After experimenting with diferent CSs I tries to use a a simple LM317 one evening!
It worked verry well to my surprise! 🙂
Here is the schematics:
I've now build an amp using IRFP450 and two LM317s running 32V DC at 2A - I think it sound really good!
PS is 2x24V/300VA C-L-C filter, no turn on thumps or buzz!
Any comments on flaws in this simple design?
/R
It worked verry well to my surprise! 🙂
Here is the schematics:
I've now build an amp using IRFP450 and two LM317s running 32V DC at 2A - I think it sound really good!
PS is 2x24V/300VA C-L-C filter, no turn on thumps or buzz!
Any comments on flaws in this simple design?
/R
Yep, it is pretty neat... I have heard people complain that it is not totaly linear throughout audiable frequency range... but my personal experiments useing LM317 CSS, did sound very nice... even used it to bias a tube's heater filament.
Since this CCS has no influence over the bias or transfer function of the source follower it supports, it's not too critical and can be replaced by a couple of green LEDs, an N type hexfet and a 0R47 sense resistor. That will sound no worse..... you could also use the LM317 in augmented mode driving a mosfet - which is useful here because of its thermal robustness.
These sorts of circuits (I think Andrea Ciuffoli has one??) sound damn good, but add H2 because of the asymmetrical distortion. The heat they create is horrific. I once built one using water cooling with a 70V supply, a ridiculously over engineered power supply (with large inductors) and around 4.5A quiescent using IRF250s. Front end was a high mu octal triode. It had a presentation of around 8 on the Richter scale, produced around 60W in single end, and sounded wonderful. But it failed traumatically one day when my clever daughter turned off the cooling tap.....
It remains unrepaired, years later, carefully stored in my garden shed. Takes two men to lift the amp, the iron is HUGE.
Cheers,
Hugh
These sorts of circuits (I think Andrea Ciuffoli has one??) sound damn good, but add H2 because of the asymmetrical distortion. The heat they create is horrific. I once built one using water cooling with a 70V supply, a ridiculously over engineered power supply (with large inductors) and around 4.5A quiescent using IRF250s. Front end was a high mu octal triode. It had a presentation of around 8 on the Richter scale, produced around 60W in single end, and sounded wonderful. But it failed traumatically one day when my clever daughter turned off the cooling tap.....
It remains unrepaired, years later, carefully stored in my garden shed. Takes two men to lift the amp, the iron is HUGE.
Cheers,
Hugh
- Status
- Not open for further replies.