Yep, I sell (at my cost) a power supply board that uses both fancy schmancy soft recovery diodes (30 amps, 200 volts, datasheet guaranteed softness ratio Ta/Tb) and CRC snubbers. People who use it are quite pleased at the sound quality they're getting. Schematic and freely downloadable Gerber CAD files are linked below for the curious.
Last edited:
All of my BEST designs now use high speed diodes, including the CTC Blowtorch (of course), but also the Parasound JC-1,2,and 3, and the Constellation phono preamps. Mark, you are offering a useful upgrade for many designs.
have you or your coterie "of ears I trust" listened to the slow bridges with the snubbers vs exotic diodes? - Blind of course
Stick with mid-fi jcx, it is a better value for you. This is not an ABX sort of sound difference.
Blind does not mean only ABX, the circle goes unbroken. Peeking is required.
Peeking is indeed required, in order to hear small differences, like diode hash. Both the CTC Blowtorch AND the Vendetta Research have a similar problem with power supply rejection in order to use a resistively loaded complementary folded cascode. Charles Hansen has had a similar problem with his circuits.
Low power supply rejection ratio works both ways. Allowing the power supply low output impedance to couple into the gain stages can lower noise.
Single figure of merits often over simplify issues.
Single figure of merits often over simplify issues.
I did offer Scott Wurcer a no-peeking, single blind experiment to listen with/without diode countermeasures (see post #90664) on his LCR MkIII. He replied in 90666.
Self administered no-peeking, no such thing. What's your point? I'm going with Jan's silentswitchers and batteries BTW. In any case John already said peeking was required as he has said before the differences disappear when you don't.
Last edited:
Why don't you guys use real power supplies instead of opening the door to AC mains noise? I mean, really! You charge enough for these things, you could at least put in a decent regulator!
At least tube gear has a reason to have poor power supply rejection. There is zero excuse for that these days.
-Chris
At least tube gear has a reason to have poor power supply rejection. There is zero excuse for that these days.
-Chris
Chris,
Just what is a decent regulator or power supply? A few years back I showed how much noise just a transformer radiates. Between line noise, transformer saturation and diode switching, the noise spectrum goes as high as I can measure. Typically it falls at 3 dB per octave.
An axial 1,000 uF electrolytic begins to rise in impedance within the audio band.
As to regulator use, in the JC gear I have looked at, he uses them as the first active filter stage.
Again low PSRR can actually have advantages for some circuits but does require a partnership with the power supply.
I don't know if differences disappear entirely? Suppose for a very subtle difference you do a hundred trials blind, and are right 65-75% of the time that way. If the designer is old, and has customers with potentially better ears than the designer (or better than the designer for any reason), I'm not sure there would be problem with peeking as design changes are made. It may be impractical to do extensive blind testing for each change when too many trials are needed to establish statistical significance.
Also, I would just add in my own case, some very subtle differences, right around the threshold my ability to discriminate, are very difficult to hear. And sometimes the differences seem not only subtle, but quite complex and hard to identify exactly and hard to memorize. I don't know if I could pass a blind test on something that difficult, but perhaps so on a really good day. But just don't know. Haven't tried it.
Perhaps more exposure to the sounds over a longer period of time would help.
I do know I have two nephews about the same age I had a hard time telling apart. That was when I didn't see them very often. After spending more time with them and getting to know them better, what was obvious to everyone else became more obvious to me. Until then, I cheated. One wore glasses most of the time and the other didn't.
Go back and look; I recommended single blind testing administered by Mrs. Wurcer. I am confident you can hear what is to be heard (or not to be heard), not what your eyes tell you is "the right answer".
Something like this, I use it in all my GainWire, mk2, mk3 preamps.
http://www.diyaudio.com/forums/power-supplies/286425-shunt-regulator-15v.html
and here is output impedance plots with different mosfets as the shunt part.
http://www.diyaudio.com/forums/power-supplies/286425-shunt-regulator-15v-10.html#post4631127
Last edited:
It is difficult to get rid of RFI being generated by the power supply diodes themselves, as well as the RF leakage of most toroid transformers. I triple regulate the CTC Blowtorch, double regulate the Vendetta, the JC-2,3,JC-80, my studio electronics, etc.
Well yeah, isn't that a tautology? That which is around one's threshold to discriminate is, by definition, hard to hear.
Again, I think you already covered that when you said these "differences" were "right around the threshold [of] my ability to discriminate".
If you cannot "pass a blind test" then the putative difference is below your threshold of detection, no? Perhaps you assume a blind test is limited in time and requires fast switching. Those assumptions are false. All that is required is that the sources are level-matched to better than 0.1dB, and that you do not know which source you are listening to when you make an evaluation (no peeking). You can take as much time as you want, days or weeks, doesn't matter. If the "difference" is audible then you will hear it and be able to distinguish it at some point.
Longer than what?
Now, all of this begs an obvious question: If you have to spend a lot of time listening carefully and comparing two sources in order to detect a difference between them, then how significant is that difference? Why should you or anyone else care about it? If the difference is that "subtle" then can you say that it ruins your enjoyment of the music played through one source vs the other? If not, what are you pursuing?
What begging the question is not: Beg The Question // Get it right.
Regarding the question you raise, the difference may be undetectable by some people, and distinctly audible to others. For example, some people can't hear the quantizing noise in undithered 16/44 audio, and some can.
- Status
- Not open for further replies.