What would cause this distortion to my sine wave?

[lilstripe]

I've got an older homebrew mono amp that runs an 807 in class A (SE) mode. 
I'm measuring around 380v on plate, 273v on screen (G2), -17v on grid compared to cathode (auto / cathode biased).
I don't have a schematic, guess I could attempt to draw one.

[shooter]

where are we looking? the speaker?

Inductors n Caps will kink up some, bias sometimes gives you some asymmetry 'tween upper n lower lobes. 
if you can trace the audio path, try working back from that pic and find where the kink happens, I have seen that just after the TS into the recovery tube.

[lilstripe]

Into a 8 ohm non-inductive resistor load
The signal coming from the prior stage is a nice clean sine wave, its something relating to the 807 or it's bias I'm guessing.  Just not very familiar with this tube and it's nuances.

where are we looking? the speaker?

Inductors n Caps will kink up some, bias sometimes gives you some asymmetry 'tween upper n lower lobes. 
if you can trace the audio path, try working back from that pic and find where the kink happens, I have seen that just after the TS into the recovery tube.

[PRR]

Welcome.

SE can do many funny things.

Does it come on suddenly or gradually as you increase level?

That kinkiness may not be a "flaw" for guitar.

What current? Load impedance? *Frequency* relative to transformer bandwidth?

[lilstripe]

Does it as soon as you power on and tubes warm up, doesn't change with time.
8 ohm non-inductive dummy load
1khz tone

Welcome.

SE can do many funny things.

Does it come on suddenly or gradually as you increase level?

That kinkiness may not be a "flaw" for guitar.

What current? Load impedance? *Frequency* relative to transformer bandwidth?

[shooter]

8 ohm non-inductive dummy load

your OT is part of that load, it's inductive

doesn't change with time.

what about amplitude of drive signal, like vol 2, vol 10

here's a could waveforms from my last build, PSE 6V6, the one labeled clean is from the plate grid, the other, speaker, notice there is a minor kink like yours, I expected some non linears as I'm using "barely legal" OT

[PRR]

> your OT is part of that load, it's inductive

A wideband audio transformer should be hardly-inductive between its bass-limit and its treble-limit. We like 1KHz because bandwidth is typically >100Hz-5KHz and 1KHz is well away from either end. (400Hz is also useful, but I doubt it will be much different.)

I don't have an SE amp handy and don't remember the many strange things I have seen as they get "close" to clipping. In fact that kink looks like an off-bias Push-Pull amp to my eye; but I accept it will happen in SE. Shooter's plot is incredibly interesting, because it shows a very similar bend.

[jjasilli]

That looks like harmonic distortion.  Lots of info on the web re how to measure it, BUT:


1.  Are you starting with a pure wave AND what does the wave look like in the prior stages of the amp?
2.  The % of THD will vary with signal voltage & output power (as stated in earlier posts).  So you need to measure at various internal signal voltage & output wattages. 
3.  Is 8 ohms the correct load?


Whether the amount of THD is normal or not cannot be determined w/o this info.

[shooter]

that kink looks like an off-bias Push-Pull amp to my eye

that's my 1st thought, and you are correct, I have found some weird waves in SE

here's an FFT of the harmonics, for the clean wave;

[bnwitt]

So is the kink is happening at 0 volts?

https://blog.zzounds.com/2017/03/31/whats-a-class-a-amplifier/

This shouldn't be happening in a single ended amp.

[jjasilli]

So is the kink is happening at 0 volts?

https://blog.zzounds.com/2017/03/31/whats-a-class-a-amplifier/

This shouldn't be happening in a single ended amp.

No, the matched double hump in the zzsounds example is crossover distortion.  In contrast, lilstripe's wave tip is deformed backwards in time; and the sidewalls ea show different amounts of squiggle which I think is harmonic distortion.

[bnwitt]

 lilstripe's wave tip is deformed backwards in time;

What??  Maybe he should check his flux capacitor.  This may be over my head

[shooter]

Maybe he should check his flux capacitor

tried it, didn't work

This may be over my head

your link uses 2 tubes PP to get kinked, we can do it with 1

In my case, I attributed it to using 2 reverb driver trannies configured parallel - series

[jjasilli]

 lilstripe's wave tip is deformed backwards in time;

What??  Maybe he should check his flux capacitor.  This may be over my head

No worries, waveforms are hard to read.  Seemingly minute visual differences can actually be very significant.  Note that the horizontal sweep (x-Axis) is Time.  Einstein proved that time is relative,  though we tend to think of it as absolute.  Anyway, if the wave is bent to the right, it's forward in time, and vice versa.  Phase shift is a backwards/forwards time phenomena, which is screwed with by capacitance (and inductance), as sluckey pointed out earlier.

[tubeswell]

What make and size of OT? Power rating? Is it a proper SE OT or is it a PP Ot that's been 'modded'? (Could this be to do with the air gap in the core?)