I have some Sanken 2SC3858 and 2SA1494s in T200 package pulled out of a Urei 6260 amplifier because one side has two of them blown. The rest of them all measure much lower hfe gain than the data sheets suggested. The NPN 3858s measure in the low 20s when the data sheet says it should be a minimum of 50.
So I was wondering if damage or age can effect HFE?
So I was wondering if damage or age can effect HFE?
There should be no change in characteristics over time whether by age or use. Measurement on hFE can depend on the test conditions such as applied voltages and test currents.
The lower binned ones, if using a cheap component tester, I wouldn't be surprised to see a really low hfe reading. 35 @ 25 deg C, maybe lower if cold? Warm them up a bit see what you get?
Last edited:
Thank you everyone. Yes, you are right. Cheap tester and an Atlas DC55 all state low hfe because of the low voltage they test at. Makes sense now.
It’s the low voltage low current that’s doing it. I’ve seen old school TO-3s measuring single digit.
B-E avalanche won’t happen in an output stage unless it’s an old transformer driven one. Severe Thermal damage just makes it leaky (increased Icbo), which can manifest itself in a higher hFE reading on a cheap tester. Initial thermal damage won’t change characteristics, but will increase operating temperature, as the thermal resistance is degraded due to weakening the die attach. Age itself does nothing in an epoxy molded package.
Older transistors themselves are more susceptible to the low current hFE issue. NJW0281’s don’t do that. They measure their full gain on my cheap DMM’s hFE range. I’m kind of surprised a lot of the highly touted Sankens, even newish ones, are still doing that. Not all do, but many do. Toshiba worked that bug out a long time ago, and it’s been ages since I’ve seen any of theirs doing it.
B-E avalanche won’t happen in an output stage unless it’s an old transformer driven one. Severe Thermal damage just makes it leaky (increased Icbo), which can manifest itself in a higher hFE reading on a cheap tester. Initial thermal damage won’t change characteristics, but will increase operating temperature, as the thermal resistance is degraded due to weakening the die attach. Age itself does nothing in an epoxy molded package.
Older transistors themselves are more susceptible to the low current hFE issue. NJW0281’s don’t do that. They measure their full gain on my cheap DMM’s hFE range. I’m kind of surprised a lot of the highly touted Sankens, even newish ones, are still doing that. Not all do, but many do. Toshiba worked that bug out a long time ago, and it’s been ages since I’ve seen any of theirs doing it.
I had an amp go faulty.
I tested all transistors with a diode test and they all tested ok.
I couldnt find a problem anywhere else.
In the end I took out all transistors and did an Hfe check.
A long tailed pair transistor had an Hfe of 1.
Replaced it and amp worked fine.
LTP is one of the places you can get the B-E avalanche problem. Leave it with a DC fault for a long time, as in “Channel in Protect - Do Not Use”, and that LTP will be toast by the time somebody gets around to fixing the bad channel. Severe overdrive can cause this too.
And then there is “2SC458 disease”, which can happen on LTPs too because several of their favorite low noise PNPs were made this way. Any of those old air-cavity Jap TO-92s with the epoxy seal at the bottom are susceptible. It’s NOT the modern over-molded package and moisture gets in eventually. Won’t usually cause hFE to fall, but they get noisy and sometimes outright fail.
And then there is “2SC458 disease”, which can happen on LTPs too because several of their favorite low noise PNPs were made this way. Any of those old air-cavity Jap TO-92s with the epoxy seal at the bottom are susceptible. It’s NOT the modern over-molded package and moisture gets in eventually. Won’t usually cause hFE to fall, but they get noisy and sometimes outright fail.
Last edited:
Absolutely ! NJW3281 all measure with their full gain. I have quite a few and most measure very close to each other too.
I'm not so sure about that when some of the output transistors are blown, depending on the circuit and the failure mode of the blown transistors.
When transistors get blown, anything can happen to the rest of them. That’s why when you blow one output you replace both, and the drivers. Something bad usually happens to the rest of them. Transformer drive can go to avalanche if you just drive it hard enough. TV horizontal outputs were supposed to be run that way.