SX-1250 Main Outputs Fail

markallen

Luddite Tendencies
This is actually the unfortunate story of an SX-5590. Quickly established the L channel main outputs were shorted. When these shorted transistors were removed, it would come out of protection (two clicks) but no tuner or panel lights.

Went through the stabilizer board first, and found three failed transistors, as well as a failed zener diode and several caps either failed or out of spec. All were replaced (including all E-caps) per parts list found here on AK. As long as I was in there, I pulled the protection board and recapped that also even though that appeared to be working.

Next, I turned to the failed L channel amp. Checked each transistor off the board. Surprisingly, found all good. Replaced all E-caps on both amps boards anyway.

Obtained two pairs of new NEC outputs (more on this below), cleaned board pins, reassembled everything. Using a DBT w/100w, gave it power - and it lit up, FM tuner came to life. Removed from DBT, set stabilizer voltage to 65, perfect. Set amps per manual, all dialed in without a hitch. Used 65mV rather than the 100 specified, as it seems this is recommended after researching here.

Finally hooked up speakers, and with some cautious optimism, gave it a bit of volume. Sound on both channels but - momentarily I heard a pop from the L channel area, minor puff of smoke, and of course, sound ceased. Crap. Removed the four outputs, all failed. Powered up (w/ DBT), and the R channel seems okay.

Here's the question. Will counterfeit transistors fail easily? Thing is, I got them from eBay, and though marked NEC, who knows? I thought I'd done due diligence on the amp board, but there's likely a problem I didn't catch. Help! Where do I go from here? Thanks for any help.
 
Sadly nothing to add, just watching for my own edification since I own a SX-1250 with a weak channel I hope to get fixed soon...
 
I cannot comment on the quality of the replacement components. When I went through my 1250 I pulled the outputs and installed MJ21193G/94Gs and never looked back.

Did you replace the trimpots on the stabilizer board? What about those on the power amp boards?
 
To answer your question about the counterfeits.

Yes, typically they fail very very easily...you'll be able to bias it up okay and then when you try any decent output swing, they will let the magic smoke out...
 
It would be nice to see the die of the eBay transistor vs. a known genuine NEC. I would bet it is much smaller than the real deal.
 
I cannot comment on the quality of the replacement components. When I went through my 1250 I pulled the outputs and installed MJ21193G/94Gs and never looked back.

Did you replace the trimpots on the stabilizer board? What about those on the power amp boards?

I replaced the stabilizer trimpots, but not those on the amp boards.

To answer your question about the counterfeits.

Yes, typically they fail very very easily...you'll be able to bias it up okay and then when you try any decent output swing, they will let the magic smoke out...

So it's likely (or just possible) it was the outputs? Would the fact that it biased up easily indicate the amp board is doing its job correctly? Not sure exactly how to proceed to establish that part of it. Prefer not to blow up any more outputs!
 
yes,yes and just get the proper Onsemi MJ2119x replacements. if you are scared, just use one set of o/p bjts to start with, then add the others later. You will be hard pressed to blow up a set of onsemi bjts unless you are cranking it into a low Z load.
 
I replaced the stabilizer trimpots, but not those on the amp boards.



So it's likely (or just possible) it was the outputs? Would the fact that it biased up easily indicate the amp board is doing its job correctly? Not sure exactly how to proceed to establish that part of it. Prefer not to blow up any more outputs!

OLD Trimpots can go open circuit, which would pop the whole output stage no matter what brand of transistors are used.
 
yes,yes and just get the proper Onsemi MJ2119x replacements. if you are scared, just use one set of o/p bjts to start with, then add the others later. You will be hard pressed to blow up a set of onsemi bjts unless you are cranking it into a low Z load.

By saying "use one set . . . add the others later", I am assuming you mean just equip the bad channel, then change out the other (currently working channel) later?

Anyway, sounds like I need a set on ON-Semi MJ2119x replacements - and new trimpots! Thanks all. Any other recommendations welcome!
 
By saying "use one set . . . add the others later", I am assuming you mean just equip the bad channel, then change out the other (currently working channel) later?

Anyway, sounds like I need a set on ON-Semi MJ2119x replacements - and new trimpots! Thanks all. Any other recommendations welcome!
No he means just use one complementary pair rather than installing all of them on one channel.
 
Either the MJ21193G/94G or the MJ21195G/96G pairings. Both are well suited for this.
 
If I understand this correctly: I only need to install one complementary pair in a channel and it will still output sound? Does it matter which sockets they're installed - like both in upper or both in lower? What about the amp settings?
 
I would double check the left amp once again. If good, install the right outputs and try it.
If it's good you'll know. And if not, the MJ's all around once you found the culprit(s)(?).
I might have a set of NEC's for the 1250 somewhere.
 
I would get good parts from Mouser.
I never buy off Ebay unless is some weird thing I can’t cross and I’m working on something that isn’t too important. I know of one good seller on Ebay but they only have what they have.

I would go through the funny amp with a fine tooth comb. It was blown when you got it, you thought you fixed it and it blew again. As Mark said, it may not have anything to do with just the outputs.

That amp uses those unobtanium STV packs and if one of those is whacky, that may be the issue. Those can test OK on a meter but still be bad. We build replacements out of fast diodes.

I just took delivery of an SX-5590.
Totally rebuilt by the previous poster.
Love those black face.
Love the SX-1250 too. I seem to be collecting them.

Luck
 
I would double check the left amp once again. If good, install the right outputs and try it. If it's good you'll know. And if not, the MJ's all around once you found the culprit(s)(?). I might have a set of NEC's for the 1250 somewhere.

Thought I'd adequately checked the left amp by removing all the transistors and diodes and testing them off board. Not exactly sure where to go from there, once I've got new trimpots installed.

I would go through the funny amp with a fine tooth comb. It was blown when you got it, you thought you fixed it and it blew again. As Mark said, it may not have anything to do with just the outputs. That amp uses those unobtanium STV packs and if one of those is whacky, that may be the issue. Those can test OK on a meter but still be bad. We build replacements out of fast diodes.

If you can't test it reliably on a meter, how do you know if the STV has failed? How difficult is it to make the replacements you reference?
 
New STV-3H varistors are available on ebay from a seller in Germany. I've used them and they were fine.
 
Your the one asking with a wonder.
We're giving our best advice which is based on experience.
Rechecking the board is what all of us would do knowing how easy it is to miss something.
Outputs shorting on their own is extremely rare.
If the STV checks good it is considered good.
Time saved, would be the question on whether you can trust the eBay purchased outputs.
 
Last edited:
If you can't test it reliably on a meter, how do you know if the STV has failed? How difficult is it to make the replacements you reference?
These are 3 diodes in series. Run 1-10mA DC thru them and measure the voltage developed. Assumption is around ~0.6V per junction, so you are around ~1.8V in total.
EW posted curves of the STV vs other diodes that he used to try to match the characteristics.
 
Back
Top Bottom