Need help with an Pioneer SX-550

[d_bolle]

If you look at the circuit there is a 430K resistor in the collector of the 2SA733 making it a low current/low power circuit. The 992 will be fine. Double check the pin-out.

Thank you for pointing out to just look at the drawing to get the current. Pin-out seemed identical. I also replaced the 945 with an 1845, it's the similar situation but there is also a 430K resistor, so I figured that your advice would work for the other one too. Is that correct?

 

[d_bolle]

Update: Thought I had it all sorted out in the amplifier section, all values looked good. I replaced all electrolytic caps just to have new ones.
Fired it up with speakers, and saw the magic smoke again. I saw it coming from the diode D6, I only replaced the left one (D5). Will replace the right one too and see what happens. Of course with a DBT again.

 

[mrk229]

Update: Thought I had it all sorted out in the amplifier section, all values looked good. I replaced all electrolytic caps just to have new ones.
Fired it up with speakers, and saw the magic smoke again. I saw it coming from the diode D6, I only replaced the left one (D5). Will replace the right one too and see what happens. Of course with a DBT again.

99.5% of the time no matter how confident I am I start things up on the DBT after most parts changes, especially power supplies and amplifier/driver boards.

Better check D4 and VR4 while you are there. The current to smoke D6 has to go through them as well. I would also check Q6 & Q8.

 

[d_bolle]

99.5% of the time no matter how confident I am I start things up on the DBT after most parts changes, especially power supplies and amplifier/driver boards.

Better check D4 and VR4 while you are there. The current to smoke D6 has to go through them as well. I would also check Q6 & Q8.

I agree. No idea what the problem is now, maybe I created a short when replacing the caps, maybe something gave way that was problematic before, maybe that is the cause for the original problems.

Or I didn't really test it with speakers before, that is possible. I did set the idle current according to the manual a few weeks ago, but that was without speakers attached. so maybe there always was an issue and I just didn't notice it.

It's back to the DBT, testing all the voltages, and checking all the transistors again.

Funny thing is, I have another SX550 on the shelf that is rusty in the back but turns on fine. I'm considering swapping the entire amplifier board at this point. But the current board has like 20 new parts now, would be a shame to waste all those pretty nichicon caps, so I'd have to take those out as well.

 

[d_bolle]

Little update:

Got some time with my favourite unit again.

Turned on with DBT and nothing else plugged in, lamp glowed for half a second and then went dim, just like it's suppsed to be.

Then attached headphones and heard a noticeable hum right and a lower more quiet hum left. volume and balance didn't affect the hum.
FM, Aux and Tape work but are mute on right, with the hum still present.

So very similar what the problem was right at the beginning.
I pulled out my new gear, a FLIR One smartphone attachement:


So Q6 is getting hot even on DBT with only headphones attached. The burn marks on R36 and R32 are from a month ago when I tested it without DBT on speakers as I thought everything was fine. They didn't get hot today as I ran on DBT without speakers. I will take those three parts out and check them.

 

[d_bolle]

Update: I replaced Q6 with an KSC2383OTA (that was the closest match in my drawer), but didn't turn the unit back on because I noticed something else: between the heatsinks of the large TO-220 transistors Q9,10,11,12 and the grounded metal case (and the screws going through the heatsinks, isolated by plastic tubes and mica plates) I measure a resistance that starts at 50ish Ohms and rises to about 500, then jumps to 50k and drops to a few k, then drops back to 50 Ohms.

Shouldn't be that OL? I had so far reused the old mica plates ones, learnt that that is a bad idea, but I can use silicone pads instead.

To test, I took the screws and the old pads out, so that Q9-Q12 are up in the air without direct connection to the heatsink. However, I still can measure a resistance. Whenever I touch a point the first time the Fluke gives me 50 Ohms and rises quickly to about 500. Then it shows OL for a second, then jumps to 50k Ohms and drops to a few k. Then it drops to 50 Ohms and starts rising again. So some capacitor somewhere is probably charging and uncharging.

Any Idea where I could look for that weird short? Or is that behaviour normal?

 

[Watthour]

If you are measuring from the tab (collector) of the 2SB596/D526 to chassis, the meter may show some measurement since the power ±30V supply rails are not completely isolated. There are also the C-B and E-C junctions connected to other parts of the unit.

This is the problem with measuring devices in-circuit.

 

[d_bolle]

Once again working on that unit. Still no clue what the problem is. I replaced the two old variable resistors VR3 and VR4 with 100 ohm blue box ones, in case they are old and corroded and cause the issue. When trying to set it, VR3 behaved as it should and I could set 20mV between 33 and 34.

But between 35 and 36 was 0 mV no matter what I set at the resistor. The resistor itself it set to about 50 mOhm, just like the other (I measured directly at the pins).

Wondering what the reason for this was, I measured all contacts of Q9-Q12 powered on (with a 40W DBT to limit current):

Q12
B18,9
C - 18,9
E 18,9

Q10
B 18,9
C 18,9
E 18,9

Q9
B 0,66
C 18,7
E 0,02

Q11
B -0,62
C -18,8
E 0,01

So both Q12 and Q10 are on the positive rail with their B and E ?!?

But the DBT didn't really light up. or maybe it did, it's a weird decoration bulb, maybe it's supposed to be so dim at 40W. But Q1,2,4,8,9-12 all measure fine in the diode tester (unplugged). No shorts. Where else could I look to find out where that rail voltage is coming from?

I did set the bias on both channels a few months ago when I thought I had it working after replacing some transistors and electrolytic caps. So there shouldn't be damage to the board itself with some short between the tracks. When testing it with speakers back then, R36 and R32 started smoking (turned it off immediately), but they measure fine with one leg out now. Something in line with them must have gotten damaged, but I have no idea what.

Any advice would be welcome!