Replace 600V 18A MOSFET with IRF640 200V 18A for testing?

[lordvader88]

I have a broken Corsair CX 600W PSU (low quality desktop PC PSU for u under-powered laptop nerds that don't carry DT everywhere).

It has a PFC sector right after the mains EMI filtering. Whatever is wrong with this thing, the 2 PFC MOSFETs were not switching.

When this PSU is turned on with a 12V fan as load, it will spin up for 1-2 seconds, then the fault protections kick in and shut down the outputs.

I think the MOSFETS and diodes are fine on this thing, from checking with DMM. However I shorted the Gate to 168V Drain of Q1 in the PFC area, with my probe, so it's toast. I wonder if it was S to D would the MOSFET survive(Only the 2nd time I fried a MOSFET) ?

So it was a K18A60V 18A 600V MOSFET, with 168V DC on its drain. Since I'm not putting this thing under load besides a 12V fan..........popping in a IRF640 200V 18A MOSFET should be ok just to keep it running so I can findout out why it's not working ?

[lordvader88]

so whats the 600V rating mean ?

[lordvader88]

I'll have to wait a while then, its the closest thing.


But I don't think the PSU cares if I use it to power a 12V fan, so I suspect I could continue looking for the problem, with a working MOSFET

[Armadillo]

I'll have to wait a while then, its the closest thing.


But I don't think the PSU cares if I use it to power a 12V fan, so I suspect I could continue looking for the problem, with a working MOSFET

I read it as kind of plug and play, wait and disappoint scenes of things. I mean if you think that the mosfet was not firing, then the least you can do was to safely make some DC measurements and make sense of things, why its not firing up. example is there and how much DC bus voltage [at least].

But I reckon, it could also be thermal protection shutdown, which means something is shorted other than powering up the fan as load.

Just my opinion and 2 cents.

[IanMacdonald]

Just in case this hasn't been mentioned, PFC circuits are a boost converter placed before the reservoir capacitor. They work on the unfiltered DC side, and boost the cyclically varying input up to somewhat more than the peak AC voltage. Thus, at least a 400v rating likely IS necessary - it will be handling more than the peak line voltage. Preferably 600v rating as original.

There's a good writeup on how these things work on the Arch Electronics site.

[drussell]

...
So it was a K18A60V 18A 600V MOSFET, with 168V DC on its drain. Since I'm not putting this thing under load besides a 12V fan..........popping in a IRF640 200V 18A MOSFET should be ok just to keep it running so I can findout out why it's not working ?

No!

The only reason the voltage is only 168 volts is because the PFC isn't working, which is probably why your supply is trying to start, sensing the under-voltage and then shutting down.  When it IS working, the voltage will be more like 400 volts, blowing the magic smoke out of your 200 volt MOSFET.  They used a 600 volt part for a reason! 

[lordvader88]

Well the main capacitor the is charged by the PFC inductor is only rated 400V. But yeah IDK what its voltage is under normal working conditions.

BUt hey I'm going to try aN IRF640 anyways, just so I can fire it up and look for the original problem. The problem is after the PFC (unless its the chip but probably not)

I haven't been able to look at what any of the feedback loops are telling the chips.

This PSU is known to have poor caps, but for fun and education I want to try and find the problem.

I think I fried a 4R7 SMD too. I need to buy some basic SMD parts, I don't have any.

[lordvader88]

I don't have a schematic or really know SMPS feedback and monitoring.

-Most SMD parts on the back seem to test fine in circuit anyway.
-the PFC MOSFETs, driver Mossies switcher mossies, and the diodes before the secondary side seem to test fine

I haven't used my little 200Khz DSO138 scope on it yet, I just zapped that K18A60V.

The PSU just reverts to standby mode after about 2-3seconds, with the 5VSB working, and the PFC being told to shutdown.

The main PWM chip, its getting power


After a failed start and it idles, I have to wait a minute for some cap to discharge before it will restart the same....where it will power a 12V fan for 2-3seconds. I never probed that yet either.

If some big capacitor on the secondary side failed or is failing , what should I be checking ?

[Armadillo]

Is the power supply connected to your motherboard or load?

If not, did you bridge the connector to signal a start command in the first place?

Since you blew some mosfets and components already, better use the incandescent light bulb in series to do your tests.

[lordvader88]

NO I popped the MOSFET testing it, just with the DMM probe, bad angle/lack of light