Fried my Alienware 18.

Came home from work and my laptop was off, tried turning it on and nothing happend, noticed the power adaptor light was off, tried taking it out and plugging it back in again and noticed small amounts of smoke coming out of the adaptor socket and a burning smell.

Opened up the laptop to see that one of my 980ms has fried.

There is a small patch of black melted plastic on the card itself and on the motherboard underneath the card.

Removed it and can't get the laptop to start, when I plug the adaptor in it shuts off the adaptor light immediately.

I take it that this is unrepairable?

 

Looks like you need (at least) a replacement motherboard and 1 x 980m. Strange to see the damage in two different areas like that, perhaps some pins on the 980m shorted out the two areas on the motherboard?

Certainly repairable, but whether its worthwhile from a cost perspective, unsure? If you end up not repairing it, consider selling me the other 980m

 

I think thats it. Board and GPU are damaged to a point where I doubt the repair costs would be cheaper than buying replaceparts.

 

Thanks for the help.

Luckily I just bought a desktop that I'm waiting on. Its a bit annoying though as I hadn't completely backed up everything off the laptop, I'm thinking I could take the hdds out the laptop and connect them up to the desktop and copy over the stuff I wanted.

I understand that I would need to get an adaptor to plug the msata ssd from the laptop into my desktops m2 slot?

Can I just connect the 2.5 inch hdds directly to my desktop?

Thanks.

 

if those are Msata SSD's then yes, youll need an adapter (msata to 2.5") and yes, the 2.5" are plug and play in desktop PC's.

 

Let me know if you want to get rid of the laptop.

 

lmao my alienware 18 blew 3 motherboards in exactly the same spot. Those power MOSFETs are absolute junk. Shame it took out your 980M with it, that's worth more than the motherboard.

 

I was using it for crypto mining so its probably my fault lol.

 

it's not your fault. I ran undervolted CPU and GPU and still blew up 3 times. it is poor quality MOSFET that is the problem.

 

I've personally never had your issue and I run a gtx 1070 on it with 150w tdp. Ironically the gtx 1070 died due to a crack in the die but the mobo is flawless and still working well even after 3 years of use and abuse.
Got my 4940mx qs overclocked to 4.2ghz too.
I bought another 2 alienware 18s not too long ago and they work well with gtx 1070s too.
Maybe you were getting unlucky.

 

Give it time, they'll blow up sooner or later.

I'd get out ASAP if I were you.

 

This is why I will likely never buy this model due to this design flaw, poor quality mosfets and no cooling never ends well. A heatsink mod might prevent this from happening again but most people wont put in that kind of effort especially when the clearance between the mosfets and the mxm card is so tight.

 

it's impossible to cool those mosfets, they are located immediately under the the 980M. There is a clearance of maybe 1-3mm.

 

It can be done and I have done similar mods like that before, also one can attempt to cool through reverse side of the board as well. As for impossible people have said the same thing about landing on the Moon and one can do a lot with 3mm, if all else fails use thermal pads and the 980m becomes the heatsink.

 

The only solution I culd think of, is using the copper back plate from the Clevo 980M, trim it to fit, and to place thermal pads in between the copper plate and the mosfets on the motherboard. However, this is passively cooled, and once the heatsink is saturated, it won't do much good.

 

You can make a vrm cooler on the mxm side of the board where all it would be is a flat piece of copper or aluminum that would extend out from underneath the mxm card then bend upwards for air flow while the fan side of the board should there be enough room and bare pcb you can add cooling there opposite of the mosfets to cool them that way. I have done mods like this to graphics cards and a couple of older laptops to get temps under control. People are often not aware the pcb can be taken advantage off for cooling purposes in a pinch due to all the copper provided there is space to apply something though thermal pads will always work. I don't do anything that harms the pcb and my mods are reversible.

 

980M components are already sensitive as it is. introducing more heat to the PCB surely travels to other components aka mosfets on the 980M, which get hot enough and already hard to cool sufficiently. But it's a personal decision, just be aware the heat all has to go somewhere and when everything is saturated, the fans and heatsink can only push out so much. I would abandon a faulty design and find something else instead of trying to correct what Dell should have done in the first place: remove those garbage mosfets and replace them with higher quality components.

 

I got some furnace cards that dump heat and learned how to deal with vrms like that, I will never trust the vrms on mxm Maxwell cards because often there is only one mosfet per phase and that screams being cheap in my book plus when vrms are setup like that they run Hot. I once had a 9800 GT that was stable at 116c while gaming no less without any issues because of the excessive modding back in the day. I only found out that it was that stable by accident lol. I really do miss the old school days of hardware modding and overclocking, I even went through the trouble of making sure that even the coils/chokes didn't get hot.