Bugs in tube / mosfet trem

[phydauex]

Howdy,
Having a few snags trying to implement Tubeswell tube/mosfet trem. See attached layout.

I wanted to try and connect this as a source follower to drive the cathode of the last preamp stage of the amp I'm working on, and when connected as shown (pot to cathode, not bias supply), I get no trem. I can hear the oscillator working, but no audible trem when I run a guitar through the amp. I should note that, the amp is fine (new build) I've hooked up the tube LFO as a bias trem and that worked. I just wanted to try something like the VibroChamp that drove a preamp tube instead of the power tubes/tube. I wanted to get away from the speaker pumping and potential bias issues for whoever might use the amp besides me.
When I lower the 1m resistor off the source to 270k and get rid of the cap , I get some trem out of it, although it's not very strong and quite sharp sounding (staccato). I also lowered the pot from 150k Lin to a 50k RA that I had in my parts drawer and that didn't make any difference. I also noticed that when I had the Fet engaged, it dropped the B+ from about 366v down to 320v. Is this circuit supposed to draw enough current to pull the B+ down that much? I have 360v on the drain, 365v on the gate, I can see a fulcuating voltage on the source (175-265v?), the preamp cathode is set at 1.48v and after the intensity pot (50k RA) I have a varying voltage of 26-45 v on the wiper (I'm pulling that from memory).

Is this trem circuit need a revision to work on the preamp cathode? I don't know alot about FET circuits, but from what I've read it should behave like it's tube counterpart. The oscillator works just fine, and when I hook it up to the bias supply gives a nice power tube wobble. I even hooked up the other half of the same tube as a CF and had a pretty strong trem when used on the bias supply. I was trying to save the other half of the tube if possible and make the amp more fool proof (future bias adjustments, etc).

Any tweaks to this trem circuit to make it work as I intend, or should I just forget about it and keep the bias mod trem and live with the speaker pumping.

Matt

[phydauex]

As a follow-up:
I returned the set up to exactly the schematic posted and wired it to the bias supply of the power tubes and it behaves and sounds fine, like it should. I was just wanting to try something different for the sake of reducing potential bias issues and to free up 1/2 a tube. I guess if it's worked for Fender and others, maybe I'll just leave it be and have a bias mod trem, with a caveat that if the speaker pumping bugs ya, lower the bias a pinch.
Matt

[loudthud]

Trying to wiggle the bias on a preamp tube through that 1 Meg is a problem. If the LFO signal is 100 V pk to pk, that's only 100 uA through a 1 Meg. Most preamp tubes run about 1 mA so that's only a very weak trem. Knock that 1 Meg down to 100K or lower. The MOSFET probably needs to run more current, drop the 220K to 100K or lower. Is this a single ended or push pull amp?

[phydauex]

That's kind of what I guessed. As I dropped the 1M down to 270k, I finally started to hear some wobble.

The amp is based on a Brown Deluxe (6G3A), so push-pull. The 220k from source to ground? I'll give those a try and see what it sounds like. I'd like to get a full, swampy trem, without the issues of burning up tubes because of the over bias that happens on the bias trem, at least from what I've read. Maybe folks that have lived with Princeton's and early Deluxes have had different experiences and it's a moot point.
Thanks for the suggestions! LT

Matt