Looks like you dumped the +48 rail into the O/P over current circuit. Not likely agreeable.... I'd look at Q21, Q212, Q214, Q216, Over curre3nt input cap C12/1 and op amp IC44 in the overload circuit. As a CAUTION note: Don't adjust anything while in protect mode! If you have try to put things back where they were before you continue.
Doc
Doc
I don't think D27 is connected to the output rail.
If the emitter of Q212 was connected to the positive rail then the Vbe max of Q212 would have been exceeded, blowing the driver transistor. This would also connect the base of the output transistor Q214 to the output through the RE resistor. The emitter of the output transistor is held @ the positive rail potential, with the base held a smidgen below. If the base of the output transistor were connected to the output of the amplifier, then the positive rail would also be connected to the output of the amplifier albeit through a 47R resistor, so the current would be limited to ~1 amp. Surely feedback would try and keep the amplifiers output @ zero. If the amplifiers output were kept close to zero the base of the power transistor would also be at zero, Veb max of the power transistor is -5 volts, ergo the base can be 5 volts below the emitter. With the emitter being held @ the positive rail voltage, if the base of the power transistor were to drop more then 5 volts, which it could quite easily do if the amplifier through feedback tried to keep the output at zero then the output transistor would die.
I suppose if the amplifier tried to compensate from having the +ve rail connected to the output via a 47R resistor then perhaps in compensating it placed too great a demand on the VAS and blew that.
Anything could have happened when you accidentally shorted things, one thing could have blown first that prevented the death of the others, like maybe the VAS blew and prevented feedback from keeping the output at zero and saved the output transistor etc, or you could have setup a cascade of silicon failure that blew a number of devices.
You say the DC offset is still quite low, but this is presumably on the outputs of the amplifier itself, at the output terminals, rather then at the amplifier PCB itself. Have you measured the DC offset before the inductor coil on the output?
Being able to set the bias seems like a good thing, have you made sure that the voltage dropped across each RE resistor is the same? The amplifier appears to test for the bias as the total across both of the RE resistors. Measure the voltage drop from TP1 to the output of the amplifier (before the output inductor L21) and then measure the voltage drop from TP2 to the output of the amplifier (again before L21), they should be almost identical. If they are not something is wrong.
If the emitter of Q212 was connected to the positive rail then the Vbe max of Q212 would have been exceeded, blowing the driver transistor. This would also connect the base of the output transistor Q214 to the output through the RE resistor. The emitter of the output transistor is held @ the positive rail potential, with the base held a smidgen below. If the base of the output transistor were connected to the output of the amplifier, then the positive rail would also be connected to the output of the amplifier albeit through a 47R resistor, so the current would be limited to ~1 amp. Surely feedback would try and keep the amplifiers output @ zero. If the amplifiers output were kept close to zero the base of the power transistor would also be at zero, Veb max of the power transistor is -5 volts, ergo the base can be 5 volts below the emitter. With the emitter being held @ the positive rail voltage, if the base of the power transistor were to drop more then 5 volts, which it could quite easily do if the amplifier through feedback tried to keep the output at zero then the output transistor would die.
I suppose if the amplifier tried to compensate from having the +ve rail connected to the output via a 47R resistor then perhaps in compensating it placed too great a demand on the VAS and blew that.
Anything could have happened when you accidentally shorted things, one thing could have blown first that prevented the death of the others, like maybe the VAS blew and prevented feedback from keeping the output at zero and saved the output transistor etc, or you could have setup a cascade of silicon failure that blew a number of devices.
You say the DC offset is still quite low, but this is presumably on the outputs of the amplifier itself, at the output terminals, rather then at the amplifier PCB itself. Have you measured the DC offset before the inductor coil on the output?
Being able to set the bias seems like a good thing, have you made sure that the voltage dropped across each RE resistor is the same? The amplifier appears to test for the bias as the total across both of the RE resistors. Measure the voltage drop from TP1 to the output of the amplifier (before the output inductor L21) and then measure the voltage drop from TP2 to the output of the amplifier (again before L21), they should be almost identical. If they are not something is wrong.
Last edited:
good news!Today i replaced Q21 and voila : green light at last!
As i had too little time i only managed to measure the offset in the outputs: 35mV on left and 37mV on right channel. I also noticed that the heatsinks of the drivers still get really hot. Except for repeating the allighment procedure again (and carefully this time), is there something i should check before connecting the amp to the speakers?
service manual
hi, do you still have that service manual?
Once I have done the same stupid thing
Someone tried to fix it, but the fix looks dodgy - it plays, but the red light nether turns green.
When I looked inside, I found the guy who was fixing it had connected D412 diode beside speaker terminal to the '-' speaker terminal. He also changed damaged resistor R230 and transistor Q28 to similar ones but bigger.
When it plays it's a nice amp, but the red light bothers me a lot.
hi, do you still have that service manual?
Once I have done the same stupid thing
Someone tried to fix it, but the fix looks dodgy - it plays, but the red light nether turns green.
When I looked inside, I found the guy who was fixing it had connected D412 diode beside speaker terminal to the '-' speaker terminal. He also changed damaged resistor R230 and transistor Q28 to similar ones but bigger.
When it plays it's a nice amp, but the red light bothers me a lot.
This story seems remarkably similar to mine. I was recapping the amp with films throughout and new electrolytics where not due to the documented problems with the decline in the electrolytics used in this amp.
I was rebiasing and I had a spark jump to transistor Q112. It has defo taken out output device Q115 and who knows what else. I guess this is a similar story but the other side of the amp. Unfortunately, being some years after these posters now I can't find some of the devices required. I've had a look about and there are some suggested devices though, I feel like I could do with a bit of advice on how to implement the repair though and if the suggested devices will fix it without any more sparks and magic smoke.
1) Replacement output device(s):
There is a definite short here (Q115) so parts must be replaced. 2SC5200, 2SA1943: Q1) Can I just replace the devices on one channel or would it be good practice to replace all output devices (there are only 4 and it's nearly as much work to replace 2 as 4)? I think the originals are Toshiba and the replacements I'm considering are ON semis...
Q2) Device pairing: do I have to buy a whole gang of these devices and pair them properly or will just buying the right grade so that they are approximately paired be good enough (got O grade in the basket)? What benefits would close matching have?
2) Driver replacements:
This is where the spark jumped to and I haven't had them out to test yet and they don't seem bust from my rudimentary tests but I can't imagine it's done them any good and maybe I'm testing incorrectly.
B649,D669: These devices are obsolete. There are recommended replacements of KSA1220 and KSC2690 (and one of these is actually in the original design). Q1) Again, can I just drop in if required? Q2) Are certain grades better matches to the original devices (got AYS grade in my Mouser basket)? Q3) How important is pairing?
Before the sparks and magic smoke incident I also found that like the above poster the drivers were getting very hot as well as the resistors R126/R127, R226/R227 to the point that there is visible discolouration of the circuit board. If there's an underlying issue that would stop this too then I'd like to sort that too.
Any help and advice gratefully received.
I was rebiasing and I had a spark jump to transistor Q112. It has defo taken out output device Q115 and who knows what else. I guess this is a similar story but the other side of the amp. Unfortunately, being some years after these posters now I can't find some of the devices required. I've had a look about and there are some suggested devices though, I feel like I could do with a bit of advice on how to implement the repair though and if the suggested devices will fix it without any more sparks and magic smoke.
1) Replacement output device(s):
There is a definite short here (Q115) so parts must be replaced. 2SC5200, 2SA1943: Q1) Can I just replace the devices on one channel or would it be good practice to replace all output devices (there are only 4 and it's nearly as much work to replace 2 as 4)? I think the originals are Toshiba and the replacements I'm considering are ON semis...
Q2) Device pairing: do I have to buy a whole gang of these devices and pair them properly or will just buying the right grade so that they are approximately paired be good enough (got O grade in the basket)? What benefits would close matching have?
2) Driver replacements:
This is where the spark jumped to and I haven't had them out to test yet and they don't seem bust from my rudimentary tests but I can't imagine it's done them any good and maybe I'm testing incorrectly.
B649,D669: These devices are obsolete. There are recommended replacements of KSA1220 and KSC2690 (and one of these is actually in the original design). Q1) Again, can I just drop in if required? Q2) Are certain grades better matches to the original devices (got AYS grade in my Mouser basket)? Q3) How important is pairing?
Before the sparks and magic smoke incident I also found that like the above poster the drivers were getting very hot as well as the resistors R126/R127, R226/R227 to the point that there is visible discolouration of the circuit board. If there's an underlying issue that would stop this too then I'd like to sort that too.
Any help and advice gratefully received.
Here's a link to the manuals. https://www.hifiengine.com/manual_library/nad/c320.shtml You need to register but it's free.
As a first step, build a dim bulb tester (google it) if you haven't already, and equip yourself with 40/60/80W bulb options. This gives you some current limiting while you repair the amp, don't plug it directly into mains until it is successfully coming out of protection. If you don't do this you'll keep blowing new parts while you figure it out, and you'll go around in circles getting frustrated.
I'd also get a cheap Mega328 component tester off ebay (or similar) if you haven't already, and start testing things. You can poke around with your meter and test for dead shorts on transistors (and diodes/zeners) in-circuit once you get a feel for things, then remove suspicious components and test properly.
Output devices don't *need* to be matched but it's good practice to match parallel devices when there are multiple BJT output pairs (eg, match the PNP outputs with the other PNP outputs on that channel). If there is only one output pair per channel (as you suggest - I didn't check the manual) then you don't need to worry - and no need to replace the good channel. I'd buy all 4 and keep the other 2 spare in case you mess up, but if there's one working channel definitely leave it untouched.
Do replace any BJT pairs as pairs though, especially when one is being replaced by a substitute, the other should be replaced with the substitute's complement (even if the original complement was working). This is probably the only case where you should replace a working part. The other case is if you suspect a zener is bad, as they can be a bit tricky to measure with basic gear.
From memory mouser doesn't often offer much choice in grades anyway, but if you have choices and want the best match - there are two quick ways to get in the ballpark. If the manual specifies grades for transistors, look up the datasheet and check the hFE values for those grades, get the highest grade off mouser that fits within the range of the old recommendation. Alternatively, use your Mega328 to measure the hFE on a known-good transistor (maybe from the good channel) and use that as your value to match new hFE from. Usually there's a big allowable range for most devices in mainstream amps like this.
hFE will rarely match between complementary BJT parts, the generally accepted method seems to be to match as close as possible for Vbe and match for hFE as a secondary concern. I wouldn't stress over it.
Finally, you can't easily measure those numbers for output transistors (as they need to be measured under load), so don't worry about those. Do measure and keep track of every part you remove though - because there are often errors in schematics and parts lists (but what's in the amp obviously works / worked). Do double check pinouts on your replacement transistors as the ECB order can sometimes be different - the component tester is your friend here.
RE transistor testing, see here:
Bipolar Junction Transistor Testing Basics | Audiokarma Home Audio Stereo Discussion Forums
I do enjoy how you keep saying you "had a spark jump" as if it happened by itself. Clearly you screwed up, don't worry everyone's done it (well I certainly have).
Hope that answers some of your questions. If you set yourself up with the precautions and testing gear, you'll have it fixed in no time, just be patient and don't take shortcuts. Also double check everything you fit/change for polarity and wrap your meter leads and screwdrivers before poking them into a powered amp.
Good luck!
I'd also get a cheap Mega328 component tester off ebay (or similar) if you haven't already, and start testing things. You can poke around with your meter and test for dead shorts on transistors (and diodes/zeners) in-circuit once you get a feel for things, then remove suspicious components and test properly.
Output devices don't *need* to be matched but it's good practice to match parallel devices when there are multiple BJT output pairs (eg, match the PNP outputs with the other PNP outputs on that channel). If there is only one output pair per channel (as you suggest - I didn't check the manual) then you don't need to worry - and no need to replace the good channel. I'd buy all 4 and keep the other 2 spare in case you mess up, but if there's one working channel definitely leave it untouched.
Do replace any BJT pairs as pairs though, especially when one is being replaced by a substitute, the other should be replaced with the substitute's complement (even if the original complement was working). This is probably the only case where you should replace a working part. The other case is if you suspect a zener is bad, as they can be a bit tricky to measure with basic gear.
From memory mouser doesn't often offer much choice in grades anyway, but if you have choices and want the best match - there are two quick ways to get in the ballpark. If the manual specifies grades for transistors, look up the datasheet and check the hFE values for those grades, get the highest grade off mouser that fits within the range of the old recommendation. Alternatively, use your Mega328 to measure the hFE on a known-good transistor (maybe from the good channel) and use that as your value to match new hFE from. Usually there's a big allowable range for most devices in mainstream amps like this.
hFE will rarely match between complementary BJT parts, the generally accepted method seems to be to match as close as possible for Vbe and match for hFE as a secondary concern. I wouldn't stress over it.
Finally, you can't easily measure those numbers for output transistors (as they need to be measured under load), so don't worry about those. Do measure and keep track of every part you remove though - because there are often errors in schematics and parts lists (but what's in the amp obviously works / worked). Do double check pinouts on your replacement transistors as the ECB order can sometimes be different - the component tester is your friend here.
RE transistor testing, see here:
Bipolar Junction Transistor Testing Basics | Audiokarma Home Audio Stereo Discussion Forums
I do enjoy how you keep saying you "had a spark jump" as if it happened by itself. Clearly you screwed up, don't worry everyone's done it (well I certainly have).
Hope that answers some of your questions. If you set yourself up with the precautions and testing gear, you'll have it fixed in no time, just be patient and don't take shortcuts. Also double check everything you fit/change for polarity and wrap your meter leads and screwdrivers before poking them into a powered amp.
Good luck!
This pic is all you need for a bulb tester, you can buy the bits from a hardware store:
https://antiqueradio.org/art/DimbulbSketch.jpg
(or you probably have them at home)
The bulb has to be incandescent, can be a bit annoying to find these days (read the "actual" wattage).
https://antiqueradio.org/art/DimbulbSketch.jpg
(or you probably have them at home)
The bulb has to be incandescent, can be a bit annoying to find these days (read the "actual" wattage).
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.