[volt-nuts] Guidance requested

[John Phillips]

I believe that 3458A will be more stable if it is never turned off.
The references are cooked in. There was a problem when production was moved
offshore, the references were stored cold after they had been cooked into
spec. When these units were shipped they did not meat spec. There was a
memo stating that if left on for 6 months the references should return to
spec.
We stored all our 3458As powered up even if we did not anticipate using
them for months.


On Sun, Jul 9, 2017 at 11:24 PM, Frank Stellmach <frank.stellmach at freenet.de
> wrote:

> So I have:
>
> 1) A Datron 4808 (not calibrated recently AFAIK, and no way to check),
> 2) A 3458A which is about 6 months out of calibration (sadly not by
> Keysight),
> 3) Three 7081s two of which are also 6 months out of calibration against
> the 3458A,
> 4) A 720A.
> The readings of the output of the Datron for the 1V, and 10V, 100V and
> 100V ranges on the 3458A (NPLC100) are:
> 1.00000177V (high limit 1.00000480V),
> 9.9999964V (low limit 9.9999720V, 99.999380V and climbing slowly (low
> limit 99.999550V)
> In spec some five or so minutes later.
> 999.98xxxV (and falling) (low limit 99.99450V) So definitely out of spec
> To be fair both the 4808 and 3458A have only been on for about 8 hours -
> things may improve :).
>
> So please how best to proceed from here?
> Thanks Dave
>
>
>
> Dave,
> at first, all instruments need a stable temperature environment, otherwise
> ppm measurements are meaningless,
> I recommend setting up the lab in the basement, if available.
>
> The 3458A is stable after about 2h, or 4h at most.
> Check its inernal temperature, and use ACAL DCV frequently.
>
> As the 3458A and the 4808 agree to <1ppm on the 10V range, there's good
> reason that the basic DCV calibration of the 3458A is still fine. Both
> these instruments do not drift that much if they are old vintage, and if
> they are not powered on 24/7/365.
> If the 3458A was unpowered most of the time, its reference very probabaly
> did not drift at all.
>
> If your 732A agrees also on 10V within a few ppm, you would have another
> fix point.
> Maybe you can send in the 732A for comparison to another volt-nut, or
> build a transportable 7,15xxxV reference with an LTZ1000, as described in
> the eevblog thread.
>
> Because the 3458A is an AUTOCAL instrument, the 1V, 100V readings are
> precise to about the same level, as the 10V range.
> The 3458A makes better ratio transfers than the 720A, butter latter is
> good for linearity check, anyhow.
> 1kV depends on the 100:1 divider of your special 3458A instrument, it can
> be precise as a few ppm only, but unprecise as 12ppm, as described in the
> specification. For 1kV measurements, all instruments need at least 1min
> stabilization before making measurements, due to power induced temperature
> drift.
>
> To use the 720A on 1kV, may not give better accuracy, either, again due to
> the high power drift.
> A 752A would be required, or another ACAL instrument, like the 5440A, or
> the 5720A.
>
> The references inside the 7081A may drift the most, so I would trust them
> less.
>
> For Ohm, you may want to follow TiNs proposals.
>
> Frank
>
>
> _______________________________________________
> volt-nuts mailing list -- volt-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/m
> ailman/listinfo/volt-nuts
> and follow the instructions there.
>



-- 

*John Phillips*