Pioneer SX-980 Protection Relay Kicking Out

Hmmm, that stinks, but it points back to power supply, as nothing else is connected. Do you still have that noise on the peaks of the +65v ripple? When you pull Q1, pull Q2 as well, then check PA3004 pin 3 to see if it is still cycling.
I do still have the ripple on the 65v supply. I rechecked for ripple on the other DC inputs to the protection board - all nice and flat. Out of curiosity, I removed the 20vac connection to the protection board to see what would happen - as this is essentially another input. The relay "on" time was reduced slightly when I did this, but the cycling continued. I was hoping that removing the AC supply would make the protection board think the receiver was turned off and disengage the relay.
 
That is interesting. Look at the ground potential difference between the ACTUAL metal of the PA3004's pin 5 and the actual metal of the chassis. Watch it through a couple of "cycles" first.

Quite frankly, grounding the AWM-124 protection board's pin 4 is just signalling a protect condition, as pulling it LOW is the function of both AWH-073 power amp channel's Q8s.



I know there is a hankering to rush and "git er dunn", but PLEASE do this in a methodical order, you would be setting up a double double blind,
so please do q1 first, then see if pin 3 and q2 are behaving.
I pulled Q1. Relay continues to cycle. Monitoring pin 3 of the IC, I'm seeing the same behavior as before - 0v when the relay is off, .7v when the relay is briefly on. Are there voltage measurements you'd like me to take from Q2?
 
That is interesting. Look at the ground potential difference between the ACTUAL metal of the PA3004's pin 5 and the actual metal of the chassis. Watch it through a couple of "cycles" first.

Quite frankly, grounding the AWM-124 protection board's pin 4 is just signalling a protect condition, as pulling it LOW is the function of both AWH-073 power amp channel's Q8s.



I know there is a hankering to rush and "git er dunn", but PLEASE do this in a methodical order, you would be setting up a double double blind,
so please do q1 first, then see if pin 3 and q2 are behaving.
I'm seeing this from pins 3,4,5,6 & 8 of the IC. It's not much, as I'm set on 0.2v/division, but thought I should mention it. Odd that I'm seeing it on pin 5 of the IC, since it's a ground pin. I did ground that pin to frame ground of the receiver, but the cycling continued. I saw this waveform in my prior testing, but I was viewing it at a higher-voltage setting on my o-scope, so it didn't stand out as anything significant.
20180913_151121 (Small).jpg
 
I do still have the ripple on the 65v supply. I rechecked for ripple on the other DC inputs to the protection board - all nice and flat. Out of curiosity, I removed the 20vac connection to the protection board to see what would happen - as this is essentially another input. The relay "on" time was reduced slightly when I did this, but the cycling continued. I was hoping that removing the AC supply would make the protection board think the receiver was turned off and disengage the relay.

It should have turned off the relay. Does connecting PA3004 pin 7 (ac detect) to PA3004 pin 5 (ground) drop the relay?
Touch the scope probe tip to its ground clip, is it a clean 0vac? Where is your scope probe ground clip attached? On the chassis? Or on one of the board ground pins? Power off: Verify 0 ohms from the PS board ground pins to the chassis and the other ground pins mentioned, use the lowest scale on your ohmmeter and turn off auto-range if it has it. Once 0 ohms is verified clip the scope ground to one of the PS board ground pins. Power on: Scope the ground connections to the PS board (pins 4, 6, 7, 21 and 39) and the Power Amp boards (pins 3, 7 and 9) the Prot board pin 3, PA3004 pin 5 and the chassis on an ac mv scale.
 
Last edited:
Also check that the Power transformer Brown lead (Center tap)is connected to ground, orange leads are connected to PS pins 10 and 11, and that FU 3 and 4 are good. Also the green leads are connected to PS pins 12 and 13.

EDIT: more info
 
Last edited:
Also check that the Power transformer Brown lead (Center tap)is connected to ground, orange leads are connected to PS pins 10 and 11, and that FU 3 and 4 are good. Also the green leads are connected to PS pins 12 and 13.

EDIT: more info
Disconnecting the AC supply from pin 1 of the protection board and grounding pin 7 of the IC does not stop the relay cycling.
Power transformer brown wire is solidly wire-wrapped to a ground pin that's screwed to the frame.
Orange wires are solidly-connected to pins 10 & 11 of the PSB, green wires are solidly-connected to pins 12 & 13 of the PSB.
Pulled one end of all 5 PSB fuses and meter-tested them - all good.
Scope probe to ground clip gives a nice flat line @ 0.2v/division. Scope probe left ungrounded gives just a few tiny bumps on the display, but nothing like I was getting from the DC voltages on the PA3004.
PSB pin 4 to ground - 0 ohms; pin 6 to ground - 1 ohm (doesn't appear to be tied to ground on the schematic); pin 7 to ground - infinite resistance (doesn't appear to be tied to ground on the schematic); pin 21 to ground - 0 ohms; pin 39 to ground - (I don't have a pin 39).

O-scope readings (all except PA3004 pin5 are 0.2v/division)
PSB pin 4 to ground
20180914_082654 (Small).jpg

PSB pin 21 to ground
20180914_082720 (Small).jpg

Amplifier boards pins 3, 7 & 9 (all pins on both boards looked the same)
20180914_083203 (Small).jpg

Protection board pin 3
20180914_083306 (Small).jpg

PA3004 pin 5 (o-scope scale changed to 20mV/division)
20180914_083623 (Small).jpg
 
When you grounded pin 7 of the IC did you connect between pins 5 and 7, or ground to another ground point?
I think that noise is getting into PA3004 and causing the problem. Do the higher and lower envelope levels correspond to the relay cycling rate? Look at the junction of R18 and R19 (0.5 ohm 5W resistors) on both power amps to see if either one is oscillating. If you expand the scope time base can you look at (and trigger on) the noise? is it random or regular frequency. Does your scope have a freq counter function? See if you can locate a point in the schematic where that noise is largest to try and localize the source..
 
When you grounded pin 7 of the IC did you connect between pins 5 and 7, or ground to another ground point?
I think that noise is getting into PA3004 and causing the problem. Do the higher and lower envelope levels correspond to the relay cycling rate? Look at the junction of R18 and R19 (0.5 ohm 5W resistors) on both power amps to see if either one is oscillating. If you expand the scope time base can you look at (and trigger on) the noise? is it random or regular frequency. Does your scope have a freq counter function? See if you can locate a point in the schematic where that noise is largest to try and localize the source..
I originally grounded pin 5 of the PA3004 to the frame. Per your suggestion, I grounded pin 5 of the IC to pin 7 of the IC after removing the AC source from pin 1 of the protection board. It made no difference. I may have mentioned this before, but removing the AC source from pin 1 of the protection board causes the relay "on" time to be slightly reduced from an already-brief amount of time. I'll check out the power amps as you suggested, but would they still affect the operation of the protection relay when I have them disconnected from the protection board? Right now, I still have Q1 removed (per Mark's suggestion) and I also have wires disconnected from pins 4,5,6,7 & 8 of the protection board.
 
I originally grounded pin 5 of the PA3004 to the frame. Per your suggestion, I grounded pin 5 of the IC to pin 7 of the IC after removing the AC source from pin 1 of the protection board. It made no difference. I may have mentioned this before, but removing the AC source from pin 1 of the protection board causes the relay "on" time to be slightly reduced from an already-brief amount of time. I'll check out the power amps as you suggested, but would they still affect the operation of the protection relay when I have them disconnected from the protection board? Right now, I still have Q1 removed (per Mark's suggestion) and I also have wires disconnected from pins 4,5,6,7 & 8 of the protection board.

I do remember the AC remove caused timing change. Something is causing that noise, which seems very abnormal to me. The ground connections should not have that much noise on them.
An oscillating power amp (or other circuit) can back feed noise into the power supplies or onto the ground, especially if there is a ground loop, which itself can cause instability. I think we have learned that the source of the problem is not the PA3004, or a "normal" fault coming into the Prot board. 2 different PBs and a third PA3004 point to it not being the PB.
 
I do remember the AC remove caused timing change. Something is causing that noise, which seems very abnormal to me. The ground connections should not have that much noise on them.
An oscillating power amp (or other circuit) can back feed noise into the power supplies or onto the ground, especially if there is a ground loop, which itself can cause instability. I think we have learned that the source of the problem is not the PA3004, or a "normal" fault coming into the Prot board. 2 different PBs and a third PA3004 point to it not being the PB.
I agree with you, there. I think we're after crappy power, crappy ground, or both. That ripple in the 65vdc supply is pretty substantial, as well. Speaking of the power supply, I'm hoping I mentioned this before, but it's probably worth restating: When I first brought the receiver into the shop, in addition to cleaning the pots, I also replaced the three regulator transistors on the power supply board. I don't have the receiver in front of me right now but, from my Mouser order, I installed one 863-MJE15033G transistor and two 863-MJE15032G transistors. These were crossed over by me. I know that two of the three original transistors on the PSB were the same. There wasn't any confusion on my part about the replacements or where they got soldered in - that was pretty straightforward. I would gladly take any recommendations for better replacements.
 
Some have reported noise issues when using the MJE15032/33 as replacements for the power supply pass transistors. The original transistors (2SD712 & 2SB682) are 4 MHz devices while the MJEs have an ft of 30 MHz which could result in an instability (oscillation). You may want to try going back to the original pass transistors if they are still in good shape.
 
Last edited:
Some have reported noise issues when using the MJE30032/33 as replacements for the power supply pass transistors. The original transistors (2SD712 & 2SB682) are 4 MHz devices while the MJEs have an ft of 30 MHz which could result in an instability (oscillation). You may want to try going back to the original pass transistors if they are still in good shape.
Thanks for that info! That's the kind of information I was looking for. I still have the original transistors from the power supply. The two transistors that match were the correct number, but the single transistor was not correct for either an original or a crossover (although it could have been beefier than the original), so I replaced it with a crossover from Mouser. I replaced the other two while I was at it. Ultimately, I'd like to find new, replacement transistors that are correct. I'm gonna feel like a total tool if I didn't mention my transistor swap earlier in this thread!
 
KSC2073/KSA940 would probably do just fine but you will want to keep an eye on the case temperatures. Q5 (+13V) in particular must be getting very hot. Information on the schematic indicates that the power dissipation would be around 2W so adding supplemental heat sinking if possible would be beneficial.
 
Last edited:
KSC2703/KSA940 would probably do just fine but you will want to keep an eye on the case temperatures. Q5 (+13V) in particular must be getting very hot. Information on the schematic indicates that the power dissipation would be around 2W so adding supplemental heat sinking if possible would be beneficial.
Okay. I've also looked at some older posts from Mark the Fixer, but it looks like at least one of his suggested transistor replacements from Mouser is no longer available. I don't have the receiver or the board layout schematic in front of me, right now, or I could be a little more specific.
 
Here is an example of an SX-980 power supply rebuild where addition heat sinks were provided. The use of shoulder washers and mica insulators are required to isolate the transistors from the heat sinks.

http://audiokarma.org/forums/index.php?threads/sx-980-project.633014/#post-8539614

Of course, I'm only speculating that the MJEs may be related to your problem which may not be the case. Reverting to the original transistors would be the easiest way to confirm this.
 
U
Here is an example of an SX-980 power supply rebuild where addition heat sinks were provided. The use of shoulder washers and mica insulators are required to isolate the transistors from the heat sinks.

http://audiokarma.org/forums/index.php?threads/sx-980-project.633014/#post-8539614

Of course, I'm only speculating that the MJEs may be related to your problem which may not be the case. Reverting to the original transistors would be the easiest way to confirm this.
Understood. When you have a piece of equipment that goes from working to not working, the first question asked should be, "what changed?" The answer is I pulled the 2 boards containing the pots and switches out of their mounts just enough so I could access them with DeOxit. The other thing was replacing the power supply transistors. Since I've isolated that stage by removing the preamp out/in jumpers, it makes sense verify that the transistor change caused/didn't cause the problem. Only thing I don't like about it is desoldering the replacement transistors, resoldering the originals back in, then, ideally, soldering in some new, good replacements (if the problem turns out to be transistor-related). I don't like putting the added stress on the board traces. But, it needs to be done. I'll be back to the shop in a couple days, so we'll find out, then.
 
I did a little more digging and found some updated part numbers for the power supply board. The transistors I'd be interested in replacing are Q1, Q4 & Q5. These are the three that were already replaced with the numbers I posted previously.

The part numbers below are all from Mark the Fixer, so I have every confidence they are correct. From two of his posts, with updated numbers at the bottom:

q1 2sd712 100v 4a 30w 55-300hfe 8mhz
512-KSC2073TU npn to-220 bce 150v 1.5a 25w 4mhz 40-140hfe $0.54
q4 2sb682 100v 4a 30w 55-300hfe 8mhz
512-KSA940TU_Q pnp to-220 bce 150v 1.5a 25w 4mhz 40-140hfe $0.54
neither side is going to come close to an amp of current, no less 4 amps

q2 2sc1735 100v 0.5a 0.8w 55-300hfe 130mhz
512-KSC2383YBU to-92L ecb 160v 1a .9w 50mhz 160-320hfe $0.17
q3 2sa850 100v 0.5a 0.8w 55-300hfe 130mhz
512-KSA1013YBU to-92L ecb 160v 1a .9w 50mhz 160-320hfe $0.17 (R=60-120, O=100/200, Y=160/320)

Q5 2sd712 512-KSC2073TU
q6 2sc945a eCb 50v .15a .25w 300mhz 40-700hfe
512-KSC2383YBU to-92L ecb 160v 1a .9w 50mhz 160-320hfe $0.17

512-KSC1845UBU >>> 512-ksc1845FTA
512-KSC2073TU_Q >>> 512-ksc2073TU
512-KSC2310YBU >>> ksc2383YTA
512-KSA940TU_Q >>> 512-ksa940TU
 
Some have reported noise issues when using the MJE15032/33 as replacements for the power supply pass transistors. The original transistors (2SD712 & 2SB682) are 4 MHz devices while the MJEs have an ft of 30 MHz which could result in an instability (oscillation). You may want to try going back to the original pass transistors if they are still in good shape.
I replaced the MJE15032/33 transistors with my original transistors this morning. I put Q1 back into the protection board, since removing it made no difference. Amazingly, I still have the relay cycling problem. I still have the same sawtooth pattern on my 65vdc supply to the protection board, but my noisy ground problem seems to have cleared up (at least on the protection board). This situation is starting to become laughable. Good lord, what's left to be done? Ugh.
 
Well, at least you were able to elminate one potential source of trouble and that's progess. As mentioned before, some ripple on the +65V is to be expected but the amplitude will be determined by the load on the power supply. When I did a quick simulation, it showed ripple in the range of 2.5 to 3V p-p with a load of around 45mA on the +50V and the relay engaged. You can get an approximate measure of the current being drawn from the +50/+33V by measuring the voltage drop across R3 on the power supply board. The +13V should be clean with no ripple.

At this point, you may want to go back to square one with all protection pins restored and Q1/Q2 installed. If it is still cycling, check all pin voltages when both in and out of protection. Then follow Mark's recommendations regarding the removal of Q1 and watch IC pin 3 for activity.
 
Here are voltage measurements from the power supply board:

1 58.5vac
2 58.5vac
3 59.3vac
4 0v
5 37.8vac
6 37.8vac
7 49.8vdc
8 -49.8vdc
9 45.4vac
10 51vac
11 51vac
12 19.7vac
13 19.7vac
14 19.7vac
15 0v
16 13.3vdc
17 13.3vdc
18 9.9vdc
19 66.4vdc
20 35.1vdc
21 0v
22 52vdc
23 24.6vdc
24 24.6vdc
25 -24.3vdc
26 -24.3vdc
27 -53vdc
28 -13.7vdc
29 45.3vac
30 3.3vac
31 0v
 
Back
Top Bottom