Is the Selenium Rectifier bad?

This question is for electronics old-timers.

My dad recently gave me his Simpson 303 VTVM. (Here's a link that provides photos and the owner's manual including schematic diagram.)
Simpson 303 Vacuum Tube Volt Meter

When I received it, I tried calibrating it using my Fluke 87 multimeter as the reference. I used lab power supplies as the DC voltage source, and was able to calibrate DC voltage measurements fairly well.

However, when I tried measuring AC house voltage, the reading was way off. 120VAC measured as ~80VAC or so.

Should I expect that a VTVM which is powered by house AC will be able to measure house AC voltage accurately, or is there some ground loop issue? I had hoped that the power transformer would isolate the VTVM circuitry from the AC line.

I measured power supply voltages. (See the schematic diagram per the above link.)

The voltage across the transformer primary is 118.3 VAC, which is house voltage.

The filament voltage is 5.4 VAC. (Is that a problem?)

The transformer high voltage secondary voltage is 120 VAC. The voltage at the selenium rectifier anode is 119.1 VAC. Since there is 0.9V drop across the 33 ohm resistor, it appears that 27 mA current is flowing.

The voltage at the selenium rectifier cathode is 90.5 VDC and 1.04 VAC ripple as measured by the Fluke.

I tried replacing the 20 mFd capacitor with a 22 mFd, 160V capacitor purchased from Mouser. However, there was no increase to the DC voltage (in fact the ripple voltage increased slightly.)

Do you think that the selenium rectifier is bad, since it shows a 30V drop? If so should I replace with a silicon diode or will that cause trouble with excessive voltage? (One problem is that the schematic does not show expected voltages.)

Thanks!

 

Hi Tom,

Well, that was my thought as well, which is why I posted the voltages that I measured. I'm not sure whether the plate voltage is good or bad since the schematic diagram doesn't show what the voltages are supposed to be.

 

Thanks for the comments. I ordered two 6AL5 and one 12AU7 (to have spares available, since shipping is $6.45 per order). The 6AL5 is $3.95 each and the 12AU7 is $7.95 (from China).

 

Dear friends and old-timer electronics troubleshooters:

Well, the good news is that I received my vacuum tube order (from Triode Electronics) today. The bad news is that after installing the RCA 6AL5, there was no change in behavior, where the AC voltmeter reading is ~80VAC when the correct value is ~120VAC house voltage.

Since the DC voltmeter and the ohmmeter seem to be operating OK, I am going to assume that the power supply (incl. the selenium rectifier) is OK.

I turned off the VTVM, opened it up, and started using my Fluke 87 ohmmeter function to probe around.

I measured resistance from pin 7 of the 6AL5 to ground. The measurement started at 0.9 megohm and ended up at 1.2 megohm.

Then I measured resistance across C2 which is 0.25 uF. (The way that I did this was to measure from pin 7 of the 6AL5 to the ACV and Ohms probe input while the selector switch was set to ACV.) The resistance was 32.4 ohms and it did not increase over time.

However, I then used the capacitance measurement function of the Fluke and measured 0.26 uF. If the capacitor is OK, why does it show this steady-state resistance of 32 ohms??

This caused me to think that maybe C1, 0.1 uF is bad since that filters the 6AL5 output before it goes to R1, 500 Kohm, the range-switching resistor network, then finally to R19, the 1 Mohm resistor in the path to pin 7, grid of the 12AU7. If C1 is shorted then the AC voltage readings would be depressed.

I measured resistance from pin 5 of the 6AL5 to ground and got 1.94 megohm. I also measured 0.22 uF capacitance from that pin to ground. How would the resistors in series with the capacitor affect the capacitance measurement? (Having trouble finding the side of C1 which is not grounded.)

Not sure whether C1 is OK, but at least it is not shorted.

The 6AL5 remains in its socket while these measurements are being taken.

Anyway, I would appreciate your comments and suggestions (pls refer to the schematic in the link provided in post #1, if you wish to follow along.)

Thanks!

 

Yes, I've turned pot C fully clockwise to show maximum meter deflection, but that is insufficient to reach the desired voltage reading. I measured the total resistance across pot C which was 1,085 ohms. Then I measured the resistance across the wiper terminal and the end terminal that is wired into the circuit, and the result was 8.9 ohms.

I had to set pot A full clockwise to reach my desired DC voltage meter reading.

I also temporarily installed a 1KV, 1A silicon diode in parallel with the selenium rectifier. Now the voltage is 122V measured across the 22 mFd electrolytic capacitor. However there's no difference in behavior, the AC voltage measurement of house voltage still reads 80V.

 

OK, I will work on replacing the capacitors which means yet another Mouser order, so you won't hear more from me on this for several days.

I tried replacing the 12AU7 with the China-made tube, no change in meter behavior (in fact the meter deflection is a bit less. The physical size of the plate structure in the China tube is noticeably smaller than the size within the old US tube.)

I also think that part of the problem is that the DC voltage calibration potentiometer has to be set full clockwise, for minimum resistance, in order to get the meter reading to be correct. I will also work on trying to figure out why that is the case. If the DC voltage calibration pot can be set mid-range, then there's more likelihood that the AC voltage calibration pot also will be within range.

What doesn't help is that the schematic diagram as posted seems to be missing a bit in the middle of the diagram, between the two pages. It's also not clear whether the schematic is accurate, with respect to my particular meter. More on that later.

I do have an old HP two-channel analog oscilloscope, 275 MHz bandwidth. I need to find a low-voltage AC source that I can hook the VTVM AC input to, for extended testing. I'm a bit nervous about using 120VAC as my voltage source to be measured, for more than a brief period of time. Maybe I can use the VTVM filament voltage as a source to be measured.

 

Many thanks to the group for your interest and support during this project.

A further update. The new capacitors arrived yesterday and are much smaller than the originals. (Isn't it great that Mouser doesn't have a minimum order $ value?)

I installed them and the good news is that the AC voltage measurement is now better. The house voltage is 118.9V and the measurement is now 105V. The spec is 5% of full-scale. Since I am using the 300V scale the allowable error is 15V.

The DC voltage is also slightly low. For example a 9V battery which produces 9.5V is measured at 8.8V or a 7% error. The spec is 3% of the 12V full-scale which is 0.4V.

The schematic diagram is not accurate. A couple of examples that I noticed:

1) The 0.1 uF capacitor is actually connected from the wiper arm of pot B to ground.

2) Potentiometers B and D are supposed to be 500 ohms each according to the schematic. However, the resistance of pot B from one end to the other measures 144K ohms while pot D measures 880 ohms from one end to the other. Note that according to the schematic, pot B and D are connected in parallel so the measured resistances should be the same from one end to the other.

Since I don't have an accurate schematic and the wiring harness is complex enough so that it won't be easy to figure out what is connected where, I plan to declare victory over the project, recognizing that the meter reads slightly low.

I attached a few photos. The new 0.01 uF capacitor is installed on the left side of the first photo because the old one was installed between the two rotary switches and was difficult to access.

The second photo shows the new 0.1 and 0.22 uF capacitors, see the white tubular parts. Also see the tan capacitor between the two rotary switches which is the old 0.01 uF capacitor with the top connection snipped away and the capacitor moved aside to the left.

The third photo shows the new 20 uF power supply capacitor which is much smaller than the original which filled the space between the power transformer and the side of the case.

The fourth photo shows the meter measuring the 9V battery (which produces 9.5V) at 8.8V.

Hi Tom,

The VTVM has a two-pin AC plug (back then, the AC outlets did not have a ground pin) so I believe that ground loops would not be an issue with the power transformer isolating the circuitry from the AC house wiring. In any event I did not use my oscilloscope since the meter now works "good enough".