The ultimate DIY? A Stax SRM-T2!

[Kerry]

Thanks

I'm using Fusion 360 for all of my modeling.  It's nice because it has CAM capabilities as well so I can build all of the tool paths right from the same program.  I've gotten good at it now.  I made all the part models including the tubes myself.  Still need to fill in the insides of the tubes.

Another benefit, is that by assembling all of the parts I can see if anything is misaligned or bumps into something else.

[Arthrimus]

Those tubes probably aren't going to do much without any guts in them. 

[bravi]

If anybody is interested in an used  T2 at eye popping price, please see the link below

http://yahoo.aleado.com/lot?auctionID=j401791916

 

Cheers!

[DefQon]

That's actually "cheap" compared to what other sellers are selling/sold it at for in the past 8 years and more. The T2 would be purely high priced due to collectors demand no matter how good the DIYT2 has improved on the original thermal and circuit semantics/schematics wise.

The original T2 was never "cheap" on the used market.

Edited August 6, 2016 by DefQon

[astrostar59]

3 hours ago, DefQon said:

That's actually "cheap" compared to what other sellers are selling/sold it at for in the past 8 years and more. The T2 would be purely high priced due to collectors demand no matter how good the DIYT2 has improved on the original thermal and circuit semantics/schematics wise.

The original T2 was never "cheap" on the used market.

Fascinating, but I wouldn't personally pay for it as how long would it last before melt down? However I can appreciate the milestone in tech this amp was back then for sure.

[DefQon]

Kevin and Birgir discussed the issues with the original T2 while back and the biggest problem was heat output within the small enclosure which will reduce the lifespan of some of the components and/or have stuff go out of spec.

The second suggestion was a regulated and better power supply. Of course the DIYT2 doesn`t just improve on these areas but most likely other aspects as well. Kevin or Birgir of course can chime in with the more specific details.

Sent from my LG-D855 using Tapatalk

[spritzer]

I would never run one of these with the cover on.  The cover reaches 60+°C (so you can fry an egg on it) but the biggest problem is the PSU.  The main power transformer will melt with enough hours on it. 

[s_r]

I've recently gone back to try and fix the noise problem in my T2, since finishing the carbon. However it seems like the batteries aren't the culprit. I've now replaced all of the sand in the right channels batteries. I've also upped the feedback caps to 15pf. Neither has made any change to the noise. I really thank Kerry for helping me try to solve this, but this is feeling more like a wild goose chase than anything.

Edited August 21, 2016 by s_r

[chinsettawong]

Have you changed out all the resisters in the batteries?  On mine, I changed all the resisters and the problem was fixed.

[s_r]

Not all of them, but I did swap out the 820k and 390k resistors in all four batteries (from takman REYs to xicons) a while ago.

[chinsettawong]

In my case, it's the problem of leaky resisters.

 

 

[sorenb]

12 minutes ago, chinsettawong said:

In my case, it's the problem of leaky resisters.

How does one knows it is "leaky"?

[chinsettawong]

You hear noises - lots of noises.

[sorenb]

do you see the noise on a scope?

[chinsettawong]

Sorry, I don't know how to use a scope.  Anyway, I learned about the leaky resistors from reading this thread.  I then changed all my PRP resistors in the batteries to a local brand resistors and the problem was fixed.

 

 

 

 

Edited August 21, 2016 by chinsettawong

[sorenb]

Just now, chinsettawong said:

Sorry, I don't know how to use a scope.  Anyway, I learned about the leaky resistors from reading this thread.  I then changed all my PRP resisters in the batteries to a local brand resistors and the problem was fixed.

No prob. I've read the same. But changing resistors does include a bit of soldering, could be interesting if someone hving the problem would run through the batteries with some wick and re-solder before attempting to replace the resistors ...

[s_r]

A few screencaps with a scope from a while ago. The first is the AC coupled voltage at the +200V output of the battery (referenced to ground). The second is the same but at the output jack. I haven't tried re-measuring the output, but after replacing most of the sand the +200V output looked identical.

[Kerry]

I finally got the last plate machined.  I had to replace the spindle on my CNC and also made a mistake on the CAD so had to redo this plate, but it's looking great now.

 

[mwl168]

Beautiful work as always Kelly. Those front panels look seriously thick.

How does it sound?

[JoaMat]

Nice!! Especially the form of top plate at the screws against the heat sinks.

[Kerry]

Thanks.

The top/bottom plates are 3/16" and the face plate is 1/2".

The sounds is amazing!

[chinsettawong]

Nicely done, Kerry.

 

[JoaMat]

Kerry, how long time does your machine need to mill the amplifier top plate?

[Kerry]

The top plate took about three and a half hours actual machining, but I do it over a few days typically.  My feeds/speeds are about 15" / minute with about .035" depth.

[Earspeakers]

I'm in the near final stage of my GeorgeP cased T2 and talked to a EE hardware manager on my team at work about heat sinking. The discussion was useful so thought others here might like to hear what he said. Turns out he has a lot of experience with heat management, since dealing with every mW is his biggest issue on the platform and they've struggled with it for years.  

What they found is that the the only factor with heat sinking is air gaps, since air is an insulator. Torquing is only specified as a poor mans way from the manufacturer of trying to ensure there's no gaps. Beyond the normal smoothness you find on a heat sink that also makes no difference - it all comes down to having a thin layer of compound (ideally it would be zero but obviously we have to suffer with minimal compound) with no air gaps. It's so important for them that they x-ray every one, and they use no hardware to clamp the parts down - just normal soldering. The key is good technique in applying the compound and placing the part. If you do that right torque, lapping or anything else doesn't help. 

I explained my project and showed him pictures. On hardware he didn't like the idea of using steel screws with the HV silicon - he recommended going with the Peek screws for arc safety and again - using good compound application technique with the screw there just to hold everything together. Further on having all that aluminum and parts screwed together (IC's, brackets, sinks) he was concerned with creep. Ideally you attach to a 'hard' surface, meaning one without creep. A thin PC board is fairly hard, but aluminum isn't - he described it as 'mushy'. You can screw to it but over time the aluminum under the screw points will "flow" away from the pressure point. So, what they do (because everybody has to suffer with aluminum sinks) is to use a wide steel washer under the clamping bolt. This works well to ensure the bracket/sink stays together with thermal cycling. Otherwise he suggested a yearly physical inspection of IC to bracket and bracket to chassis to ensure a continuing tight fit. 

So, he recommends PEEK screws, don't over torque (or worry about torque at all), but be fanatical about compound application, such as using a non silicon based compound like AOS 52022KH for long term stability if you like. For attaching sinks and such he recommends using a wide steel washer (not a 'lock' washer) to spread out the pressure point. 

FYI -

Edited September 22, 2016 by Earspeakers