This is probably a very easy question and I’m 99% sure I know the answer, but my brain isn’t allowing my to move forward without confirmation.
I have this circuit where I need to replace a 2SA839 and 2SC1669
Generally the recommended replacements are MJE15032g and MJE15033g and they’re definitely overkill for the job. I have plenty of the 15032/15033 transistors on hand, because of an ordering mistake that I made I have 3-4 lifetimes worth of TIP41C and TIP42C, so I’m trying to use them wherever possible.
I understand why the TIP devices are not usually recommended as a replacement because of the lower Vce and Vcb, but that’s of course circuit dependent. The TIP devices exceed the specs of the originals in every way except for the Vce and Vcb, being rated at 100v for each.
With the voltages being all the same polarity for each transistor the TIP devices should easily handle this circuit correct? With the collector to base voltage being around 13.5v and the collector to emitter voltage being around 14v. My brain is says “duh, you know this” while at the same time “maybe you should confirm”.
Thank you,
Dan
I have this circuit where I need to replace a 2SA839 and 2SC1669
Generally the recommended replacements are MJE15032g and MJE15033g and they’re definitely overkill for the job. I have plenty of the 15032/15033 transistors on hand, because of an ordering mistake that I made I have 3-4 lifetimes worth of TIP41C and TIP42C, so I’m trying to use them wherever possible.
I understand why the TIP devices are not usually recommended as a replacement because of the lower Vce and Vcb, but that’s of course circuit dependent. The TIP devices exceed the specs of the originals in every way except for the Vce and Vcb, being rated at 100v for each.
With the voltages being all the same polarity for each transistor the TIP devices should easily handle this circuit correct? With the collector to base voltage being around 13.5v and the collector to emitter voltage being around 14v. My brain is says “duh, you know this” while at the same time “maybe you should confirm”.
Thank you,
Dan
Absolutely, here is its immediate circuit
And then it looks like they handle the voltages for the rest of the circuit, they drive the lines labeled B+
May I ask, what are you looking for? Is there more in the circuit that would tell you what the Vce and Vcb would be?
Dan
Okay great, thanks. Can you explain to me why the 100v rating would be an issue. I thought that the Vce rating meant that it could have up to 100v different between the collector and emitter. With there being a 14v difference between those two pins I’m confused I guess.
Thank you, I appreciate it,
Dan
In this circuit, the collector voltage is said to be 87V, which can vary (being unregulated),
and may be even higher considering the AC line variation and transients during power on/off.
There are large capacitors in the regulator circuit at its input and output terminals,
but no protection diodes are used to clamp the device voltages during power up/down.
Most engineers will strongly derate semiconductors down from the mfr ratings, to account
for device variability, and degradation with time and use, to improve reliability.
All things considered, even with all the parts you already have in stock, it is highly advisable
to use the higher rated parts. If the original 150V parts failed, that is a strong sign that
the 100V parts are not a good idea, and also could cause cascading failures.
It's only two parts, even though they are critical, and the cost won't be that much.
Which audio amplifier is this?
and may be even higher considering the AC line variation and transients during power on/off.
There are large capacitors in the regulator circuit at its input and output terminals,
but no protection diodes are used to clamp the device voltages during power up/down.
Most engineers will strongly derate semiconductors down from the mfr ratings, to account
for device variability, and degradation with time and use, to improve reliability.
All things considered, even with all the parts you already have in stock, it is highly advisable
to use the higher rated parts. If the original 150V parts failed, that is a strong sign that
the 100V parts are not a good idea, and also could cause cascading failures.
It's only two parts, even though they are critical, and the cost won't be that much.
Which audio amplifier is this?
Last edited:
Okay, that makes sense, thank you for explaining it, that actually made sense to me. And actually the originals didn’t fail, so they’re perfectly fine in this circuit as far as voltage is concerned. This is a pair of Kenwood L-07M mono amplifiers. The PNP transistor has to have its base pin spread wide to jump over a trace, I’m guessing the original person to place the transistor got over zealous. When I removed them to apply new thermal compound to their heatsinks I could see that the base pin was barely hanging on. I could add solder to it, so there would be a solder blob right where it meets the body, but I figured it would be better to place new transistors. I’m sure I’ll have plenty of more places to use the TIP devices.
Dan
The originals are just a basic vertical output transistor, absolutely nothing special. 3 MHz fT so a fast replacement won’t matter. The circuit even has fold back current limiting so no SOA concerns either. So why was that one originally selected? Cheap, available, sufficient. A new TIP41C would be preferable to an original with a fatigued lead, or a roll-the-bones part from Ebay. The most direct replacement that you can get easily and from trustworthy sources is KSC2073. Not expensive, if DK or Mouser have any stock.
Thank you for your input. So you feel that it’s possibly the TIP41C could work here? I took Rayma’s advice and put the MJE parts in. They’re so overkill and the fact that I have a few hundred each of the TIP devices I was just looking for a place to use them if they worked. Looks like Mouser has a version of the KSC2073 on hand, but not sure I’d order any unless they would work somewhere that either the TIP devices or the MJE devices wouldn’t. I have another pair of L-07Ms that I should hopefully have soon, I’ll try the TIP devices in it when I rebuild it. Regardless, with the advice given I should know that these parts should last a very long time in this spot.
Dan
If you are working on these amplifiers for yourself, it's one thing, but if you are repairing them for
a customer, never use inadequate parts. You could be liable for failures leading to injury, death,
and/or property damage, like a fire destroying a residence. Not to mention the ethics of the situation.
There are plenty of expert witnesses who can explain in court why a repair was improper.
If you sell repaired equipment on ebay, etc. with inadequate parts you should describe it as for parts only,
even if it functions. Even your own insurance may be void if you have personal damages from such
improper repairs, since that could be considered negligence.
a customer, never use inadequate parts. You could be liable for failures leading to injury, death,
and/or property damage, like a fire destroying a residence. Not to mention the ethics of the situation.
There are plenty of expert witnesses who can explain in court why a repair was improper.
If you sell repaired equipment on ebay, etc. with inadequate parts you should describe it as for parts only,
even if it functions. Even your own insurance may be void if you have personal damages from such
improper repairs, since that could be considered negligence.
Last edited:
100% agree with you. This pair is actually for my father. Which is why I asked the question. If you had come in and said “the TIP41C and TIP42C will be more than overkill in that position” then sure I would use them, but since it’s even a possibility that they may be stressed then it’s not even worth the chance and the Onsemi’s go in. Now this other pair that are coming my way, they’re for me. I may put the TIP devices in, but it would be strictly for observation and testing. If they last 5 years and no issue then I could see them as a viable substitute. But really nobody else is probably going to make the mistake of ordering more TIP devices than they’ll know what to do with. I do on occasion fix audio equipment that leaves my house, I would never take a chance on any of that equipment. I only by premium brands (Nichicon, Panasonic, Nippon for larger filters) and never from eBay unless the persons product has been vetted by folks here. No this current pair I have here already have the MJE15032 and MJE15033 and they’ll stay, appreciate the advice.
Dan
Dan
The TIP41C probably has better SOA than the MJE15032 at 87 volts. Either is overkill. The only thing wrong with using the MJEs for a regulator like this is it brings you one step closer to not having any, potentially when they are the only thing suitable, not in stock at Mouser, and have a year lead time. There are many places you just can’t use TIPs or vertical outputs because the 3 MHz fT just won’t cut the mustard and you wish you still had all your MJE15032’s. Even the humble Dynaco ST120 sounds like crud with TIP41/2 drivers.
That regulator circuit is pretty fault tolerant. 100 mA current limit, folding back to near zero into a short (ie, cold start with that 470 uF output cap). There isn't a TO-220 bipolar made with a VCEO>87V that will have heartburn with that, except perhaps some high voltage switching types like the MJE13007 which probably don’t have enough gain and may drop out of regulation. What usually fails in these situations are the solder joints on the PCB getting oxidized and the board turning to carbon over time if the heat sink is too small for the handful of watts it dissipates - continuously.
That regulator circuit is pretty fault tolerant. 100 mA current limit, folding back to near zero into a short (ie, cold start with that 470 uF output cap). There isn't a TO-220 bipolar made with a VCEO>87V that will have heartburn with that, except perhaps some high voltage switching types like the MJE13007 which probably don’t have enough gain and may drop out of regulation. What usually fails in these situations are the solder joints on the PCB getting oxidized and the board turning to carbon over time if the heat sink is too small for the handful of watts it dissipates - continuously.
Lol, that sounds like a good idea down the road a few years when they’re no longer available and worth $5 a piece lol.
Cool, I really appreciate you confirming this. This is exactly why I was doing it. I mean I still have about 100 of the 15032 and 40 or so of the 15033, but absolutely I’d prefer to conserve them if I have multiple times more of another TO-220 that would work. I’ll throw the TIPs into my personal pair and be confident that they’ll work great.
Dan
I would actually love your opinion on one other transistor swap. I’ll actually make the swap anyways, but would like to hear your thoughts.
I powered the second amp up and had no bias voltage. Went through the amp driver board and going through replacing the problematic transistors I did on the other board (2SA810 and 2SC1452) and found a faulty driver, it’s measuring as a diode. The driver happens to be a 2SC1913. This and the 2SA913 are high speed transistors and there are not a lot of options with transitional frequencies as high.
I ran into this exact issue before when repairing a Marantz 2330b
https://audiokarma.org/forums/index...markthefixer-and-anyone-else-willing.1019084/
To make the repair I stepped from the TO-220 to a smaller TO-126, specifically the KSA1220 and KSC2690. These are superior in every way to the TO-220 as far as voltage, Ic, etc, except in a smaller package. It was something another well known member had been contemplating since there were no readily available replacements without going to eBay.
Long story short, the 2330b has been running not quite daily, but it gets used a lot and the 1220/2690 combo in there have never gotten more than warm. They’re easily taking on the task and the amp sounds fantastic. The 2330b circuit looks to be pretty similar voltage wise. 66/-66v on the collectors in the Kenwood and 67/-67v in the Marantz. Similar voltages on the bases and emitters.
Here they are in the schematic below, Q11 and Q12.
Curious how you guys think the KSA1220 and KSC2690 would do as being drivers in this amplifier?
Dan
That’s one where the answer is yes and no. Normally, TO-126 drivers are not appropriate to replace TO-220 drivers. BUT that particular original is only a 15W device instead of the usual 25-30, so it’s not any “bigger” than the modern TO-126’s. It drives 3 pair, which is ok if you’re not expecting 2 ohm loading, but not enough if you are (don’t use them in a big QSC). The original is no better (or worse). In normal home stereo use on supplies under 70 volts the peak currents dont get stupid enough to cause the output transistor beta to start dropping. Which keeps the load on the drivers reasonable. Driving only one or two pair on those voltages, especially old 4MHz TO-3’s, you would overheat them - either blowing the drivers or causing bias runaway. The originals would have gone poof in short order.
That Fairchild pair is NLA, BTW. Another victim of the SMD craze. Toshiba is still making theirs (TTC011/TTA006, which are actually their last discontinued TO-220 pair stuck in a TO-126). They also have a couple of other choices. For the time being. I bought a stash of C4793/1837 before they went away, along with the C3902/A1507 from ON before they hit the chopping block.
In your particular situation a larger driver could be used, and if the amp is going to be run hard I would. Thats one place where I would put in ME15032/3 without hesitation. It’s an EF3, which can handle a lower fT (30 vs 100-ish) and a higher Cob (from the bigger die) without doing anything that would degrade the sound. In an EF2, the VAS has to deal with that higher capacitance even if the speed is equivalent - which it may or may not always do. In that case you want to match the original die size.
That Fairchild pair is NLA, BTW. Another victim of the SMD craze. Toshiba is still making theirs (TTC011/TTA006, which are actually their last discontinued TO-220 pair stuck in a TO-126). They also have a couple of other choices. For the time being. I bought a stash of C4793/1837 before they went away, along with the C3902/A1507 from ON before they hit the chopping block.
In your particular situation a larger driver could be used, and if the amp is going to be run hard I would. Thats one place where I would put in ME15032/3 without hesitation. It’s an EF3, which can handle a lower fT (30 vs 100-ish) and a higher Cob (from the bigger die) without doing anything that would degrade the sound. In an EF2, the VAS has to deal with that higher capacitance even if the speed is equivalent - which it may or may not always do. In that case you want to match the original die size.
That’s really the only thing that gave me confidence to try it in the Marantz was the fact that it’s such a low wattage TO-220 device. That and Markthefixer who I wholeheartedly trust was also contemplating making it a recommended swap, though when I asked him about it he had not gotten around to testing.
So am I reading this correctly, at a given voltage the less outputs the drivers are driving it’s harder on the driver? So being on the same voltage these should have an easier time in the Kenwood since the driver pair are driving 3 pair vs the Marantz only having 2 pair?
Yeah I know about the Fairchilds being gone, sad really all of these useful through hole devices disappearing. When I heard about it I was lucky enough to snag 200 of each from Mouser before they disappeared. I still have about 180 pair or so left, only because I now use the TTC004b and TTA004b where possible and save the 2690/1220 for where necessary. I was actually considering the Toshiba devices for this task, but felt the Fairchilds to be better suited. They 2690 is actually still available, but once those PNPs are gone, they’re gone sadly. I should pick up some of those TTC011/TTA006.
You’re lucky to have grabbed the 4793/1837 and 3902/1507. Those went bye bye when I was still fairly fresh into electronics and was only buying the commonly used things like 1845/992 and such. I actually do have 4 pair of the 4793/1837 and a few dozen pair of C5171/A1930, but they’re pulls from a few Onkyo surround receivers so they legs are a bit short. I’d have to solder extensions onto the legs, maybe I could save them for something somewhere where the Tao-126 wouldn’t work??? Though I don’t know when that would be. I’m going to of course have to do the other amp so that they match, so two pair will be used.
So the 15032/15033 come up again. Not sure if you saw in the thread I linked, but I did. Run into issues with using them in the Marantz, but you already addressed that I wouldn’t in these amps. If I did go with the 15032/15033 then I’d probably replace the pair I already used in the other amp with the TIP41/42 like we discussed earlier so that I only use two pair for this pair of amps instead of four pair. It looks like once again the MJE15032 is out of stock and won’t be back for nearly a year. So I’d like to use as few as I can.
I’m going to throw the 2690/1220 in just to see how they do. I don’t imagine these amps being run extremely hard, lowest load they’d ever see is 4 ohm and my father doesn’t own any hard to drive speakers, but at the same time I’d want to future proof them for the next guy that ends up with them whenever that is. As I mentioned, I have driven the Marantz pretty darn hard and not an issue at all. Would you put the extensions on the C5171/A1930? Or use the 15032/15033? I’ll work on getting the TO-126 in and report back.
Dan