Ugly distortion with tube amp when you play

[g1]

Here's one for the power amp, hopefully it matches:
http://www.drtube.com/schematics/hiwatt/hwpwr100w.gif

[Enzo]

The scope and the audio signal tracer are used similarly in this situation. Inject a clean signal like a sine wave into the input, then follow it stage by stage through the amp. You say your 1kHz signal goes through, but does it sound clean all the way through? Or does it sound like your problem at any point? And scoping, we look at the clean sine input, then l;ook to see if something similar appears at each stage. we are looking for some point where up to it, the signal looks/sounds OK, and after that point it no longer does. That point will be at or near the problem.

[The Dude]

No better time than the present to learn how to use that scope. First and foremost, there are plenty of o'scope primers on the web. A Google search will find you plenty of reading material.

As it relates to this repair, basically you inject a signal and follow it stage to stage looking for an obvious problem (i.e. distortion, loss of signal, etc.). The scope will show you a "picture" of the wave along with accurate measurements of the waveform. Since you're not familiar with the scope, in simplest terms, you'll see a sine wave. At some point, the sine wave will become distorted and it won't look like a sine wave any more, or rather than the wave being larger after a gain stage, it will be smaller. Most of the time you don't need to overanalyze the scope screen too much. You are just looking for something "unexpected".

Most importantly, know the voltage limit of your scope and do not touch the probe to anything above that limit. Aside from that, again, I recommend doing a bit of reading first.

[Ampzone]

No better time than the present to learn how to use that scope. First and foremost, there are plenty of o'scope primers on the web. A Google search will find you plenty of reading material.

As it relates to this repair, basically you inject a signal and follow it stage to stage looking for an obvious problem (i.e. distortion, loss of signal, etc.). The scope will show you a "picture" of the wave along with accurate measurements of the waveform. Since you're not familiar with the scope, in simplest terms, you'll see a sine wave. At some point, the sine wave will become distorted and it won't look like a sine wave any more, or rather than the wave being larger after a gain stage, it will be smaller. Most of the time you don't need to overanalyze the scope screen too much. You are just looking for something "unexpected".

Most importantly, know the voltage limit of your scope and do not touch the probe to anything above that limit. Aside from that, again, I recommend doing a bit of reading first.

This helps, so far the wave looks normal around V1 after that around V2 it starts to look jagged or hacked off so to speak around the 220pf cap around the treble pot. It is not longer an even sine wave. Now maybe this is normal? for this type amp. Now with the audio probe plugged into an external amp I get signal all the way to the grid stoppers on V5-8 pin 5 input. The sound is the 1k tone and not unlike the earlier stages.

[Ampzone]

Here's one for the power amp, hopefully it matches:
http://www.drtube.com/schematics/hiwatt/hwpwr100w.gif

looks to be similar to the one I posted, thanks

[J M Fahey]

This is where the scope really shines.
The audio probe will show you *some* problems, specially gross ones as signal becoming buzzy (probably clipping) , maybe excessive hum or hiss, or plain not reaching some stage (or not leaving it) but the waveform (shape) will tell you lots more.
Of course you must learn to "read" it.

In principle, inject a clean 1kHz 100mV signal at the input, set all tone controls halfway, rise volume slowly, and check signal along the way; you should find a stage where either shape changes radically or signal plain disappears (being 1/10th to 1/100th what expected counts as disappearing too).

I bet it's a build problem, either miswiring (check it against the layout 100 times) , poor/cold/non soldering or , unusual but not impossible, sometimes a part (say, a cap) gets way overheated (clumsy soldering) and it breaks open *inside* the part, it often looks quite normal on the outside, so you have to test for signal on both sides of coupling caps, don't assume that if you have sound on one end, it will be on the other.

Also dirty parts legs, coupled to poor soldering, do not really "take" the solder, which cools and forms a ring around it but there is a thin grime layer isolating solder from wire.

EDIT: ask the builder for the layout HE used, and post it here, maybe he followed it faithfully but the layout itself is bad.

[eschertron]

AFAIK no it did not, voltages look good, bias is good, that much I can see with the meter. I have a tektronix scope which is a new tool that I have yet to master, e.g translating what I seeing in to a diagnosis. The audio probe and tone generator are on hand and show signal 1k getting through that I can confirm.


Thinking out loud what could cut the signal level at V1 input grid by 2/3rds? Normally when you touch this with a non conductive item such a chopstick it get really noisy with hum and the sound of your contact. On this amp it's barley noticeable when you make contact.

I've been reading with interest; I hope you get this sorted out, and gain knowledge along the way!

Can you post the voltages that you read on the plates and grids of the EL34s? Just another log on the fire, but my hunch is that the noise you hear is crossover distortion from extremely cold bias on the output section.

[TigerAmps]

In addition to checking the wiring and components, I once wired the primary of the output transformer wrong and it caused all sorts of distortion and grief.
After checking it over and over, I tried a different transformer and the amp works, so I went back and realized I mixed up the color codes.
In my case it was a Marshall 50 watt, but I was looking at the 100 watt colorings. Also, had a Peavey Old Butcher that the jacks were causing
shorts across one or more of the taps, which can yield weird and distorted signal at the output stage ...
it's a long shot but though I would throw it into the mix as well. The jack issue also had the output tubes drawing a LOT of current with signal applied.
Might want to check the idle current of the output tubes just to see if there is another symptom/issue.

[Ampzone]

I've been reading with interest; I hope you get this sorted out, and gain knowledge along the way!

Can you post the voltages that you read on the plates and grids of the EL34s? Just another log on the fire, but my hunch is that the noise you hear is crossover distortion from extremely cold bias on the output section.

Thanks for chiming in 460 plates grid appx 458/457. Draw is appx 32-35mv per tube, -44v on the input grids. Heaters 3.15v AC.

[Ampzone]

Big help here. I'm getting it setup for another scope check, and thanks for everyone's patience, getting into scope world is new to me. Baptism by fire right! Well, lets leave the fire out.

[Ampzone]

In addition to checking the wiring and components, I once wired the primary of the output transformer wrong and it caused all sorts of distortion and grief.
After checking it over and over, I tried a different transformer and the amp works, so I went back and realized I mixed up the color codes.
In my case it was a Marshall 50 watt, but I was looking at the 100 watt colorings. Also, had a Peavey Old Butcher that the jacks were causing
shorts across one or more of the taps, which can yield weird and distorted signal at the output stage ...
it's a long shot but though I would throw it into the mix as well. The jack issue also had the output tubes drawing a LOT of current with signal applied.
Might want to check the idle current of the output tubes just to see if there is another symptom/issue.

Thought of this and switched the primary's around, it was horn-siren city!!

[TigerAmps]

In my case the CT was on one of the tube plates ... and unfortunately it did not howl. It just sounded like distorted dodo ...

[eschertron]

Baptism by fire right! Well, lets leave the fire out.

Right.

Heaters 3.15v AC.

Either that's a misquote or a mistake. The El34s should take the whole 6.3vac across the heaters. I'm going the check the schem for that. You should too.

edit: well I'm sure you have the 6.3vac, since otherwise the tubes wouldn't be conducting.

[TigerAmps]

Yes EL34 should have 6.3 on the heaters - not sure if they would even glow at 3.15 Volts ...

[eschertron]

Thanks for chiming in 460 plates grid appx 458/457. Draw is appx 32-35mv per tube, -44v on the input grids. Heaters 3.15v AC.

Wait. You measured (as opposed to calculated) the 32-35ma per tube, right?