Wouter van Ooijen wrote:
> Unfortunately, I seem to have been on the *selling* end of such 
> transistors (MJ15003). What would be the best way to identify such 
> transistors from measurable parameters? I measured the Beta at 
> reasonable current (3A) but it was OK. Higher current? Higher 
> dissipation? Or maybe a dynamic parameter?

A curve-tracer, some careful examination and a bit of "sacrificial blast test".

Check that the characteristic curves (Vc vs Ic, for a range of Ib values) are 
close to the datasheet spec. Don't expect them to be exactly right, but they 
should be damn close.

Also check Hfe at various ratings, and junction breakdown voltages. AIUI, 
genuine parts tend to land somewhere in the middle of the Hfe range; I haven't 
personally seen any parts from a reputable manufacturer that are "just barely" 
in spec (i.e. Hfe very close to minimum permissible Hfe).

The blast-test is basically a controlled overload. Find out what the maximum 
power rating is, and run the transistor at that for a few minutes. I've heard 
the pulse ratings are most likely to fry a fake.. :)

Or you can smash open a known-genuine part and a suspect part. Compare the die 
sizes, construction, and other attributes of the genuine part against the 
suspect (or suspects).

At one point I was looking at building some form of "fake transistor detector" 
-- basically a bunch of A/D converters and some high-power programmable power 
supplies, built up to form a combined pulse-tester and curve-tracer. I never 
got round to fleshing out the design and building the thing (though I do have 
a near-complete design for a Zener diode tester that can go up to 200V and 
runs from a 3.7V Canon digital camera battery).

Funny how most of my projects seem to gravitate towards test and measurement...

-- 
Phil.
piclist@philpem.me.uk
http://www.philpem.me.uk/
-- 
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist