AA864 Vibrolux Reverb noise issues

[argonaut]

I've been pulling out my hair trying to solve a noise issue with a '65 Vibrolux. I was hoping I could get some assistance. The amp making an undisireable noise when its volume is turned above 3 or so when playing through either channel. It sounds like a low 'farty' distorted noise riding atop the notes. It's most noiticeable when the amp is up in volume and when playing between the F and A notes on the low E string.

I suspected a problem with the PI as the noise is present in both channels. I quickly confirmed this with an audio probe. As both channels have a nice clean signal until it reaches the PI grid that connects to the NFB side. There it becomes all squashed and nasty. I did find a few out of spec resistors and replaced them and also tested all coupling caps. Not to mention trying all sorts of tubes, cleaning sockets, checking grounds, and the usual. I replaced all the electrlytics 5 years ago. So, I'm confessing my ignorance........I'm stumped. What else should I look for?

Here's some additional info that relates to the PI:

All measurements were taken with a VTVM

323V B+ feeding PI (12AT7)

1- 190V
2- 68V
3- 72V

6- 167V
7- 69V
8- 72V

I ran a 1K test signal into the amp and took a pic of the PI grids. The pic labeled PI-neg is pin7 (see attached). Also attached is an AA964 schem that, at a quick glance, looks just like my 864. Thanks in advance for any help!

[Jana]

Are you sure the problem is in the PI? There seems to be a lot of crossover distortion in the lower scope trace. Have you tried different power tubes? What does the trace look like if you measure the signal at the junction of the 100 ohm, .1 cap. I guess what I am getting at is, is that half of the AT7 generating this distortion or is it merely reproducing it.

[argonaut]

Hi Jana,

No, <haha> I'm not sure it's the PI. I've tried new power and pre tubes with the same results.

The trace looks exactly the same at the junction of the 100 ohm and .1 cap as it does on pin 7.

Thanks for asking.

[Jana]

The trace looks exactly the same at the junction of the 100 ohm and .1 cap as it does on pin 7.

This seems to indicate that the PI is reproducing that distorted waveform rather than creating it.

Can you disconnect the NFB loop and check the output of the PI again?

Also, check the waveform at each power tube plate, what do you see?

What does the waveform at the speaker output look like?

The lower waveform you posted looks like power tube clipping with crossover distortion. Have you checked the bias of the power tubes? (I'm sure you have, but had to ask).

[Jana]

Also, what are the waveforms like at the plates of the PI (pins 1 and 6)?

[argonaut]

Attached are the traces for the PI plates, speaker output, and power tube plates.

When I disconnect the NFB loop, I no longer see a signal on Pin 7 of the PI; however, there reads a clean waveform on both plates.

Bias is set correctly.

Thanks!

[Jana]

okay, those waveforms are looking normal. The noise on some of the waveforms is, I presume, noise being picked up from the probe being on those plates.

Reduce your signal level slightly so that you get a clean waveform on the plates of the PI, send a signal of a level so that the PI is just on the verge of clipping, but still clean. Then, what do you see on the speaker output? How much power is the amp putting out at that level?

Yes, you would no longer see a signal at pin 7 with the NFB disconnected.

[Jana]

have you tried hooking the amp up to an external speaker cabinet and having the chassis isolated and far enough away from the speaker so that it isn't subjected to any of the vibrations? Does it still do it? I am starting to wonder if there is a cold joint or a part with a broken internal connection.

If you do a play test as described above and the amp is okay, then I would put it back on the bench, connect the signal and scope and see if you can get the output waveform to react to poking around with a chopstick, rapping the chassis with a wood block, etc.

[rfgordon]

Traditional Fender layout, with long grid wires running from the tone capacitors to the control pots can result in a parasitic oscillation that sounds exactly like what you describe. I've fixed a couple of SF and BF amps by moving the tone caps/resistor from the board to the backs of the pots.

[argonaut]

Jana- The amp acts the same through an external speaker and the noise can be heard with an audio probe. I'll inject a clean signal into the PI input just before clipping and will report back when I get the chance. Thanks again for the suggestions.

Rich- Thanks for the input! I've been wondering if that's what I've been chasing. I'll definitely give that a try.

[argonaut]

double

[argonaut]

triple

[rfgordon]

Injecting signal is a really good way of figuring out exactly where the nasty little noise critter lives--whatever the cause.

Sometimes, though, when injecting a signal into later stages, like the PI, you need to really jack it up. I learned that this weekend. I was finishing up a build for a client--He had brought me the cab and semi-ravaged chassis of a 1961 tweed cab Maestro mystery amp that "used to kinda work" when he was a teenager. After enlarging the speaker hole to go from 8" to 10" speaker (BTW, the aged cone Weber 10 is MAGNIFICENT!!), the layout was pretty odd, cuz he really wanted to keep the bell cover on the speaker. Anyway, I was worried about lead dress issues, and the amp had an icky distorted thing riding on the attack and fizzling out. Sounded JUST like the parasitics in BF Fenders!

So I traced and I chased, and finally pulled the preamp tube (it only had one) and slammed a huge signal into the 6V6 grid. Lo and behold, there was the noise. It was a bad-sounding tube. But I had to pump a big signal in there to hear it.

[argonaut]

I've been meaning to report back for awhile but got totally side-tracked with the holidays. I moved the tone caps/resistor to the back of the pots and straightened up the lead dress and it cleaned it right up. Game on!

Thanks again Rich and Jana for taking the time!