Calibration stability

[gar]

090726-2059 EST

In 63 or 64 I bought a Ballantine 420 AC and DC Calibrator. This used an OA2 voltage regulator tube for the reference. Clearly these were aged and selected for stability.

For convenience in May 1964 I built a small DC voltage reference box with outputs of 12.1, 10.0, 1.0, and 0.1 V. This was a voltage divider from a nominal 12.1 V temperature compensated Zener diode. This box was calibrated to the Ballantine at 10.0 V using a voltage difference method.

Today my voltage reference reads the following using a Beckman 4410 meter:
Input DC
AC V Ouput

100 --- 9.976
110 --- 9.978
120 --- 9.980
130 --- 9.982
100 --- 9.978
130 --- 9.982

15 minutes later
120 --- 9.981

At 120 V input to both the Ballantine and my box the reading of my box was
0.0054 V higher than the Ballantine.

Rather good stability of the two devices over the years.

These readings do not give me an absolute value for either box. A Fluke 27, lower resolution than the Beckman, produced a reading of 9.98 V. I would conclude that the Ballantine was slightly off calibration when new and still is. Did the Ballantine drift 0.0054 over the years, or my Zener, or both. No way to know. It is more surprising that the OA2 was this stable over this period of time. However, the Zener diode was from the early days of making these devices.

.

 

[SAC]

Did the Ballantine drift 0.0054 over the years, or my Zener, or both. No way to know.

Were any other components used in the voltage difference method that was originally used in the calibration, and which could contribute some error to the original measurement? It may be interesting to repeat the calibration today and see how much error results.

I agree that the results seem quite good for 45 years. My guess is that is mostly a measure of the storage life of the devices and not the operational life of the devices? I guess that is a pretty well-built tube to be holding the vacuum so well after so long and with little contamination.

 

[gar]

090810-2225 EST

SAC:

The original meter used to measure the voltage difference was an HP 412A with an 0.001 V full scale range on the lowest range. This meant that 10 micro volts could be seen. This meter is not in operational condition today.

The OA2 voltage regulator tube is a gas discharge type regulator. As I believe I mentioned it was probably aged and tested for short time stability at the time of manufacture.

The 412A used a rotating light chopper to modulate photocells to convert DC to AC for amplification. These may have been photo-conductive cells.

.

 

[SAC]

Ahhh - one of these:

http://www.messmuseum.de/hp412a.htm

1% of full-scale 1mV is pretty good!