1. IMPROVED AC-DC CURRENT TRANSFER STANDARDS FROM 10 mA TO 100 A
Torsten Funck

2. NCSLi 2006 AC-DC current transfer standards
Contents
Introduction
Quartz-PMJTCs for current transfer
Calculation of the current transfer difference
Validation of the model used for calculation
Measurement setup using driven guards
Step-up chains
Special developments: transconductance amplifier and current transformer
Improved measurement uncertainties
Conclusions
NCSLi AC-DC current transfer standards

3. NCSLi 2006 AC-DC current transfer standards
Introduction
Electrical quantities are defined, realized an maintained at direct current (dc).
For alternating current (ac) calibrations, traceability is given due to ac-dc transfer.
For current, the measured quantity is d i , the ac-dc current transfer difference:
NCSLi AC-DC current transfer standards

4. NCSLi 2006 AC-DC current transfer standards
Quartz-PMJTCs
Planar multijunction thermal converters (PMJTCs) on a quartz substrate are used as a calcuable standard
Very low conductivity of the substrate
Low dielectric constant er
Low, calculable transfer difference
Difficult to manufacture
Low thermal time-constant
Low sensitivity
NCSLi AC-DC current transfer standards

5. Calculation of d i due to changing heater impedance
/10 R C 2 1 G
vs (f)
NCSLi AC-DC current transfer standards

6. Calculation of d i due to stray capacitances
NCSLi AC-DC current transfer standards

7. Current transfer differences
NCSLi AC-DC current transfer standards

8. Validation of the calculation
NCSLi AC-DC current transfer standards

9. Uncertainties using built-in Tee
NCSLi AC-DC current transfer standards

10. Measurement setup using driven guards
Symmetrical setup
Upper standard connected „upside down“
Constant voltage across parasitic capacitances => equal transfer differences in “upper position” and “lower position”
NCSLi AC-DC current transfer standards

11. AC-DC current transfer standards
Low currents: PMJTC with low heater resistance
Problem: Technology limits 90 W < RH < 900 W
Medium currents: Shunt + PMJTC
Problem: Shunt inductance and current dependence
High currents: Cooled shunt + PMJTC
Problems: Shunt inductance and current dependence, compliance voltage of current source
NCSLi AC-DC current transfer standards

12. NCSLi 2006 AC-DC current transfer standards
Shunts used at PTB
Up to 300 mA: PTB design (top)
500 mA to 5 A: Holt design (middle)
10 A and 20 A: Fluke design (bottom)
NCSLi AC-DC current transfer standards

13. Shunts in Justervesenet design
+ Low cost materials
+ Large number of resistors ensure equal current distribution
+ Very small area of current path yields low inductive coupling
- No case
Justervesenet
NCSLi AC-DC current transfer standards

14. NCSLi 2006 AC-DC current transfer standards
High current shunts
EL-9800
(NIST Design)
Set of four shunts:
10 A to 30 A (25 mW)
30 A to 50 A (10 mW)
50 A to 80 A (5 mW)
80 A to 100 A (3 mW)
NCSLi AC-DC current transfer standards

15. Step-up chains using different standards
5 A 7
2 A 6
1 A 5
3 A
500 mA 4
200 mA 3
300 mA
100 mA 2
50 mA 1
30 mA
20 mA
start C B A
Step \ Chain
Chains A & B: Self-made and Holt shunts
Chain C: Justervesenet-Shunts (two sets)
Comparison of the chains: 1 A at 1 MHz:  < 10 µA/A
5 A at 100 kHz: < 5 µA/A
NCSLi AC-DC current transfer standards

16. Measurement of the current level dependence
Assumption:
Current level dependence of shunts results from self-heating
Measurement approach:
Shunt with low voltage drop used as a standard
=> Low power => low self-heating => low current level dependence
Low voltage drop => voltage-amplifier necessary to operate PMJTC
NCSLi AC-DC current transfer standards

17. Measurement of the level dependence: Shunt with Amplifier
Device
under test
Standard
NCSLi AC-DC current transfer standards

18. Transconductance amplifier
Commercially available current amplifiers suffer from
Low bandwidth (< 100 kHz)
Insufficient dc stability
Varying gain factor (+/-)
Design goals
Bandwidth 1 MHz
Low dc reversal difference
Stable gain, equal for both polarities
Transconductance adjustable by changing external shunt
NCSLi AC-DC current transfer standards

19. Principle of transconductance amplifier
NCSLi AC-DC current transfer standards

20. Measuring the level dependence at high current
NCSLi AC-DC current transfer standards

21. NCSLi 2006 AC-DC current transfer standards
Current transformer
1: Coaxial connector (outer conductor), 2: Core with secondary winding, 3: Primary winding (inner conductor), 4: Secondary connector
NCSLi AC-DC current transfer standards

22. Uncertainty of measurement
The following sources of uncertainty of the step-up procedure can be identified for each step:
u (std) Standard uncertainty of the (calculable) transfer difference of the standard used, i.e. uncertainty of previous step
u (C) Standard uncertainty of the comparison procedure
u (A) Type A standard uncertainty, i.e. the standard deviation of the mean
u (Lev) Standard uncertainty due to the current level effects
in the shunt
u (LF) Standard uncertainty due to PMJTC low frequency effects, which are level dependent
NCSLi AC-DC current transfer standards

23. Improvements in uncertainties of measurement
Red : old (2002)
Blue: new (2006)
NCSLi AC-DC current transfer standards

24. NCSLi 2006 AC-DC current transfer standards
Conclusions
PMJTCs on quartz substrate developed at PTB are well suited for current transfer at low current up to 1 MHz
Potential driven guarding reduces the uncertainty of the comparison measurements
New shunts with low current level dependence and low transfer difference allow further reduction of the step-up procedure's uncertainty
Extension of the measurement capabilities to 1 A at 1 MHz using new transconductance amplifier developed at PTB
New approaches for the determination of the current level dependence allow for corrections and therefore for further reduction of the uncertainty
NCSLi AC-DC current transfer standards

25. Thank you for your attention!
Torsten Funck
NCSLi AC-DC current transfer standards