But it sounds amazingly 🙂, now (after almost 4 years of using and liking it) I am not going to change anything 🙂. But I can see that anyone going to build it now, should do some testing.
One idea that I have for some time now , is using hawksford error correction in a voltage regulator. In conjunction with some regular feedback, it can have very very low impedance output. I have designed a power amplifier output stage with that topology with very low output impedance.
Why strive for low impedance, when you can have that plenty-fold in the following cap..??
-If you choose a Low ESR cap and make it big enough to support a low frequency cut-off at 1 mHz or something like that, you'll have really low impedance in all the frequency span. Personally i believe one of the pitfalls is the stray capacitance in almost all devices that allow diode switching elements to pass through.
The best PSU's, I have heard have been Tube regulated L-C-R-C (more RC's) passive supplies, they simply have more balls scale and tonal colour. I have a DAC with such a tube supply, and to me it not only kill's, but really murders the newest DCS- clock this and clock that I also have..🙂
-If you choose a Low ESR cap and make it big enough to support a low frequency cut-off at 1 mHz or something like that, you'll have really low impedance in all the frequency span. Personally i believe one of the pitfalls is the stray capacitance in almost all devices that allow diode switching elements to pass through.
The best PSU's, I have heard have been Tube regulated L-C-R-C (more RC's) passive supplies, they simply have more balls scale and tonal colour. I have a DAC with such a tube supply, and to me it not only kill's, but really murders the newest DCS- clock this and clock that I also have..🙂
In my dac I use 20000uf for the pcm1794 power supply, and I like it very much. also use ferrite beds , as I think ferrite is the best solution to the switch noises, not only from the diodes but from the power line. Sometimes the best solutions is the simplest one 🙂
I agree on ferrite beads being good, and almost the only way to stop switch and power line HF noise, The tube rectifier is just a way to not add any more transient noise by eliminating the recovery spikes of diodes.
I must confess that never give much attention to the diodes, but what you are saying makes sense to me. One solution may be using fast diodes with resistors in serie with the diodes , that should make the commuting current low enough for not causing problems. Essentially is this what the tube diodes are. I will investigate this further.
Maybe an RF core with a few windings in series with each diode, could be a good medicine.
Last edited:
Yes inductance in serie with the diodes will limit the di/dt , it seems a good soluction. for low power applications a resistor may be enough. But nothing like testing , will do that tomorrow
Tube rectifier can not have large capacitor connect directly. So for limiting current spikes that could damage them , it was mandatory to use inductance before the main capacitor. wen solid state diodes start to be use , the inductor was not use anymore as they are bulky and expensive, but now that I think on this, a inductor before the capacitor can be a good thing.
Yes, it is.
I wind coils since i was 11.
It is the forgotten missing link, and yes, not using them lowers cost.
I wind coils since i was 11.
It is the forgotten missing link, and yes, not using them lowers cost.
The first coil I did was a guitar pickup wen I was a kid , I did that with magnets from the motors of toys 🙂 ...Long time ago. Not much money back than , we had to be creative , now the kids have everything they want.
No need. Use a Electrolytic Capacitor and you will be fine.
If you do not feel comfortable with that , use lm8261 or lm8272 they have a different internal compensation mechanism , and are design for unlimited capacitance load drive.