Mystery Component

On a sunny day (Wed, 30 Jan 2013 20:16:28 -0800) it happened "Paul Hovnanian P.E." wrote in :

Why not measure one? PNP-NPN? Beta Build a curve tracer (Vce breakdown).

:-)

Oh, Ft..

Quick NPN PNP determination with a meter in diode test mode. Allow for base in center or to one side. Fets show a body diode in reverse and touching the meter between G-S will store enough charge to measure some forward conduction D-S. Grab the leads with your fingers to drain the charge and D-S should look open. Even nicer if your meter has a transistor tester function that gives a figure of merit on beta. Nothing beats a curve tracer though. I've got a 30 year old sencore one.

Hard to say if it's a house number but looks like it. When you buy enough or pay enough, vendors will private label. Saw that all the time when I worked with mil-spec stuff.

Which begs ( for ) the question: Has anyone attempted to develop a method for identifying components simply from their characeristics?

Plug a BJT into a test socket, drive the pins, measure the traces, feed the results into a database, and out pops...

"It's a 2N2222A, dummy!"

Hm... is there a way of testing ( say ) a BJT's BVce "gently", that is, so that it doesn't _stay_ broken down?

Jes' curious.

Frank McKenney

--
  Nearly every feature of the American system of manufacturing, from 
  the elements of the new textile machinery to the concept of 
  interchangeable parts, had actually been conceived earlier by 
  Europeans.  But while a few Europeans could see the possibilities, 
  their communities kept them powerless to give their ideas a free 
  trial.  Too many had a stake in the older ways.  Industrial progress 
  in Europe required extraordinary courage to break the prevailing 
  pattern; in America it required a willingness to try the obvious. 
  American genius was less for invention or discovery than for 
  experiment. 
    -- Daniel J. Boorstin / The Americans: The National Experience

On a sunny day (Thu, 31 Jan 2013 13:00:32 -0600) it happened Frnak McKenney wrote in :

Ferry Lon Taim AGOH I designed and build a transistor curve tracer. Was just not much work, and boss wanted one. To keep dissipation in limits you need to use low impulse duty cycle. It could find NPN or PNP automatically. No micros at that time, just relays and logic, output to a normal xy scope.

I am sure with a PC these days you could look up things in a database, and have a better display.

Sure, a real BJT (perhaps excepting some RF exotica) can be pushed beyond BVce so long as current is kept low. You can see this quite nicely on a curve tracer if you have access to one. I haven't heard of any problems with devices "abused" in this way, not even MTBF.

BVeb is another matter. Breaking down EB junctions reduces beta, increasing noise and may lead to shortened MTBF. Supposedly forward biasing "anneals", but I don't know if it's completely reversible. IIRC its a function of total-reverse-charge, so you should be able to go into a very limited breakdown for a short time without "significant" damage - perfectly sufficient for a tester.

-F

Apparently Peak-Atlas are marketing a (massively) upgraded version of their

3-terminal component analyser, it has a USB port for connection to a PC (I think it may actually do curve tracing).

The original version I bought about a decade ago will at least tell you if a device is transistor, MOSFET or just a pair of diodes - anything more complex and it'll just report "unrecognised or faulty component".

Also how to tell you have a "digital" transistor in front of you?

On a sunny day (Fri, 1 Feb 2013 10:42:07 -0000) it happened "N_Cook" wrote in :

WTF is a 'digital transistor'?????????????????

is,

beyond

a

problems

go

Exactly, thats how they are referred to in UK listings, and without any " "s

On a sunny day (Fri, 1 Feb 2013 11:36:40 -0000) it happened "N_Cook" wrote in :

Well, cannot help you there, but a transistor is a transistor, no matter what is written on - or about it. A rose is a rose by any other name (to keep it UK related).

Transistor with base resistor(s) built in. Eg.

Best regards, Spehro Pefhany

--
"it's the network..."                          "The Journey is the reward" 
speff@interlog.com             Info for manufacturers: http://www.trexon.com 
Embedded software/hardware/analog  Info for designers:  http://www.speff.com

On a sunny day (Fri, 01 Feb 2013 08:06:21 -0500) it happened Spehro Pefhany wrote in :

OK, but that is actually an integrated circuit. Just mis-labeled 'transistor'. :-)

You can still test those with a multimeter and lab supply. On a curve tracer for transistors it would look strange I'd guess.

If you see reverse conductance in base emitter junction with an Ohm meter, that would be an indictation. If that resistance was the same both ways, more so.

I'd probably toss it :-)

On a sunny day (Fri, 01 Feb 2013 13:17:42 GMT) it happened Jan Panteltje wrote in :

PS, some of us here will remember RTL logic. That was pretty much like that, but more transistors to make gates, and output R too. Integrated circuits.

Its a regular transistor with built in resistors - some are exactly described that way rather than "digital", the most common arrangement is a B-E bleed resistor + a base series current limiting resistor.

They are often used directly as an open collector inverter gate.

My first job was component level fault finding on Olympia desk calculators, they contained about 4 boards of DTL - a 5th board at the back was critically sensitive MOS shift registers. The boards were about the same area as S100, but wider & not so high, the front board had about a dozen nixie tubes.

I must've just missed out on RTL by not all that long, it was only just becoming scarce in component catalogues of the day.

Any idea what $169 is converted to GBP?

Ian Field Inscribed thus:

Just over £105.00p

--
Best Regards: 
                        Baron.

On a sunny day (Fri, 1 Feb 2013 16:46:09 -0000) it happened "Ian Field" wrote in :

Thanks. Spehro even gave a link to a datasheet.