6V6 marshal/basman style amp farting out at high volume

[elipsey]

Hi, I just wanted to follow up by saying that I settled on a configuration that I'm pretty pleased with, and posting an updated schematic. I made changes one at a time so I could hear the individual effects, and the changes to the power supply were really the missing ingredient in all of my previous experimentation. Increasing the first filter and lowering PI voltage tightened things up at high volumes. The lower preamp voltages also really chilled out the tone of the whole amp in a good way, like daz talked about in his other thread. This whole experiance really illustrates the importance of filtering and power supply. I feel like I kind of get it now about why Marshalls moved to diode rectifiers and heavy filtering after the JTM45. The amps I have been comparing this one to all have solid state rectifiers, and they just seem to hold together better at high volume, even the supersonic 22, which otherwise has the same tubes, power supply and PI as I have tried. Also, I guess most gear will sound sloppy if I turn it all the way up and clobber the low E string on a neck humbucker, so maybe I need to moderate my expections.

I have heard people on the forum say that a big part of the marshall sound comes from boosting treble in the early stages, and then emphasizing bass in later stages and at the power amp. I think this strategy, plus the lower preamp voltages really got me closer to that sound. Even lower preamp voltages might sound good, but I'm happy so I'm leaving it alone.

Having a switchable treble cut cap across master volume really works for me. I like the high gain tone, but it's not too harsh or thin at lower volumes like people sometimes say about the jcm800 preamp. The clean sound is nice and balanced, but I can still use the 'bright' switch to get some twang now and then, even out of darker pickups.

Thanks for your input everyone.

[Chuck H]

changes to the power supply were really the missing ingredient in all of my previous experimentation. Increasing the first filter and lowering PI voltage tightened things up at high volumes. The lower preamp voltages also really chilled out the tone of the whole amp in a good way, like daz talked about in his other thread. This whole experiance really illustrates the importance of filtering and power supply. I feel like I kind of get it now about why Marshalls moved to diode rectifiers and heavy filtering after the JTM45.

The preamp is where a lot of the fun stuff happens but indeed and amplifier is basically a modulated power supply. Your guitar and the preamp control the modulation but if the power supply itself isn't up to a task in one way or another no amount of tweaking will get you there. That isn't to say that a burly power supply is always best. It's precisely the deficiencies in some designs that give them their particular compression, envelope of attack, A certain dynamic and some EQ effects that become dynamic with current. So it's really about making the right sort of power supply for the right sort of amp. At least WRT guitar amps.

[mhuss]

The problem is grid leakage current (due mainly to the fact that the tubes do not have a perfect vacuum inside). If you don't have a low enough resistance, the leakage can make the grid slightly more positive, which increases the idle current, which ionizes more atoms, which increases the leakage, which makes the grid even more positive...thermal runaway.

The vacuum and leakage current vary between manufacturers, and even batch to batch, so a high value will work fine until you get the wrong tube -- or the existing tube leaks a tiny bit. Having a too large value will generally work but will increase the change of (catastrophic) failures.

[elipsey]

The problem is grid leakage current (due mainly to the fact that the tubes do not have a perfect vacuum inside). If you don't have a low enough resistance, the leakage can make the grid slightly more positive, which increases the idle current, which ionizes more atoms, which increases the leakage, which makes the grid even more positive...thermal runaway.

The vacuum and leakage current vary between manufacturers, and even batch to batch, so a high value will work fine until you get the wrong tube -- or the existing tube leaks a tiny bit. Having a too large value will generally work but will increase the change of (catastrophic) failures.

That's interesting. I think every 6V6 amp I have ever played had 220k bias resistors, for example DR, PR, and various other fenders like supersonic 22, Dr. Z Z28, and maybe more I can't think of... so it seems like it must be close enough for most tubes. Have you ever seen thermal runaway happen, that didn't seem to be caused by bias failure or some other problem outside of the tube? Do you think maybe the old tubes were both better and cheaper so the value was regarded as less critical?

[g1]

Have you ever seen thermal runaway happen, that didn't seem to be caused by bias failure or some other problem outside of the tube? Do you think maybe the old tubes were both better and cheaper so the value was regarded as less critical?

I've seen something similar happen several times and am not sure this was the cause but it makes sense. Only with Fender 6V6 amps.
Deluxe being played hard starts to redplate, user notices and shuts amp down. Upon re-start tubes work fine and player finishes gig. Problem does not occur again until pushed hard enough, or maybe does not happen again. No fault in bias circuit.

[elipsey]

Yikes. Well, does this mean I should be trying to replace the bias resistors in every 6V6 amp I see? I don't want to be that guy who's trying to fix stuff that somebody doesn't think is broken when they bring me an amp.

Maybe just the ones I build...

[Joe L]

Another direction to attack the problem is by adding a power dampening control like the Mojave Scorpion. All it is is a 50K reverse audio taper pot placed between the 470 ohm bias resistor and the cathodes of the phase inverter. It works amazing well for being so simple in dropping the PI drive to the outputs without changing the distortion quality. A linear pot works just fine except for being a bit touchy. Just make sure to wire only two terminals of the pot so it is a varying resistance.