What kind of rectifier used in analog AC voltmeter?

[schmitt trigger]

At my old company (Paton Electrical) we used 4 x OA91 in a bridge format.

https://www.universitypaton.com.au/page/10/our-history-university-paton

Love the factory floor. Those huge ceiling windows must have provided plenty of natural light on the work area

[MrAl]

Schottky diodes should be reasonable similar to a GE diode.
The meters usually have a scale that is nonlinear at the low end to compensate for the drop at the diodes. The problem with replacing a diode (or both) is that the type of diode effects the details of the nonlinear scale.
Both the Ge diodes and Schottky diodes can deviate a bit from the ideal curve and there is also the current scale from the size of the diode. Larger diodes have naturally more current at the same voltage.

If only one diode is bad one could measure the drop at a low current (like 1/3 the meter movement, ~ 20 µA) and look for a diode with a similar forward voltage. If needed parallel diodes could be used for a slightly lower voltage.

Thanks but it's too late now. However, when I check it with the diode check function of the Fluke 189 it's 1.4V. Strange.

Hi,

If you change the diodes I would recommend recalibrating the meter.
I am not sure what the scale looks like on your meter but it may involve changing that if nothing else works.

[Kleinstein]

One should definitely check the accuracy after a diode change. However by now the old analog meter would no longer be used for high accuracy AC. The AC accuracy was never really good for the analog meters, especially not at the lower end of the range (like < 20%).

[radiolistener]

Old-school AC voltmeters, such as the В3-48, measured AC voltage by amplifying the alternating signal and feeding it to a thermoconverter -essentially a heating element. The measurement was performed by detecting the temperature in the thermoconverter. And they was pretty precise in AC RMS measurement.



With diode rectifiers you can't measure proper AC voltage. You can use diodes for a rough estimate of AC voltage with a known and clean waveform shape, but not for precise AC measurements.

These days you can use true-rms DMM or oscilloscope for such kind of measurement.

[floobydust]

Simpson 1-115970 1N87A gold-bonded Ge diode. That diode goes back to 1958, used as a video detector. Used in 260-2 to 260-5. Then changed to 1N270/276 in 260-6. Then changed to 1N100 in 260-7.

Weston Copper Oxide Rectifiers Used in Measuring Instruments 1946 source url
Attached Vf curve for it, as old hat.

[TimFox]

Old-school AC voltmeters, such as the В3-48, measured AC voltage by amplifying the alternating signal and feeding it to a thermoconverter -essentially a heating element. The measurement was performed by detecting the temperature in the thermoconverter. And they was pretty precise in AC RMS measurement.



With diode rectifiers you can't measure proper AC voltage. You can use diodes for a rough estimate of AC voltage with a known and clean waveform shape, but not for precise AC measurements.

These days you can use true-rms DMM or oscilloscope for such kind of measurement.

Thermal converters are very accurate, since they use a servo with two matched resistor-thermocouple (or thermistor) combinations in a thermally-insulated enclosure.
The servo adjusted the DC voltage across the second resistor until the two thermocouples gave equal outputs:  the DC voltage across the second resistor therefore equalled the rms voltage across the first resistor, and could be measured accurately with a DC voltmeter.
They can tolerate a large "crest factor", the ratio between peak and rms voltages, compared with solid-state rms converters.

[BeBuLamar]

The 1N34 germanium diode came in today. The forward bias voltage is 0.25V. Since the original R23 was bad so I replaced it with a 1K 10 turn trim pot. With the parts replace and some adjustment of the trim pot it now works very well. Well now I want to see how I can improve the accuracy in the DC voltage measurement. The 260 series 2 doesn't have any adjustable potentiometer like the newer version.

[floobydust]

If the meter movement is out, it affects all functions. I think Simpson later trimmed for the meter movement resistance using associated 50uA full scale calibration, and another trimpot for full scale.
But the 260-2 meter is 2kΩ and later models 1.8kΩ and you don't have a 50uA range. So I'm not sure best how to improve the DCV cal.
I would consider also adding the two protection clamp-diodes that later models incorporated, across the meter movement. Schottky or Ge as full scale is 100mV.

https://www.simpson260.com/downloads/downloads.htm