Repair and restoration of my STAX SRA-12S

[DefQon]

An update, the SRA-12S that had a dead transformer has been replaced with one from another 12S. 

 

Now I haven't touched anything on the two output boards except replace the 50kohm resistors.

 

I powered on the amp and all measures were fine and plugged in a spare SR-3 to test. Left channel works, right channel does not.

 

So it could be a bad HV transistor or trimpot balance going off, swapped the output boards and issue still persisted with right channel.

 

Made a measure for dc offset and balance.

 

Measure the working left side first and it was floating around 0.4-7vdc. Measured the right side and holy shit 195-225vdc, measured the dc offset on right channel (-/+R) and it stabilised at about 380-400vdc. What the fuck is going on?

 

Checked the underboard wiring and all is correct, tried turning the trimpots and no values changed.

 

Any ideas?

 

I have a feeling the trimpot has gone bad, as I did clean the boards last time, one is 50kohm and other is 500ohm. Which is dc offset and which is for balance?

 

 

Edited July 7, 2015 by DefQon

[DefQon]

Measured the trimpots, both are fine.

 

It's quite odd why I'm seeing raw 200vdc+ when measuring the balance via red probe in +R and black probe in -R and dc offset measure by black probe on ground post almost floating at 400vdc. But when I test the HV output 2sc1167's they are all fine. Measured all resistors and values within spec range. Swapping the boards doesn't flip the bad channel. 

 

Any ideas spritzer?

[spritzer]

You shouldn't see any DC on the output since you can't adjust the amp that way.  It is AC coupled so no DC should be on the output. 

[DefQon]

Ok well found out the problem, tried flipping the output boards again and a few times with my testing purposes only sr 3 the right channel would work, flipped the board and the problem replicated on the left side for the problems. This was a lesson to be taken out the book when I repaired my SRM 1 MK1 that ended up with a intermittently working HV 2SC1828 output device and a dead 2sk130. So i desoldered the transistors and doing from base to collector and emitter diode test since the 2sc1167 is npn I found the collector to have double the values of the same test done on all other transistors. So most likely case a faulty 2sc1167 on its last legs. Am a bit afraid the high hv might`ve taken out the 4 2sk30a gr and o grade jfets in the input side which are near impossible to source.

Sent from my LG-D855 using Tapatalk

[DefQon]

Man it took me almost half a year to source 2sc1828s when I was repairing the SRM 1 and only needed one. Fast forward 5 months and 3 show up on ebay from a US seller for 25bucks shipped. Can't find any 1167s at the moment. Only Ceitron Consolidated has it but not sure if from Toshiba. Going to email Steve

Sent from my LG-D855 using Tapatalk

[Tinkerer]

You shouldn't see any DC on the output since you can't adjust the amp that way.  It is AC coupled so no DC should be on the output. 

I have a dumb question then. How do you measure the balance and offset on this?

[spritzer]

On the top side of the large resistors.  Balance them together and then the offset (well it isn't the offset at all) is adjusted to 390VDC. 

[Tinkerer]

On the top side of the large resistors.  Balance them together and then the offset (well it isn't the offset at all) is adjusted to 390VDC. 

Thanks. That really helped out. I didn't realize whoever put in the new transformer before me was undervolting the boards at 320VDC. Had to adjust the pot in the back of the amp to get the voltage up on the output boards before I could get those numbers.

 

EDIT: I just saw on the previous page that 290VDC had been recommended before. Don't know how I missed it when I must have read this thread five times over the last few months. My replacement power resistors are rated 10 watt and the replacement transformer is much beefier than the old one so whatever I throw at either of them shouldn't be a problem I think. Pretty sure everything on the output cards are rated 500V+ as well. Should everything still be fine at 390VDC?

 

Just want to make sure I'm not missing anything.

[spritzer]

Ehhh might have been 290.  I was just doing this off the top of my head but the full data should have been posted here by me at some point.  290V sounds plausible as it is half of the B+ minus the natural loss. 

[Tinkerer]

Thanks. Everything's running great now. EDIT: Set output boards to the proper 290VDC and keeping them there.

 

I'm mostly posting to show what mine looks like since it's one of those that's certainly had an interesting past. Something may really off that I'm not seeing.

 

As far as I can tell, this is what was done to it at some point before I got ahold of it:

New electrolytic caps

New bando transformer with winding combinations for any mains voltage

Pro bias breadboard epoxied to chassis and pro socket added

Phono board removed for some reason

Main transistors replaced and heatsinks added for them

 

Then this is all I did:

Rewired from 100v to US voltage

Replaced caps on output boards

Replaced 3/4 power transistors (the ones replaced were about 10% off and now they're all within 1%)

Replaced pots with multiturn ones

Reflowed everything on output boards

 

Edited October 2, 2015 by Tinkerer

[Tinkerer]

Well, I wound up needing some more help. The reason I was even replacing anything on this was because after a couple months, I started hearing slight static in the right channel that would come and go. It had been working perfectly again the last time I posted, but it's started up again. Here are the symptoms.

 

1. Problem switches sides when I switch boards.

2. Leaving the meter to read balance after it has been zeroed when it happens, I get a 4-15 vdc difference when I hear the static. Then it goes back to 0 when the static goes away.

3. It's not a cold solder joint. I must have reflowed everything twice not counting replacement parts.

4. It's not the two main transistors. It still did the same thing when I tried some replacements.

Any ideas? I'm hoping it's not the smaller transistors. EDIT: Swapped jfets in from the 30db extra gain board (2x K30A O and 2x K30A GR), which accounts for all but one on the output board. Didn't fix the problem.

I also have some measurements for the resistors but they were taken on the board so not terribly accurate sometimes. When I replaced some, they usually would measure within tolerance when removed from the PCB.

 

Edited October 7, 2015 by Tinkerer

[DefQon]

Tinkerer I saw your posts with the pictures on Head-fi and was very close to creating another account there to PM you with some information on the transformer specs.

Anyway it is not surprising that you're getting some problems because the transformer is not even for the 12S, the output boards have been modified with different output transistors that don't even belong there.

The transformer is the exact same unit found in the SRM-3 amplifier and the original output transistors are TO-3 cased Toshiba 2SC1167, near impossible to obtain but you can find them if you know where to look as I purchased 5 NOS replacements for 1 of my output boards for my 12S which is still on the bench for repairs.

The static noise you hear on the channel is either an out of spec leaky resistor or bad output transistor. The input jfets (2sk30's) would not be the problem. You need to source the original TO-3 transistors and replace those ones (I have some NOS stock I can sell if you need some). Also you can't measure the resistors as it is as they will all show up out of tolerance especially the 150kohm resistors because the 500ohm trimpots adjust the balance of the output. To measure the resistors properly you need to isolate it from the trimpots or desolder them one by one and measure it that way, and most likely it will measure fine (but the dielectric properties can be out of spec when high voltage and heat is presently going through the resistors). You should only see 290-300vdc between the base and collector of the output transistors. I'd replace all the resistors with uprated equivalents and replace those transistors with original Toshiba TO-3 units. 

Btw since I can't measure the winding voltages for my SRM-3 due to a broken transformer, what are the voltages and mA current supply for the secondary side windings?

Edited October 8, 2015 by DefQon

[Tinkerer]

Btw since I can't measure the winding voltages for my SRM-3 due to a broken transformer, what are the voltages and mA current supply for the secondary side windings?

230-0-230 and 18-0-18 IIRC. I think the original was supposed to be 260-0-260 and 26-0-26 looking at the boards? But it seems to be putting out what it needs to with the pot adjusted in the back. The pro bias measured 579v in front of the ballast resistor and the normal was adjusted to 230v. Don't have current readings unfortunately. Will try to get some when I get a new battery for my multimeter.

 

Also just bit the bullet and grabbed a full set of higher rated resistors from mouser except for the power ones I already replaced. All dale and xicon so quality should be no issue.

[DefQon]

Can you show me a picture of what voltages is where on the winding side of the transformer? I know the top row of windings is primary and the bottom is secondary.

 

Thanks

 

And yes the original transformer is a CT 260 and CT 26 volt transformer. Not sure about the pot you are talking about?

[Tinkerer]

There's a pot you can use to adjust some stuff in the PS by the two big caps next to the transformer.

 

As for the transformer, the primaries are like a mirror of the one on the SRM 1, right to left instead of left to right

 

120 100 G  120 100  G CG

  |___|___|     |____|__|__|

 ________     _______

230 CT 230   18-CT-18

[DefQon]

You mean the pot on the daughter board? That pot is to control the +-/15v supply to the Phono, front input PCB with the relay and Intermediate Stage ISA cards.

[Tinkerer]

It affects the amp boards as well. Or maybe it just does on mine since mine is missing the phono board. Anyway, mostly just posting the transformer current read 250ma and 50ma on the secondaries. Figured it would be a good time to take the readings as my mouser order just came in for a full output board rebuild minus the fets and my soldering iron was warmed up anyway.

 

EDIT: Installed full resistor replacements, all Xicon and big ass Dale RN60D's. As everyone probably figured, the static was after all a bad resistor. Still have the original TO-3 replacement transistors coming in too thanks to some help from you guys, but it sounds great right now with the SR-007. I actually like it better than my old feedback-modded 727A. We'll see if the original configuration changes that.

Edited October 16, 2015 by Tinkerer

[DefQon]

So I was bang on the money with the resistors. Your transistors shouldn't take too long to arrive as it only took less then 2 weeks for me.

[DefQon]

How did you adjust for 290vdc from the collector? The 50k or 500ohm pot on the output board?

I still can't fix this damn unit, have replaced everything on the PSU boards, everything on the output boards except the trimpots, 910ohm input resistor, the input stage single 2sk30 jfets and the 2sk734. I can't have the IS button not pressed down without huge distortion, when IS is engaged any song with bass will distort the transducers past 2'O clock on the volume dial. With the black probe in the ground post at the back, meter adjusted for vdc if I measure the case of the HV transistors with the amp on and not playing any music it measures -250vdc but I can adjust for -290vdc through the pots. If I put the dmm to vac it reads about 130vac. Why is it negative on vdc? Although both channels read -290vdc once adjusted for it just doesn't make sense. 

 

I'am tempted to bring out the hammer.

Edited October 26, 2015 by DefQon

[Tinkerer]

The 50k is the balance pot for each card. The 500 pot is to adjust the voltage. I got distortion kind of like that when the right channel card was unbalanced. I also measured balance and voltage from the top of the power resistors and not the transistors, but I don't know if that makes a difference.

 

Why are you turning on the extra 20db amp board though? Isn't that supposed to add lots of extra noise? I know it raises the noise floor on mine. Dead quiet without music at any volume level without it, faint static at high levels with it on.

[charlo89]

Are you sure that the channel is dead and it is not a switch issue?  Measure the top of the large resistors for DC voltage.  You should have roughly 0VDC between them and each one relative to ground is +290VDC. 

 

Replacement parts for the output devices are all but impossible to source.  You need low Cob which is the truly hard part. 

 

Did you fix the problem?

 

Sorry for the delay, and thank you for your help.

I was in paris for two months, and when I come back the amp worked, I adjusted the offset and checked the DC voltage between the resistors, I found +6v on the left channel, +3V on the right channel, and arround 285-295V between the resistor and the ground.

I bought a SRM-727A last week and I found +12V between R+ and Ground, it was very difficult to adjust this tension to something under 1v, it was the same for R+/R- L+/L- and L+/Ground.

I do not know if this is normal, but I found +290V between the Bias and the ground for the 727A and +360V between the bias and the ground for my SRA-12S. (on the jack for both)

Edited October 27, 2015 by charlo89

[spritzer]

Higher bias when loaded like that is normal on the SRA-10/12S. 

[Tinkerer]

Just thought I'd chime in that I got my NOS 2SC1167's in today and replaced the stopgap C3148's. Sounds good. Voltage drifts more than it used to though, but still less than a volt. Haven't listened enough to tell how different it is. Now the only original parts left on the output boards are the small fets. I think the old girl is about as good as she's gonna get at this point.

 

Now off to work on the KGSSHV and try not to blow it up.

[DefQon]

What fets did the previous owner change them to? Should be 2 2SK30A O and GR idss grade jfets and a single Y idss 2SC734.

Edited October 28, 2015 by DefQon

[Tinkerer]

What fets did the previous owner change them to? Should be 2 2SK30A O and GR idss grade jfets and a single Y idss 2SC734.

They're the exact same as the schematic, just as you listed. But I think they were slightly newer production as the O, GR, and Y identification were on the side label instead of the top like the originals on the IS amp board.