Quote from: drdn0 on May 06, 2025, 10:00:53 PMWell, I'll be.

Looked at it with fresh eyes today, and out of interest removed Ck from the first triode - instantly, there was about a 50% reduction in oscillation.

I then changed the 1k5 to 1k8 (for anything tubeish I've always used 1k8) just to see what would happen.

Dead quiet.


Tested half a dozen different tubes, and not a single skerric of oscillation. Out of interested I changed back to 1k5, and every tube oscillated again.

I'll blame the ghosts.

Removing the cap and up'ing the resistor both reduce the gain.

If you reduce the gain you will reach a point where oscillations cannot be maintained.

(Your efforts didn't go to waste because you removed quite a bit of noise in the debug process.)

Quote from: Rob Strand on May 06, 2025, 11:08:07 PM

Quote from: drdn0 on May 06, 2025, 10:00:53 PMWell, I'll be.

Looked at it with fresh eyes today, and out of interest removed Ck from the first triode - instantly, there was about a 50% reduction in oscillation.

I then changed the 1k5 to 1k8 (for anything tubeish I've always used 1k8) just to see what would happen.

Dead quiet.


Tested half a dozen different tubes, and not a single skerric of oscillation. Out of interested I changed back to 1k5, and every tube oscillated again.

I'll blame the ghosts.

Removing the cap and up'ing the resistor both reduce the gain.

If you reduce the gain you will reach a point where oscillations cannot be maintained.

(Your efforts didn't go to waste because you removed quite a bit of noise in the debug process.)

I understand how it works mechanistically, just wild to me that a voltage gain reduction from 32 to 29 with the changing of Rk (given everything else is already fairly conservative and relatively...normal) could be the difference between stability and obscene amounts of oscillation - hence ghosts 

Quote from: drdn0 on May 07, 2025, 12:49:13 AMwild to me that a voltage gain reduction

Wild to me too. Most tube-work is +/-20% makes nearly no difference.