an sx-950 im trying to work on. any advice is appreciated.

just wanted to give a little update on this, I know i take a painfully long time to actually get any work done, or report back on it. but as ive mentioned (and/or made obvious) in other posts this is all a learning experience for me. which I do primarily as a hobby because I find it fascinating.

SO, to that end, ive learned how to read the little diagram on the boards for transistors, that has made it a lot easier to know which pin goes where.
so heres basically what has happened since my last post.

I had 8 transistors installed incorrectly, which blew 2 fuses (the ones from the orange wires of the transformer), ive put all transistors in the correct places, i triple checked that all transistors are the right ones for the spot, they were, they are all installed correctly, and my solder joints look good. so i replaced the fuses, put some cardboard inbetween the still loose amp board so the heatsink doesnt arc against the chassis, shook a little bit with nervousness, plugged it in and turned it on. the 2 or 3 seconds it takes feel like an eternity on the first power up. when i hear, click.

it comes out of protection now.

I still have to fix a wire that broke, one of the ones that go to the outputs broke, i guess spin that around the pin a bit and solder it on, that should be good right.
i have to start reassembling things, have to adjust the bias and offset, question on that, i dont see how i can adjust the offset without the outputs connected, but what about the bias? ive never done that before.

things are looking good so far though.

disclaimer: im working on this one for a friend who knows i take forever, but that i also dont charge;/
 
It's good that you've stuck with it. I've had several projects which I've set aside for some time and returned to later after "clearing the cranium" for a fresh perspective.

For the broken wrapped connections, a twist and solder should take care of that.

As for adjustments, the offset can be set without speakers connected. You can also forget about the service manual instructions to install 5.1KΩ resistors in the input jacks. Simply turn the volume to zero and set the offset voltages.

Do the same for the bias (idle) currents, but understand that it may take longer to warm up the output stages for final adjustment - again at zero signal and zero volume. You will want to check initial bias and may perform several adjustments as the unit reaches operating temperature.
 
ive been checking my work as i go and ive run into something im not sure about...

so after installing each output transistor i measured the case of said transistor to the bracket of the heatsink to check for shorts, there were none, but i decided to measure it again once i screwed the heatsink back into the chassis, is there supposed to be ~0.2 dropoff between the case of the outputs and ground? because there is..

anyway, heres some pictures of my work so far, i now have THREE broken wires to resolder but whatever. its just so friggin hot tho. 95 with a heat index of 106 today.
 

Attachments

  • 003.JPG
    003.JPG
    125.7 KB · Views: 19
  • 004.JPG
    004.JPG
    128.8 KB · Views: 19
  • 005.JPG
    005.JPG
    119.3 KB · Views: 17
, is there supposed to be ~0.2 drop off between the case of the outputs and ground? because there is..
I am not sure what you mean by drop off, do you mean ohms?
if so the TO-3 transistor (bjt) cases are to be isolated or high resistance from chassis ground.
use some sandpaper to scrape off the oxidization on the wire wraps b4 you solder on the broken wires, you could do the same to the wire that t you are going to solder on the terminal, wrapping one or more turns of the bare wire around the terminal first.
i use a small piece of heatshink over the wire insulation to stop the insulation from wicking back with the heat of the solder iron.
 
with the meter on the diode function it shows 0.2 between the case(collector) and ground. i didnt measure resistance as i understand what the diode numbers mean in terms of a short circuit better.

ive been sitting here thinking about it and im leaning towards that being normal, seeing as all 8 outputs are measuring that now. and considering that the left part of that amp board is isolated from the right side,
 
i just went and measured 2 of them on the ohm setting, it slowly counts up, doesnt seem to ever stop, i quit after they reached 1.5k ohm
 
yes do not measure in diode mode, counting up means that there is capacitance in the circuit and it the cap is charging up from the small voltage developed in the meter, so sounds like you are okay there
 
I here a relay click, which is a good sign, what else have you tested? offset,bias, PS voltages?
 
I here a relay click, which is a good sign, what else have you tested? offset,bias, PS voltages?

thats where i called it a day yesterday, offset and bias still have to be adjusted, that power up test is in its "its a complete amp again" stage. ill be getting back to work on this a little later this morning.
 
so far in trying to adjust the bias there seems to be a problem.

steps i took: removed preamp-amp jumpers, turned speaker selectors off, aux input, volume all the way down.
vr 3 and 4 fully ccw

vr 1 and 2 seem to work and i got those down to around 10mv on pins 9/10 and 24/25, couldnt seem to find an exact 0.000, vr3 as measured on pins 12 and 13 will not go below 40mv turning it cw will raise it but fully ccw is 40mv, vr4 as measured on pins 27 and 28 does absolutely nothing. its 0.000 fully ccw and a half turn cw was still 0.000. i reverted back to fully ccw for that one as well.

I also noticed, because i turn the unit off each time i have to move my probes, i noticed that the dc offset pins both shot up in voltage whenever the unit is turned off, they dart up to several volts, i think i saw a peak of 8v for an instant and then immediately 2v and then drain back down to nothing, now im assuming that wouldnt be translated into the speakers since the relay turns off with the unit but i dont know if its normal for it to have that voltage or not.

I will measure the voltages and report back. in the meantime, any ideas on why vr3 and 4 would be behaving so differently from each other?
 
vr3 seems to be working okay, the bias is on the high end 40mA, it is probably because of the new device, what devices are you using?
vr4 shows no bias in the o/p stage, check the bias voltages against the schematic and each channel to each other. Q6,8 voltages? orientation? do some resistance measurement checks
would not worry about the turn off transients, it depends on the time that they settle and when the relay breaks contact.
 
vr3 seems to be working okay, the bias is on the high end 40mA, it is probably because of the new device, what devices are you using?
vr4 shows no bias in the o/p stage, check the bias voltages against the schematic and each channel to each other. Q6,8 voltages? orientation? do some resistance measurement checks
would not worry about the turn off transients, it depends on the time that they settle and when the relay breaks contact.

devices as in outputs?
863-MJ21195G pnp output
863-MJ21196G npn output

im using everything from markthefixer's post in this thread.

as far as checking voltages, its been storming, its still storming. it looks like be doing that tomorrow.
if it is normal for it to be higher with these outputs, then it appears that whichever channel the left side of the board is, is working. that side of the board did not fail so it makes sense that it would be working despite having new parts, the right side however... out of boredom while ive been waiting on the thunder to go away, i traced pins 27 and 28 back in the service manual, those go directly to the emitters of the outputs, which suggests to me they arent being powered on at all, wouldnt those be powered by the other transistors? and according to the diagram it looks like all of those transistors report to the protection board, doesnt the protection board look for everything to be a certain voltage... anyway im hoping its something simple like a 4th wire broke that i didnt notice. im going to be checking those too when i go to measure the voltages
 
512-KSA992FBU Q2,Q4 diff 2sa726 (just in case - they're cheap!!) (check orientation)
512-KSA1381ESTU Q6 pre-driver (check orientation)
512-KSC3503dSTU Q8 pre-driver (check orientation)
512-KSC2073TU Q10 driver (check that you have the correct position)
512-KSA940tu Q12 driver (check that you have the correct position)
863-MJ21195G pnp output (check that you have the correct position)
863-MJ21196G npn output (check that you have the correct position)
bottom line the 2 channels should match in bias voltages with tolerances

yea you have to make sure that voltage is getting to the o/p collectors to begin with.
pins 27,28 tap off the emitter bias resistors, acts as the over current detectors to send the V developed across them to the protection ckt. You have no bias I so no V developed across the emitter resistors
the protect ckt checks for DC offset,over current and AC present/fail

important to know, what is the voltage across Q6,8 collectors since it controls the bias generator which is placed between them.
 
Last edited:
i did some quick checks just so i could get this information to you sooner, i have a lot of other work to do this morning before i can do thorough voltage testing or anything on this, im going to have to devise a safe way to get to the collector of Q6 because my hand shakes and that one is difficult to get to, but on Q8, the base and emitter are in the -61 to -62 volt range, and the collector is -200mv

Q6 base and emitter are in the -49.5v range, collector unmeasured because im too chicken to reach for it right now, ill get back to you on this.
with negative probe on chassis

I looked over my positions of the transistors and they are all in the right spots, they also match the positions on the other channel. as far as orientation goes, i looked at each ones datasheet on mouser, made notes of which pin is which, and then placed them according to the diagram on the board for the spot, so arrow is emitter, thick line parallel with hole is base, and thinner line running from thick line to a hole being collector.
 
I need to measure these again i think, just to be certain i didnt miss/double read/ or outright skip one but here are the voltages going into the amp board.

pin 1 and 2 are -60.5, their counterparts pin 16 and 17 are -60.6
pin 3 is 0.5mv, pin 18 is -0.4 volts
pin 4 and 5 are 35.97 volts, pin 19 and 20 are 36.2 volts
pin 6 is 0 pin 21 is 0
pin 7 and 8 are 51 volts, pin 22 and 23 are 51.3 volts
pin 9 is 0, pin 24 is 0
pin 10 and 11 are -10 mv, pin 25 and 26 are -21mv
pin 12 is possibly in error, i may have measured the wrong thing, or i failed to observe/write down mv, i got 35 volts written down, pin 27 is -21mv
ping 13 is -10mv, pin 28 is -21mv
pin 14 and 15 are 48.2, pin 29 and 30 are 48.1
 
Last edited:
so ive been reading the manual this morning, tracing things back,

if im reading it right, pin 30 takes a direct path to pin 31, which is the collectors of the outputs, which do show voltage, but they also take a second route, through a diode, through a capacitor, into the base of Q12, where it appears to take a trip down the board, where it splits, one end of the fork in the road goes to the bias adjustment trimmer, the other goes to the collector of Q8, and through a capacitor to the base of Q8
the collector of Q6 also traces back to that bias adjustment trimmer, with a trip through a diode first
so..
perhaps i should go measure that diode
the voltage seems to be everywhere, until it comes time to bias it
and thats where its mysteriously missing

ok so i just measured both channels bias diodes (in circuit) (those ones that mount to the heatsink just so we're all on the same page)
that asked more questions than it answered, but theres a difference in readings
the one for the left side is open in both directions, I also tried a resistance check in one direction and got nothing.
the one on the right side is reading 0.03 one way and 0.04 the other in diode test mode.

now, i would think this should be closed in order to complete the connection to the bias trimmer, so why is the one thats open the only one that SEEMS to be working? (actual status of channel is unknown, but it does respond to trimmer adjustments at least)


i may be barking up the wrong bush but, i dunno, theres another difference between channels though.
 
Last edited:
just to update this thread since im now done with my other thread, i took these, varistors? out of circuit and measured them again, no change. one is open both directions, one is shorted.
 
Back
Top Bottom