[volt-nuts] What is Transfer Accuracy?
Marv @ Home
marvin.gozum at comcast.net
Sun May 4 18:41:43 EDT 2014
I think its fuzzy but most all good DMM have it specified using some term.
Its a term I see HP or Agilent use mostly.
Fluke 5808a uses the term 'Transfer Uncertainty', which is inline
with metrology replacing the term 'accuracy' with 'uncertainty' _and_
a confidence interval [ which was undefined by the term 'accuracy'].
A common understanding of 'transfer accuracy' presumes it could be
used as a transfer reference in a pinch. Transfer references in
general are secondary references used to transfer the primary
reference value to other devices; a Fluke 732b is a transfer
reference, its portable and has defined accuracy to 30 days, while
the primary reference, say a Josephson Junction, cannot be moved from
Per manual, the 3458a 'transfer accuracy' are specified for 10
minutes to a primary reference and requires the original calibration
temperature +/- specified limits [ this is vital to control for ppm
drift/temp]. For 10V scale, 10 minute accuracy is 0.05ppm, its 24h
spec: 0.5 ppm + 0.05 ppm of scale. The other quoted accuracy is the
1 hr accuracy of 0.1ppm.
As Dr. Frank points out very well the great value of this spec is
making ratio measurements. However, its not implied by the stated
manual specification. OTAH 10 minutes or even 1hr is barely enough
time to run from the primary standard to your lab, so it must be used
in other ways.
Ratio measurements are independent of DMM drift since calibration:
its most dependent on the shortest term accuracy, aka 'transfer
accuracy', ambient temperature, linearity of the device and
naturally, the accuracy of your reference. Based on the manual
specs, the longer you delay making the ratio measurement the more
likely the 3458a will drift, so keeping it under 10 min give you the
best results; beyond it you're then at the 1 hour accuracy, and
beyond 1 hr, you'd use the 24 hr accuracy.
At 12:44 PM 5/4/2014, Jan Fredriksson wrote:
>Is there a common understanding about what "transfer accuracy" means or is
>it more of a fuzzy term?
>volt-nuts mailing list -- volt-nuts at febo.com
>To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
>and follow the instructions there.
More information about the volt-nuts