More fun with scope image interpretation

[Tom Phillips]

We are assuming that there is no input signal applied when you made the measurements "at idle." The concern is the much higher than normal AC ripple on the screens and grids. Unless the AC frequency is ultrasonic I would expect the amp to have a loud hum at idle. Especially with 1.5VAC on the grid.

[Chuck H]

Input signal voltage is about 300mV pk-pk for 1KHz and 100Hz.

I usually run 100mV (200 pk pk). Some guitar amps get tender above that for bench testing. Probably not the issue with that amp though.

I'll try my other signal gen tonight when I get back to it too just in case. I don't htink it's that though, since this big modulation is only becoming visible in the output section.

That does make it seem moot. I wouldn't bother. Besides, the sig gen your using and the scope are right there. You can just scope the input of the amp and see what the sig gen is doing.

Looking at the vids I see what looks like signal doubling. It's the paired and warbling lines at mid frequencies and the traveling node along the waveform at lower frequencies. Like the beating of two notes that are out of tune causing a ring modulator effect. I can't figure how the power amp would do this by itself. Hum was mentioned. This doubled signal could be a sum/difference overtone from excessive ripple in the signal chain.

[Tom Phillips]

[QUOTE=Chuck H;396858]

...Looking at the vids I see what looks like signal doubling. It's the paired and warbling lines at mid frequencies and the traveling node along the waveform at lower frequencies. Like the beating of two notes that are out of tune causing a ring modulator effect. I can't figure how the power amp would do this by itself. Hum was mentioned. This doubled signal could be a sum/difference overtone from excessive ripple in the signal chain.

I think that some of the effects could be caused by the scope trigger setup. Try using the external scope trigger hooked to the sig gen output. That might eliminate some of the jumpiness you see. Also try longer sweep times in addition to those you showed before so we can see more duration of the output signal.
Cheers,
Tom

[nsubulysses]

It's all good y'all. thank you for your input. As usual many people were right on.

I traced AC around and replaced C8, D13 and D11. Slightly shotgun but it's kind of a pain to pull the tubes and pull the board over and over. I have a very stable signal now and fractions of a VAC on grids. Have never listened to this amp through a speaker yet but I bet when I do tomorrow it will sound alright.

VAC RMS is misleading because I had it plugged into 8 ohm load at the time and the head only goes as high as 4 ohms.

Sorry this new scope is such an eyesore compared to the old analog ones but still fun to post vids and hear the old masters comment

[Tom Phillips]

Glad to hear the progress.
The video in post #19 won't play. It brings up a "Please sign in to view this video." None of the other videos did that.

[nsubulysses]

I think it's because I forgot to click Publish. Should be good now.

[J M Fahey]

It's all good y'all. thank you for your input. As usual many people were right on.

I traced AC around and replaced C8, D13 and D11. Slightly shotgun but it's kind of a pain to pull the tubes and pull the board over and over. I have a very stable signal now and fractions of a VAC on grids. Have never listened to this amp through a speaker yet but I bet when I do tomorrow it will sound alright.

VAC RMS is misleading because I had it plugged into 8 ohm load at the time and the head only goes as high as 4 ohms.

Sorry this new scope is such an eyesore compared to the old analog ones but still fun to post vids and hear the old masters comment

Good!! That's what's expected out of a regular amp clipping
JMF

[Chuck H]

What is the shake on the clipped portion of the wave about? I've never seen that before. Could it be an artifact of another problem that was exacerbating the issue you already repaired?

[nsubulysses]

Chuck that is power supply ripple. At full power output the signal gets a little modulated because the power supply is being taxed to the max

Check this -- http://music-electronics-forum.com/t37757/

[Chuck H]

Chuck that is power supply ripple. At full power output the signal gets a little modulated because the power supply is being taxed to the max

Check this -- http://music-electronics-forum.com/t37757/

Hmmmm.?. Ok. Everyone seems to agree that it's normal. I just wonder what I may be doing wrong then since I haven't seen it before.

[Tom Phillips]

I'll bet that if you slow down the sweep speed to show many more cycles of the output waveform, then you will clearly see the 120Hz ripple.

[J M Fahey]

Yes.
And the other way, which happened here and in the thread posted by nsubulysses, is to have a test frequency which is a multiple of ripple frequency.

The wiggle is the beat (difference in this case) between both, or between test signal and ripple exact multiple.

It happens all the time when clipping a power amp, because there you *always* have important ripple , not the case in a preamp, even including the PI in some cases, because those few mA are very easy to filter, while power amps need from half to a full Amp if tube and up to tens of Amperes in SS .

Now you can't physically see wiggling above, say, 15/20 Hz , eyes are slow (thanks God or we wouldn't have Cinema,TV, etc.) so in practice you must be within, say, 5Hz to see it.

In nsubulysses' thread test frequency was nominal 600Hz, varying between 599 and 601, while line frequency is either 50/60Hz, ripple 100/120Hz.

Probably by chance he chose "600Hz" which is an exact multiple of all of 50/60/100/120 Hz ; if he had tested with, say, 440Hz wiggle would still have existed, but impossible to see.
In this case wiggle frequency was 1 or 2 Hz, a fast tremolo (literally) but still visible.

In any other case, ripple is visible, specially a thigh frequencies, if you slow scope sweep down long enough to see the ripple peaks themselves, riding the test signal, as Tom Philips suggested.

[loudthud]

Most of the more expensive analog oscilloscopes have a control called Holdoff. This adjusts the time between the end of a horizontal sweep and when the trigger is enabled for the next sweep. If it is adjusted while viewing a 600Hz waveform, a place can be found where the waveform will sync up with the line frequency and the waveform will start wiggling up and down showing the ripple. If the 600Hz is exactly in sync with the line frequency, the wiggling will stop and you will be able to see where each cycle is clipped slightly different.

[nickb]

Most of the more expensive analog oscilloscopes have a control called Holdoff. This adjusts the time between the end of a horizontal sweep and when the trigger is enabled for the next sweep. If it is adjusted while viewing a 600Hz waveform, a place can be found where the waveform will sync up with the line frequency and the waveform will start wiggling up and down showing the ripple. If the 600Hz is exactly in sync with the line frequency, the wiggling will stop and you will be able to see where each cycle is clipped slightly different.

And in the same vein, many scopes have a 'line sync' selection where they trigger off the mains. In that case the line frequency modulation becomes steady and the signal ripples through.

[nashvillebill]

My scope has both variable holdoff and line sync, I guess I never learned how to use either feature to their full advantage!