Hello everyone,
I posted a little while ago regarding my repair of a Rotel 980-BX power amplifier. (Dave, thanks for the pic) I have a little experience in DIY audio, but have never attempted to tackle something like a amplifier repair. I think I found the fault in the amplifier, but would like to check with all of you because I am decidedly a non-expert.
Problem: On amplifier power-up, the fuse attached to the transformer (AC fuse) blows.
Diagnosis:
1) Performed a diode test on all output devices. I don't know an emitter from a base or collector, but none of them were 0V. No shorts.
2) Performed a diode test around each rectifier bridge. Went "around the square" for each. Meter read 0V across two pairs of legs around the rectifier. This is asymmetric compared to the other side. Similarly, the resistance across these legs is 0V.
3) If I measure the resistance across each of the power supply capacitors, only one of them shows a much lower resistance than the others. This capacitor appears connected to two of the broken legs of the rectifier.
Is what I have here just a blown rectifier bridge, or are there other components that I should check? Is the symptom of a blown AC fuse consistent with this?
Thanks for all the help! I'm really looking forward to getting this thing running. 🙂
Doh
I posted a little while ago regarding my repair of a Rotel 980-BX power amplifier. (Dave, thanks for the pic) I have a little experience in DIY audio, but have never attempted to tackle something like a amplifier repair. I think I found the fault in the amplifier, but would like to check with all of you because I am decidedly a non-expert.
Problem: On amplifier power-up, the fuse attached to the transformer (AC fuse) blows.
Diagnosis:
1) Performed a diode test on all output devices. I don't know an emitter from a base or collector, but none of them were 0V. No shorts.
2) Performed a diode test around each rectifier bridge. Went "around the square" for each. Meter read 0V across two pairs of legs around the rectifier. This is asymmetric compared to the other side. Similarly, the resistance across these legs is 0V.
3) If I measure the resistance across each of the power supply capacitors, only one of them shows a much lower resistance than the others. This capacitor appears connected to two of the broken legs of the rectifier.
Is what I have here just a blown rectifier bridge, or are there other components that I should check? Is the symptom of a blown AC fuse consistent with this?
Thanks for all the help! I'm really looking forward to getting this thing running. 🙂
Doh
Hi Doh,
I had the same problem with my amp. The output transistors were shot. I replaced them and resoldered the entire PCB. Thanks to mrfeedback's suggestions the amp works fine now.
I think you should remove the output transistors and check them with the transistor Hfe test function with a digital DMM.
Cheers,
Vivek
I had the same problem with my amp. The output transistors were shot. I replaced them and resoldered the entire PCB. Thanks to mrfeedback's suggestions the amp works fine now.
I think you should remove the output transistors and check them with the transistor Hfe test function with a digital DMM.
Cheers,
Vivek
Hi, Doh
I suspect the cap or the bridge
Unsolder the suspicious parts and measure them out of the circuit.
One or both of them will be shorted.
/Hugo 🙂
I suspect the cap or the bridge
Unsolder the suspicious parts and measure them out of the circuit.
One or both of them will be shorted.
/Hugo 🙂
IME bridges and psu caps rarely go wrong, at least not as to blow a fuse. Shorting power transistors are a much more common cause of the symptoms you are seeing. Since the rectifier, psu caps and the amp circuit are all in parallel it will be difficult to draw a conclusion.
If practical, I recommend disconnecting the amp boards from the psu. Replace the fuse and power up. Measure the psu voltages. This will exhonorate the psu if it works ok. Then connect one amp board at a time, etc, etc.
If practical, I recommend disconnecting the amp boards from the psu. Replace the fuse and power up. Measure the psu voltages. This will exhonorate the psu if it works ok. Then connect one amp board at a time, etc, etc.
Some More Iideas
Hi, here are some more ideas. Hope they help.
Those shorts you measured on the rectifier diode sound like the resistances of the secondary windings of the power transformer. As mentioned, bridge rectifiers rarely go bad. Out of 50-100 amp repairs I have done, I may have seen one bridge rectifier go bad.
I recommend using a short lamp in series with the power line. That way, you can plug the amp in and limit the current drawn to the amount the filament of the light bulb can pass at full voltage drop. Try a 60 watt bulb to start out. You may need to go to a 100 watt bulb or probably at most 2 60 watt ones in parallel. Then you can isolate various parts of the circuit by working your way back from the output transistors.
Beware that sometimes blown outputs can indicate blown parts earlier in the signal path. It is a good idea to diode test a little "upstream", checking the output drivers, bias circuit, or more.
Hi, here are some more ideas. Hope they help.
Those shorts you measured on the rectifier diode sound like the resistances of the secondary windings of the power transformer. As mentioned, bridge rectifiers rarely go bad. Out of 50-100 amp repairs I have done, I may have seen one bridge rectifier go bad.
I recommend using a short lamp in series with the power line. That way, you can plug the amp in and limit the current drawn to the amount the filament of the light bulb can pass at full voltage drop. Try a 60 watt bulb to start out. You may need to go to a 100 watt bulb or probably at most 2 60 watt ones in parallel. Then you can isolate various parts of the circuit by working your way back from the output transistors.
Beware that sometimes blown outputs can indicate blown parts earlier in the signal path. It is a good idea to diode test a little "upstream", checking the output drivers, bias circuit, or more.
Thanks for all the advice!
I was a bit worried that it might be the output transistors that have shorted, but when the diode test failed to ID a faulty transistor (perhaps I'm doing it wrong?) I got my hopes up and started measuring other parts of the amplifier. The rectifier and the cap are the only two parts that measure asymmetrically with the other side of the amp.
I really like the idea of disconnecting the amp and power supply to see what happens. Unfortunately, the amp and power supply are built on the same PCB. 🙁
From what I gather about the diode test, it's not 100% diagnostic of a blow output device when applied directly to the transistors.
I am a little baffled by how to use a short lamp to isolate different parts of the amplifier. In fact, I'm not really sure what this device looks like and how it would interact with the amp to let me isolate parts of it. It does, however sound like something that would be very useful.
Another post I read stated that if one of the output transistors on an amplifier dies, they may all be reaching the end of their useful life. As such, I think I will go on the presumption that one of the output transistors is actually blown, and all need to be replaced. I have specified the following substitutions:
2SC3856 --> MJL3281
2SA1492 --> MJL1302
The parameters for voltage and current ratings are virtually identical, and another post indicated that they can be used as drop-in replacements. The only difference seems to be the Hfe and Ft on the replacements is better?
The replacements appear to be well favored by those on the forum, and from the instructions on: http://www1.jaycar.com.au/images_uploaded/transubs.pdf it appears that they are suitable replacements, as well as being readily as samples 🙂 I was able to find replacements for every other transistor in the amplifier for less than $20 and they are on the way as well.
I'll pick up a new set of bridge rectifiers as well, since I need to order more fuses for the amp. I figure if I'm going to desolder it, I might as well replace it. I believe that I can specify a higher specification rectifier without harm to the amp, but cannot specify one with lower ratings. The highest rated bridge rectifier I have found in the appropriate size package is: 10A, PRV 1000, Ifm 200. Do you think this will be adquate? The amplifier is class AB 120WPC, as specified by rotel.
For now, I would like to think that the source of the problem is not the capacitor. (capacitor failure usually occurs with the caps exploding, no?) as a 10,000uF 63V capacitor may be quite expensive. I will perform the same diode tests after I have finished repairs and replacements on the rest of the amp and hope that it is not faulty since the amp has 4 of them, and I would prefer to have all of the same type in the amplifier at once.
Thanks again for all the help!
I was a bit worried that it might be the output transistors that have shorted, but when the diode test failed to ID a faulty transistor (perhaps I'm doing it wrong?) I got my hopes up and started measuring other parts of the amplifier. The rectifier and the cap are the only two parts that measure asymmetrically with the other side of the amp.
I really like the idea of disconnecting the amp and power supply to see what happens. Unfortunately, the amp and power supply are built on the same PCB. 🙁
From what I gather about the diode test, it's not 100% diagnostic of a blow output device when applied directly to the transistors.
I am a little baffled by how to use a short lamp to isolate different parts of the amplifier. In fact, I'm not really sure what this device looks like and how it would interact with the amp to let me isolate parts of it. It does, however sound like something that would be very useful.
Another post I read stated that if one of the output transistors on an amplifier dies, they may all be reaching the end of their useful life. As such, I think I will go on the presumption that one of the output transistors is actually blown, and all need to be replaced. I have specified the following substitutions:
2SC3856 --> MJL3281
2SA1492 --> MJL1302
The parameters for voltage and current ratings are virtually identical, and another post indicated that they can be used as drop-in replacements. The only difference seems to be the Hfe and Ft on the replacements is better?
The replacements appear to be well favored by those on the forum, and from the instructions on: http://www1.jaycar.com.au/images_uploaded/transubs.pdf it appears that they are suitable replacements, as well as being readily as samples 🙂 I was able to find replacements for every other transistor in the amplifier for less than $20 and they are on the way as well.
I'll pick up a new set of bridge rectifiers as well, since I need to order more fuses for the amp. I figure if I'm going to desolder it, I might as well replace it. I believe that I can specify a higher specification rectifier without harm to the amp, but cannot specify one with lower ratings. The highest rated bridge rectifier I have found in the appropriate size package is: 10A, PRV 1000, Ifm 200. Do you think this will be adquate? The amplifier is class AB 120WPC, as specified by rotel.
For now, I would like to think that the source of the problem is not the capacitor. (capacitor failure usually occurs with the caps exploding, no?) as a 10,000uF 63V capacitor may be quite expensive. I will perform the same diode tests after I have finished repairs and replacements on the rest of the amp and hope that it is not faulty since the amp has 4 of them, and I would prefer to have all of the same type in the amplifier at once.
Thanks again for all the help!
Ok,
I desoldered the rectifier bridge and measured it out of the circuit. It appears to have a short across one pair of legs. The capacitor no longer measures a short, so I am hoping it is ok.
Is it possible that components upstream of the rectifier are shorted out and that it is simply the victim of cascading failure?
To put it another way, is plausible that the rectifier bridge is the only part that has blown? The package appears to be a leaded D-72 type device. All the parts I could locate of this package type max out with ratings of 1000V 10A or so, and the part from international rectifier maxes out at 1000V 8A. They certainly don't appear to be as durable as the 400V 35A units that we see more commonly in DIY.
My gestalt is that an 8A rectifier would be operating within its limits in a 120W class AB amplifier, but that it does experience quite a bit of strain as it charges 2 x 10,000uF capacitors as it powers up. Does this sound about right?
Thanks again!
I desoldered the rectifier bridge and measured it out of the circuit. It appears to have a short across one pair of legs. The capacitor no longer measures a short, so I am hoping it is ok.
Is it possible that components upstream of the rectifier are shorted out and that it is simply the victim of cascading failure?
To put it another way, is plausible that the rectifier bridge is the only part that has blown? The package appears to be a leaded D-72 type device. All the parts I could locate of this package type max out with ratings of 1000V 10A or so, and the part from international rectifier maxes out at 1000V 8A. They certainly don't appear to be as durable as the 400V 35A units that we see more commonly in DIY.
My gestalt is that an 8A rectifier would be operating within its limits in a 120W class AB amplifier, but that it does experience quite a bit of strain as it charges 2 x 10,000uF capacitors as it powers up. Does this sound about right?
Thanks again!
Glad you found it.Doh said:
You have a good chance that the only problem will be the rectifier.
What exactly is in there now?
A 1000V/10A will certainly do the trick.
/Hugo🙂
The part appears to be unmarked, and I don't have a service manual from Rotel. I ordered a pair of the the international rectifier part (1000V 8A) so I can replace both rectifiers. I guess I'll also have a bag full of parts in case anything else in the amp blows in the future. 
I'm really excited to hear how this thing sounds. The overall layout and parts selection in the amp looks pretty impressive to me. Big beefy power supply, solid binding posts, nice quality resistors (Look like Beyschlags), Blackgate capacitors in the amplifier itself, and 20,000uF per rail of Rotel's own branded electrolytics in the power supply. Of course, not nearly as impressive as some of the stuff I've seen built on this forum (especially Peter Daniel's gorgeous craftsmanship)
Thanks for all the help! I'm sooo happy that this thing is going to be up and working.
Doh
I'm really excited to hear how this thing sounds. The overall layout and parts selection in the amp looks pretty impressive to me. Big beefy power supply, solid binding posts, nice quality resistors (Look like Beyschlags), Blackgate capacitors in the amplifier itself, and 20,000uF per rail of Rotel's own branded electrolytics in the power supply. Of course, not nearly as impressive as some of the stuff I've seen built on this forum (especially Peter Daniel's gorgeous craftsmanship)
Thanks for all the help! I'm sooo happy that this thing is going to be up and working.
Doh
It lives!!!!
I replaced the rectifier bridges in the amplifier last night and the amplifier lives!!!
I also got carried away and rebuilt the output stage as well. The amplifier case has so many screws that I never want to have to open it again. (yah right) 😀 😀 😀 😀 😀
I'm letting it burn in now, and will have to give it a good listen when I get home. Thanks to everyone for all the help!
I replaced the rectifier bridges in the amplifier last night and the amplifier lives!!!
I also got carried away and rebuilt the output stage as well. The amplifier case has so many screws that I never want to have to open it again. (yah right) 😀 😀 😀 😀 😀
I'm letting it burn in now, and will have to give it a good listen when I get home. Thanks to everyone for all the help!
Doh said:
I was just about to tell you to have a look at all the solderings before closing the box 😀 😀
Just kidding...glad it works again.
What did you rebuild in the output?
/Hugo
This is rather embarassing, but I rather indiscriminately replaced every transistor on which I could read the part number with a new one from the same manufacturer. This basically comprised every transistor in the chassis with the exception of one pair per channel of potted components, which I didn't feel like digging out. (Hey, it is over 10 years old)
Overkill to be sure, but like I said, there are a _lot_ of screws in the chassis (perhaps two dozen to get the cover and bottom panel off) and I already had the parts on hand.
All the transistors were made by Sanken, with the exception of one per channel by NEC, and they were all available from bdent for less than $20, so the cost was not an object
The only substitutions I made were for the output devices:
2SC3856 --> MJL3281
2SA1492 --> MJL1302
It turns out that the old rectifier bridges _did_ have a part number stamped on them. A little google revealed that they were 400V 8A parts. 😀
Curious, though, the DC-offset of the amp measured 2mv in one channel and 60mv in the other before I swapped all the devices, immediately after the rectifier bridge replacement. After all of the transistors were swapped, I got the same readings. I would presume that DC-offset is caused by mismatching of transistors, but it's a real head-scratcher why I would get the exact same results with a whole new transistor set. I'm not all that worried about it since its a tolerable value, but it still bothers me how this could be so. Coincidence? Any ideas?
Overkill to be sure, but like I said, there are a _lot_ of screws in the chassis (perhaps two dozen to get the cover and bottom panel off) and I already had the parts on hand.
All the transistors were made by Sanken, with the exception of one per channel by NEC, and they were all available from bdent for less than $20, so the cost was not an object
The only substitutions I made were for the output devices:
2SC3856 --> MJL3281
2SA1492 --> MJL1302
It turns out that the old rectifier bridges _did_ have a part number stamped on them. A little google revealed that they were 400V 8A parts. 😀
Curious, though, the DC-offset of the amp measured 2mv in one channel and 60mv in the other before I swapped all the devices, immediately after the rectifier bridge replacement. After all of the transistors were swapped, I got the same readings. I would presume that DC-offset is caused by mismatching of transistors, but it's a real head-scratcher why I would get the exact same results with a whole new transistor set. I'm not all that worried about it since its a tolerable value, but it still bothers me how this could be so. Coincidence? Any ideas?
Are there any trim pots on the PCB?Doh said:
If yes, you could have 1 or 2 per channel.
The first one could be a DC adjustment, the second one the bias setting.
First thing to do ...oh, no...I forgot you closed the amp already...anyway, first thing to do is measure the bias across the emitter resistors of the output transistors.
If a second pot is there it could be the DC adjustment.
Unfortunately, without technical data it's difficult to tell how to set the bias. Maybe someone has a schematic over here?
/Hugo
Ahh yes, the trim pots. There is one trim pot per channel, but turning it appears to have no effect on the DC offset at the speaker terminals. I presumed that it is used to set the bias when the amp was open and left them pretty much alone since, after ten years, turning them produces a not-so-pleasant scratching sound.
As far as schematics, I posted a while ago asking for one and got a response with what I think is the schematic for another Rotel amp. The link is here...
http://www.diyaudio.com/forums/showthread.php?s=&threadid=16697
I believe it is a lower power version, with my amplifier having different devices and an additional output pair, but sharing the same basic topology. At very least, looking closely at the schematic, I see a great deal of resemblance in the way the devices are layed out from the transistors to, of course, a single 2.2k trim pot per channel.
I'm by no means proficient at reading schematics, but such a pot would appear to set the quiescent current, no?
This, unfortunately, puts us no closer to adjusting the DC offset. Any thoughts?
As far as schematics, I posted a while ago asking for one and got a response with what I think is the schematic for another Rotel amp. The link is here...
http://www.diyaudio.com/forums/showthread.php?s=&threadid=16697
I believe it is a lower power version, with my amplifier having different devices and an additional output pair, but sharing the same basic topology. At very least, looking closely at the schematic, I see a great deal of resemblance in the way the devices are layed out from the transistors to, of course, a single 2.2k trim pot per channel.
I'm by no means proficient at reading schematics, but such a pot would appear to set the quiescent current, no?
This, unfortunately, puts us no closer to adjusting the DC offset. Any thoughts?
Hey Man,
Like they said, measure the voltage across the emitter resistors ie .22,.33, or.47 ohm 5 or 10W, and see what the voltage is. It will be in the millivolt reigon. I'd set the pot so there is around 20-30mV across the highest reading one, assuming multiple output devices.... I built a 500W amp that requires 17mV across each. If the transistors aren't matched very well, use the reading that is the highest and set it down to 20 mV or so and see. If the pot is scratchey, replace it or you will have bocoup problems and instability...... NOT GOOD...... Why are you adjusting it with speakers attached, anyway? Are you crazy or independantly wealthy? ;-) If you are not able to adjust the bias current, you have other problems. 60mv on the output is a little high but I have heard of worse. Usually, there are resistors attached to the differential pair on the input, that should be 100-200 ohms. They may be on the collectors. Check that they are the same value. If one is different by a lot, you may see that drift in voltage. It means that the error amplifier function of the diff. pair can't work properly due to the inbalance. This will also cause problems too.....
Good luck!
Chris
Like they said, measure the voltage across the emitter resistors ie .22,.33, or.47 ohm 5 or 10W, and see what the voltage is. It will be in the millivolt reigon. I'd set the pot so there is around 20-30mV across the highest reading one, assuming multiple output devices.... I built a 500W amp that requires 17mV across each. If the transistors aren't matched very well, use the reading that is the highest and set it down to 20 mV or so and see. If the pot is scratchey, replace it or you will have bocoup problems and instability...... NOT GOOD...... Why are you adjusting it with speakers attached, anyway? Are you crazy or independantly wealthy? ;-) If you are not able to adjust the bias current, you have other problems. 60mv on the output is a little high but I have heard of worse. Usually, there are resistors attached to the differential pair on the input, that should be 100-200 ohms. They may be on the collectors. Check that they are the same value. If one is different by a lot, you may see that drift in voltage. It means that the error amplifier function of the diff. pair can't work properly due to the inbalance. This will also cause problems too.....
Good luck!
Chris
Ok, I measured the voltage across the emitter resistors, and it appears that there is a great deal more variation on the side with the DC offset problem. In fact, I have a great deal more trouble getting consistent values for this side of the amp. Two curious observations:
1) The voltage across the emitter resistors for the 3281 devices almost uniformly reads from low to high from the left to the right. Specifically, the values run as follows: 21.6, 22,8, 26.2. This appears to persist despite the replacement of the left-most device.
2) The left side of the amp is layed out as the mirror image of the right side of the amplifier, with one exception. There appears to be an extra resistor in the left side of the amp. (The side with the offset) This resistor, R671 simply does not have a counterpart on the other side of the amp though the other side of the amp has unmarked through-holes where the corresponding resistor should be. Is there any reason that such an asymmetry would be built into an amplifier?
I do not believe that the offset voltage is a result of mismatched output devices. This offset voltage was present before I rebuilt the output stage, and after I replaced the one output device. The voltage differences between the emitter resistors persists as well. Is it possible that the measured DC offset is a result of this extra resistor?
I suppose the next step would be to remove this resistor and see what happens. I just wish I had noticed the difference before I put the bottom panel back on. 😛 What do you think?
1) The voltage across the emitter resistors for the 3281 devices almost uniformly reads from low to high from the left to the right. Specifically, the values run as follows: 21.6, 22,8, 26.2. This appears to persist despite the replacement of the left-most device.
2) The left side of the amp is layed out as the mirror image of the right side of the amplifier, with one exception. There appears to be an extra resistor in the left side of the amp. (The side with the offset) This resistor, R671 simply does not have a counterpart on the other side of the amp though the other side of the amp has unmarked through-holes where the corresponding resistor should be. Is there any reason that such an asymmetry would be built into an amplifier?
I do not believe that the offset voltage is a result of mismatched output devices. This offset voltage was present before I rebuilt the output stage, and after I replaced the one output device. The voltage differences between the emitter resistors persists as well. Is it possible that the measured DC offset is a result of this extra resistor?
I suppose the next step would be to remove this resistor and see what happens. I just wish I had noticed the difference before I put the bottom panel back on. 😛 What do you think?
Hmm...looks like the PCB does differ locally at R671. It and its neighbor attach to the base of one transistor, while the corresponding resistor on the other side attaches to the emitter of another transistor.
Don't take R617 out.
There's probably a good reason for it being there.
Not to confuse you or make things to complicated but maybe replacing some small electrolytic caps might solve your problem.
Looking at Dave's schematic I would replace the input cap (C602)
and C606 (guess) 100µF/6.3V. Possible other electrolytes around the driver transistors could best be replaced.
/Hugo 🙂
There's probably a good reason for it being there.
Not to confuse you or make things to complicated but maybe replacing some small electrolytic caps might solve your problem.
Looking at Dave's schematic I would replace the input cap (C602)
and C606 (guess) 100µF/6.3V. Possible other electrolytes around the driver transistors could best be replaced.
/Hugo 🙂
The capacitors, huh?
These are the values and types of the capacitors found in the circuit (per channel):
.1 uF WIMA green polypropylene x1
220pf styroflex x3 (well, it says 220 on it)
330pf styroflex x1 (ditto)
50V 10uF electrolytic x1
50V 4.7uF electrolytic x1
15V 100uf electrolytic x1
Are you suggesting that the offset voltage could be the result of dried out electrolytics? Also, I've read that styroflex capacitors can be rather delicate. Are they candidates for being the culprits here as well?
If the styroflex capacitors are suspect, I think they'll need to be replaced with comparable value polypropylene units. I'm really not looking forward to soldering styroflex types. 😛
These are the values and types of the capacitors found in the circuit (per channel):
.1 uF WIMA green polypropylene x1
220pf styroflex x3 (well, it says 220 on it)
330pf styroflex x1 (ditto)
50V 10uF electrolytic x1
50V 4.7uF electrolytic x1
15V 100uf electrolytic x1
Are you suggesting that the offset voltage could be the result of dried out electrolytics? Also, I've read that styroflex capacitors can be rather delicate. Are they candidates for being the culprits here as well?
If the styroflex capacitors are suspect, I think they'll need to be replaced with comparable value polypropylene units. I'm really not looking forward to soldering styroflex types. 😛
- Status
- Not open for further replies.