M17x R2 strange issues (doesn`t boot)

Hey guys,

I recently bought off a second hand M17x R2 with some problem.Actually the problem was that the display didn't work every time when turned on, but everything was fine when attached to an external monitor.I thought it was a bad connection with LCD cable. Anyways I received the laptop, and everything was working just fine for a couple of days. The display didn't fail at all. But then suddenly one evening the VC fan switched to high speed.There was no actual reason for it because the notebook was just standing idle with just a browser running.I checked the temperatures , but everything was OK ( i cant remember exactly , but CPU was ~30c , and gpu was maybe 48 - 50c ) I waited for a couple of minutes, but fanspeed didn't go down.So i restarted it and after that fan speed was ok.When i restarted the display wend grey for a sec, in between switching from desktop to win loggoff screen Then my gf decided to turn the screensaver on, and after that the display never lit up again, but I could run some programs on blind via some hot keys, it played music and so on. I then restarted it again and that was the end. It would not boot up anymore. When I power it up , the display stays off, then i hear 3 loud beeps and CAPS, SCROLL flashing and NUM constant on. Yesterday I took the laptop apart, and everything seemed fine.The only strange thing is that there are no pads between the VC cooler and the RAM chips, and they are starting to look more white than black. After that I repasted everything with AC paste, got it together , did a cmos reset, but still the same. Now I am wondering if i should buy a VC first ( plan to buy directly a 7970 ) or should i go for MB. I suppose the VC ( its an 4870m btw) got fried, but then again its strange to fail so suddenly after 5 days of working without problems, and without running pressure on it ( games, benches, etc.). Would be happy to hear your suggestions.

Regards
Daniel

edit: one thing i just discovered is that when I power it on, the CPU fan just turns for a second and then goes off again, and the VC fan is working at high speed ( pretty noisy ). I guess that thingy with the CPU fan is normal for some machines?

 

Hi Anal,

It does sound like a video card problem. However, if you have any chance to test with a known, working card first I would do that. the 7970 can be expensive, and there still could be damage to the motherboard, LCD, RAM, etc.

Good Luck!

 

Hey rsgeiger,

Thanks for the reply! Now i see that i probably misread the light code.I hope the RAM fault code did not appear after i reassembled my laptop. I have tried switching the RAMs , but with no success. I guess i ll have to buy some new RAM and then try again...if it fails again , then its the MoBo.

P.S. if only one of the DIMM`s is broken, can i try to power the lappy on with only 1 installed in order to determine which one is broken?

 

The RAM faulty code might refer to RAM or the mobo. Have you checked with different RAM?

 

I guess its probably faulty MoBo. Today I tried swapping my 920xm with 620m, in order to check if the internal memcontroller in my CPU has not failed...but without success...It has never happened to me that ram could so suddenly just die...always it starts with some errors under load, some blue screens and so...but never so suddenly. And the ram is no more than a year old. The previous owner changed the 4gb factory ram with 8gb Samsung modules....i just cant believe that both modules can die like this simultaneously and unfortunately i don't have other RAM to test, so i just go and buy a new MOBO...but in Europe finding a mobo for R2 is impossible, so probably i have to order it from USA. Would be very thankful to any1 that can give me some info on where to buy R2 Mobo from Europe. Thanks in advance

Regards

 

I am a bit confused...this thread http://forum.notebookreview.com/alienware-m15x/555061-flashing-scroll-num-caps-lock.html says that if NUM is constant on and other ones are blinking, its VC failure and not Memory failure! On the German dell-support website, it says the oposite - constant NUM + others flashin is Memory failure....what is the truth then?

 

This is the most accurate info.

 

This is the LED code table I have always used. Hasn't been wrong yet: Alienware

 

Yesterday I managed to get it runinng. I just changed the CPU to 920xm again and reseated everything ( VC , LCD cables, DIMM's ) and when I started it for the first time there were only 2 beeps and no light code anymore. I tried different combinations with only 1 ram module, and I got to see the picture of that beautiful alien head I thought the ram was faulty, bit still I installed the second one and nothing changed. Unfortunately I could get into bios but no further...it would not boot windows, stayed just on black screen. After few restarts the VC fan stayed on high speed for a long time, and then I knew i will not get any picture...its that way ever since. The LCD is working by the way, coz I can see it changing different aspects of black when I turn it off and on. And I get no beeps or light code anymore...just the VC fan on high speed and black screen

 

Two beeps, eh? That indicates the system does not detect any RAM installed. If the RAM was faulty it would have four beeps. What did the BIOS screen say when you had both RAM modules installed? Did it detect both DIMMS and show the correct amount (4GB or 8GB) of RAM available? Two things also to check: which BIOS version are you running?Also, when you boot up again, try pressing F12 at the Alien head to go into the startup options. If the list shows up, try selecting Diagnostics. This is your PSA (pre-boot systems assessment). It checks the CPU, RAM, and motherboard/chipset for hardware errors. Run thru all the tests if you can and record any errors it spits out.

 

Yes it detected both modules.When I had only one it showed 4gb in bios, and when I added the second it changed to 8gb , so it was showing ram correctly.I didn't check the bios version unfortunately, but a sticker on the board is A04. I guess if the previous owner has not flashed a newer one ( which I doubt ) then its still A04. When I tried to run diagnostics, it started and then the screen was black and I had to reboot...I bought a new 6970m that should arrive in a couple of days. Guess if nothing gets better after installing the new VC, then the lappy goes to service diagnostics and motherboard repair/change

 

Does it work with an external monitor at this point?

 

The sticker doesn't mean that is the BIOS version of the motherboard. You'll need to verify in the actual BIOS what revision it's running. The diagnostics started running, then it went black, you say? Did the diagnostics make it to the VGA test. That's where it flashes a rainbow of colors on the screen and asks if you see the color rainbow and it's displaying properly. If you didn't get that far, it's possible the GPU has issues.

 

Update : while I am still waiting for my 6970m to arrive I left the R2 without battery for a couple of days and also I pulled the small MoBo battery out.Today I decided to give it a go and start it, but without connecting the small battery to the mobo.It has gone trough POST without any beeps or lights but after the initial screen after the alien head it just hangs up...there is no image on the screen, nor any HDD function ( the eyes of the small alien head stop blinking ), nor does it want to eject my CD ( on alien head screen eject works). I can get into the bios, which is A10 version by the way, but no further. Bios recognizes the 920xm, both RAM modules, both HDDS...everything...it just wont even get to the windows loading screen. I have tried running diagnostics, but it happened just the same - nothing. The screen is working , because i can clearly see its back-light... it just does not display anything. I have tried everything - boot from both HDD's , from a windows boot memory stick, from CD...I guess the GK just hangs up when it sees a bit of stress, or maybe the problem is elsewhere? RAM maybe? I don't know what to think anymore...would be glad to hear your opinion

 

Try removing the hard drives and running the system again. If it gets thru POST then says there's no bootable devices, then we have an answer.

 

So whats the answer then? With both HDD's out it happens exactly the same...goes through POST and then hangs up...whats the problem?

Edit: I don't know if it is relevant but with HDD's installed, when I go into the bios, both HDD's are recognized...now that they are out, Bios says [none], so it recognizes that there is nothing installed, but still nothing changes and the situation remains the same

 

Tells us its not the hard drives or any of the software or OS' installed on them. One thing to remember when troubleshooting, EVERY piece of info is relevant in some way. The BIOS is detecting the hard drives correctly (and detecting when they are missing). The more I look at the symptoms and what we've tried so far, the more I think its the CMOS (BIOS chip), or the BIOS itself. You have two options: you can wait for the new video card and see if installing the new one fixes the problem. OR...you can try something a little more radical. But before that, if you can get into the BIOS screen again, the first BIOS page will tell you which BIOS version is installed on your R2.

 

Its the A10

Edit: another strange thing i noticed is that sometimes it's detecting my processor as 1 physical CPU with 3 cores, sometimes with 4 cores...and it is not detecting the model.I will make a screenshot later and upload it

 

That's not a good sign. Maybe you should try reseating your CPU. And hit the bottom of the die and ZIF socket with a quick blast of air from the can.

 

It managed to boot!!! All i did was to blow/clean the socket/pins AND replace the CPU with the old i7 620m...now I don't know if the problem was because of the cleaning or because of the CPU...guess the only way to know for sure is to exchange the CPU's again...I did various tests and I am running Diagnostics from the boot menu at the moment.Everything seems OK, no errors so far, temperatures are fine...the GPU is pasted with MX-2 both on mem and core...on idle are: GPU 39, mem and shader - 45...main CPU is 35. The 6970 is probably arriving tomorrow and honest to say I'm a bit afraid...I fear that everything will break down again if I change the card again.Any suggestions?

P.S. Radji, big thanks for all the help! ++++ rep

 

As long as your careful, nothing should happen. You want to make ABSOLUTELY sure the new GPU is firing on all cylinders, and given your current 4870m card has been fried by not using thermal pads, the new 6970m card will go a long way to getting your system to perform at a high level again. Swap the CPUs around and see if the XM CPU is the culprit or its installation. Make sure to hit the socket and the chip bottom with air from the can again. Also, visually inspect the pins on the bottom of the CPU (but carefully and wear an ESD bracelet...they're like $20 but the protection they provide is invaluable). Check the pins on the chip to make sure none show any sign of damage, wear, distortion, or dirt/dust. Those things are tiny so it would not take much to mess up those pins. Also check the ZIF socket. See if perhaps a bit of dried thermal paste or something was lodged into one corner blocking a few of the sockets. If the XM CPU still gives problems, then it is your perpetrator. The point is to determine if you can keep the XM CPU in and then do so. You want to repaste the CPU and not have to worry about it again for a long while...a.k.a, until it needs to be re-pasted again. The CPU should not have issues like what your having. It leads me to believe either one, the XM CPU wasn't handled properly at some point in it's life and a bit of static damaged it, or two, a bit of dust got in between the CPU and socket or the CPU was a c-hair off when you swapped them out. Once you install a CPU in the socket, there is no reason to remove it unless it's being replaced or upgraded. The seal between it and the ZIF is meant to keep contaminants out so there is no reason to remove it for cleaning or repasting. If you can get the system working properly with the 920-XM CPU, then make sure to not remove or adjust it unless you absolutely have to. That will bring the chances of it failing again due to manhandling down to zero. As far as the 6970m card goes, if that is the only thing you will be replacing right now, you have nothing to worry about. You can replace the GPU in the R2 so easily. I said get an ESD bracelet, right? Well, times that by 10! GPU cards are especially susceptible to static damage. You may want to think about an anti-static mat to put the system on while working on it too. I know this all sounds extreme, but I can't tell you how many times I see users here in the forum cry about their new MXM cards being DOA. Everytime I think, "Well, did you use and ESD bracelet? NO? Sucks to be you!" As for replacing the GPU, just make sure to clean the heatsink very thoroughly before you mate it to the 6970m card. And it can't hurt to hit the MXM socket with a quick blast of canned air to get out any dust that might have made it's way in there. Here is the forum guide written about installing the 6970 cards in our R2s: http://forum.notebookreview.com/ali...-6990-your-m17x-r2-single-gpu-crossfirex.html.
It's a good read, so follow the instructions in it. Let us know what happens>

 

I have some news....unfortunately bad...With the new card the computer does not even get to POST I suppose...I turn it on, everything lights up, GPU fan goes max speed, CPU fan does not turn at all, and then just nothing happens...the eyes of the small alien head are not even blinking, which indicates it is not even trying to go on...And on top of it I can not even shut it down. I have to remove the battery...I don't know what to do anymore...I am tired of all this

ok...I got the reason...on closer look there were some broken transistors on back of the card...I ll need another one...the question is , can these broken transistors have affected my motherboard and damaged it somehow?

 

Possibly, if they caused a surge of power to the motherboard. It's very unlikely though since the motherboard has some safeguards against internal power surges and the GPU gets its power directly from the PSU without having to go thru the motherboard...or so I've been told . Which card was the culprit anyway?

 

There are so many sign that you have faulty hardware but at this point they all point to the mobo or cpu.
At this point i can see you only had a succesfull post with your original HW combination:
620m + 4870?

If this was just a random act then i really do think that all the problems originate from the motherboard or Bios even...

Since you swapped out everything....i suggest finding a mobo, but as you already stated: finding a motherboard in Europe is a difficult task.

 

Finally, everything works fine! Except for some minor issues...I have the feeling that the VC is throttling. Its just unusual to achieve between 10 and 17 FPS in Starcraft 2 on high. I will have to probably undervolt it to 1.0v ( its 1.1v now). Another issue that is bugging me is that the colors on the keyboard, alien-head, etc are messed up. Seems like the red colored leds are staying permanently on. And this results in not pure colors. When I select blue, I can see the red leds glowing underneath.On white its even worse...it results in a white-redish color which is really annoying.I uninstalled and then reinstalled Synaptics driver, Command Center, .NET Framework. And still it wont change...could this be a hardware problem or? The second problem is that with the 920xm installed, the system hangs after POST, just before the windows screen.It is an ES, and BIOS recognizes it and everything (both ram sticks, etc ). It hangs even when i try to run Diagnostics from the boot menu. With the old 620m processor, everything works like a charm. I wonder could all these problems be related and a BIOS issue ( I somehow doubt it but you never know) or its just a bad Processor? Bios is latest version - A10

 

Like i said: everything works "fine" with your old setup so your CPU is messed up i am afraid...
Good to hear you got everything working again.

As for the leds on the Kboard, when you load your preset profile via the command center, does it reset the colours?.
Does designing NO colours turn the keyboards black/off?

 

I am now having to agree with unphoto. Its looking more and more likely your motherboard is going bad.