Alternate tremolo signal injection?

[potatofarmer]

Has anyone tried injecting a tremolo signal into the (normally) unused second grid of a long-tailed pair phase inverter?

I'm hoping it would result in a bias-modulation-esque sound without the awful (for DC) condition of having an additional 250k between the grid leaks and the bias supply.

So in other words take this:
http://www.lynx.bc.ca/~jc/6l6vibrato.gif

and use, say, a 1M pot for the intensity control, then connect that to an LTP grid through a second large coupling capacitor.

[loudthud]

You need a cap to ground from the unused input and inject the tremolo signal into the tail. That will make the tremolo signal in-phase at the power tube grids so it will cancel in the output transformer. You may have trouble driving the tail without some kind of buffer for the LFO signal.

[potatofarmer]

Thanks! Oh man, was THAT a bonehead mistake.

[potatofarmer]

Well, no; wait. What about the vibrochamp: http://vibrochamp.org/wp-content/upl...-schematic.jpg

I suppose I need to find clips of the vibrochamp to see how this sort of tremolo sounds but it seems cancelling the low frequency isn't absolutely essential. Seems safe to assume it would sound different, but that's not necessarily bad...

And hey, if I keep that grid-grounding capacitor small enough, I can maintain the same gain for the regular input (80 Hz and up) but it'll effectively be out of circuit at the tremolo frequency.

Though I imagine the tremolo speed would effectively be doubled by injecting it into an LTP grid instead of an earlier single-ended stage.

[loudthud]

From a theoretical perspective, a Tremolo should only change the gain of the guitar signal with little or no trace of the LFO in the output. The Fender blackface amps with the tremolo bug are a good example of this. In most lower cost amps where the LFO is injected into a preamp tube, the LFO signal needs to be filtered out. In several Vox amps you will see a high pass network, a number of caps in series with resistors to ground between the caps. The filtering can be subtle, just carefully selecting the coupling caps downstream from the point where the LFO is connected. This is the area where the amp is tweeked to make the tremolo work as good as it needs to be. So when you try to cut and paste a tremolo circuit into an amp, it may take more changes to the rest of the circuit to really optimize the tremolo's operation.

The output stage bias wiggle Tremolo is somewhat unique because the push pull circuit cancels the LFO at the output. But the cancelation isn't perfect and some LFO signal gets impressed on the power supply where it can leak back into the signal path. Injecting the LFO into the phase inverter should be possible and achive similar results.

[potatofarmer]

Quick update: I never got around to trying this, but I did convert my 135W Twin's tremolo to bias modulation by way of a "Type 1" PPIMV - dual-ganged volume pots with coupling caps before and after. I then connected the intensity pot to the ground of the PPIMV.

Basically what they did on the Fender Pawn Shop Excelsior. Not quite getting enough depth (~20v P-P), but I haven't started fiddling with the oscillator triode yet so I'm pretty sure that's not going to be an issue for much longer.

And while I'm at it, the "type 1" master volume is great for volume reduction. I'm sure the LarMar / "Type 2" is great as well, but this way I get to maintain the bias balance control.

Only quirk so far is that the tremolo intensity, of course, is independent of the master volume setting. Definitely a fair trade off for keeping the power tubes' grid leak resistance low though!