Hi everyone,
I have a HH electronics MXA150 mixer/amp combo that I am trying to repair. It came to me with a suspected blown bridge rectifier (which was massively under-specced anyway) and suspect output transistors.
I have replaced the output transistors (2N3773 x2) with CDIL (Continental Device India) replacements instead of the standard Motorola devices and have also replaced the bridge rectifier with a 25A type. I checked the old output devices for shorts and failed, appearing to be blown, so I decided to replace them them just for good measure (they are only £3 each). I checked the driver transistors TIP31/32 pair and they tested fine so I proceeded to start it up and check the bias and whatnot.
I turned it on and everything appear to be fine, no sparks or smoke and tried to adjust the bias to 10-20mv across the two emitter resistors. The reading would not stay stable and drifted a lot (probably due to the lack of a multiturn trimmer) so I turned the amp off with the bias turned all the way down.
The heatsink got strangely warm (not too hot to touch) in this first turn on (possibly the cause for the drifting readings) quite quickly. I then turned the amp on again, and within a few seconds, the two rail fuses blew up. They were 5A types and were completely obliterated (very little wire left at all).
I am still wondering what the problem could be so have attached the schematic to help. The power amp circuit is at the bottom of the page.
I have a HH electronics MXA150 mixer/amp combo that I am trying to repair. It came to me with a suspected blown bridge rectifier (which was massively under-specced anyway) and suspect output transistors.
I have replaced the output transistors (2N3773 x2) with CDIL (Continental Device India) replacements instead of the standard Motorola devices and have also replaced the bridge rectifier with a 25A type. I checked the old output devices for shorts and failed, appearing to be blown, so I decided to replace them them just for good measure (they are only £3 each). I checked the driver transistors TIP31/32 pair and they tested fine so I proceeded to start it up and check the bias and whatnot.
I turned it on and everything appear to be fine, no sparks or smoke and tried to adjust the bias to 10-20mv across the two emitter resistors. The reading would not stay stable and drifted a lot (probably due to the lack of a multiturn trimmer) so I turned the amp off with the bias turned all the way down.
The heatsink got strangely warm (not too hot to touch) in this first turn on (possibly the cause for the drifting readings) quite quickly. I then turned the amp on again, and within a few seconds, the two rail fuses blew up. They were 5A types and were completely obliterated (very little wire left at all).
I am still wondering what the problem could be so have attached the schematic to help. The power amp circuit is at the bottom of the page.
Yes, I made sure they were isolated from the heatsink (mica washers, plastic bushes etc.). Plus, if they weren't properly isolated, the fuses would've blown straight away at the first power up. The way this amp works, one of the mounting bolts is used to carry the collector voltage from a pad on the PCB and is uninsulated from the device (but is insulated from the heatsink). The second bolt is used to secure the other device mounting hole and is insulated from the heatsink AND the metal case.
I'm afraid of powering it up again without any kind of protection though. I'm contemplating using the light bulb limiter trick or maybe using resistors instead of fuses. Also, I might replace the emitter resistors too, because this amp is old and they are only 5% tolerance.
I was just wondering what I should do if I were to power it on again, bearing in mind that it doesn't have any kind of protection circuit at all.
BTW, I have an oscilloscope, DMM and function generator available for testing the amp.
I was just wondering what I should do if I were to power it on again, bearing in mind that it doesn't have any kind of protection circuit at all.
BTW, I have an oscilloscope, DMM and function generator available for testing the amp.
I think that I was measuring DC offset when I was testing the amp using my scope on the output. From what I remember, there wasn't any at all. I turned the bias all the way down when I powered it up again (I think, the pot seemed to be working so that anti-clockwise was increasing bias current).
I know there is still a fault but am looking for some suggestions as to what it might be. There are no blown or leaky capacitors on the board and no visibly burned parts.
One thing I will remember to do next time I power it up is to disconnect the cable going to the front panel boards to rule out the possibility of a problem with the opamps and inputs.
I know there is still a fault but am looking for some suggestions as to what it might be. There are no blown or leaky capacitors on the board and no visibly burned parts.
One thing I will remember to do next time I power it up is to disconnect the cable going to the front panel boards to rule out the possibility of a problem with the opamps and inputs.
I don't know your experience from the post, but for testing, I would make a signal ground ref. for the input if you pull the input connector. Considering the J112 muting fet, using the 1k resistor to the "Send" socket would seem OK for this purpose.
Signs are that there will be a DC problem. Did you measure output offset at the junction of the emitter resistors R23,4 rather than say at the speaker or phones sockets? Regarding the unstable bias settings; it's not clear but could the preset or surrounding circuitry be duff and possibly the source of problems?
Certainly the light bulb trick is the cheapest protective option for firing up unknowns and the visual indication is at least convincing but some pros swear by variacs. I guess that in experienced hands, that's correct too. Don't concern yourself about issues with emitter resistor tolerance. Most commercial amps use 10% tolerance WW. If your concerns were fanatical precision, audiophile dogma or optimal current sharing in a high power array, maybe you would think about it. If they are intact here in an instrument amp, then they are just fine.
Signs are that there will be a DC problem. Did you measure output offset at the junction of the emitter resistors R23,4 rather than say at the speaker or phones sockets? Regarding the unstable bias settings; it's not clear but could the preset or surrounding circuitry be duff and possibly the source of problems?
Certainly the light bulb trick is the cheapest protective option for firing up unknowns and the visual indication is at least convincing but some pros swear by variacs. I guess that in experienced hands, that's correct too. Don't concern yourself about issues with emitter resistor tolerance. Most commercial amps use 10% tolerance WW. If your concerns were fanatical precision, audiophile dogma or optimal current sharing in a high power array, maybe you would think about it. If they are intact here in an instrument amp, then they are just fine.
Last edited:
Thanks for your reply Ian. I am quite experienced at building amps and although I haven't repaired a transistor amp before, I still have the knowledge to do so.
Unfortunately, I was only keeping an eye on offset at the speaker sockets using my oscilloscope. Next time I power it up, I will put my spare DMM at the junction instead of on the output socket. I had two meters checking the amp at the time, the better one was measuring the bias whilst the other one (from what I can remember) was just there in case I needed to measure anything else. The oscilloscope was on the output, but because of its position on my bench, I can't see the screen unless I move away from what I was working on.
I need to get some fuses and put together a light bulb limiter before I power it up as I'm sure it will come in handy repairing other amps.
Unfortunately, I was only keeping an eye on offset at the speaker sockets using my oscilloscope. Next time I power it up, I will put my spare DMM at the junction instead of on the output socket. I had two meters checking the amp at the time, the better one was measuring the bias whilst the other one (from what I can remember) was just there in case I needed to measure anything else. The oscilloscope was on the output, but because of its position on my bench, I can't see the screen unless I move away from what I was working on.
I need to get some fuses and put together a light bulb limiter before I power it up as I'm sure it will come in handy repairing other amps.
Last edited:
Hi guys, sorry to resurrect this old thread but the question this time is on exactly the same topic.
I have finally got some new (hopefully genuine) output transistors along with the two driver transistors (TIP31/32C).
I now have a 0-30V/0-3A dual bench supply which I am able to use for testing purposes so I was thinking of setting the current limit to around 100ma and slowly upping the voltage and see how things go.
I would really like to know if someone could tell me in light of my first post which way the bias pot works on this amp (see attached schem. in the first post). In the same post, it says I turned the bias down to what I thought would be minimum (anticlockwise-all the way) and instead blew the fuses on the next power up. I am just thinking does the pot on this amp work the other way round-clockwise to turn the bias down instead of up?
This time I will definitely have a multimeter instead of oscilloscope at the junction between the two emitter resistors, one measuring the bias, the inputs grounded and as mentioned earlier, slowly increase the voltage with my multimeter. 20mV (worst case bias) through two 0.22ohm emitter resistors is around 90mA through the output devices in my calculations so 100mA current limit should be OK to give a bit of headroom (assuming things work OK) to set the bias and allow for current drawn in other stages. It will also be enough to limit current in the case of a problem.
I have finally got some new (hopefully genuine) output transistors along with the two driver transistors (TIP31/32C).
I now have a 0-30V/0-3A dual bench supply which I am able to use for testing purposes so I was thinking of setting the current limit to around 100ma and slowly upping the voltage and see how things go.
I would really like to know if someone could tell me in light of my first post which way the bias pot works on this amp (see attached schem. in the first post). In the same post, it says I turned the bias down to what I thought would be minimum (anticlockwise-all the way) and instead blew the fuses on the next power up. I am just thinking does the pot on this amp work the other way round-clockwise to turn the bias down instead of up?
This time I will definitely have a multimeter instead of oscilloscope at the junction between the two emitter resistors, one measuring the bias, the inputs grounded and as mentioned earlier, slowly increase the voltage with my multimeter. 20mV (worst case bias) through two 0.22ohm emitter resistors is around 90mA through the output devices in my calculations so 100mA current limit should be OK to give a bit of headroom (assuming things work OK) to set the bias and allow for current drawn in other stages. It will also be enough to limit current in the case of a problem.
Well everything turned out OK. I put the new transistors in and set the bias pot to the middle (so either way it wouldn't blow up the output transistors), set the current limit to 100ma and slowly advanced the voltage.
No sparks flew at all even as I advanced my PSU to 30V, everything was fine. I re-adjusted the bias (turns out it does work in the opposite direction) and let it idle for a bit. Around 1mV of offset was on the output which was perfectly acceptable. I then turned everything off, removed the current limit and power up again. Once again, no smoke or sparks and everything was good.
Connected a 4ohm speaker to its and music came out! This thing is loud! Now I just need to power it from the mains (while I was at it I thought I might as well rewire the mains connections with some decent spade terminals) and readjust the bias once again for the higher power supply voltage.
Thanks everyone for all the help. So glad I could get this up and running again!
No sparks flew at all even as I advanced my PSU to 30V, everything was fine. I re-adjusted the bias (turns out it does work in the opposite direction) and let it idle for a bit. Around 1mV of offset was on the output which was perfectly acceptable. I then turned everything off, removed the current limit and power up again. Once again, no smoke or sparks and everything was good.
Connected a 4ohm speaker to its and music came out! This thing is loud! Now I just need to power it from the mains (while I was at it I thought I might as well rewire the mains connections with some decent spade terminals) and readjust the bias once again for the higher power supply voltage.
Thanks everyone for all the help. So glad I could get this up and running again!
Yeah, the PSU was a really handy tool and I'm sure it will be again in the future (it wasn't really a `present` as such because I paid for it). It was expensive, but at the end of the day, you pay for quality and it is certainly well built.
Even though I did replace the driver transistors as well, I think the problem may have only been limited to the outputs but replacing the drivers too didn't do any harm either.
I am fairly certain that I got genuine parts this time as they have the same batch and date code layout as a genuine ON Semi device should, the lettering doesn't rub off easily and they feel heavier than the previous devices. Surprising considering I got them from a back street-not very well known-but good electronics shop in my local town. I will definitely be buying from them again (not that keen to pay Farnell's £20 minimum order!).
Hopefully with future faults, they should only be limited to the output transistors as this is a fairly rugged amp and I think, had it not been for fake parts and a bias pot operating the other way round, I may have fixed it first time.
Even though I did replace the driver transistors as well, I think the problem may have only been limited to the outputs but replacing the drivers too didn't do any harm either.
I am fairly certain that I got genuine parts this time as they have the same batch and date code layout as a genuine ON Semi device should, the lettering doesn't rub off easily and they feel heavier than the previous devices. Surprising considering I got them from a back street-not very well known-but good electronics shop in my local town. I will definitely be buying from them again (not that keen to pay Farnell's £20 minimum order!).
Hopefully with future faults, they should only be limited to the output transistors as this is a fairly rugged amp and I think, had it not been for fake parts and a bias pot operating the other way round, I may have fixed it first time.
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.