Electronics > Repair
Real or fake IGBT FGL40N120?
<< < (6/6)
T3sl4co1l:
You may find this of interest:
https://www.eevblog.com/forum/projects/diode-recovery-tester/msg1314881/#msg1314881

A high voltage power supply may be nontrivial, but at a low repeat rate, not much current is needed at least, easily sourced from an electrolytic capacitor.

The type of test they use is the same double-pulse test which my circuit performs; see e.g.
https://www.infineon.com/dgdl/Infineon-Double_pulse_testing-Bodos_power_systems-Article-v01_00-EN.pdf?fileId=5546d46271bf4f920171ee81ad6c4a1f
https://download.tek.com/document/75W-61623-2_AFG_Oscilloscope_Double_Pulse_Test_App_Note.pdf
https://www.iea.lth.se/kel/2021/Lectures/PE20_GuestLecture%20OnDoublePulseTesting_Philip.pdf
etc.

Switching speeds will be roughly similar at say 100V 10A, far from precise, but also representative enough.

BTW, make sure to check everything around them; blown transistors easily take out gate drivers or controller ICs, perhaps even on the far side of a transformer (if they're using a gate drive transformer instead of isolated driver circuits).  Likely something secondary side failed, causing a cascade including primary side transistors, fuse, etc.

Tim
masterx81:
Whoa, nice project! Way an overkill, both for my use (i not "play" too much with electronics, not much as years ago, i repair sometimes something for my use), and an overkill for testing a chinesium igbt 😅
In any case, i've found something nice to read about. It's an argument that i've never explored.

For the repair process, luckily it have an optoisolated driver (A3120) and only few components on the gate, already all measured. Had to replace a resistor and both the drivers. Also found some failed semiconductors and electrolitic caps on other boards. I not know the past of this ups, i've bought it time ago quite cheap as used/working, but never tested. In this days i've tested it, and got the surprise. It's identical to one that i already own, it's a "backup" unit in case mine stop working. The working one is also a repaired item, that had way less damage on it. They use the same battery, that is an home made lfp package... They are 5000va units, i have all the sensitive loads of the house protected, and also the well water pump (so that i not risk water in the basement in case of blackout).
boB:

First I would check to make sure it is even a  transistor.

Then, if you have several of these, I would cut the package open to look the die size or just run it in circuit to make sure it doesn't heat up excessively compared to the real one. 

boB
masterx81:
So, as i was curious to know if throw away the fakes, i decided to use the belt sander+dremel and diamond discs.
And now, the truth..


On the left a fairchild genuine exploded igbt, on the center a fairchild fake (the one of the previous photos), on the right an aliexpress item, marked as "original", with really good onsemi markings and appearence (compared to one onsemi bought from lcsc, hope genuine... they where quite pricey).
The center one isn't that bad as die size, but definetely not genuine, the one on the right, maybe will handle half the power of the genuine...
A thing that i've noticed is that the genuine one had softer plastic casing, the fake ones had a really strong plastic.
A genuine one from lcsc on cap meter (de5000) measured around 4nf between gate-source. The center one quite near (4.9nf if i well remember), the one on the right 1.9nf. I not know if mean something che gate soirce capacitance, bat was easy to measure.
Long story short... get them from good sources. The market have more fakes than genuine items.
Navigation
Message Index
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod