Hi.
I'm new to high power amps, but I've been repairing electronics for a long time, mostly replacing burnt output trt's, bad capacitors, etc.
Last month, a cousin brought a 1kw per channel rack, one channel faulty.
After opening the rack I saw about ten resistors burnt and when checking the output transistors (MJL21196) they were shorted, so I replaced them and all the resistors, some MJE350 and MJE340, a BC238C. I've seen these amps in different cases and I suspect they're made almost identical in different brands. Anyway, each channel has 12 mjl21196, 6 on each side of a heatsink, and I've seen logic gates IC's and op-amps as in a crown schematic available on their website as "discontinued". Anyway, the issue is that since I've never touched a PA like this I have no idea where to start, two resistors keep burning, and I really have no clue where to start, I have the equipment (oscilloscope, signal generator, etc), but since I can't found the schematic for this amp, It's a bit frustrating. The fault is not on the power supply nor in the control circuit, I plugged it on the other channel and still the same. I guess you will suggest I study first the basis of the amp, but I have no idea if it's a class H or a class G. I've read that these amps are not worth repairing, but since the output trt's had been replaced, It seems worth trying. I've purchased some op-amps in case they are the problem, more trt's (small signal) and the optocouplers are difficult to find at reasonable prices.
I can read and study the topology, but as I said I have no clue where to start, I have the sloane book and I think It could be a start, but well, you guys know better than me these amps and maybe could give me a few (or a lot) of tips on what to read, and If I'm lacking some equipment to properly repair (I hope) this amp.
Here are some pics to help:
Rack Top view:
![a>](https://www.diyaudio.com/forums/<a href="<a href=http://www.flickr.com/photos/88082778@N07/9626411373/" target=_blank>http://www.flickr.com/photos/88082778@N07/9626411373/"</a> title="SP2000_2 por TallerSeverino, en Flickr"><img src="<a href=http://farm8.staticflickr.com/7312/9626411373_044bfe4e3a.jpg" target=_blank>http://farm8.staticflickr.com/7312/9626411373_044bfe4e3a.jpg"</a> width="500" height="337" alt="SP2000_2"></a>)
Power amp side view (1):
Power amp side view (2):
Power amp top view:
Power amp input section and burnt components (there is a BC238C on the bottom side of the pcb with thermal paste to the heatsink)
Kind regards.
I'm new to high power amps, but I've been repairing electronics for a long time, mostly replacing burnt output trt's, bad capacitors, etc.
Last month, a cousin brought a 1kw per channel rack, one channel faulty.
After opening the rack I saw about ten resistors burnt and when checking the output transistors (MJL21196) they were shorted, so I replaced them and all the resistors, some MJE350 and MJE340, a BC238C. I've seen these amps in different cases and I suspect they're made almost identical in different brands. Anyway, each channel has 12 mjl21196, 6 on each side of a heatsink, and I've seen logic gates IC's and op-amps as in a crown schematic available on their website as "discontinued". Anyway, the issue is that since I've never touched a PA like this I have no idea where to start, two resistors keep burning, and I really have no clue where to start, I have the equipment (oscilloscope, signal generator, etc), but since I can't found the schematic for this amp, It's a bit frustrating. The fault is not on the power supply nor in the control circuit, I plugged it on the other channel and still the same. I guess you will suggest I study first the basis of the amp, but I have no idea if it's a class H or a class G. I've read that these amps are not worth repairing, but since the output trt's had been replaced, It seems worth trying. I've purchased some op-amps in case they are the problem, more trt's (small signal) and the optocouplers are difficult to find at reasonable prices.
I can read and study the topology, but as I said I have no clue where to start, I have the sloane book and I think It could be a start, but well, you guys know better than me these amps and maybe could give me a few (or a lot) of tips on what to read, and If I'm lacking some equipment to properly repair (I hope) this amp.
Here are some pics to help:
Rack Top view:
An externally hosted image should be here but it was not working when we last tested it.
Power amp side view (1):
An externally hosted image should be here but it was not working when we last tested it.
Power amp side view (2):
An externally hosted image should be here but it was not working when we last tested it.
Power amp top view:
An externally hosted image should be here but it was not working when we last tested it.
Power amp input section and burnt components (there is a BC238C on the bottom side of the pcb with thermal paste to the heatsink)
An externally hosted image should be here but it was not working when we last tested it.
Kind regards.
Like this,
To add a photo, files or non standard files.
First click "go advanced" in the box below the "quick reply" message box. Doesn't matter if you decide half way through a message to do that, it carries it forward.
Then click "Manage attachements". Maximise the new Window so that you can see all the text.
Click browse in the first box at the top and find your picture. Repeat for any more pictures.
Click upload... a message appears "uploading"
When complete the files will show as being attached. Now click the small text that says "close this window"
The pictures should now be attached and when you submit your post they will appear.
Make sure your pics aren't too big, a couple of 100k is plenty, and many members object when they are massive and it alters the margins
It tells you in the attachments window what max sizes are allowed.
If you want to attach a file that has a non standard format for example excel, circuit simulation etc then try putting the files in a zipped folder and attaching that.
To add a photo, files or non standard files.
First click "go advanced" in the box below the "quick reply" message box. Doesn't matter if you decide half way through a message to do that, it carries it forward.
Then click "Manage attachements". Maximise the new Window so that you can see all the text.
Click browse in the first box at the top and find your picture. Repeat for any more pictures.
Click upload... a message appears "uploading"
When complete the files will show as being attached. Now click the small text that says "close this window"
The pictures should now be attached and when you submit your post they will appear.
Make sure your pics aren't too big, a couple of 100k is plenty, and many members object when they are massive and it alters the margins
It tells you in the attachments window what max sizes are allowed.
If you want to attach a file that has a non standard format for example excel, circuit simulation etc then try putting the files in a zipped folder and attaching that.
Does it have multiple rail voltages or only one? Is it +/-X volts, or +/-X and 2X volts? I don't *see* anything on the boards that can be immediately identified as a rail switch or commutating diodes. If it's only a single rail (per side) it's just class B. I don't see driver transistors either, unless it's using one of the TO264 flatpacks as a driver.
Connect a 60w lamp in series with your 110v/240v supply, and disconnect the ok module.
Try running it up and see if ok at idle.
If not,the lamp should help prevent burning/blowups while you determine the fault.
Dan.
Try running it up and see if ok at idle.
If not,the lamp should help prevent burning/blowups while you determine the fault.
Dan.
Last edited:
Well, the manufacturer is "space equipment" and the model is SE-2000. I've seen them often in Spain, the pcb says It's made in the EU.
Did you mean this?
Did you mean this?
So it's just +/-110V, with no +/-55V mid-rails? Running class B at those voltages is ASKING for trouble unless the outputs are in series-parallel. Even then you'd want a few more.
I've checked the power supply inputs on the module, there are two wires for each +/-110V, but checking continuity they're the same, I'm going to check the power supply though, It seems that this feature is to avoid thicker wires on the power supply rails.
I'm a little confused, class B is push-pull, and all of the output transistors are npn, correct me if I'm wrong. It just doesn't add up If this a 1kw PA.
Edit: Must be a quasi-complementary design, the large inductor before the relay to the speakers now makes sense.
I'm a little confused, class B is push-pull, and all of the output transistors are npn, correct me if I'm wrong. It just doesn't add up If this a 1kw PA.
Edit: Must be a quasi-complementary design, the large inductor before the relay to the speakers now makes sense.
Last edited:
Probably 1kw into 2 ohms. Supply voltage is a little high for that - all you need for that is +/-95 and it would be more reliable. That's probably why the outputs keep burning out. The old QSC 1310 is 1k into 2R, and it runs on +/-95 and has 8 TO-3 outputs per bank. That's what you'd expect to see unless it's a class H - but that needs 4 rails which you don't seem to have. I wouldn't go blasting it or running subs on less than 8 ohms. For lighter duty maybe 4 ohms - but for that you don't need such a big amp.
All NPN designs don't make much sense for these things. The price on TO-264 NPNs and PNPs are the same. And I can't see the production volumes being so high that they'd get the massive breaks on say, a million pieces.
All NPN designs don't make much sense for these things. The price on TO-264 NPNs and PNPs are the same. And I can't see the production volumes being so high that they'd get the massive breaks on say, a million pieces.
The other module works with this supply voltages, so I guess It must be my voltmeter that needs adjusting and/or replacement.
Anyway, the first two pairs of mjl21196 are not connected in parallel. As seen on the pictures there's 12 of them, the first row has 5 of them in paralel, same on the lower row, the 2 on the left must be the drivers, anyway the components that keep burning are two resistors which are associated with a mje340, a bc238C and a mje350. The BC is under the pcb and there's something wrong there, why they chose to put it in such a strange place I cannot understand, the BC burns and then the resistors, I will solder a new one on the top but It must run hot when it was put in contact with the heat sink. I think a better option would be a better trt in to-220 or similar and attach a small heatsink. The resistors burnt at the point that a track underneath was pulverized, I soldered a wire but that wasn't the problem. I'm running out of resistors to check.
I will keep posting when something besides burnt resistors happens.
I really don't know if they're really 1kw per channel, I'm missing the manual and on the back, where you expect to find this information there's nothing about that. The manufacturer changed since 2001 and now It's called mcm audio, located in Spain, It was the first thing I did, contacted them a month ago requesting information. No answer. This morning I tried to open their website and It's been hacked, so it's a dead end on this. Any suggestion on getting a replacement for the bc238c ?
Anyway, the first two pairs of mjl21196 are not connected in parallel. As seen on the pictures there's 12 of them, the first row has 5 of them in paralel, same on the lower row, the 2 on the left must be the drivers, anyway the components that keep burning are two resistors which are associated with a mje340, a bc238C and a mje350. The BC is under the pcb and there's something wrong there, why they chose to put it in such a strange place I cannot understand, the BC burns and then the resistors, I will solder a new one on the top but It must run hot when it was put in contact with the heat sink. I think a better option would be a better trt in to-220 or similar and attach a small heatsink. The resistors burnt at the point that a track underneath was pulverized, I soldered a wire but that wasn't the problem. I'm running out of resistors to check.
I will keep posting when something besides burnt resistors happens.
I really don't know if they're really 1kw per channel, I'm missing the manual and on the back, where you expect to find this information there's nothing about that. The manufacturer changed since 2001 and now It's called mcm audio, located in Spain, It was the first thing I did, contacted them a month ago requesting information. No answer. This morning I tried to open their website and It's been hacked, so it's a dead end on this. Any suggestion on getting a replacement for the bc238c ?
It must be the bias regulator, when needs to sense the temperature of the outputs. When the outputs/drivers/predrivers fail, it can take this out. When I have a failed output stage I always replace this device (and its associated resistors) because it's so critical and easily destroyed. Just be mindful of the pin-out if you replace it with something different. The BC's are supposed to be CBE, but I've seen them both CBE and EBC. It's NPN, so collector is positive (it will connect to the base of an MJE340 predriver - with maybe a resistor in between).
Options...
Yesterday I checked some more components, There was a quad op amp lm354 shorted, the tl074 is fine as are the dual op-amps. I forgot to say there's also a temp sensor attached on the bottom of the pcb, but it's located more to the right and as for now there's no time to check this since the resistors burn before even the relay to the speakers connect.
I was checking one of the optocouplers and seems faulty. I have to talk to my cousin. I'm thinking that besides the output trt's that have been replaced It would be a good option to completely replace all the components, I have most of them anyway, check the PCB properly for burnt tracks, and most important, make an schematic that could allow me to understand this amp. Anyway, the suspect now is a vlc5c3 optocoupler (the led side).
About the schematic I remember that once I did this using the PCB, locating all components in altium and then back to an schematics, just the opposite of the workflow, It was for a 80's computer interface and worked fine, lots of tracks (digital), took me a while but, well, after all we are in diy electronics because we like it.
I also suspect the vias in the pcb not making appropiate contacts so I will start a "bom" and then unsolder and make a schematic. Thanks for your tips and I will keep working.
Have a nice weekend!
Yesterday I checked some more components, There was a quad op amp lm354 shorted, the tl074 is fine as are the dual op-amps. I forgot to say there's also a temp sensor attached on the bottom of the pcb, but it's located more to the right and as for now there's no time to check this since the resistors burn before even the relay to the speakers connect.
I was checking one of the optocouplers and seems faulty. I have to talk to my cousin. I'm thinking that besides the output trt's that have been replaced It would be a good option to completely replace all the components, I have most of them anyway, check the PCB properly for burnt tracks, and most important, make an schematic that could allow me to understand this amp. Anyway, the suspect now is a vlc5c3 optocoupler (the led side).
About the schematic I remember that once I did this using the PCB, locating all components in altium and then back to an schematics, just the opposite of the workflow, It was for a 80's computer interface and worked fine, lots of tracks (digital), took me a while but, well, after all we are in diy electronics because we like it.
I also suspect the vias in the pcb not making appropiate contacts so I will start a "bom" and then unsolder and make a schematic. Thanks for your tips and I will keep working.
Have a nice weekend!
One method that proves useful in certain circumstances like this one, and taking advantage that you have one perfect amp channel for refference is to make cold measurements.
Replace all known bad components on the bad board. Then do cold measurements in both the good and the bad boards. Do all type of measurements with your DMM, in resistor, diode, capacitance settings. Compare back to back.
You will most likely find the issue without applying any voltage.
Replace all known bad components on the bad board. Then do cold measurements in both the good and the bad boards. Do all type of measurements with your DMM, in resistor, diode, capacitance settings. Compare back to back.
You will most likely find the issue without applying any voltage.
Yes, that's a strange method because all the components are soldered, but I did that in diode mode and that's how I found the LM quad opamp shorted. I kept checking on ohms mode the power rails connectors +/- vs gnd, It was 12K on the + side and 20k on the - side, I kept reading 6K on the faulty one, replaced the quad amp, and a mpsa92 with a BF223 (or something like that, I cannot remember now), the suspect optocoupler keeps reading (on diode) around 200k vs the working one 1,2M.
Once the power rails connectors did change to 12k and 20k I was fool enough to think that the problem was somehow solved and that the reading on the optocoupler was due to a resistor wich wasn't the exact replacement, 1,8K and I put a 2K one.
Once again I was fool enough and tired, connected the module to the rack connectors and this time the resistors burnt faster than in any previous tests I made.
I've finally located the manufacturer of the amp, CE Studios, in Spain, It was written on the back of the amp, It must be some kind of oem amp because they are "ram audio" and have their own catalogue of power amps, I contacted them, let's see where this takes me in order to maybe spare parts, schematics, properly biasing, etc.
I started to unsolder the output trt's to take out the pcb and start the schematics but stopped in the middle, thinking that the cold measurements where working. But, in this is a big If, I have no idea on how to bias the amp I will keep working blind, and the small potentiomenters are suspects too, they're small, plastic case and, well, I don't know.
Should I continue to make the schematic? Or keep checking for "suspects" with cold measurements?
I don't find it a realistic possiblity to pickup an squematic from such a complex double sided board like this one, unless you have a year to put into it.
First off, did your cousin bring the amplifier for repair right after it failed or he had a stop in some other service workshop first?
Second, I recommend you stick to perform cold measurements by now. Remove all screws securing the transistors to the heatsink and remove the board. There might be blown traces on the back side.
Then, try to find sense in all measurements you do. As e.g., you say the reading at the optocoupler in the good board is 1.2M and 200K in the bad board, but you suspect that difference could be because you substituted a 1.8K resistor for a 2K. We have to agree this is far from making any sense.
Do back to back checks between the good and the bad boards. First of all, that all power transistors are in their correct positions. Someone might have removed any of them and re soldered in the wrong location 😱
Check out all dil 8/14 devices to be inserted in their proper locations and in the correct orientation.
Start making cold measurements in both boards in an area far away from the damaged zone. Must be coincident up to normal component tolerance. (10% difference would be acceptable, but under normal conditions and if all components would be fine there should be almost zero difference between both boards.
Approach with you measurements towards "ground zero"🙂 ANY appreciable difference in measurement between the good and the bad boards should be tracked to the faulty component. Do not apply any power until ALL measurements match or YOU WILL continue blowing up parts and frustrating yourself.
Be methodical and as I recommended work from the outside towards the damaged area.
Take notes for your refference.
First off, did your cousin bring the amplifier for repair right after it failed or he had a stop in some other service workshop first?
Second, I recommend you stick to perform cold measurements by now. Remove all screws securing the transistors to the heatsink and remove the board. There might be blown traces on the back side.
Then, try to find sense in all measurements you do. As e.g., you say the reading at the optocoupler in the good board is 1.2M and 200K in the bad board, but you suspect that difference could be because you substituted a 1.8K resistor for a 2K. We have to agree this is far from making any sense.
Do back to back checks between the good and the bad boards. First of all, that all power transistors are in their correct positions. Someone might have removed any of them and re soldered in the wrong location 😱
Check out all dil 8/14 devices to be inserted in their proper locations and in the correct orientation.
Start making cold measurements in both boards in an area far away from the damaged zone. Must be coincident up to normal component tolerance. (10% difference would be acceptable, but under normal conditions and if all components would be fine there should be almost zero difference between both boards.
Approach with you measurements towards "ground zero"🙂 ANY appreciable difference in measurement between the good and the bad boards should be tracked to the faulty component. Do not apply any power until ALL measurements match or YOU WILL continue blowing up parts and frustrating yourself.
Be methodical and as I recommended work from the outside towards the damaged area.
Take notes for your refference.
I did this with altium, It's a lot of work, I know, It's doable, locate all the parts and start making tracks. Altium does the "backwards" job of make an schematic. I'm inclined not to do it though.
No, It came after the PA stopped working, maybe a 2 days delay.
I removed the pcb unsoldering the trt's, putting them on the heatsink was a lot of work with thermal paste and so on, just to check the lower side of the pcb, I didn't remember there was a small resistor there, the temp sensor and the bc238c. I will clean the pcb bottom layer and then look for blown tracks.
On second though I will solder the trt's back and remove the thermal paste and micas on both working amps, so the cold measurements are reliable and for a change in thermal paste on the working unit.
I agree, doesn't makes sense from any point, I do stupid things, blaming that resistor is one of them.
No, nobody touched the damaged module, all the power trt's are the same type and It's hard to solder them wrong, their solder pads are on the edges of the pcb and correctrly labeled ECB I think.
The 8/14 dil are correctly located, lookig at the working unit when I changed the lm354 I checked the rest, ok on this.
Thanks man, I will do that, my DMM has no capacitor measurement, I will get one and of course and keep making cold measurements starting from the power rails connectors area/relay/coil, this is the far side of the resistor that keeps burning. An excel document will do fine to annotate the results.
Thanks for all your tips and sorry for the stupid thing on the optocupler.
Let us know your findings - Status
- Not open for further replies.