Odd Distortion Noise

[Daver]

I did sub out the field coil for a resistor. The chassis is out of its' cabinet. I have already replaced the filer caps again one at a time. I'm using 22uf F&T 500vDC caps. I will try monitoring the OT B+ node. Any chance the PT could be at fault? Thanks.

[Daver]

If I put the scope on the B+ node at the OT center tap (no signal input) I get a fairly nice sine wave. Looks to be about 120 Hz. If I use my probe to "listen" to that node I get a just a hum. Just for fun I did the same on the cathode resistor. when I play a guitar I hear the distortion more clearly. It's almost the foremost sound. Putting the scope on it without a signal reveals a sine wave but it has a bunch of furry hash on it. Hmmmmm. It has to be on the plates too. So I go back to the plates of the power tubes. With no signal input there is a furry waveform. It's also kind of triangle shaped. But this wasn't the way it was when I looked at it before. ??? If I play the guitar the waveform "cleans up" initially then gets funny again as the note fades. WTF? I'm going to start over. I've replaced or subbed out every component in the output other than the sockets and wiring. Any other ideas? Thanks.

[okcrum]

All the losses in the old parts that were preventing oscillation are gone now that you have replaced them all. Now it's time to cure the oscillation. Add screen and control grid stoppers and you should be fine. A snubber across the OT primary will likely work as well.

[Daver]

I have added screen and grid resistors. It doesn't sound like an oscillation, just distortion. If the old parts were preventing oscillation, why did the problem exist in the first place?

[Daver]

Here are some pictures of the scope taken at the OT. It's a 1KHz sine wave at about 500mv AC input. The first is with the volume on about "2". The distortion is heard with a guitar at this level. The second is the volume halfway up and the third is the amp dimed. I'm using the Mic input. The distortion is heard through the normal channel as well. Remember I can hear the distortion from the output tubes only. Notice how the first scope picture doesn't show anything unusual but distortion is heard. Weird.

[nickb]

I don't see the fuzz on the traces you mentioned.

On even the first trace there is significant even harmonic distortion but that could be simply because your are driving with 500mV. That would give around 100 * 1.4 *2 x 0.5 =140 Vpp on the first 6SQ7's plate. That would be pretty distorted. Also there a lot of gain on the mic input. How does it sound if you plug into the instrument input?

[Daver]

Same on the Instrument input. I know this thread is long and confusing. I used a probe I built to sample the circuit and output it to another amp to monitor it. The distortion I hear is not present all the way up to and including the power tube grids. It is present on the plates and cathodes. It's happening in the power tubes somehow.

The fuzz I was referring to is shown in the first picture. It's taken from the plate with NO input signal. The second picture is the same plate with the signal applied. I backed down the input to 300mv.

[nickb]

Same on the Instrument input. I know this thread is long and confusing. I used a probe I built to sample the circuit and output it to another amp to monitor it. The distortion I hear is not present all the way up to and including the power tube grids. It is present on the plates and cathodes. It's happening in the power tubes somehow.

The fuzz I was referring to is shown in the first picture. It's taken from the plate with NO input signal. The second picture is the same plate with the signal applied. I backed down the input to 300mv.

[ATTACH=CONFIG]32611[/ATTACH]

[ATTACH=CONFIG]32612[/ATTACH]

At least using the instrument input got rid of the even harmonic distortion. The first 'fuzzy' scope shot is pretty meaningless without scales: what V/div and sec/div are you using?

EDIT: Afterthought: do you see the same 'fuzz' on the output of the transformer?

[Daver]

I was still using the Instrument input, but lowered the input signal strength. The fuzz is not present at the OT. The scope was set for 5mV/Div. and 2mS/Div. The probe was on X10. I guess it's pretty small fuzz.... I think I'm over analyzing everything. Thanks.

[nickb]

I was still using the Instrument input, but lowered the input signal strength. The fuzz is not present at the OT. The scope was set for 5mV/Div. and 2mS/Div. The probe was on X10. I guess it's pretty small fuzz.... I think I'm over analyzing everything. Thanks.

20mVpp on the plates is very small. The fundamental is 120Hz so this looks kinda like PSU ripple but there is a strong 240Hz component which is unexpected . In any case you say it's not on the output. If it's common to both plates it will be cancelled in the OPT. I suspect this is not the gremlin you're looking for. So, we are still no further forward

How about an audio sample direct from the output of both 1Khz and Guitar? It might be interesting to look a the frequency spectrum of the 1KHz sample (use PC based FFT function on tool of choice). People laughed at me the last time I suggested doing an FFT here but here I go again

[Daver]

I don't have a good way to record. Where can I get an FFT?

[nickb]

The best way would be to attenuate the 1Khz (500Hz might be better) signal to about 1Vrms and use the line in of the audio card on your PC. Record using tool of choice - Audacity is free and does decent job. It also includes an FFT (Analyse->Plot Spectrum). Post the audio too- it would be good to hear what you hear.

I know this is a bit 'out there' but since we can't see anything on the waveform we are going to have to try a different way. I'm not optimistic since there is not even a hint on anything on the waveform. BTW FFT's are often built in to cheap DSO nowadays, including PC based ones.

If you haven't got access to a PC with a line in, there are spectrum analyzer Apps for phones.

[stokes]

Yet all of them have been replaced. And, again, this is happening in the power tubes and not the phase inverter so that should eliminate any coupling cap. Thanks.

I've read everything here and dont recall seeing if you checked the grids of the power tubes for dc voltage.Sorry if I missed something here.If the caps from the PI to the power tubes are leaking you wont see the distortion till you get to the plates

[Daver]

stokes, there is no positive voltage on the power tube grids. The amp had the original coupling caps replace with some Nichicon radials. I replaced those with some axial caps because the Nichicons were poorly cobbled together in the amp. Not very safe. So we should be good there. Thanks!

[Daver]

Here's a recording I did of the distortion. It doesn't record very well. You can hear it as the chord trails off. Yes, I'm not the only one that heard it..... http://www.soundclick.com/player/sin...&q=hi&newref=1