Measure junction capacitances for one.
Characterising transistors isn't really an easy thing to do. It takes time and you basically set up an experiment. I don't have the time to mess around doing this unless I really have to.
Hi NanoFarad,
Well, physics doesn't care about cost and availability. The part is either right, or it isn't. You either have a failure or you don't. Nothing else matters. Talking about other factors only attempt to explain a decision you made. I'm not trying to be difficult, but reality comes down to things that do not involve cost and availability. You either get smoke or music, it really is that simple.
Characterising transistors isn't really an easy thing to do. It takes time and you basically set up an experiment. I don't have the time to mess around doing this unless I really have to.
Hi NanoFarad,
Well, physics doesn't care about cost and availability. The part is either right, or it isn't. You either have a failure or you don't. Nothing else matters. Talking about other factors only attempt to explain a decision you made. I'm not trying to be difficult, but reality comes down to things that do not involve cost and availability. You either get smoke or music, it really is that simple.
The reality of the situation in much of the world, or within many peoples budgets, is that you simply have to design and build around what is availavle. Which simply may not be good enough to build something 100+ watts per channel. But if you can get 20 or 40 without smoke, smuggling in power transistors like they were a couple of Ki‘s, or spending more than your $20 weekly allowance, one does what they have to do.
Availability (at least in the developed world not run by petty dictators) of parts that are the next step beyond generic is 50,000 times better now than when I started in the 70’s. If I could have gotten real C5200’s or 2N3773’s for what Mouser charges NOW back in 78 I’d have been all over that so fast it would make your head spin. I’d have dumped all my salvaging operations and not looked back.
Availability (at least in the developed world not run by petty dictators) of parts that are the next step beyond generic is 50,000 times better now than when I started in the 70’s. If I could have gotten real C5200’s or 2N3773’s for what Mouser charges NOW back in 78 I’d have been all over that so fast it would make your head spin. I’d have dumped all my salvaging operations and not looked back.
You are muddying the water, that's the problem. If the part says BD139, it is BD139 & it will work in most situation. You can't expect identical, same to same specifications from different manufacturers, there must be some tolerance in spec which is definitely undesirable but we're helpless.
Btw my earlier post was about something which is not available in my country & i think vertical fets are not identical to laterals.
Regards
Hi NanoFarad,
I'm sorry, but with all due respect ... I clarified the situation.
If a part says it's a BD139 and it is out of manufacture, chances are it is not a BD139. Period, especially if stamped with a manufacturer that did not originally make it.
Different manufacturers are not identical, but they are normally darned close in parameters that matter. If they are not, the part cannot be marked as that particular part number and is a forgery at that point in time. There is no discussion or argument, we are dealing with facts here. Availability, price, how you might feel has zero to do with the object in question. So, if you can't get a currently manufactured part, a fake is not acceptable and the repair shouldn't be completed. That's up to you, but wishful thinking is not a good business plan. Up to you.
If something is not available to you in your country, that is unfortunate. Sometimes you can't have what you want. I would have loved to have some real uPC1225H, but all I got was three entire batches of different "real" uPC1225H that don't work. They are all fakes. So do I cry about it? No, that repair is scrapped. I don't like it, but that is the way of the world. Too bad for me and my customer. I'm the only one that lost money in that situation, never mind the time. The customer merely had to accept that equipment is unrepairable.
Vertical FETs are completely different in behavior to horizontal FETs. You might get them to work, but not well. Horizontal FETs have a much larger die area, so they were always expensive. They were discontinued leaving a ton of product as unrepairable. Again, too bad. Without the right part you cannot properly repair an item, much less build a project using them. Move on.
I'm sorry, but with all due respect ... I clarified the situation.
If a part says it's a BD139 and it is out of manufacture, chances are it is not a BD139. Period, especially if stamped with a manufacturer that did not originally make it.
Different manufacturers are not identical, but they are normally darned close in parameters that matter. If they are not, the part cannot be marked as that particular part number and is a forgery at that point in time. There is no discussion or argument, we are dealing with facts here. Availability, price, how you might feel has zero to do with the object in question. So, if you can't get a currently manufactured part, a fake is not acceptable and the repair shouldn't be completed. That's up to you, but wishful thinking is not a good business plan. Up to you.
If something is not available to you in your country, that is unfortunate. Sometimes you can't have what you want. I would have loved to have some real uPC1225H, but all I got was three entire batches of different "real" uPC1225H that don't work. They are all fakes. So do I cry about it? No, that repair is scrapped. I don't like it, but that is the way of the world. Too bad for me and my customer. I'm the only one that lost money in that situation, never mind the time. The customer merely had to accept that equipment is unrepairable.
Vertical FETs are completely different in behavior to horizontal FETs. You might get them to work, but not well. Horizontal FETs have a much larger die area, so they were always expensive. They were discontinued leaving a ton of product as unrepairable. Again, too bad. Without the right part you cannot properly repair an item, much less build a project using them. Move on.
So I did some capacitance measurements on a few different transistors with my cheap handheld capacitance meter.
I have an old gray philips BD137 as benchmark, and first compared that to a new BD139 from CDIL: very little difference, capacitance between different pins is vitually identical for the two devices.
Next I tried a new BD140 from STmicro. Here all capacitance values are maybe 30 percent higher. Looking at the datasheets this is to be expected because it is a PNP part, they have higher capacitance and lower Ft than the NPN parts.
On the basis of these simple measurements, nothing seems clearly wrong with the new parts.
I have an old gray philips BD137 as benchmark, and first compared that to a new BD139 from CDIL: very little difference, capacitance between different pins is vitually identical for the two devices.
Next I tried a new BD140 from STmicro. Here all capacitance values are maybe 30 percent higher. Looking at the datasheets this is to be expected because it is a PNP part, they have higher capacitance and lower Ft than the NPN parts.
On the basis of these simple measurements, nothing seems clearly wrong with the new parts.
Hi gijser,
Seems right then. I also test for breakdown voltage. I have a current source with compliance to over 300 VDC (HP 6186C), so I can set the current low enough not to damage other things. I wouldn't use the part after a breakdown test. You could run the part up to the max rating in the datasheet to prove it's okay. There is normally safety headroom on top, so the part should never break down at the max rating. In service or design, you avoid that maximum even for peak conditions.
Seems right then. I also test for breakdown voltage. I have a current source with compliance to over 300 VDC (HP 6186C), so I can set the current low enough not to damage other things. I wouldn't use the part after a breakdown test. You could run the part up to the max rating in the datasheet to prove it's okay. There is normally safety headroom on top, so the part should never break down at the max rating. In service or design, you avoid that maximum even for peak conditions.
The bigger of those two dies will fit in a TO-92. There are a lot of 625 and 800 mW TO-92’s and that’s about how big they are. The BD’s - even the Philips - are rated 1.5 amps max collector current. That little bitty one can’t fit a big enough bond wire on it to do 1.5 amps emitter current.
So CDIL part turned out to be up to the mark. My experience was the same. But reality is most people would prefer ST or ONsemi part because of their big portfolio & so called consistency on the specification. I understand that because i always follow the same 
But the point is if availability is the problem then using CDIL, UTC or KEC is far wise choice than using parts from unreliable source with "ST" stamp on it.
But the point is if availability is the problem then using CDIL, UTC or KEC is far wise choice than using parts from unreliable source with "ST" stamp on it.
If they are good, then great. However when my reputations and reliability are on the line I will always go with those with a long track record of excellence. It doesn't matter what country they are from.
I've dealt with On Semi since they were Motorola, as most here have. We have watched them steadily improve over time. So ... they are a very well known quantity. I wouldn't buy TI transistors simply due to their past track record. Good thing they don't produce transistors anymore (that I know of).
I agree. A known fake is never an option even if that is the only thing apparently available. But then, a fake isn't available really. Using those is like knowingly using parts not rated for the job.
I've dealt with On Semi since they were Motorola, as most here have. We have watched them steadily improve over time. So ... they are a very well known quantity. I wouldn't buy TI transistors simply due to their past track record. Good thing they don't produce transistors anymore (that I know of).
I agree. A known fake is never an option even if that is the only thing apparently available. But then, a fake isn't available really. Using those is like knowingly using parts not rated for the job.
We prefer ST or ON because they can be tacked on to a Mouser order over here, when ordering a couple hundred bucks of other stuff. It’s more trouble for me to get CDIL than ST, because it requires going out of my way to do it. Maybe I’ll intentionally try out theirs some day but for now I’ve got too much on my plate.
Most of my experience with TI transistors is their horrible versions of TIP TO-220’s. They were the originals, but the worst in the business. Even ungraded Motorolas (no bloody A,B,C, or D) from radio shack beat TIP C’s from TI.
Most of my experience with TI transistors is their horrible versions of TIP TO-220’s. They were the originals, but the worst in the business. Even ungraded Motorolas (no bloody A,B,C, or D) from radio shack beat TIP C’s from TI.
The problem is a large number of non- professional people not good at identifying fake components. On the other hand some people knowingly uses such parts for repairing work because of non availability of the specific component, more profit margin or to make customer happy who doesn't want to spend much on repairing their television or receivers. This happens mostly in developing countries & sometimes i find nothing wrong in it although i'm strongly against counterfeiting. The most interesting thing is, here 90% technician are magician because they can repair almost everything without having proper instruments (forget about having an oscilloscope).
Hi NanoFarad,
We have exactly the same problems in Canada. Untrained idiots decide they can repair electronics (incorrectly) and charge a lot for ruining equipment. The few remaining good techs have to clean up afterwards.
Greed is the #1 problem, followed by ignorance and false pride. Earlier (decades ago) we did not need government intervention. Today, I would say we need to have the industry regulated. I'm going to bet that the problem is worldwide, the same in all countries.
We have exactly the same problems in Canada. Untrained idiots decide they can repair electronics (incorrectly) and charge a lot for ruining equipment. The few remaining good techs have to clean up afterwards.
Greed is the #1 problem, followed by ignorance and false pride. Earlier (decades ago) we did not need government intervention. Today, I would say we need to have the industry regulated. I'm going to bet that the problem is worldwide, the same in all countries.
Hi Rick,
Agreed on the industry. It is the wild west now, a complete mess.
Yes, but those are not large die silicon. TI does fine with integrated circuits, and these dual transistors are basically that same process. Thanks for letting me know about those devices, I'll check them out. But I'll never buy a power device from TI unless they have corrected the issues they had back then.
Gee, enough protection diodes in the JFE150? wow! The JFE2140 being a dual device. Cool. $6.25 each single quantity at Digikey, and in stock. How do you find them?
Agreed on the industry. It is the wild west now, a complete mess.
Yes, but those are not large die silicon. TI does fine with integrated circuits, and these dual transistors are basically that same process. Thanks for letting me know about those devices, I'll check them out. But I'll never buy a power device from TI unless they have corrected the issues they had back then.
Gee, enough protection diodes in the JFE150? wow! The JFE2140 being a dual device. Cool. $6.25 each single quantity at Digikey, and in stock. How do you find them?