Pioneer SX-980 Protection Relay Kicking Out

[Brivan]

I believe the voltages posted on the schematic next to the PA3004 pins are the steady state voltages with no audio output from the power amps, i. e. 0.0vac; no dc offset at the power amps, i.e. 0.0vdc; no overload condition i.e. Q1 in cutoff; and Q2 on (saturated) with relay pulled in; steady AC present on pin 7; and negative voltage present on pin 6. SM PA3004 Logic diagram shows 18vac w current limiting resistor to pin 7 and -13vdc with current limit resistor to pin 6. If those conditions are true the logic diagram S1 should be open, pin 8 should be a steady +7.2 vdc.

All voltages in the IC are measuring roughly the same as the values shown on the schematic IC pinout. I have a rather wide trace on the o-scope when connected to pin 3:

At 0.2v per division, here's my brief reading at pin 3 when the relay closes - about 0.7vdc (hard to capture this):


Here's my voltage reading at pin 7 of the IC - about 1.5v peak-to-peak. Measures about 0.68v with the digital meter.

 

[Brivan]

AC voltage at PB pin 1 - about 19.6vac on my meter:


.....and my ripply-looking voltage on PB pin 10:

 

[merlynski]

Okay, I grounded PB pin 4. IC pin 8 voltage does remain stable at about 0.5v, so you're prediction was correct. I was remembering a different situation, apparently. I'll check some voltages with the o-scope.

With the voltage stable at +0.5 vdc on IC pin 8 the relay should have stopped cycling. Yes? If so the Protection circuit is probably working correctly and the problem is one of the inputs to the Protection board triggering the cycling. Since the replacement board behaved the same way this reinforces that conclusion.
Now we have to eliminate the inputs 1 by 1.
If PA3004 pin 4 is grounded, does the cycling stop? If not then the problem is not caused by a signal coming from the Power Amp outputs or overload circuits. If it does we must dig deeper.
Is your scope dual trace? If so probe CH1 to PA3004 pin 3 (or base of Q2, set scope to trigger CH1 on the neg edge, DC triggering. Adjust triggering to be as close to the top of the neg edge as possible. Set the sweep time base so that the negative pulse is less than half the screen width. This may be a slow sweep. Then use CH2 to look at every pin of the PA3004 and the pins of the protection board for something changing at the time the pulse on CH1 changes state. That may be a narrow event. I expect that the width of the negative pulse is the charge time of the cap on PA3004 pin 8.

 

[Brivan]

With the voltage stable at +0.5 vdc on IC pin 8 the relay should have stopped cycling. Yes? If so the Protection circuit is probably working correctly and the problem is one of the inputs to the Protection board triggering the cycling. Since the replacement board behaved the same way this reinforces that conclusion.
Now we have to eliminate the inputs 1 by 1.
If PA3004 pin 4 is grounded, does the cycling stop? If not then the problem is not caused by a signal coming from the Power Amp outputs or overload circuits. If it does we must dig deeper.
Is your scope dual trace? If so probe CH1 to PA3004 pin 3 (or base of Q2, set scope to trigger CH1 on the neg edge, DC triggering. Adjust triggering to be as close to the top of the neg edge as possible. Set the sweep time base so that the negative pulse is less than half the screen width. This may be a slow sweep. Then use CH2 to look at every pin of the PA3004 and the pins of the protection board for something changing at the time the pulse on CH1 changes state. That may be a narrow event. I expect that the width of the negative pulse is the charge time of the cap on PA3004 pin 8.

The cycling stops with the relay de-energized when pin 4 of the PB is grounded. Also, when pin 4 of the PB is grounded, the voltage at pin 8 of the IC stays at a constant 0.5vdc. I can perform the checks you suggested tomorrow, but I have a question: since I now have pins 4,5,6,7 & 8 of the PB disconnected from the receiver, eliminating external sources as part of my relay cycling problem, shouldn't the relay work normally? Or do I need to have one or more of those pins connected to the receiver for normal operation of the relay? That's been my question as of late - since PB pins 5,6,7 & 8 are inputs and outputs, I didn't think they needed to be connected. But pin 4 of the PB was my question - grounding it causes the relay to stay de-energized, so should there be voltage there to cause the relay to energize? The schematic would lead me to believe there isn't, but I'm not sure.

 

[merlynski]

The cycling stops with the relay de-energized when pin 4 of the PB is grounded. Also, when pin 4 of the PB is grounded, the voltage at pin 8 of the IC stays at a constant 0.5vdc. I can perform the checks you suggested tomorrow, but I have a question: since I now have pins 4,5,6,7 & 8 of the PB disconnected from the receiver, eliminating external sources as part of my relay cycling problem, shouldn't the relay work normally? Or do I need to have one or more of those pins connected to the receiver for normal operation of the relay? That's been my question as of late - since PB pins 5,6,7 & 8 are inputs and outputs, I didn't think they needed to be connected. But pin 4 of the PB was my question - grounding it causes the relay to stay de-energized, so should there be voltage there to cause the relay to energize? The schematic would lead me to believe there isn't, but I'm not sure.

There is not, the absence of a signal is the normal condition.
The overload detection is a current controlled process.
When pin 4 of the PBoard is grounded it pulls current through diode D2, this pulls current through the BE junction of Q1, turning it on. This causes current to flow through the collector of Q1 and R1 and R7. This creates a voltage drop across R7, causing pin 4 of the PA3004 to go positive. This positive voltage is seen by the PA3004 as an indication of an overload, so it turns off Q2 and drops out the relay (off). As long as there is enough current through BE on Q1 to pull the PA3004 pin 4 positive the relay will stay off. When the PB pin 4 is no longer grounded it goes to high impedance, D2 stops conducting, Q1 turns off and R7 pulls PA3004 pin 4 back to 0vdc. This allows the PA3004 to start charging the cap on pin 8. As soon as the voltage on pin 8 rises enough the PA3004 turns the relay on again. This controls the delay time of the relay coming on. Second function of Pin 4: if it detects a high voltage dc offset on the power amp outputs, whether it is positive or negative, through R5 and/or R6. If one amp output goes toward about 10vdc positive or negative the voltage on pin 4 will increase to about 1 volt causing the PA3004 to turn off the relay. Note that PA3004 pin 4 is a voltage 'window' detector, any voltage + or - outside its window makes it turn off the relay. I don't know the exact voltage window, which is why I started the 'PA3004 pin 4' thread, to see if anybody else knows. Whew, that was a bit long winded .
EDIT: clarification

 

[merlynski]

since PB pins 5,6,7 & 8 are inputs and outputs, I didn't think they needed to be connected.

This is true, and it also tells us the problem is not coming IN on one of those pins. That means the problem has to be coming in on one of the pins that is connected to the system. 1, 2, 3, 9 or 10. Make sure your ground to pin 3 is solid. If you ground PA3004 pin 4 while the relay is cycling and it does not stop cycling that eliminates PB pin 4, as well as 5, 6, 7 and 8. Try adding (just clip leads) a large value cap from PB pin 10 to ground (80 volts rating or higher) to see if the oscillation on the peaks of the ripple decreases significantly. Try the same with PB pin 13, 25 volt rating should be high enough. Check PB pin 2 for clean dc, look at the ripple with your scope, try adding more capacitance there too.

 

[Brivan]

First off, thanks for the fantastic write-up of the operation of the protection board! I'll definitely save it for future reference! Okay, here we go: Grounding pin three to a ground point on the frame made no difference. Grounding pin 4 of the IC did not stop the cycling. Adding a 470uf cap between pins 10 and ground, 9 and ground, 2 and ground made no difference. Yes, I did remember to reverse the cap leads when connected to pin 2 and ground (you learn these things the hard way). DC on pin 10 looked choppy; but less so with the cap connected. Adding a cap to 9 and ground, 2 and ground made no difference in waveform - it was nice and flat the whole time.

Yesterday, I ordered a new Pioneer PA3004 IC from Ebay. Hoping to have it next week some time. I did this based on the fact we're getting down to everything else checking out. The spare protection board I purchased was obviously used, but was claimed to work. Well, you never know - there's a chance it has the same issue as my original board. As long as that chance exists, I need to eliminate it, so I plan on installing the replacement IC on my original board, which has had Q1, Q2 and all of the caps replaced. That way, I'm still preserving the all-original board that I bought. If we find something else wrong, I'm only out 20 bucks.

Pin 10 without extra cap:


Pin 10 with extra cap:


Pin 9:


Pin 2:

 

[merlynski]

You are welcome, writing it up helps me understand it better also. The scope traces look fine, pin 10 reacted about like I expected, but not as desired. With the scope look at all the Power Supply pins for similar signals or oscillations. Look at the power amps' pin 11 (outputs). Output transistors' E, B & C. With ohmmeter and power off check continuity to ground (chassis) for every ground pin on the protection board, power supply board and the power amp boards. Those old wire wrap wires are brittle and may have broken off.
Is your scope dual trace? If so probe CH1 to PA3004 pin 3 (or base of Q2, set scope to trigger CH1 on the neg edge, DC triggering. Adjust triggering to be as close to the top of the neg edge as possible. Set the sweep time base so that the negative pulse is less than half the screen width. This may be a slow sweep. Then use CH2 to look at every pin of the PA3004 and the pins of the protection board for something changing at the time the pulse on CH1 changes state. That may be a narrow event.

 

[Brivan]

You are welcome, writing it up helps me understand it better also. The scope traces look fine, pin 10 reacted about like I expected, but not as desired. With the scope look at all the Power Supply pins for similar signals or oscillations. Look at the power amps' pin 11 (outputs). Output transistors' E, B & C. With ohmmeter and power off check continuity to ground (chassis) for every ground pin on the protection board, power supply board and the power amp boards. Those old wire wrap wires are brittle and may have broken off.
Is your scope dual trace? If so probe CH1 to PA3004 pin 3 (or base of Q2, set scope to trigger CH1 on the neg edge, DC triggering. Adjust triggering to be as close to the top of the neg edge as possible. Set the sweep time base so that the negative pulse is less than half the screen width. This may be a slow sweep. Then use CH2 to look at every pin of the PA3004 and the pins of the protection board for something changing at the time the pulse on CH1 changes state. That may be a narrow event.

Will do. Also, I was getting ready to check the -65vdc on the power supply board, when I noticed a glaring mistake. In my attempt to replace caps C11 & C12 for the +/- 65vdc supply, I somehow replaced C15 & C16, instead. So, I went ahead and replaced C11 & C12. Unfortunately, no difference in the relay problem and not much difference in the waveform on the scope. Curiously, the schematics called for 220uf caps for C11, C12, C15 & C16. On my board, C11 & C12 were only 100uf. I went ahead and replaced then with 220uf, anyway.

 

[Brivan]

Will do. Also, I was getting ready to check the -65vdc on the power supply board, when I noticed a glaring mistake. In my attempt to replace caps C11 & C12 for the +/- 65vdc supply, I somehow replaced C15 & C16, instead. So, I went ahead and replaced C11 & C12. Unfortunately, no difference in the relay problem and not much difference in the waveform on the scope. Curiously, the schematics called for 220uf caps for C11, C12, C15 & C16. On my board, C11 & C12 were only 100uf. I went ahead and replaced then with 220uf, anyway.

I was able to get the scope to trigger as you described and checked all the pins of the PB & IC with the other channel. I didn't see anything out of the ordinary happening at the time of the trigger. I'll go through them again to be sure. I did see that I had a screwy-looking waveform on pin 1 of the PB. When checked with a meter, I found I had lost my 21vac at that point. I checked connections to the power supply board and AC input to it from the transformer. I found voltage in, but nothing out. I know I checked fuses before, but I found a blown 1-amp fuse. I must have inadvertently popped it during some of my testing. I replaced the fuse and got my 21vac back. Relay still cycling (of course). I measured AC voltage at pin 7 of the IC and got 0.652vac - about what I've had in the past. Interestingly, the relay cycles slower with the fuse replaced than it did with the fuse blown. The difference isn't huge, but it's noticeable. Makes me wonder what would happen if the 0.652vac to pin 7 of the IC was bumped up to a higher value.

 

[merlynski]

I was able to get the scope to trigger as you described and checked all the pins of the PB & IC with the other channel. I didn't see anything out of the ordinary happening at the time of the trigger. I'll go through them again to be sure. I did see that I had a screwy-looking waveform on pin 1 of the PB. When checked with a meter, I found I had lost my 21vac at that point. I checked connections to the power supply board and AC input to it from the transformer. I found voltage in, but nothing out. I know I checked fuses before, but I found a blown 1-amp fuse. I must have inadvertently popped it during some of my testing. I replaced the fuse and got my 21vac back. Relay still cycling (of course). I measured AC voltage at pin 7 of the IC and got 0.652vac - about what I've had in the past. Interestingly, the relay cycles slower with the fuse replaced than it did with the fuse blown. The difference isn't huge, but it's noticeable. Makes me wonder what would happen if the 0.652vac to pin 7 of the IC was bumped up to a higher value.

Which fuse? FU6? That feeds one of the rectifiers that powers +24vdc supply to the +13vdc regulator that feeds the PB pin 9.
Was the relay cycling while the AC was missing from PB pin 1? It should not be, because missing AC on PB pin 1 tells the PA3004 that the power switch was opened and that it should shut down the relay (off). That is the 'mute at power off' function.
The current through pin of PA3004 is about 40 uA, (18vac/470kohms). That clipped wave form there looks like a BE transistor junction. Trying to increase the voltage on pin 7 significantly might be difficult, but you could pull more current by paralleling R8 with another 470k (or so) resistor. Don't go too low, you might fry the IC.

 

[Brivan]

Yes, it was FU6. If I was missing 13vdc, I didn't notice - I know I measured that voltage (and others) umpteen times. My assumption was that the blown fuse was a recent development. At any rate, the relay cycled all through the whole blown fuse episode, albeit a little faster. But, yeah, you have a good point - the relay should have stopped cycling (assuming the IC is behaving normally). Paralleling that 470K resistor would have been my first guess. I'll piggyback another one on to it and check the voltage at the IC pin 7. I probably don't want to go higher than 1.25vdc or so.

 

[Brivan]

Yes, it was FU6. If I was missing 13vdc, I didn't notice - I know I measured that voltage (and others) umpteen times. My assumption was that the blown fuse was a recent development. At any rate, the relay cycled all through the whole blown fuse episode, albeit a little faster. But, yeah, you have a good point - the relay should have stopped cycling (assuming the IC is behaving normally). Paralleling that 470K resistor would have been my first guess. I'll piggyback another one on to it and check the voltage at the IC pin 7. I probably don't want to go higher than 1.25vdc or so.

Well, I started with 470K, then 220K, then 110K, which gave me a reading of 0.85vac at pin 7 of the IC. The relay cycling speed stayed the same throughout, so that doesn't appear to make a difference. Also, on a side note, I did scope the -65vdc supply and compare it to the +65vdc supply - the oscillations form both supplies look the same, so I'll assume that's normal for this receiver. I'll check more power supply voltages tomorrow, just to see if there are any similar fluctuations.

 

[merlynski]

Well, I started with 470K, then 220K, then 110K, which gave me a reading of 0.85vac at pin 7 of the IC. The relay cycling speed stayed the same throughout, so that doesn't appear to make a difference. Also, on a side note, I did scope the -65vdc supply and compare it to the +65vdc supply - the oscillations form both supplies look the same, so I'll assume that's normal for this receiver. I'll check more power supply voltages tomorrow, just to see if there are any similar fluctuations.

I did not expect it to change the cycling rate, the ac on pin 7 is a presence/absence detection, current flowing through pin 7 enables relay to be on.
Was the relay cycling when the 21vac was missing from PB pin 1? When missing I expect the relay to always be off. You could lift one end of PB R8, that should keep the relay off always.
EDIT: more info

 

[merlynski]

Yes, it was FU6. If I was missing 13vdc, I didn't notice - I know I measured that voltage (and others) umpteen times. My assumption was that the blown fuse was a recent development. At any rate, the relay cycled all through the whole blown fuse episode, albeit a little faster. But, yeah, you have a good point - the relay should have stopped cycling (assuming the IC is behaving normally). Paralleling that 470K resistor would have been my first guess. I'll piggyback another one on to it and check the voltage at the IC pin 7. I probably don't want to go higher than 1.25vdc or so.

If the +13vdc was missing the relay would never pull in, as that is the power supply for the PA3004. PS board D6 and D7 each are responsible for half of the rectification of the ac input to that supply. The blown fuse would have created greater ripple and lowered the +24vdc supply voltage to the +13vdc regulator, but maybe not low enough to cause the regulator output to drop. You could look at that with FU6 in and out to see.

 

[Brivan]

I did not expect it to change the cycling rate, the ac on pin 7 is a presence/absence detection, current flowing through pin 7 enables relay to be on.
Was the relay cycling when the 21vac was missing from PB pin 1? When missing I expect the relay to always be off.

Yes, it was. I didn't even notice the missing 21vac until I started going through the pins on the board and the IC with the o-scope. The relay cycles continuously, unless I ground something to make it stop. And it cycles with pins 4,5,6,7, & 8 of the protection board disconnected. All voltages going to the board and IC look normal - except for that sawtooth-pattern 65vdc, which we've concluded is probably normal.

 

[Brivan]

If the +13vdc was missing the relay would never pull in, as that is the power supply for the PA3004. PS board D6 and D7 each are responsible for half of the rectification of the ac input to that supply. The blown fuse would have created greater ripple and lowered the +24vdc supply voltage to the +13vdc regulator, but maybe not low enough to cause the regulator output to drop. You could look at that with FU6 in and out to see.

Tomorrow I can pull FU6 and see what I measure for voltages.

 

[petehall347]

for fun you could replace q2 .unless you tried that already ?
failing that make sure the grounding is good .

 

[Brivan]

for fun you could replace q2 .unless you tried that already ?
failing that make sure the grounding is good .

I did replace Q2 (and Q1 and all the caps) on my original protection board. When I connected up the replacement "good" used board, I got the exact same problem - relay cycling off and on. Since I've been doing additional testing, I reconnected my original "Frankenstein" board, saving my all-original replacement board for future use, if needed. So the board that's currently connected up to the receiver has had Q2 replaced. I did run a separate jumper from the ground pin of the protection board to the receiver frame, but it didn't make any difference for me. I'm at the point where I'd like to get the replacement PA3004 installed to see what, if any, difference it makes. I'm hopeful it solves this problem, since I've been unable to dig up any other issues.

 

[petehall347]

both boards same symptoms .and if protect signals removed pin 8 and pin 5 says power supply .but that appears ok to you .
you could lift R6 instead of pis 8 and 5 ..
sorry if i missed you trying this ..