will
|
Hey Brad,
I am afraid I am not a good one to answer your question. So I was holding off responding, hoping someone who knows about power supplies would come in.
I have done a lot of modification on my amps, but using Steve's design as a pattern, and tending not to diverge very far from it. Working this way, I did not feel the need to learn the deeper intricacies of amp design, more experimenting with parts and wire sounds, and using relatively minor value changes (for the most part) if I reasoned it might help me reach my goals.
My biggest value changes were using .082 Miflex Copper/Oil coupling caps in the place of 0.10 Jupiter Copper, and changing 47uF IC electrolytic caps to Jensen 40uF. My hope with both was related...less capacitance shifting the balance a little away from my Toriis leaning a little toward bass thickness (by my tastes and based on my systems), and in turn, tightening and speeding up bass, while allowing more open/spaciousness in the balance. I think both worked, but having changed the caps used as well as the values, it is hard to be definite about what did what.
From my rudimentary understanding, I think of power supply caps as storing energy, and as filters. As I recall it, our full wave rectifiers convert AC sine waves (positive and negative alternating waves) as they come from the transformers, to all positive waves in sequence...a problem being, though all positive now, they still retain the wave form, so there are power dropouts between wave peaks. Then it seems the stored cap power is used to smooth those waves by providing power needed downstream during the wave dropouts, filling in the spaces between the wave peaks with stored power. This flattened combination is DC, but apparently this process does not fully smooth the outgoing waves, leaving a sawtooth pattern... residual AC "ripple" causing noise, especially noticeable if the saw teeth are too deep.
It seems cap values are determined based on creating a decent balance between transformer AC output, the filtered DC load needed, how much remaining ripple is acceptable, and using a cap value that does not pull too much/too fast from the rectifier...Bringing us to the issue you are looking at with the 16 uF rated U52 charging a cap bigger than it is rated for....
I have not tried to get further than this generalized view of how rectifiers and power caps seem to work, so I can't say if you can change your reservoir caps from 47 to 16 or less without negative effects. I kind of imagine this on its own to be so far out of Steve's design parameters that it would create audible problems, but I could only speculate on how, not knowing enough to speak with confidence about it. I"m guessing that there may be good ways to compensate for the lowered capacitance also, but again, that is not something I have looked into or tried.
Hopefully someone else will offer more conclusive information for you. Or, if you figure it out, please let us know.
On the other hand, there are a lot of really nice rectifiers that do work with Steve's power supplies.
|