The ultimate DIY? A Stax SRM-T2!

[starcat]

Swapped a ref102 from the positive rail PSU while turning it off. 

Now measuring -259.27 for the -260V rail and confirming the problem was a malfunctioning ref102, so the PSU is now fully operational. Will let it run for several hours. Any procedure for load testing it? 

Kerry, JoaMat, James, Craig - thank you very much!

Now off to making the umbilicals and finishing the amp 🙂 soldering the LEDs next - anything special about testing and matching them?

Edited June 20, 2021 by starcat

[Kerry]

That’s great news. Congrats. 

I just use resistive loads mounted on heat sinks. Make sure they can handle the power and voltage. 

[starcat]

Btw, to you connect in the PSU the circuit ground to chassis ground? And if yes, directly or through an inrush current limiter with 10R or just via a 10R resistor? 

I would opt for an inrush current limiter with 10R, question is how many Amps would be best for the T2 PSU and what fuses are you using? I have read somewhere Craig isn't using a fuse at all 🙂 in his rhodium plated inlet. 

Edited June 20, 2021 by starcat

[Kerry]

I don’t ground the PS to the chassis because there is a switch on the amp board that ties ground to the chassis or can leave it floating. 

[starcat]

Phew, completed the wiring. Thinking about wrapping the wires or leave them as is for better cooling.

Edited June 20, 2021 by starcat

[Craig Sawyers]

Umbilicals - takes more time than you might think! IIRC mine took a solid 8 hours plus.

Because the heater feeds to the power tubes float at -500V I ran the wires inside an insulating sleeve in the umbilical. Glass fiber I think...

Well done with the fault diagnosis BTW - nice job!

[starcat]

Yeah, they take a long time but are fun to make 🙂 I just completed the umbilical for the BHSE I am building (with original HeadAmp boards) and it took long as well despite using an AMP connector with crimp type pins. To be honest, I liked that connector much more than the Amphenol we use for the T2. It is like cutting wires, crimp pins on both sides and insert into the connector housing and voila. The pins are automatically isolated within the connector housing as well, so no fiddling with shrink tubing or soldering into the connector and the pins are available for different AWG sizes as well. A very nice system. 

Craig, do you have any part number for that special glass fiber insulation and how long are your umbilicals? I am planning for 90cm, 3ft. 

Edited June 21, 2021 by starcat

[Craig Sawyers]

I have no idea where my fiberglass sleeving came from - but it was almost certainly Farnell or RS. Mine is black.

I'd have to root around at the back of my rig to measure precisely how long my umbilicals are - but three or four feet, no more than that.

[starcat]

I am using this 15mm polyester sleeve from HellermannTyton for the cable above and considering using it for the T2 umbilicals as well. I don't have any additional sleeving inside the outer sleeve for special wires. I have twisted the 6.3V AC heater wires inside. Anything else I should take care for? 

This sleeve is very nice and is available in 25mm and 35mm as well. 

Edited June 21, 2021 by starcat

[Craig Sawyers]

Provided the insulation between the 6.3V that floats at -500V and the line at +500V, fine. If there is a chance that the wire insulation cannot handle 1000V difference, you are going to need some sort of additional barrier. Hence the fiberglass sleeving over the twisted and floating 6.3V.

[starcat]

I am using silver plated stranded copper wire, PTFE insulation rated for 600VAC, mil-spec, wall thickness .25mm, temperature rating up to 200°C (392°F). Probably good enough for both wires against each other.

Edited June 21, 2021 by starcat

[jamesmking]

personally I use cal test silicon test lead wire. Its cat IV rated at 1KV , very flexible and can withstand 180C. Its also just copper strands with no plating. (I find almost anything silver plated to be bright sounding and have banned anything silver or silver plated from hifi).

I also arrange the wiring to the plugs and sockets so that + voltages go in the top half, grounds and zero volts in a line through the middle and - voltages in the bottom half. This keeps the potential difference between any two pins to a minimum. when possible I leave unused pins around any especially high voltages.  I also test with a insulation tester.  At 1000V the insulation resistance must be off the scale and there must be >50Gohms resistance at 2500V between all adjacent wires.

https://www.uni-trend.com/html/product/tyyq/Electrical/UT510_Series/UT513A.html

 

Edited June 22, 2021 by jamesmking

[Craig Sawyers]

They have replaced the UT513 with the UT513A. But it seems to be a bolted on bargain - in the UK £227. And unusually for a Chinese product the CE mark conforms to the correct logo, so it looks as if it was properly certified.

Thanks for the heads up!

I regularly use an AVO RM290 Megohmmeter, but that only goes up to 1kV, but it does go to 10^12 ohms. Good old analog meter.

[starcat]

I thought that the 600V PTFE wire was that everybody used, KG as well, no? That extra insulation would be nice to know what part no is it and where to get. I was randomly searching at Mouser but couldn't find anything suitable. 

Found here the Cal Test silicone wire 0.75mm 20AWG 10m black, pretty thick at 3.7mm OD. Looks pretty thick for that Amphenol connector, btw. 

Edited June 22, 2021 by starcat

[Craig Sawyers]

Yes - I use the same PTFE coated wire, 600V rated. So from one conductor to another it should be 1200V. The maximum difference is from -500V to +500V - so 1000V. In fact it is a bit more, because there is a -560V line, so 1060V.

Now in all likelihood the insulation will be OK. But I decided to double insulate the heater wires at -500V just to be sure.

But hey guys - just telling you what I did with mine. Whatever works for you is OK with me.

[starcat]

Thanks, Craig! I also thought wire to wire with 2x 600V should be ok. But would get the extra fiber glass insulation if I could find it. 

[starcat]

1KV and 10A rated silicon wire from Cal Test, 20AWG, 0.5mm2 conductor and 2.7mm OD for use with the heater AC wires: https://www.mouser.at/ProductDetail/510-CT2884-0-10

Compared to the PTFE wire it has double the thickness, 2.7mm vs. 1.37mm, both 20AWG, while 18AWG PTFE would be 1.67mm OD. 

Using two wires each 1KV makes it 2KV wire to wire. Using two 600V wires make it 1.2KV which is pretty much good to go for both the +/- 500VAC heater wires, as they are both insulated; correct me if I am wrong.

Edited June 22, 2021 by starcat

[Craig Sawyers]

Interestingly the data sheet is inconsistent. It says 20AWG and 0.05mm^s in the title block, but 0.5mm^2 in the main drawing (which is the correct number)

Anyway, assuming it is 0.5mm^2, that is 0.033 ohms per metre or 0.066 ohms loop resistance assuming 1m long umbilicals. The output tube pair takes 0.9A, so the voltage drop along the wires will be 0.9 x 0.066 = 0.06V.

So that wire looks just fine

=================================

Correction

Wrong current! I looked up 6AC7 rather than 6CA7. Correct current per pair is 3A. So the voltage drop is 3 x 0.066 = 0.2V, which is a bit more significant.

Edited June 22, 2021 by Craig Sawyers

[starcat]

The EL84 would be even more with current draw at 4.5A or 9A per pair.

But because each tube sits on its own wire and transformer winding it is 1.5A respectively 4.5A per 6CA7 or EL84 tube. 

 

So, Craig out of both wires given, would you for the umbilical go with the 

1. Cal Test 0.5mm2 20AWG 1KV 10A silicon wire or with the
2. bulkwire.com PTFE 600VAC wire which I have 20AWG for the smaller tubes and 18AWG for the EL34s? 

Edited June 23, 2021 by starcat

[jamesmking]

33 minutes ago, starcat said:

The EL84 would be even more with current draw at 4.5A or 9A per pair.

But because each tube sits on its own wire and transformer winding it is 1.5A respectively 4.5A per 6CA7 or EL84 tube. 

 

So, Craig out of both wires given, would you for the umbilical go with the 

1. Cal Test 0.5mm2 20AWG 1KV silicon wire or with the
2. bulkwire.com PTFE 600VAC wire which I have 20AWG for the smaller tubes and 18AWG for the EL84s? 

el84? - the spec sheets says thats 0.76A per valve for the heater.... http://www.r-type.org/exhib/aaa0028.htm

T2 and most of kevins designs are EL34. EL34 is around 1.5A to 1.6A heater per valve. http://www.r-type.org/exhib/abv0058.htm As craig says its around 3A for a pair and separate heater windings for each channel... I use transformer windings rated at 4.5A and the extra 1.5A is simply a safety margin. Heater current draw will be higher when the heaters are cold (less resistance) but the current draw will decrease within a minute or as the heaters heat up and reach steady state.

I aim for 50% margin on the transformer heater windings. I like to run the el34 heaters at 6V rather than 6.3 for longer life.

 

 

Edited June 22, 2021 by jamesmking

[starcat]

Yeah, EL34. I have as well 4.5 and 1.5 amps per tube for the transformer secondaries. Sorry for typo, doing too many things at once right now 🙂

Edited June 22, 2021 by starcat

[starcat]

Phew, soldered some 50x LEDs... Checked the C3675 and A1468 for breakdown voltage. Will use the ones with the highest in the batteries. 

Are there any fakes of the C3381 and K246 and is there any special procedure for testing those?

I mounted only 4 of the total 8 standoffs as the ones close to the front panel are in the way of the Stax connectors and the ones in the back are exactly where the Amphenol connectors are. 4 of them should be fine I think. 

Edited June 27, 2021 by starcat

[GeorgeP]

2 hours ago, starcat said:

Phew, soldered some 50x LEDs... Checked the C3675 and A1468 for breakdown voltage. Will use the ones with the highest in the batteries. 

Are there any fakes of the C3381 and K246 and is there any special procedure for testing those?

I mounted only 4 of the total 8 standoffs as the ones close to the front panel are in the way of the Stax connectors and the ones in the back are exactly where the Amphenol connectors are. 4 of them should be fine I think. 

If the 2sc3381 and 2sk246 are the ones you got from me, then they are real. That said, I have seen fake 2sc3381. 

[JoaMat]

Nice progress, starcat.

Some builders put LEDs on the other (tube) side...

On 8/13/2015 at 11:20 PM, spritzer said:

 

 

On 10/31/2017 at 8:51 PM, Whitigir said:

Also, putting battery 22K resistors on the tube side elevated half an inch makes it perfect for using clip leads when setting the voltage to 6.55V.

[starcat]

What pot impedance has the T2 be designed for? I can get the quad RK50 in 50 ohms only but not in 10. Anyone has one for sale?

14 hours ago, JoaMat said:

Also, putting battery 22K resistors on the tube side elevated half an inch makes it perfect for using clip leads when setting the voltage to 6.55V.

Thanks, JoaMat. I see me already changing the 22K R42 to the other side of the board 🙂 

Edited June 28, 2021 by starcat