Alienware 17 R1 (2013) won't turn on, power cable light goes out when plugged in

I was working on Photoshop and not doing anything particularly hardware intensive when my computer suddenly just completely shuts off without warning--everything went dark, including the power cable light. Replugging the power cable from the wall makes it light up again, but when I plug it back into my computer, it instantly goes dark.

Video of the problem: https://vine.co/v/eIDJirVhILW

Tried multiple power cables from different Alienware models, all exhibit same behavior. Power cables still work when plugged into an Alienware M17x-R1 from 2009 that's been mostly stripped for parts.

Alienware 17 R1 (2013)
Windows 10 Pro (64-bit)
Core i7 4900MQ 3.7Ghz
Nvidia GeForce GTX 780m
32GB RAM (8x4?)
256GB SSD + 1.5 TB HDD + 750GB HDD
120Hz 1920x1080 screen

Was plugged into a 4K and 2K monitor, keyboard, mouse, speakers, and (through USB3 hubs) about 10 external storage devices.

Got this laptop near the end of 2013, been running the Win10 TP since its first release, upgraded to Win10 Pro on July 29th, just did a clean reinstall yesterday. Warranty expired.

Worth noting that, for the last four months or so, the battery has completely failed to hold a charge. Also, 4K over miniDP would often have screen tearing and lost signal, while 2K over HDMI had constant flickering and thin bands of noise.

-----

If anyone can help me find a solution, I'd really appreciate it. I do VFX work and this laptop is basically my lifeblood. I have an upcoming deadline in about 5 days.

 

You have a short in your laptop. Do not try to turn the machine back on. Disconnect the battery.
If you are feeling adventurous, try connecting the cable once the battery is disconnected, but understand that this could potentially damage your computer.

*IF* it starts up without the battery, it should be just a case of getting a new one. Otherwise, you will have to do some additional troubleshooting in the sense of starting to replace various parts in the hope of finding the one that is causing the short.

I'm fairly sure that will take longer than 5 days though.

If you do decide to connect the power cable again, take out the hard drive so it doesn't get damaged (further than it may already be at this point).

 

Did this. Tl;dr: It's the GPU.

I pulled drives and everything out the back, then started putting components in one by one and testing with the plug. Added RAM, WiFi card, optical drive, CPU, CPU fan with no trouble--power adapter light stayed on, computer lit up when I hit the power button.

Plugged the GPU, heatsink, and GPU fan back in. Pluhged in power, light went out instantly.

Goodbye 780m, I guess...

Thanks so much for the advice--it feels so much better knowing that the problem appears to be isolated to a single component (even if it is probably the most expensive one...).

What would you advise I do now? Is it safe to add the drives back in? How would I best go about replacing the GPU? Would it be unwise to use the computer without a GPU for light casual use while I await a replacement?

Edit: Tried booting the computer with everything except the GPU and HDDs. I'm getting no boot screen, so I'm assuming that using it without a GPU is out of the question...

 

You are using the 120Hz screen, so the internal GPU is disabled. The only option is source a gtx780m or gtx880m and plug that in. It should be plug and play. (You can also get a gtx980m or gtx970m, but not sure in what timeframe you can get that up and running)

You could try calling Dell and seeing whether they have a gtx880m available. You could also see whether @woodzstack has a 780m or 880m available, and at when he can get it to you.

See http://forum.notebookreview.com/thr...0m-8gb-mxm-3-0b-videocard-for-laptops.778828/ for @woodzstack's for sale thread.

I am unsure what other retailers are near you, or which can get you what you need on short notice.

And yes, it should be safe to add the HDD's back in.

 

I have all videocards available, but they are all NEW.
I do have 1 person who has a 780M they say works(Used), they want $400CAD for it.
I will not take responsibility for used hardware unless it was my own at one time.

Anyways - keep in mind when I sell new, it is NOT pulled from a laptop. Vendors like HIDevolution or XoticPC they sell USED cards as new. They are technically refurbished.

That is why the quality of the cards they get are the lower end. Just like OEM processors, you get better quality from the ones you buy individually rather then then ones included in a computer.

Anyways, what I am getting at is, its usually not worth buying older cards from me, the 680M had such demand not too long ago, it's price cost more then the 980M. Its not me that sets these prices.
But, a 680M for instance was an expensive card to begin with. These cards - unlike the myth everyone thinks- never go down in price. Year after year people fail to realize that. Its used/broken/second hand stuff that goes down in price.
With the US dollar going steadily up, I have warned people who tell me they will wait for a price decrease. When we were selling them for 700CAD, I said buy them now.. Now that is nearly 1,000 almost a year later, I did warn people.
But lower end stuff like 860M or 965M are practically giveaway's. I had a few lying around and threw them out because they couldn't sell. Course vendors keep trying to pawn them off on me.
I have even had people just give them to me because they couldn't sell them either.

 

Your lost screen tearing is because your over your bandwidth probably. You would for instance get the same thing trying a single link or let me abbreviate DVI-D to DP and trying 2560x1600(4MP) or 4k(8MP) even. It will work, but its not going to have enough picture for every frame. If for instance your losing some signal somewhere, your bandwidth would lower, like Noise CAN theoretically cause interfernce if its a cheap DP cable. Other types of cables usually don't have this issue - though I think HDMI did too, back in the day. like pre-1.3 version.

I would check the thing used to connect to the 4K from your alienware. If I had to guess your using a MiniDP converter. It might have limitations. It would definitely need more power, it is basically just a thunderbolt plug/miniDP, it would need more juice to do 4K, wheras it can technically do 4K cause its a 1.2DP standard, its the miniDP version. Has less power. I use the same miniDP and have had to find all kind of things to get it to work for say triple monitor setups, or for 4K. You can get a MiniDP MST hub from startech for instance, that is powered, and can hold three full DP cables, and up to two of them can even still do 60Hz

 

I would recommend just upgrading to a newer 970M or 980M.

I can vouch for @woodzstack , bought a GTX 970M for my AW17 R1, no issues whatsoever.

 

Sold hundreds to people on these forums, I forget more people I've sold to then I remember, lol. But thank you !

The 970M is a favorite of mine - if you are NOT going to use an external monitor and your upgrading an older laptop - and just going to use what LCD is built in, then don't go nuts and get a 980M. The 970M has less compability issues then the 980M did at first, it takes up less power, it fits more heatsinks, and can use either a duel or triple pipe(overkill) and will not rob the system of as much power as a 980M will - yet still powerful like a 780M would be (even more so, seriously, no joke!) So, if you have an extremeCPU or could like to hold those turbo clocks longer - you can actually up the wattage abit and such and get abit of extra performance whereas you wouldn't have otherwise.

Another thing : USb's FAns' Lights , external monitors and the connections the make even.. the speakers whatever these all take power.. If you put 4 HD's 4 Sodimms of RAM have all lights on , a 120Hz screen 980M with extreme cpu and all USB plugs are connected to something, even including hubs, mouse keyboards all external... etc..

You will get such a crappy performance, because all your Turbo states are being robbed of performance, most of your clock cycles are down clocked, there will be so many times your USB's stop functioning, or your sound card stops working or your external monitor flickers off and disconnects... or an HD has an error or an issue happens with the RAM and you end up with a BSOD... all because not having enough power. The 970M eliminated like half of that - just becauise it is saving you like 25-30W's easily !

You only gain small improvements from going to a 330W PSU when you had a 240W before hand. There is a thing called vDroop, which effects load on components inside the laptop. When a load is generated, based on efficiency, when the current moves somewhere, say when a game loads up, there is like this split second vacuum that will cause an unhealthy spike in voltage and amperage across the entire system. Over time, this wears out your components faster, and can even cause them to fail, sooner, if not right away. The capacitors designed for these systems were only intended to hold a charge based on a certain amount of power. Increasing that power speeds things up before they release the charge, and create the logic high or 1/0's that they do, this forces many other components to do more work, faster and soon, and since some components do not benifit from extra capacity - they can instead be damaged by it, instead. So, really, a 980M and extreme cpu and a 240W psu as an example - can limit what you can use with the laptop. its a balance of power. Clevo's have this same issue, so does MSI, and DELL.

 

As for your issue, some of us have had the same issue. I do not know what caused it, but a bios reset fixed it once for me. I mean a hard reset, and totally discharging all power. Then just plugging it back in, and things worked again. This happened to me when i tried using a different adapter, i firth thought it was the adapter, but I have since ruled it out, and also never had the issue again, and well, I also sold my AW17 like last November.

 

So since its 120Hz, you should just triple check there is nothing contaminating contact with the gpu and the motherboard, be careful and re-install the CPU, and reset it all (should auto reset anyways if the CPU was removed,and you didn't full discharge it regardless)
So here is a funny thing, Put the GPU in first, then put back the CMOS battery then put back the CPU then put back everything else, including battery and powercable, see if that makes a difference, see if you can go to bios, if you can restart and do a full/long diagnostic. Don't go into windows. This should test harddrives/ram/cpu/gpu etc.. you shouldf get a slight error somewhere if there is an issue, that diagnostic is fairly decent, even if not very detailed, i have always noticed it picking up everything though.

 

it also tests battery and PSU

 

so that's 2 things i need to buy: a pair of torx screws and a dremel

 

Alright, so how long and how much would it take to get a replacement 970m?

 

@DivideByIchi: http://www.ceg-hardcorecustom.com/product/ceg_nvidia_gtx_880m could be an alternative with a short turn-around time as well. They are more expensive than @woodzstack though.

 

I bought mine from rjtech, they are super awesome.
They shipped mine on the same day.
Card was brand new and overclocks like a beast.
I'm currently using it on my m17x r4 3d.
good luck