The ultimate DIY? A Stax SRM-T2!

[kevin gilmore]

I have some of those for the dynafets i am going to build soon, got them from marc.

the ceramics seem like a much better idea, lower capacitance and guaranteed higher breakdown voltage. (i think)

[Inu]

I have some of those for the dynafets i am going to build soon, got them from marc.

the ceramics seem like a much better idea, lower capacitance and guaranteed higher breakdown voltage. (i think)

Capacitance

With 4171G = 4PF

With Silicone sheet = 11PF

[Petter]

Word of warning - I have worked with PPS isolators in 80-130 degC environment (typically 1hour), and the PPS isolators would quite often loosen - meaning that the assembly holding them would have to be tightened. I am all for PPS, but be careful to doublecheck that you don't get into a situation where you get too loose after a while. This is probably nothing to worry too much about, but it would be sad if it happened and went unnoticed. I too will go for this setup by the way.

Beryllium Oxide washers are not really that much better than Al2o3 units as I recall, and last time I checked, probably 10 years ago, BeO units were still available from Thermalloy or someone like that. I too am all for excess, but in the case of BeO, it is probably not worth it unless you have a nice source already. The thing to watch out for with ceramics being almost completely rigid, though is that they are less tolerant to "un-flat" heatsink than pads.

Petter

[kevin gilmore]

The 7721's are just a slight bit too big for the ixys parts. So either drill them out .002

or use the standard plastic washers. I went with the standard plastic washers cause

i did not want to drill out the parts while soldered to the circuit board.

The ceramic insulators are good for about 4kv.

Beryllium oxide is a slight bit better. But its also toxic...

Perfect.

Still in search of the ultimate in insulating screws.

Power supply back under load.

[nattonrice]

Are those nylon cap screws a candidate?

[Kerry]

What about Glass Filled Nylon. Here's one at McMaster-Carr. Search product 91221A220.

They don't come in 4-40 so you'd have to use 6-32, which will work with everything except the 2SA1486 devices.

They have an operating temperature range from -40 - +230 deg F. I don't see there electrical properties listed, but I have used them on the Blue Hawaii.

[Petter]

I have been playing around with thermal grease quite a bit for computers and come to the conclusion that I like thermal grease which is very easy to apply in super-thin layers. Toothpaste quality if you like.

Another important item for me is that it should not only be non-electrically conductive, but also not contain conductive material (so that it can become conductive in the future). While this is typically not a problem for the almost unspreadable crap that you normally would pick up from an electronics place - at least the "high performance" items, it is a problem with typical high performance pastes intended for CPU cooling, and given the voltages we operate on not something I want to bother messing with.

Bottom line, Arctic Silver Ceramique is the best I have found. Super easy to apply in thin layers, does not contain conductive material (belying the name). In these HV days, you may want to take a look at it. It is not really that costly when you consider that it spreads a long way (2.5g at $5). www.arcticsilver.com

Using easily applicable matererials and good technique will probably get you more than moving to more esoteric materials such as BeO (from Al2O3)

Your mileage may vary.

Petter

[nattonrice]

Yes ASC is a commonly discounted item at online comp stores.

You'll prob need more than 2.5g though.

[kevin gilmore]

Drilling out the bracket and circuit board holes to accept #6-32 is certainly something that can be done.

I'm not going to do it. I left the power supply run under 50% load overnight, and everything looks great, the

nylon screws did not loosen at all. I would sure like to find some glass filled nylon screws, i'm even willing

to have a batch made.

[luvdunhill]

Your mileage may vary.

There are a lot of much better options out there than Artic Silver. Here are two:

http://www.aavidthermalloy.com/products/options/ultrastick.pdf

Home

[Voltron]

I have been playing around with thermal grease quite a bit...

***

Your mileage may vary.

Petter

Funny coincidence. I have been playing with thermal grease quite a bit myself.

[Les_Garten]

There are a lot of much better options out there than Artic Silver. Here are two:

http://www.aavidthermalloy.com/products/options/ultrastick.pdf

Home

I got a Woody reading that PDF, was that wrong of me?

[Fitz]

Wow, and it's not even that expensive either.

[Petter]

I got a Woody reading that PDF, was that wrong of me?

I think it may well have been. The way I read it, Ultrastick is a paraffin based substance which has been on the market for at least a decade (which can still mean it is good of course). I am not in a position to say whether it is better or worse than anything else, and indeed I was planning to use it with BeO isolators about a decade ago when I was looking at it. It seems lovely application wize. However, there is nothing on initial thermal performance, only a suggestion that thermal performance is constant over time, and better than something else (and they don't say what this is). Besides, based on paraffin it is actually flammable, and will evaporate at high temperatures - but sure, it cleans up nicely.

The way I read the second reference (tapes), I do have some experience with thermal tapes, also from the CPU world, and I don't like them one single bit. It is true that new materials that have recently entered the heatsink market may be very very good indeed - graphite based materials can be almost as good as copper (in one plane) etc, but I don't believe in them for paste, possibly they might be good for film. Besides, if you look at "research" done by overclockers, it all goes the way of thermal compounds a la Arctic Silver, or at least it did when I was keeping up on these things.

My point is that you select a technique that suits you, and if you execute well, you should have a great long-term solution. It helps if the materials are easy to apply, and all these + my Berquist films with integrated "thermal compond" fit that bill. I like the concept of non-conductive materials that do not evaporate, and thick ceramic spacers + flat mating surfaces.

Again, your mileage will vary, and I am probably spending way too much time on this minor point which will minimally influence the sound quality of the finished product.

[Hopstretch]

Well, we all have our own thermo-sexual preferences and far be it from me to discriminate on the basis of lubricity or any other semi-relevant parameter.

[kevin gilmore]

final power supply picture with the ceramic insulators and the nylon cap head screws (#4-40 x .5)

Also used the special double thickness nylon nuts which really hold things tight...

Screws came from Mcmaster-Carr, Not sure where the nylon nuts came from.

http://gilmore.chem.northwestern.edu/t2power3.jpg

Will replace the washers on the amplifier tomorrow or sunday.

Going to need longer screws for the 2sa1486's.

By the way, the power supply is approximetly 29 lbs finished.

Amplifier chassis is about half that.

Edited March 19, 2010 by kevin gilmore

[Kerry]

Were you able to use the thicker ceramic insulators without having to re-solder the devices? It seems like you were just able to re-bend the leads a bit to get them to sit right.

I love the black nylon screws.

final power supply picture with the ceramic insulators and the nylon cap head screws (#4-40 x .5)

Also used the special double thickness nylon nuts which really hold things tight...

Screws came from Mcmaster-Carr, Not sure where the nylon nuts came from.

http://gilmore.chem.northwestern.edu/t2power3.jpg

Will replace the washers on the amplifier tomorrow or sunday.

Going to need longer screws for the 2sa1486's.

By the way, the power supply is approximetly 29 lbs finished.

Amplifier chassis is about half that.

[kevin gilmore]

Yep, no resoldering of any kind. The leads only get bent an additional .050 inch anyway, and there

is more than plenty of room for that.

just finished replacing all the ceramics in the amplifier, now wiring the stax jacks.

thermal grease all over the place

Too bad there is not a .035 inch thick ceramic insulator that is .5 wide.

[kevin gilmore]

By the way as of yesterday i have

9 sets of heatsinks drilled and tapped and 4 sets of fronts/backs fully machined.

All of the fronts and backs have been cut to size.

Soon i will have tops and bottoms, then 3 sets get shipped

for polishing and finishing.

More stax jacks and knobs also on the list.

You will be able to do your own knob if you want to

[kevin gilmore]

There is a sizing error on the board for the dact part. Don't know how it could have come out that

way, but it did. Since you always had to extend the pins anyway to get it to match the center line

of the rk50, a little creativity will get it to fit.

the dact has .1 pin spacing, and .580 from first set of pins to second, .800 from second to third and .580 from the third to the fourth

somehow i ended up with .15 spacing, .575,.850,.575

[kevin gilmore]

Change D8 in the power supply from regular diode to 30 volt zener.

updated schematic that actually matches the circuit board here.

http://gilmore.chem.northwestern.edu/t2schempower.pdf

In fact you can also change D2,D5,D11 to the same 30 volt zener.

Reason, in testing (and doing stupid things with the bias) i found a way to zotch the

-500 volt fet. The zener prevents this.

[nattonrice]

Soon i will have tops and bottoms, then 3 sets get shipped

for polishing and finishing.

That's moving along pretty quick.

So you have decided on a final finish for the cases and heatsinks?

Are you going to go with your paint shop?

[kevin gilmore]

So i highly recommend that everyone test their led's. Which i thought i had done.

But one either was bad from the start or died infant mortality. It was causing all sorts

of dc instability.

And yes i know its out of focus, the result of a 3 second hand held exposure.

http://gilmore.chem.northwestern.edu/t2lites1.jpg

Troubleshooting this thing is real easy as you can do it one channel at a time

and the led's point to where the problem is.

updated power supply schematic, i decided to change out all 4 diodes for zeners.

http://gilmore.chem.northwestern.edu/t2schempower.pdf

the last thing to get fixed will be the bom

Edited March 22, 2010 by kevin gilmore

[kevin gilmore]

And in full glory... ok, i need a 4 inch tripod, that or put the thing on a table

(need decent horizontal surface that is not already filled with layers and layers and layers of crap)

(LARGE horizontal surface!)

http://gilmore.chem.northwestern.edu/t2lites2.jpg

So what you all figure?? About 500 hours of burnin

The power supply cables are going to take at least 300 hours just to settle down the electrons

Edited March 22, 2010 by kevin gilmore

[Voltron]

AND . . .

Friggin awesome. What's it sound like!?