VIII SEMETRO

Towards Uniformity of DC Voltage Metrology within SIM

Yi-hua Tang Quantum Electrical Metrology Division

National Institute of Standards and Technology Gaithersburg, MD 20899

VIII SEMETRO

OutlineDC voltage metrology status within SIMJVS Comparison

MAPIndirect comparisonDirect comparison

Summary Future perspective

VIII SEMETRO

Lab / Country No. of JVS

NIST / USA 4 (20)

NRC / Canada 1

CENAM / Mexico 1

INMETRO / Brazil 1

INTI / Argentina 1

VIII SEMETRO

Principle of JVS

V = nf / KJ-90

f: frequency

KJ-90 = 2e/h = 483 597.9 GHz/V Josephson constant

n: integer

BIPM.EM-K10.a JVS at 1.018VBIPM.EM-K10.b JVS at 10 VBIPM.EM-K11.b Zener at 10 VSIM.EM.BIPM-K11.a Zener at 1.018 VSIM.EM.BIPM-K11.b Zener at 10 V SIM.EM.BIPM-K10.b JVS at 10V

VIII SEMETRO

MAP Example 1: NIST-BIPM JVS comparison at 10 V

3 system comparison:BIPMNIST-GaithersburgNIST-BoulderOct. 1998 – Jan. 1999BIPM.EM-K11.b

Pressure corrections applied.

VIII SEMETRO

d (μV) uc (μV) (k = 1)

NIST(G) - BIPM 0.26 0.14

NIST(B) - BIPM 0.22 0.17

NIST(G) – NIST(B) 0.04 0.13

Results of BIPM.EM-K11.b

Uncertainty: 1 - 2 parts in 108

VIII SEMETRO

INMETRO JVS system

Established in 1998

VIII SEMETRO

y = -0.00565391x + 295.00550325

97.1

97.2

97.3

97.4

97.5

97.6

9/3 9/8 9/13 9/18 9/23 9/28 10/3 10/8 10/13

Off

set f

rom

1.0

18V

(μV)

Date

Z3 INMETRO

NIST

Linear (INMETRO)

y = -0.02007355x + 827.94375914

125.6

125.8

126.0

126.2

126.4

126.6

9/3 9/8 9/13 9/18 9/23 9/28 10/3 10/8 10/13

Off

set f

rom

1.0

18V

(μV)

Date

Z5 INMETRO

NIST

Linear (INMETRO)

y = 0.01066458x - 280.46730707

92.1

92.2

92.3

92.4

92.5

92.6

9/3 9/8 9/13 9/18 9/23 9/28 10/3 10/8 10/13

Off

set f

rom

1.0

18V

(μV)

Date

Z6

INMETRO

NIST

Linear (INMETRO)

y = 0.00224443x + 35.74730979114.0

114.1

114.2

114.3

114.4

114.5

9/3 9/8 9/13 9/18 9/23 9/28 10/3 10/8 10/13

Off

set f

rom

1.0

18V

(μV)

Date

Z7INMETRO

NIST

Linear (INMETRO)

INMETRO-NIST JVS Comparison at 1.018VSept. to Oct. 1999Protocol: MAP SIM.EM.BIPM-K11.a

VIII SEMETRO

Source Uncertainty (μV)Pooled Type A of INMETRO, uA

INMETRO 0.003Pooled Type A of NIST, uA

NIST 0.006Standard deviation of mean of four Zener differences uB

INMETRO -NIST0.011

Type B uncertainty from INMETRO and NIST JVS systems, uB

INMETRO, NIST0.016

Expanded uncertainty (k = 2) 0.049

cu = 2 2AINMETRO(u ) + 2

ANIST(u ) +

2

B

INMETRO−NIST(3.18

2 u ) + 2BINMETRO, NIST(u )

Z3 Z5 Z6 Z7

INMETRO - NIST 0.023 0.052 0.001 0.013

Δ(INMETRO− NIST ) =(Vi(predict) − Vi(NIST )[ ]

13∑

13

VNISTINMETROMean μ022.0)( =−

VIII SEMETRO

Question:

Is MAP good enough for JVS comparison?Answer:

No

Why?1. JVS uncertainty: few parts in 1010

MAP uncertainty: few parts in 108

2. Uncertainty using MAP limited by Zener characteristicsEnvironmental effects• Pressure• Temperature• Relative humidity

Zener intrinsic noiseShipping impact

VIII SEMETRO

Question:

How can we improve the situation?Answer:

Develop a transportable JVS

8.5”

Net weight: 50 lbGross weight: 120 lb

VIII SEMETRO

NIST-CENAM SIM.EM.BIPM-K11.bNIST CJVS in CEMETRO, March 2006

VIII SEMETRO

Dionisio HernándezEnrique Navarrete David AvilésYi-hua Tang

VIII SEMETRO

Results of NIST-CENAM SIM.EM.BIPM-K11.b

No pressure correction needed.

VIII SEMETRO

d (μV) uc (μV)

CENAM - NIST -0.035 0.043

NIST - BIPM 0.26 0.28

CENAM - BIPM 0.23 0.28

DCENAM-BIPM = (dNIST-BIPM + dCENAM-NIST )

UCENAM-BIPM = (UNIST-BIPM2 + UCENAM-NIST

2)1/2

To establish a link between CENAM and BIPM

VIII SEMETRO

VIII SEMETRO

Question:

Is indirect JVS comparison good enough?Answer:

No

Why?JVS uncertainty: few parts in 1010

Indirect JVS comparison uncertainty: few parts in 109

Uncertainty limited by Zener noise

Environmental effects• Pressure• Temperature• Relative humidityZener intrinsic noiseShipping impact

VIII SEMETRO

Question:

How can we improve the situation?Answer:

Do not use Zener as transfer standards. Make direct JVS comparison.

Protocol:

BIPM JVS comparison Option a; BIPM JVS comparison Option b;Alternative protocols

VIII SEMETRO

An alternative protocol

Step 1 Step 2

Step 3 Step 4

VIII SEMETRO

NIST-NRC SIM.EM.BIPM-K10.b

Barry Wood of NRC

NIST CJVS at NRC August 2007

VIII SEMETRO

VIII SEMETRO

Link between NIST and BIPM via NRC

VIII SEMETRO

-11

-9

-7

-5

-3

-1

1

3

5

7

9

11

0 5 10 15 20 25 30 35 40

Meas. number

U (N

IST)

- U

(BIP

M) /

nV #1 to #5#6 to #8#9 to #10#11 to #15#16 to #25#26 to #30#31#32 to #37

NIST-BIPM JVS comparison BIPM.EM-K10.b

BIPM protocol Option B

VIII SEMETRO

Histogram of the JVS comparison

BIPM protocol Option B

VIII SEMETRO

-10

-8

-6

-4

-2

0

2

4

6

8

10

12

0 5 10 15 20

Measurements

U(N

IST)

- U

(BIP

M) /

nV

#1 to #11#12 to #22

NIST-BIPM JVS comparison BIPM.EM-K10.b

BIPM protocol Option A

VIII SEMETRO

Option B Measured by NIST

Option A Measured by BIPM

Difference (nV) -0.074 -1.53

Type A (nV) 0.68 0.75

Type B (nV) 0.93 0.77

Combined uncertainty (nV) 1.15 1.07

U (NIST) – U (BIPM) / nV

VIII SEMETRO

Direct link between NIST and BIPM

VIII SEMETRO

BIPM JVS in NIST March 2009

VIII SEMETRO

Stephane Solve Regis ChayramyYi-hua Tang

VIII SEMETRO

Vd = Va1 – Va2 = (N1 f1 – N2 f2 ) / KJ-90

Vd - Vm = Vo + mt + Vn + δ

δ

= Vd – Vm – Vo – mt – Vn

Model of NIST protocol

Automatic JVS direct comparison

VIII SEMETRO

Automatic JVS direct comparison

VIII SEMETRO

Example: automatic comparison

nominal voltage: 9.997 V

217 points in 14 h. U (NIST) – U (BIPM) / nV ua / nV

0.02 0.24

VIII SEMETRO

Not listed in BIPM KCDB

BIPM.EM-K10.b

SIM.EM.BIPM-K11.bSIM.EM.BIPM-K10.b

BIPM.EM-K10.a

BIPM.EM-K11.b

Summary

VIII SEMETRO

MAP(Indirect)

CJVS(Indirect)

JVS vs. JVS(Direct)

PJVS vs. PJVS(Future)

Voltage range Up to 10 V Up to 10 V Up to 10 V Up to 10 V

Uncertainty 2 x 10-8 2 x 10-9 1 -

7 x 10-10 parts x 10-11

Limiting factor Zener Zener(1/f noise)

Null detector Null detector

Time needed Weeks Days Hours Hours

Expense Low High High Potentially low

Comparison of various protocols

VIII SEMETRO

Future perspective

Coming soon: new 10 V PJVS

Expected uncertainty: few parts in 1011

at 10 V

Detection of small system error is possible