Asteroid unpleasing distortion

[arjepsen]

Actually the ppimv is/was all the way up for these measurements, so that shouldn't be the culprit.
Anyways, I'll see if I can't get around to messing with the output plate leads, and the nfb tail when I get home tonight.

[arjepsen]

ok, first I tried reversing the OT primaries - but that made the amp oscillate, so I put them back where they were.
Then I disconnected the 47K feedback resistor.
It did help with the waveforms, showing that the pi now distorts after the power tubes.
However, the gritty raspy decay on notes are still there (even though it doesn't seem to end as aprubtly as before.)

Here's the newest scope pics - yellow is across output jack, blue and purple are the powertube grids:

DS1Z_QuickPrint1.png

DS1Z_QuickPrint2.png

DS1Z_QuickPrint4.png

DS1Z_QuickPrint5.png

DS1Z_QuickPrint6.png

[romberg]

If the amp went straight into self oscillation when you reversed the power tube grids then you definitely had them right the first time. So, that was a false alarm. You can reconnect the negative feedback as well.

However, on these shots, I think I see a fuzzy high frequency signal on the top of the clipped waveforms. I may just be wanting to see fuzz there. But they don't seem near as clean as the sides (more vertical bits). One can fiddle around with the vertical and horizontal knobs on your scope to zoom in on the tops a bit.

If you don't have a fizz cap, try putting a 50pf cap across the PI plates. You could even try 100pf. You will want to use a cap rated for the voltage. So, probably a ceramic of 500-1000V rating.

Mike

[arjepsen]

Here's acouple of zoom-ins on those tops.
Not sure which function shows it most correctly - "High-res" or "average":

High-res:

DS1Z_QuickPrint3.png

Average:

DS1Z_QuickPrint4.png

Will try reconnecting the nfb (wouldn't that bring back the pi distorting before the output tubes?)
and the fizz cap

[arjepsen]

Here's a few more clips:

No nfb, no fizz cap:

withou fizz cap - no nfb.wav

82pF fizz cap, still no nfb:

with fizz no nfb.wav

With fizz cap and nfb reconnected:

fizz plus nfb.wav

I still hear buzzy distortion - particular on lower notes, and the raspy/gritty decay.

[Reeltarded]

I would always advise people to build the amp and make it work before installing PPIMV. That has to be the fail here.. has to be..

Whats that guitar? Why am I hearing what sounds like an antenna picking up Sun spots? at those settings even flipped over and unshielded I think that is too much cosmic microwave background. heh

Unplug your cable and hold the connector and shake it against the floor.. I hear microphony that I want to choke

[arjepsen]

I would always advise people to build the amp and make it work before installing PPIMV. That has to be the fail here.. has to be..

Whats that guitar? Why am I hearing what sounds like an antenna picking up Sun spots? at those settings even flipped over and unshielded I think that is too much cosmic microwave background. heh

Unplug your cable and hold the connector and shake it against the floor.. I hear microphony that I want to choke

Hmm
.. I'd think the LarMar ppimv wouldnt matter, when it's fully up. However, I can try to pull it out for these testings.
The guitar is a warmoth strat build - fralin vintage hot pickups. It is quite susceptible to noise, but not more than other strats I've tried.
Turning the guitar volume all the way down makes it all quiet.
Not sure I completely understand your last line.. I belive the cables are alright, but I could be wrong?

[Reeltarded]

If you unplug the guitar end and rattle the cable around and its noisy: problem.

it sounds like "gwabble guabvle gwabble".. probably jist hearing the fitzing of an open amp with single coils

[romberg]

Did you try the suggestion of increasing the value of the grid stoppers on the power tubes? The scope shots you have clearly show something
"funky" going on between the PI input and the power tube grids. It looks like a cap charging/discharging to me. Larger grid stoppers can limit the
current that does this. Merlin mentions you can go as high as 100k on these. Might be easy enough to swap some higher values in and see if
they clean up the signal on the power tube grids.

Also, double check the treble pot in your tone stack. I've seen amps where this pot goes dodgy and starts to cause all kinds of havoc with the
audio path (usually when the center wiper contact is flaky).

Lastly, is your LarMar a 250k dual gang pot with no extra caps? A value on this pot that varies too far from the grid leak resistors it is replacing
can throw off the load the PI sees. What exactly is your PPIMV setup?

Mike

[arjepsen]

@reeltart:
Ok, I think I understand. No, my cable doesnt “gwuable”

@Mike:
The onlly cap between pi and powertubes is the cut control
I really hope my treble pot isn´t bad - its a pec pot. Theyre kinda expensive
Anyway the tonestack seems to work as intended, so Im hoping thats ok.

The Larmar is a stereo 250k pot, with a 2m2 from wiper to bias-point. When it´s fully up, I measure 221k and 218k from wiper to bias-point, so I assume that´s close enough to the regular 220k´s in the schematic. There´s no extra caps.
(I assume there's no difference between what side of the pi plate coupling caps gets connected to the cut control cap, and which gets connected to the cut control pot?)

[romberg]

The onlly cap between pi and powertubes is the cut control

There must be coupling caps on the PI plates? It would be extremely unusual for the PI to be dc coupled to the power tubes. This is one of the
problems with trying to get help with one of these amps where there is no schematic. It is hard to tell what is going on. I'm going to assume that
the PI plates do have coupling caps of some sort.

Is this cut control like an AC30 cut control? That is a cap and pot in series across the PI output?

The Larmar is a stereo 250k pot, with a 2m2 from wiper to bias-point. When it´s fully up, I measure 221k and 218k from wiper to bias-point, so I assume that´s close enough to the regular 220k´s in the schematic. There´s no extra caps.

That sounds fine and normal for this type of MV. I doubt it is the trouble assuming it is working correctly.

(I assume there's no difference between what side of the pi plate coupling caps gets connected to the cut control cap, and which gets connected to the cut control pot?)

Yes. First assuming that it is the AC30 arrangement, then connecting it to the plate side of the PI will put plate voltage onto your pot. Which may make it both scratchy (hmm one of your symptoms) and more dangerous. It should be connected to the side opposite the PI plates of the PI coupling caps. Or better still, removed entirely.

If it is the AC30 type cut control, I would try disconnecting it and see if it is the cause of all the funkyness going on with
your PI. This circuit looks really, really similar to what I think is referred to as a "type III" master volume or cross line master.

This master is just a 1M pot wired across the PI output. It "shorts" the out of phase signals together to reduce volume. And every single one of them I've heard sounds really, really, really bad. Especially when overdrive is involved. It is really an awful sounding circuit. The reason is that the PI output is not all that equal an opposite. So, it never completely cancels each side. Especially when overdrive occurs. The AC30 cut control looks like it with the addition of a cap which makes the awful work at only high frequencies. Disconnect one side of this type of cut control and see if it makes your amp sound better and the scope traces should also look better.

If this is your problem, then you could always add a switch (maybe a pull pot) to enable and disable this. Then only use it when you won't be using overdrive (where it is not as horrid sounding).

Mike

[arjepsen]

There must be coupling caps on the PI plates? It would be extremely unusual for the PI to be dc coupled to the power tubes. This is one of the
problems with trying to get help with one of these amps where there is no schematic. It is hard to tell what is going on. I'm going to assume that
the PI plates do have coupling caps of some sort.

There is - my mistake.
You can see the coupling caps, and the cut control in this schematic:

komet schematic.pdf

Yes. First assuming that it is the AC30 arrangement, then connecting it to the plate side of the PI will put plate voltage onto your pot. Which may make it both scratchy (hmm one of your symptoms) and more dangerous. It should be connected to the side opposite the PI plates of the PI coupling caps. Or better still, removed entirely.

Yes, sorry for being unclear. The cut control is after the coupling caps from the PI - just like in the schematic.
I am familiar with the type III master - this is the regular cut control. I can try to remove it from the circuit, but I assume that might make it even brighter.
Will try it out, and scope some pics.

And yet again thanks for the continued help in troubleshooting!

[arjepsen]

ok, tried disconnecting the cut-cap. No difference in the buzzyness.
Tried upping the grid stoppers on the power tubes to 5k6. Still no difference.

The buzzy distortion is most evident when plucking the lower strings, E, A, D.
The gritty raspy decay mostly on chords.

[arjepsen]

Managed to put together a headphone amp thing, for testing out the signal at various places.
The distortion does seem to appear simultaneously at the pi plates and the ot, so I presume the problem originates around the pi.
Reading up on Merlin, among other things he suggests putting a smaller cap on the second pi grid, than the usual 100nf - I think I'll try that next.

[Roe]

try to move the grid and plate wires (away from each other) if you have not been doing this already