Joe, right you are. That 12AX7 needs to be biased near cutoff so that the effect will work properly. Regarding using SS devices to derive the low frequency waveform, there is no reason that shouldn't work well. If you look at the circuit between the oscillator and the mix tube grids, it's low passed like hell. They obviously want a sine wave modulating signal free of higher order harmonics. Exar makes (made??) a nice oscillator IC and as I recall is has square, triangle and sine wave outputs. You could wire that up, and even use op-amps for the phase split function. I would use low passing on those phase split op amps just to make sure any harmonics were in the noise. All you'd have to do then to drive the grids of the mix tube is make sure the amplitudes are right. Shouldn't effect the tone at all. In fact, I suspect that even reverb could be implemented with SS devices and if done properly, it would sound fine.

yeah.. the 8038 right? The one thing you might want to do when going to a SS osc would be just add in an adjustable duty cycle - what I mean is that the tube osc sine wave is asymetrical if memory serves and that may effect the end result. It may not though. The tube osc is LP'd like crazy (all those .1uF caps..) to get some semblance of sine-ness. 
 
Speed:

    What about using IRF820 MOSFETs instead of at least some of the 12A_7 tubes? Mike Donovan had drawn up some circuits using them that is on my schematics page and I was thinking that it would be cool to keep the same basic architecture as the original, but just replace some of the glass with silicone...

Steve Ahola

P.S. Maybe when you get a chance you could check the schematic and tell us what that "obvious" error was... (BTW was it also on other brownface amps using this tremelo circuit?)

P.S.S. Here's another idea- how about a standalone brownface tremelo unit that would plug into a passive fx loop ahead of the PI? Have you looked at the Matchless Vibrobox schematic- I was wondering if there were any similarities...