ECS K7S5A mosfet issue

**retiredcaps** · 06-25-2010, 12:57 AM

Re: ECS K7S5A mosfet issue

I'm not an expert, but that could be leftover flux residue from the factory soldering process on those chips?

**370forlife** · 06-25-2010, 01:03 AM

Re: ECS K7S5A mosfet issue

That definitely just looks like flux residue.

**kaniki** · 06-25-2010, 01:40 AM

Re: ECS K7S5A mosfet issue

i cant say that is is flux residue, but i know i have seen boards that looked like that in the past with no issues and ran for years later.. not something to worry about.. except for those swollen caps that i can see.

**Per Hansson** · 06-25-2010, 01:46 AM

Re: ECS K7S5A mosfet issue

In any case when the caps fail (like yours have) the Mosfets have to work much harder.
So they become allot hotter (there are even examples on this site where the Mosfets have desoldered themselves)

So either they have become really hot or it's just leftover flux residue from the factory, in any case with new caps the board should work fine

**Dammon** · 06-25-2010, 04:32 AM

Re: ECS K7S5A mosfet issue

Thanks, guys. I'm going to go ahead and install the new caps. I'll post some pics later when I get it to work.

**retiredcaps** · 06-25-2010, 10:17 AM

Re: ECS K7S5A mosfet issue

Since the consensus is that it is flux residue, you may want to clean that off using alcohol. I believe flux, over time, will eat away at the board (I think I read that somewhere). A q-tip and isopropyl should do the job.

Originally posted by Dammon

Thanks, guys. I'm going to go ahead and install the new caps. I'll post some pics later when I get it to work.

**seanc** · 06-25-2010, 12:48 PM

Re: ECS K7S5A mosfet issue

Test them with the diode function on your multimeter.

I re-capped an Intel board with failed Rubycons (2004 date range) - one of the mosfets shows a dead short in circuit (all show dark flux residue on the mosfets like yours), board won't power up, I may try and find a replacement, but I ended up substituting another socket 478 board.

**Toasty** · 06-25-2010, 01:32 PM

Re: ECS K7S5A mosfet issue

Clean up any of the "solder balls" such as those around Q16. Those balls sure make it appear that it overheated, but hopefully they weren't destroyed.

DigiKey has these, but you'll need to cut & bend the leads (L package vs. S package):

http://search.digikey.com/scripts/Dk...IRL3103LPBF-ND

Datasheet:

https://cdn.badcaps-static.com/pdfs/...556264df4f.pdf

Toast

**Scenic** · 06-25-2010, 02:19 PM

Re: ECS K7S5A mosfet issue

most of the time when i see surface mount mosfets with bubbly solder and blue-ish discolored (and sort-of dull) metal parts those are dead shorted from overheating.

check them with the diode test on your multimeter as seanc said.
ohm/beep test may also work to find shorted ones.
*carefully* lifting the 2 pins up from the pads prevents false readings from somewhere else in the circuit.

if one of them is in fact shorted, replace all of them at once. i did that a couple times salvaging similar spec'd mosfets from otherwise damaged and unfixable boards.
just don't mix up different types (ex.: 4x identical mosfet replaced with 2x mosfet ABC and 2x mosfet XYZ)

worst case scenario.. (was still fixable though)

**Dammon** · 07-24-2010, 11:01 AM

Re: ECS K7S5A mosfet issue

Sorry for the late update, I had to wait for more caps to be shipped.

I didn't replace all the crap caps, just the ones that had swollen tops. Didn't see the point in recapping the whole board with good caps since it would only see infrequent use. There were 11 swollen caps: 9x 2200uf 6.3v, 1x 1000uf 10v, and 1x 470uf 16v.

Was a little worried when I powered it on because the soldering iron slipped once or twice on the motherboard.

Fortunately, it still works.

Ran through memtest and Prime95 for a couple hours with no problem. Prior to recapping, the board would hang in less than 10 seconds.

It's quite a thrill to see a board that I had written off as dead many years ago come back to life.

Attached Files