Using Oscilloscope to Track Oscillation

[surfsup]

I have a digital oscilloscope and a current build that has just a slight trace of oscillation somewhere. It only appears when I have everything dimed, and the 1st gain pot at 8.5 or higher.

Using the oscilloscope, I don't really get perfectly clear traces. They all have some fuzziness to them, wherever I probe. The electricity in my amp room is pretty crappy too, I should mention. If I dime an amp in the room, at around 6 or higher there is a high pitched whine on top of the amp. If I move the amp into my dining room, the amp has no such whine when dimed.

So perhaps I am being hurt by this electrical issue, but I'd like to use the scope to track oscillation better. What ideas do you guys have? I have googled endlessly but finding any tips on tube amp oscilloscoping is pretty much a needle in a haystack event. Thanks in advance as always.

[Colossal]

Not related to the scope protocol, but have you run an extension cord from a known good electrical source in another part of the house (such as the other room where you get no oscillation), BUT, leave the amp in the room where you are getting oscillation? Just wondering if the source is environmental or truly electrical/mains on that one circuit in your place. Shield plate in place when the amp is oscillating?

[tubeswell]

Hi surfsup

Has your amp got a NFB loop? (If so, try disconnecting that first and see if that is the cause of your oscillation. I'm not thinking about +ve feedback here, but rather the possibility of the 'wrong amount' of NFB)

[surfsup]

Colossal, no I did not try that but I have nothing else electrical on besides the amp in that room so I don't see how it can be environmental. The room has a window and with a sunny day I can see fine.

No NFB loop. this is the Xits 10 clone that I did a while back and tamed the previous oscillation by moving the EQ leads around. The circuit is a liverpool except the TS is after 2nd stage instead of 1st. I modded the amp to do a few things. The 3rd stage grid leak is supposed to be 56k, I have a 100k switch in series for 156k. I also modded the slope resistor on a switch for 33k/56k/100k options. Vox TS but 0.01 mid cap. I also put a mid cap in series 0.01/0.0047 (crunch switch - me likee). Bass pot is 1M with a 1M across it for 500k to reduce the bass slightly - this circuit has a lot of bass (possibly due to the film PS caps I used?). With all of it dimed at 10 I get some oscillation starting at around 7.5 (3pm) on the gain pot. At 7-ish its not there even thought there is a lot of gain still, unless I play a full chord, but I would never do that anyway with everything maxed like this. This is producing quite a bit of gain with the 156k to ground.

Single strings are fine. Power chords (strings 1-2-3) are fine. If I play a double stop on "some" frets, the oscillation I describe above is there. but only two strings or three strings and only when it is the high strings and only at some areas on the neck. it is subtle but I can hear it. It slighly pulls the sustain down a bit too. I noticed if the guitar is slightly out of tune, this oscillation is very unforgiving and gets quite a bit worse. But overall it is subtle. Single strings sound insanely good. Guitar Vol down, the cleans are very separated and piano-ish. I really like this amp.

Here's what I tried.

-I chopsticked the EQ leads quite a bit. Seems they are the best they will be as is now
-I put various caps (18pf to 250pf) between g-p, k-p and as snubbers across Rp on stages 1, 2 and 3 (currently have a 250p across Rp on stage 2 only but a 18pf across Rp of stage 3 seems to slightly help)
-I have the gain pot rotated 90 deg with the bright a DPDT with 100p and 500p, both caps are super short sitting against the top/bottom of switch with short leads to pot lugs
-Moved the MV to PT grid wires, no change (thinking about shielding these)
-Ran a coil around the 50p treble cap in TS and grounded one end to chassis
-grid leads are all shielded (except Ptube grid runs - these are twisted unshielded leads running from faceplate to back of amp across PS caps)
-Paul Ruby diode mod across 220k Ptube grid leaks (2xel84 amp)

I guess, is this what I should expect from a high gain 2xel84 amp with the gain this high? Or can this be perfected? I have been working on this on-off for a while now and feel i've exhausted my knowledge and am sick of working on this but I'm 98% of the way to the promised land! I just wish I could "see" the oscillation on the scope!

[surfsup]

Couple other questions besides the scope stuff...

1) Could playing through an attenuator (airbrake) set low cause oscillation?
2) If I have the grid run shields grounded at the tube socket rather than the pot, could that cause a difference?
3) Will a grid run parallel to a grid leak to ground (0.5"-1" in length) cause oscillation? I don't see how on this one but what the heck I'm asking...

[overtone]

Sometimes i look for a lead-dress parasitic by moving slowly between the leads with a totally-insulated big, fat screwdriver.
When you get to the trouble area the parasitic drops off the scope. Somehow the screwdriver soaks up the naughtiness.
Best, tony

[surfsup]

Overtone nice thanks. I'll try that. Any suggestions on how to test doublestops with a signal/dummyload? The reason I ask is the signal of two notes combined changes shape due to two frequencies not matching constantly. I guess i can freeze the frame on the screen. Anyway, any suggestions to test double stops are helpful and appreciated...

[tubeswell]

If there's no NFB, it must come down to lead dress then

[overtone]

...was thinking about the frequency aspect mentioned
how about a signal generator first of all, do a few sweeps to try and provoke the oscillation, then take it from there?

[markr14850]

I sometimes use an old keyboard/synth for a signal generator. Among its "instruments" are sine, square, and triangle waveforms. I place a weight on a key, and it generates a nice stable output. If I want "double stops", I just put another weight on another key.

[Phil_S]

Colossal, no I did not try that but I have nothing else electrical on besides the amp in that room so I don't see how it can be environmental. The room has a window and with a sunny day I can see fine.

The environmental thing could be something you can't see. It could be a florescent lamp in another room, a cable in the ceiling/floor/wall, your proximity to a cell phone, RF from overhead wires on the pole that runs through your back yard, aliens spying on you...the list is endless. Try the extension cord, which will at least tell you if it is the location or the power supply. You have nothing to lose by trying it.

[Colossal]

The environmental thing could be something you can't see. It could be a florescent lamp in another room, a cable in the ceiling/floor/wall, your proximity to a cell phone, RF from overhead wires on the pole that runs through your back yard, aliens spying on you...the list is endless. Try the extension cord, which will at least tell you if it is the location or the power supply. You have nothing to lose by trying it.

Phil,

Hope you are not roasting back there in Baltimore.

Some good points here for sure. Yes, Wi-fi definitely is a source of noise and I keep my amps away from my desktop and laptop computers (when the wi-fi is on). I was thinking a bad ground perhaps in that room only (hence the extension cord idea to rule it out). Compact fluorescent lamps or other high-frequency ballasts are a potential source. I know Mark Fowler reported running amok of a compact fluorescent. Those cheap chinese Feit Electric imports are notorious for inducing terrible line noise. I don't know if that name brand exists anymore but used to see them at Home Despot.

[surfsup]

ok I have the amp in one room, the PC in another. The PC is running Reaper with a 1kHz tone file looping over and over.

The amp is plugged into an extension cord running from the other side of the house. The scope is next to the amp with it plugged into a wall socket in that room. The PC is plugged into a wall socket in another room.

So basically all three are plugged into separate rooms, each has a separate breaker in the box so they are separate lines aside from the mains supply from the ground into the house.

I still cannot get rid of the slinky effect unless i turn on the scope's built in filter and turn it down to the lowest setting >25kHz

Here is a screen shot of the input at the pin on tube one with the output of the square wave at one PI plate. You can see the input is very fuzzy. the output - a separate question - at a gain of 0-25% has a snake like stairpattern in it...??? Above 30% gain it looks great. Amp sounds great single strings. Some double stops are perfect, some are slightly blatty on the high strings. Phasing? Oscillation?

[surfsup]

Here is the input at stage one vs one power tube grid at the pin...1kHz and 440Hz tone files playing simultaneously to imitate a double stop...the waveform looks fairly smooth, I don't really see oscillation. What's interesting is the fuzziness of the input is gone later down in the amp.

[surfsup]

Ok so I turned my PC off and just plugged a guitar in and ran the scope. The fuzziness is still there. Seems to just be there no matter what. What's interesting though is the amp doesn't have any oscillation or double stop problem with it not plugged-in in that amp room I have so I guess it is problem solved.

I would still like to use this scope more precisely so I need to figure out this fuzz. Any other opinions?