Text preview for : nf_ncsl02.pdf part of Fluke nf ncsl02 Fluke appnotes nf_ncsl02.pdf
Back to : nf_ncsl02.pdf | Home
VERIFYING PRECISION VOLTAGE DIVIDERS AND HIGH RESISTANCE
RATIO BRIDGES USING A JOSEPHSON VOLTAGE STANDARD
Neil Faulkner
Fluke Corporation
Abstract:
Directly verifying the ratio accuracy of precision voltage dividers which have a specification
around 0.1 to 0.2 ppm at a division ratio of 10:1 is very challenging. A Josephson Voltage
Standard (JVS) has the required uncertainty for the measurements but the output resistance of
these dividers is high enough (>40 kOhm) that a direct measurement of the divider output with
the JVS has high uncertainties. A method is described using a second low resistance divider
which yields acceptable uncertainties with the JVS. The results of the measurements on two
models of dividers is given. This method can also be used to verify high resistance ratio bridges.
Also discussed are refinements that were made to the Fluke JVS in Everett, WA to reduce the
uncertainty of measurement at 1 V and below.
1. Introduction
The verification of the division ratio of a voltage divider can be accomplished by comparing it to
another divider whose ratio is accurately known or by using a device such as a DMM that has
excellent linearity or by measuring the input and output voltage with a sufficiently low
uncertainty. Any of these methods becomes difficult to do when the divider has a division ratio
specification as low as 0.1 ppm and an output resistance as high as 65 kOhms as does the Fluke
720A. The Fluke Primary Standards Lab was requested to verify this divider in a way that was
traceable and accredited. NIST does not require traceability for self calibrating ratio devices such
as the 720A yet some customers wanted some kind of a verification through instrumentation that
was traceable and accredited. The Josephson Voltage Standard (JVS) System in the Lab had the
required uncertainties to measure the input and output voltages of the divider and these
measurements were traceable and accredited. The problem was that the uncertainty of measuring
the output voltage was too high due to the affect of the high output resistance of the divider. A
method was developed using a voltage source with a low output resistance divider that solved
this problem. This divider was built so its division ratio could be adjusted until its output voltage
was equal to the output of the divider under test as indicated on a null detector. Then the output
of this divider was measured on the JVS. The JVS was used to measure the input voltage to the
divider under test and the division ratio was then calculated.
This paper discusses the problems that were overcome by using this divider, some changes made
to the Fluke JVS to improve uncertainties at 1 V and below and some results achieved with this
technique.
2002 NCSL International Workshop and Symposium
2. Improvements made to the JVS to reduce uncertainties at low voltage
A JVS is an adjustable source of voltage from 0 to