Need help with serious M15x problem.

Hey I just found this forum a day ago and I really like the looks of it. You guys have some really great info here.

My M15X was at Alienware UK for a hinge replacement recently and about a week after I got it back I started having a freezing problem once in a while. After a couple weeks the problem began to be extremely frequent to the point where the computer is practically useless right now. However all they did at the last service was flash the bios, replace the cover, and run 3d mark 06 according to the rep I spoke with, so they won't take any responsibility for this now out of warranty machine.

Originally the screen froze (all to one color and sound started skipping) while I was playing WoW. Then a week later again, a few days later again, a few less days again, until now where the computer will stay on for only a few min (less than 10) before freezing NO MATTER what program I'm using (including BIOS and a diagnostics CD). I can only get away from the blank, frozen screen by manually shutting off the power.

If I try to turn the computer back on directly after the problem occurs in a significantly shorter time (now being before windows even boots). If I let the computer sit for a while I have longer before the problem occurs (making me think something is overheating... but what?) I have been able to us HWmonitor to check temps for the few minutes it runs and everything seems pretty normal except my acpi temp (could someone tell me what this is reading on the M15x?) which is always reading 80 o C no matter what. I noticed on someone else's screenshot using the same program on an M15x that they had a proper reading for their acpi temp that fluctuated with load like it should.

A couple times I was monitoring temps when the computer froze, and there was no consistent temp of cpu cores, gpu core, or hdd that made it freeze - and they were all within normal ranges when it happened.

I was thinking my next step would be to reseat all of the components that I could and if that didn't work I would try to hook it up to an external display using minimal components to try and narrow something down. I bought some thermal compound already, but read somewhere that I shouldn't replace the Alienware thermal stuff - Would any of you suggest that I hold off on this step, because I don't feel like I have many options at this point. Not within my budget anyways. I really appreciate any and all help/advice I can get here. Sorry about the long post and please let me know if I missed any important info.

*Oh, almost forgot to mention that I put back in the original 2gigs of ram to make sure that wasn't the problem, and I also checked each bank with each stick to make sure it's not a ram issue. I've also made it through a couple diagnostic tests including video and motherboard with nothing coming up. I have only gotten half way through a hdd test but nothing showed up during that time either. I'm also pretty sure dust isn't the problem.

 

I have a few ideas. When you turn on your system. Hit Stealth mode see if the problem stops. The cpu and intergraded gpu are on one fan. The other gpu has it own fan. This might help detecting the location of the issue.

 

Looks like you are going to need to run Stress Prime 2004 next:
http://sp2004.fre3.com/

Also, how did you test the integrity of your different RAM sticks and slots?
http://www.memtest.org/

Finally do you remember what BIOS version you were running on prior to the last depot, and which are you running on right now?

 

Hey Atol: currently I'm running my computer in stealth mode and it has been running half an hour or more so far (just running CPUID HWmonitor and Speedfan) without crashing yet. This is a record for my computer this last week. What all does stealth mode do exactly?

Sneaky: I didn't actually run a memory test (though I do have memtest on my bootcd that I can try). What I did do was first change my 4gb ram back to my original 2gb ram and the problem persisted. Then I took out one stick and ran it with one stick - problem persisted. I used the same stick in the other bank - problem persisted. Then I used the other stick in both banks - no change, problem persisted. I don't think all 4 sticks/both banks are messed up, but I will still run the memtest to be sure. I will also try running the stress test. I will let you know my results as soon as I get a chance.

*Edit: I am slightly ashamed to say I don't know what version my original bios was and I can check my current one as soon as I'm done with the stress test that is currently running in stealth mode.

 

You have 2 Grapics card on the M15X. It sounds like your main grapics card or grapics card fan not working right.

What stealth Mode dose is drop your computer in a power saving mode and turns off your main grapics card and use the intergraded grapics card.

 

If you're not sure if the problem is hardware related, you can always try a clean re-install of your OS (reformatting the HD) before you send off your system for repair or open it up. Can't hurt-just takes a little time.

 

try one GPU external (the gamers one)
and one onboard (the net surfer one)

while you may have meant this, your wording seems to lead he has two GPU slots which is incorrect.

 

ok... Here' an update: yesterday when I tried going into stealthmode right after getting into windows, it worked... It ran for 8 1/2 hrs before I shut it off to go to bed. 4 1/2 of those, it was running the Orthos stress test and passing everything. Temps on the cpu cores were fine and I even ran WoW for a bit (thats where the problem initially occured) before running the stress test. I never got my gpu over 72 o C the whole night and mostly it would get up to like 66 then the fan would take it back to 55. During the stress test it would stay at around 57. This would all make sense if stealthmode switched it to the integrated 965 express, but I was under the impression that u had to press fn + F7 to get it to switch cards and that all stealthmode did was underclock the gpu and cpu.

Anyways, today I tried to start up to do that memtest I was advised to, and the computer froze almost immediately. So, I tried to start windows to switch to the integrated card, but couldn't even get to windows before it froze... Anyways, I ended up taking my 8800m GTX all the way out and started up my computer... It's running fine. So I'd say it's very very likely that it's my gpu... But is it fixable?

Oh! Here's another funny thing. Since my card is all the way out of the laptop, speedfan and HWmonitor aren't picking up any gpu temp. Why then was my gpu temp fluctuating yesterday (according to load) when I was in stealth mode? If stealth mode really only decreases the 8800's clock speed what would that mean in terms of my experience of stealthmode working? Some fan was working to decrease the gpu temp which was reading as my 8800...

I hope this new info helps u help me

Oh, and sneaky, my current BIOS version is vX36 P3A37.

 

Ok, here's another support to my last post. I loaded up WoW with my 8800m gtx still out of the laptop and unlike yesterday in stealth mode where I had up to 60fps, with the same and then even lower settings now I'm getting 11-18 fps.

Soo, my question now is why would my 8800 work in stealth mode, but not out of stealth mode?

*Edit: Also I still wonder if it's something other than overheating because I have seen it freeze before when my gpu temp. readings were still well within normal bounds.

 

What are you running as far as graphics drivers? Also, have you tried doing a clean driver reinstall by using Driver Sweeper? That may take care of your problems right there. Ive been a lil out of the loop with versions, Im sure someone around here with a m15x will tell you which one works best for them.

Also, have you tried replacing the thermal paste that you mentioned you purchased? Just follow the instructions you can find around the web, do a clean job and there wont be any problems, but it might help clear out your freezes.

Please keep us posted.

 

sneakyFspy: No I haven't replaced the thermal compound yet. I'm getting some alcohol today to clean off the old stuff. I've checked out some instructions already and just need to do it after I get the alcohol.

I'm not sure what driver they had installed when I got my computer back, but shortly after the problem started occuring, I tried the latest driver from Nvidia.com 185.85 I believe, but that didn't work at all. I got the screen flashing (like it kept trying to reload the res) except for when I alt+ctrl+del and opened task manager, then it stopped... Anyways whenever I tried to change the res after doing that it happened, so I installed the one from Alienware.com. It was the 175.80 beta driver. Actually when I installed this one, it seemed to work well again for a while, then started acting up again after a couple days. I suppose it's very possible this is the same driver that was already on the computer...

I didn't use Driver Sweeper though, and I didn't even uninstall the previous drivers before installing the new ones. I will look to see what people around here recommend for drivers, thanks for the idea. I will keep updating as I try things.

 

Big no-no. Look no further, you really need to run a sweep and follow the "e-wrecked steps". I think you will be fine after a clean driver install.

 

Good to know - I've still got a ton to learn. I'm downloading some new drivers and driver sweeper as I type. Is it possible for me to uninstall the old drivers, clean, and install new drivers while my 8800 is out of the computer? Or does it need to be in while I install the new drivers. Also looking through my Programs and Features window, the only thing I see that I think could be my video card driver is only called Nvidia Drivers... Also there's Nvidia PhysX - does that need to be uninstalled too?

Would a driver problem also explain why it was freezing in Bios and my bootcd testing programs and not just after windows loaded?

 

Well Uninstalled the Nvidia Drivers, then shut off my computer. Put the 8800m gtx back in and started back up in safe mode. Got to the point where I put my password in, and right after that the screen went all that blue/teal color and froze yet again. So I can't even get into windows to run driver sweeper and install the new driver Maybe I'll let it sit for a while and see if I can get into windows in a bit. Anything I can do in the meantime?

 

Read up on this procedure in the meantime and try again in a while.

http://en.community.dell.com/forums/t/19261160.aspx

 

definitely the gfx.

"Stealth Mode" - downclocks GPU, CPU, etc.
"Binary GFX" - Fn + F7, changes to the integrated graphics.

Start in Binary GFX using right-click, by holding the right click button, while the system is off, until the system turns on. Warning - sometimes this mode is hard to shut-off... it will require a full power off, *small delay* then power on via power button.

definitely does sound like the graphics no matter what, I would like you to see this KB article about the proper procedure to install the heatsink and how to apply the thermal paste to the gpu.
http://4help.alienware.com/cgi-bin/alienware.cfg/php/enduser/std_adp.php?p_faqid=5670

Please ensure it's being done right.

if you're going to run memtest, install both banks of RAM, start on integrated graphics (use right-click method if necessary), and run through the RAM... since memtest has nothing to do with the GPU, it'll run fine on integrated. Be sure to test both setups for memory. (2G and 4G)

carefully reinstall your graphics card and try to boot up, in stealth mode. safe mode won't really have much effect... only use it if you're cleaning drivers or something.

after doing a stealth-mode start-up and driver install, (after all necessary reboots), get something like ATItool's Graphics test going (an alternative is furmark in stress testing mode), and run that, non-stealth. it should crash the system if the GPU is the culprit.

if you can't successfully run atitool/furmark for an extended period of time on your external (nVidia) graphics, without crashing, contact Alienware. you're going to need a new graphics module. if you're out of warranty, the last I checked, the 9800M GT was under $400 direct from AW.

and for the performance... less than $400 is pretty cheap for the repair.

 

So, I was able to be in windows long enough to get the driver installed mostly properly. I didn't uninstall the PhysX though, so I'll have to try again in a while when I can get into windows for that long again. I also did the thermal compound procedure and I believe that went well. Anyways though, problem is still happening. Holding the right click button isn't working for me - that would be really useful though, because right now I need to take out my 8800m gtx to get the computer to start with the integrated graphics.

 

You might want to try reinstalling the On-Screen Display Utility as well if you're having problems switching to chipset gfx.

 

the right-click procedure is supposed to be controlled by the bios, and therefore out of the hands of the OS and any driver setups.

It usually takes about 5 seconds before the system powers on from the right-click method, and it has to be the touchpad's right click button.

Good luck.

 

Alright I'm a bit of a dufus apparently. I thought I had to hold the right click and then hit the power button and keep holding right click until windows started up instead of JUST holding the right click until the system power turned on. All clear now though. Thanks for that, never knew that one before.

 

Well I'm going to bed for the night. Haven't been able to get a display driver "properly" installed because now even when i'm booting up in stealth mode I get the problem before I have a chance. I'm thinking the driver probably won't do the trick anyways, but if that doesn't work I don't think I have many more options - unless you guys can think of anything. Am I already to the point where I should try and "bake" my 8800 as a last shot?

Oh and I did end up doing a memory test and a hdd test just to triple check and passed both.

Anyone know someone who has upgraded their 8800 or something else that will fit my machine and doesn't need it anymore/will sell it for a significant discount. Right now I'm living in Finland, but I'll be in the States visiting family this summer. Unfortunately with travelling and school I'm really strapped for cash right now and can't afford a new 9800 gtx like you suggested Mystik (though that's only like 280 euros when you convert it). Would be really nice - actually if I came up with that kind of money I could get a Nvidia GTX 280M for like $500 Wishful thinking though I'm sure...

 

Hey, a computer repair friend of mine just emailed me this after I explained my whole situation to him:
_________________________________________________________________

"Sometimes new drivers upgrade the graphics card making it unstable.

It just happened with me today with a fresh install of Windows / reinstalling new ATI drivers. A game graphics out and bugs out from the new enhancements the driver offers. Switching to an older driver might resolve some issues. There is an over clocking application that can lower the physical clock. I suggest lowering the ram clock speed instead of the core clock. Most issues are caused by higher memory clock. This would explain the temperature it only monitors the graphics card CPU.

http://www.guru3d.com/index.php?page=rivatuner
application to lower memory cloak speed."
_________________________________________________________________

*Edit: Ok I was finally able to get some drivers installed properly AND... no luck same problem using the 167.whatever on alienware's site and the 179.48. I was going to try downclocking the memory with rivatuner, but I could only get my computer to run for a little bit in stealthmode and it's already downclocked. I'm pretty sure the card is overheating, but it's not the core that's overheating (according to HWmonitor, Speedfan, and Rivatuner). As suggested to me, it could be the card's ram, but what would be the cause if it's not the drivers? Are there any other possible fix solutions (besides the oven) or am I screwed here?

_________________________________________________________________

Alright, this is a way late update, but I've been out of the country travelling for over a month and I'm finally back home. Just wanted to thank everyone who participated in trying to help me. I learned a lot and feel much more 'connected' with my m15x now. I ended up running out of all options and did the home solder reflow (otherwise known as "the oven trick") and my 8800m gtx is alive once again. It runs a bit hotter than it did before, but I'm not sure if it's a fan problem or the card itself. Either way, with a laptop cooler, it's fine and I'm able to run anything I want to without issue. I can go back through my phone pictures to find proof if anyone wants it