Hopefully in the correct forum.....
I'm going a bit mad trying to trace a problem with what I think is false triggering of the protection circuit on a Pioneer A-400X amp.
There seem to be no apparent problems with DC offset (about 2mV on each channel). After being left switched off for a while the amp will happily work for maybe 40 minutes (with a signal) then the protection kicks in. With no signal the protection kicks in after about 10 minutes from cold but when warm the amp might never come out of protection or will almost immediately go back in to protection after the startup delay. (Is the no signal thing a clue?). There seems to be either a temperature or residual charge element to this perhaps?
However, when the protection kicks in there's no indication why that I can see - no DC offset, overloading didn't happen.
I'm thinking its the protection circuit false triggering and so I removed transistors Q303 and Q304 which I think detect DC offset and also the resistor R308 which connects to each channel's overload detection. That changed nothing and I can't see how!
Am I right in thinking that protection is triggered by either the DC offset detection or overload circuits pulling the base of Q305 low so that it turns on and discharges C305 - that in turn results in the relay being deactivated. Once Q305's base has been pulled low it latches itself on by virtue of its base current through R309?
I've replaced C304 incase that was leaky, Q305 itself and C305 but no effect. With the components named above removed I can't see what is able to switch Q305 on. Once triggered the voltage at C305 cycles varying between about 4 and 18 volts.
The supply voltages on the B+ and B- are both 42.9v of the specified 43.3v and the protection circuit has its own supply bang on at 41.9v (albeit with about 1v of ripple despite having changed its filter cap too).
I can't see how this fault might be caused although D306 and D305 are a bit mysterious to me - some kind of over-voltage protection?
Any ideas? - I can no longer see which components might be responsible!
Notes on the attached circuit diagram - the power for the protection circuit comes via D308. The "bottom" of R311, R316 et al go to ground.
I'm going a bit mad trying to trace a problem with what I think is false triggering of the protection circuit on a Pioneer A-400X amp.
There seem to be no apparent problems with DC offset (about 2mV on each channel). After being left switched off for a while the amp will happily work for maybe 40 minutes (with a signal) then the protection kicks in. With no signal the protection kicks in after about 10 minutes from cold but when warm the amp might never come out of protection or will almost immediately go back in to protection after the startup delay. (Is the no signal thing a clue?). There seems to be either a temperature or residual charge element to this perhaps?
However, when the protection kicks in there's no indication why that I can see - no DC offset, overloading didn't happen.
I'm thinking its the protection circuit false triggering and so I removed transistors Q303 and Q304 which I think detect DC offset and also the resistor R308 which connects to each channel's overload detection. That changed nothing and I can't see how!
Am I right in thinking that protection is triggered by either the DC offset detection or overload circuits pulling the base of Q305 low so that it turns on and discharges C305 - that in turn results in the relay being deactivated. Once Q305's base has been pulled low it latches itself on by virtue of its base current through R309?
I've replaced C304 incase that was leaky, Q305 itself and C305 but no effect. With the components named above removed I can't see what is able to switch Q305 on. Once triggered the voltage at C305 cycles varying between about 4 and 18 volts.
The supply voltages on the B+ and B- are both 42.9v of the specified 43.3v and the protection circuit has its own supply bang on at 41.9v (albeit with about 1v of ripple despite having changed its filter cap too).
I can't see how this fault might be caused although D306 and D305 are a bit mysterious to me - some kind of over-voltage protection?
Any ideas? - I can no longer see which components might be responsible!
Notes on the attached circuit diagram - the power for the protection circuit comes via D308. The "bottom" of R311, R316 et al go to ground.
Hi Mooly, yes I changed that cap as I wasn't sure if the 1v ripple was more than to be expected. It didn't seem to make any difference.
Thanks.
Thanks.
Excellent. So you have a scope?
I would begin by working back from the relay driver Q306. When the relay drops out and so it is in the faulty state check offset for both channels, maybe use the scope to make sure nothing weird is present. Check the relay supply is OK and that all those series dropper resistors are OK. Look at the base voltage of Q306 with the scope.
Is Q301 (and presumably similar off the page for the other channel) pulling Q305 base low?
There should be no voltage across R306. If there is then Q301 (or the other channels one) is conducting.
Its not impossible for something like R329 (22k) to be high. Highish voltage and small high value resistors can and do fail.
If you still get nowhere then perhaps carefully record all the transistor voltages in that part of the circuit and list them clearly and we'll take a look.
1 volt ripple is fine. The cap is small to ensure a fast drop out on mains disconnect
Thanks Mooly - a few replies in bold. I'll learn to use the quote feature next time!
Excellent. So you have a scope? I do
I would begin by working back from the relay driver Q306. When the relay drops out and so it is in the faulty state check offset for both channels, maybe use the scope to make sure nothing weird is present. Check the relay supply is OK and that all those series dropper resistors are OK. Look at the base voltage of Q306 with the scope. Will do although as I mentioned I resorted to removing R308 and Q303 and Q304 so that that neither the offset and overload detection circuits are connected and still the protection kicks in after some period.
Is Q301 (and presumably similar off the page for the other channel) pulling Q305 base low? I don't believe it is - especially with R308 disconnected.
There should be no voltage across R306. If there is then Q301 (or the other channels one) is conducting. Again, I'll check but see note re R308.
Its not impossible for something like R329 (22k) to be high. Highish voltage and small high value resistors can and do fail. Will check. The relay does switch on but gets turned off by the C305 voltage dropping from 22v to some varying value less presumably due to Q305 switching on.
I may be barking up the wrong tree but it's how Q305 is still being switched on with R308, Q303 and Q304 out that's confusing me. At least I think it is - I need to check the voltage across R309 when C305 is being pulled low.
If you still get nowhere then perhaps carefully record all the transistor voltages in that part of the circuit and list them clearly and we'll take a look. Thanks!
Excellent. So you have a scope? I do
I would begin by working back from the relay driver Q306. When the relay drops out and so it is in the faulty state check offset for both channels, maybe use the scope to make sure nothing weird is present. Check the relay supply is OK and that all those series dropper resistors are OK. Look at the base voltage of Q306 with the scope. Will do although as I mentioned I resorted to removing R308 and Q303 and Q304 so that that neither the offset and overload detection circuits are connected and still the protection kicks in after some period.
Is Q301 (and presumably similar off the page for the other channel) pulling Q305 base low? I don't believe it is - especially with R308 disconnected.
There should be no voltage across R306. If there is then Q301 (or the other channels one) is conducting. Again, I'll check but see note re R308.
Its not impossible for something like R329 (22k) to be high. Highish voltage and small high value resistors can and do fail. Will check. The relay does switch on but gets turned off by the C305 voltage dropping from 22v to some varying value less presumably due to Q305 switching on.
I may be barking up the wrong tree but it's how Q305 is still being switched on with R308, Q303 and Q304 out that's confusing me. At least I think it is - I need to check the voltage across R309 when C305 is being pulled low.
If you still get nowhere then perhaps carefully record all the transistor voltages in that part of the circuit and list them clearly and we'll take a look. Thanks!
Worth also checking R310. That going high could make it appear like Q305 was conducting and pulling the voltage down across C305. I know small resistors never fail 🙂 but occasionally they really do.
If you remove R306 (is it? blurry) you remove the input from the the overcurrent detection. That would eliminate one route.
You have pulled Q303 and Q304 so that eliminates the offset sensing.
Quick simulation of the relay driver. AC is supplied here from 0 to 7 seconds. The relay coil supply I left as permanent because in practice the rail feeding that will fall slowly.
You can see how quickly the circuits supply falls (that tiny reservoir cap) and also how the switch on is 'clean' and 'sharp'. The drop out is reasonably well behaved.
This should click and run. Relay current and supply shown. I can't see any obvious reason for the two series diodes other than being a clamp of some sort perhaps under a fault condition of high offset. There is zero current through them at power on and power off in the simulation.
You can see how quickly the circuits supply falls (that tiny reservoir cap) and also how the switch on is 'clean' and 'sharp'. The drop out is reasonably well behaved.
This should click and run. Relay current and supply shown. I can't see any obvious reason for the two series diodes other than being a clamp of some sort perhaps under a fault condition of high offset. There is zero current through them at power on and power off in the simulation.
That's great. Thanks so much for your help. I will check the resistors as suggested. R310 going high is my hope.
Yes, the blurry resistor R308, or is it R306, was pulled to eliminate the overload (quoted from memory but did trace on the board at the time though will double check that).
I can't help thinking I've missed something bloomin' obvious here but we'll see!
Yes, the blurry resistor R308, or is it R306, was pulled to eliminate the overload (quoted from memory but did trace on the board at the time though will double check that).
I can't help thinking I've missed something bloomin' obvious here but we'll see!
Voltage checks (but perhaps done with the scope on this one) should show where it falls down. With the amp on and in a faulty state begin by confirming the offset is zero on both channels and that the over current sensing (Q301/302) are not on.
Next just work back from Q306 base and see what is pulling that voltage down.
Also make sure there is enough relay current available and those 5 series resistors are OK.
Next just work back from Q306 base and see what is pulling that voltage down.
Also make sure there is enough relay current available and those 5 series resistors are OK.
The boomin obvious thing to check first is to make sure the power supply is OK and stays that way during fault conditions. Electronic service 101.
OK, have done some more readings:
These are with the fault condition present and with Q303 and Q304 removed so no DC offset sensing and R308 removed so no overcurrent sensing. R308 is the blurry one above Q303 and Q304 on the diagram which I've traced on the board.
To answer another poster's question there is no change on the power supply rails with the fault present - the B+ and B- are 43.2 and -43.2 volts respectively. The separate supply to the protection circuit is 41.9 volts.
Q306 collector 43.2 volts (this comes from B+, not the protection circuit supply).
Q306 base 0 volts - the 22v Zener is not conducting.
Q307 collector 41.9 volts.
Now the odd one - At the "top" of zener D307, either side of R328 and either side of R309 there is a voltage cycling between about 6 and 18 volts. The cycle is irregular - sometimes rasing by a volt or two then collapsing back to about 6 volts but sometimes climbing to about 18 volts. The periodicity is about 2 to 3 seconds.
Using both scope channels, the left-hand side of R309 is about 600mV lower than the right so am I correct in thing that Q305 is on and keeping itself switched on by virtue of the current through its base and R309? (Q305 and C304 have been changed). If Q305 is on I'm curious how it gets into that state when the amplifier is switched on but warm - if fully cold it takes a few minutes for protection to kick in. So it does appear to be Q305 pulling the voltage down but to me the question is what switches it on in the first place.
Despite being disconnected I did check Q301 and Q302 which have 1-2mV on their bases and therefore off. The offsets on both channel's big overcurrent detection resistors was around 6mV.
These are with the fault condition present and with Q303 and Q304 removed so no DC offset sensing and R308 removed so no overcurrent sensing. R308 is the blurry one above Q303 and Q304 on the diagram which I've traced on the board.
To answer another poster's question there is no change on the power supply rails with the fault present - the B+ and B- are 43.2 and -43.2 volts respectively. The separate supply to the protection circuit is 41.9 volts.
Q306 collector 43.2 volts (this comes from B+, not the protection circuit supply).
Q306 base 0 volts - the 22v Zener is not conducting.
Q307 collector 41.9 volts.
Now the odd one - At the "top" of zener D307, either side of R328 and either side of R309 there is a voltage cycling between about 6 and 18 volts. The cycle is irregular - sometimes rasing by a volt or two then collapsing back to about 6 volts but sometimes climbing to about 18 volts. The periodicity is about 2 to 3 seconds.
Using both scope channels, the left-hand side of R309 is about 600mV lower than the right so am I correct in thing that Q305 is on and keeping itself switched on by virtue of the current through its base and R309? (Q305 and C304 have been changed). If Q305 is on I'm curious how it gets into that state when the amplifier is switched on but warm - if fully cold it takes a few minutes for protection to kick in. So it does appear to be Q305 pulling the voltage down but to me the question is what switches it on in the first place.
Despite being disconnected I did check Q301 and Q302 which have 1-2mV on their bases and therefore off. The offsets on both channel's big overcurrent detection resistors was around 6mV.
Thinking out loud here but could it be something like a burst of noise being passed by C304 that switches Q305 on which then holds itself on?
Those all sound 'correct' for being in the faulty state
it should be steady of course and 18v won't bring the Zener and relay into conduction.
Maybe. The devil is in the detail. Even if the base were lifted you would still measure around 500mv less at the base lead simply because you are reading through what is effectively a diode junction.
This is extremely odd... for sure.
Question? You've done everything I would have done and you seem to know exactly what to expect and your theory is good 👍 so I'm going to ask this...
Has the amp got a hidden history? Is it yours or one you 'acquired'. One possibility is spillage of something conductive on the board for example. Not always easy to detect either. There is only C304 that could be pulling the base down and you changed that.
(The circuit I was working from is blurry and I've corrected a couple of mistake in the simulation... the Zener and a couple of component references. I'll reattach the sim file later)
So with these removed there is nothing obvious to turn on Q305
I've acquired this amp so don't know its history too well. The board was dusty but otherwise clean and I gave the solder side a clean with a soft brush and some alcohol. I've had a good look for any strands of wire, solder splashes etc but nothing to be seen.
I've since put my scope on the base of Q305 and I do see very brief and intermittent spikes both positive and negative of up to 1 volt which I wonder could be switching on Q305 when it's off if they're fast enough for C304 to pass.
There's still the question of why all is ok for a few minutes and then once the problem has occurred the amp won't come out of protection unless at switch on until allowed to thoroughly cool. Next step is to resolder the joints I've not already done in case it is a heat-sensitive bad one somewhere.
I've since put my scope on the base of Q305 and I do see very brief and intermittent spikes both positive and negative of up to 1 volt which I wonder could be switching on Q305 when it's off if they're fast enough for C304 to pass.
There's still the question of why all is ok for a few minutes and then once the problem has occurred the amp won't come out of protection unless at switch on until allowed to thoroughly cool. Next step is to resolder the joints I've not already done in case it is a heat-sensitive bad one somewhere.
What I would do next is desolder just the base of Q305 so it is floating and not touching anything and see if the relay pulls in. You could also look with the scope at the print where the base would connect and see if anything is there. Also scope the cathode of D308 just to check the supply there is stable.
Problem looking like it's solved thanks to Mooly's suggestion and not so much the service manual!
After lifting the base of Q305 the relay stayed on thus showing that Q305 being switched on was turning off the relay as though a genuine trigger (over current, DC offset) had occurred even though those circuits were disconnected.
On connecting the scope to the R309-C304 junction, i.e. where Q305's base would have been, there were a lot of random noise transients spread out by a few seconds. Some were about 500mV and others far larger and off the scale so over 2 volts.
I disconnected R309 and D306 in turn which made no difference which confused me a bit as I had also pulled R308 so I thought there were no other components involved and wondered what else could be generating the noise. I then started tracing tracks and came across D309 which is NOT on the circuit diagram. Thanks Pioneer....
D309 goes cathode towards R308 and is directly on the protection circuit side of it albeit some distance away on the board which is why I had not noticed it. That's my excuse anyway...
Pulling D309 resulted is the noise disappearing. I'm not sure what type it is but replaced it with an 1N4148 and everything is now stable - no protection kicking in!
I'll reconnect the DC offset and over current circuits and hopefully all well.
Hi Mooly - D309 is in series with R308 cathode towards R308, anode towards R309, Q303, Q304.
I feel slightly better that it was a component not on the diagram! It was amazing what that diode was generating without any current flow!