Odd Distortion Noise

[Daver]

I have an interesting problem with a circa 1940 Gibson EH-150. When playing through any input, at any volume level, there is a distortion that is sort of behind the note or chord. It's almost like having an overdrive pedal mixed in at a low level with the clean signal. You can hear it best after the initial hit of the strings. Even if you roll the volume on the guitar way down it still exists. I recapped (every electrolytic and coupling cap) this amp about 6 months ago for the owner, cleaned and tensioned the sockets, jacks etc. The weird thing is if I feed the amp a sine wave there is no distortion I can see on the scope. It will clip if it's cranked up enough, of course, and that looks normal, but at the same level the waveform looks clean at the output a guitar played through it exhibits this distortion.
I have tried another speaker, guitar, guitar cord, all tubes, replaced a few out of spec resistors and tried varying the frequency of the signal generator to see if the distortion was frequency dependent. I subbed out the field coil for resistors. I'm sure I'm forgetting other things I've tried. Any ideas? Thanks.

Dave

[Mick Bailey]

What you describe sounds like crossover distortion. Have you scoped just the crossover point to see if there's a notch? Sometimes getting the whole waveform on the display won't show it. Check the idle dissipation of your output tubes.

[Jazz P Bass]

I would try to 'scope' the stages while strumming a chord on the guitar.

I have run into 'can't see it with a signal generator' before.
Sometimes it takes that varying sine wave to show the true nature of the problem.
It's usually a capacitor.

[g1]

Attached schematic shows 2 versions, which one do you have?

[Daver]

I have the 6L6 version, g-one. I'll give the other suggestions a try and report back. Thanks!

[The Dude]

The sound you are describing sounds like what I hear when a B+ filter is failing. I know you said you replaced the electrolytics 6 months ago, but that doesn't mean you couldn't still have a bad one. I would start with scoping B+ and see how much ripple you have and/or if the DC goes low when signal is applied.

[Daver]

OK, I scoped the amp while playing a guitar through it. I obviously got a more complex waveform, but nothing furry, spiky or squared off. I replaced the 4 filter caps one by one. No change in the distortion. I don't see any crossover notch either. The tubes are drawing 60ma each at idle with 300volts between plate and cathode. The B+ doesn't drop but a volt or two when I'm playing through it at a volume that produces an otherwise clean tone. Could it be the output transformer? Thanks for all the help!

[52 Bill]

The tubes are drawing 60ma each at idle with 300volts between plate and cathode.

Really 60mA each?

[km6xz]

60ma? That noise is simply the tubes screaming "HELP, I am burning up, save me"

[Daver]

Yes, 60. It's cathode biased with a 200 ohm cathode resistor. They're 6L6GC's with 300 volts between plate and cathode. That's about 17 watts dissipation on a 30 watt tube. New tubes that run cooler don't change the noise. I've changed the cathode resistor to one as low as 150 ohms to as high as 320 ohms. No change.

[nickb]

How about you signal trace? Run into a dummy load and then listen to the signal using a second amplifier working from grid to grid up the chain? At least you could narrow it to a single stage.

[Daver]

nickb, I don't quite follow. Are you suggesting taking the output from each stage and running that into the input of another amp? Thanks.

On another note, I subbed out the output transformer and the distortion remains.

[J M Fahey]

It's the speaker.

Try the amp with another certified good one.

[Daver]

Nope. That was the first thing I tried. I listed it in the original post. Too much to read, I know.

[nickb]

nickb, I don't quite follow. Are you suggesting taking the output from each stage and running that into the input of another amp? Thanks.

On another note, I subbed out the output transformer and the distortion remains.

Yes, exactly. At least you'd be able to narrow it to a single stage - not that there are many of them! Edit: Don't forget an attenuator.

Sorry for slow response - been having major email problems.