JTM45 Bias Voltage moving around & bad distortion

[OrbitalLex]

Hello all,

I have a JTM45ish amp on the bench and everything is working fine (sort of), voltages all check out within reason, considering my slightly lowered wall voltage here.....except for the bias. I can set the idle current to say 40ma, but it jumps around as high as 70ma while playing through it. I have never played through an amp while biasing before, so maybe this is normal....but...

There is also a bad farting distortion (blocking/crossover) that is matches the bias voltage spikes. If I let a note ring, the distortion falls off before the sustain of the note does.

Just so my voltages don't throw anyone for a loop, this amp has no "Normal" channel that the traditional JTM45 does, only the "Bright" channel and it also has no rectifier tube. B+ is SS rectified. This gives me the impression that the flailing bias voltage is not so normal.

Here are my voltages if it helps and thanks for any input/advice.

V1
p1-0 (not used)
p2-0 (not used)
p3-0 (not used)
p4-3.1ac
p5-3.1ac
p6-178
p7-0
p8-1
p9-3.1ac

V2
p1-166
p2-0
p3-1.1
p4-3.1ac
p5-3.1ac
p6-300
p7-166
p8-166
p9-3.1ac

V3
p1-238
p2-24
p3-43
p4-3.1ac
p5-3.1ac
p6-232
p7-26
p8-43
p9-3.1ac

V4
p1-40mv
p2-3.1ac
p3-400
p4-395
p5-neg44
p6-397
p7-3.1ac
p8-40mv

V5
p1-40mv
p2-3.1ac
p3-400
p4-395
p5-neg44
p6-397
p7-3.1ac
p8-40mv

Thanks again!

Chris

 

[OrbitalLex]

And if it helps, I checked every component before and after installation for value, dc leakage, continuity, etc. The only part of this contraption that I am unsure of is the choke. It is ancient, but reads 20H. I don't know how to test it any further than inductance.

 

[JMFahey]

I can set the idle current to say 40ma, but it jumps around as high as 70ma while playing through it.

It's called "Class AB" 😉

 

[OrbitalLex]

Cool, so that answers that part of it. Thanks, JMFahey 🙂

Now, I'm curious about this horrible distortion. I'll keep digging at it.

 

[JonSnell Electronic]

Check the anode load resistors. You cannot check quiescent current with a signal flowing.

 

[OrbitalLex]

I reduced the cathode cap at pins 3 & 8 of V1 from 220uf to 1uf and the problem sorted itself. Nice, thick and clean tones. Sighhhhh. I have spent weeks poking and prodding at this thing. Now on to the next project. Thanks guys!

 

[JMFahey]

Then it was probably what's called "blocking distortion" : a strong signal peak gets rectified by following grid, shifts bias a lot, so much so that the tube gets cutoff form a fraction of a second, then it drains off and tube comes on again.
Sounds horrible.

 

[OrbitalLex]

Yeah, I figured it was blocking or crossover distortion and the scope led me to V1 cathodes. My old HP scope took a lightning strike in October and was able to borrow one a few days ago.

 

[Bob Richards]

The effect of blocking distortion (ugly fizzle) and the crossover distortion glitch it creates can be significantly minimized by putting a "conjuctive filter" across the speaker wires (I use a 20ohm 5W in series with a 0.22uF cap for an 8 ohm 15 watt amp circuit). It compensates for the substantial rise of impedance that guitar speakers have when you go up in frequency (above about 6kHZ), which allows the OT to ring, which amounts to a substantial amplification of the crossover distortion glitch, the spectra of which is not musically related to the signal of interest. There's also the Paul ruby zener mod that many people get wrong. I tried it and it worked real good.