Need help troubleshooting an Ampeg VT-22

[SILVERGUN]

Amp came in with the screen supply resistor (R55) making intermittent contact with the board so when the kid would bang on top of the amp it would make intermittent contact and he would get some output.
I found that resistor had melted the solder away at the board connection and had some heat damage so I replaced it and soldered in a new resistor and I'm now witnessing this problem:
- At idle, all appears well, with 575V plates and 572V screens  , each 6L6GC is showing between 25-30mA plate current
- As I increase the input signal the plate voltage drops proportionately to the amount of signal that I feed it....falling from 575V down to 480V (ish) at max signal, and obviously dragging all other voltages with it.
- At this time the plate current rises, and eventually will cause a redplate condition if amp is continued to operate at high output.
- Also the screen current rises to a point where I'm dropping 30+ VDC across that 470R screen supply resistor,,,with one side of the push/pull pair appearing to draw more current than the other side.
- I'm watching the bias voltage drift more negative and goes from it's idle reading of -65VDC to a max reading of -90VDC on one side and -78VDC on the other side of the push-pull pair.
- Watching it on a scope with dual probes it becomes obvious that one side is taking the hit more than the other.
- I removed one power tube on each side, and swapped some known good tubes in to eliminate the tubes and sockets.
- I also cleaned and retensioned the sockets.
- I have checked all diodes, resistor values, caps, etc., etc....
- I took a break from "troubleshooting" to do a cap job, and replaced all main filter caps and it has not affected the situation.

What am I seeing and what other info do you guys need to help me figure it out?
I consider myself a decent troubleshooter and I'm very interested to find out what I'm missing, and learn something new.
So far I have learned the term "bias excursion"
Thanks, in advance for any help you may be able to give.

I have confirmed that the attached schematic is correct because it matches the one inside the amp.

[shooter]

After a lot of reading last year I turned down a request to resurrect one of those amps for someone.  The consensus is they are a B*&ch.

I removed one power tube on each side

did you replace pairs on each side with known goods also?
When you did caps I'm guessing you did bias supplies also?

looks to me like something is wanting to much current, that could be any-all the tubes, but could also be a partially shorted OT.

that's all from a lame armchair throubleshooter

[Willabe]

I would say that you have added a lot of very good posts to threads here shooter.


             Brad     

[SILVERGUN]

After a lot of reading last year I turned down a request to resurrect one of those amps for someone.  The consensus is they are a B*&ch.

Yeah, over-engineered, cross wired, way too high voltage, PC board crap.....ehh

More info:
- I tried all good tubes in all spots. I eliminated one tube per side just to limit the variable.

- I performed the "super secret transformer test" on the OT and it came back OK according to that test. (it's where you flash the primary windings with low voltage DC and have a neon bulb attached to the secondary,,,the bulb flashed, which is supposed to indicate that the OT is OK. I had all leads from the OT disconnected when I performed the test and saw no shorts across any of the windings.

- I am obviously pushing grid current in those 6L6s, and forgot to measure that, but I'm assuming that is what is causing the bias excursion.

- I did not change any components in the bias circuit yet. (including caps, everything tests OK)

- Just to confirm what I was seeing, I removed all power tubes and scoped the AC output of the PI and it appears to be fairly even and balanced, and only mostly square, with no voltage drop as the signal increased.
- It's only with the power tubes in that one side gets grossly mis-shapen, as if the grid current is dragging it down,,,and the bias shift is happening on both sides, just more to one side.

[Willabe]

When you did caps I'm guessing you did bias supplies also?

If those bias caps are dying they mess with the bias.

Because the amp is old and because of the hi dcv stresses I'd change all the screen grid R's and all the grid stop R's for good measure too.

New JJ's are said to be able to handle those hi dcv's.


                     Brad   

[SILVERGUN]

Thank you sir!...that's on the agenda for this weekend.
 

[TubeGeek]

I just spent a bunch of time working on a VT-22 and V4b this past year.  The VT-22 just left my shop with a new PT.


Not easy amps to work on/figure out.  Watch out, the outer shell of the filter caps are HOT.  I found that out the hard way.


Now I have a B-15S in for overhaul.  Ampeg's are so much fun to work on


I don't know if I'd be much help with your troubleshooting, sorry.

[SILVERGUN]

Thanks for stopping by TG. What was the failure that led you to replace the PT?
I understand I'm on my own on this one. This WILL BE the last pc board amp I will ever open up, except to rip it out and start over.
It would be different if I was trying to make money doing this. Maybe I'll learn something more than 'don't work on pc board amps' 

Updates: (did all the things that an amateur would do while they were guessing what the problem was and trying to eliminate one component at a time)
- Replaced coupling caps between PI and power tubes
- Replaced all screen R's and raised values to 1.5K
- Replaced all grid stopper R's feeding 6L6s and raised values to 100K
- Replaced both bias caps
- **Momentarily disconnected OT wires from plates of output tubes and quickly took measurements and still saw voltage drop. In my mind that completely rules out the OT.....right?**How stupid was that?** use a 1-10 scale 

Problem still exists.

I'm seeing .750 DCV volts between the ground side of the bridge rect. and the cathodes of the power tubes at idle that raises to .875 under load.
I was seeing that same reading before I replaced the filter caps.

It seems to me that either the rectifier or PT are failing under load, but all diodes check good with no load....guess I'll just replace them and see what happens.
 

Nothing like spending hours on something that you cant fix to make you feel like an idiot.
Humbling for sure.

Every once in a while I take a break and jam a little to blow off some stress.

[shooter]

It's Easter, so get in the amp and Easter egg hunt

you should be able to see increased ripple on the B+ rail if the diodes are breaking down.  They do like to fail short 1st, I was kinda thinking the plate stack diodes as long as you're hunting Eggs with a shotgun.

[TubeGeek]

Thanks for stopping by TG. What was the failure that led you to replace the PT?
I understand I'm on my own on this one. This WILL BE the last pc board amp I will ever open up, except to rip it out and start over.
It would be different if I was trying to make money doing this. Maybe I'll learn something more than 'don't work on pc board amps' 

Updates: (did all the things that an amateur would do while they were guessing what the problem was and trying to eliminate one component at a time)
- Replaced coupling caps between PI and power tubes
- Replaced all screen R's and raised values to 1.5K
- Replaced all grid stopper R's feeding 6L6s and raised values to 100K
- Replaced both bias caps
- **Momentarily disconnected OT wires from plates of output tubes and quickly took measurements and still saw voltage drop. In my mind that completely rules out the OT.....right?**How stupid was that?** use a 1-10 scale 

Problem still exists.

I'm seeing .750 DCV volts between the ground side of the bridge rect. and the cathodes of the power tubes at idle that raises to .875 under load.
I was seeing that same reading before I replaced the filter caps.

It seems to me that either the rectifier or PT are failing under load, but all diodes check good with no load....guess I'll just replace them and see what happens.
 

Nothing like spending hours on something that you cant fix to make you feel like an idiot.
Humbling for sure.

Every once in a while I take a break and jam a little to blow off some stress.

The PT had an open in the filament winding.  Found a direct drop in replacement made by Heyboer.


I spent a lot of time on the one I had too.  Much more time spent than billed for.  Good thing this doesn't pay my bills.

[Jacobysmalls]

Hey Silvergun, did you ever figure this one out? I was very curious after reading the entire post how it ended. I'm about to tear into a VT-22 myself and am a little nervous ...

[SILVERGUN]

This was a while ago but I remember chalking it up to a bad board.
I started talking to the owner and he wasn't happy with the amp the way it was, so I pulled the board and made a custom board based on what he was looking for.

[punkykatt]

Its always nice to hear a happy ending

[SILVERGUN]

I couldn't find pictures from that repair but here is the scrap board that I messed up in the process.
This was the amp that prompted me to start buying G10 in sheets, so at least this one didn't hurt so bad.

That amp has a big hole cutout in the chassis where the tube sockets are mounted to the board so rather than try to plug the hole I just figured I would make a custom board size and mount the sockets right on the G10.