Repair and restoration of my STAX SRA-12S

[chinsettawong]

I just love to see your rebuild. Keep more photos coming.

Wachara C.

[spritzer]

Glad you like it. Here are some more:

The RCA sockets all soldered and hopefully all the correct wires went to the right spots... The PowerKon is a tight squeeze and required some cutting of the support bar next to it but it went better than expected. I still need to install two 392K resistors on the PRE OUT1 sockets as that's how this unit was configured.

100nf/1000V caps added to the B+ line right on the sockets. This is what we do on the most current Stax amps and given the rather basic PSU design at hand here it makes sense. I also used the empty tabs on the sockets to install a voltage divider for the Pro bias. The B+ is around 640-650VDC so a 10% decrease is needed to get 580V and 22K/220K does that perfectly. I also tweaked the normal bias voltage divider so that it now outputs 230V instead of the stock 200V. Both outputs have a 100nf cap between the divider and 5M ballast resistor.

[chinsettawong]

Very nice! More pictures please.

[John Buchanan]

Beautiful work, Birgir. If the SRA-10S runs as hot as the SRA-12S, everything in there will get mighty hot.Mine never sounded as good as the SRA-3S however. Is the SRA-3S circuit the same as an SRM-T1?

[spritzer]

AFAIK the SRA-10S is identical to the SRA-12S. The service manual shows no differences aside from what one sees over a long production cycle. This unit is relatively old which is why I'm going totally overboard in rebuilding it. The design is that far off from the old Stax tube amps, in this instance we have just a low voltage SS stages feeding the outputstage which runs off a 640V B+. Simple resistor loading of the output stage so not much anybody can do there and the PSU is a bit too small for my liking. Since I will be ordering a custom transformer for this one I'm contemplating PSU changes. Now the space available is pretty much zero but a simple regulation like the old KGSS PSU should work here and the voltage can be managed simply by regulating each of the two stacked supplies independently. In stock form it is just a 260-0-260VAC transformer feeding a bridge and two stacked 47uf/350V caps. I can easily fit much larger caps so that will be the first step but it is always better to add regulation...

As for the SRA-3S, the basic building blocks of the amp maybe similar to the T1 but the circuit is quite different. The T1 is fully DC coupled but the SRA-3S has both input and output capacitors . Now on the amp I have here now I changed all of these (plus all other caps just to be safe) and it sounds just lovely driving this SR-Lambda I just made from Sigma parts.

[luvdunhill]

Birgir, how much current through the plate resistors? I might have something you could use to upgrade these... Any metal around where the resistors are?

Edited September 24, 2011 by luvdunhill

[spritzer]

The amp was utterly fubar when I got it so no measurements possible but the math places it around 13mA. This goes in hand with the schematic I found online.

No metal around it except the metal steel brackets used to secure the HV boards in place. Here is a picture of the inside which shows how bloody small this thing is about 30cm long by 17cm wide...

Now my bloody soldering station just died...

Edited September 25, 2011 by spritzer

[livewire]

Man that shit looks old, I see it has good vision.

Is your relic older than thou Spritzer?

Are you gonna change the flux capacitater too?

[spritzer]

This one has about 10 years on me so yeah, it is rather old.

New soldering station landed today so I'm back in business and as it turns out, the pre and phone stages are nothing like the schematic and service manual. This is truly a very early unit...

[spritzer]

Slowly coming together...

[as-audio]

The diagram, shown in post 23, has some mistakes.

I checked part PB-300, single channel output boards.

 

- The connection from R324 and C303 to the drain

  of Q307 is wrong, these are on the collector.

 

- "R311" appears two times, on Q310 drain read R307.

 

There may be more errors that I did not see.

 

Please note that the bord layout for Q304 and Q305 is for

a japanese bipolar transistor, but they installed the JFet

devices shown on the diagram which have a different pinout.

 

The circuitry of this output amp has no stable operation and 

I will probably comment on it later.

 

The host for the diagram picture is not always available. I

can not upload it here again, because the file size is twice

of that allowed. Available from me on request.

[Voltron]

Thank you for your timely and important updates.

[as-audio]

Thank you for your warm acknowledgnet.

Everybody who is in need for repair of this

item will appreciate your contribution.

[morphsci]

So that would be like 2 out of 8,000,000,000 people?

Edited September 17, 2013 by morphsci

[livewire]

They're beatin feet to necroposter's door as we speak!

 

The circuitry of this output amp has no stable operation and 

I will probably comment on it later.

 

Could ya? Would ya?

Maybe it's because there's a bomb between the 2nd & 3rd stages?

Huh? Didja think of that?

 

[as-audio]

So that would be like 2 out of 8,000,000,000 people?

Is this to point out the relevancy of you or this forum? Thanks.

[as-audio]

They're beatin feet to necroposter's door as we speak!

 

Please translate.

 

Could ya? Would ya?

Maybe it's because there's a bomb between the 2nd & 3rd stages?

Huh? Didja think of that?

 

I did not find a solution yet.

 

What do you mean by "a bomb between the 2nd & 3rd stages"?

[as-audio]

 

[wink]

Some on this site don't like newbies bumping 2 year old threads.

 

Your last post has left some of us speechless too.

[spritzer]

Well the schematic being wrong doesn't surprise anybody, this is Stax after all.  They always leave errors in the schematics. 

 

This amp was also in production for a long time and during that time semiconductors came of age so to speak. 

 

Ohhh and you can still buy the design today.... it's called the KingSound M-10 and is fucking awful. 

[kevin gilmore]

don't forget that the kingsound has been "mosfetized" also known as "cavallied"

[morphsci]

Is this to point out the relevancy of you or this forum? Thanks.

No, it is to point out that you are a douchbag cunt™ with an overinflated sense of importance. K Thx Bye.

Edited September 18, 2013 by morphsci

[spritzer]

don't forget that the kingsound has been "mosfetized" also known as "cavallied"

 

 

Ohh so true.  Give me BJT's or give me death!!! 

[DefQon]

Sorry for reviving this thread but I felt it best to consolidate all posts/faq's regards this beast in this single thread since schematics and other technical discussion have been made in the previous pages over a year ago.

 

So I've managed to snag a 12S from a buyer who didn't know what it was for 100 quids (bloody bargain). The amp works fine except some hissing and collapse of the left channel, which seems to work intermittently. So I popped off the hood (which look's like somebody tried to scramble eggs on it), the pair of 51kohm 5watt red resistors on each output board have left scorched burnt marks but measure within the 5% of the specced resistance.

 

Consulting with the schematics and pictures of the restored 10S pictures few pages back by spritzer, where you mentioned you had a voltage divider of two resistors in series for pro-bias output connecting the output stage green pcb slot connectors underneath, 22kohm and 220kohm was the 22kohm originally 2.2kohm? The spec sheets says 2.2kohm and my 12S particularly confirms this as well with a parallel 100uf 25v axial cap next to the 2.2kohm 1/4 resistor. What is the 100uf 25v axial cap for? I've disassembled the amp, cleaned the brass chassis and reassembled the slots, trafo and other stuff such as rewiring the sockets and have ordered parts only to rebuild the output and both power supply boards completely.

 

Another thing on the schematics and the PS board underneath the amp, the rectifier is made by 4 RA-1 diodes and then another 4 round diodes (similar to the one I replaced in my SRM-1 except bigger) what can the RA-1 and other round diodes be replaced with? The usual 1N4007 suspects applicable in this case?

 

Thanks

[spritzer]

The voltage divider is off the B+ so you need to lower it from +640V down to +580V.  The resistors do that but remember to add the ballast resistor and a filter cap before the divider. 

 

No idea what parts you are talking about there though. 

 

UF4007's will work perfectly for that.