How to identify a part in a TO92S, TO92 and SOT54 packages?

Thread Starter

dreamer44

Joined Nov 19, 2015
36
Hi all,
I have several components here that look like transistors in a TO92S package. They were glued together, face to face, so it's not possible to see the marking. Does anybody know how to:

1. identify if they are transistors, mosfets, j-fet, regulator, etc
2. identify exactly what kind of transistor, regulator, etc. they are?
 

bwilliams60

Joined Nov 18, 2012
1,442
Do you have any pictures, schematics or anything else to go by? Right now, we have no idea what you have. Always good to dd pictures, diagrams and as much info as you can give us? Otherwise we are blind.
 

Thread Starter

dreamer44

Joined Nov 19, 2015
36
I had the schematics of the original circuit where the part was installed, but it is in a very bad shape so I am re drawing it with the information I have. I don't have a picture of the circuit, otherwise I would have posted it.

I suppose there must be a method to find out what the part is, perhaps with an ohmeter?
 

Dodgydave

Joined Jun 22, 2012
11,284
They're probably Matched Pairs, used in audio output stages, one will be npn, and pnp, the only way is to separate them to get the markings, or use a multimeter with hfe tester on it.
 

Thread Starter

dreamer44

Joined Nov 19, 2015
36
They are already separated and there's no way to see the markings. They were glued face to face, but inverted (in one the leads were to one side, in the other to the other side) so I believe they are the same part.
 

dl324

Joined Mar 30, 2015
16,840
If you have a schematic, it will indicate what type of device they are. Part numbers should also be on the schematic.

If you had devices, but no schematic, you can use an ohmmeter to check junctions on BJTs. For JFET and MOSFET, you'd need a test circuit. You could get a ballpark figure for some characteristics (breakdown voltage, gain, etc) with a curve tracer, but other parameters (frequency, leakage, etc) would have to be determined from usage in the schematic.

If you don't have functioning devices, you could determine important parameters by studying the schematic.
 

Thread Starter

dreamer44

Joined Nov 19, 2015
36
Sorry if this sounds like a puzzle, but by looking with a magnifier I can only see the following, where "x" stands for a letter or number that cannot be read:

xxx17
x
xx 2
 

hp1729

Joined Nov 23, 2015
2,304
Hi all,
I have several components here that look like transistors in a TO92S package. They were glued together, face to face, so it's not possible to see the marking. Does anybody know how to:

1. identify if they are transistors, mosfets, j-fet, regulator, etc
2. identify exactly what kind of transistor, regulator, etc. they are?

Just a guess, but is there an op amp close by? It may be a pair of matched J-FETs to make a J-FET input op amp stage. Physically joined for thermal stability reasons.
 

ian field

Joined Oct 27, 2012
6,536
Hi all,
I have several components here that look like transistors in a TO92S package. They were glued together, face to face, so it's not possible to see the marking. Does anybody know how to:

1. identify if they are transistors, mosfets, j-fet, regulator, etc
2. identify exactly what kind of transistor, regulator, etc. they are?
A transistor tester such as the Peak Atlas will identify the leads for you.

Alternatively - a transistor looks to a DMM like a back to back pair of diodes, in reverse leakage, the B/E junction will zener somewhere around 5 - 8V - but you have to keep the test current very low to avoid damage.

A JFET should look like 2 diodes - but only up to a point, the S/D terminals should conduct either way unless you put a reverse bias on the gate.

MOSFETs can squirm a bit! They have a body diode, so S/D will always conduct one way - whether it conducts the other way depends on the polarity of any static charge on the gate capacitance. The gate should read as infinite resistance either way round.
 

Thread Starter

dreamer44

Joined Nov 19, 2015
36
Here are the readings with a fluke multimeter:

Between leads 2 and 3 (with the red probe on 2) I have 0.65V in diode mode and around 427K resistance (with this number varying +/-1k between readings).
Between leads 1 and 2 (with the red probe on 2) I have 0.65V in diode mode and around 422K resistance (with similar variation).
All other possible placements of the probes yield no reading both in diode and resistance mode.

What can be assumed from this?
 

Thread Starter

dreamer44

Joined Nov 19, 2015
36
Thanks Dave. I'm not sure I follow what Ian mentioned above about the way of determining whether it is a mosfet, jfet, or BJT. Any clue?
 

d0ughb0y

Joined Aug 18, 2015
5
are you trying to identify the part because you are repairing it and need to replace them? or are you salvaging part to use elsewhere?

I'd say your best bet is to reverse engineer the circuit and guess what the part is and try substituting with common part equivalent.
 

hp1729

Joined Nov 23, 2015
2,304
A transistor tester such as the Peak Atlas will identify the leads for you.

Alternatively - a transistor looks to a DMM like a back to back pair of diodes, in reverse leakage, the B/E junction will zener somewhere around 5 - 8V - but you have to keep the test current very low to avoid damage.

A JFET should look like 2 diodes - but only up to a point, the S/D terminals should conduct either way unless you put a reverse bias on the gate.

MOSFETs can squirm a bit! They have a body diode, so S/D will always conduct one way - whether it conducts the other way depends on the polarity of any static charge on the gate capacitance. The gate should read as infinite resistance either way round.
Yep, NPN transistors. Probably a matched pair. Physically joined for thermal reasons,, so they drift equally.
They both appear to be good.
Are you up to the challenge of de-engineering the circuit? Trace out that part of the board to get a schematic?
 

ian field

Joined Oct 27, 2012
6,536
Yep, NPN transistors. Probably a matched pair. Physically joined for thermal reasons,, so they drift equally.
They both appear to be good.
Are you up to the challenge of de-engineering the circuit? Trace out that part of the board to get a schematic?
I used to hand trace VGA monitors on a regular basis.

Mail order schematics were usually so many generations old, the final photocopies were unreadable.

Even the readable ones were an older revision of the chassis on the bench - and if you ever got an up to date and readable schematic, the next one on the in pile would be completely revised design.
 

hp1729

Joined Nov 23, 2015
2,304
I used to hand trace VGA monitors on a regular basis.

Mail order schematics were usually so many generations old, the final photocopies were unreadable.

Even the readable ones were an older revision of the chassis on the bench - and if you ever got an up to date and readable schematic, the next one on the in pile would be completely revised design.
I've done it far too often. Usually the manufacturer was out of business or in China or someplace. Surface mount took all the fun out of that. Not impossible, just a different challenge. BGA? Impossible.
 
Top