 
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
|
|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
88.19.142.194
I am planning to replace the stock, general-purpose rectifier diodes of my power amp with FRED diodes (IXYS DSEI120-06A [gold-doped], chosen for its low voltage forward), but I have read some disconcerting reports from a few members of this forum in the sense that these diodes in reality add *edginess* to the sound, causing an imbalance between the mid/high frequencies and the bass frequencies.
How coud this be? According to the datasheets of these parts, they have *SOFT* recovery characteristics and lack of ringing, thus generating less RFI/EMI.
That being the case, I understand that they should have the opposite effect, a *lessening* of the glare, edginess and grit of the sound, that is.
Any comment about this?
Follow Ups:
I used the Stealth (hope this helps).
I found them SS toned and edgy. Couldn't get rid of the reverse spike either.
Increasing C1 grossly distorted the waveform as well.
I went back to the UF5408; the reverse spike filter worked flawlessly and could not be happier. I posted the scope pics on TDIY.
http://www.audioasylum.com/cgi/t.mpl?f=hirez&m=17573
http://db.audioasylum.com/cgi/m.mpl?forum=bottlehead&n=7137&highlight=hexfred+'s
http://db.audioasylum.com/cgi/m.mpl?forum=set&n=11755&highlight=hexfred+'s
http://www.audioasylum.com/audio/tubes/messages/11/113677.html
Here is another similar experience:
"While FREDs do offer better performance in power supplies, they do not necessarily offer better sonic results though, as I've discovered later. Overall tonality moves a little higher, thus increasing moving lower treble a little higher and making it sound a little unbalanced, when treble notes start to stand out from overal picture. It may sound nice and feels like system sounds clearer, but get my ears tired quick. I have witnessed similar result with Schottky diodes, but not to such extent. I've decided to leave original D801-804 in place with 10nf cap shunts, but have the D811-814 as FREDs, since DAC and the rest of digital circuitry should benefit from a lower noise power supply."
http://artio2m.com/index.php?2a0e7d109581c4e5fb009b326f0c963e#2009Aug23
of sonic observations and post-hoc fallacies.
There are enough sonic observations to make the question a serious one, but little in the way of engineering insight to explain them.
Every part in a piece of audio equipment may add some of its own sonic character. How the diodes are made and how they were retro-fitted to the equipment could be critical.
"HexFRED" diodes were designed to reduce the stored charge required to support forward conduction in a p-n junction. This charge comes out of the diode as a pulse of reverse current when the diode bias changes from forward to reverse. When the diodes are used in switching power supplies, these pulses of reverse current represent a major waste of energy and loss of efficiency.
Audio tweakers discovered the HexFRED diodes reduced the noise generated by the reverse current pulses in comparison to standard junction rectifier diodes. This noise arises from the steep falling edge of the pulse which sets up ringing in the transformer and associated power supply circuits. Tube rectifiers and Schottky barrier rectifiers do not have stored charge and do not generate reverse current pulses. Solid-state rectifier power supplies deliver a lot more current than tubes, but were known to cause degraded sonic performance. Substituting HexFREDs helped to minimize the degradation.
However, there is still some charge storage and reverse-current pulse magnitude and steepness of the falling edge with the reported HexFRED diodes. Being kind to the experimenters and assuming they did not cause their own problems through sloppy workmanship, it may the case that the HexFRED pulses were still sufficient to excite electrical resonances in the power supply circuits.
You will note that there are now a lot of alternative fast- and soft-recovery diodes available from several manufacturers, as the power loss issue has been joined by EMI requirements on manufactured equipment. I don't have experience with your gear, so I can only offer general advice. Your diode choices may offer improved performance, as long as they are compatible with the turn-on (surge current) characteristics of your gear. The fancy diodes may not have as much overload capacity as conventional rectifiers, and you may burn them out when you switch the gear on from a cold start.
Bypassing each diode in the bridge with a snubber (series R-C network) may further reduce the noise and improve performance. Do some searching for design methods and be prepared for some trial-and-error.
Just a word about Schottky diodes. While they do not generate a reverse recovery spike there is still a resonant tank circuit formed by the leakage inductance of the power transformer and the junction capacitance of the diode. In the case of some Schottky diodes this capacitance can be quite high. the secondary of the power transformer still needs snubbing in my view.
I have replaced a number of diodes with HexFreds and Freds. A favorite was the 4 amp Harris diodes. I have not experienced any added brightness. What I have heard is a quieter background and better definition.
Use the Fairchild STEALTH, high speed, soft recovery diodes. I have had the most consistent sonic results with this variety of HEXFREDs than others (Mouser.com carries them). Have not encountered the sonic harshness with HEXFREDs that you have heard about...I think that it is most dependent on the elements in the audio chain and the voicing of your equipment than the HEXFREDS.
Have had good results with Schottky diodes also. However, most of the really good sounding ones are of the lower voltage variety. Haven't heard the Cree Schottkys carried by Sound Connexions of Canada. The Schottky diodes actually have non-existent switching noise (best scenario) and are very high speed. In most power supplies, they will out perform the HEXFREDS. Hope that this helps and doesn't confuse you.
Hi,
Thanks for your reply.
It is true that some HEXFRED/FRED diodes have *NOT* soft recovery?
Is it possible to compare the "softness factor" of different diodes from the specifications given by the manufacturers in the datasheets ?
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:
Top of Page ]
[ Contact Us ]
[ Support/Wish List ]
[ Copyright © 2025, Audio Asylum, all rights reserved ]
