Need Explanation Of This Tremelo Circuit of 1956 Premier 120 Amp

[BrianC]

I'm working on a 1956 Premier 120 tube guitar amp. I've seen many tremolo circuits, but this one I can't quite get what it does in the circuit to produce the trem. I have it working and it pulses the speaker cone at the speed the trem is operating at...which is kind of odd, although I've heard of this before. I changed all the cap and no difference. Just curious if someone knowledgeable of tube circuits can give a outlilne of what's going on in this tremolo circuit to produce the effect, and why the speaker pulses with the signal. Unfortunately there are only re-draws of the circuit and there may be mistakes or circuit changes from year to year. Here's the best circuit renditions I could find....
https://www.jt30.com/premier/model120schematic.pdf
https://www.jt30.com/premier/Premier_120Schematic.gif

[R.G.]

Looks like one of the output tube's cathode is connected to the speaker. That would do it I think.

[BrianC]

That's a mistake on one of the schematics I posted..the other schematic doesn't show it. I found another schematic and that cathode connection to the speaker isn't there either. I still would like to have a tech explanation of how that trem circuit works, as it's different than most.
https://www.jt30.com/premier/premier120schematic3.gif

[trobbins]

A quick look shows the LFO stage is modulating the B+ supply to V2, which has both triodes in cathode follower config, as well as modulating the voltage at F. The V3 triode connected to V2 circuitry is providing the drive signal to paraphase PI stage but mixed with Ch 1 via Ch 1 vol. V2 first triode is fed from Ch.2 controls, and its cathode signal is driving V3 triode, with interaction from V2 second triode cathode voltage which will be modulated from F and from anode B+. It looks like the V3 B+ supply is 5Hz low pass filtered, to suppress LFO harmonics and signal frequencies that would come in from B+ supply to V2.

The output stage coupling has a 17Hz high pass filter action, which is going to let through a lot of the LFO. The global feedback is unlikely to supress the LFO, as the output transformer will roll-off feedback at that low frequency, and there may well be a hump in low bass response that accentuates the LFO beating. You would need to be careful if wanting to change any PI/output stage circuitry as it may cause problems due to the feedback loop.

To suppress the LFO beating, you could step down the PI/output stage gain at LF by using a shelf network for the 20nF coupling caps. You could also add in some CR high pass filter sections before the PI input.

[BrianC]

Thanks!..that's good info.