[volt-nuts] DMM calibration

[John Phillips]

I did some work for a resistor manufacture  and was told about resistor
drift. It does not keep going the same direction forever it does hit a
limit and start back the other way. 4 years is just an approximation. The
material and environment will have a lot to do with the period. Agilent
burns in  their  voltage references and the main difference is how long
they "cook" them. They also screen to get a better reference. evenhandedly
the drift rate and amplitude drops. In a few hundred years things will be
stable or broken. Broken is very stable. Like wire resistance will not keep
going down till it hits 0 or up until it opens.
Agilent says do not turn your 3458A off or it will not be as stable as if
it has been running for years.
Agilent move some reference manufacturing off shore, when the references
met the stability numbers they were harvested and put in stock bins. Latter
when they were used they no longer met the spec. There was a memo sent out
that if you had one of these references its stability should be withing
spec if the meter was not turned off for 6 mo. (that is how I remember it).