The ultimate DIY? A Stax SRM-T2!

[GeorgeP]

take a look at the aavid stuff too - lots of choice.

[nopants]

I have grease but the keratherm conductivity is an order of magnitude higher. Less mess too but I guess that's a secondary concern

[luvdunhill]

Lapping the heatsinks and brackets beforehand and a tiny bit of paste and you will barely be able to pull them apart.

[8street]

Yea, I ordered some 22ohm resistors to put in the -500V supply as well. I assume you mean touch the legs of the 10k PTF56 or the 2SK246 and not the point where it's soldered, right? Do you mean the noise would maximize in the other channels batteries, or in the other battery in the same channel? Either way it sounds like I'm looking for a change in the noise by touching these points.

No. Touch the isolated copper in PCB between ptf56 and sk246. It will add capacitance for stability current source. Noise would maximize in the other battery in the same channel. In any case, just try it.

[G600]

My PSU is alive ! I put a 15VA Talema toroid for 15-0-15 and it's fine.

All voltages test fine with no load, except the -12V at -15.XX volts.

I'll replace the 7812 and 7912, along with the overvoltaged caps.

Then I'll heatsink the amp part and test it too, but still waiting for proper screws (the joy of sourcing US screws in metric Europe...).

Bad news is the postal services have lost a quad Sylvania Green Hornets... so short of input tubes too.

 

Edited October 13, 2015 by G600
typo

[nopants]

Are you guys twisting the filament runs between PSU and amplifier? I noticed the pin legend specifying C-J and D-H for each of the filaments, wondering what the reasoning was for this as opposed to using adjacent ones.

Quite a painful part of this build, have to admit

[GeorgeP]

twisted between for sure.

I went with adjacent, but don't know if there was any logic to the legend that some folks worked off of.

and yes, this is part is pretty painful.

Edited October 29, 2015 by GeorgeP

[nopants]

Can I confirm the power domains for each umbilical? I took these from the schematic:

A	Bias
B	N/A
C	6DJ8 6.3
D	EL34 6.3
E	+12
F	+250
G	-12
H	6DJ8 6.3
J	EL34 6.3
K	Chassis Ground
L	-560
M	+500
N	-500
P	Circuit Ground
R	-260

Can I get away with 24 AWG for the filaments? 

Edited October 29, 2015 by nopants

[Craig Sawyers]

I used http://www.alphawire.com/en/Products/Wire/EcoGen/EcoWire/6715 (18AWG, 600V 105C) for the EL34 heaters.  I twisted pairs and then pushed each pair into braided glass sleeving.  Reason is that the EL34 heaters are referenced to -500V, so you need to be real sure of the insulation of those heater wires and sleeve the solder joints to the pins in the connector.  In fact sleeve all the pins (with Hellerman sleeves) - there are some fearsome voltages kicking around you badly want to keep away from each other and the connector body.  The 6DJ8 heaters are close to ground, take way less current and don't need the same level of insulation.

And yes, wiring the power umbilicals is a PITA, but you need to take your time and do it real well.  Figure on at least four hours work, take frequent breaks and figure the best part of a day.  And then buzz through to make sure of the pin-pin correspondance, and that there are no shorts.

I went through the same sort of nonsense more recently with active speakers I built (Linkwitz LX521).  Since I didn't want an error to find its way into the wiring in the speaker itself and in the connection to the 4 power amps per channel I used 8-way speakons.  Checked wiring of speakers with a battery and confirmed which cone kicked.  Then wiring the cables, again multiply checked.  The consequences of feeding the bass up the tweeter was such a horrid thing to contemplate I was determined to get it right, and have a connector system that was impossible to get wrong.

 

 

[nopants]

is it possible to get a run that doesn't cost a kidney? luckily they sell this at mouser...

 

I looked up the current limits for 24 AWG and it seems I need 3A capacity, this should still work right? The insulation is definitely an issue though. I opted for some ptfe/kapton wrap affair but I guess I should reevaluate this. Thanks for the tip on the hellerman sleeves, I was going to do some normal stuff with heatshrink.

 

On the note of  I'm sort of concerned about thoroughly cleaning the flux off the boards, what did you guys do? I had some pcb flux remover but it seems to just create some weird flux precipitate that is pretty hard to remove. I tried to scrub it down with rubbing alcohol but it started coming back green so I'm a bit concerned haha

[spritzer]

Minimum 18AWG for the heaters.  Those current limits are free air and at serious loss.  You don't want that in a 6.3V heater line. 

I used 1000V rated SPC wire that I twisted and then insulated all the pins on the plugs with 600V heatshrink.  I hate doing this though which is why I'm trying my best to cram a Megatron into a single chassis... 

For the flux I use a foam cleaner from Elecrolube.  Spray it on the boards and let it sit for a minute or two.  Then brush them and then rinse under a running tap.  Let them sit for a couple of days to dry. 

[GeorgeP]

You definitely NEED 20 gauge for the filament wires !

And a 16 gauge wire for the master ground.

I did this.

Also, if you use the Amphenols, they are spec'd for 20ga other than one 16ga pin. But you can probably comfortably fit a couple of 18ga wires.

 

 

Oh, and I would not leave that one pin empty in your umbilicals. You may not have need for it on the amp side right now, but you will be kicking yourself if you need it down the road.

[Pars]

I used http://www.alphawire.com/en/Products/Wire/EcoGen/EcoWire/6715 (18AWG, 600V 105C) for the EL34 heaters.  I twisted pairs and then pushed each pair into braided glass sleeving.  Reason is that the EL34 heaters are referenced to -500V, so you need to be real sure of the insulation of those heater wires and sleeve the solder joints to the pins in the connector.  In fact sleeve all the pins (with Hellerman sleeves) - there are some fearsome voltages kicking around you badly want to keep away from each other and the connector body.  The 6DJ8 heaters are close to ground, take way less current and don't need the same level of insulation.

<snipped>

 

Sounds like a lot of work, but probably appropriate. I had to google Hellermann sleeves... tyton or rubber? Not sure these are available in the US from my brief search.

[gepardcv]

For the flux I use a foam cleaner from Elecrolube.  Spray it on the boards and let it sit for a minute or two.  Then brush them and then rinse under a running tap.  Let them sit for a couple of days to dry. 

This one?

[Craig Sawyers]

Sounds like a lot of work, but probably appropriate. I had to google Hellermann sleeves... tyton or rubber? Not sure these are available in the US from my brief search.

I have no idea - on my side of the pond I just buy them from any of the usual suspects - mainly Farnell or Canford, but RS, Rapid etc all do them too.

They may be called something different in the US.  But here is a link to what they look like http://www.canford.co.uk/HELLERMANN-CABLE-BINDING-SLEEVES

Craig

[spritzer]

This one?

No, Safewash 2000 but in a spray can which turns it into foam. 

Edited October 30, 2015 by spritzer

[s_r]

No. Touch the isolated copper in PCB between ptf56 and sk246. It will add capacitance for stability current source. Noise would maximize in the other battery in the same channel. In any case, just try it.

I've tried this now, and got no audible change through headphones. I've also been waiting for a replacement set of input tubes, which I got the other day, to see if that cures the noise. Still there, however.

[G600]

Don't you have a friend with a scope ?

[s_r]

Wish I did. Although I really wouldn't want to chance blowing up someone elses scope.

[GeorgeP]

I've tried this now, and got no audible change through headphones. I've also been waiting for a replacement set of input tubes, which I got the other day, to see if that cures the noise. Still there, however.

That is no fun.

For those who had noise in the batteries, do you recall if the noise level tracked with the volume or whether it was independent of the volume control? Might help s_r in trying to track this down.

[chinsettawong]

That is no fun.

For those who had noise in the batteries, do you recall if the noise level tracked with the volume or whether it was independent of the volume control? Might help s_r in trying to track this down.

Mine was independent of the volume control.

[8street]

I've tried this now, and got no audible change through headphones. I've also been waiting for a replacement set of input tubes, which I got the other day, to see if that cures the noise. Still there, however.

Maybe it helps:

Add 750-1,5k series resistors in gate all 2sk216. For me it was awesome helpful.

Add 0.1uF MKP or NP0 capacitors parallel 22k resistors in batteries.

 

Also for fantastic sound I turned batteries to 708V.

Edited November 2, 2015 by 8street

[GeorgeP]

What was your +200 and -540 adjusted to in order to get 708v? And were your mods needed to achieve this?

[spritzer]

Just get a super cheap scope (Rigol, Siglent etc.) and some HV probes.  Compared to the cost of the amp it costs next to nothing. 

[Craig Sawyers]

Yup - the T2 was for sure the most expensive bit of kit I've ever built.