M17X R2 GTX 580M Overheating Issue

I have an M17X R2 that I have been restoring for the past year. It's kind of a Frankenstein monster of parts from eBay, but it all works fantastically. (1200p screen FTW!)

I decided to take it a step further and try to equip it with a 580m that I got for a decent price. I installed it and modified the drivers to make it work, and aside from the fan speed always at max, it seemed to work great.

However, any time I do anything remotely graphically intensive (games, movies, youtube), the computer shuts off unexpectedly and reboots. It'll happen as soon as a minute or as long as an hour into gaming.

I use HWINFO64 to get the temps, and monitor them while gaming in windowed mode, and the temps seem to idle around 69-72 degrees before it dies. I have 5870s in crossfire and they don't seem to have the issue.

Is there something special you have to do to get them to work properly? Has anyone had success with gaming on these cards? I've actually tried two different cards and they both have the same issue.

Any help would be appreciated. Big fan of NBR forums!

Thanks,

Matt

 

Hi man, nice project .

Unfortunately the 580m (and 675m which is an identical card) are some of the worst chips to come out of NV in a long time (IMO). I had one that did exactly the same as yours and always happened at around 72C. That was when dell sent me the machine in my sig. The 675m has already failed once and I'm not even OC'ing it!

Although this chip is supposed to operate up to 78C before it throttles down I have found that a good pad/paste/putty setup results in far lower temps. Mine games at around 65C with the fan going at high speed at that point and dropping it to 61C. With yours running full blast all the time I would expect lower numbers. When you do get the temps down you can look at using HWInfo to create your own fan table.

Did you get the x-bracket with the new GPU? The AMD core may be thicker and this could cause a poor contact?

Have a look here:

A real saga but should give you some good info: http://forum.notebookreview.com/ali...x-580m-78c-throttling-see-post-191-later.html

Good luck.

 

Hey, thanks for the reply! I read through most of the thread you referred me to, and it looks like I am missing quite a few thermal pads >.<

I guess the 580m cards need more thermal pads then the 5870m cards. I had assumed that it just needed pads on the ram modules. I have some coming in the mail soon, so hopefully this will solve the issue.

Thanks again!

 

You're very welcome. I thought it might be that. It's a tricky beast to get working anyway so any temp increases just shorten the life - I am seeing 580m's dying now on base machines - too old .

This is how mine arrived from dell with everything pre-applied. That putty area is important. This covers the VRM's and they are the hottest component. Now HWinfo shows their temp, even though my GPU core never exceeds 65C, the VRM's are above 90c!

View attachment 110516

 

Yikes, I hope I didn't damage anything already messing with it. I already put the 5870s back in for the interim.

Is there any padding that needs to be placed on the bottom of the card? Except for the ram modules of course. I saw some pictures of 580ms with thermal pads on the bottom underneath where VRMs are.

 

Nothing on mine, but maybe some benefit? Use the latest version of HWInfo and you can see the VRM temps to compare .

FYI, I was told the VRM's are rated at 105c but the problem is passing heat to local components

 

My end goal I think it to get two 580s and put them in SLI with my modded 330w PSU. I have HyperSLI at the ready lol.

I kinda got discouraged with the shutting off issue, but hopefully this works out. I'd love to push this R2 to its limits, but I'm almost to the edge of trying to sell it and buy an m18x. I'd hate to get rid of the 1200p screen though...

I'll report back in a few days with my results. With any luck this will work out and I'll have a working 580m. I only paid 30 bucks for the card. Thanks for all your help!

 

I wish you all the luck and wait for your results

I've seen a lot of people with the R2 1200p screen complaining that nothing else touches it for clarity and definition. I will probably check out the 120hz panel as this is not anti-glare but truelife ( Dell TrueLife™ LCD Screens - What is it?) if dell ever get the 3D problems sorted out....

You should keep an eye out for a 680m. These have proved reliable and overclockable - there might be a good reason for the $30 .

 

Ah very true. The card was bundled with another card that had burnt capacitors. I think the seller though both were shot, as they were both in the same machine. I just bought them on a whim hoping for a good find. I guess only time will tell.

 

Yeah, at that price it's gotta be worth a shot - good luck.

 

I would have bought it also! That price is very low. Let us know how it goes with the SLI.

 

Well, I received my thermal pads in the mail, and arraigned them the best I could to mimic how they are set up on the R3 heat sink. Sure enough, I am now able to play games without the computer shutting off.

I downloaded the latest HWINFO and it shows two VRM temps. The first one seems to get to about 68C-72C under load. The other stays at 47C no matter what I do. But I haven't noticed any significant FPS drops, or artifacting of any kind.

Thanks again for your advice. It seems that this card was a decent purchase after all!

 

Congratulations, result!

Mine failed again last night. First after a restart I got the 8 beeps. Reseated the card and then it started up but the display went black. Plugged in an external monitor and indeed it was working so I removed the video card again and this time used IPA swabs to clean the edge connector. It started OK but no output at all from the mini-port or HDMI now. Then it would suddenly go blank and the fans would ramp up. No response from the machine at all so force power down.

Now it just randomly happens and I still have no external output . Engineer with new MB and video card is due Monday. With one of my VRM's running over 90c most of the time could have been a life shortening problem

 

Ẁow, not again! If this keeps up, and your warranty is long enough, then aren't you going to be ending up with a brand new top of the range A17 due to the fact that you've got a 120Hz screen? Sounds like it could be a good result!

 

Yeah, unbelievable isn't it? This will be my fourth card - two in my R3 and two on this one! Of course you have to ask 'what am I doing?' I run a very mild OC and temps on the core are very good, never exceeds 65c. I also never use benchmarking programs any more (except Heaven which the dell guy wanted me to use for testing) - only in-game ones.

I think the engineer was thinking about a replacement as after coming up with MB and video card he went away to 'talk to some other experts' for 5 mins. I didn't want to mention it since this one is already a new replacement! I only have one month left on the warranty and I asked about an extension, strangely the guy said I couldn't until it was fixed

Any more problems like this and I'll start a push for a replacement (again!). At least I'm on first name terms with my on-site guy, and he is very familiar with the m17x....

 

Yep, good luck with that, sounds promising!