Frequency doubling

[Helmholtz]

The scope signal doesn't reveal anything unusual.
It's basically a square wave with treble emphasis.
The square wave shape is due to heavy clipping on both sides, mainly caused by the DCCF. (If you scope the CF output (AC coupling) you should see the square wave without peaks.)
The overshoot peak should be variable with the treble control.

When analysing an amp I prefer to start with a clean setting, allowing me to see a sinewave throughout the preamp.

[Helmholtz]

I just noticed, is that actually a shared 10k cathode resistor at the power tubes??
I would have expected something like 100R here.

[loudthud]

Or maybe just select the used channel as trigger source?

It's a simple audio signal. Shouldn't be hard to trigger.

My thinking is that if there is some kind of frequency doubling, it shouldn't be on the trigger output of the generator and you can instantly see it when you have a rock stable trigger. If there is noise or a parasitic oscillation on the audio signal, it can be a challenge to find a stable triggering point.

[Helmholtz]

My thinking is that if there is some kind of frequency doubling, it shouldn't be on the trigger output of the generator and you can instantly see it when you have a rock stable trigger. If there is noise or a parasitic oscillation on the audio signal, it can be a challenge to find a stable triggering point.

Agreed.
What I meant was that an untrained scope user might forget to select the active scope channel as trigger source.
Using scope channel 1 with trigger source channel 2 won't work.
With noise or HF oscillation, HF reject mode typically helps.

[g1]

I just noticed, is that actually a shared 10k cathode resistor at the power tubes??
I would have expected something like 100R here.

I see that on the schematic, which does not make sense (the value 10K). But in the chassis picture, it looks like the 1R's are going to chassis ground.

[Helmholtz]

I see that on the schematic, which does not make sense (the value 10K). But in the chassis picture, it looks like the 1R's are going to chassis ground.

THat would be a fatal error.
With zero grid bias the amp depends on cathode bias.

Also the NFB/presence wiring in the schematic looks wrong.

[loudthud]

I thought I saw that the 1 Ohm resistors went to a 100 Ohm 2 or 3 Watt to ground.

[35L6]

100 Ohm 3 Watt . The schematic has an error , Don't know I was thinking . I did have a sketch . I just wanted to get something built so I could test it . As Tolkein said " the tale grew in the telling " .I as using a high signal because I couldn't get it to oscillate . I usually adjust the time base until I get the best sine wave but when I saw the number of peaks double I thought " that's it ! " not thinking it couldn't be right . It puts out about 3.5 W @ onset of clipping . Only has 152V on the plates . 50L6 has 200V max . Spec sheet says 4.3 W @10% with 200 V . I scoped it the other day with 3kHz and 4kHz but still couldn't get it to oscillate . Getting back to the layout I have changed parts and added parts and I just don't have enough tie points . Going to make another with 3 terminal strips for the preamp .

[35L6]

Helmholtz , your right , I left out a cap !

[J M Fahey]

Now, scoping a guitar signal is a different thing and means quite a challenge for the trigger, as the signal contains a number of frequencies with changing phase and amplitude.
In other words, a guitar signal is not really periodic, rather changing constantly.
Scoping something like that often requires a storage scope (DSO).

Preamp distortion (especially a cold clipper) indeed ADDS the doubled frequency as well as higher harmonics.
Also with more than one note sounding at the same time (or with power supply ripple) - you get intermodulation distortion (IM), adding difference and sum frequencies.

As long as there's still a fundamental, I woudn't speak of frequency doubling even though the distorted signal contains some doubled frequency.
But if your system (including speaker) has some strong treble resonance, higher notes might sound like frequency doubled indeed.

That.

I scope Guitar signals every day, thatīs what I build or service, and am quite used to waveforms jumping all over the display.

Not owning a storage scope, I got used to picking and remembering one out of many waveforms jumping around, but also setting sweep frequency slow, so the general envelope triggers it, the fundamental frequency if you wish, and thatīsnreasonably stable.

Now, trying to see an individual waveform? .... forget it.

[Helmholtz]

The schematic has an error.

I think more than one. Check the OT secondary grounding and NFB connection.
And yes, the NFB coupling cap to right side PI grid is missing.

Spec sheet says 4.3 W @10% with 200 V .

That's for a single tube (SE class A).

[35L6]

Helmholtz , yes it's time to draw a new schematic . There are changes to component values and possibly mores errors . Juan , you inspired me to start scoping guitar signal . One day they might make sense . Loudthud , this project was meant to be an indoor amp . 3.5 W is plenty loud for my living room but I have wondered how much power I could get from it . Would 10 W require fixed bias ? Isn't there a simple formula to estimate the theoretical limit for power ?

[Helmholtz]

Datasheet suggests 2.5W to 3W with your B+.
That is for single tube class A operation.

Your circuit is class A PP, so output should be doubled, meaning 5W to 6W.

With fixed bias class AB, around 10W should be possible with a 4k primary, new tubes and ignoring OT losses and B+ sag.