sjep120r100 faking from China

[Antoinel]

Originally Posted by Tea-Bag

dont go by just the lettering differences. The original R100's and some R100As used the more 'white' letter printing. They also used a more 'shiny' legs and drain substrate backing.

That being said, I've had a couple of parts searchers seek these out lately. They are saying any offer right now are from seedy sources and likely fake. So really buyer beware.
The IC district of China can unfortuntely produce whatever you want for a part in printing in a couple of days notice.

R100's of any sort will have a much lower VgS test then any IRF240.

If the fakes are truly N channel JFETs, one may use them in a diy F6 set up which will be the appropriate and acid test method.

[flg]

As for the lettering, I've seen a lot of laser marked parts lately. When viewed from strait on, you see the white looking letters. When angled properly you can see the laser damage to the surface of the plastic (the writing) and no white? Very similar to what I'm seeing in these photo's.
As for the lead tin'ing, that is alsom a very process related variation. I don't think SemiSouth actually had any of their own processing facitlities. If both the processing and packaging/marking might be done at various facilities we might have several slightly different looking parts?

[qusp]

Originally Posted by Tea-Bag

dont go by just the lettering differences. The original R100's and some R100As used the more 'white' letter printing. They also used a more 'shiny' legs and drain substrate backing.

That being said, I've had a couple of parts searchers seek these out lately. They are saying any offer right now are from seedy sources and likely fake. So really buyer beware.
The IC district of China can unfortuntely produce whatever you want for a part in printing in a couple of days notice.

R100's of any sort will have a much lower VgS test then any IRF240.

well its not just the printing, the logo is wrong and too large/not crisp, the lettering is in courier in the wrong size, letter spacing etc, rather than an actual vector graphic that scales in proportion. its all wrong, not just the process

[flg]

Well, your making me want to go test the dozen or so I have...
Maybe, I'm not promissing notin, but, maybe I'll work up a few tests to check out a "good one".
Every one would like that I supose...

[flg]

Well, thinking about a few easy, non destructive types of tests to see if a few of the parameters are within spec on a possible Fake R100, I'll start with this. The Vgsth test is somewhat unique in value amongst the devices we typically play with. It is spec'd (Loosly) at 1V (typical) with 1Vds and 34mA Id.
Using the circuit described below, you should be able to apply approximately 4.4V at the +V node. Measure 3.4V across the 100 ohm resistor giving you 34mA Id and 1 volt across the g-s/d-s. The 1 volt is your gate to source threshold voltage. It is spec'd from .75V-1.25V however so don't get too excited if it's not 1V.
You could try this first with a IRF part but you would probably need to raise the expected voltage g-s to about 4V so you would need to apply approximately 7.4V to the +V node for an IRF device with an approx 4V threshold. Don't forget to lower the voltage to 4.4V for the R100 conditions.
I don't know of to many parts with a 1V threshold so if you can't pass this test I would be worried what I actually have???

[Tea-Bag]

flg, I should be able to test the real deal for this as proof of concept with some #'s.
Can I get away with a 1/4 resistor at this current and voltage? I can dig for higher power ones

[flg]

1/4 W should be good. You can limit your pwr supply current to 100mA or something safe and bring the +V up from 0. It might be a good idea to use a gate stopper R of a couple hundred ohms. it should not affect the test.

[Antoinel]

Originally Posted by flg

1/4 W should be good. You can limit your pwr supply current to 100mA or something safe and bring the +V up from 0. It might be a good idea to use a gate stopper R of a couple hundred ohms. it should not affect the test.

The distributor(s) of the alleged fake will need to be alerted once its identity is determined. What is its selling price?
How's one fake a high power semiconductor [TO-247 package]? Fabricating any semiconductor demands specialized equipment, clean rooms, and skilled personnel. Going through this exercise of faking does not make sense. It still may be the real deal; but is off spec from startup or prototyping.

[qusp]

you dont think they can make lumps of plastic and cheap metal for less than $40each in China? in fact now that they are unobtainium, probably worth more

besides many are reclaimed fets, sandblasted and reprinted with trendy high dollar part #

[buzzforb]

I have had almost 4000 fake fets come through my hands. Some in very official looking boxes. Venturing outside of known channels or purchasing agents will get in you trouble. I went toe to toe with big us EOL dealer who had fakes get into their chain. Said it could not happen...that is until the manufacturer told them they were incorrect. China has a bad reputation. I have no interest in being politically correct on the topic. There are known worldwide for stealing patents and making substandard goods, as well as fakes. My brother works for a major agricultural company that has to double check product that they buy from a contracted source. EVen thought they have done business for many years, they still try to slip crap through. Caveat Emptor. Trust me.