Hi Do,
These transistors are normally close in gain anyway when you buy them. Refresh my memory, how many of what do you need?
After a search, Mouser seems to have stock on the NJL0302, and hopefully the NJL0281 also. Buy more than you need and you can check matching by looking at the voltage across the emitter resistors. If any are way off, you can just change it out. Alternatively, you can rig up a quick jig to test them before mounting, but if my memory serves me, they were mostly really close in beta with the odd flyer - and it wasn't too far off either. This will be less expensive than buying a "matched set".
Mouser wants $5.15 each in Canadian dollars in single quantity. $4.38 each if you buy 10 at a time.
-Chris
These transistors are normally close in gain anyway when you buy them. Refresh my memory, how many of what do you need?
After a search, Mouser seems to have stock on the NJL0302, and hopefully the NJL0281 also. Buy more than you need and you can check matching by looking at the voltage across the emitter resistors. If any are way off, you can just change it out. Alternatively, you can rig up a quick jig to test them before mounting, but if my memory serves me, they were mostly really close in beta with the odd flyer - and it wasn't too far off either. This will be less expensive than buying a "matched set".
Mouser wants $5.15 each in Canadian dollars in single quantity. $4.38 each if you buy 10 at a time.
-Chris
Hi Chris,
I just can't remember if the Ns and Ps had to be matched together or just Ns with Ns and Ps with Ps..?
I'll have to look into the manual.
I still would prefer an already matched set though, save me the trouble.. I'm lazy! LOL
Ciao!
Do
ok then... That will cost me lots of BJTs to be able to match them. I don't even know if I buy 20 of each that I'll be able to match them...
Maybe I put this amplifier in the closet until someone can sell me their match transistors kit.
Thanks
Do
Funny enough, they seem to have an office just 5 minutes from where I live...! If it's still there, I will go there directly.
Do
Matching closer than his limits is pointless between compliments. It would be nice, but what is critical is the match between NPN parts, and between the PNP parts. Normally the devices matching in a single lot is pretty close.
Matches between parts for MJ21195 and MJ21196 is a little wider between PNP and NPN parts, but very close within each polarity. I just bought 25 pairs from Arrow that are due in tomorrow to bring stock levels back up and allow for a decent yield for matching. I still have 9 pairs to go through, but it doesn't pay to wait until you're close to out of stock these days! My order is coming from two different warehouses as it is.
Expecting any better is hand wringing for no reason. Back in the 80's, parameter spread was huge, even for premium Japanese parts. Stocking outputs was no joke in those days no matter from Motorola or Toshiba. I had thousands tied up in parts inventory. Gee, I have that today as well! I guess you get numb to stocking parts when you are used to times that were a lot worse. By comparison, devices today are a dream come true.
The fact is that today, we are spoiled by how good parts like output transistors are. 2:1 beta ranges between NPN and PNP parts are common and always have been. The dream devices were MJW0302 and MJW0281 as they were all clustered around 100 for a beta reading. I wish they had not discontinued those parts. They were the best I had ever seen by far.
The voltage amp section will compensate for the 2:1 beta difference between those parts. You only need to worry about each polarity, and that is where matching really counts. Polarity matching is nice to have but completely unrealistic to expect.
Relax, build and be happy.
-Chris
Matches between parts for MJ21195 and MJ21196 is a little wider between PNP and NPN parts, but very close within each polarity. I just bought 25 pairs from Arrow that are due in tomorrow to bring stock levels back up and allow for a decent yield for matching. I still have 9 pairs to go through, but it doesn't pay to wait until you're close to out of stock these days! My order is coming from two different warehouses as it is.
Expecting any better is hand wringing for no reason. Back in the 80's, parameter spread was huge, even for premium Japanese parts. Stocking outputs was no joke in those days no matter from Motorola or Toshiba. I had thousands tied up in parts inventory. Gee, I have that today as well! I guess you get numb to stocking parts when you are used to times that were a lot worse. By comparison, devices today are a dream come true.
The fact is that today, we are spoiled by how good parts like output transistors are. 2:1 beta ranges between NPN and PNP parts are common and always have been. The dream devices were MJW0302 and MJW0281 as they were all clustered around 100 for a beta reading. I wish they had not discontinued those parts. They were the best I had ever seen by far.
The voltage amp section will compensate for the 2:1 beta difference between those parts. You only need to worry about each polarity, and that is where matching really counts. Polarity matching is nice to have but completely unrealistic to expect.
Relax, build and be happy.
-Chris
Do, I have measured/matched hundreds of NJL0302/NJL0281 transistors in the past years.
I have attached the results of matching about 50 transistors of each type a year ago.
I have measured the transistors with a test-rig that I have set up especially for measuring/matching.
I have measured with the collector current resulting in about 170mA.
According to my experience it is "impossible" to "cross-match" NPN and PNP.
You will be very lucky, if you find 2 "matched" triples of each, when you buy 10-12 pieces of each.
Best regards - Rudi_Ratlos
I have attached the results of matching about 50 transistors of each type a year ago.
I have measured the transistors with a test-rig that I have set up especially for measuring/matching.
I have measured with the collector current resulting in about 170mA.
According to my experience it is "impossible" to "cross-match" NPN and PNP.
You will be very lucky, if you find 2 "matched" triples of each, when you buy 10-12 pieces of each.
Best regards - Rudi_Ratlos
Attachments
Maybe you should just put two new LEDs. It is possible you damaged one of them from excessive heat while soldering. Small diodes are fragile to heat. If one of them is damaged, it could just be out of specs. Also check the current limiting resistors connected to the LEDs to make sure you have the proper resistors on both sides
Do
Hi Do,
That was a solid suggestion (replacing the LEDs).
I think you will be fine with your transistors. They also don't absolutely have to be exactly the same (but that's nice). 20 years ago I was buying 30 of each to get the same matches you'll see with today's parts and processes.
When you measure the beta, the transistors must be at the same temperature. That means you don't touch or breathe on them. Just attach the clips of the tester without touching any transistors. Write the beta on a pad instead of on the part until you are done. Of course, this also means you don't do this in the direct sunlight or near air vents.
-Chris
That was a solid suggestion (replacing the LEDs).
I think you will be fine with your transistors. They also don't absolutely have to be exactly the same (but that's nice). 20 years ago I was buying 30 of each to get the same matches you'll see with today's parts and processes.
When you measure the beta, the transistors must be at the same temperature. That means you don't touch or breathe on them. Just attach the clips of the tester without touching any transistors. Write the beta on a pad instead of on the part until you are done. Of course, this also means you don't do this in the direct sunlight or near air vents.
-Chris
Yeah, 20 years ago I was 24 years old and just starting my career.. I had been interested in electronics for many years before and was buying parts at radio shack at the time, didn't quite have the money to buy what you were buying. Hehe
Definitely not working with the same budget these days. LOL!
Ciao!
Do
Also keep in mind that it is a shunt reg, if you test it with no load it will have to dissipate all that energy through the heatsinks. The shunt has been calculated specifically for the IS load so when it is attached to the proper load it should be dissipating less heat in theory
Also, as anatech suggested, check for oscillation.
Do
Also, as anatech suggested, check for oscillation.
Do