Swirling Fizzy Tone on Decaying Notes

[syscokid]

Shielded grid wires didn't change anything.

Bypassing the V2 stage reduced the fuzzies a tiny bit, while reducing gain and volume of course.

[nuke]

Do the grounding stuff already mentioned, that's often the most difficult and most effective way to start. Grounding schemes will drive even very experienced engineers crazy at times especially in today's high-speed digital circuits.

In looking at it, I might offer that the blue and brown leads from the output transformer to the plates, should be twisted tightly together if possible and routed well away from the control grid wiring. If the wires run near each other, have them cross at 90-degrees and separate them as much as possible.

[syscokid]

Do the grounding stuff already mentioned, that's often the most difficult and most effective way to start. Grounding schemes will drive even very experienced engineers crazy at times especially in today's high-speed digital circuits.

Yes, I did try the alternative grounding schemes that were suggested, but there was no change in the audio response. I reverted the amp’s grounding scheme back to the original layout.

In looking at it, I might offer that the blue and brown leads from the output transformer to the plates, should be twisted tightly together if possible and routed well away from the control grid wiring. If the wires run near each other, have them cross at 90-degrees and separate them as much as possible.

Thanks, nuke! Yeah, those wires are a bit sloppy. The OT that is in those pics, is installed on a temporary basis. None of the leads have been cut. When I decide on which OT to keep for the amp, I’ll try to do a tidier job with the wiring.

[R.G.]

There is the occasional amp that does not respond to cut-solder-replace techniques. You might be nearing the point where an oscilloscope is necessary.

On another thought, you don't by some chance live near a big radio transmitting tower, have a neighbor that likes to transmit massive CB signals, or use over-the-AC-mains control signals do you?

On yet another thought, star grounding is prescriptive - if it's done properly, the amp grounding system will not be the cause of internal oscillation. It's a bitter pill to swallow, but reworking the entire ground network would completely eliminate grounding as an issue. Huge amount of work at this stage, but it would eliminate questions.

[syscokid]

There is the occasional amp that does not respond to cut-solder-replace techniques. You might be nearing the point where an oscilloscope is necessary.

Nearing the point? I think that I have been tripping over that “point” lately, ha ha! Even before I started this thread, I already was saying to myself that this situation might be a good example for why I need to understand how to setup and use an oscilloscope, and trace the signals too. Procrastination runs deep in my blood.

On another thought, you don't by some chance live near a big radio transmitting tower, have a neighbor that likes to transmit massive CB signals, or use over-the-AC-mains control signals do you?

Nothing like that near me.

On yet another thought, star grounding is prescriptive - if it's done properly, the amp grounding system will not be the cause of internal oscillation. It's a bitter pill to swallow, but reworking the entire ground network would completely eliminate grounding as an issue. Huge amount of work at this stage, but it would eliminate questions.

Amp is very quiet, even at max settings, but I might just give it another try. This time I’ll establish the ground connections on a more permanent basis instead of using alligator-clipped wiring. I have plenty of time! … I hope!

[R.G.]

Nearing the point? I think that I have been tripping over that “point” lately, ha ha! Even before I started this thread, I already was saying to myself that this situation might be a good example for why I need to understand how to setup and use an oscilloscope, and trace the signals too. Procrastination runs deep in my blood.

My blood too!

Understanding how to set up and use a 'scope is not a trivial exercise if you haven't done it before, though, so going through all you can to avoid having to go learn a new set of instrumentation is/was a good idea.

Thinking about how people get help here, I envision it as a group of amp enthusiasts who have a fair amount of experience; from some symptoms and a picture, a schematic, voltmeter readings, they can often guess what's wrong, or guess what to measure to pinpoint it. Tube guitar amps are remarkably alike these days - from simplest to most complicated, they are amazingly similar in how they work, so guessing based on other amps we've seen can be very good in most cases. But sometimes, the problem is rare enough that you have to dig deeper and bring on instruments more complicated than ears, soldering iron and voltmeter.

An o'scope is just a visual voltmeter with a calibrated time scale. But the controls to get the trace to the right size to fit on the screen, the right speed to capture whatever is going on, and specifically the triggering of a trace to catch something funny when it happens can get complicated. But in this instance it would tell you exactly where in the loudness profile of a note the oscillation starts and stops. That's an important clue. I once had an amp that did something similar, and NOTHING told me what was happening, including scoping the signals at every gain stage; then I changed speakers, and the problem vanished. The speaker I was using rubbed a little with small signals, but big signals either stopped it or covered it up. I would have sworn it was crossover distortion. Well, I guess it was - only the speaker was doing it, not the biasing of the output tubes. Scopes can provide negative information too - this-and-such is NOT the problem, for instance.

Another thing a scope can do is tell you when there's things going on that you can't hear - like a continuous oscillation at, say, 100kHz that gets dragged down into the audio range by notes as they decay.

Amp is very quiet, even at max settings, but I might just give it another try. This time I’ll establish the ground connections on a more permanent basis instead of using alligator-clipped wiring. I have plenty of time! … I hope!

Hmmm. I just dragged out another debugging technique that might help. Do you have a signal generator other than your finger or a guitar? Sometimes you can ground the grid of successive stages and inject a signal after the muted stage. In your case, you can't really mute stages because it's a fizz on decaying notes, not silence. For instance, you might mute the stage before the PI, feed a signal into the PI, and listen for fizz with just the PI and power stages running. If this happened (for instance), the issue would almost certainly be local to the PI and power stage, including the feedback from the speaker.

[xtian]

As will all our digital tools, modern oscilloscopes have taken all the troublesome setup issues and reduced it to a single button click. I am loving my inexpensive FNIRSI 1013D scope. Small, lightweight, touch screen, powered by USB-C so easily portable with battery pack:

https://www.amazon.com/FNIRSI-1013D-Plu ... B0BHHYMVKV

[Stevem]

Here’s another check out I used once to help find out that the amp had an oscillating issue I could not hear before I had a scope to confirm that .

Measure what your output tubes are idling at current wise if you can and make note of that.

Also hook up your voltmeter set for ac volts across your speaker output and note how many mv you read, and hopefully your reading only mv and not much higher numbers !

Now unhook the PI coupling caps.

In doing doing this do you see a lowering of the idle current of the outputs?

If so you have an oscillating condition driving the outputs off of true idle .

[syscokid]

Here’s another check out I used once to help find out that the amp had an oscillating issue I could not hear before I had a scope to confirm that .

Measure what your output tubes are idling at current wise if you can and make note of that.

Also hook up your voltmeter set for ac volts across your speaker output and note how many mv you read, and hopefully your reading only mv and not much higher numbers !

Now unhook the PI coupling caps.

In doing doing this do you see a lowering of the idle current of the outputs?

If so you have an oscillating condition driving the outputs off of true idle .

Both volume controls at zero, and tone control 50%.
With PI caps connected, the power tubes are measuring 22.6 mA each. The ac volts across the speaker output are wildly fluctuating back and forth between 3 mV to 10 mV.
With PI caps disconnected at the grid resistors, tubes are measuring 22.4 mA each. And the speaker output ac voltage is a bit more stable but still fluctuating at 2.8 mV to 3.1 mV.

[syscokid]

Hmmm. I just dragged out another debugging technique that might help. Do you have a signal generator other than your finger or a guitar? Sometimes you can ground the grid of successive stages and inject a signal after the muted stage. In your case, you can't really mute stages because it's a fizz on decaying notes, not silence. For instance, you might mute the stage before the PI, feed a signal into the PI, and listen for fizz with just the PI and power stages running. If this happened (for instance), the issue would almost certainly be local to the PI and power stage, including the feedback from the speaker.

I have a tone generator app or software on my PC. I can transfer the signal from the PC's headphone jack to a guitar instrument cable and into the input jacks of the amp and program any frequency. But I don't understand how to "mute the stage before the PI, feed a signal into the PI"

[syscokid]

I have a tone generator app or software on my PC. I can transfer the signal from the PC's headphone jack to a guitar instrument cable and into the input jacks of the amp and program any frequency. But I don't understand how to "mute the stage before the PI, feed a signal into the PI"

Never mind... I think I figured it out how to do it. Disconnected the coupling cap from V2 to the PI. Fed a 300 Hz signal at the grids of the PI. All I heard was pure tone. Maybe I should be using a different frequency?

[R.G.]

Disconnected the coupling cap from V2 to the PI. Fed a 300 Hz signal at the grids of the PI. All I heard was pure tone. Maybe I should be using a different frequency?

Great! Now - can you turn the level of the signal to the PI down, simulating a decaying note? Does that have a swirling fizzy note at some level?

[syscokid]

Disconnected the coupling cap from V2 to the PI. Fed a 300 Hz signal at the grids of the PI. All I heard was pure tone. Maybe I should be using a different frequency?

Great! Now - can you turn the level of the signal to the PI down, simulating a decaying note? Does that have a swirling fizzy note at some level?

I could not detect any of that fizzy stuff. I tried different initial output levels from the PC, and gradually turned the signal down trying to simulate a note dying out.

Not sure if all this is playing tricks with my hearing, and my mind, but it almost sounds like this issue is getting a little bit worse. I don’t remember noticing the fizzies after I completed the initial version of this build a little over three months ago. Definitely started noticing the fizzies about a month ago. Frustration blues… ugh!