I feel that this is a great video and I hope everyone here can watch it.
Not a bad video, but not exactly a "tell all" either. Main point might be stated as, "the machine measures what it does, up to the user how to make good, honest sense of it." This assumes the user knows how to correctly operate the machine, of course.
IMHO the machines are pretty darn good at what they do. The problem, if there is one, is often with interpretation of results by humans...
I had a chance to vet your story with some old timers, including JC. They report you did contribute as you say. At the time you wrote about diode commutation noise as it relates to audio it was generally not a well known issue, according to JC. He said to the best of his recollection he and Walt Jung helped to get your letter published, since at the time you were unknown to the magazine editors and they weren't sure if your report was legit.
Back in 1989 and 1990, HiFi News and Record Review published a series of articles written by Ben Duncan about his AMP-02 project. It was kind of a combination of a tutorial on some design ideas and a DIY project. There was a way to buy pcbs, too.
Some of the things he wrote about were the concepts of common mode signal rejection, AC line noise, and a bunch more topics that are often scoffed at even today. (That's progress for ya...)
There was a single throw away line in one of the articles in the series, actually just part of a line, suggesting that fast soft recovery diodes in the rectifier circuit might be an area for investigation in order to get improved audio performance.
Walt Jung and Rick ran with that idea, developed it, and published what they found, complete with some measurements.
Ain't history cool?
Some of the things he wrote about were the concepts of common mode signal rejection, AC line noise, and a bunch more topics that are often scoffed at even today. (That's progress for ya...)
There was a single throw away line in one of the articles in the series, actually just part of a line, suggesting that fast soft recovery diodes in the rectifier circuit might be an area for investigation in order to get improved audio performance.
Walt Jung and Rick ran with that idea, developed it, and published what they found, complete with some measurements.
Ain't history cool?
First things first: replace the crapy diode/diode bridges with fast-switching soft recovery diodes.
I don't know if that should be first. Using better diodes certainly improves things, a lot, but it doesn't eliminate the ringing problem with the transformer inductance. You still need an RC snubber of some kind to kill it dead. I think that "bad" diodes damped properly may be better than "good" diodes not properly damped. Doing both is the best solution.
Look at the article authored by Mark Johnson here: Soft Recovery Diodes Lower Transformer Ringing by 10-20X
He's presented lots of other information on the subject elsewhere here on diyAudio.
Thank you Markw4 and CG. It is nice to see you have investigated my work with diodes and know what I have done and then posted about it. So many other people here seem to be here to argue and fight with other members. We all have the same interest - Audio and enjoying music. There is not much talk on this forum about how our equipment sounds after we build it or make changes to it and that is so strange to me as the end result is to listen to music thru our equipment.
Merry Christmas!
Merry Christmas!
I have doubts about need for snubbers. Yes, ringing is there. However, I have a hard time to find any additional noise caused by rectifiers, even at first reservoir capacitor. Here is an example with PS used previously in this thread. It has fast rectifiers, in fact MOSFETs and LT4320 control circuit, 15V output and 2A load.
With no snubber, ringing frequency is about 167 kHz and about 1 Vpp. However, nothing is visible at first PS capacitor. Ringing appears after rectifier diodes switch-off, isolating primary from rest of the PS, so any noise can enter only through small rectifier diode junction capacitance, and it is absorbed by first reservoir capacitor. MOSFETs used here have larger parasitic capacitance than plain diodes.
I have PS circuits with properly adjusted snubbers, using Quasimodo, but I start to ask myself is there any real benefit.
With no snubber, ringing frequency is about 167 kHz and about 1 Vpp. However, nothing is visible at first PS capacitor. Ringing appears after rectifier diodes switch-off, isolating primary from rest of the PS, so any noise can enter only through small rectifier diode junction capacitance, and it is absorbed by first reservoir capacitor. MOSFETs used here have larger parasitic capacitance than plain diodes.
I have PS circuits with properly adjusted snubbers, using Quasimodo, but I start to ask myself is there any real benefit.
How much of that ringing is electromagnetically coupled to the audio circuitry? In other words, not by conduction from the terminals of the reservoir capacitor. I think you'd need to use a current probe to measure that.
But, if you're happy saving the pennies that snubbers cost, go for it!
Maybe it could be put it this way: If you try snubbers does it help the sound at all? If not, then maybe you don't need them, or maybe you have bigger problems to work on first. No offense intended, of course. Its just that sometimes Murphy can get you 🙂
They sell extremely over priced power cords of course there marketing is BS. Almost everyone of there misconceptions is actually true. Find there measurements paper to see what a joke there "research" is.
As I suggested, I think you may need to use a current probe on the wiring of the amplifier section proper. It sure could be that your amplifier embodiment doesn't have anything bad coupled to it. If that's the case, great job, and let us know your secret!
To me an early sign that expensive power cords and such are likely to be snake oil is that they only exist in domestic HiFi.
They do not exist in recording and recording engineers are pretty much the only people who get to compare original to the recording.
They do not exist in medicine where it could be a matter of life and death.
They do not exist in experimental physics where the tiniest margins could mean billions of wasted dollars/euro/pound/yen.
No they exist only in a limited market where ultimately it does not matter as no lifes or tax dollars are at risk and that is rife with marketing BS.
Regarding power cords and EMI/RFI filtering, there is more than one way to skin a cat. In medicine and physics research nobody cares if ferrites can sound grainy. They care about specs and measurements, and grainy is not usually something with a specification, even if it might be audible in some hi-fi cases. In recording and mastering, for the most part people are not interested in the super utmost fidelity. They are interested in making money, selling records, loudness wars, translation, what the customer wants (record company, band, producer), etc. There are few labels that specialize in best in industry SQ, but not the majority. My two cents worth anyway.
- Status
- Not open for further replies.