schematic..
I looked at my poor schematic.. (do you have better quality)
the R123 it is a 20k.. but looks like it is on the - side and the Q's you are refering to is on + side, how does that connect with them blowint..
- just curiouse, i want to know what might had happend
..
i can see how most of it connects.. it has blown because of to high gatevoltage on output fets??.. and by blowing the fet's it is like shorting the fet, and sending higher voltage back in to the circiut?.
I looked at my poor schematic.. (do you have better quality)
the R123 it is a 20k.. but looks like it is on the - side and the Q's you are refering to is on + side, how does that connect with them blowint..
- just curiouse, i want to know what might had happend
..i can see how most of it connects
.. it has blown because of to high gatevoltage on output fets??.. and by blowing the fet's it is like shorting the fet, and sending higher voltage back in to the circiut?.Sorry... typo. I meant R132.
No, I'm afraid I do not have a better schematic.
The blowing occurs because the load is too high and the outputs can't pull the output voltage up to where it needs to be. The feedback circuit then increases the drive voltage in an attempt to compensate. Quickly the drivers are pushing more voltage into the gates than they can take. The Zeners are supposed to short this out and protect the gates, but, in my case at least, it didn't work that way. You may have better luck. If the Zeners worked the way they were supposed to, then your outputs may be OK (but I wouldn't bet on it).
Also, no, there really aren't any other suitable transistors that can be used in this amp. TMOS is a special transistor topology, suitable, like lateral mosfets, for linear use. The substitutes listed for MTM20N10 are only good for switching applications, like motor controllers or power supplies.
The only sort of similar devices currently available are the Exicons, but they're only rated 16A. These are 20A devices. If you wanted to swap in all Exicons you might be able to rework the amp without too much other work, while limiting its current capability a bit. I don't think it would be worth it, and the Exicons are pretty expensive anyway.
If you need to buy more than 2, then you should check with Rayland. I got the N channels from them for $22 each, but I had to buy enough to meet their $150 minimum. From there I was able to match out the two pairs I needed.
What did you find regarding the Zeners? Are they short? If you want to check one of the N channels, you can ground yourself (to protect from static) then pull one out. Put the meter in diode check mode and apply the red lead to the gate and the black lead to the source. Then quickly switch the red lead to the case; for a good FET, this should read short. If so, apply the black lead to the gate and the red lead to the source, then immediately check with the red lead on the case and the black lead on the source; it should now read open. Red to source and black to case should read a forward diode drop (about .6V).
This test works by using the voltage from the meter to charge or discharge the gate. The gate has enough capacitance so that it will stay charged for a few seconds while you check connectivity between source and drain. P channels can be checked the same way but with red and black leads reversed.
No, I'm afraid I do not have a better schematic.
The blowing occurs because the load is too high and the outputs can't pull the output voltage up to where it needs to be. The feedback circuit then increases the drive voltage in an attempt to compensate. Quickly the drivers are pushing more voltage into the gates than they can take. The Zeners are supposed to short this out and protect the gates, but, in my case at least, it didn't work that way. You may have better luck. If the Zeners worked the way they were supposed to, then your outputs may be OK (but I wouldn't bet on it).
Also, no, there really aren't any other suitable transistors that can be used in this amp. TMOS is a special transistor topology, suitable, like lateral mosfets, for linear use. The substitutes listed for MTM20N10 are only good for switching applications, like motor controllers or power supplies.
The only sort of similar devices currently available are the Exicons, but they're only rated 16A. These are 20A devices. If you wanted to swap in all Exicons you might be able to rework the amp without too much other work, while limiting its current capability a bit. I don't think it would be worth it, and the Exicons are pretty expensive anyway.
If you need to buy more than 2, then you should check with Rayland. I got the N channels from them for $22 each, but I had to buy enough to meet their $150 minimum. From there I was able to match out the two pairs I needed.
What did you find regarding the Zeners? Are they short? If you want to check one of the N channels, you can ground yourself (to protect from static) then pull one out. Put the meter in diode check mode and apply the red lead to the gate and the black lead to the source. Then quickly switch the red lead to the case; for a good FET, this should read short. If so, apply the black lead to the gate and the red lead to the source, then immediately check with the red lead on the case and the black lead on the source; it should now read open. Red to source and black to case should read a forward diode drop (about .6V).
This test works by using the voltage from the meter to charge or discharge the gate. The gate has enough capacitance so that it will stay charged for a few seconds while you check connectivity between source and drain. P channels can be checked the same way but with red and black leads reversed.
transistors..
so i cant find a fet wich simular ohm rating and simular s rating
perhaps same rise and fall time and then replace it with that one ?..
it is late here. so i will go to bed, but i will return in a couple of days with a reply, so hope you will give the post look sunday or monday..
but Thx so much so far..
so i cant find a fet wich simular ohm rating and simular s rating
perhaps same rise and fall time and then replace it with that one ?..
it is late here. so i will go to bed, but i will return in a couple of days with a reply, so hope you will give the post look sunday or monday..
but Thx so much so far
..Still haven't heard from my friend. But I have been following this post, as far the output transistors go my cadaver had a mix of four Motorola MTM20P10s and four IR IRF150s. As I had stated before this amp had been repaired at Sumo before I got it. So MAYBE IRF150s and IRF9150s would work. I haven't checked to see if they are less expensive or more widely available. Last time I searched for TO3(TO204) type Mosfets I found few available new(ie non-surplus) and the ones I did find were expensive. The person that I think took over SUMO after Sir James was Andrew Hefley, but I won't swear to it. I know he did a few white papers for Motorola or ON, don't know if he worked there though. I also heard he was designing for Monarchy Audio in San Francisco, CA. Maybe if he can be tracked down he can give you some info if indeed he was the one. It's been a few years since I've even thought about any of this so I hope my memory serves me right.
Craig
Craig
As far as I can tell, IRF150's aren't in production anymore either. Their NTE sub is $27 from Mouser.
Somewhere around here I have a Motorola power MOSFET book from 1993 or so which includes the MTM20N15 (which replaced the MTM20N10) and the MTM20P10. If I find it I'll scan and post the typical response curve so you have some idea of what you'd have to find to sub it out. Motorola's description of these, in that book, was as vertical MOSFETs that had special characteristics that provided much more linear response, making them suitable for audio output use.
For me it would not be worth it to gamble. Best case, sure, you find that some $5 current production HEXFET is a suitable sub. (This I find very unlikely, in that just a few years ago people didn't think it was at all possible to make a decent audio amp with HEXFETs.) Worst (and much more likely) case, your $5 HEXFETs blow some more of the currently good Motorola parts in the amp. Then you'd have to replace those as well and be out the $10 for the IR's.
Somewhere around here I have a Motorola power MOSFET book from 1993 or so which includes the MTM20N15 (which replaced the MTM20N10) and the MTM20P10. If I find it I'll scan and post the typical response curve so you have some idea of what you'd have to find to sub it out. Motorola's description of these, in that book, was as vertical MOSFETs that had special characteristics that provided much more linear response, making them suitable for audio output use.
For me it would not be worth it to gamble. Best case, sure, you find that some $5 current production HEXFET is a suitable sub. (This I find very unlikely, in that just a few years ago people didn't think it was at all possible to make a decent audio amp with HEXFETs.) Worst (and much more likely) case, your $5 HEXFETs blow some more of the currently good Motorola parts in the amp. Then you'd have to replace those as well and be out the $10 for the IR's.
I just pulled out the Motorola TMOS selector guide and cross reference from 1991, the cross for an IRF150 is their "similar" MTM25N10E, there is NO cross for the IRF9150. In 1991 Motorola had only two P channels in the TO-204 case, 12P10 and 20P10, not much to choose from. Maybe there were more later. I agree that the originals would be best and I would NOT have mixed brands as was done in the amplifier I have. Sponkii needs to figure out whether his outputs are good or bad and bite the bullet if they are bad. I least I had no problem deciding my amps fate because I only had one channel to start with!!!!
Craig
Craig
decision
Huh.. it is a tough choice..
it sounds like it is a good amp, but havnt got a lot of money at the moment, so $30 is quit a bit for each output fet..
but i will see if the oputput fets are dead..
i know that is is not the same type and modifications have to be made, but is it possible to change it with a TO-220 case fet if the rating fits ?..
other wise I think I'm going to take the amplifier apart and use the single channel as a sub woofer amplifier.. (which would be saaaad)..
Huh.. it is a tough choice..
it sounds like it is a good amp, but havnt got a lot of money at the moment, so $30 is quit a bit for each output fet..
but i will see if the oputput fets are dead..
i know that is is not the same type and modifications have to be made, but is it possible to change it with a TO-220 case fet if the rating fits ?..
other wise I think I'm going to take the amplifier apart and use the single channel as a sub woofer amplifier.. (which would be saaaad)..
When I was working on mine I looked into all sorts of replacements. I didn't find anything in any other case that was suitable and cheap either. i.e. No TO-220, TO-3P, or TO-247 that would work. You could maybe parallel in a bunch of the Renesas 8A devices, but it wouldn't be any cheaper.
FWIW, you could also add a pair of ground taps and run one channel as 2 x 50W into 8 Ohms, by running the phase positive and phase negative inputs with L and R signals and running the outputs each between one original binding post and chassis. A bit of a shame, but better than nothing.
FWIW, you could also add a pair of ground taps and run one channel as 2 x 50W into 8 Ohms, by running the phase positive and phase negative inputs with L and R signals and running the outputs each between one original binding post and chassis. A bit of a shame, but better than nothing.
Hi Sponkii,
In searching around a bit recently I came across the Fairchild FQP33N10 and FQP22P10 TO-220 mosfets. These actually do like like they might be an electrically reasonable replacement for the MTM20N10 and MTM20P10. They're also very cheap (About $1 each).
You might want to try removing all the existing fets from the blown channel and swap in a set of these (in addition to replacing the other blown parts). By pulling all the exisiting fets, you're at least protecting the good ones in case this retrofit doesn't work. Of course, you'd also need to modify the mounting to allow the retrofit from TO-3 to TO-220. You'd also probably want a Variac so that you could power the amp up slowly and see how it biases.
Cheers,
Paul
In searching around a bit recently I came across the Fairchild FQP33N10 and FQP22P10 TO-220 mosfets. These actually do like like they might be an electrically reasonable replacement for the MTM20N10 and MTM20P10. They're also very cheap (About $1 each).
You might want to try removing all the existing fets from the blown channel and swap in a set of these (in addition to replacing the other blown parts). By pulling all the exisiting fets, you're at least protecting the good ones in case this retrofit doesn't work. Of course, you'd also need to modify the mounting to allow the retrofit from TO-3 to TO-220. You'd also probably want a Variac so that you could power the amp up slowly and see how it biases.
Cheers,
Paul
Thanks to all
Hey everybody, after some further measuring i found out that there is more damage that expected.
so now i have decided to take the amplifier apart and use one chanel as a sub woofer driver...
if anyone could recommend a good driver to use that would be nice..
(and not something REALLY expensive)
Hey everybody, after some further measuring i found out that there is more damage that expected.
so now i have decided to take the amplifier apart and use one chanel as a sub woofer driver...
if anyone could recommend a good driver to use that would be nice..
(and not something REALLY expensive)
I'd be more interested in the chassis if it weren't so far away. Besides the one channel I do have is just like the one Sponkii has, shot. Do we see a trend here, between Sponkii's, SQLGuy's and mine the only three that I know about since no one else has chimed in are or were shot. Mine has been "repaired" more than once at Sumo!!!! FWIW, went by the last site of the infamous Sumo factory in Agoura Hills just yesterday, didn't see any spare parts though, haha.
Craig
Craig
I wouldn't say too much of a trend. My amp ran very well for about 13 years, until I tried to get it to drive a short (and I was pretty insistent about it!).
Since repairing it, it's been on 24/7 and running well for over a year. The outputs are hard to find, and the design is rather complicated, but I don't think it's inherently flawed or anything... just uncommon.
So, if I could get Craig's partial and Sponkii's blown channel/board/heat sink, I could maybe make another working AII...
Paul
Since repairing it, it's been on 24/7 and running well for over a year. The outputs are hard to find, and the design is rather complicated, but I don't think it's inherently flawed or anything... just uncommon.
So, if I could get Craig's partial and Sponkii's blown channel/board/heat sink, I could maybe make another working AII...
Paul
Crappy...
S. A. 2 might possibly be the worst amp ever!!!!...
right now im about to make an subwoofer with the last channel, and i was about to test it with a speaker, when it suddenly decided to blow..
same problem as last time, haven had the time to check it properly, but i could see that the same resistor has burned.... :S..
so perhaps i have to do some rearing..
if i could do some measuring and explaining of my problem in further details, might some of you be able to help me get it working again..
S. A. 2 might possibly be the worst amp ever!!!!...
right now im about to make an subwoofer with the last channel, and i was about to test it with a speaker, when it suddenly decided to blow..
same problem as last time, haven had the time to check it properly, but i could see that the same resistor has burned.... :S..
so perhaps i have to do some rearing..
if i could do some measuring and explaining of my problem in further details, might some of you be able to help me get it working again..
SQLGuy said:I wouldn't say too much of a trend. My amp ran very well for about 13 years, until I tried to get it to drive a short (and I was pretty insistent about it!).
Since repairing it, it's been on 24/7 and running well for over a year. The outputs are hard to find, and the design is rather complicated, but I don't think it's inherently flawed or anything... just uncommon.
So, if I could get Craig's partial and Sponkii's blown channel/board/heat sink, I could maybe make another working AII...
Paul
I have a SUMO And II that is a later production run. On these runs, they switched to the 1RFP9243 and 1RFP243R devices. My unit is perfect on one side and rebuilt on the other with one problem I can't see to track down. It breaks into oscillation about 1/2 power (you can see it on the sine waver, just at the peak, when you cross a threshold, it starts to oscillate on the back side of the sine wave, both polarities of the sine wave.
I have repaired hi-end hi-fi for years and this was a project I was working on 7 years ago and then stuck it on a shelf. I would like to fix it, it is a challenge, but also would entertain offers if someone is interested.
There are NO burn marks on the board, just this oscillation problem. It biases fine.
The reason the one side is rebuilt is that I didn't realize, when I hooked this up to my test set up, that it was a floating output, so I grounded one side via my scope and speaker load common ground.
I had purchased (if I recall) about 12 output devices per output half in order to match up 4. I built a matching set up and they are within 2% of being perfect (i got a run of devices from the same batch, so was pretty lucky).
This has LED's in the bias circuit and it may be that there is where the problem lies....as you increase the input, just as it starts it's unstable behaviour, they light.
This can be powered on full power and idle all day with no issues and even run at low volumes, it is just when you bring it up to a point, this behaviour starts.
It has really kicked my butt, as over a 30 year carreer in repair there are only one or two items I couldn't fix.
Email for pictures, they are too big to post.
Attachments
I would assume that the biases between the working and rebuilt sides match? Do the LED's on the working side also light when driven to high enough levels?
What about the NULL adjustment? Do you have a procedure for adjusting NULL?
The output is differential/bridged, so you can test by running the two phases as two separate 50W-ish amplifiers with ground taken from the chassis and inputs driven through the separate + and - inputs (assuming your version has those). If you read the two phases separately like this, do you see oscillation on both phases or just one?
What about the NULL adjustment? Do you have a procedure for adjusting NULL?
The output is differential/bridged, so you can test by running the two phases as two separate 50W-ish amplifiers with ground taken from the chassis and inputs driven through the separate + and - inputs (assuming your version has those). If you read the two phases separately like this, do you see oscillation on both phases or just one?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.