My Motherboard Released the Magic Blue Smoke!!! Update: It Lives!!!

[iamdarkyoshi]

Hello everyone. My gaming computer released the magic blue smoke today when I plugged in a faulty DVI cable into the onboard graphics. This cable had given me troubles before when I used it on my normal graphics cards, but the cards just simply shut off instead of exploding things. I decided to mark the cable as defective using the Photonicinduction method that way I wouldnt mix it up with my other 50 spare cables.

Anyway, when the cable was plugged in, after about 10 seconds I smelled the smell that we are all familiar with. However, the PC kept going. My eyefinity monitors weren't disturbed, and the audio I was playing continued to play. I shut the machine down immediately and unplugged the cable once I smelled smoke.

Is this machine still safe to operate? Or should I desolder the blown component? I have had SMD soldering experience. If I desolder anything, what should replace it? The failed component appears to be D8, a diode?  I have some low-tech marco photos of the only visibly failed part as well as the annihilated cable in a .zip file, and they can be found full size here: https://drive.google.com/file/d/0B-LjUeh-cirIeVJTbEYwa1pUMnc/view?usp=sharing
Motherboard is an Asrock Z77 Extreme4 if anyone has the same board.
Thanks!!

[romantronixlab]

Picture would be best to attached to the post for later reference.  Was not able to open zip.
The diode most times is not just what could be gone, check also for regulators and resistors near the dead Diode. and component.

[iamdarkyoshi]

Sorry about the .zip, I did not want the photos compressed.

[iamdarkyoshi]

Images can be found as direct uploads here:

Sorry, the forum wants 1mb max per file, and 2mb total. My photos are about 2.5mb each.

[Grapsus]

You can fit a lot of pictures in 2 MB, the trick is to crop, resize and use 80 as JPEG compression factor, not 95.

Did you test the diode with a multimeter ? Is it internally shorted or open ?

[iamdarkyoshi]

I have not. A bit later today I will remove the board and get out my multicore desoldering braid and electrolube flux and removs the part. It must not be critical considering that it kept running...

[JBaughb]

I have not. A bit later today I will remove the board and get out my multicore desoldering braid and electrolube flux and removs the part. It must not be critical considering that it kept running...

Im not sure if thats the correct way of looking at it...

[iamdarkyoshi]

I have not. A bit later today I will remove the board and get out my multicore desoldering braid and electrolube flux and removs the part. It must not be critical considering that it kept running...

Im not sure if thats the correct way of looking at it...

Ya probably not... I do think that the onboard graphics are nonfunctional now because after the smokd it played the "hardware unplug" sound.

[iamdarkyoshi]

Now I think that the component was an "LNW 24" as there are many elsewhere, but the board designators are inconsistent. See photos

[SL4P]

D8 is unlikely to be the same component as Q anything...
Look for another D in the same package - and you MAY be half way there - but not for sure.
Ideally find someone with another board and check a known uncooked part

[iamdarkyoshi]

D8 is unlikely to be the same component as Q anything...
Look for another D in the same package - and you MAY be half way there - but not for sure.
Ideally find someone with another board and check a known uncooked part

I have found someone on the linustechtip forums. Awaiting response.

[Rasz]

why not use continuity tester and check pins of this component against dvi pins? it looks like a clamping diode, maybe on +5v DVI provides?

[rob77]

why not use continuity tester and check pins of this component against dvi pins? it looks like a clamping diode, maybe on +5v DVI provides?

+1

actually it looks like a dual diode in sot23 - and apparently only one side is blown.

[bktemp]

LNW = 2N7002 = N-channel mosfet
This matches the Qxx on the board.

[macboy]

Now I think that the component was an "LNW 24" as there are many elsewhere, but the board designators are inconsistent. See photos

Highly unlikely. "Q" indicates a transistor, in this case a MOSFET. "D" indicates a diode. Either can be in a SOT package.

[iamdarkyoshi]

Now I think that the component was an "LNW 24" as there are many elsewhere, but the board designators are inconsistent. See photos

Highly unlikely. "Q" indicates a transistor, in this case a MOSFET. "D" indicates a diode. Either can be in a SOT package.

Which is what made me skeptical. However, the blown part has a clearly visible L, a somewhat visible N, and possibly a fainly visible W? Also, the board's designators are hard to understand. I cant find anything else in the area that could be a diode though. I am in touch with another owner of the same board, and he is going to get me some photos of his.

[senso]

I can tell you the obvious, PQ is Power Transistor/Mosfet, SQ is Signal Transistor/Mosfet.

[macboy]

I am in touch with another owner of the same board, and he is going to get me some photos of his.

That is the best course of action. No guessing.

[iamdarkyoshi]

I am in touch with another owner of the same board, and he is going to get me some photos of his.

That is the best course of action. No guessing.

Or perhaps sparkfun's "magic blue smoke refil kit"

[iamdarkyoshi]

Alright everyone, got some details from the Asrock forums:
parsec wrote:
"
OMGosh, a photo of that part would require a macro lens on a digital SLR camera.

Using a high power magnifying glass and two lights I was able to see the markings, clearly after several attempts. It is:

LNW 22

The 22 is rotated 90 degrees counter clockwise relative to the LNW. This part is labeled D8 on the board, so apparently is a diode.

You'll need better luck than I had reading that part, trying to replace it IMO. You must have experience with surface mount devices as well as the tools to work with them.

Good luck with this project!
"

So that is what we have for information. Any chance I would find this same part on another older scrap motherboard? How does one read the marking on these parts?

[iamdarkyoshi]

It lives!! Both the onboard and the crossfired graphics cards work! I used a mosfet that had similar specs stolen from a scrap Pentium 4 mobo , and aside from a different number, it looks new. Thanks everyone. Now to get back to CAD work. And games 

[gildasd]

It lives!! Both the onboard and the crossfired graphics cards work! I used a mosfet that had similar specs stolen from a scrap Pentium 4 mobo , and aside from a different number, it looks new. Thanks everyone. Now to get back to CAD work. And games 

Cool! Job well done!

[jaycee]

If this was on the DVI output pins, it was likely a dual diode wired up as a "catch" diode.. one to +v and one to ground. It's likely your faulty cable caused a short to a higher supply. You can also get this happening if you have no power ground and a chassis is floating at half mains.