 6083-1064 24444 24444 lMZ-68A lMZ-68A 1.5R68B 1.5R68B NO. 6193-100J 6619-1600 6350-3649 6350-2649 6350-2649 0350-2649 6250-1604 6250-0715 6049-0193 604 a: fF _ iJOf,F _~ If! 100a~~"F t:::t::: r::,J;c:" , 1'0: I __I=t---=-1- _:~~ -1:,:::t~ " i 0J 0J ~5pF.-.05pF I .x •• I xX III 1 -:;----=1 z' I .XXpF I - ~> 10J 005~F --l;;;- "" >- i~. 't> I •• ro 10 ,0J 0J 'I;,'_,f [ZM!V] ,lo.F co oL .o len a: loc 0 1 '° 1 I L_-=r. I ! 112. 0 III 110 109 NOTE / A.UXllJAI:?V CONTACT SI/OWN ~i f-fj,,/1~r~~fJ!:f£~J£j~()N5 THIS' OF CIRCUIT IS lFPICAL S 5 - S.2 / . lOB 107 10.. v--t-~. ---t-- o ----+ __c-__ -t----.,-- + .. ('----+---,~)! v--+-----c'---+---o--._--_.+----<~-+-. "'r- 105 c --.._-t-----c~_+--o---t----~-t-----c--t--o----+--··- .. !3.>! 104 c· - -·\·-c~--+--o---+---·()·_--+--~,.~-t~·-o---t--'.2') 103 102 10/ IYTI2. WTI3 W! 5 112III iC,lCD PRECI SIGN CAPACITANCE BRIDGE SCHEMATIC DIAGRAM 110 10'3 108 107 lOb 105 104 103 102. 101 0,11 0,12 313 IC~PI\C-I Cli i)l>(lntatlons of R5, R9, and e17 are significant; properly oriented by factory. 514 TFlNC::E] SI5 51C, WTIG r-{ Ci'lBLE BR /_0 .;1 .8 7 -, TERMINI\L SELECTOR RE"'OoUT U'F;/e .", EXT o MULTIPLIER ~IO X I C"'L XIOO )\10 ~I -,--I ~~J .i i o 12. :0 co a: o " ~,1'Li'PI....,E~ • 5 f----- ------ -+-+----QwT310 WTIE> Ci'l8LE Wl-i-VT~ JII Figure 5-9. Schematic diagram of the 1616 Precision Capacitance Bridge. SERVICE & DIAGRAMS 5-13 AppendixA G900 CONNECTORS ~ a 3-TERMINAL UNKNOWN PORT / \ LOW GR 1616 BRIDGE = CA IF] Cx SPECIAL CABLES Cs - \. TEST FIXTURE Figure A-l. Test setup for 2-terminal measurerrent,with fringing eliminated. 1616·38 Appendix A An Interpretation of Millea's Method* to Eliminate Fringing from Two-Terminal Measurement. If a suitable 2-terminal standard is not available, and the uncertainty of ±.008 pF in the fringing of the G900 c. Provide a terminal for connecting a wire (274- LM) to the outer conductor or each G900 connector (points a. b). connector is unacceptable for your measurements, method is recommended. (Refer to para. 3.9) Working Capacitor. Use a 2-terminal coaxial capacitor with a G900 connector. This is C it may be your unknown x: capacitor or one you wish to calibrate as a standard. this Test Fixture. The assembly of adaptor, tee, cables, and wires is shown in Figure I The special cables (2 required) are not commercially available, but may be constructed as follows. a. Remove the G874 connector from one end of a Type 874- R22 LA Patch Cord. b. Install a G900 cable connector instead, Type 900C58. but do not connect its outer conductor to the cable shield. Leave a gap as shown, by cutting the shield braid back as far as the cable Jacket. Touch the cut ends of the braid back under the Jacket enough to assure insulation. The two should not be cut so short as to expose the inner conductor to external fields. The rubber sleeve should provide some strain relief and the cable retainer (by overlapping the shielded portion of the cable) sufficient shielding. It is imperative that the inner conductor of the connector be rigidly supported so that CA and C are constants. B A-l Procedure. a. With the connections as shown, i.e.. Cx connected to CA , a to h, and b to g, measure. C 1 = C A + CX' b. With the work ing capacitor transferred to the other arm of the fixture, i.e.. Cx connected to C B ' b to h, and a to g, measure C2 = C B + C x. c. With the fixture closed on itself, i.e., C A connected to C B ' a or b to h (neither to g), measure C3 = C A + C B d. Calculate the capacitance without fringing (C 1+C 2-C 3)/2 • Millea, Aurel, "Connector urement of Two-Terminal Research, 3, of the National 4, July-Dec., 1970. Cx = Pair Techniques for the Accurate MeasLow-Value Capacitances," Journal of Bureau of Standards, Vol 74C, Nos 3 & _GenRad GR1621 Capacitance Measurement System Supplement to GR1616 Precision Capacitance Bridge Manual 1 Introduction This supplement to the GR 1616 Precision Capacitance Bridge instruction manual (form number 1616--0100-(0) provides additional information that applies to the recalibration procedure. Keep this supplement with the basic manual and make an appropriate note in the margin on page 5-8, just after step e of paragraph 5.4.1. 2 Supplementary Procedure If in paragraph 5.4.1 step e, you find that the lo-nF and/or 100-nF capacitance trimmers lack sufficientrange to obtaina balance,use switches S22 and/or S23 as follows. SeeFigure 5-5. To Extendthe Range of the 1G-nF Trimmer 2.1 • Remove the instrument from its cabinet, as described in paragraph2.4 or 2.5. • Perform a trial balance, using only the 1-pF, lo-pF, and 100-pF lever switches not the trimmers,andlimiting the 100-pF switch to just two positions: 0 and -1. • If the error at balance is positive (lOO-pF lever switch set to 0), tum S22 one step counterclockwise; but, if it is negative (lOO-pF switch set to -1), tum S22 one step clockwise. • Set the 1-pF, lo-pF, and lOO-pF lever switches to their zero positions. • Returning to paragraph 5.4.1, step e, if the lo-nF trimmer now has sufficient range, reinstall the instrument in its cabinet andrepeat the calibration procedure of paragraph 5.4.1. But, if the lo-nF trimmer still has insufficientrange, repeat the above supplementaryprocedure. Part Number 1616-0150-00 Page 1 1616 PrecisionCapacitanceBridge,Supplement 2.2 To Extend the Range of the 10G-nF Trimmer • • • • • Remove the instrument from its cabinet, as described in paragraph2.4 or 2.5. Perform a trial balance, using only the 1-pF, lO--pF,and 100-pF lever switches not the trimmers,and limiting the 100-pF switch to just two positions: 0 and -1. H the error at balance is positive (l00-pF lever switch set to 0), tum 823 one step counterclockwise; but, if it is negative (lOO-pF switch set to -1), tum 823 one step clockwise. Set the 1-pF, 1Q-pF, and lOO-pF lever switches to their zero positions. Returning to paragraph5.4.1, step e, if the 1QO-nFtrimmer now has sufficient range, reinstall the instrument in its cabinet and repeat the calibration procedure of paragraph 5.4.1. But, if the 1QO-nFtrimmer still has insufficientrange, repeat the above supplementaryprocedure. CBox 822 FrontPanel 61000.0 Figure 5-5. Top View, Cover RemOVed, Showing locations of Switches S22 and S23. _GenRad >""\ 300 BakerAvenue,Concord, MA01742-2174 Page 2 Printedin U.S.A. Part Number 1616-0150-00 \2~~:;::;'/ For more information mil or the location of your local GenRad sales office, contact: GenRad U.S.A. Germany Asia Pacific 1-800-4-GENRAD TEL. 089/431990 TLX. 529917 GEND ~kn~a~~~~78-4400 Canada Italy TEL. 416890-0160 TLX. 06-986766 GENRADCO MSGA TEL. (02) 502951 TLX. 320 373 GENRAD I United Kingdom Switzerland European Headquarters TEL. 062839181 TLX. 848321 GENRADG TEL. 01/552420 TLX. 816828 GENRA CH France TEL. (040) 450605 TLX. 50535 GRBEN NL TEL. (1) 47970739 TLX: GENRA 220991 F Netherlands TLX. RS37808 GRASIA All Other Countries International Department 300 Baker Avenue Concord, Massachusetts U.S.A. TEL. 508 369-4400 TLX. 92-3354 01742 mtIIIIIIGenRad 300 Baker Avenu e , Concor d , Massac h usett s 01, Printed in U.S.A.
Source Exif Data:
File Type : PDF File Type Extension : pdf MIME Type : application/pdf PDF Version : 1.6 Linearized : No XMP Toolkit : Adobe XMP Core 4.0-c316 44.253921, Sun Oct 01 2006 17:08:23 Modify Date : 2007:09:10 11:52:10-04:00 Create Date : 2006:06:04 20:44:43-04:00 Metadata Date : 2007:09:10 11:52:10-04:00 Creator Tool : Adobe Acrobat 7.0 Format : application/pdf Document ID : uuid:c84bc80a-12e2-4b9b-875f-88fa833bea1b Instance ID : uuid:7de31133-bf70-46ee-ae2e-9c887192af8c Producer : Adobe Acrobat 8.0 Paper Capture Plug-in Page Count : 84 Creator : Adobe Acrobat 7.0EXIF Metadata provided by EXIF.tools