• The move to the new server is done. There are some software and database maintenance updates in process. This has us passing the hat around to help out. We appreciate any donations. Seriously, even a dollar helps. The payment page may be found here - https://www.audiokarma.org/support.html

SX-950 amp board removal update--

Detailman

Detailman

Addicted to tweaking
Just wanted to post an update on an earlier question of mine.
Thanks go to Echo Wars for making my life easier with board removal
and also to Mark The fixer for instructions also.
Now for progress.
Replaced all the burned resistors and pulled the 8 outputs.
2 bad 427
1bad 557
should I order all 8 MJ to keep them the same ?(probably the right thing)
Or just replace the bad ones?
The two to-220 transistors were also blown as expected but I pulled them but didn't replace yet cause I have a question.
When I took them out neither one used an insulator. Just plenty of heat sink compound that was actually still soft!!!!!!!
Any other 220's I've had experience with have used an insulator.
I thought they may have melted but I pulled the good side and none there either.
I assume this is the way it is.

Pics of burned area this thread
http://www.audiokarma.org/forums/showthread.php?t=131560

I guess I should probably finish the board and do the protection board and test it before I order Output transistors.
Sound right?

When I finish the amp board I'll repost before making measurements.
Thanks for the help so far.
 
Register to hide this ad
The TO-220 heatsinks don't contact anything else, hence, no need for insulation of the collector tab.

I'd replace at least all the output devices on the bad channel. Up to you whether you want both channels to match each other.
 
3 dead outputs of 8? yes I would do all 8.. about 37 bucks shipped from mouser.

each to-220 has it's own heatsink, so as long as they don't touch anything they're ok. You COULD add an insulating kit: mica flat washer and plastic shoulder washer, it won't hurt anything ... they just cheaped out.

As for testing, the protection can only be partially tested, and I would not operate the amp without ALL the transistors connected. There's enough power connections laced through there that it would not be wise (there's no feedback without them all in place). Then I would first carefully check to see that the power supply was behaving itself.


In the amp with all resistors testing ok, diodes all check ok, all transistors test ok (or swap in all new, all that's left is the cheap stuff) I would put it ALL together. The level of testing is tedious, but at the end you KNOW everything is ok. Then I would dial the bias pot down to zero ohms (minimum bias) and slowly crank the thing up on a variac, letting the variac's current reading settle back to zero each time I increased it. IF there's a soft start relay it could induce a little intestinal distress in the tech as it clicks in. The key is if the current goes up quickly in sync with the dial and does not drop. If that happens there's a problem, and the damage should either be eliminated or limited by this procedure.

edit: drat, too much research, EW beat me!!
 
EchoWars said:
The TO-220 heatsinks don't contact anything else, hence, no need for insulation of the collector tab.

I'd replace at least all the output devices on the bad channel. Up to you whether you want both channels to match each other.
Click to expand...

Thanks again EW.
If I'm ordering I might as well spring for the whole 8.
Appreciate the clarification on the 220's.
I'm sure more help will be needed as I progress.
 
markthefixer said:
3 dead outputs of 8? yes I would do all 8.. about 37 bucks shipped from mouser.

each to-220 has it's own heatsink, so as long as they don't touch anything they're ok. You COULD add an insulating kit: mica flat washer and plastic shoulder washer, it won't hurt anything ... they just cheaped out.

As for testing, the protection can only be partially tested, and I would not operate the amp without ALL the transistors connected. There's enough power connections laced through there that it would not be wise (there's no feedback without them all in place). Then I would first carefully check to see that the power supply was behaving itself.


In the amp with all resistors testing ok, diodes all check ok, all transistors test ok (or swap in all new, all that's left is the cheap stuff) I would put it ALL together. The level of testing is tedious, but at the end you KNOW everything is ok. Then I would dial the bias pot down to zero ohms (minimum bias) and slowly crank the thing up on a variac, letting the variac's current reading settle back to zero each time I increased it. IF there's a soft start relay it could induce a little intestinal distress in the tech as it clicks in. The key is if the current goes up quickly in sync with the dial and does not drop. If that happens there's a problem, and the damage should either be eliminated or limited by this procedure.

Appreciate it Mark.
So you don't think it's wise to power up without everything all together huh?

I've got a copy of the SM so it should be helpful.
I haven't heard this 950 playing except as pre amp as I bought it this way but I see a relay so I assume there is speaker protection?

You all have been a great help and this is a great learning experience for me.
Just finally finished my 2230 after screwing up a good item myself so that taught me a lot.
Hope not to repeat my mistakes on the 950 so I'm taking it slowly.

Take care.

edit: drat, too much research, EW beat me!!
Click to expand...
LOL
 
Last edited:
I've ordered MJ's for replacements and other parts suggested by Mark.
I've already replaced all four To-220 transistors and the burnt resistors.
Just waiting till parts arrive from Mouser.
One question?
When reading thru the SM I noticed procedure for setting bias etc. uses two jumped rca's with 5.1kohm resistors. I've got 4.7 and 5.6. Are they okay to use or do I have to series some to get to exactly 5.1K?
Just preparing ahead. Anxious to get this thing singing.
Like a kid in a candy store.
 
Just pull the pre-out power-amp jumpers. When there are none in a unit, turn to aux and go to minimum volume.
 
These AREN'T phono inputs where sub millivolt signals are significant, but rather range from hundreds of millivolts to near a volt.

Normally there is minimal pickup. but if you were to shove a wire in there and touch it, well ..... ouchhhhh......

I've used zero ohm shorting junpers, 5k shorting jumpers and no shorting junpers without seeing any difference.

And I have a Broadcast Band FM station somewhere in the 88-91 mhz range nearby that blasts in significantly hotter than ANYTHING else on my spectrum analyzer, even with just a 24 inch piece of solder shoved into the input connector as an antenna!!
 
markthefixer said:
These AREN'T phono inputs where sub millivolt signals are significant, but rather range from hundreds of millivolts to near a volt.

Normally there is minimal pickup. but if you were to shove a wire in there and touch it, well ..... ouchhhhh......

I've used zero ohm shorting junpers, 5k shorting jumpers and no shorting junpers without seeing any difference.

And I have a Broadcast Band FM station somewhere in the 88-91 mhz range nearby that blasts in significantly hotter than ANYTHING else on my spectrum analyzer, even with just a 24 inch piece of solder shoved into the input connector as an antenna!!
Click to expand...

Gotcha
Thanks for clearing that up.
I've got lots of shorting rca's around but I'll probably just leave empty.
Be back when parts arrive.

Arrivedrci.
 
OK
Next update.
Got my parts from Mouser and installed new finals MJ's 93 94's. Also replaced all burned resistors and checked all diodes in burn area. OK.
Also replaced the 2sd381's and 2sb536 with suitable replacements on both channels. Also small signal transistors a726 all four.

Dialed VR1 2 3 +4 all the way down and fired it up. VR1 and 3 adjustments no problem.
VR4 ok 20mv. But vr3 looks like it was charred and I didn't notice it before and it is very jumpy. Can't get it to settle near 20mv. It's either 4 or 5mv and then it sort of jumps to four or five hundred mv or so or higher.
Didn't use any jumpers or shorting rca's when making measurements.

Nothing smelled like burning etc. when I fired it up but when I let it idle for a while and came back and tried to adjust the bad pot again I could hear a slight hum when I turned the pot up or down. Up made it louder.
Even though no inputs were selected and no speakers either the heat sink was very hot to the touch as were the outputs. Couldn't leave my fingers on them at all especially the old bad side.
It's getting late but I wanted to post progress before going off to sleep.
Something is not right about this so far so all help is welcome.
Don't think I did anything wrong when installing new parts but you never know.
 
Detailman said:
OK
Next update.
Got my parts from Mouser and installed new finals MJ's 93 94's. Also replaced all burned resistors and checked all diodes in burn area. OK.
Also replaced the 2sd381's and 2sb536 with suitable replacements on both channels. Also small signal transistors a726 all four.

Dialed VR1 2 3 +4 all the way down and fired it up. VR1 and 3 adjustments no problem.
VR4 ok 20mv. But vr3 looks like it was charred and I didn't notice it before and it is very jumpy. Can't get it to settle near 20mv. It's either 4 or 5mv and then it sort of jumps to four or five hundred mv or so or higher.
Didn't use any jumpers or shorting rca's when making measurements.

Nothing smelled like burning etc. when I fired it up but when I let it idle for a while and came back and tried to adjust the bad pot again I could hear a slight hum when I turned the pot up or down. Up made it louder.
Even though no inputs were selected and no speakers either the heat sink was very hot to the touch as were the outputs. Couldn't leave my fingers on them at all especially the old bad side.
It's getting late but I wanted to post progress before going off to sleep.
Something is not right about this so far so all help is welcome.
Don't think I did anything wrong when installing new parts but you never know.
Click to expand...

DON'T TURN IT BACK ON!!!!!!!!!!!!!!!!!!!!!!!!

At this point the robot would be saying "DANGER!!! WILL ROBINSON DANGER!!!!"

You did nothing wrong, but VR3 is poised to destroy ALL your hard work. The heatsinks should ONLY get slightly warm after a while.

VR3 is in exactly the right/wrong place to do the most possible damage.

Since you will be REPLACING VR3, you may as well do VR1, 2 & 4 as well.

VR1,2 (10k ohms) go in with the wiper at the HALFWAY POINT.

VR3 and VR4 (100 ohms) go in with the wiper set to zero ohms. The bias reading (read across (12 & 13) or (27 & 28) ) of 20 mV shouldn't go over 50 millivolts even accidentially.

At this point I don't have any parts suggestions (i'm lucky enough to just get BACK online... and find this, I'm still laid up although boredom is starting to outweigh the irritation that shut down the tv ) and I hope someone will chime in with what they use/used. I would recommend 10 turn miniature trimmers although 25 turns may be more common.

UNTIL you get the pots it would be better for the amp, at the BAD VR3 position, to short the pads to zero ohms and remove that AMP-KILLING pot....
 
Thanks for taking the time Mark.
Do hope your situation improves. I know it's starting to get pretty chilly where you are and it seems like this is the time when sickness hit me when I lived up north. I don't know if that's the case with you but I hope you get it under control.

I may even have some pots around or can find some out of carcasses of old radios etc.
This morning I was looking at the schematic and to me it looks like vr-3 controls the resistance between the collectors of 2sa818 and 2sc1628 thru the bias diode.
I'll check that diode.
I also didn't replace either of those transistors but checked them out of circuit and they measured ok. I have the replacements(818-1142ystu and 1628-2682ystu from Mouser) that you suggested for them but the pinouts were different and the old ones tested ok so I put em back in.
Aside from the pot I'm thinking maybe the transistors are weak or that diode is blown.
Thanks again and I won't do anything till I hear from you.
 
markthefixer said:
DON'T TURN IT BACK ON!!!!!!!!!!!!!!!!!!!!!!!!

At this point the robot would be saying "DANGER!!! WILL ROBINSON DANGER!!!!"

You did nothing wrong, but VR3 is poised to destroy ALL your hard work. The heatsinks should ONLY get slightly warm after a while.

VR3 is in exactly the right/wrong place to do the most possible damage.

Since you will be REPLACING VR3, you may as well do VR1, 2 & 4 as well.

VR1,2 (10k ohms) go in with the wiper at the HALFWAY POINT.

VR3 and VR4 (100 ohms) go in with the wiper set to zero ohms. The bias reading (read across (12 & 13) or (27 & 28) ) of 20 mV shouldn't go over 50 millivolts even accidentially.

At this point I don't have any parts suggestions (i'm lucky enough to just get BACK online... and find this, I'm still laid up although boredom is starting to outweigh the irritation that shut down the tv ) and I hope someone will chime in with what they use/used. I would recommend 10 turn miniature trimmers although 25 turns may be more common.

UNTIL you get the pots it would be better for the amp, at the BAD VR3 position, to short the pads to zero ohms and remove that AMP-KILLING pot....
Click to expand...

I went ahead and pulled the vr-3 and it is shot. open. vr-4 is ok.
Looking for 100ohm pot in my scrap stuff but closest I could find was a little large and I broke it trying to make it work. Of course Rat shack doesn't have any that size.
I've got more stuff to go through so I may find a workable one.
Also changed the 818 and npn that goes with it to new subs and made the pin outs right. When I find the trimpot I want to check and see if it's ok and then I'll change the same stuff on the good channel.

I hope I don't have to make another order just for those pots.
 
Detailman said:
Thanks for taking the time Mark.
Do hope your situation improves. I know it's starting to get pretty chilly where you are and it seems like this is the time when sickness hit me when I lived up north. I don't know if that's the case with you but I hope you get it under control.

I may even have some pots around or can find some out of carcasses of old radios etc.
This morning I was looking at the schematic and to me it looks like vr-3 controls the resistance between the collectors of 2sa818 and 2sc1628 thru the bias diode.
I'll check that diode.
I also didn't replace either of those transistors but checked them out of circuit and they measured ok. I have the replacements(818-1142ystu and 1628-2682ystu from Mouser) that you suggested for them but the pinouts were different and the old ones tested ok so I put em back in.
Aside from the pot I'm thinking maybe the transistors are weak or that diode is blown.
Thanks again and I won't do anything till I hear from you.
Click to expand...

Patience... don't risk all your work by using old crap in the most critical part of the circuit.

The 0 to 100 ohms SLIGHTLY increases the bias voltages at your pnp and npn output transistors. You will notice the outputs sit at +0.6v and -0.6v. This is the point where the transistor begins to operate. Since each transistor's voltage is different, it is necessary to incorporate an adjustment to set that point correctly. With the fact that there are other circuits in between the VR3 adjustment and the outputs, the voltages opposite each other at the collectors of Q5 and Q7 have to be +1.2v and -1.2v which is generated mostly by the bias diode D5 (STV4 about 2.4v) and increased slightly by VR3.

There is a current flowing through (Q5, D5, VR3, Q7) and audio makes that 1.2 swing all over the place, with the other voltage following at 2.4v lower (VERRRRYYY IMPORTANT!!!) and this CONTROLS the output transistors through the audio waveform cycle.

BUT when that circuit (Q5, D5, VR3, Q7) is broken (think of it as a spring anchored at both ends... in balance ), huge voltage differences happen (+20 -20 etc) turning on BOTH outputs at the SAME time in an uncoordinated fashion. The transistors get red hot internally and self destruct.

This is why VR3's reliability is very important.
 
Great explanation.
I'm getting kind of frustrated as you sensed.
I'll do some checking and see what pots I can find.
There are plenty of choices online.
Hope the heat didn't hurt the new ouputs already.
Had it turned on originally for about 10 min. but won't put it on again till I get the vr.
 
Bourns 3266 series @ 2.92 ea... 12 turns sealed cermet 1/4 inch square, .177 thick 2.5mm triangular spacing
3266p sits flat on the board
3266w sits up like a briefcase on the board with the adjust straight up
3266x sits up like a briefcase on the board with the adjust parallel to the board (side adjust)

3299 series is 25 turn @ $ 2.85 ea
lots more of the same , different manufacturers, perhaps slightly lower prices and some prices a LOT higher.
 
I was looking along these lines.
Mouser p/n652-3309P-1-101 less expensive version but haven't done anything yet since it's friday evening. I've seen a couple others in the same type but different maker.
I'l go back and look at your suggestions and learn the difference.
Thanks
 
You must log in or register to reply here.
Back
Top Bottom