m18x motherboard problem (i guess)

Hi Guys,

I hv this 2nd hand m18x R1, it is having screen freeze problems, system is also not responding. it will happen about 30 mins after it is turned on. the screen freezes and also the cap/num lock lights are not responding when pressed the buttons, so it should be a system freeze rather than graphical. software reads 80+ cpu temperature, shortly after system turned on.

i opened it up, and see whats in the picture. somebody attempted to fix it apparently.

it seems that one of the fan power outlet is broken so somebody tried to connected to the other one with the yellow wire. clever, but air coming out of these 2 fans are not as hot as the air coming from the fan on the other side (not in the picture).

I suspect that this arrangement is causing both fans to not work as efficient as before, thus getting overheat and freeze. im thinking instead of replacing the mobo. is it possible to take one of the GPU out (it has a crossfire set up), disconnect the fan on top of that GPU (left fan in the picture) and then plug in the CPU fan properly.

does anyone know if this might work and fix the overheating and screen freeze problem？

Thanks.

 

It's hard to know for sure. That wiring arrangement may or may not be causing an issue and the overheating could be just a matter of the heat sink radiators being clogged up with lint and dust or needing thermal paste replaced. The left GPU (nearest the VGA port) is the primary GPU. You can try it with the right GPU removed. You're going to need to use the process of elimination. Start with the easiest solution (like cleaning the heat sink radiators) and work your way from there. Let us know how it goes.

 

Hi Mr Fox,
Thank you for your fast reply. I will check the radiators and heat sink 1st. Do you think that the yellow wire could have been a control wire （instead of the power cable）? i watched them working, these 2 fans turn on together while the other one (not in the picture) was always on. do you think may be these 2 fans only turn on when the GPU is having a high temperature? since they are connected to the GPU connector? which might be the reason why CPU is having a high temp, cuz that fan is not controlled by cpu temp? it would be helpful that we have a blue print of the mobo somewhere. so we know which pin of the ports does what.

Thanks again for you help.

 

Yes, that is a distinct possibility. Is the connector on the motherboard physically broken? If it is a PWM control wire and it is using the GPU signal, that definitely could be the cause. The CPU will be hotter than the GPU in most cases and if the fan isn't coming on that could easily cause the CPU to overheat.

Try using HWiNFO64 for manual fan controls and set the fans to run full blast and see if the CPU temps are still too high under heavy load.

 

Hi Mr Fox,

yes i tried with max fan speed, it then seems alright. gone thru graphics test and stress test no problems. but it sure is noisy.

i noticed that im only able to control fan1&2 not fan 3. dont know why that is.

also, i had 8 beeps problem today before i could set up the fans.

the screen went on with green and sometimes pink particles, all over the screen. computer freeze shortly after that. i force shut it down, and tried to restart it, it gave me 8 beeps. i left it there for a few mins and then started it again, 8 beeps gone. should def be heat related then, since the pictured fans are above the primary gpu and cpu.

after this restart i set up the fan to max manual set, then it went thru the test.

Do you think that the primary gpu might be damaged by the heat? should i swap those 2 and see if theres a difference? but since it starts up no problem and can go thru test, it should be ok. just heat, i guess.

may be i will take out the slave GPU and plug the cpu fan in that intact port? but then there is the whole sensor and fan control problem again, since the slave gpu is gone and there wont be any heat, right?

Thanks.

 

oh actually, damn it... i got it wrong... so the damaged port, is actually the slave GPU, which is under the numpad side of the keyboard. do you think it will be helpful if i just retire the slave gpu and see if it makes a diff? but i thought the 8 beeps problem is for primary gpu?

Thanks.

 

update, just disabled the 2nd display adapter in device manager, restarted, and ran the alien autopsy graphic test again, without the fan control, it still gave me the green particles and froze

 

Test your GPUs one at a time in the primary slot. First remove the secondary GPU (the one under the number keypad). If the primary GPU has problems running by itself it is defective and needs to be replace. Whether it works correctly or not, swap the secondary GPU into the primary slot and leave out the GPU you remove from the primary slot and see if things work right. If everything is OK with both GPUs working solo in the primary PCIe slot, you may have a bad SLI bridge cable. You could still have a motherboard problem as well, and with the goofy hot-wired fan thing going on, you're going to have to go through the process of elimination to isolate the component causing your issues.

 

Hi Mr Fox,

i took out the slave, it seems working fine now, passing video tests without fan control.

at 1st i accidentally took the master out, put everything back and started wonder why i only hv intel graphics.... then figured that i took the master out, cuz i flipped it too many times and confused myself.......

i am going to run it for a few days and see if it is stable.

hopefully now i hv got a working laptop, and a set of crossfire upgrade kit lol.

Do you know how i can tell if the gpu is a 6970 or 6990?

Thanks.

 

You're welcome. You can tell if it is a 6970M or 6990M looking at the specs in GPU-Z. Have a look here: http://forum.notebookreview.com/threads/6970-or-6990.618628/