Sometimes it pisses me off when I see schematics like this published on the web as tested, functional circuits.(There is even a photo of the "prototype") Why makes this someone happy remains a mystery to me.... and this is not the only one, sadly. No I didn't build it (fortunately), but if someone does will wonder why it does not work.
Attachments
The DC feedback via the 15k will push up the opamp output voltage until the top FET both conducts the bottom FET's constant(ish) current and sits the output at 0v to match the +in pin. I.e the opamp output pin won't be at 0v but a few volts higher.
Looks OK to me apart from the obvious explosion risk (Lol, risk !... its a certainty)
I tried some class A mosfet combination, which was almost same, like this. The biggest problem was, that increasing the bias current, results much better sound. So I made some nice fried mosfets, even with fan cooling
Sajti
In the meantime.... being curious about it, I knocked up the circuit on a veroboard with whatever I had in my spares box (IRF540-s, then IRFP150-s and OPA134) supplies +/- 16V, double checked everything, aaaaaand...
Nothing , no workee, opamp inputs, output at 0V, Idle current zip, this thing is dead as a dodo. Still, maybe I am doing something wrong (need other type ((non fet input)) opamp, different FET-s ??
Nothing , no workee, opamp inputs, output at 0V, Idle current zip, this thing is dead as a dodo. Still, maybe I am doing something wrong (need other type ((non fet input)) opamp, different FET-s ??
The type (JFET, or bipolar input) of the opamp is not important. Maybe it change the offset, with few mV-s.
There must be another problem. Once Your ccs is working, the output of the opamp must be somewhere +2-4V.
Sajti
Hi Guys
dragonweed - the mosfet types make no difference to the circuit operation. The current source won't conduct if the upper mosfet is not on.
The opamp type won't matter too much as far as having functionality. Do you have power to the opamp? most schematics do not show that.
The mosfets should also have gate protection zeners. The CS only needs a single 12V with cathode to the gate and anode to the source. The upper mosfet should have back-to-back zeners. These do not effect operation but protect the gates from ES discharge.
If you pull the opamp and ground the gate of the upper mosfet, both mosfets should warm up instantly as there will be a path through both.
Have fun
dragonweed - the mosfet types make no difference to the circuit operation. The current source won't conduct if the upper mosfet is not on.
The opamp type won't matter too much as far as having functionality. Do you have power to the opamp? most schematics do not show that.
The mosfets should also have gate protection zeners. The CS only needs a single 12V with cathode to the gate and anode to the source. The upper mosfet should have back-to-back zeners. These do not effect operation but protect the gates from ES discharge.
If you pull the opamp and ground the gate of the upper mosfet, both mosfets should warm up instantly as there will be a path through both.
Have fun
As others said, the circuit should work as is: it has a few minor issues, but nothing that would prevent it from singing.
Issues I noticed:
-It a SE amp, meaning huge heat loss for little result (but if you chose that kind of topology in the first place, that's something you have to be ready for
)-The 100K is much larger than the 15K FB, meaning some offset, in particular with the 5532.
-The 220pF looks a bit large, and if the pole it creates is lower than the one caused by the MOSFet, it could unsettle the 5532.
-The positive swing will be severely handicapped due to the threshold voltage of the MOS and the OP voltage of the opamp. With low supplies, it could become problematic
None of that should be fatal, and if the amp simply doesn't work, there must be some obvious error, like wrong pinout or similar
Last edited:
Conclusion for next time I post something stupid...... This morning I realized, that the transistor in the CCS was wired reversed , that was what made it not working.
Now it is OK, works as it was hoped for. Sound is OK, but not outstanding, ACA beats it by large.
Thanks for all of your valuable help and advices.
, that was what made it not working. Now it is OK, works as it was hoped for. Sound is OK, but not outstanding, ACA beats it by large.
Thanks for all of your valuable help and advices.
I've tried, (built ,tested, listened) a few class A amps in the past, but unfortanetly none of them came even close to the modified (dual supply) ACA I am using now... It is a shame, that -due to parts limiting factors- this thing cannot be made with much higher power. Just a few of them: Hiraga 20W, Le Monstre, several JLH-s, with bjt-s and mosfets, the MJamp, a few Nelson Pass designs.
Anyone has experience beyond these, that might be worth to try? Preferably SE types.
Anyone has experience beyond these, that might be worth to try? Preferably SE types.
To lower the dissipated power of this kind of circuit, it has been sometimes suggested to modulate the current delivered by the current source which is then not constant anymore.
Two schemes are possible, the White follower and the Taylor follower. Here are the ideas :
Taylor Source Follower
Two schemes are possible, the White follower and the Taylor follower. Here are the ideas :
Taylor Source Follower
You can also use the correlation of stereo channels to reduce the dissipation :
http://www.diyaudio.com/forums/chip-amps/192934-se-class-regulator-chip-amp-madness.html
:http://www.diyaudio.com/forums/chip-amps/192934-se-class-regulator-chip-amp-madness.html
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.