Diodes EEEYAHHH!

[4krow]

 If diodes are so simple in function, how come there are so many types? I understand the need for different voltages and current ratings, but beyond that, there are a mountain of part numbers, types(UF, shotsky,zener,etc,etc!). When I'm thinking of upgrading a diode in a power supply with the intent of reducing ripple, a host of options overwhelm me.

[adamct]

You are singing my song, bro. When I was trying to pick an LED for my Quickie, I damn near drove myself bonkers. On some level, it didn't even matter. Just pick one and move on. But I was frozen by indecision. I think I have about 15 different red LEDs at home as a result...

[HF9]

From my very limited understanding, some have higher or lower noise from switching. UF are the most basic, IN are a little better, Schottky are very quick switching, etc. Someone is welcome to correct me if I'm wrong. 

[Doc B.]

Thanks, you're wrong. A 1N type is the original recipe. These rectifier diodes have a very noisy reverse recovery spike that gave them a reputation for sounding harsh and grungy compared to tube rectifiers. The UF (ultrafast) series have reduced this spike drastically. The Schottkys have no reverse recovery spike.

[Grainger49]

Zeners are intended to "cascade" when they reach a certain reverse bias voltage.  Forward biased they drop a standard 0.6 or 0.7V.

So for a 15V Zener it will hold you at 15V when you apply 20V.  They have a maximum Zener current and a maximum reverse voltage that can't be exceeded without releasing all the magic smoke.  This is how Dan and Paul use a string of (IIRC) 75V Zeners to keep the voltage at a respectable low value till the Paramount voltage regulator comes on and starts to work.

LEDs work in forward bias.  Our favorite is the HLPM 6000.  It seems to hold a steady forward current with varying voltage inputs.  This is a cool trick someone at Bottlehead came up with.  Probably Paul J.

[earwaxxer]

Interesting you bring it up Greg... I'm waiting on some Schottkys, tants eye of newt etc. to play with the PJCCS. Interesting circuit. Messing with it is a good way to learn about it. I will, for sure, share my experiences. I'm running out of stuff to play with! Looks like a BeePre may be the next step>

[kgoss]

My degree is in  computer science not electrical engineering, but I have a question on this topic.  I read a lot about putting Hexfred diodes in power supplies to upgrade them.  I never see it mentioned for Bottlehead amps, but I have read about installing them in older gear and my  Cary SLP-98P F1 preamp has Hexfreds in the power supply as one of the F1 upgrades.  I was just curious about the Schottkey vs. Hexfred comparison.  Thanks in advance for the education.

Ken

[corndog71]

I've had issues finding those schottky diodes.  Can anyone send me a link?

[4krow]

So, as an upgrade for 1N4001, maybe a UF4007 would be a fit?  Just ordered some for that purpose.

 And I forgot all about those Hexfreds....maaan

[earwaxxer]

I've had issues finding those schottky diodes.  Can anyone send me a link?

http://www.partsconnexion.com/rectifier_diode_cree.html

[saildoctor]

I've used some of the new SiC (Silicon carbide) Schottky diodes with good results.  Only problem is that I could only find them in a to-220 size.

[HF9]

Haha, that's what I get from replying from memory Just to reiterate the correct information: UFnnnn diodes generate less switching noise than their 1Nnnnn counterparts. I recently dropped a pair of UF4007s in my ST70 to help increase the life of the tube and the UFs were more ideal than the 1N4007s. The difference in cost is pennies, so no worries there.