[volt-nuts] What is Transfer Accuracy?

Frank Stellmach frank.stellmach at freenet.de
Sun May 4 13:41:49 EDT 2014


this term is quite fixed.
You will find that specification for all "premium" instruments; where 
this parameter is lacking, it's no good instrument, either.

It describes, how precisely one can compare two different artefacts 
against each other.
As an example, you want to transfer the "accuracy" of a 732B 10V 
standard to your home standard.
The transfer accuracy of an HP3458A specifies 0.1ppm accuracy within 10 
minutes and 0.5°C.

That means, your home standard will be be measured within 0.1ppm 
compared to the 732B, and that, despite the fact that the 3459A in all 
other stability/accuracy figures is about an order of magnitude worse.

This is also valid for non-equal values, i.e. if you compare 7,2V 
(directly out of a LTZ1000) against10V from the 732B, and that's due to 
the ultra high linearity of the 3458A (typ. 0.02ppm).

It's a pity, that for this instrument the Ohm function transfer 
stability is not specified.
I tried to specify it on my own by external very stable resistors 
(VHP202Z), which were temperature stabilized or temperature-measured. 
And I got about 0.2ppm within 10min and 0.2°C.

But in the end, what I am also missing, is an official document or 
procedure, how this transfer accuracy is to be measured correctly / 

Perhaps somebody can name a source for that, perhaps at NIST?

Same goes for the definition of linearity, integral and differential, 
for example.
I also had to guess that, when I measured the linearity of my 5442B.


More information about the volt-nuts mailing list