Zenith ground loop problem?

7/7/2009 9:28:47 AMJoe S.

Got a curious problem I've never seen before. Recapped a Zenith 4G903 that I found. It plays OK on the isolation transformer but in wall outlet, no reception. Background hum seems normal. I tried switching the polarity of the AC, but that didn't help. Replaced selenium rectifier with diode, but that didn't make a difference. Caps connecting common connection to chassis replaced and double checked connection and wiring. When I unplug it with the power on, the signal comes through for a fraction of a second but dies as the B+ falls. I'm thinking I've got some kind of ground loop thing going on here, but not really sure. Any hints on how to troubleshoot this one?

Thanks

7/7/2009 12:51:54 PMLewis L

:Got a curious problem I've never seen before. Recapped a Zenith 4G903 that I found. It plays OK on the isolation transformer but in wall outlet, no reception. Background hum seems normal. I tried switching the polarity of the AC, but that didn't help. Replaced selenium rectifier with diode, but that didn't make a difference. Caps connecting common connection to chassis replaced and double checked connection and wiring. When I unplug it with the power on, the signal comes through for a fraction of a second but dies as the B+ falls. I'm thinking I've got some kind of ground loop thing going on here, but not really sure. Any hints on how to troubleshoot this one?
:
:Thanks

Geez Louise man, this is s weirdo. How 'bout taking Voltage readings of the A and B Voltages in both conditions, and lessee where that will lead us?

Lewis

7/7/2009 1:55:19 PMWarren

When using the transformer, I don't think you have a full 120VAC going to the radio. When plugged direct you do. It sounds like the B+ is to high. Check the B+ readings both ways, and any bias readings too. Maybe a dropping resistor value problem. Did you use a resistor after the diode?

7/7/2009 6:36:21 PMEdd

Sir Joe . . . .S

I also tend to concur on the excess voltage aspect. SINCE the set is using coated, fine- filament, battery type powered tubes, so things happen in a hurry.

If using a conventional heater- cathode tube lineup, the hot- cold. . . .emission - no emission . . . inertia would be greatly
p r o l o n g e d out with that type of tube. With YOUR sets tube types, that burst of sound cycle is probably ocurring within just a few seconds.

So, I can see the unplugging of the set and the diminishment of the DC supply voltages, with that critical voltage threshold being approached and the then presence of sound . . . ..but .. BUT.. . the filaments have already chilled down, so there was but a very- very short timespan of any sound coming forth.

You did not specify if the isolation transformer was also incorporating a Variac feature, or even opting for taps to adjust the
AC output voltage. . .like the RCA Vari-Tap series does. . . .assuredly not.. . .or you would have tried reducton.

If that was my set to analyze, I would be hooking it up to the isolation transformer where it was working and then consulting
the schemtic , we now see the specified benchmark for the power supply written just above first electrolytic C13 as being 100 VDC.

Log that down and then move on up to the plates and screens of all of the tubes and log in their voltages.
Those being in the 87 and 92 vdc range at the 3V4 output, 18vdc for the 1S5, 92vdc for the 1U4 and 92 and 66vdc for the 1R5.

Now, when you power up the set from the full AC LINE source , somehow, I am expecting the 100 vdc test point to be up quite a bit, in fact so much that it resilts in the shifting of the biasing paramenter of tubes such that they're driven into heavier conduction and a shifting of those plate voltages and a cessaton of amplification.

Now, when you unplugged the set, there was a DC power reserve stored in the electrolytics , but no incoming power refresh, so the DC level started being pulled down. . . .until drain resulted in voltage decline. . . . UNTIL that magic threshold was hit where set amplification initiated again, but, by that time the filaments were cold.

Considering that you replaced that selenium rectifier with a modern !N4007 family, you went from an ~ decade of voltage drop in passing thru a selenium rectifier to a mere .7 volt in its passage thru a silicon diode.

That R8 140 ohm 3 watt surge / dropping resistor, needs to be inceased in value, in order to bring the B+ supply level back down to that specified 100 VDC at its test point.

Wouldn't hurt to also check each filament voltage to also see how close they are to spec and them also not being oversupplied with voltage.. . .with like due consideration of potential filament longevity, as well as not excessively depleting the filament coating emissivity chemicals.

73's de Edd

7/8/2009 8:09:15 AMJoe s.

:When using the transformer, I don't think you have a full 120VAC going to the radio. When plugged direct you do. It sounds like the B+ is to high. Check the B+ readings both ways, and any bias readings too. Maybe a dropping resistor value problem. Did you use a resistor after the diode?

I checked the B+ at the first filter cap - 100 volts. I had a Variac running from my isolation transformer set to 120v. Measured B+ at this point both - transformer and without - same voltage. Neglected to go further down the line as my focus was mostly on looking at the inout - which probably was a mistake, but tunnel vision sometimes takes over. I only replaced the selenium on the outside possibility that was the problem. Had same symptoms either way. Yes, I used a resistor after the 1N4004 I tried as a replacement. I'll rewire the selenium back in since it doesn't seem to be the problem, and re-examine the other biasings. Yes, when I unplug the radio I only get a short burst of signal before it naturally dies out. I'll try lowering the voltage with the Variac on wall current to see if that has any effect. If so, then sounds like voltage too high somewhere along the line.

Many thanks to all.

7/8/2009 1:21:03 PMJoe S.

Thanks again to all. You sent me on what looks to be the correct path. Went home at lunch time and checked a few voltages with my digital meter this time (as opposed to my old Simpson). Turns out my Variac at the 120v setting actually puts out 125v. AC line voltage measured 120v. Reduced Variac to actual 120v output, and set stopped playing. Checked B+ at first filter cap, and it was low at 88v. Brought Variac up to where the set began to play, and B+ reads about 94v. Looks like I need to get the B+ correct at proper line voltage, and I should be good to go.

::When using the transformer, I don't think you have a full 120VAC going to the radio. When plugged direct you do. It sounds like the B+ is to high. Check the B+ readings both ways, and any bias readings too. Maybe a dropping resistor value problem. Did you use a resistor after the diode?
:
:I checked the B+ at the first filter cap - 100 volts. I had a Variac running from my isolation transformer set to 120v. Measured B+ at this point both - transformer and without - same voltage. Neglected to go further down the line as my focus was mostly on looking at the inout - which probably was a mistake, but tunnel vision sometimes takes over. I only replaced the selenium on the outside possibility that was the problem. Had same symptoms either way. Yes, I used a resistor after the 1N4004 I tried as a replacement. I'll rewire the selenium back in since it doesn't seem to be the problem, and re-examine the other biasings. Yes, when I unplug the radio I only get a short burst of signal before it naturally dies out. I'll try lowering the voltage with the Variac on wall current to see if that has any effect. If so, then sounds like voltage too high somewhere along the line.
:
:Many thanks to all.

7/8/2009 1:54:30 PMEdd

Sir Joe. . .S

If in the testing sequences, and the power supply rectifier, at a time, was using a 1N4007 series of silicon diode, and that output DC voltage was actually as low as specified; give serious consideration of the actual "health" of that can electrolytic filter
capacitor.

Especially its 60 ufd input section.

An immediate analysis could be had by shunting that 60 ufd section with a loose, like value, single tubular section stock capacitor, for an evaluation.

73's de Edd