Tweet
I have been helping a friend repair an old Ampeg R12RM This is close enough:
Motorboat at about 2Hz. Spent hours pulling out hair I no longer have. Pulling the PI stopped it. Pulling either power tube changed the freq. FOund that letting pin 2, the input grid - of the phase inverter stopped the boat, but just sat there humming. Terminating that grid by the 470k in the circuit or even grounding that grid allowed it to oscillate. SO all prior stages not involved, even with those tubes removed, still did it.
No DC leakage on any caps, B+ filter caps were new. Even so, paralleling more filter cap with the existing had no effect. Removing NFB had no effect. Resistors all intact, mostly near value.
On a whim, we removed both power tubes, so only one tube in the amp, the phase inverter. Scope showed it still oscillated. The phase inverter oscillated all by itself.
Ultimately, we decided to replace the two resistors dividing to provide signal to the lower triode. The 470k was measuring around 510k, and the 510k was up to about 605k. Normally, I don't sweat resistors being 20% off, most amps work with all manner of variations. But in this amp, the values in that divider needed to be pretty close. Two new resistors on value, and the amp is now working and stable.
Motorboat at about 2Hz. Spent hours pulling out hair I no longer have. Pulling the PI stopped it. Pulling either power tube changed the freq. FOund that letting pin 2, the input grid - of the phase inverter stopped the boat, but just sat there humming. Terminating that grid by the 470k in the circuit or even grounding that grid allowed it to oscillate. SO all prior stages not involved, even with those tubes removed, still did it.
No DC leakage on any caps, B+ filter caps were new. Even so, paralleling more filter cap with the existing had no effect. Removing NFB had no effect. Resistors all intact, mostly near value.
On a whim, we removed both power tubes, so only one tube in the amp, the phase inverter. Scope showed it still oscillated. The phase inverter oscillated all by itself.
Ultimately, we decided to replace the two resistors dividing to provide signal to the lower triode. The 470k was measuring around 510k, and the 510k was up to about 605k. Normally, I don't sweat resistors being 20% off, most amps work with all manner of variations. But in this amp, the values in that divider needed to be pretty close. Two new resistors on value, and the amp is now working and stable.
Comment