Unbalanced Bias Current Problem

[MikeSafreed]

No, I didn't change out the PT. I realize the frequency range is right to possibly contribute to the problem, but I have measured voltages under load and the PT seems to be holding up. However, like all of you, I've had stranger things happen, so, I'll change it out - I should have one in the shop somewhere. I'll try it later this afternoon.

I also need to see what happens this afternoon with the NFB disconnected and the PI feeds to the power section reversed. I'll post the results of both later tonight.

And, thanks to everyone for their suggestions. Like the old joke says, we'll find it in the last place we look!

[MikeSafreed]

Firestorm, I disconnected the NFB and it had no impact. It didn't make the symptoms worse, nor did they get better. Same amount of distortion... I switched the PI outputs and the bias symptoms moved to the other power tube - now the 6L6 closest to the PI tube is changing bias wildly - pulling up to 55 MA more than the 6L6 closest to the rectifier. All this proves to me is that the issue follows the 2nd half of the PI (pins 6, 7, . There are all new components on that half, as well as the first halve. Do you have any ideas for followup to this test?

I searched, but I do not have a replacement PT. I ordered one just now and should have it in a day or two. I need it anyway for a new build for another customer, so it's not wasted money.

New scope on the way...

[eddie25]

The phase inverter tube has been replaced a couple times right?

[Firestorm]

Okay, so it's not NFB related and the problem stays with grounded grid side of the PI. That's the side where you were seeing a 46VAC signal compared to a 36VAC signal from the inverting side.

Since you ruled out tubes, caps, resistors and the OT by replacing them, the logical candidate remaining is that one side of the PI is actually getting a bigger signal than the other, amplifying it and overdriving its respective power tube.

All of which implies phantom coupling of an in-phase signal to the grid on Pin 7. I think I'd partially lift the circuit board enough to kind of get a look get a look under it; I've seen a lot of stranded wire clippings floating around in SF Fenders. I might also suggest temporarily assembling some of the PI components up in the air, off the board, in case some part of the board is conducting. In particular, I'd lift the ground end of the .1 cap on Pin 7 and ground it somewhere else (like an input jack) to make sure it's not picking up a stray signal. It is sitting 100R above ground where it is.

Aside from that, I think you've got to wait for the new scope. (Which one did you pick, BTW?)

[alvarezh]

Mike, Firestorm's suggestions are excellent. Now that you know that the problem has changed sides, I would wait changing the P.T. as a last resort.

A strong signal coming in a stray manner is definitely a logical happening. It has happened to me if I let the plate wires from the same preamp tube come close together, they start to oscillate at supersonic frequencies and make one or a few power tubes red plate. You have not seen the power tube red plating probably because you are just playing a guitar through it, if you maintain a constant signal long enough with a tone generator, the affected tube will most likely try to fry itself.

Best wishes.

[MikeSafreed]

Guys, I have made some positive progress. It's not 100% better - maybe 90% - but the problem has been significantly reduced. I switched the 220K power tube bias supply resistors and the PI plate resistors from the top of the board to the bottom, and vice-versa. This way, I was able to shorten the power tube grid wires quite a bit. It lengthened the PI plate wires. The distortion has been reduced - still present but not nearly as bad. This kind of proves that part of the issue IS some type of unwanted coupling. Later today I will take the earlier advice about moving the ground for the .1uf cap that connects to pin 7 of the PI and see if that also improves the situation. Regardless, I'm going to need to trace out the entire amp and find where the source of the remaining distortion resides. What I did helped the situation, but it's not the total solution. Something else has to be wrong - a stable amp doesn't go unstable unless something changes. It could be the board itself changing with age, it could be corrosion on that stupid brass ground strip, or a failed component somewhere else. The new scope ships today, from New Jersey to Ohio, so I should have it on Friday. As was suggested, I got the Rigol DS1052E. I know it's not a high-end scope, but it will do until i can afford to toss another couple of grand at a really good scope to replace my old HP scope (it was old, but man, was it cool! - I worked at HP for 40 years and one of my HP tech buddies hooked me up with the scope when HP was replacing older scopes with newer models). It did things that newer scope makers haven't even heard about doing...

To answer an earlier question, yes, I've changed the PI tube SEVERAL times. Always to a known-good tube, and it's never made any difference.

Even with the improvement in the situation, I'm still leaning toward a rebuild with a new board and a new grounding buss - I like to use a 12 gauge copper wire soldered to the back of the pots and chassis ground it near the input jacks. I need to get that old brass grounding plate out of the amp so I can check the grounds on the input jacks, and to do that I will have to do a lot of unsoldering. After all that work, it makes sense to me to go ahead and do the rebuild. The customer seems to be leaning that way now and I believe we've settled on a price.

So, even with the progress, I'm not totally satisfied, and I think a rebuild in the safest, long-term solution.

I will post again once the scope gets here and I've completed some traces...

[Firestorm]

Later today I will take the earlier advice about moving the ground for the .1uf cap that connects to pin 7 of the PI and see if that also improves the situation.

I forgot about the other PI "input." You might also try to get PI tail resistor and the PI bias resistor off the board temporarily. If a signal that is in-phase with the input on pin 2 gets into the tail or the bias resistor, it adds to the non-inverted signal and subtracts from the inverted signal, which is what you were seeing. The phantom signal would have to be pretty large, which is why I was thinking of stray wire clippings under the board.

Good luck with the new scope.

[MikeSafreed]

Firestorm, I will check out the PI tail and bias resistors later today and see if lifting them off the board and isolating them helps. I did lift the entire board and check for anything weird beneath it, but I didn't find any wire clippings or any obvious problems. There were a couple of solder blobs from previous repairs and I removed them. I saw nothing that should have contributed to the symptoms... I'm leaning toward the coupling occuring because of faulty grounds all along that brass strip. I guess I could go ahead and leave the strip in place and solder in a new grounding wire to the back of the pots, regound the input jacks to that new buss, run some new ground wires to the buss from all the bypass caps, and see what happens... That way I don't have to rewire all the pots. Just a thought...

Phil, the scope should be here later today or tomorrow - hope the learning curve is short. I've been reading about the Rigol scope online and there have been some software issues that require downloading some updates, but everyone seems to get them up and running without too much trouble.

[Phil_S]

Phil, the scope should be here later today or tomorrow - hope the learning curve is short. I've been reading about the Rigol scope online and there have been some software issues that require downloading some updates, but everyone seems to get them up and running without too much trouble.

Hey Mike! I hope I was clear, that was simply the cheapest one I found at that place. It looks way more than adequate for what we do with guitar amps, but I have no experience with it. I'm having some difficulty imagining software issues are a problem for guitar amp work, but who knows? In any case, you didn't spend much for it, and it comes with a warranty. I hope it works out for you.

Unlike you, I hold simple amateur status. My scope is a BK 1472C, 15MHz 2 channel. Almost as old as I am, LOL. I don't use it much, but it sure is a wonderful tool to have in the tool box.

Anyway the reason I suggested it, is that you have been burning hours you can't recoup for lack of the tool and the accountant in me says it is penny wise and pound foolish to be without it. I am reasonably sure it will move you past this in the flash of an eye. Your swift action suggests you agree. FWIW, I think you did the right thing. Good luck with it.

Phil

[MikeSafreed]

Phil, I certainly wasn't complaining about your suggestion for a scope. It was much appreciated! Heck, this new scope may turn out to be more than I'll ever really need. The HP scope was actually overkill. It had features that I never used, it was VERY LARGE and took up a lot of real estate, and you're right about what we need for guitar amps. We don't need too much scope capability. I'm kind of an old-school guy - I can do the vast majority of my troubleshooting with my eyes, ears, and even nose. Plus, after a while, you get to the point where you've heard just about every problem and you kind of know without even looking at the inside of an amp what needs to be done. Of course, once in a while, you get one like this Vibrolux Reverb that drives you crazy and you need all the suggestions and technology you can get your hands on to help with a solution. Thanks, again, for the pointer to the scope...

[MikeSafreed]

Guys, the Vibrolux Reverb amp has been fixed. Really just one thing cured the problem, and that was what I posted previously - I shortened the grid wires to the power tubes by moving the resistors in the PI. Even with the scope, I could never determine what, if anything, had changed in the amp to cause the parasitic to start out of nowhere. I'm guessing that it was simply a symptom of aging parts - the circuit board, a potential breakdown in the effectiveness of the insulating board between the circuit board and the chassis, and the grounds becoming a little less effective.

On the scope, the signal looks good with the shorter grid wires to the power tubes. If I switch the PI resistors back to their original locations, and lengthen the grid wires, the symptoms return and the waveforms on both of the grid wires go crazy. But, at the same time, the waveforms prior to the grid wire connections always look correct, no matter the length of the grid wires. So, something is coupling with the grid wires, but whatever it is won't reveal itself. I can fix the symptoms but not the cause. At this point, that's good enough for me.

The customer has his amp back, he played it out last night, and I'm moving on to something on which I might actually make a buck. Thanks to everyone for their help! And, the new scope actually works great... Thanks for the URL and the promting to get off my rear and get a new scope, Phil.

[martin manning]

Hallelujah! Which resistors exactly did you move? And, which power tube was showing the high current draw?

[alvarezh]

I am really glad for you Mike, it definitely was a tough one, thanks for sharing the experience, very helpful indeed.

Cheers!

[MikeSafreed]

Martin, I moved the 220K resistors in the PI to the bottom of the board, and the 100K/82K PI plate resistors at the bottom of the board to the top of the board, and their corresponding wires. I lengthened the plate wires for the PI so they would reach the top of the board, and then I shortened the grid wires for the power tubes. I had to move the bias connection and the DC power connection for the PI to the new locations of the resistors. Overall, I was able to remove about 3" of wire for each power tube grid. This made all the difference in the world. The stupid thing is that I've known about this simple mod for a zillion years. In fact, I've built a lot of new amps based on the AB763 circuit using this PI resistor configuration right from the start, just to keep the power tube grid wires short. But, I got fooled by the fact that this amp was very stable to start with. This type of issue usually doesn't just start showing up overnight. But, it did. Glad it's gone for now...

[martin manning]

MIke, ok I see... You just flipped that section of the layout front-to-back, meaning with respect to the front and rear of the chassis. It is indeed hard to imagine how this instability could appear out of nowhere when it was functioning without it previously. Thanks for taking the time to document the solution.