Asteroid unpleasing distortion

[arjepsen]

I would stop all drive stages up to the PI with a 33k on each and see if that totally fixes it. Input stage included..

There's already a 33k on the input. The thirds stage has a 220k/330k voltage divider right in front of it - I assume that one doesn't need a stopper?
That leaves only the second stage - which I put a 39k in front of, and the 10k on the pi. I'll try increasing the one in front of the pi to 33k.

Another question: could it be crossover distortion instead??

[Reeltarded]

yYes, very much so.. but..

Unless there is a stopper on the tube pin: its not a stopper. A little piece of wire is the same as a long one there.. speaking of frequency..

Don't raise the PI stopper. That should be great at 10k. It will affect the outcome, which we are trying to avoid, so everything before it is where to focus that.

About crossover distortion. Yes. What are you thinking? We talked about it a little and it also sounds like something in the negative feedback.

yesterday a friend had me look at an amp I built he was having trouble with. I told him to bring his cables and everything. He was using audio cables I told him to throw away 7 years ago. I retubed and biased his amp, plugged my guitar in and it sounded like pure hash highend finger noise nonsense. 30ft of new signal cable later it was perfect.. make sure this isn't an issue with you input jack or your signal cabling. If a cable is microphonic it does a thing just like that with whatever the fundamental the amp vibrates at.

Things to consider.

[arjepsen]

Thanks a lot for your continued input!
I think the cables are ok. They are about a year old - planet waves.

About crossover distortion: Not really thinking a lot - I don't know too well how blocking and crossover distortion sounds, since most of what I find is other people's description by word.
Anyways, I'll try a 33k after the voltage divider. I might also try to change the NFB resistor out for a pot and play around with that.

[romberg]

Here's a quick noodling - hoping it shows what I mean.
https://www.dropbox.com/s/wuq3tc6wu806t ... p.wav?dl=0

Amp settings:
Gain, treble, mid, presence: 1 o'clock
Bass: 11 o'clock
HiCut: between 2-3 o'clock.

Your clip sounds a bit like what I had going on in one of my express amps. There was some "bonus fuzz" in the background that would abruptly stop as a note decayed. The extra fuzz was not very musical and did not show up at all with clean notes (low guitar volume or low amp gain settings). The bonus fuzz seemed to appear when the amp was wound up a bit. I don't have an amp with your circuit, so I don't really know how much gain is involved with a 1 o'clock setting. But I'm guessing it is a fair amount.

I found the fuzz in my amp by using an oscilloscope. In my case it was an oscillation that occurred right at the moment the power tubes hit saturation. Right at this moment a brief (maybe 1/10 of a waveform) burst of high frequency noise appeared on the grids of the power tubes. The PI was not distorting at all (or even close to it) right before this set in. And the power tubes were not severely distorting yet. Just hit the point where the top of the waves were starting to flatten. As soon as the signal dropped below the power tube saturation point the fuzz/oscillation vanished. My guess is that this has something to do with the grid current in the power tubes. It is possible the same thing is going on in your amp.

In my case, I "fixed" it by increasing the fizz cap on the output of the PI from 50pf to 100pf. So, you might try something similar. It is pretty easy to clip in another 50-100pf cap across the fizz cap and see if it makes things better.

Mike

[arjepsen]

ok, small update:
When looking at trying to put a 33k grid stopper on the 3rd stage, I started wondering why the 220k mentioned earlier seemed to actually be right there as a grid stopper.... oooh...
I had put the 330k to ground "before" the 220k, instead of from the grid to ground.
Fixed that - starting up the amp again, hoping it would now be magical.....

Alas it's still not.
It might be a bit brighter, which makes sense - now there's not a 220k grid stopper on the third stage....
But the annoying fuzzy distortion and gritty/raspy decay is still the same.
Will try the fizz cap you mention (about 100pf between the plates right?) There isn't one there in the original design, but I'll try it and see.

[statorvane]

If you can't nail it down, I would suggest looking at your PPIMV circuit, or bypassing it entirely to eliminate that as a potential culprit.

[stephenl]

Try using 10k for your power tube grid stoppers. As Reel mentioned, right on the grid pins. It might help if it's blocking distortion...

It doesn't roll off the high's audibly, I used 10k grid stoppers on my K50 Concorde and it has no lack of treble.

[M Fowler]

Don't you guys think that the problem is within the bias of the amp, which could be the PPIMV as Steve suggested?

Mark

[Reeltarded]

Yes I think that absolutely could be it too.. also tapping grid wires to see if microphonics are doing it.. does seem like one of like five things.

I still didn't look inside the amp.

[arjepsen]

ok, with Rombergs instructions, I got some scope snapshots made.

Yellow trace is signal generator
Blue is across the OT secondary.

First at clean, gain at noon:

OT out - gain at noon - clean.png

Second is gain at about 1 o clock:

OT out - gain at 1 o clock.png

Third is same setting - pink trace is first 6L6 grid:

pink trace 6l6-1 grid - gain at 1.png

Fourth same, showing second 6L6 grid:

pink trace 6l6-2 grid - gain at 1.png.png

Fifth same setting, showing PI input. Here the probe is loading down the signal

pinkt trace PI grid.png

Am I right in guessing that the blue trace shows crossover distortion??
As far as I can tell, it doesn't seem to be at the input of the PI, but shows up at the grid of the 6L6's, so the problem would be narrowed down to around the PI, right?

[romberg]

Check out this page:

http://home.polstra.com/amps/wreck1/scope/

On these you can see crossover distortion which is the discontinuity in the trace around the zero volts point (thus crossover distortion). I don't see much (if any in your amp).

The funky signal your PI is putting out (looks like a nipple on top of the waveform peaks) I have seen before. I think it is related to the negative feedback loop breaking down. When does this start to happen? Before or after the power tubes start to clip?

[Reeltarded]

There is a phasing issue with your NFB.

[arjepsen]

I think it shows up at the same time

Here's a few more shots, turning up the gain.
Light blue is across OT secondary, dark blue and purple is power tube grids.

0.png

1.png

3.png

4.png

5.png

[romberg]

It looks like your PI is distorting before your power tubes. Which is not typical for a Ken Fisher type amp. Your PPIVMV must not be dimed is this correct? And it could very well be your negative feedback is the wrong phase and is supplying positive feedback to the PI. Swapping either the OT primary leads or the power tube grid wires will reverse it.

Usually when the feedback is backwards the amp self oscillates and it is very obvious. But you have a PPIVMV on there. So, it may be masking the worst of the symptoms. Also a PPIVMV will pretty much kill your negative feedback and presence control when it is not turned close to all the way up.

[Reeltarded]

LIft the NFB tail and see what that sounds like.

Dave totally nailed this. (shakes fists at sky and drinks corn mash)