First time doing this kind of things, so I'm quite scarely cat. 
Bought this faulty NAD 3020i (3225PE actually, but same board and schematics), owner claimed only one channel working when playing CD and phono + possibly the whole preamp is out.
I don't consider something inside glowing and smoking as "one channel working" as it would take a lot of guts to continue operation. Maybe the other channel is working, how'd I know.
Garage sales integrity is low in my country.
The one thing that was glowing was Q409 (BD139), which I saw burn mark before powering up. After the smoking the burn mark became bigger and it cracked so I'm pretty sure it's dead.
I've removed the transistor in question (marked with red cross) and is going to get a replacement. Now, question is, is it safe to turn on the amp so that I can check the rest of the amp? Also, if I replaced the transistor, would the new transistor fail like the old one due to unknown reason, in which case how do I check for that?
R443 was set at 25.5ohms, while in the other channel it was 24.5ohms, could a wrong biasing caused the eventual failure of this transistor?
And R445 (also marked with a cross) looks to have some kind of burn mark though I'm not sure if it was glowing. Given the schematics, would it have glown (to coin a term)?
Attached is part of the schematics:
Bought this faulty NAD 3020i (3225PE actually, but same board and schematics), owner claimed only one channel working when playing CD and phono + possibly the whole preamp is out.
I don't consider something inside glowing and smoking as "one channel working" as it would take a lot of guts to continue operation. Maybe the other channel is working, how'd I know.
Garage sales integrity is low in my country.
The one thing that was glowing was Q409 (BD139), which I saw burn mark before powering up. After the smoking the burn mark became bigger and it cracked so I'm pretty sure it's dead.
I've removed the transistor in question (marked with red cross) and is going to get a replacement. Now, question is, is it safe to turn on the amp so that I can check the rest of the amp? Also, if I replaced the transistor, would the new transistor fail like the old one due to unknown reason, in which case how do I check for that?
R443 was set at 25.5ohms, while in the other channel it was 24.5ohms, could a wrong biasing caused the eventual failure of this transistor?
And R445 (also marked with a cross) looks to have some kind of burn mark though I'm not sure if it was glowing. Given the schematics, would it have glown (to coin a term)?
Attached is part of the schematics:
An externally hosted image should be here but it was not working when we last tested it.
wwenze said:
Dead just because it's cracked and smoking? Are you not rushing to conclusions?
It is actually quite unlikely it's the only device blown. The drivers and outputs may not show cracks and a glow but have likely moved to a better place too.
Just check all semis on the board with a diode tester and suspicious looking resistors with an ohmmeter. And once you've replaced all the dead bodies make sure you power up carefully - through current limiting resistors or a lightbulb.
Re: Re: NAD 3020i repair help?
25.5 vs 24.5 is not reeally a difference to consider (4%)
We can say this is within the 5% tolerance factor.
In audio many times even a difference of 20% or more in telrance can be quite normal and good working.
This is not a bad advice.
And what you should do first.
Then post back and tell more.
Surely we have many guys that knows NAD 3020i in this forum.
I have read plenty of topics about this nice old CLASSICAL amplifier.
regars
Lineup your true auudio friend
wwenze said:
25.5 vs 24.5 is not reeally a difference to consider (4%)
We can say this is within the 5% tolerance factor.
In audio many times even a difference of 20% or more in telrance can be quite normal and good working.
analog_sa said:
This is not a bad advice.
And what you should do first.
Then post back and tell more.
Surely we have many guys that knows NAD 3020i in this forum.
I have read plenty of topics about this nice old CLASSICAL amplifier.
regars
Lineup
your true auudio friendRe: Re: NAD 3020i repair help?
When I removed it one of the pin along with part of the transistor had disengaged itself from the main body. The other 2 remaining pins measured 100ohms in both directions.
This thing looks like it has been repaired before, albeit shabbily. At least one other transistor has been replaced (and it is green-color, strange), and this broken one had been replaced from the original ST part to something that doesn't even have a brand on it. So I'm not surprised it failed again, but suspect something might be causing it to fail.
Thanks for the suggestion of testing all the parts to make sure. They're all still on the PCB but I'll try my best at measuring.
analog_sa said:
Dead just because it's cracked and smoking? Are you not rushing to conclusions?
It is actually quite unlikely it's the only device blown. The drivers and outputs may not show cracks and a glow but have likely moved to a better place too.
Just check all semis on the board with a diode tester and suspicious looking resistors with an ohmmeter. And once you've replaced all the dead bodies make sure you power up carefully - through current limiting resistors or a lightbulb.
When I removed it one of the pin along with part of the transistor had disengaged itself from the main body.
The other 2 remaining pins measured 100ohms in both directions.This thing looks like it has been repaired before, albeit shabbily. At least one other transistor has been replaced (and it is green-color, strange), and this broken one had been replaced from the original ST part to something that doesn't even have a brand on it. So I'm not surprised it failed again, but suspect something might be causing it to fail.
Thanks for the suggestion of testing all the parts to make sure. They're all still on the PCB but I'll try my best at measuring.
That transistor is the bias control transistor. If that's smoked you can guarantee the output transistors and drivers are smoked too. This has probably, in turn, smoked the bias device.
So, you are looking at a replacement for the outputs Q415/417, drivers Q411/Q413.
Which transistor is the green one? I bet it is the 2SB649.
You would be best pulling ALL transistors in the channel and testing them out of circuit. Similarly test the resistors and diodes.
DO NOT turn the amp on with this transistor removed! If the outputs and drivers arent dead yet they WILL be with a large bang if you do this!
So, you are looking at a replacement for the outputs Q415/417, drivers Q411/Q413.
Which transistor is the green one? I bet it is the 2SB649.
You would be best pulling ALL transistors in the channel and testing them out of circuit. Similarly test the resistors and diodes.
DO NOT turn the amp on with this transistor removed! If the outputs and drivers arent dead yet they WILL be with a large bang if you do this!
You're so right about that! 
I measured the above 4 mentioned trans in circuit and they're all shorted. And the green one is indeed 2SB649.
Guess I'll be pulling out everything around that 5 dead trans to check for collateral damage. That means over 20.
Great training for my currently inept soldering skill, if I don't kill anything more by heat.
jaycee you really know this amp! Why is the 2SB649 green? It's on the other channel actually so I'm pulling out at least the trans to check just in case.
And out of the 4 dead MJ2955, 2N3055, 2SD669A, and 2SD649A, Farnell only has MJ2955. And since I'm going to pull out those in the right channel too I might as well replace them all one shot. Since Farnell has free delivery over $30 in my country and the trans cost max ~$3 a piece.
Any replacement for those?
I measured the above 4 mentioned trans in circuit and they're all shorted. And the green one is indeed 2SB649.
Guess I'll be pulling out everything around that 5 dead trans to check for collateral damage. That means over 20.
Great training for my currently inept soldering skill, if I don't kill anything more by heat.jaycee you really know this amp! Why is the 2SB649 green? It's on the other channel actually so I'm pulling out at least the trans to check just in case.
And out of the 4 dead MJ2955, 2N3055, 2SD669A, and 2SD649A, Farnell only has MJ2955. And since I'm going to pull out those in the right channel too I might as well replace them all one shot. Since Farnell has free delivery over $30 in my country and the trans cost max ~$3 a piece.
Any replacement for those?
PNP transistors from that kinda of era were often made in green casings.
I dont really know the amp, just how to read a schematic
For the MJ2955/2N3055 you can also use MJ15003/4, or MJ15024/5.
The 2SB649/2SD669's could be tricky. I think they are discontinued. Fairchild's KSA1220A/KSC2690A look like good replacements. Unfortunately Farnell dont seem to carry them. MJE15032/3 might work but they are larger TO-220 devices and aren't as fast.
Watch out for fake transistors if you shop elsewhere, especially the 649/669's.
I dont really know the amp, just how to read a schematic
For the MJ2955/2N3055 you can also use MJ15003/4, or MJ15024/5.
The 2SB649/2SD669's could be tricky. I think they are discontinued. Fairchild's KSA1220A/KSC2690A look like good replacements. Unfortunately Farnell dont seem to carry them. MJE15032/3 might work but they are larger TO-220 devices and aren't as fast.
Watch out for fake transistors if you shop elsewhere, especially the 649/669's.
BD139 and 140 would work in a pinch if you can't find the proper Jap driver transistors. The ratings are fine, they will be a little slower (but that may not make any difference). The only 649/669's you'll ever find are fakes, you'd have to replace them with something current production to get real ones.
Any TO-3 audio output pair in production (and a lot of obsolete types) will replace and upgrade the 2N3055/2955. Even if you used 3055's, with newer devices the outputs need to be fit with 10 to 33 ohm resistors in series with the base because the original circuit lacked emitter resistors. This warning is well documented in this forum. Older 3055's had high enough internal resistance not to need them, but new ones are "better".
Any TO-3 audio output pair in production (and a lot of obsolete types) will replace and upgrade the 2N3055/2955. Even if you used 3055's, with newer devices the outputs need to be fit with 10 to 33 ohm resistors in series with the base because the original circuit lacked emitter resistors. This warning is well documented in this forum. Older 3055's had high enough internal resistance not to need them, but new ones are "better".
I repaired a NAD receiver with a similar amp circuit. The lack of emitter resistors on the outputs (Q415, Q417) makes this design susceptible to thermal runaway. The outputs runaway and blow, taking the drivers (Q411, Q413) and the bias (Q409) transistors out too.
My fix was 0R33 (5W) in series with the emitters of Q415,Q417 and 6R8 in series with the bases.
Apply the same fix to the channel that's still working, it is just a matter of time until it goes up in smoke too.
My fix was 0R33 (5W) in series with the emitters of Q415,Q417 and 6R8 in series with the bases.
Apply the same fix to the channel that's still working, it is just a matter of time until it goes up in smoke too.
analog_sa said:
Which is why base stoppers are recommended instead of emitter resistors. Either will stop thermal runaway. Using emitter resistors changes the nature of the crossover region. Older 2N3055's from that era had a high intrinsic base resistance. So you need to put it *back* when using modern transistors.
Getting confusing...
Base stopper is the resistor in series with the base? So just that instead of the emitter degenerator is enough?
I've more or less finalized the order list:
MJ15003/4 to replace the 2N3055 pair
BD139/140 to replace the 649 pair
And BD139 to replace... the BD139.
Do I get the plain vanilla models without the . and letters before/after? The MJ15003G has a G behind, environmental-friendly production?
And do I need to heatsink the BD139/140 that is replacing the 649 pair, which are not sinked?
Base stopper is the resistor in series with the base? So just that instead of the emitter degenerator is enough?
I've more or less finalized the order list:
MJ15003/4 to replace the 2N3055 pair
BD139/140 to replace the 649 pair
And BD139 to replace... the BD139.
Do I get the plain vanilla models without the . and letters before/after? The MJ15003G has a G behind, environmental-friendly production?
And do I need to heatsink the BD139/140 that is replacing the 649 pair, which are not sinked?
"Base ballasting" works almost as well as emitter ballasting. And certainly better than nothing. I wouldn't think the drivers would need heat sinking on +/-30V.
G or non-G parts - just depends on what you can get. The G parts are less likely to be fakes. But , even if you end up with fake outputs, what they will be are 2N3055/MJ2955. Which is what you started with.
G or non-G parts - just depends on what you can get. The G parts are less likely to be fakes. But , even if you end up with fake outputs, what they will be are 2N3055/MJ2955. Which is what you started with.
The G parts just mean "Lead free". If you're getting them from Farnell you should be ok. Beware if they are branded Motorola - they are either very old stock, or fake. They should be branded On Semiconductor ("On" in a circle).
I forgot to mention emitter resistors. Yes, put those in. Some 3W 0.33R resistors will do and you should be able to put them between the emitter pins and where the emitter connects on the PCB. You will need to adjust R443 after this to set the bias correctly.
I forgot to mention emitter resistors. Yes, put those in. Some 3W 0.33R resistors will do and you should be able to put them between the emitter pins and where the emitter connects on the PCB. You will need to adjust R443 after this to set the bias correctly.
analog_sa said:
and also arguably (with more supporting evidence) the reason it is prone to thermal runaway. Without it quiescent bias current isn't stable due to Vbe changing with temp.
As pointed out by others, the original 3055/2955 had high enough internal resistance to provide this, the newer versions don't. That's why when I had a repair shop fix my NAD the first time it smoked, it just blew up again. Had to learn amplifier design to understand and fix it myself.
Ok, the parts arrived, and now I'm already stuck -
Those who repaired the amp before, how did you put the emitter resistor between the emitter pin and the PCB? I tried to isolate the pin from the PCB but fail, even without solder it's making contact. Do I need to paint it to insulate the pin?
Those who repaired the amp before, how did you put the emitter resistor between the emitter pin and the PCB? I tried to isolate the pin from the PCB but fail, even without solder it's making contact. Do I need to paint it to insulate the pin?
Thanks for the reply. Pardon me not having a decent camera.
I figured cutting the tracks was the only way. Cut it at 4 places too to minimize capacitance from redundant tracks. Love 1-layered PCB else I wouldn't know how deep to go.
I did thought of a way to install the base stopper without cutting any tracks though, but I'll have to cut unused tracks anyway, stray capacitance. So now I'm pondering whether to use the base stoppers or not since I'm installing emitter resistors.
I figured cutting the tracks was the only way. Cut it at 4 places too to minimize capacitance from redundant tracks. Love 1-layered PCB else I wouldn't know how deep to go.
I did thought of a way to install the base stopper without cutting any tracks though, but I'll have to cut unused tracks anyway, stray capacitance. So now I'm pondering whether to use the base stoppers or not since I'm installing emitter resistors.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.