Welcome! Need support, you got it. Or share your ideas and experiences.
184.76.111.134
I'm refreshing most of the components in my stock PAS-3 (except that the
filament selenium rectifier has been replaced with a pair of diodes),
including the two 2000uf 18V caps in the filament voltage doubler. I've
searched the web to see what people do with these, and I saw everything
from two 1800uf caps to a total of 100,000uf (he didn't say what the two
caps actually were). 4000uf is actually substantial capacitance for a
circuit not expected to carry substantial load, or be subjected to large
transients. Increasing capacitance here is just logarithmically reducing
the ripple. I'd think 1/10 the capacitance (2x220uf say) would be
minimally adequate and probably not have a detectable impact on
performance.
I can find a pair of 47,000uf (or 56,000uf) 35v caps that'll fit in the
space to make up ~100,000uf, but the cost is unseemly: $42. If I were
going to spend money on capacitance, it wouldn't be here. A better answer
to reducing ripple here than humungous capacitance is a better circuit.
I can design an 8-component voltage doubler with substantially reduced
ripple using just two 220uf caps.
Most designs I've seen up the doubler caps to between 2200uf and 4700uf
with the claim of tighter bass and etc. I doubt it, because these
components are not in the signal path. Still capacitance in this range
is cheap and readily available (I have a pair of 35v 3900uf caps in my
junk box). What's really needed here? (Skip the lecture about a regulated
PS - I know, I know).
Stuart
Follow Ups:
My experience has been that the more capacitance on the PAS filament supply, the better it sounds.
I've done quite a bit of experimenting with this in both my modded PAS (Vorhis Last PAS) and a stock (other than diodes replacing the selenium rectifier) PAS.
Forget the theory of whether or not the extra capacitance should make a difference. The experimenting is easy enough to do. Get yourself a couple of high capacitance caps of appropriate voltage, and temporarily clip them into the voltage doubler circuit, using test leads. Turn the PAS on, let the caps charge, and listen for a while.
Once you've got a handle on the sound, remove the alligator clip leads from the PAS circuit, and listen again. Leave the external caps charged, and after you've listened again, clip them back in. There's no issue disconnecting and re-connecting the external caps, as long as they're at the same voltage potential. Never try and clip in uncharged caps to a hot circuit, unless big sparks don't startle you. It's also bad for the caps!
If you don't hear an improvement over the stock 2000 µF caps, then there's no need to go further. But I'm betting you'll hear an improvement. I currently have two 120,000 µF (60,000 µF in series) external caps connected via an umbilical to my Vorhis PAS, in addition to two 4,700 µF on-board caps. I've tried adding two more 120,000 µF caps, and yes, I could hear further improvement, although to a lesser degree.
No need to get that crazy though. I have recently purchased a couple of Nichicon KW 22,000 µF, 25V caps from Mouser, for $7.44 each (link below), and they'll fit nicely inside the PAS chassis.
Try it for yourself. I'd like to read about your results.
PS: Something else you might want to consider are the diodes in the filament supply. I originally built my Vorhis PAS using 1N4007 diodes in place of the selenium rectifier. A few years ago I replaced the 1N4007s with two Schottky diodes (about 50¢ each), which further improved the sound. FYI, I don't consider myself a hopeless Tweaker, and I'm not shy about admitting when I don't hear any improvement in sound. Like raising wires off the floor, for example. Since it was free, I tried it, and didn't hear any difference.
But the Schottky diodes and extra capacitance? That worked for me.
Do you have a part number for the diodes...please.
Thank You
The part number of the Schottky diode is 31DQ10
It’s hard to know. There are a lot of reports on-line, as you have noticed, but not a lot of direct correlation with sonic results. If memory serves, Frank Van Alstine at one time used something like 40,000 here. More recently I exchanged e-mail with him where he quoted something more like 10,000. Who knows why he might have dropped it, if that is indeed what happened.I rebuilt my PAS using something like 24,000 total – and it doesn't sound right. I am wondering if it is too much. There could be other issues. Waiting for time to investigate further.
Edits: 10/11/13
Peter,
After a certain amount of UF there is a point of diminishing returns. Once something is quiet, it is quiet. You can't get anything more quiet than quiet.
I am skeptical about the other responder saying that he hears differences when he goes or clips in more UF and then unclips them. But it's his ears and who am I or anyone else to judge what he hears. I wish I had hearing that good.
To me(getting off subject a bit), I hear more of a difference(for the better in my opinion) when I took an ST-70 or a pair of MK III's and took the choke out and put in another filtering section, than I ever would putting in huge amounts of UF on a filament supply. That choke deadens the sound immensely. Most will argue that it is needed or a choke of some nature that can handle the current. I am not into an argument. All I can say is what I hear.
I feel like this is an insufficiently explored topic in forums, at least in the forums I frequent. Either in DIY projects or in factory-assembled big name high end products, it does seem like upgrades or higher end models increase the microfarads. I do not think it is an issue of reducing the ripple to infinitesimally lower and lower levels. There are probably other reasons, perhaps less “sag” or something else that may not be widely understood or discussed. I have considered that increasing the energy storage might obviate the need for regulation.The question seems to be the threshold: how much is too much? I am not even really concerned about a question of diminishing returns. What I am most interested in is at what point larger amounts of capacitance can actually adversely affect the operation of the amplifier and/or the sound (of course, what constitutes an adverse effect on the sound to one person may not be the same to another).
For instance, Stereophile magazine reviewed the Octave Audio RE 290 amplifier that sells for $10,000 - $13,500 depending on whether you purchase the “Super Black Box,” which is a module that increases the capacitance in the power supply. The reviewer described his perception of the differences in sound with and without the Super Black box, which improved the sound in most ways but not in all ways, subjectively. I don’t recall if the review specified the amount of capacitance in question, but I am including the link.
Then there are the $5,600 each Decware Zen Tori monoblocs that sport 4,400uF for the B+ (not sure if that’s 4,400uF per monobloc or combined between the two).
Well, these amps presumably had designers who knew how to design a power supply with this much capacitance. However, adding huge amounts of capacitance to a Stereo 70 or PAS might be a different matter. We know that the time constant could be an issue as well as other performance parameters that could be adversely affected. I do not know how to calculate this stuff, but I can say that increasing the capacitance in my Stereo 70 from the original 90uF to 500uF sounds great - but at what point might it actually cause problems if I wanted to increase it more?
100,000uF for the PAS filaments? You've got me on that one. I’ve never tried it, and, no disrespect intended, but I have my doubts about it. But, I guess I will have to defer to someone else who has actually tried it.
With regard to the choke, it is there to reduce ripple in the B+. It's an inductor in series with the power supply, and I have wondered if it would also affect the high frequency response of the amp for the same reason it reduces the ripple? Certainly worth experimenting with.
Edits: 10/18/13 10/18/13 10/18/13 10/19/13 10/19/13
Hi Paul,
You have every right to be skeptical. I don't blame you at all. I've always felt the same way about you removing the ckokes. Not because you don't hear and like the improvement in sound. Only because it goes against conventional wisdom. That doesn't mean it's wrong though, only that you like it better, and I've always wondered if I would too.
One thing I will take exception to (unless I'm making an incorrect assumption here) is when you say,
"To me(getting off subject a bit), I hear more of a difference(for the better in my opinion) when I took an ST-70 or a pair of MK III's and took the choke out and put in another filtering section, than I ever would putting in huge amounts of UF on a filament supply."
Have you tried actually temporally clip leading in the extra capacitance into a stock ( meaning unregulated) PAS supply and listening to it that way? Maybe I misread your post, but when you say, "than I ever would" it sounds like contempt prior to investigation. I've always wanted to try your method of removing the choke from my ST-70, and listening for myself. Unfortunately, my work schedule allows me enough time to listen to music, but not enough to solder as often as I'd like. That's one of the things I admire about you, you get a lot of projects done.
Back to the subject at hand. Temporary adding extra caps to the PAS filament supply (via clip leads) will only take you a couple of minutes. I'd like to hear about your results. I'm not suggesting this will give you more improvement than removing the choke, let me be clear about that, because I simply don't know. Only that caps are cheap, the original poster had them in his "junk box", and adding extra capacitance is easy.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: