Thanks for the compliments. I wondered if that was you I sent that info to. Yes, everything to the right of C6 and up until C2 is the regulator circuit. If you print the schematic and compare it to Nick's, you'll notice it's verbatim his work. I just used different parts to support the through hole design, adjusted a resistor to get higher voltage, and a larger inductor from your work.
Sounds like you are having the same problem I was with that design. In my work with the circuit, the addition of the regulator circuit did not appear to cure this little issue, but I am still waiting to really confirm that (I haven't touched that project since that competition, too many other irons in the fire). I mainly added the regulator so that I could crank up the SMPS voltage (I was able to get 550v out of it no problem), which also raised the switching frequency, and use the regulator circuit to cut the hash off and leave me with steady DC. In thinking about this problem since then, I think the ability for the inductor to source current is too great which has allowed lower (audio range) switching frequencies. I plan to look into this possibility by evaluating inductors with lower inductance and/or higher series resistance. I think I bought down to 10uH or 15uH for this project, and series resistance is pretty easy, I just need to get unlazy, buy some more 1771s and switching MOSFETs, and get to it. This approach, if it's successful, would eliminate the need for the regulator circuit entirely.
D2, 5, 6 and 7 are 100v zeners. D3 and 4 are 12v zeners for gate protection. C9 and C5 are the same type of low ESR filter cap.
In theory, you can also use a voltage divider to set the voltage at the gate, but I suspect that any noise on the rail will also end up present on the gate unless you do adequate filtering in parallel with the gate/ground resistor (it would take the place of the series string of 100v zeners).
Do you have a scope?
