Amp issue - PI distortion and low volume - request for help

[mlp-mx6]

Hello friends,

I have a small (built in a Vibro Champ chassis) amp I built for a friend that now has an issue I haven't successfully diagnosed. I'm hoping you can help. My friend said it suddenly lost volume and he "smelled something" so he shut it down immediately.

The topology is: gain stage; volume; gain stage; tone stack; PI, power. (2 12AX7s and 2 6V6) Solid state rectifier, runs ~420V on 6V6 plates.

When I input a sine wave and check the signal on a scope, input to V1a is good, output of V1a is good, input of V1b is good, output of V1b is good. Input to PI is good. Output of either side of the PI starts to distort at volume=3 and just gets worse from there. No headroom, *terrible* tone.

All voltages seem fine, both plate and cathode. I've subbed all the tubes with no change. Also subbed a new power supply cap for the PI. Nothing looks wrong except the signal post-PI.

I can show pics this weekend. In the meantime, any ideas?

Thanks,
Michael

[pdf64]

Was it being cranked when it lost volume?

[mlp-mx6]

Nope. It was not cranked. But the owner did say he "smelled something" and then turned it off immediately. I suspect the PT, but voltages still look okay.

[xtian]

I'd guess the power tube failed, burning something in the process. Screen resistor?

[pdf64]

Is there a global feedback loop?
Scoping signals within a loop (eg phase splitter plates) can give strange results, due to the NFB error correction mechanism.
It can be best to open the loop when fault finding with a scope.

[Stevem]

How much does that 420 volts of V+ pull down when the amp starts to drop volume?

[mlp-mx6]

Thanks, folks. I'll be able to investigate and take pictures this weekend, perhaps this evening. I'll get back to you.

[mlp-mx6]

One side of the output xfmr measures 88 ohms, the other side measures 72 ohms. Voltages aren't very off, but this just looked weird.

Powered up -
Brown wire/power tube - Blue wire power tube
No signal - both at 437V
@3 on vol - both at 383V
@10 - brown=362V, blue=360V

Screen resistors are fine. Power tubes are new. Voltages are steady.

I can open the fb loop because I installed a switch to do just that. Pics that follow are open-loop. Volume setting is the pic name. Starts to square on 3, worse on 4, BAD worse with ringing on 10. These are all at the higher voltage PI plate.

New PT is on order. (One of the mounting screws was stripped. No idea how that happened. It was delivered to me that way.) Do I need a new OT also? Open to all suggestions. Again, thanks. Also, if I can provide any other pics/info, just let me know.

[Stevem]

You can run it open loop but your gonna have at least 6 db more signal than you need feeding the PI and that will carry over to too much feeding the output tubes to have much clean volume range.

Your big drop off in power supply voltage with the amp on only 3 is a sign that you have too much drive signal hitting the output tubes,I would try out a at7 or ay7 in the Pi socket.

Double check the resistor values in the PI section.
Can you post a schematic ?

[mlp-mx6]

Here's the schem

[mlp-mx6]

The output xfmr was blown. I subbed in a known good one and all was well. Just installed the real new one tonight. Rockin' once again.

Just FYI - the sound of a blown output xfmr is horrible.

[Firestorm]

IMHO, transformers should always be overspecced. Very little sonic goodness comes from pulling too much current through an OT and its an expensive "fuse." PTs, too. If you want to simulate poor regulation, a power resistor works fine.

[mlp-mx6]

IMHO, transformers should always be overspecced. Very little sonic goodness comes from pulling too much current through an OT and its an expensive "fuse." PTs, too. If you want to simulate poor regulation, a power resistor works fine.

The original was a Hammond 1608. Definitely not underspec'd.

[Firestorm]

Probably infant mortality. Happens in the best of families. If nothing untoward happened, it should be on them.