It should be FJP2145 shouldn't it? Also I see two C3675's near the bottom left that should be 2145's.

It should be FJP2145 shouldn't it? Also I see two C3675's near the bottom left that should be 2145's.

http://www.mouser.com/ProductDetail/Fairchild-Semiconductor/FJPF2145TU/?qs=%2fha2pyFadui9BXYxjh0%252blwo7N7LskIUoGMKoPSrDrpk%3d

 

FJPF2145 is in TO220F package which has insulated tab

 

It might be nice to have space for small gate stopper resistors on the 2sk216s?

And maybe flipped...emitter goes to 510R?....the label should change that I think.

Also I see two C3675's near the bottom left that should be 2145's.

Edited October 12, 201411 yr by GrindingThud

label was flipped, wiring was correct.

3675 label also fixed.

Edited October 13, 201411 yr by kevin gilmore

/

 

Edited November 12, 201510 yr by JoaMat

I'll take a stab at cleaning up the board too.  I think it would be best to have a left and right version of the board instead of one large one unless we can find a chassis that works for this. 

Edited October 13, 201411 yr by spritzer

/

Edited November 11, 201510 yr by JoaMat

The modushop 2U/300mm chassis should work just fine for this but then we need split boards.  I also prefer to use angle brackets to hold the boards in place.  Makes testing that much easier. 

 

Also makes the PSU easier to deal with as there isn't a chance in hell we can make it fit into 300*200mm. 

Are the 2u heatsinks enough for these devices? What's the size compared to the original DIYT2 sinks?

The heat sink is rated at 0.45°C/W each so it should work. 

The designer of Stax SRM-T2 he knew what he was doing.

Edited November 11, 201510 yr by JoaMat

He sure did.   

for those of us too poverty for angle brackets, will there be issues with the additional  parasitics if we mount devices near the sink and use wires to get to the main amplifier board?

Given the issues we had with the resistor brand I'd guess so.  Angle brackets are a good idea though.  They support the boards and make testing far easier.  Easy to drill by hand too. 

/

Edited November 11, 201510 yr by JoaMat

Nice work JoaMat!

 

Calling Kevin,

I wonder if all the components of the psu could fit a smaller pcb to match the size of the modded board (including transformers)?

 

I think I’m dreaming....

Edited October 18, 201411 yr by johnwmclean

how's the hum (if any) with torodiy?

that power supply looks insane

there is a verion of the power supply board that is 11.95 inch wide x 9.2 inch deep

 

with all the electrolytics in there, just don't see a way to make it much smaller.

 

joamat's may be slightly smaller, and 2 boards might be easier to build.

Nicely done

joamats power supply boards are single sided and are 8 x 5.83 inches each

 

so you would save about 1.2 inches.

 

http://gilmore.chem.northwestern.edu/t2psnew1.jpg

 

http://gilmore.chem.northwestern.edu/t2psnew2.jpg

 

pretty sure the boards are done, someone please check this one carefully

 

http://gilmore.chem.northwestern.edu/t2modified.jpg

Edited October 18, 201411 yr by kevin gilmore

The 10M90S on the top center is not centered properly. It's about 1/10th of an inch to the right.

I haven't carefully checked the rest yet.

 

Edit:  never mind,  just looked again and it looks good now.

Edited October 19, 201411 yr by Kerry

Wait, did the original from joamat have the servo trimmers? If so, crazy as that is what I did to my T2 to fix the biasing shenanigans... Assuming you did, can you describe your biasing procedure and give an idea of how far over the original 6.2K you went? I am glad to share pics of mine and procedure as well.

How do the new darlington servo transistors fair in terms of gain? More or less than 2sc3675?

Edited October 19, 201411 yr by luvdunhill

Seems they are much lower gain compared to the original. Anyways, I can just PM you as I am curious about a few related items

/

Edited November 11, 201510 yr by JoaMat

It might be better to put the trimmer on the larger resistor. I have found 6.2K with 1K in series to work well. I can even get many of the tubes working that others couldn't with this setup, for example Amperex 6DJ8. I think in the original T2 transistors these two transistors were selected to have certain properties which are now lost to time (maybe... See more on this below)

Using the third resistor has allowed me to get offset and balance in the mV range. The mystery I haven't solved is why others didn't need this and are still claiming 0 offset and 0 bias. Seems that transistors would have to be chosen exactly the same way as Stax did for this to happen. Regardless, evidence is there for example with Kerry's experience of setting things at 6.55V and everything zeroing out.

Would it be possible to get a survey of your voltages across the 6.2K and 510R resistors? This will give the gain of the transistors. It would be useful (for me) to compare this to your original T2 as well. If the gain is as low as the datasheet suggests, I can't see them working in my T2 which means perhaps I have some other marginal component that is causing "problems" I haven't found. Again, just don't see a way for this circuit to be as well behaved as it is for people.