HP 6920B Meter Calibrator

Thanks to Hylaphone over on the Antique Radio Forum, I finally have a real meter calibrator.

IMG 5708.jpg

The really useful function that I couldn’t otherwise produce in my lab is AC up to 1000 volts. I also suspect the high DC voltages are more stable than my Power Designs model 2K-10 HV power supply.

IMG 5713.jpg

Hylaphone did a nice job of recapping it. It checks out to be working fine, but it is a little out of spec according to my (not recently calibrated) HP 3468A, 

IMG 0181.jpg

and compared against my Geller 10 volt reference (also not recently calibrated).

HP Agilent Keysight still has a manual scan available on their web site. It has a January 1976 revision date.

IMG 5706.jpg

There is also an older manual scan at the HP Archive site with an October 1965 revision date, but that scan has less information and the calibration instructions appear to have at least one error. 

For example, they tell you to connect a scope between test points 1 and 4, with the scope ground on test point 4. That would short out the 12.6 volt DC reference circuit. I think that is a typo and it should have been test points 1 and 2.

On the other hand, the calibration instructions in the newer manual also sound a bit sketchy. It would have you set the inverter phasing by eyeballing the output on a scope. The older manual has a detailed procedure for balancing the DC offset using a scope.

Does anyone have advice on calibrating the 6920B?

IMG 5707.jpg

IMG 5709.jpg

IMG 5710.jpg

IMG 5711.jpg

IMG 5712.jpg

IMG 5713.jpg


Over on the Antique Radio Forum, Alan S cautioned me that the HP 3468A, being a true-RMS meter, would diverge from the 6920B’s average-sensing AGC loop if the AC power line had much harmonic content. I tried to calibrate the AC ranges using the HP 3468A anyway, but I was never confident in the result. Also, I was never satisfied with the inverter phasing adjustment using the instructions in the newer manual. I couldn’t seem to get rid of all the glitches in the waveform.

IMG 4231.jpg

I finally took a stab at driving it from my General Radio 1310-B audio oscillator. I tuned it to 60 Hz using my Heathkit counter. I thought that would resolve any problems from harmonics, but when I re-did the inverter phasing adjustment, I couldn’t get a stable clean output signal. Glitches would come and go over time; I would adjust them away, and then they would reappear and grow over time, then slowly disappear, only to come back again.

Finally the light bulb went on in my head, and I recognized that the GR 1310-B was slightly off frequency from the power line and so I was seeing beats between the power line ripple and the HP 6920B inverter frequency, which was controlled by the GR 1310-B.

IMG 4232.jpg

I don’t have a photo of the setup handy, and I’m not at home to take one right now, but the GR 1310-B has a phono plug on the left side that is both a sync signal output and a sync input. I used a 6.3 volt filament power supply to drive the sync input to sync the 1310-B to the exact power line frequency. Bingo - I was able to easily and stably adjust the inverter phasing by eyeballing the scope, and I now had confidence that the HP 3468A true-RMS reading matched the average sensing output from the HP 6920B.

So, I finally have a calibration setup that I trust. Yay!

© Steve Byan 2011-2016