• Please note that Audiokarma will be offline briefly for updates early on Monday morning, September 28th.

Any way to identify unknown transistors?

elnaldo

elnaldo

Lunatic Member
Hello. Cleaning my parts box I found I have several transistors without any label, but they look too nice to trash them, They are TO-220 devices, red body, golden leads, probably old motorola or RCA, without anything printed on them.

Is there a circuit I could build to check some basic specs as Vce, Ic and gain?

Thanks !
 
Register to hide this ad
I'm currently embarking on building one of these (which so far look like it hasn't been ripped off yet by the Chinese). Could it help at all? Alternatively there is another curve tracer as a kit on eBay manufactured by a Taiwanese company that is available as a kit- it looks very good, but you also unlike the first, need a scope. https://www.instructables.com/id/Transistor-Curve-Tracer/
 
pinout usually can be found easily
polarity is easy to find
gain is easy to measure once you find pinout
Ft is also easy to measure but requires a signal generator.
Vcemax / Vcbmax is kind of hard, you could use a curve tracer to sort of find it, but it's still sort of like the last hard ones:
Ic/Iemax and Pmax are the hardest of all, have to destructive testing if you don't have the most sensitive of tools...
 
I'd start with one of those cheap component testers. Besides a regular transistor, it could be a dual diode, a regulator IC, an FET, triac, or ???? (although if you're sure its a regular transistor then you have a leg up.)
 
Could it be that the red paint is hiding the printing underneath? Maybe carefully wiping paint thinner on the paint will reveal the printing.
Some manufacturers will do this so you have to buy replacements from them. I have a Sony power amp with proprietary output transistors. They have Sony part numbers printed on them. They are in fact US mil. spec. transistors. Bought as surplus by Sony and used in this amp model.
 
I have a cheap component tester tester (but never tested this devices), pinout is not a problem at all.

I also have a signal generator.

The hard part is to see the Vce and Ic.

The devices are not painted, at least they look like the material itself is red.

I'll post picts later.

Perhaps it's time to build an Octopus Curve Tracer !
 
FWIW I do recall some red transistors, with gold leads, however they may not be the same. GE D44 series...picture from web
 
Hello,
There are cheap options to measure the breakdown voltage with this DY294 chinese transistor tester. It puts a current limited high voltage across the junction and gives you the breakdown voltage this parameter may rule out certain devices out of all the possible similar looking transistors. I found that the curve traces are some sort of a fingerprint. I used to check TO-220 parts that i purchased in local shops to see if they were fakes. I used a home built a curve tracer for this. There is a way to simultaneously apply a known C-E voltage and increase the base current and then register the colector current to see if the device has a SOA according to that of its supposed datasheet. I used this too, but the way I implemented it is cumbersome. For a while i was searching for a way to measure the transition frequency, and I did find some schematics, but i never built it.
 
Even with a curve tracer, you don't know the die size and power capability. Without a different set of tests you don't know how fast they are. I might build a test breadboard or something that was really noncritical, but I wouldn't use them to fix anything or build anything new. They're also very old as that round lead style hasn't been used in near-to-forever.
 
It's not hard to safely "saturate" a transistor that only needs to run a low collector current. Pretty much even a low β transistor you only need a slight amount of base current to saturate, just 100µA base current will easily saturate any transistor at 100µA collector current.

The trick is to find the SOA of the transistor without "destructive testing" - running the transistor at its limits before it gets "too hot" which is subjective and not necessarily easy to tell as the junction temperature is what needs to be measured, not case temperature.

(The avalanche/reverse bias voltage limits are another thing but those usually can be current limited enough to not cause damage while testing. If it arcs internally it's too much and becomes inadvertent destructive testing.)
 
You must log in or register to reply here.
Back
Top Bottom