I have just seen MJ15024 used by MANY people from various websites. I have no experience with them myself. Chris, no need to say sorry. I wouldn't consider it disagreeing. I simply state an opinion, some from experience, learned from others, observations etc. and am very glad to have it confirmed, shot down, corrected etc. I'm glad others put stuff out on the table for us all to learn.
Edit: this is 1 such source
http://synthetizer-sche.chez-alice.fr/power ampli/phase linear/repair.htm
Edit: this is 1 such source
http://synthetizer-sche.chez-alice.fr/power ampli/phase linear/repair.htm
Hi Richard,
I stand corrected there. Some used some slower outputs I thought.
Quasi complementary uses NPN and PNP drivers and outputs of one type only (NPN or PNP). One pair is a follower, the other a Sziklai configuration.
Full complimentary uses NPN and PNP drivers and outputs as followers (NPN driver to NPN output). On Semi's newer devices have been designed with this configuration in mind to reduce distortion.
I am not a fan of Quasi complimentary output configurations (sorry Quasi!
), but I'm sure they can be made to work well. My personal bias (yuk yuk yuk).
I hope that helps you out a little. I have attached the data sheet for MJ21195/96.
-Chris
Okay, if those are the originals, use the MJ15024 and MJ15025 parts. 2SD555 are very fast transistors. If my memory is correct, they are good for 10A, so you are adding a good 50% in current capacity.
I stand corrected there. Some used some slower outputs I thought.
I'm pretty sure they are slower (4 MHz gain bandwidth product). The gain vs collector current curve is flatter and the SOA is greater. I personally love how tightly On Semi is holding these parts close to each other.
I'd have to research that. I have not had the need to replace any lately and can't remember what I did years ago. Check them first. Leakage would be my major worry if there is no massive failures (open, short).
It might be clear silicon grease. The new stuff is normally white. Replace the mica insulators while you are at it, you need to buy thermal compound anyway (not from Radio Shack).
I like full complimentary. Whether it's worth changing or not is dependent on the execution of the circuit.
Quasi complementary uses NPN and PNP drivers and outputs of one type only (NPN or PNP). One pair is a follower, the other a Sziklai configuration.
Full complimentary uses NPN and PNP drivers and outputs as followers (NPN driver to NPN output). On Semi's newer devices have been designed with this configuration in mind to reduce distortion.
I am not a fan of Quasi complimentary output configurations (sorry Quasi!
), but I'm sure they can be made to work well. My personal bias (yuk yuk yuk).I hope that helps you out a little. I have attached the data sheet for MJ21195/96.
-Chris
Hi Bigred,
That's a good link for these amps. Thanks for posting a link. This information is accurate (I remember reading this site a year ago approx.).
I do sometimes get confused between threads and amps. Too many amps across my bench to remember them unless I got burned badly on one. Then I remember it clearly.
Thank you for clarifying.
-Chris
That's a good link for these amps. Thanks for posting a link. This information is accurate (I remember reading this site a year ago approx.).
I do sometimes get confused between threads and amps. Too many amps across my bench to remember them unless I got burned badly on one. Then I remember it clearly.
You know what? I do have to say I was sorry. I was mistaken and you were quite right. You should always offer information you believe is correct. Keeps guys like me in line!
Thank you for clarifying.
-Chris
anatech said:
The ORIGINALS are MJ15024. The 2SD555's were put in when it was repaired in 1988. According to the service bulletin dated 10-81, 2SD555's can be used in both Series 1 and 2 models, while the MJ15024 are to be used only in Series 2 model amps.
My amazingly ignorant question(s): where does the grease go? (and why do the insulators need replacing?)
Yes, very much, though I am still a bit confused. I also downloaded the data sheet for the MJ15024 (attached) and cannot tell what the differences really are. Most of the specs are the same. A few people have mentioned the MJ21196 as an upgrade; if it is slower, is that really an upgrade?
Richard
Richard,
the MJ15024 is definitely the way to go. I "think" Chris's concern was oscillation issues using a faster device in a unit which used slower devices. I believe we have cleared that up.
The insulators are easily damaged. It is just good practise to use new insulators even if the old ones appear okay. Apply a THIN EVEN coating to the back of the transistor and either to the heatsink where the transistor mounts or the insulator itself. You want a good mating surface for heat transfer here and don't reef down on the mounting screws.
the MJ15024 is definitely the way to go. I "think" Chris's concern was oscillation issues using a faster device in a unit which used slower devices. I believe we have cleared that up.
The insulators are easily damaged. It is just good practise to use new insulators even if the old ones appear okay. Apply a THIN EVEN coating to the back of the transistor and either to the heatsink where the transistor mounts or the insulator itself. You want a good mating surface for heat transfer here and don't reef down on the mounting screws.
Let me confuse the issue more, then!
I found this on http://hometown.aol.com/phasetek
--------------------------
Q: Is it ok to use the newer ON Semiconductor MJ 21193/94 or MJ 21195/96 as substitutes for the TP 9054/DB 15025 (MJ 15024/25)?
A: Yes, if the amp was originally built with theTP 9054 and MJ 15025. If the amp was built with PL 909's or 910's, see the above caution. ON Semiconductor introduced these parts several years ago. The specs are nearly identical to the MJ 15024/25 except for the Safe Operating Area (SOA) spec. The MJ 15024/25 are rated at 2 amps @ 80 volts, the 29113/14 at 2.5 amps @ 80 volts and the 21195/96 at 3 amps @ 80 volts. According to ON, they have the largest die of any TO-3 power transistor on the market which means they can dissipate more heat and operate more reliably. Cost of all three is roughly the same so the MJ 21195/96 is the obvious choice if available.
---------------------------
The TP9054 is a rebadged MJ15024. So this seems to say that the MJ21196 is preferred. Maybe I should write Ed and Dean and getthe full scoop.
Richard
I found this on http://hometown.aol.com/phasetek
--------------------------
Q: Is it ok to use the newer ON Semiconductor MJ 21193/94 or MJ 21195/96 as substitutes for the TP 9054/DB 15025 (MJ 15024/25)?
A: Yes, if the amp was originally built with theTP 9054 and MJ 15025. If the amp was built with PL 909's or 910's, see the above caution. ON Semiconductor introduced these parts several years ago. The specs are nearly identical to the MJ 15024/25 except for the Safe Operating Area (SOA) spec. The MJ 15024/25 are rated at 2 amps @ 80 volts, the 29113/14 at 2.5 amps @ 80 volts and the 21195/96 at 3 amps @ 80 volts. According to ON, they have the largest die of any TO-3 power transistor on the market which means they can dissipate more heat and operate more reliably. Cost of all three is roughly the same so the MJ 21195/96 is the obvious choice if available.
---------------------------
The TP9054 is a rebadged MJ15024. So this seems to say that the MJ21196 is preferred. Maybe I should write Ed and Dean and getthe full scoop.
Richard
Very interesting. Looks like your gonna have to find out what your originals were. Although the series II is more compatible with MJ15024's, Chris's concern of oscillation is still a real possibilty even with the newer MJ 21193/94 or MJ 21195/96 depending on the originals used if I'm understanding correctly. My head hurts. 
Bigred said:Very interesting. Looks like your gonna have to find out what your originals were. Although the series II is more compatible with MJ15024's, Chris's concern of oscillation is still a real possibilty even with the newer MJ 21193/94 or MJ 21195/96 depending on the originals used if I'm understanding correctly. My head hurts.
My originals are the restamped MJ15024 (PL took them and put on TP9054; according to the service manual they are one and the same). The NECs were installed back in 1988.
Hi Richard,
If the 2SD555's were installed and they didn't oscillate, you can use the MJ1502X types. The MJ21196 types should work fine as well. There is always the chance that the 2SD555's were borderline stable. In this case you really should have an oscilloscope and dummy loads.
The MJ21196 has a larger SOA than the MJ15024, but that shouldn't be a concern unless you pound the poor thing. I would probably convert it to full complimentary and use MJ21195 and MJ21196 only because the process is better. Higher SOA on these would not be my over riding concern.
Concerning the mica insulators, it's best to use new ones. They do crack and split easily. Get a package of new ones. If your were to crack one, or split it through the thickness, human nature would have me assume that you would use that insulator anyway. Just get a stack of new ones. Enough to do both sides and more (spares). Get some new thermal compound at the same time. Don't worry about getting the best, just use Wakefield's normal product.
Hi Bigred,
-Chris
If the 2SD555's were installed and they didn't oscillate, you can use the MJ1502X types. The MJ21196 types should work fine as well. There is always the chance that the 2SD555's were borderline stable. In this case you really should have an oscilloscope and dummy loads.
The MJ21196 has a larger SOA than the MJ15024, but that shouldn't be a concern unless you pound the poor thing. I would probably convert it to full complimentary and use MJ21195 and MJ21196 only because the process is better. Higher SOA on these would not be my over riding concern.
Concerning the mica insulators, it's best to use new ones. They do crack and split easily. Get a package of new ones. If your were to crack one, or split it through the thickness, human nature would have me assume that you would use that insulator anyway. Just get a stack of new ones. Enough to do both sides and more (spares). Get some new thermal compound at the same time. Don't worry about getting the best, just use Wakefield's normal product.
Hi Bigred,
I expect all really old amps to oscillate. I've even seen film capacitors open in supply bypass locations. Some don't oscillate until they are delivering a signal, and then only on one part of the waveform. If you expect the worst, you will be pleasantly surprised more often.
-Chris
Hi Richard,
I'll agree with Bigred here. As far as I'm concerned, I'm only happy with mica and grease. I think that rubbery stuff was added later.
Hi Ron,
I might use the MJ2119x in my Marantz, probably not the MJ15024 / 25 since the beta between NPN and PNP were so far apart. The original transistors were much better in that regard (2SD665?). I do recall the Japanese parts were quite tough as well.
The bias currents are set by non-adjustable components, mostly VBE drops. How did yours measure? Bias current?
-Chris
I'll agree with Bigred here. As far as I'm concerned, I'm only happy with mica and grease. I think that rubbery stuff was added later.
Hi Ron,
I might use the MJ2119x in my Marantz, probably not the MJ15024 / 25 since the beta between NPN and PNP were so far apart. The original transistors were much better in that regard (2SD665?). I do recall the Japanese parts were quite tough as well.
Tell me this wasn't a Huntron like thing. The only leakage I measured one these was due to contaminants on the case. I measure leakage directly. It's possible those were damaged because the Toshiba parts were far better than US parts until very recently. Just for matching purposes alone, I used Toshiba over US parts. That and the Toshiba ones ran rings around US transistors for speed. You did match your parts, right?
The bias currents are set by non-adjustable components, mostly VBE drops. How did yours measure? Bias current?
-Chris
Hi Chris,
My test gear showed me a better pattern (flatter line) on the newer devices. There wasn't any leakage that would effect performance.
Interesting that the beta was a closer match than the MJ. Does it really make a big difference in the design of this amp?
Thanks for the heads up.
Ron
My test gear showed me a better pattern (flatter line) on the newer devices. There wasn't any leakage that would effect performance.
Interesting that the beta was a closer match than the MJ. Does it really make a big difference in the design of this amp?
Thanks for the heads up.
Ron
Hi Ron,
The current product from On Semi pretty much addresses those concerns. Those Toshiba parts that were original ran pretty hot for 20 years, so they may have degraded.
I'm not sure how you are testing your parts, but that's okay as long as you know how to interpret the display. I will say that it is important to match the drivers and outputs (all of them really) as that does improve performance. Did you check the bias current?
-Chris
I'm pretty sure it does. That is the one thing that bugged me about Motorola parts for years. You could match NPN and PNP, but the two were normally a factor of 2.5 ~ 3 apart. The Japanese parts were normally pretty close and faster on top of that. My matching yield was much better using the Japanese parts too.
The current product from On Semi pretty much addresses those concerns. Those Toshiba parts that were original ran pretty hot for 20 years, so they may have degraded.
I'm not sure how you are testing your parts, but that's okay as long as you know how to interpret the display. I will say that it is important to match the drivers and outputs (all of them really) as that does improve performance. Did you check the bias current?
-Chris
Hi Ron,
DC offset and bias current are two things you should always check on a repaired amplifier. You might measure under bias THD if you were to load the output with 4 ohm and about 1 Vrms output at 10 KHz or so. Higher powers and lower frequencies tend to measure much better.
The 'scope is only useful for looking at residuals.
-Chris
Edit: Do you still have the original parts? What was the reason for the replacement?
The 300DC is a high feedback amplifier. Any bias would make it measure okay.
DC offset and bias current are two things you should always check on a repaired amplifier. You might measure under bias THD if you were to load the output with 4 ohm and about 1 Vrms output at 10 KHz or so. Higher powers and lower frequencies tend to measure much better.
The 'scope is only useful for looking at residuals.
-Chris
Edit: Do you still have the original parts? What was the reason for the replacement?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.