That sounds ok, always measure anything like this from 0 volts. Pin 8 will be +15 and pin 4 -15. However measure those junctions directly across the transistor, dont measure these from ground and try and work out the drop, for various reasons it does not always work as you expect.
I really have to disappear now but post the results and I will write down a sequence of readings to take that should help you find the fault if you have no luck. Even though the voltages are wildly wrong the answer is there.
I really have to disappear now but post the results and I will write down a sequence of readings to take that should help you find the fault if you have no luck. Even though the voltages are wildly wrong the answer is there.
O.K. Had time to think about this one. There are a couple of things to try to eliminate. The problem may be caused by circuit around the OpAmp. This is configured as a DC servo to hold the output at zero volts with no signal. Could you short out that FET Drain and Source with a bit of wire so that it grounds the junction of R49/50 and R19. Measure the speaker volts again, is there any change ? This will eliminate the servo and the amp should settle close to zero.
If there is no change leave the short in place and read the volts on the collector of Q7. With the output at -37volts the voltage here should be around -35 ish.
Read the volts on collector of Q8, and again with output at -37 this should also be around -35.
If there is no change leave the short in place and read the volts on the collector of Q7. With the output at -37volts the voltage here should be around -35 ish.
Read the volts on collector of Q8, and again with output at -37 this should also be around -35.
I'll give that a try when it stops working again!!!
I switched it all off last night, and went to make some dinner. Decided to come back to it and go through systematically trying to take voltage readings and voltage drops across as much as I could find... except the output was now at zero volts.
So after prodding for a good half an hour, and the amp behaving perfectly well with zero volts on the output, I decided to go for broke and replace the 60W lamp with the fuse and try it again under normal circumstances. All fine and happy, although after just under an hour, the scratchy noise was back in the left channel (but no pops, clicks or humming).
I'm actually away today with work commitments, so won't get much chance to investigate more today, but I'll be back onto it as soon as I can.
Thanks for the next suggestion... I'll let you know the information once it's failed again.
Regards,
Wayne
I switched it all off last night, and went to make some dinner. Decided to come back to it and go through systematically trying to take voltage readings and voltage drops across as much as I could find... except the output was now at zero volts.
So after prodding for a good half an hour, and the amp behaving perfectly well with zero volts on the output, I decided to go for broke and replace the 60W lamp with the fuse and try it again under normal circumstances. All fine and happy, although after just under an hour, the scratchy noise was back in the left channel (but no pops, clicks or humming).
I'm actually away today with work commitments, so won't get much chance to investigate more today, but I'll be back onto it as soon as I can.
Thanks for the next suggestion... I'll let you know the information once it's failed again.
Regards,
Wayne
Or the OpAmp itself, unlikely but you never know. Swap the FET with the other channel and that would eliminate it, if it was duff fault would move to other channel. My other gut feeling is that it is a driver transistor playing up, Q11/12, they read O.K. when off but breakdown under voltage.
Edit, Freezer spray may help, dont blast everything but drip it onto suspect part, and watch your speakers, under fault conditions.
Edit, Freezer spray may help, dont blast everything but drip it onto suspect part, and watch your speakers, under fault conditions.
Yes, I'll be watching the speakers. I would say that the amp has actually been on for the best part of the last fifteen years or so - the exceptions being holidays, power cuts and moving house... at the moment, however, it'll be off if I'm not actually using it (which is probably worse for it with the heat-up - cool-down cycle that entails). That way I can quickly turn it off when it starts popping again.
I need to head to our local Maplin store today to pick up some more rubber feet for it - the originals had turned to jelly and had started to go all gooey. I'll pick up some freezer spray while there... ready for round two.
Many thanks again. Much appreciated.
Wayne
I need to head to our local Maplin store today to pick up some more rubber feet for it - the originals had turned to jelly and had started to go all gooey. I'll pick up some freezer spray while there... ready for round two.
Many thanks again. Much appreciated.
Wayne
Latest info...
The amp just hasn't wanted to lose that channel again, although the scratchiness was back after an hour or so of use (the scratchiness was faint, sounding a little like it does when you move a worn volume control - the scratchiness doesn't change volume with the volume control, however).
After hearing the scratchiness coming back in the left channel, I thought I'd re-solder the joints around the resistor which had gotten rather hot (R40) just to ensure that they aren't dry joints, or cracked due to the excessive heat from that resistor.
The first attempt, and the speakers emitted the 'broken' hum, so I rather rapidly switched the amp off and tried to resolder again.
This time, after double-de-soldering and re-soldering, it seems to have made a good joint and the hum is gone and, after having the amp on in the background for five hours so far (touch wood), it's as quiet and smooth as you like - no scratchiness and no pops and hums.
(As a note, it looks as though the circuit board track leading to one side of R40 has lifted from the actual board itself - excessive heat from when it initially emitted the smells maybe? That could explain why my first solder joint didn't work properly. What I might do is to solder a fat piece of wire directly between that leg of the resistor and the nearest solder point for that track on the underside of the PCB - just to ensure a good connection given heat changes).
Now I'm still not quite sure where this leaves me and the problem... on the one hand, a dry joint could have caused a break in the circuit, causing a bunch of stuff to go to -36V and giving the hum on the speaker. The pop in the speaker is just the output swapping straight to -36V and back again. That bit I get.
What I don't understand is that if the break in the circuit due to the dry joint really was one of the legs of R40, then surely that resistor would effectively be open-circuit on one end, and thus no current flowing through it, and thus it wouldn't get hot.
If one of the sides of R40 wasn't quite broken, but wasn't quite joined would surely increase the resistance at that joint, effectively increasing the total resistance for R40 which would obviously affect the circuit behaviour, but I don't quite inderstand how this would cause R40 to overheat?
Perhaps the problem still exists somewhere and will come back to haunt me another day?
Many thanks again, Mooly, for all your help and suggestions with this... I may be back in touch again in the future when/if it goes pop again!
Wayne
The amp just hasn't wanted to lose that channel again, although the scratchiness was back after an hour or so of use (the scratchiness was faint, sounding a little like it does when you move a worn volume control - the scratchiness doesn't change volume with the volume control, however).
After hearing the scratchiness coming back in the left channel, I thought I'd re-solder the joints around the resistor which had gotten rather hot (R40) just to ensure that they aren't dry joints, or cracked due to the excessive heat from that resistor.
The first attempt, and the speakers emitted the 'broken' hum, so I rather rapidly switched the amp off and tried to resolder again.
This time, after double-de-soldering and re-soldering, it seems to have made a good joint and the hum is gone and, after having the amp on in the background for five hours so far (touch wood), it's as quiet and smooth as you like - no scratchiness and no pops and hums.
(As a note, it looks as though the circuit board track leading to one side of R40 has lifted from the actual board itself - excessive heat from when it initially emitted the smells maybe? That could explain why my first solder joint didn't work properly. What I might do is to solder a fat piece of wire directly between that leg of the resistor and the nearest solder point for that track on the underside of the PCB - just to ensure a good connection given heat changes).
Now I'm still not quite sure where this leaves me and the problem... on the one hand, a dry joint could have caused a break in the circuit, causing a bunch of stuff to go to -36V and giving the hum on the speaker. The pop in the speaker is just the output swapping straight to -36V and back again. That bit I get.
What I don't understand is that if the break in the circuit due to the dry joint really was one of the legs of R40, then surely that resistor would effectively be open-circuit on one end, and thus no current flowing through it, and thus it wouldn't get hot.
If one of the sides of R40 wasn't quite broken, but wasn't quite joined would surely increase the resistance at that joint, effectively increasing the total resistance for R40 which would obviously affect the circuit behaviour, but I don't quite inderstand how this would cause R40 to overheat?
Perhaps the problem still exists somewhere and will come back to haunt me another day?
Many thanks again, Mooly, for all your help and suggestions with this... I may be back in touch again in the future when/if it goes pop again!
Wayne
Hi, Yep, intermitant faults can be a real pain. I tell you whats worth doing, with a clean well tined iron go over the joints particularly Q11/12 and Q7 and Q8. In normal running does Q8 run hot. The solder should flow cleanly round the joints and bond to the component instantly, if the solder won't take instantly or looks crusty this could be the problem. Just by chance is that I/C in a socket. If it is could be poor contacts. Even though it's working now I would be tempted to swap that FET with the other channel, so if fault reappears you know for sure it's not that.
Have fun and if you have any questions post them here. It's great working on stuff like this, it's like a police investigation, all the evidence is there, you just have to piece it together
Regards Karl
p.s. When it's all 100% how about swapping those NE5532's for something audibly better.
Have fun and if you have any questions post them here. It's great working on stuff like this, it's like a police investigation, all the evidence is there, you just have to piece it together
Regards Karl
p.s. When it's all 100% how about swapping those NE5532's for something audibly better.
Yes... I know all about intermittent faults - mainly of the software form these days, as I work as a software developer. The number of customer bug reports we get that are simply 'it crashes' is unbelievable. I hope I've been better at giving relevant info that that!
The op-amp IC's are soldered straight onto the motherboard - I used to repair ZX Spectrums as a kid, and making sure the ICs were seated was the first thing I'd be trying, so I think they're OK.
Under normal running conditions, Q8 is fairly warm; not overly hot, and about the same as Q108 for the other channel. I can leave my little finger on it for a good five seconds without having to remove it.
By comparison, R40 (and R140) run hotter - I have to remove my finger after some 3-4 seconds on both.
Speaking of temperatures - and I'm not sure if this is normal, having never actually tested prior to having the problems - but the amp does feel rather warm around the back where the Q13/Q14/Q113/Q114 heatsinks joint to the casing. I can only just about hold my fingers on the casing on the back after it's been running for a while. Given that the sound is pretty clean, I would imaging it's normal.
Regards,
Wayne
PS. I'm open to suggestions for the NE5532s. I think it's one of those things I've never really considered while the amp was working (obviously in the early days it was in warranty). But now that it's probably only worth £30 on e-bay, I wouldn't have much to lose by trying something.
PPS. I've actually got hold of an Alpha 9 via e-bay as a replacement anyway - perhaps I can migrate the Alpha 3 to the kitchen!
The op-amp IC's are soldered straight onto the motherboard - I used to repair ZX Spectrums as a kid, and making sure the ICs were seated was the first thing I'd be trying, so I think they're OK.
Under normal running conditions, Q8 is fairly warm; not overly hot, and about the same as Q108 for the other channel. I can leave my little finger on it for a good five seconds without having to remove it.
By comparison, R40 (and R140) run hotter - I have to remove my finger after some 3-4 seconds on both.
Speaking of temperatures - and I'm not sure if this is normal, having never actually tested prior to having the problems - but the amp does feel rather warm around the back where the Q13/Q14/Q113/Q114 heatsinks joint to the casing. I can only just about hold my fingers on the casing on the back after it's been running for a while. Given that the sound is pretty clean, I would imaging it's normal.
Regards,
Wayne
PS. I'm open to suggestions for the NE5532s. I think it's one of those things I've never really considered while the amp was working (obviously in the early days it was in warranty). But now that it's probably only worth £30 on e-bay, I wouldn't have much to lose by trying something.
PPS. I've actually got hold of an Alpha 9 via e-bay as a replacement anyway - perhaps I can migrate the Alpha 3 to the kitchen!
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.