Mullard 7 watt clone with oscillation problems

[Fazer]

Yes, I noticed the phase shift too. But I have to rule out operator and strumentation errors first. I plan to swap scope and signal generator channels, to confirm the shift is in the amp.

If so, any hints to what I should check to find the cause?
Anyway, first We have to make the amp stable with the feedback in. Then We'll address the other issues.

Thanks for the help,
Giampaolo.

[nickb]

Just go thru stage by stage comparing L with R until you see the difference.

[Helmholtz]

Just go thru stage by stage comparing L with R until you see the difference.

Maybe not necessary, but I would compare left and right square wave responses at corresponding stages.

[loudthud]

If there is any question about the scope or probes, connect both channels (via probes) to the same signal and compare the waveforms.

[Baronkarza]

Hi,
at last the phase shift was due to operator error : an uncalibrated probe.
We still have a gain difference (To check later)

https://1drv.ms/u/s!AlelTeAIElpOhrwXa50TE8f3mA81tQ

I've separated the ground. Much less hum, still some.
The feedback is still not connected, so maybe the residual hum is due to the higher gain.
We made a brief music test with disposable speakers and a CD player. No strange noises or distorted sound.

Next week We will reconnect the feedback and hope for the best :-)

That's all for now

Emiliano

[Chuck H]

You could check the max power at the onset of clipping. If it's the same for both channels I might suspect that the imbalance is in the grid drive.

[Baronkarza]

IT'S ALIVE !

Hi everybody,

So:

1) We reconnected the feedback. Both channel are now at the same amplitude, with no phase shift.



2) Checked the stability with a squarewaves at 100Hz, 1kHz, 10kHz. the lower ones seems good, but the 10kHz was too much rounded off.
3) So, We had changed the feedback capacitor from 600 pF to 350pF, shifting the cut off frequency.
4) We rechecked for stability. No issues. The square at 10kHz is better:



I don't have commercial tube amps to compare, but the 10kHz square seems in the ballpark to me, with a -3dB around 25kHz. My other builds are the same or slightly worse (but with worse OTs). Now we have a (calculated) -1.6 dB at 20Hz and circa -2 dB at 20kHz.
5) With a sinewave signal of 1Vpp, we have a max ouput before visible round off of about 19,55 Vpp resulting in a max power output of 7W on a resistive load of 6.8 ohm, spot on target.

It's sound quite good, with a pair of Klipsch R-15M and an Onkyo C7030 CD player.

Thanks to all for the advices, expecially to nickb for the hint on the primary wires issue.

Regards,
Emiliano
(and Giampaolo)