Parasitic Oscillations/ Glowing Screens

[JMPGuitars]

Like I said, lots of opinions on both sides of the subject.

[TriodeLuvr]

Well, the other potential issue with mismatched output tubes is 100Hz/120Hz hum. That's not just an opinion, it really happens!

[JMPGuitars]

That's an if, and also depends on how mismatched, right? Matched by definition can be between 5% to 20% variance depending on whose opinion on what matched means.

I haven't had any issues either way. I get tubes matched if it's free, and not if it isn't. I can test my tubes and match them better than any tube matching service anyway if I want. I test for actual voltage & current levels rather than arbitrary numbers a tube matcher made up.

[Kegger87]

For what it's worth, I have tried the hotter tube in both positions and the screen problem pretty much stays with V4 regardless of what tube is in it. I threw my good testing xf4 Mullard glass in and biased up for a test. One tube has alot of blue glow with signal from the interior of the tube, but this follows that tube when swapped, and from what I've read is somewhat normal for old production El34's. These Mullards however, regardless of which tube is in V4, will now redplate ever so slightly in the very center of the plate. It does not redplate until you start throwing large signal at it, v4 will idle with no signal off standby and not redplate. I only noticed this as I had the lights down low while investigating the blue glow from the inner structure. Had the lights been on I don't think the red plating was intense enough to have been noticed on V4. And of course swapping the tubes between positions, the problem is (any tube regardless of brand) in v4 only. I'm really at a loss here, but have pulled the Mullards back out and gone back to the Jj's for further testing.

[Bieworm]

Analyze it.. the screens are fed from the same source of the B+, so that circuit is not really a suspect. The plates are fed by the OT primary. Maybe there is something wrong with the OT? Try to swap the wires from the OT primary feeding pin 3. Does this change the problem or problematic tube socket number?
I know it’s a hard one on your build. It looks like you don'thave enough length on one wire. Rookie mistake. Always cut both wires at the same length (longest run) for you might need to swap those to avoid oscillation
By the way… what OT are you using? Most OT’s have the same primary colours (red to B+, brown and blue to the plates) Are you sure you used the correct wires?

[Kegger87]

The OT is a Modulus. You are correct, I had the red and white on the opposite sockets as per the layout document and had horrible oscillation on first sound test. I swapped them after extending the lead and the amp was up and running. I read on Rob Robinette's website there's a 50% chance of the leads being correct on the OT and learned from this for next time. White heat shrink would have been better as well. Mike at Modulus has instructed me to take resistance readings on the OT, I will be doing this and confirming negative bias again hopefully by the end of the Long weekend. What would this circuit's design spec out for proper resistance readings?

[Bieworm]

Ask modulus about the resistance values of the OT to be sure. Theoretically both halves should have the exact same resistance. Redplating is a sign there is way too much current being drawn. You can check the current by measuring it across the 1 ohm cathode resistors. But also by measuring the voltage drop across the screen resistors. Check all resistors for having the right value.

[Kegger87]

Ok I finally had a chance to take some readings, measuring from the center tap to each pin 3, I got 42 and 42.2 ohms. Measuring the secondary side, I measured between the OT secondary lead going to the ground on the output jacks (assuming this is the center tap), and to the OT secondary wires going to the impedance selector. I got values of 0.9, 0.6, and 0.5 ohms. Does anyone have insight on these values? I have contacted Modulus but am waiting to hear back as we have a large time zone difference being on opposite sides of the world!

[Bieworm]

Those measurements seem ok to me. I think you should put all your bets on the current draw of the redplating tube vs the not redplating one. Maybe there is some kind of coupling going on, or something touching something like a ground or… also look for faults in the grid circuit of the power tubes

[Kegger87]

Ok, after some guidance from Modulus, the negative feedback resistor was disconnected from the output, and the OT leads on pins 3 on V4 and V5 were swapped. Play testing led to a redplate still on V4, so I think that has eliminated the probability of a fault being in the OT. After the amp cools off I'll double check the negative bias still exists but last i tested I still had -40 odd volts.

[Bieworm]

Have you tried tightening the tube socket pin grabbers? When that is not connecting properly you can get redplating too

[Kegger87]

Back from a hiatus with amp repair, baby on the way and had a busy fall getting ready for winter.

Back to the plexi amp after a long break, I changed out the coupling cap in the phase inverter on the v4 side grid although I could not make the cap test bad. The negative feedback and the OT leads were all put back to proper orientation, and a slightly cooler bias at 36mv at 450 plate volts was redone.

The amp on the test bench with a signal generator to amp, then through a dummy load outputting into my oscilloscope, performs flawlessly. My dummy load is only 50 watt rated and was hot to the touch and I could not make the amp redplate after repeated abuse.

Moving on to play testing with a guitar into a 4x12, the v4 tube will show a redplate while actively playing a chord or note progression (harder playing equals more intense redplate). As soon as you back off the intensity or guitar volume the red plate dissipates.

It was suggested I try to retention the tube sockets but as the redplate isn't in a runaway condition I don't think this is the cause. Another outside suggestion was I switch pin 5 of both power tubes and see if the redplate follows that side.

Is there anything else I can try? I appreciate everyone's patience with a newbie.