Strange oscillation?

[Chuck H]

I've been a weekend warrior on this one. Did a ton of checking and testing last weekend and was unable to locate an oscillation loop. No leaky caps either. I did find out that the problem actually starts at the treble cap, where I hadn't detected it before. The visible distortion under the same test is the same, but less severe. I think this is the root cause and it's simply being amplified by the following triode. Sooo...

I eliminated the tone stack and ran the treble cap straight to the volume control and the distortion is still visible. Still suspecting a potential loop of some kind I grounded the following grid and tested straight from the 330pf cap and the distortion is there. With only the first preamp triode in operation and the rest of the circuits eliminated from the signal chain. I tried two other caps, another silver mica and a ceramic and the distortion is still there. Sooo...

I tried a larger, .01u film cap and the problem went away. The problem only manifests with a smaller value capacitor. And it only manifests with increase or decrease in LF at the input. I think that changes in LF are creating large enough currents in the small value capacitor that any change from steady state are causing the distortion during charge/discharge rather than oscillation. Why this should be so profoundly audible in THIS amp I can't say. It's a very common circuit. As I said, I've heard a little of this phenomenon before in other amps and was never able to troubleshoot it before. It's bad enough in this amp that I have to or it won't be acceptable. My next move will be to open up a different amp with a similar circuit and look for the phenomenon. I'm not sure what I'll do if this turns out to be "normal" behavior. More to come.

[Chuck H]

Well I just opened up another amp with similar enough voltages and preamp topography and the same tone stack. The same exact thing is happening but there is no audible distortion artifacts. I always wondered why some amps crackle a little on low notes and I always found it annoying as hell. I still don't exactly know, but I know at least one of the elements. The amp I'm currently working on will just need to get a preamp update that avoids channeling that early stage through a small value capacitor. A pita, but not a problem.

[nickb]

It sounds like that first stage is oscillating.

It reads to me that you replaced the 330pF treble cap with a 0.01uF. Did you mean something different since that doesn't quite make sense to me due the effect on the treble control operation. The rest of this is predicated on this so I'll plough on anyway.

A small plate coupling cap means more gain at lower frequencies, not what you would normally associate with instability. Did you try moving the grid stopper up tp the grid? What is the spatial relationship between the wires for the tone stack and the input (grid) wiring? You could lower the gain with a smaller plate resistor ( AC coupled if you need the headroom) and see what that does.

I suspect the reason it shows up on LF it that it's only just unstable and so takes a little time to build up oscillatory amplitude. If the instability is dependant on the operating point, as is often the case, then the LF waveform goes through the critical point slowly enough for the oscillation to get established.

[Chuck H]

Thanks Nick. I lifted the plate end of the 100k slope resistor and ran the treble cap straight to the volume control. So it was now effectively a coupling cap.

I moved the grid stopper to the grid of the input triode.

IMHE there is nothing about the layout or grounding to oscillate. Further, I checked another amp with the same preamp topography and the distortion was visibly present but not audible.

"A small plate coupling cap means more gain at lower frequencies,.." I'm not sure how to interpret this.?.

The distortion occurs when the LF input is increased OR decreased. This behavior was also present in the control sample amp I opened. Hmmm... I've never seen a decrease in gain trigger an oscillation. I'm pretty sure this is an artifact of current changes within the cap at frequencies to which it presents a high impedance. There is no distortion present at any time during steady state at any reasonable frequency even at elevated input signal levels.

I've heard this crackling on low notes before on several amps. This amp happens to be especially bad and so must be corrected for. This may be because of it's simplicity. There are fewer stages, decoupling circuits, phase anomalies, etc. and this 'small value cap charging/discharging artifact' is not swamped in subsequent circuits? I don't easily give up on problems, in this case though I need to make an exception. I've troubleshot it as far as I can and reduced it to a single gain stage and, further, observed it as common behavior in other amps, though not usually with audible consequences. At this point I have the option of trying to find out why all amps do this, but only some amps suffer OR I can change the circuit in THIS amp to one that avoids the problem entirely.

I'm not being curt. I've just been consumed with this anomalous issue in my peripheral thought for a couple of weeks now.

[nickb]

Thanks Nick. I lifted the plate end of the 100k slope resistor and ran the treble cap straight to the volume control. So it was now effectively a coupling cap.

So is all the wiring to the tone stack now isolated?

I moved the grid stopper to the grid of the input triode.

That is a good thing but I wasn't holding my breath.

IMHE there is nothing about the layout or grounding to oscillate. Further, I checked another amp with the same preamp topography and the distortion was visibly present but not audible.

I would move it around to see what happens. Can you take a piccie?

"A small plate coupling cap means more gain at lower frequencies,.." I'm not sure how to interpret this.?.

The stage driving the cap will have a little more gain. It's being loaded by whatever the cap is driving at HF so the gain will go down. Not by much, but it does.

The distortion occurs when the LF input is increased OR decreased. This behavior was also present in the control sample amp I opened. Hmmm... I've never seen a decrease in gain trigger an oscillation. I'm pretty sure this is an artifact of current changes within the cap at frequencies to which it presents a high impedance. There is no distortion present at any time during steady state at any reasonable frequency even at elevated input signal levels.

I've heard this crackling on low notes before on several amps. This amp happens to be especially bad and so must be corrected for. This may be because of it's simplicity. There are fewer stages, decoupling circuits, phase anomalies, etc. and this 'small value cap charging/discharging artifact' is not swamped in subsequent circuits? I don't easily give up on problems, in this case though I need to make an exception. I've troubleshot it as far as I can and reduced it to a single gain stage and, further, observed it as common behavior in other amps, though not usually with audible consequences. At this point I have the option of trying to find out why all amps do this, but only some amps suffer OR I can change the circuit in THIS amp to one that avoids the problem entirely.

I'm not being curt. I've just been consumed with this anomalous issue in my peripheral thought for a couple of weeks now.

It certainly can get to you. So, do I need to post more kitten pictures?

[Chuck H]

Ok, for any who are following...

First, sorry I didn't come through with pics. It really would have required a video to capture and since I'm a techno tard (thanks Justin) I don't have current video capability, software for uploading it or knowledge of how and no membership at any sites that I could upload it to. I just didn't want to go to the trouble of putting it together. It's just new stuff I'm not into. Plus I was too busy telling those damn kids to "Get offa my lawn!"

Did some more testing, circuit changes, observation, etc. and determined that the problem wasn't manifesting exactly as I thought. The distortion at small value bypass caps IS normal, but it looks more wavy and is only phase distortions. I was looking soooo close that I thought this phase distortion might be the culprit. But there is a more static, sharper distortion that appears analogous to the phase distortions with changes in LF amplitude. Reading signal off the board itself demonstrates the static distortion only. I'm theorizing that the millivolts of DC floating on the board must be conducting it into the signal chain. Even though I floated half the damned board I'm not convinced it isn't the problem. I scraped off about a square half inch of the board until I was down to resin powder and it reduced reading by about half. Not good enough, the board has to go. Maybe next weekend. I'll post back results.

[eschertron]

Ok, for any who are following...

Yes, we are playing at home. Thanks for the update. I've read a few threads about conductive fiber boards. Is this another example of that phenomenon?

[Chuck H]

Yes, we are playing at home. Thanks for the update. I've read a few threads about conductive fiber boards. Is this another example of that phenomenon?

I sure think so since there's nothing wrong with the amp, but it doesn't work right We'll see. The problem never presents the same way twice because the resistive paths are random. Some people think it's caused by moisture and the boards can be dried out. Maybe sometimes. Not always IMHE. Hardly a solution in the PNW where the humidity is always high. I've also seen it happen that conductive pathways actually form permanently in the carbon the boards are dyed with. If this turns out to be the case this will make five problematic black paper boards out of seven or eight I've dealt with. Not a real good record. I know a lot of techs here haven't had the same ratio of problems with these things.

[g1]

Assume that the extra waveform artifact was not present in the other amps that did not have the audible glitch?

[Chuck H]

Assume that the extra waveform artifact was not present in the other amps that did not have the audible glitch?

Dunno. This is the first one I've tackled with a scope. I've only had a scope since about 2007 (2008?). In fact it was the scope that confused this case. Without it I'd have just stuck my DMM probes into the board and just defaulted to where I am now.

[g1]

Ok. I wasn't sure when you said another amp had the same thing:

Well I just opened up another amp with similar enough voltages and preamp topography and the same tone stack. The same exact thing is happening but there is no audible distortion artifacts.

[Chuck H]

Oh! I thought you were asking about other amps with the conductive, POS, awful, "I wish they were all at the bottom of the ocean, black fiber board.

The control amp that I referenced earlier is one of mine. It has the same preamp, basically. And yes, there is an extra wave form artifact on that amp. But it's more wavy. From what I've read it's phase distortion. You could probably see it for yourself if you have a BFish topography amp around. Get tight on the signal at the coupling cap of the second triode, turn up the treble, turn down the bass and mid. Then put a low frequency signal into the input and turn it up and down rapidly at the signal generator. As the treble cap charges and discharges with the relatively large LF signal the time constant isn't nearly fast enough and phase errors occur. I think. See for yourself. Something is certainly going on.

So that WAS happening with the control sample amp. But not the artifact with the more irregular, more triangular static sort of wave form. Both artifacts do appear to ride the sine wave, so I was thrown off.

[Justin Thomas]

It really would have required a video to capture and since I'm a techno tard (thanks Justin) I don't have current video capability, software for uploading it or knowledge of how and no membership at any sites that I could upload it to

No problem!

Justin

[Chuck H]

FINALE

Interesting observations. The problem continued to exist even after changing the board!?! Though it was diminished some. Through the speaker it can be heard as a light static sound when LF amplitude at the input is increased or decreased. What the new board seems to have done is minimize this anomaly to the point of insignificant when actually playing a guitar through the amp. The low note crackle is gone. Since I've been over the circuit repeatedly, actually rebuilt it a couple of times in this process, idealizing grounds, moving grid stoppers to actual grid pins, etc. I'm certain there's nothing miswired. The layout is good and I can't identify any parasitic problems at all. I plugged the "control amp" back in and while I CAN see the same phenomenon on the scope (to a slightly lesser degree) it isn't audible in the OT "sing" or at the speaker. Still, I've done all I know to do, checked all I know to check and rebuilt everything there is to rebuild. At this point it runs real quiet, sounds great both clean and overdriven and I'm calling this done.

Here's the final schematic after voicing. The zeners don't kick until after the power tube is clipping. Their purpose is to minimize asymmetry and improve clipping. They do. The 1500p cap is pinching off a little voltage spike. Ever notice that every damn Champ sounds like a hashy mess overdriven, even though they're said not to??? Well, this one doesn't sound like that. The tone from this amp is really good and it moves gracefully into heavy clipping now. Sounds more like a baby Deluxe than a Champ. Puts out 8 watts. Pretty sweet.