EEVblog Electronics Community Forum

Electronics => Beginners => Topic started by: Cliff Matthews on August 10, 2016, 06:47:50 pm

Title: Mystery TO-3's with built-in diode protection?
Post by: Cliff Matthews on August 10, 2016, 06:47:50 pm
I have 10 (now 9..) all aluminum TO-3 mystery transistors - NPN with gain from 30 to 55. I cracked open one with the least gain to see the die inside, because both my DVM and ESR testers (v2.1) show the presence of a protection diode across EC. However, the ESR tester only shows this diode if hooked up a certain way.. 
Is this just a bug in software? Also given the die size, is there a quick way to test with a heat sink if these could be used as series pass at 5+ amps and 100 watts? I'm a bit confused with Vf showing over 1-volt..
Title: Re: Mystery TO-3's with built-in diode protection?
Post by: madires on August 10, 2016, 07:04:31 pm
I'd guess it's a 2SD388. I'm not sure about the diode, but it could be caused by the leakage current. IIRC, the test for the flyback diode only checks if there are three diodes.
Title: Re: Mystery TO-3's with built-in diode protection?
Post by: bktemp on August 10, 2016, 07:08:37 pm
I'm a bit confused with Vf showing over 1-volt..
It looks like a darlington transistor: You can clearly see two cascaded transistors on the die. Therefore you have a saturation voltage of at least 0.7V.
The low current gain is either a measurement error because of the leakage current caused by buildin resistors (quite common with darlington transistors), or maybe it is a high voltage switching transistor.
Title: Re: Mystery TO-3's with built-in diode protection?
Post by: Cliff Matthews on August 10, 2016, 07:26:43 pm
Just measured the EC diode drops with a DMM. It's steady from 0.44 to 0.45v on all, if that helps. So if the 388's are 2SD388 (80w), what would the 185's be? (that's the type I cracked open).
Title: Re: Mystery TO-3's with built-in diode protection?
Post by: Enigma-man on August 10, 2016, 07:52:20 pm
What you need is access to a transistor curve tracer like the one I have pictured below and a good scope. You would then be able to determine every characteristic without guessing.
Those $20 testers are not the end all to test equipment.  They are good for quick tests of unknown types like you have.

Motorola never made a 2SD anything with date codes of 1974-1980... 2SD are Japanese designations as are 2SA,2SB and 2SC
I have a Motorola Power Device Data book from 1980 that is at least 1000 pages. No 2SD devices in it.
The 338 marking is meaningless unless you have the equipment where the transistors came from.
An MJ4247 is a 2SD388 substitute but it is not a Darlington device.

bktemp is most likely correct as what you have are Darlingtons. Diodes were connected across the collector and emitter.  It's guesswork on your part as
to the voltage and current ratings of the device.
Title: Re: Mystery TO-3's with built-in diode protection?
Post by: Cliff Matthews on August 10, 2016, 07:56:43 pm
Opened one 388 and the die is 50% smaller than the 185 I opened previous. I did a side-by-side and this one too, looks more complex than an NPN (not that I know jack about die features..)  :-//  So the 3 remaining 185's (whatever there real number is) could be good to go at 100 watts then?
Title: Re: Mystery TO-3's with built-in diode protection?
Post by: Cliff Matthews on August 10, 2016, 08:11:35 pm
What you need is access to a transistor curve tracer like the one I have pictured below and a good scope. You would then be able to determine every characteristic without guessing.
Those $20 testers are not the end all to test equipment.  They are good for quick tests of unknown types like you have.

Motorola never made a 2SD anything with date codes of 1974-1980... 2SD are Japanese designations as are 2SA,2SB and 2SC
I have a Motorola Power Device Data book from 1980 that is at least 1000 pages. No 2SD devices in it.
The 338 marking is meaningless unless you have the equipment where the transistors came from.
An MJ4247 is a 2SD388 substitute but it is not a Darlington device.

bktemp is most likely correct as what you have are Darlingtons. Diodes were connected across the collector and emitter.  It's guesswork on your part as
to the voltage and current ratings of the device.
Great info on part numbering. These came to me from my uncle some years back when Litton Canada closed shop in Brampton, ONT. so that explains my inability to look them up. A curve tracer has been on my to-do list for a while.. maybe it's time to make a basic one, thanks!
Title: Re: Mystery TO-3's with built-in diode protection?
Post by: danadak on August 10, 2016, 08:13:36 pm
A short discussion of saturation V in a darlington -

https://en.wikipedia.org/wiki/Darlington_transistor (https://en.wikipedia.org/wiki/Darlington_transistor)

A very common darlington -

http://www.onsemi.com/pub_link/Collateral/TIP120-D.PDF (http://www.onsemi.com/pub_link/Collateral/TIP120-D.PDF)             sat V 2 - 4 V at current


Regards, Dana.
Title: Re: Mystery TO-3's with built-in diode protection?
Post by: edpalmer42 on August 10, 2016, 08:28:28 pm
Is Vf the collector-emitter or the base-emitter voltage?  If it's a darlington, the base-emitter voltage will be two diode drops rather than one.  This will be independent of anything else in the circuit.  Depending on how the die was processed, you might even be able to probe the die and measure each transistor seperately.

Ed
Title: Re: Mystery TO-3's with built-in diode protection?
Post by: Enigma-man on August 10, 2016, 08:37:15 pm
Cliff:
What you probably have are common 2N-WHATEVER Motorola transistors with Litton numbering.
You may end up blowing all the devices and everything hooked up to them before yer done.
Stop the transistor massacre, buy some or find others you can readily identify...  ;D
Title: Re: Mystery TO-3's with built-in diode protection?
Post by: Cliff Matthews on August 10, 2016, 09:17:55 pm
...Stop the transistor massacre, buy some or find others you can readily identify...  ;D
Ha! Guess you're right, ripping open devices is cruel. There ought to be a PETA out there somewhere for sensitive EE's  :)
But it's just what I need to push me into rigging up a curve tracer.   
Title: Re: Mystery TO-3's with built-in diode protection?
Post by: T3sl4co1l on August 11, 2016, 12:59:03 am
Yeah, Darlington for sure.  The small pad (with bond wire) is the base, the "floating" metal is the base-emitter, and the other metallization (with bond wire) is the emitter.  Substrate collector (common to both transistors, handy!)

Normally, B-E resistors are present, to help discharge the internal B-E junction, speeding up switching.  (For the same reason, you can observe Vce(sat) less than 0.7V, but only for low currents, under 100mA or so.)  If your component tester doesn't know the difference between a resistor and a diode, it might think that's what's in there.

Transistors sometimes have a E-C diode (common on high voltage TV transistors, called a "damper diode"), or a E-B diode across the 'drive' transistor(s) (so base drive can be used to assist internal B-E turn-off).  They rarely, if ever, have a full E-B diode.

Tim
Title: Re: Mystery TO-3's with built-in diode protection?
Post by: David Hess on August 11, 2016, 01:56:49 am
I agree with the other posters that they are Darlington transistors.

You can get an idea of the power rating from the die size but will need some known comparison parts for that work.  Since they are TO-3 parts, I doubt the small one is any lower than about 100 watts making the large one 200 watts or more.

If you do not have a suitable curve tracer, then the Vce can be estimated by shorting the base to the emitter and driving about 1 milliamp through the collector to the emitter using a resistor in series with a high voltage DC source.  This will actually return Vcb and Vce will be lower than this.
Title: Re: Mystery TO-3's with built-in diode protection?
Post by: madires on August 11, 2016, 10:31:23 am
Normally, B-E resistors are present, to help discharge the internal B-E junction, speeding up switching.  (For the same reason, you can observe Vce(sat) less than 0.7V, but only for low currents, under 100mA or so.)  If your component tester doesn't know the difference between a resistor and a diode, it might think that's what's in there.

Must be a quite old tester firmware. Both firmwares (k & m) detect a base-emitter resistor since 2014.