Id add couple of diodes and drop the load of the opamp down to 50k and separate opamps supply from output stage.
Its a badass design, it must oscillate.
Its a badass design, it must oscillate.
The operating point is cute - should be about 1/4A standing current, just enough to keep the MOSFETS on. (And therefore stablise the capacitances)
Op amp (and it's ability to drive 2K5 load) will drive stability of the whole thing.
Which op amp did you use?
Bet you it can do 0.01% at 1kHz at rated power. Which might be about 10W into 8 ohms. Should double into 4 too.
(Yeh, I'd like to see gate stoppers too but the 1K resistors are probably doing that job nicely)
Last edited:
"The operating point is cute - should be about 1/4A standing current, just enough to keep the MOSFETS on."
Have you ever looked at the gate voltage vs temperature on IRF FETs?
Vgs threshold drops like a rock with temperature (BANG!).
HTTP 301 This page has been moved
Have you ever looked at the gate voltage vs temperature on IRF FETs?
Vgs threshold drops like a rock with temperature (BANG!).
HTTP 301 This page has been moved
This one is an old design from Elektor (late 80's), with just a few more parts you can easily get 50-60 watts on 4 ohm, it is thermally stable, adjustable bias, with C3 phase margin can be set for different opamps. Years ago I've tried this circuit with OPA 604, and IRF 540-9540, I remember it was surprisingly good (for the cost of it.....)
Attachments
I also tired many of those opamp power amps... They mostly sound wonderful but output power is somewhat limited. Im thinking of building something similar already... But I would use adjustable bias instead. Vbe multiplier would be a good choice and it would add some stability. Im already designing one with IRFP240/9240
- Status
- Not open for further replies.