fast vs. slow rectifiers

[Franz]

Hi everybody!

I have been following this forum for a while now and I am about to build Aleph 5 Monoblocks for the mid/treble range in a biamped system.
One of my main concerns is about PSU design. I just read Nelson Pass´es article about the XA series. It struck me that he uses high speed /soft recovery rectifiers here in contrast to what he says in the older articles about amp psu design, where he favours slow diodes aka standard bridges bypassed by caps.

There must be some sonic benefit otherwise Pass labs wouldn´ t spend the extra dollars for those components.
Does anybody have an idea why Mr. Pass changed his mind about this issue?
Anybody ever tried both and did a comparison?
Maybe the Maestro could himself give some comments.


Thanks for your replies


Franz

[till]

Somewhere, i think in the alephx thread is a statement of Nelson, they use the "fancy diodes" because Wayne wants them. If you search a little bit in the board, you will find also the comment, best choice would be schottky diodes.

And you will find comments, there are much much more important things than which kind of diode, capacitors, resistor you will use.

[protos]

My recent experience in replacing normal diodes for FREDs in my DAC psu was quite positive. I perceived a tighter bass and more dynamics. I haven't tried it yet on the Alephs but I am thinking about it after the Dac mod. I think Nelson has said he doesn't like fast recovery types but I am not sure if that includes soft recovery types.

[JDeV]

Originally posted by till
Somewhere, i think in the alephx thread is a statement of Nelson, they use the "fancy diodes" because Wayne wants them. If you search a little bit in the board, you will find also the comment, best choice would be schottky diodes.

And you will find comments, there are much much more important things than which kind of diode, capacitors, resistor you will use.

Here is some more comments and Mr. Pass's comment on Wayne's choice.

http://diyaudio.com/forums/showthrea...e&pagenumber=1

[Klaus]

Hi,

step by step I replaced over the last years diodes in my consumer devices by soft-recovery ones. In DIY projects I used them right from the start on. In all devices the benefits were the same. I think they are not standard due to there higher price.

It is not the most importent thing - without a One and Only design no start could be made - but to achieve 100% performance this tricky tweaking stuff is absolutely necessary. As well as best Caps with bypasses, inductance-free power resistors and so on, and so on. When I e.g. compare my DIY-Aleph with a Pass-production one a clear difference can be heared. My budget for material was allowed to be higher, that´s the difference. Lucky DIY-guys...

Klaus

[JDeV]

Hi Klaus,
Can you specify the data of the diodes you used eg: make, model number - please.

[Klaus]

Hi Baz,

I used and tested various Diodes, e g

http://www.schuro.de/Daten/IR/31dq10.pdf

http://www.schuro.de/Daten/IR/hfa08tb60.pdf

http://www.schuro.de/Daten/IR/hfa25pb60.pdf

I think all simular diodes are pretty good. The inner resistance could be helpfull with small value = good damping. But the main thing is to my understanding the much lower reverse switch off charge energy. The best should be the Schottkies, see type 1 - but "just" 3,3A nom current. And this all has nothing to do with "ultra/hyper-FAST"... We rectify 50/60Hz - but with hard class A current peaks.

What´s better: to have a Schottky or to reduce inner resistance? I don´t know yet, with my Aleph P in Progress I´ll test and report.

Regards to the Summer-half

Klaus

[hifiZen]

Have a peek at MBR3045WT and kin from IR. These are what I used for my prototype Aleph-X. Can't comment yet on how they compare sonically to anything else, but these Shottky's have two nice characteristics:

1. low price (for a high performace diode)
2. very low forward voltage drop (roughly 0.55 to 0.6V), which translates directly to greater efficiency and reduced heatsinking requirements. Quite possibly this also means greater reliability. You'll have a hard time finding another diode type with such a low forward voltage drop.

[Circlotron]

I have never heard of anyone using fets as synchronous rectifiers in a 50 / 60 Hz power supply. They are the ultimate in low forward voltage drop. The idea is to put the current in the reverse direction to normal through the fet, and turn the gate on (positive as normal) when you are doing this. When the voltage is in the "normal" direction (source negative, drain positive) you keep the gate off so the thing doesn't conduct. This makes an almost ideal diode. Watch out though - the body diode of the fet while quite ok for rectifying low frequency ac, can have a very sudden, hard turnoff so it would be quite noisy. When the gate is turned on, any current that would flow through the body diode flows instead through the normal drain to source path. Only if the voltage drop (amps x rds_on) is greater than ~0.6v will current then flow through the body diode as well.

[jackinnj]

this is where it's helpful to have a good oscilloscope -- you can see the problems of some fast diodes -- more of a problem for switchers in which every millimeter of lead length means extra inductance.