"Black Flag" JTM100 - the build has begun!

[Flames1950]

Grid wire on V3 too close to the plate resistor/voltage? 5.6K swamp resistors aren't too close to anything tucked under everything like they are? (I stuck with the old style with the swampers strung between the tube pairs and no swampers on the inside tubes on my replicas so I didn't have to deal with that.)

[shakti]

Actually, the grid stoppers *are* very close to the screen wires. You can see a detailed photo on page 2 of the thread. The buss that connects them is running parallell with the screen wires too. There's not really any room to separate them more. Could that be it?

Still I don't see how reading DC at PI plates could elicit the problem if that's it...

[shakti]

Some interesting new observations:

- with this new multimeter, measured plate current (see, I'm learning! ) is slightly more stable, but still not entirely. I am also not able to elicit any wild plate current swings by measuring DC at PI plates. Maybe that was an artifact of my old multimeter? So let's disregard that for now.

- measured plate current at idle (i.e. when switching from standby, no signal in, all controls at 0) is sometimes stable, sometimes not, swinging by an order of 1-2 mA. Playing the amp with bass and treble on 0, mids on 10, can start off a plate current swing, but not as great as before. This time a swing on an order of maybe 8 mA. Again, may have been something to do with my old multimeter, but it still swings, and with it I see the tubes (all 4) flashing/pulsating.

- when measured plate current swings, voltage drop over the screen resistor also swings. I guess that's normal under normal playing, but does it tell us anything more about where the problem comes from? Right off of standby, the screen resistor voltage drop also fluctuates, so I guess that eliminates once and for all the actual bias circuit as the problem?

[shakti]

BTW, moving V3 grid wires around doesn't affect anything. Neither does the presence wire. Haven't tried the tone control wires yet, but that seems like a long stretch.

I may just have to go the route of replacing tone stack caps one by one...would really suck if one of the 0.022s is bad...

[Flames1950]

Playing the amp with bass and treble on 0, mids on 10, can start off a plate current swing, but not as great as before.

Playing the amp at all should generate cathode/plate current of course.......

I usually can read some small variance at idle on the cathode current, but seems like it's only in tenths of a milliamp. Meter sensitivity and averaging/speed may make this look better or worse I suppose. Honestly, bias is usually set with all controls on zero so I couldn't say if there's more or less variance there when turning the tone controls -- I've just never bothered to try it.

You may be to the point where it's time to concentrate more on the farting out in the sound, at least until you can prove a relationship between the farting and the variance in cathode current at idle.

[shakti]

Playing the amp with bass and treble on 0, mids on 10, can start off a plate current swing, but not as great as before.

Playing the amp at all should generate cathode/plate current of course.......

Of course...but it doesn't continue to swing up and down long after you stop playing, unplug the guitar, turn off the lights...

I agree though that the most likely explanation is a bad component somewhere that causes both farty bass, less output than expected, and unstable plate current. The idle current swing is greater than on any of my other amps, but I guess that *could* be a burn-in issue. Less output (measured highly unscientifically by checking how hot the attenuator gets) I guess *could* be a burn-in issue as well, but not as much as this? The amp seems lod, but I can't play it here without attenuation. The attenuator only gets lukewarm, even with everything on 10 for 10 minutes. Playing my Hiwatt through the same attenuator makes it scorching in 5 minutes.

[shakti]

Just one more thing before I pick this poor thing apart...isn't it strange that the oscillation/modulation of plate current swings both up *and* down? So if idle plate current is at 36mA, when the modulation occurs, it will swing down to 32 mA and up to 40. Playing normally, plate current will only shoot up from the idle setting, right?

[Flames1950]

Just one more thing before I pick this poor thing apart...isn't it strange that the oscillation/modulation of plate current swings both up *and* down? So if idle plate current is at 36mA, when the modulation occurs, it will swing down to 32 mA and up to 40. Playing normally, plate current will only shoot up from the idle setting, right?

That's been bothering me too.

[shakti]

Hm...funny thing: I removed the grid stopper resistors, and went with a single 1k5 between the sockets, like originals. Now the plate current swings mostly upwards, i.e. idle reading is 36 mA, now it'll swing up to 40, and dips down to 35. So it's not a symmetrical swing anymore. I measured this on both V4 and V5, and it's the same range of swing on both (although different idle reading of course). I can't say with absolute certainty that this change coincided with the change of grid stoppers. At any rate, the swing seems to be slightly smaller as well.

Is it perhaps a burn-in issue after all? Maybe the mustards just really need to be broken in to stabilize?

[shakti]

One guy at the Plexi Palace forum suggested that twisting the OT wires like I did could sometimes lead to oscillation. Any truth to that?

However, they are pretty close to the rectifier and mains filter.

Thinking out loud: if twisting the OT wires can cause a problem like this, and removing V3 stops the problem, then it must start as soon as a signal is running through the OT. However, turning up bass (which presumably will send *some* signal through the OT, if only hiss) will stop the problem, perhaps by "overriding" the oscillation (which by nature seems to be a low frequency one, due to the slow cycle).

[neikeel]

I imagine that you know the reason why it continues after you have turned off? It is because on these amps the HT switch going off only isolates the power coming in. This means that the preamp cans, the PI cans and screen cans will still hold their 500v or so and will only discharge slowly. The later amps with both bleeders and the on/off switch before the diodes allows all the caps to discharge. If you have an oscillating signal this appears to be being fed by these cans once you have switched off (bit like the old trick of playing out when you turn off to discharge the caps - and no I prefer the croc clip on V1 plates but it does work I have measured with and without doing it, although someone suggests cathode stripping may happen if you do it regularly).

Any way I still think that it is a problem around the tone stack (probably a leaky cap allowing mains or ac to contaminate your signal), why not sub in some spare sozo's or mallory's you have lying around between the pots and tone stack and temporarily replace the tone cap with a modern mica, although I would still go back to separating your screens from the bias resistors too, but that is because I lack innovation when I am at the limits of knowledge .

Flames and SDM are much more experienced at this than me so I will lurk only now

Good Luck

[shakti]

Sorry, I wasn't clear on it...I meant that the problem continues if I stop playing or even unplug, so the amp continues to produce an inaudible signal which modulates the power amp. I didn't mean literally switch off the amp, although I guess it could happen the way you describe though. I am painfully aware that the B+ from the mains filter is before the standby switch and doesn't drain when you turn it off, unless you leave the standby switch open. Don't ask me how I know...look at my hairdo!

I have a fellow trying to help me on the Plexi Palace forum as well. I am going to track down this parasitic oscillation (which it appears to be) no matter what! I'm just reluctant to go changing parts on a whim, but if that's what I have to do, so be it.

[shakti]

A few more things tried:

- disconnected the feedback loop (removed 27k feedback resistor): problem gone. Hard to say how the amp sounded like that, as it is of course very loose and gainy with no feedback. This suggests that the problem, wherever it starts, feeds back into the phase inverter. Someone at the Plexi Palace forum suggested that if I had a high frequency oscillation/parasitic oscillation that feeds back via the feedback loop, the returned out of phase signal would be ever so slightly delayed compared to the "original" oscillation, that I could get a low frequency "beating" modulation similar to what I saw with the plate current swinging up and down.

- reconnected the feedback loop. Removed the two 0.022uFs and 56k slope resistor in the tone stack. No difference, turning bass up still stabilizes it. Removed 250pF treble cap in the tone stack as well, still no change.

- voltage drop over the screen resistors is also unstable, and stabilizes when I turn up the bass.

[shakti]

Removing PI input cap (0.022) stops the problem. However, subbing in another cap of the same value, the problem is back.

What next? PI filter caps?

[neikeel]

Removing PI input cap (0.022) stops the problem. However, subbing in another cap of the same value, the problem is back.

What if you put another back in and remove the 0.1uF input to the PI (the one that leads to the presence?

I guess a couple of different PI caps might be worth a try as that is where your 'noise' is probably coming from.