Champ pulsing DC

[goldenGeek]

Hello
I am currently building a kit for someone, I think its a "STF supercharged champ". I have spent quite a few nights trying to figure out some initial problems but now I'm left with one weird case: the amp plays fine up until about 2-3 o clock, where it starts to act like a tremolo. The sound is pulsating (probably at 50hz), and when I am checking the voltages all voltages pulsate. Has anybody experienced this before? The output from the 5y3gt is stable at about 290V (might be a bit low?) until the pulsating starts. then it varies up to about 380V. The plate voltage, the other B+ voltages and cathode bias also varies (of course). Do you think this is the rectifier or power transformer?

[strelok]

If it's pulsating and sounds like a tremelo effect its too slow of an oscillation to be 50hz, which sounds like a hum. First thought is it's an oscillation of some sort, though those are usually more in the high frequency range. I would first try double checking the wiring then poking and moving wires around a bit with a chop stick while the amp is doing its pulsating thing to see if it has any effect. Pictures and a schematic will go a long way to help solve the problem if you're able.

[goldenGeek]

Ive been poking and chopsticking quite a bit, but the varying voltages is something i've never experienced before. I counted the pulses, and its about 50/minute. I can take a gut-shot later, but its all messed up now because i've soldered out the master volume and tone circuit. The schematic (which I noted down from the layout I was handed over) is attached here.

[frankdrebin]

It could be the rectifier,I guess you don't have an oscilloscope, try substitution with a good one or a couple of 4007 diodes.

[goldenGeek]

I tried to replace the redtifier with 4007s, but the issue is still present. I also checked the heaters this time and that voltage is also pulsating (about 1V up/down). So, is my PT screwed... I guess so. The weird thing is that I can play it softly at full volume, but when I strum hard it gets kicked into that tremoloish state. I have to turn the volume down to make it stop... hmmm... some feedback-oscillation going on?? Or PT??

I actually have a scope, but I have never taken the time to learn how to use it... I guess this is a good time to pull it out and start poking around. Any good advice on where to start looking. Or should I try to get a new PT...?

[JazzGuitarGimp]

The described behavour doesn't sound like a PT issue. I'm not saying the PT isn't at fault. But it isn't the first thing I'd look at. Have you tried replacing all the tubes with known good ones?

[goldenGeek]

Yes, I've tried to replace the tubes with tubes I know work. I soldered the PT out and I'm gonna try it in another SE amp I have on the shelf. If it works in the other amp I'm thinking that it could maybe have something to do with the dual (parallel) output tubes..? Maybe I should put in a 470r cathode resistor instead of the 250r and pull one 6v6... I'm soon out of things to try, I have replaced filter caps and pots, double checked resistor values. Oh well, I'll try the PT first and post back.

[martin manning]

Try adding some screen stopper resistors. Perhaps you have an oscillation going on between the two tubes in the power stage. At 50Hz, it seems likely that it has something to do with the power supply filters.

[goldenGeek]

Try adding some screen stopper resistors. Perhaps you have an oscillation going on between the two tubes in the power stage.

What value would you recommend?

[martin manning]

Try something like the typical 470 ohms. Kind of a shot in the dark, as the frequency is so low, but you should have them there anyway.

[JazzGuitarGimp]

This strikes me as possibly inadequate power supply isolation between stages. I'd be inclined to measure the 10k and 22k power supply string resistors to make sure they are the marked value and within tolerance. Then I would measure the DC idle voltage accross the three power supply filter caps - use clip leads and clip on to each of the caps' lead wires (as opposed to just grounding the meters' negative lead to chassis): this will ensure both leads are soldered into the circuit. If you have a cap that measures 0V accross it, you've likely found the culprit.

If that doesn't get you there, you might try clipping in additional capacitance across the three filter caps, just to rule out the possibility of a bad cap.

[goldenGeek]

This strikes me as possibly inadequate power supply isolation between stages. I'd be inclined to measure the 10k and 22k power supply string resistors to make sure they are the marked value and within tolerance. Then I would measure the DC idle voltage accross the three power supply filter caps - use clip leads and clip on to each of the caps' lead wires (as opposed to just grounding the meters' negative lead to chassis): this will ensure both leads are soldered into the circuit. If you have a cap that measures 0V accross it, you've likely found the culprit.

If that doesn't get you there, you might try clipping in additional capacitance across the three filter caps, just to rule out the possibility of a bad cap.

I desoldered one leg of both resistors and measured about 9.9k on the 10k and 22.4k on the 22k. I also replaced all the filter caps (I didnt have extra 10uf, so I used 16uf for those positions), but of course I could be really be out of luck and have multiple bad filter caps. I will try to measure across the caps clipped to the cap leg. The last time I measured it was clipped to the ground bus right between the caps.

Edit: and by the way, thanks for the good advice, I'm certain the amp sooner or later will come alive.

[goldenGeek]

Ahh... I figured theres one component I havent replaced yet... the 22uf across the 250R on the cathode. Can that sucker cause trouble like this?

[goldenGeek]

Fortunately you were right The PT does work fine in the other amp. Then I guess it's back to the soldering board, I'll try the 470R stoppers, try with one output tube, and possibly replace any other component that I haven't replaced already. How about the output transformer... can that one cause trouble like this? Its a Hammond125CSE rated at 8W, running 2x6v6 in parallel... but it was supplied with the kit so it should be okay, right?

[Firestorm]

There are so many things that can go wrong in SE designs. There's no phase cancellation in the output, the OT has to handle the tube current without saturating, the preamps will hum since there are no bypass caps. In my parallel SE amps, I use two output transformers to handle the current (but mine are designed for 6L6s and EL34s, not just 6V6s). If your OT is robust enough (at least a Hammond 125DSE or a 125ESE), that's fine. You're getting 50 pulses per minute. That's a parasitic oscillation. Fender Tweeds "breathe," the cones move in and out less than once per second. Also parasitic. Always a lead dress problem. In a small chassis, it's hard to address.

Try these: put a bypass cap across the first stage cathode resistor. The gain will rise but hum in the cathode will shunt to ground. Give each 6V6 its own 470R bias resistor, bypassed with at least 25uF. Each 6V6 screen should have a 470 to 1K resistor. Take the screen supply and make a voltage divider to knock it down to 50-100 VDC and add that to the heater center tap. Put a cap on it if you want to stabilize the voltage, doesn't really matter. These fixes will eliminate most random hum, which is working to generate your 50/minute oscillation.

And fix the lead dress. Count the signals. Input is plus 1, after 1 gain stage minus 1 (out of phase). Never let plus 1s get close to plus ones, never let minus 1s get close to minus 1s. In a small chassis, you might have to shield a wire.