I have a Marantz 250M here from the original owner who says it worked fine until he put it in a new system and the speaker leads were shorted blowing one channel.
He had it worked on, but it is still not working. I think it was checked out but not completed.
I have the schematic. Does anyone know the story behind these Motorola SJ2519/sj2520 output devices?
Are they simply beta selected?
The beta range is listed for many of the transistors.
This is a nice looking piece, however it is a rather unusual design. Seems it does not tolerate a shorted output even though it has protection.
Pete B.
He had it worked on, but it is still not working. I think it was checked out but not completed.
I have the schematic. Does anyone know the story behind these Motorola SJ2519/sj2520 output devices?
Are they simply beta selected?
The beta range is listed for many of the transistors.
This is a nice looking piece, however it is a rather unusual design. Seems it does not tolerate a shorted output even though it has protection.
Pete B.
Driver Substitutes
Looking at a Motorola Product Selector Guide and trying to determine what the TO-66 drivers might be,
I see that there are the:
2N3584/2N6421 which are 2A, 35W, 250V, hfe 25 min at 1A, ft = 10 MHz
And the:
MJ3248/MJ3238 which are 8A, 75W, 150V, hfe 40 min at 3A, ft = 20 MHz
The drivers drive the load through a 5.6 ohm resistor but are also boosted by the output transistors.
The outputs offer no help when the output is shorted to ground and the drivers would see
very high currents until the current limiter turns on. This makes me lean toward the 8A part, especially
considering thermal derating and SOA.
MJE15034/35 should be a good substitute.
Pete B.
Looking at a Motorola Product Selector Guide and trying to determine what the TO-66 drivers might be,
I see that there are the:
2N3584/2N6421 which are 2A, 35W, 250V, hfe 25 min at 1A, ft = 10 MHz
And the:
MJ3248/MJ3238 which are 8A, 75W, 150V, hfe 40 min at 3A, ft = 20 MHz
The drivers drive the load through a 5.6 ohm resistor but are also boosted by the output transistors.
The outputs offer no help when the output is shorted to ground and the drivers would see
very high currents until the current limiter turns on. This makes me lean toward the 8A part, especially
considering thermal derating and SOA.
MJE15034/35 should be a good substitute.
Pete B.
Here is the schematic:
http://www.audio-circuit.dk/Schematics/Marantz-250M-pwr-sch.pdf
I don't think they could have been based on the Toshibas, since this amp was sold in 1976 - for just one year.
Pete B.
http://www.audio-circuit.dk/Schematics/Marantz-250M-pwr-sch.pdf
I don't think they could have been based on the Toshibas, since this amp was sold in 1976 - for just one year.
Pete B.
With only a single pair of output device per channel beta selecting (matching) isn't a big deal. If the beta is low it will likely just increase THD a bit at high power levels.
What might be an issue is the stability (compensation) if you put modern faster output devices in. I've worked on some old Marantz amps that hated newer transistors but I don't have experience with the 250M. If you're not familiar with repairing power amps, you might want to either let someone else do it, or read up on how to try to avoid lots of collateral damage if things don't go well when you power up a repaired amp.
If the amp oscillates on power up, for example, it will probably self destruct before you realize it--even with such tricks as a lightbulb in series with the AC power. And a blown output transistor can take other parts with it. So it can be both expensive and frustrating. There are lots of tricks to help avoid problems, but they vary in their effectiveness, applicability, and depend on what test equipment you have access to, etc. If you don't have the right equipment, it can be very hit or miss.
Especially if someone else has attempted a repair, and failed for some reason, I'd be cautious. If it was a simple repair, it probably would be working now. If it left the shop still dead, it was likely either an expensive repair, a failed repair, and/or the shop knew from experience substitute parts wouldn't likely work, etc.
What might be an issue is the stability (compensation) if you put modern faster output devices in. I've worked on some old Marantz amps that hated newer transistors but I don't have experience with the 250M. If you're not familiar with repairing power amps, you might want to either let someone else do it, or read up on how to try to avoid lots of collateral damage if things don't go well when you power up a repaired amp.
If the amp oscillates on power up, for example, it will probably self destruct before you realize it--even with such tricks as a lightbulb in series with the AC power. And a blown output transistor can take other parts with it. So it can be both expensive and frustrating. There are lots of tricks to help avoid problems, but they vary in their effectiveness, applicability, and depend on what test equipment you have access to, etc. If you don't have the right equipment, it can be very hit or miss.
Especially if someone else has attempted a repair, and failed for some reason, I'd be cautious. If it was a simple repair, it probably would be working now. If it left the shop still dead, it was likely either an expensive repair, a failed repair, and/or the shop knew from experience substitute parts wouldn't likely work, etc.
I had the non metered version of this amp - the 240 - for about 18 months in the late 1970's before it blew the output stage (I think RF ingress caused it to oscillate, fryng the output stage).
Bcause of the CFP output, the output devices are probably quite critial in that if you stuck a low ft device in there, it migh t not behave very well. For th e small signal devices, I don't think anything special is needed. This amp could do with a bit of degeneration in th e long tailed pair, but that's not a topic for here.
However, left with little info and not much else, I'd just go for the highest FT devices you can find consistent with the package type (T03). This will probably b e On types, and back in th e 70's, Ft's in thes e big device s were not much bette r than 4 of 5 MHz.
I still have the caps (2 x Sprague 22mfd) from this amp!
If this amp had lasted and blown on me now, I would probably know enought to ge t it going again - but not back in the late 70's - I pretty green then.
Bcause of the CFP output, the output devices are probably quite critial in that if you stuck a low ft device in there, it migh t not behave very well. For th e small signal devices, I don't think anything special is needed. This amp could do with a bit of degeneration in th e long tailed pair, but that's not a topic for here.
However, left with little info and not much else, I'd just go for the highest FT devices you can find consistent with the package type (T03). This will probably b e On types, and back in th e 70's, Ft's in thes e big device s were not much bette r than 4 of 5 MHz.
I still have the caps (2 x Sprague 22mfd) from this amp!
If this amp had lasted and blown on me now, I would probably know enought to ge t it going again - but not back in the late 70's - I pretty green then.
Thanks for all the tips; I've had plenty of experience repairing, debugging, and scratch building amps, usually ones others were unable to fix.
I am curious as to if these were in any way stable.
Perhaps I should first look at the working channel to see if I can measure the open loop transfer function and how well the output protection works.
There are a lot of compensation caps around the output stage and it looks like they had trouble stabilizing this amp.
Pete B.
I am curious as to if these were in any way stable.
Perhaps I should first look at the working channel to see if I can measure the open loop transfer function and how well the output protection works.
There are a lot of compensation caps around the output stage and it looks like they had trouble stabilizing this amp.
Pete B.
Lumba Ogir said:
I am certainly aware of all the issues, however the front end runs at a lower regulated supply voltage and the output stage has gain, so it would be a complete rebuild to do what you say. It is interesting that the 250M output stage has a sort of, virtual ground (AC wise anyway) input to the output stage. Interesting, actually makes a lot of sense - if it could be made reliable.
I agree, on the other hand, that I don't want to waste a lot of time on a hopeless rebuild. How about that 6800 pF cap on the base of the upper outputs? Hmmm .....?
I made a lot of contributions to this thread, you might keep in mind - so I know about CFPs:
http://www.diyaudio.com/forums/showthread.php?s=&threadid=41926&highlight=
Sorry if I called your skills into question 
I don't regularly follow all the threads here enough to know which members have what skills, etc. So it's safest to be cautious when offering advice that could lead to fried parts or worse.
Having a working channel is a good thing. I also agree the compensation in your amp looks like it is a house of cards. And, all too often, amps like that are challenging to get stable with modern transistors.
Marantz did some really innovative things with their old amps and they were impressively well made. If you have ever read any of the old Marantz owner's manuals they are kind of fun. They wax poetic about their unique topologies, etc. But some of what Marantz did in the interest of better performance (or perhaps just to be different for marketing reasons) was at the expense of stability and/or reliability.
Depending on what your ultimate goal is, sometimes 70's solid state amps are best off being upgraded to a much more modern design. It's like restoring a 50's or older car. Do you keep the authentic DC generator and 6 volt battery with all its problems or do you put in a more modern alternator and 12 volt electrical system? It depends on if you're a purist and want everything authentic or you want it to work well and be reliable.
I've often found it better to upgrade old amps using only their chassis, power supply, metering, etc. They stay vintage on the outside but are modern under the hood. You end up with much better performance and reliability.
My current favorite is to use the LME49810/LME49830 drivers from National. They offer better performance than even Douglas Self can get from his optimized "blameless" discrete input and VAS stages and will support a wide range of output stage designs. You can also get ready made PCB's for them that often easily fit within old amp chassis saving the time and expense of a custom PCB.
But, to some, such a conversion is blasphemous. And I can appreciate the challenge in repairing/restoring a vintage amp while keeping it as authentic as possible. But if I was going to actually *listen* to it much, I'd probably go the upgrade route.
I don't regularly follow all the threads here enough to know which members have what skills, etc. So it's safest to be cautious when offering advice that could lead to fried parts or worse.Having a working channel is a good thing. I also agree the compensation in your amp looks like it is a house of cards. And, all too often, amps like that are challenging to get stable with modern transistors.
Marantz did some really innovative things with their old amps and they were impressively well made. If you have ever read any of the old Marantz owner's manuals they are kind of fun. They wax poetic about their unique topologies, etc. But some of what Marantz did in the interest of better performance (or perhaps just to be different for marketing reasons) was at the expense of stability and/or reliability.
Depending on what your ultimate goal is, sometimes 70's solid state amps are best off being upgraded to a much more modern design. It's like restoring a 50's or older car. Do you keep the authentic DC generator and 6 volt battery with all its problems or do you put in a more modern alternator and 12 volt electrical system? It depends on if you're a purist and want everything authentic or you want it to work well and be reliable.
I've often found it better to upgrade old amps using only their chassis, power supply, metering, etc. They stay vintage on the outside but are modern under the hood. You end up with much better performance and reliability.
My current favorite is to use the LME49810/LME49830 drivers from National. They offer better performance than even Douglas Self can get from his optimized "blameless" discrete input and VAS stages and will support a wide range of output stage designs. You can also get ready made PCB's for them that often easily fit within old amp chassis saving the time and expense of a custom PCB.
But, to some, such a conversion is blasphemous. And I can appreciate the challenge in repairing/restoring a vintage amp while keeping it as authentic as possible. But if I was going to actually *listen* to it much, I'd probably go the upgrade route.
No problem rocket, what you say makes perfect sense; I have said similar things myself. But this is a fairly collectable piece, they seem to get a good amount on ebay, at least a few years ago. It seems well worth it to try to get it right if possible.
Those blue meters are so cool, by the way!
I know I could upgrade the design and keep the meters.
Pete B.
Those blue meters are so cool, by the way!
I know I could upgrade the design and keep the meters.
Pete B.
I just found the repair manual today and it seems that the short circuit test is to simply put an ammeter across the output while driving it to 60W at 200 Hz and look for less than 9.5 Amps. There are pots to limit each output stage half, and that test is only going to provide an average value.
Don't know how I missed this thread:
http://www.diyaudio.com/forums/showthread.php?s=&postid=1901166#post1901166
If 9.5 amps is the design limit, thinking now it is probably the RMS value. I'd calculate the peak value, and drive the amp into 1 ohm increasing the level and watching for current limiting on both pos and neg cycles at the correct peak value.
Pete B.
NP: Grace Potter - 2_22: http://www.myspace.com/gracepotterandthenocturnals
Don't know how I missed this thread:
http://www.diyaudio.com/forums/showthread.php?s=&postid=1901166#post1901166
If 9.5 amps is the design limit, thinking now it is probably the RMS value. I'd calculate the peak value, and drive the amp into 1 ohm increasing the level and watching for current limiting on both pos and neg cycles at the correct peak value.
Pete B.
NP: Grace Potter - 2_22: http://www.myspace.com/gracepotterandthenocturnals
The Crown service manuals for some of their huge reference amps suggest all sorts of abusive things including bringing it up to full power with a sine wave and then shorting the output for 10 seconds. I think I'd have a hard time doing that with a $5000 amplifier! 
For current limiter testing I use my arbitrary waveform generator with a 2 cycle very low duty cycle 1 khz tone burst into a 1 ohm load. I slowly bring up the level until the output voltage is around where it should be current limiting. If the tone burst doesn't clip somewhere close to the expected level, I usually investigate further rather than just crank up the level further. This method also usually shows any instability when current limiting.
For current limiter testing I use my arbitrary waveform generator with a 2 cycle very low duty cycle 1 khz tone burst into a 1 ohm load. I slowly bring up the level until the output voltage is around where it should be current limiting. If the tone burst doesn't clip somewhere close to the expected level, I usually investigate further rather than just crank up the level further. This method also usually shows any instability when current limiting.
Hey PB2
Did you get this amp going yet?
I used to work on these puppies years ago, and still have some info tucked away here, Let me know
I do remember the SJ2519/SJ2520 outputs are beta matched to within one beta, hand selected. I know there was a range from around hfe 20-25, 25-30, 30-35, ish, and they did not go above 50 I think. Each color dot on the originals indicate the exact range of beta. Also it was never a good idea to change any two but all 4 at the same time, otherwise parasitic oscilations occurred, and is part of the excess capacitance on the design. If they were not matched properly they would fail in short order. Note the schematic shows the order number for the outputs, partial on each side of the compliment, as an assembly. That was how Marantz provided these outputs.
Specs I remember are output of 16 Amps, Voltage was 160 or 180 VCE Base drive was 4 amps.
These were special manufactur for Marantz by Motorola, there never was a good cross back then, Possibly the MJ15000 series if available might work for these. Also check the Q801 as it was known to occasionally fail when the outputs shorted, check for leakage.
Good luck, its a great amp really, just make sure it is biased critically, and run it under clipping.
I should drag my 250 out and work on it too LOL....... next year!
Harlan
Did you get this amp going yet?
I used to work on these puppies years ago, and still have some info tucked away here, Let me know
I do remember the SJ2519/SJ2520 outputs are beta matched to within one beta, hand selected. I know there was a range from around hfe 20-25, 25-30, 30-35, ish, and they did not go above 50 I think. Each color dot on the originals indicate the exact range of beta. Also it was never a good idea to change any two but all 4 at the same time, otherwise parasitic oscilations occurred, and is part of the excess capacitance on the design. If they were not matched properly they would fail in short order. Note the schematic shows the order number for the outputs, partial on each side of the compliment, as an assembly. That was how Marantz provided these outputs.
Specs I remember are output of 16 Amps, Voltage was 160 or 180 VCE Base drive was 4 amps.
These were special manufactur for Marantz by Motorola, there never was a good cross back then, Possibly the MJ15000 series if available might work for these. Also check the Q801 as it was known to occasionally fail when the outputs shorted, check for leakage.
Good luck, its a great amp really, just make sure it is biased critically, and run it under clipping.
I should drag my 250 out and work on it too LOL....... next year!
Harlan
Hi Harlan,
Sorry I missed this post. No I've not worked on it yet, other than to confirm that the outputs are blown. I don't like the design, so I'm having trouble bringing myself to rebuilding it stock. I'm not sure what I'd prefer to build in the chassis, so it is on hold for now. On the other hand I do think that it is a piece worth collecting so in that sense it seems I should keep it stock. I'd have to look around for replacements for the TO-66 drivers.
Did you work on them because you owned one or was it for your business?
