[tubesforever]

Bill Epstein is building his new Hagerman with Fairchild HEXFRED's.  These are fast epitaxial diodes with  soft recovery features that help to eliminate ringing.  While these are used quite a bit with solid state gear and tube gear for bias DC circuitry, this is the first time I have seen them for heater circuits.

These are cheap.  1.25 each at Percy Audio (no affiation) for the 4a 600V Fairchilds.  That makes them a tempting tweak for an OCD modifier like myself.

The question is, how do these orient vs the Shottke Diodes. 

When I read up on cathodes, the cathode is where positive ions flow out.  However in a diode, the cathodes are listed as the negative pole of the device.

So which way do I arrange the cathode of this HEXFRED?  Does it go to the same side as the Shottke  silver stripes or the opposite side.   

I am sure I can figure this out by reading the manufacturer's literature and re-examining everything on the Hagerman board.  But if Bill or anyone else knows the proper orientation I would appreciate some advice on this. 

I will cover mine in clear shrink wrap so it might be more viewable from my pictures when this tweak is completed.  I plan to have this done by Sunday night.  Right now I am taking a two day vacation just listening to some fine music.   

[Bill Epstein]

Here's the data sheet with pic that I used to orient them, taking the cathode with the extra metal bit as the striped side of a conventional unit. Sure hope that's right!

http://www.fairchildsemi.com/ds/IS/ISL9R30120G2.pdf

[tubesforever]

Bill, thanks for the tech sheet.  I looked up the Shottke diode material on Digikey and found that the silver band is the cathode.  The cathode goes to the square holes on the Hagerman board.

Tonight I removed the Shottke diodes (I had to cut off the leads off the Schottke's so this became a sacraficial no going back kind of tweak).  I used some shrink wrap to protect against shorts, carefully bent the legs and soldered them into place.

Ordinarily for many of you with working Clarinets, and Cornet/Cornet2 this would be easy with the stock specified Nichicon 10k uf caps.  With my Panasonic TSHA caps, I had to lean them on their side.   

BTW for those wanting to do this, the Fairchild HEXFREDS are just 1.25 each for the 4a 600V model from Percy Audio.  You need to use some plyers to bend the leads to fit the diode traces.  On the right side the cathode surfaces meet.  So you need to use shrink wrap to prevent shorts.  On the left side the black cases meet which is safe, but I would shrink wrap them all. 

Looking at the hexfred from the black surface facing up and the cathode surface down, the cathode leg is on the left.  This goes to the square pegs.  The anode leg is the one to the right and goes to the round shaped traces.

The large cathode surface on the back is there so the HEXFRED can be mounted to a heat sink to protect it in high current applications.  I understand that the Schottkeys are seeing approximately 18 watts from what Jim emailed me.  Because these are seeing so little current, I am going to try mine without a heat sink.  If I need to use them, I will need to replace the TSHA with something smaller. 

If you feel you have to use a heat sink (I do not recommend it) then make sure you use an insulator pad between the cathode surface and the heat sink because if you don't that heat sink is a big accident waiting to happen--It will be fully conductive!

I will give you more information tomorrow on how they sound and post some pictures.  I was playing music late tonight at the second and third step of my 24 step attenuator.  Even at extremely low volumes the overall sound quality was very promising.

More to follow.

[Bill Epstein]

Here's a pair of the little devils showing the extra bit of metal on the cathode, which would be the banded side of a conventional diode



Whether you order from Percy, or go to Digikey for the Cree Silicon carbide Schottkys http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=CSD04060A-ND

Both sources have the mica spacers, conductive grease (use sparingly) and TO-220 heatsinks. As Tubes says, the sinks aren't necessary and probably won't fit, anyway.

  DO insulate them with heat shrink, not just from each other, but from inquiring fingers

As long as I'm doing links, here's another one to Mouser for the KOA Speers winners of the Kiwame look-alike contest

http://www.mouser.com/Search/Refine.aspx?FS=TRUE&Ntt=*spr2lt52*&N=1323038&Ntx=mode%2bmatchall&Ns=P_SField&OriginalKeyword=spr2lt52&Ntk=Mouser_Wildcards

Just 30 centavos, they look like Kiwames, smell like Kiwames and just maybe, quack like'm.


***Now who can explain how to use the decals that come with the kits?***

[Niteshade]

I have never had a ringing problem while using silicon diodes. Could you describe the problem?

[PatOMalley]

Is there + and - designations on the diodes so you know how to orient them? I don't see that printed on the diodes in any of the pictures.

[hagtech]

It's a complicated issue involving much more than diodes.  One can damp out the ringing with snubbers, which I do.  The diodes are merely the stimulous.  They generate the impulse energy that kicks the LC tanks into ringing.  The judicious approach is to attack both ends - use proper diodes and add snubbers together.

www.hagtech.com/pdf/snubber.pdf

The typical hack method is to put small caps across the diodes.  That merely shifts the resonant frequency, which may or may not help.  It does not actually dissipate the energy.  Diodes that hang on and do not turn off fast end up introducing the largest energy glitch into the circuit.  Hence, we prefer not to use slow diodes.  Schottkys operate on a different principle than a typical PN junction diode, and do not suffer from stored charges (or something like that).  The silicon carbide diodes are again different, but I have not looked into the device physics yet.  Of course, vacuum tube rectifiers tend to be the cleanest types.

jh

[tubesforever]

Thanks for you input Jim.  I think I can confirm a lot about your design strategy and its effectiveness.

After removing the Schottke diodes and replacing them with the Hexfreds, I left my system playing music quietly all night long.  This morning at 9 AM I started seriously listening to the hexfreds.  First off,  I cannot hear a significantly demonstrable change in the sound of my system.  This is probably because the Bedini amplifier I am using has a higher noise floor than my Clarinet line stage. 

From this baseline, while I know the Hexfreds are providing less ripples and ringing in the heater circuit my amp doesn't give me the ability to hear any change in terms of the overall noise floor.  The highs might be a slight bit better however this is purely subjective analysis on my part.  Humidity, temperature or other variables could be impacting my system to enhance the highs this morning. 

If you want to evaluate your own noise floor, this is how to do it.  Jim Hagerman posted about this method as a good way to track down ground loop hums and it works to define your noise floor equally well.

First, disconnect the line stage rca leading into your amps line input.  Power it up.  Listen carefully, close up to the speaker drivers.  You are now hearing the noise floor of your amplifer.  Now plug in the line stage interconnects with no other devices connected and do the same test.   Simply add each source one at a time and listen for changes. 

In my system the Hagermans were so silent with either the Schottke diodes or the Hexfreds that they were easily quieter than my amplifier. 

On my recent road trip my Hagerman was clearly more quiet than any preamp and phono stage out there other than the Manley Steelhead, Pass Xono, and Wave Light pieces.   These pieces were as quiet.   BTW these other pieces all cost upwards of 6,000 dollars retail. 

For most of the people who heard my Hagermans in their own systems, my pieces were so silent they were wondering if the equipment was turned on and connected.  Some with solid state gear could not believe their ears.  They assume tube gear is noisy and that is not the case.  Jim did a great job on these power supplies in my opinion.

This morning I kept trying to hear differences in low level bass detail, room response  and other micro detail type information but it sounds like it did before with the exception of a very slight improvement to my high frequency response which could be completely subjective.   Humidity, air temperature and other factors could easily account for this slight improvement.   I see no need to remove the Hexfreds now that they are in.  They will do no harm and there can only be an upside.

If your current Hagerman pieces are running and you are happy then you might want to follow the following hot rodder axiom.  If its not broken, don't fix it. 

The teflon caps?  Now that is a tweak that will turn your listening room upside down.