A novice calibration question

[gray5596]

A novice question.

Using a stable 10 volt reference, I was checking the accuracy of two 5.5 digit multi-meters.  One read slightly high, 10.0001 volts. The other read slightly low 9,9999 volts (occasionally flickering to 10.0000). Because the 1st meter was reading slightly high, I was able to get 6 digits of information but only 5 digits of information from the 2nd meter. 

To prevent this from happening, why aren’t common voltage references slightly higher? For example 10.1 volts would allow both meters to stay in the 6.5 digit mode and it would be easier to compare.

I’m just curious. 

[Bud]

They would Not be references then 

[bdunham7]

Using a stable 10 volt reference, I was checking the accuracy of two 5.5 digit multi-meters.  One read slightly high, 10.0001 volts. The other read slightly low 9,9999 volts (occasionally flickering to 10.0000). Because the 1st meter was reading slightly high, I was able to get 6 digits of information but only 5 digits of information from the 2nd meter. 

Although 10.0001 does actually have 6 digits and 9.9999 only has 5, there's no real significance to that difference in terms of precision or how much "information" you have.  You still have the same number of signficant figures after the decimal and the LSD (last digit) represents the same amount, 100µV per count. 

[Kleinstein]

The actual resolution is still the same.

A few meters may actuall only show a result of 10 V at the may or may even max out at 9.999 or so. For these meters it can actually be better to have a reference voltage a little lower than the rounded value.
This is less an issue at 10 V, but can be a real point at some 2 V, with many meters having +-1999 counts. For this they have ref. voltages at 1.9 V and similar values with calibrators.

[dobsonr741]

The 10V reference is not really 10V sharp. It’s a zone, defined by the initial or calibrated accuracy, the long term and thermal stability, and the load regulation.

To transfer calibration, your source should be better, I’d say by a magnitude, to have a meaningful ability to judge the accuracy of a target.

How accurate is your 10V reference?

[gray5596]

Thanks for all the answers and ideas.

The 10 volt reference I was using was not anything spectacular, a AD587LN in a sealed case. It looks stable when warmed up and when measuring it using my HP34401A in the 6.5 digit mode.