[volt-nuts] Calibrating an HP 34702A?

[Marvin E. Gozum]

If you don't have access to formal calibration equipment, then what 
can be done is a best approximation by calibrating your DUT against a 
known calibrated DMM.

You'll need voltage sources down to resolution of the DUT, and 
resistances for each range, and you'll use these to transfer the 
calibration of the DMM to the DUT; even non-metrological sources may 
be stable enough to transfer measurements made from seconds to 
minutes of measurement from the calibrated meter, to the DUT.

Your DUT will not be calibrated to spec, but incudes the errors of 
the calibrated meter, and the error of your transfer references.

For each range, use the voltage, current or resistances in the DUT's 
calibration manual, but make adjustments to match the calibrated DMM reading.

For a final check you can compare both meters reading random voltage 
within every range, sampling untested resistors and currents, all 
appropriate per range.

Why do this?  At least the repaired DMM will be in a known 
measurement confidence area compared to totally unknown 
territory.  Even if you don't have variable voltages or resistances, 
at least you can get some ranges in order with whatever you have, 
assuming the DUT does not have ranges that are dependent on the 
status of a one or several other ranges that are not calibrated.

The same rationale can be used to get the AC ranges in order.

In the absence of formal calibration equipment, its better than 
leaving it as is.



At 01:37 PM 2/23/2011, Charles wrote:
>First post :)
>
>Having recently repaired a nonworking $5 hamfest special HP
>34702A/34740A (4-1/2 digit DVM/display unit), I am interested in
>calibrating it. Without paying many times my purchase price to a
>lab!
>
>I obtained some sample MAX6325EPA (2.500V +/- 0.02%) which gets me
>"pretty close" for the 10V range. I also have a handy 596.5 ohm,
>0.01% resistor for the Ohms ranges.
>
>The basic DC accuracy spec for the HP is something like +/- 0.03%
>plus 0.01% of range, so I need a better standard than that. Also I
>don't have a good way of generating very stable 10 KHz AC signals
>to cal those ranges...
>
>Much of the discussion on this list is 7.5 or 8.5 digit meters, so
>a 4.5 should be easy for you "Nuts". Any tips?
>thanks
>-Charles



Sincerely,



Marv Gozum
Philadelphia, PA