On 14/01/2014 23:55, Gareth Magennis wrote:
"Phil Allison" wrote in message
...

"Gareth Magennis"
"Phil Allison"


I recently bought some TIP35C transistors from Farnell UK. Just before
I fitted one to an amplifier I thought I'd just check the pinouts.
Good job I did because instead of BCE they measured BEC.


** That is just not possible.

The middle leg is always the collector and metal tab used for
heatsinking.

FFS - get an ohmmeter and verify that simple fact.

Seems your dopey microprocessor ( ie software) based POS has gone ape
and confused Collector with Emitter.



I did suspect my dopey checker, but even with a fresh battery it ALWAYS
reads EXACTLY the same,


** ROTFL - since when is THAT any reason to trust it ?????

" It must be right cos it always tells the SAME lie "

Wot an idiot you are.


which is: 3 of them measure BEC with an Hfe of 5, the other one BCE with
an Hfe of 20.

** No fooling.

BJTs have very low Hfes if C and E are reversed.

Like 2 or 3 instead of 100 or 200.

They *also* show very low readings if the meter is no ****ing good.

See how that might be confusing ???


(All tabs connect to centre pin)


** Then there is simply **NO** problem with the devices.
-------------------------------------------------------------------

YOU are the problem with your dopey meter that is not able to tests power
BJTs properly.

FYI Try it with a TV horizontal output transistor like the BU208.


The replacements I finally got from Farnell with a different batch code
measure an Hfe between 40 and 70.


** Testing the Hfe of *power transistors* is a whole DIFFERENT ball game
to small signal devices.

That is why I designed a special device just for power devices -
published as a project in Electronics Australia magazine in May 1988.

Test are done at a constant Ic of 0.6 amps instead of some fixed and way
too small Ib.

Results are spot on and match maker's data which follow the same idea.

Works fine with TV horizontal output devices, power darlingtons and any
BJT rated for high power.



I know Hfe is kind of meaningless on these types of checker voltages, but
was still wondering why the checker would give a different pinout result
rather than fail the device.

** Cos it is using the wrong approach for power BJT devices.

I suspect all 4 of the "old" transistors are faulty,

** ********.

You have a ******ing useless** meter that only works with those examples
of power BJT devices that exceed maker's specs at the low end of the Ic v
Hfe range.

Take a LOOK at some ACTUAL power BJT data sheets some time !!!!!!!!!

Hfe is NOT a constant !!!!!!!!!!!!



.... Phil





Yes I know Hfe is not constant, that is why I said the reading is
meaningless on such a low voltage checker.

I do not use this device to measure Hfe, I don't usually give a **** what
the Hfe of a transistor is, I assume all new ones I buy from Farnell are
within manufacturers spec.
I usually just want to know if a transistor is bad, what pinout it is, and
whether or not it tests on this meter in the same ballpark as a known good
one.

These 4 "faulty" transistors do NOT test in the same ballpark, they probably
DO exceed the maker's spec if they work properly at all. In that respect
the tester has done all I have asked of it.

I am not going to put one of these devices into a Power Amp (which I was
about to do before I tested them) and have it blow up in my face. Farnell
can damn well send me 4 good transistors which ARE in the right ballpark,
which is exactly what they have done, one of which went into the Power Amp
which did not blow up in my face.



Cheers,


Gareth.



Some transistors work surprisingly well upside-down. Not well enough to
rely upon, but well enough to confuse people.

Why don't you measure the breakdown voltage when you reverse-bias the
base-emitter junction and then do the same with the base-collector junction.

One of these junctions will break down at a few volts, that will be the
one with the emitter. The other one will break down at some higher
voltage and will be the collector.

This test is slightly destructive in that it can permanently worsen the
noise figure (and probably other parameters) of low-noise small-signal
transistors, especially if you don't limit the current to a very small
value, but in this case I guess you don't want the suspect transistors
any more so you might as well do this test, and see whether your meter
is right or wrong.

Also, the tab is always the collector, unless the counterfeiters went to
extreme expense to re-package exotic RF devices as some bog standard
part and sell them for a huge loss, which is the reverse of the usual
business model. The bulk of the chip of silicon (the substrate) is the
collector, and this is stuck with conductive glue, or soldered, to the
tab. Therefore unless someone put a thermally-conducting,
electrically-insulating layer in between the chip and the tab, (which is
expensive and uncommon), the tab is the collector. This applies to both
NPN and PNP transistors.

It is easy to believe that measuring instruments with a digital display
are always right but sometimes they aren't.

Chris