Is the Selenium Rectifier bad?

Hi Tom,

One pot is used for DC measurements while the other is for AC. Both are maxed out.

I put the old 12AU7 tube back in because the new one from China did not show any improvement.

You are right that the filament voltage is 15% low: 5.4V instead of 6.3V. Unfortunately I don't have a way to fix that.

The resistor in the DC probe is 1.23M ohms (vs. 1 M ohm) and the 4M ohm resistor in line with the DCV input is actually 4.2M ohms. The resistors that are below the above-mentioned two are actually pretty close to the specified values.

 

No, there are no extra terminals on the transformer.

Note that the cathode resistors are supposed to be 80K ohms each, according to the schematic. However, I measured 104K ohms from pin 3 of the 12AU7 to the negative side of the power supply electrolytic capacitor. The reading from pin 6 to the negative side of the electrolytic capacitor was 104.4K ohms.

As the cathode resistance value increases, would that provide greater or lesser current flow to the meter (and thus greater or lesser meter deflection), for a given voltage being measured?

I found the following webpage that has a simplified schematic of a "typical" VTVM. Maybe this would help our thinking about the balanced bridge circuit.
Tales From The Tone Lounge; The Incomplete Idiot's Guide To VTVM'S!

 

Final update: I found a very good publication about VTVM operation:
http://www.cfp-radio.com/documentations/Sylvania-VTVM.pdf

Page 10 of that document discusses how the balanced triode push-pull VTVM circuit works. It makes the point that the cathode resistors are large so that they will dampen cathode current changes, swamping out small variations in plate current due to tubes and the power supply.

You may recall that I had determined that experimentally when I tried to increase the high voltage supply from 90 to 122V by putting a silicon diode in parallel with the selenium rectifier, yet there was hardly any change in meter behavior.

I thought more about the fact that the cathode resistors measured 104K ohms and decided to look for them. They were marked 100K ohms, so they were not too far off.

However since the schematic showed 80K ohms, I thought I should try lowering the resistances to see whether that would encourage higher tube gain in response to changes in grid voltage. I found a couple of 330K ohm resistors and put them in parallel with the two existing 104K ohm resistors. The net result was ~79K ohms.

What a difference! Now I have plenty of calibration range on both DC and AC voltages, and was able to set both calibration pots according to my reference voltages.

The bad news is that a piece of the mode wafer switch has just broken. When I looked at it I realized that it had been broken previously. I assume my dad used clear cement to repair it. I will look for a suitable epoxy or similar adhesive tomorrow.