TUT 3 Plexi Build Parasitic Oscillation Problems - Strange one!

[priesty]

I've been building a Plexi copy a-la Kevin O'Connor's TUT3 Plexi. I have made it as close as possible to his suggested layout in the book (photo attached).

The main differences from TUT3 Plexi is that I have made it a 4 x 6V6 output section, similar to the TUR3 Laney project.

The amp has a strange parasitic oscillation. The amp is hooked up to a 8ohm dummy load, but thanks to the vu Meter I fitted, I know this amp is oscillating. I did the following things in this order and nothing changed:
1 - Swapped OT output leads - No Change
2 - Disconnected negative feedback loop - made no difference
2 - Added 150pF capacitor between plates of PI tube (there was no cap in the design) - No Change
3 - Added 680pF Grid-to-Ground caps on all output valves - No Change
4 - last ditch effort, removed 150pF from PI anodes and replaced with 1000pF - No Change other than the side of the OT with most oscillation seems to swap.

None of the above worked, but one interesting thing does happen. The control knobs are not yet fitted, and when I touch the volume pot shaft to alter gain, the oscillation disappears! What the hell is going on here?? Very confused!

[soundguruman]

First, try installing a 7pf 1000V cap between plate and grid of the first preamp stage.
BUT, It's really hard to tell from the dark photo. Can only see about half of it.
Is the inputs grounded when you test it? Must be to make a judgement.
So, post more photos, more close ups, better light.

When you have this type of oscillation, it's usually in the first 1-2 stages of the preamp.
Then the rest of the stages amplify it.

Pull out the first preamp tube.
Does the oscillation stop?
I thought so.

[Chuck H]

It might be worth noting that WRT high gainers they can often be unstable outside the cabinet, but stable inside the cab, when the top shield is in place. If this a high gainer you might try placing a pot lid or some other metal plate like object on the chassis. So that it covers the preamp area. Otherwise, I would guess by the symptoms that it is a ground scheme problem. Either the input is not grounded well enough, some things are "daisy chained" that shouldn't be, or there is a poor filter arrangement for the preamp. Any of this could cause instability. Since touching the bare pot body of the volume control stops the oscillation, this means that including your body as a secondary ground makes the problem go away. Therefor, an examination and refinement of the ground scheme may well fix the problem.

[priesty]

Thanks for the comments, I will try to post improved photos later.

Filtering - each preamp triode has a 22uF capacitor with good subdivision of triodes with 10k resistors between triodes.

Ground scheme - a "galactic" style combination of star earths and ground 'bus'. I think by what you are saying the ground will be the issue, so I will invest some time in it tonight and will see what happens! Thanks again.

[priesty]

Update, I quickly checked ground resistances from chassis ground to inputs and their volume pots, all filter capacitors, middle pot, power tubes and got 0.2ohms for everything, so I think it may be a ground interaction thing. Will have more of an in-depth look later and report back!

[Enzo]

I am a litle confused. You say the only way you know it is "oscillating" is because a VU meter indicates a level?

Have you connected a speaker to hear what is going on? Or scoped it? If touching the metal controls stops this signal, might it be hum?

I tend to reserve the term parasitic oscillations for little glitches perched on the signal waveform. Parasitic to it. If the amop just sits there continuously oscillating, I don't call that parasitic oscillation, I just call that oscillation. But if it is 60Hz, it isn't even that.

[priesty]

It certainly sounded like parasitic oscillation, even with a dummy load, you could hear the plates screeching their electrons off.

I ended up re-making some solder joints in the grounds that didn't look as good, but ultimately, I've removed the HF bypass capacitor on the bright channel outlet mixer resistor and replaced V1 (12AX7) with a 12AY7. Still tonnes of gain, but no metal feedback. Oh yes, I also removed the 1000pF cap from between PI plates and replaced with a 100pF cap. A 390pF cap was put across the anode resistor of the gain stage between volumes and Cathode follower too. Amp plays nice and feeds back just the right amount when a note is sustained.

Thanks again for all your help!

[Enzo]

OH, if you could hear the parts singing, that is further evidence of what was happening.

[Gtr_tech]

Check your lead dress. Keep plate and grid leads away from each other. You can check this by unbundling wires and moving them around.

[soundguruman]

Theory:
The oscillation is caused by the VU meter.
Ye gotta admit, it's time to find an oscilloscope.