Toshiba Qosmio X505-Q8102 random shutdown, surely not a cooling issue

Hello,

I've bought X505 two years ago and last year it started to shutdown randomly under load. Im not a newbie in laptops and i've fixed the overheating issue just after I've bought the laptop. Temperatures are low and sometimes laptop it's really cold but shuts down itself.

Yesterday story:
It runs nice for 2 hours on windows desktop
Im starting Borderlands 2
CPU: 55C, GPU: 62C
Playing for 1-2 min (MINUTES) and it shuts down
CPU 61C, GPU 65C
Surely it's not overheating issue, i've fixed it long ago by replacing thermal paste to the high class one and cleaning fans.

I've read somewhere on the internet that this could be the power board issue, I bought one and replaced the old but the problem still persist. When replacing the thermal paste I was looking precisely on the motherboard and CPU but I couldn't find any visible damage causing this shutdowns. Maybe it's cold solder joint on the motherboard but that will be worst thing that it could be.

It cannot be software issue too, system shuts down even on fresh Linux install. This is just instant power cut, no 'shutting down' screen, nothing. Most interesting thing is that on the external power it shuts off after 1-2 minutes but on the battery it shuts down much more often and sometimes just after 10s of heavy load. So it cannot be power adapter fail too. I've even tried running it without battery but no luck, still shuts down.

Im really desperated, there are almost no parts on the internet for this laptop as it was expensive and small amount of qosmio x505 with Sandy Bridge CPUs was produced. So finding power board was hard but i've tried to find new motherboard and it's too expensive for me so im trying to fix the issue by myself.

Laptop specification:
i7-2630QM Sandy Bridge
GTX 460M
HM65 MoBo

Laptop exact model - PQX34U-00M00V
MoBo service product numer (if useful) - A000054130, second generation

Have You any ideas what the issue could be? I've run out of ideas. If you don't know how to help me please point me to the place where I can get help on PM but I think that's biggest notebook forum.

Regards

EDIT::

I've made a video that's better describing the problem -

 

So does this happen when under GPU load or always? I couldn't figure that out from your OP. You did mention Linux though so I suppose it happens all the time?

 

Hello,

Thanks for reply.

It happens when CPU or GPU is under load, i can't really test CPU without GPU and GPU without stressing CPU so I don't know. I couldn't reproduce problem using Furmark stress test or Prime95. Furmark 100% GPU, no restart. Prime95 100% CPU and no restart. Sometimes when system is idle it shuts down too but very rarely. But as far as i know if GPU is damaged then artifacts should appear and there's no artifacts, display is fine. I've tried underclocking GPU with MSI afterburner and testing but result is same, shutdown.

GPU is replacable but it's on Toshiba modified MXM slot so only Toshiba GTX 460M would fit and toshiba specific hardware is very expensive so if it it's gpu issue then no luck for me.

I've done a memory test but memory is fine, GPU memory test was passed too. I've tested CPU in AIDA64 CpuQueen and results are correct for an i7.

Regards

 

It could be that your GPU memory is damaged, at least some part of it. Games usually use a lot more VRAM than Furmark when browsing the internet and the chances that the damaged area is accessed are higher when playing games.

You could test this by seeing if the shutdowns stop or occur less frequently when lowering the resolution and using lower graphics options.

 

Hello,

As You said I did some testing.

I've lowered screen resolution and level of details in games but result is the same - shutdown

I also did a video memory test with OpenGL and DirectX using Video memory stress test program, results here:

Code:

Changing video mode to 640x480x16...OK
[2013-10-31 08:14:22] Test started for "NVIDIA GeForce GTX 460M"...
Trying 16bpp mode...NOT SUPPORTED
Trying 24bpp mode...NOT SUPPORTED
Trying 32bpp mode...OK
[2013-10-31 08:34:29] Pass completed (0 errors found).


Changing video mode to 640x480x16...OK
[2013-10-31 08:38:13] Test started for "Primary Display Driver (NVIDIA GeForce GTX 460M)"...
Trying 16bpp RGB:565 mode...OK
Trying 16bpp RGB:555 mode...OK
Trying 16bpp BGR:565 mode...NOT SUPPORTED (Code: 80004001)
Trying 32bpp RGB:888 mode...OK
Trying 32bpp BGR:888 mode...NOT SUPPORTED (Code: 80004001)
[2013-10-31 09:06:16] Pass completed (0 errors found).

Finally i went to the BIOS and tried to turn off CPU cores to test how it'll go on 1 core and laptop shutdown in BIOS while changing settings. Luckily bios settings aren't damaged so i went to the BIOS settings again.

When I tried second time to turn off CPU cores, it worked, I did some testing using just 1 core of the CPU. Laptop didn't restart but i don't know that is some of the CPU core damaged or it just wasn't loaded much to show the issue but i can finally exclude a GPU fail. It also still could be a MoBo fail. My BIOS allows me to keep the one core running or the all four so i can't test cores separately.

After this testing I again went to the BIOS to turn on remaining cores because laptop was very slow on the only one core and notebook shutdown again in BIOS, luckily the settings weren't damaged again. I suppose that GPU is fully working and MoBo or CPU is the issue. I won't go to the BIOS again because these shutdowns in BIOS can brick the computer completely.

Have you any more ideas how to check which part is damaged?

Regards

 

I am not sure what it could be. May be one of the other users could be of more help.

It shouldn't be the mobo because that would mean the laptop should shut down regardless of whether it is under load or not. It shouldn't be the CPU (despite your last post) because Prime95 ran just fine. Prime95 runs multithreaded and uses all available forms of SIMD on x86 processors which means it pretty much utilizes the CPU as much as possible.

It shouldn't be the GPU because the tests ran fine.

If you combine the memory tests with your stress tests, you simulate a gaming environment in that the usage of components is similar.

The one thing I can think of is a bad voltage regulator. You said the laptop shuts down quicker when on battery. Heavy use dictates for a lot of instantaneous current. If a voltage regulator is bad, it could cause a larger power draw than what the battery can sustain. What do your core voltages show for the CPU and GPU?

 

Hi,

Thanks for fast reply.

Im attaching voltage of CPU in GPU in AIDA64 extreme stability test. It's very complex test that combines RAM, HDD, CPU and GPU so laptop should shutdown, but it didn't...



Test was running fine for about twelve minutes, shutdown should occur much earlier. I don't know how to check voltage while running the game, even if it's too high and laptop shutdowns i won't be able to check that because it'll shutdown faster than any program could check.

On the beginning it started to shutdown ONLY on battery but im not using the battery often so i ignored it. Then after some time, 1-2 month it started to shutdown even on the external power but after 1-2 shutdowns it stopped for another week so i ignored it too, I thought that's software problem. Now it's shutdown everytime when playing games or working in 3D designers. Maybe that's some clue that will help You to identify the problem.

Also i tried it running without the battery, just on external power and you can suppose what the result is - shutdown

Unfortunately I can't even work on it because it'll shutdown while testing the project (im developing 3D Games in Unity3D)

One more thing, voltage regulator is located on the MoBo, isn't it?

Regards

 

I think Throttlestop allows for recording voltages which it saves to a file. I'm not sure if it saves on the go or when you stop the test. However, that will give you a voltage trend right before the shut down (assuming it records on the go). If voltage is steadily increasing, that would point to a VR problem.

Also, yes the VR is on the mobo.

 

Hello,

I tried everything second time and lowered GPU core to very low 400 Mhz, memory to 600 MHz and laptop stopped restarting! I've played borderlands 2 on medium details because of clocks for 2 hours and no shutdown.That's really weird because stress tests or gaming on 1 cpu core didn't showed this issue. It looks like the GPU card is degradated and can't work on stock clocks but it should show artifacts, not shutdown the whole PC. Maybe toshiba vBIOS is working in little different way than normal nvidia one.

Do you think that reballing could help to fix this? My friend could do a professional reballing service, this is only way to fix this laptop as Toshiba GTX 460M with modified mxm slot is very expensive and rare.

I can also try to edit vBIOS and set lower voltage and clocks permanently for GPU. Im not an expert in overclocking or downclocking. If card is unstable at stock clocks should i increase the voltage and try set stock clocks or just down clocks because increasing voltage could burn this card? Current voltage for 3D mode is 0.9125V

Best Regards

 

Power brick dead ? I'm not sure if you mean power brick by power board

 

Hello,

Not power brick, it's fine. Power board connected to the MoBo that contains power button and red LED light.

Can you help me with this voltage question? I can set 0.9625V instead of 0.9125V but card is degenerated, it can't go stable on stock clocks so i dont know what to do. Reballing will help? Maybe increase voltage? Maybe decrase voltage? I can do this with NiBiTor as it supports GTX 460M mods.

It's stable on Core: 500 MHz vs 675 MHz stock and Memory:466 MHz vs 625 MHz stock

Regards

 

How can you be sure its fine, it can be not working well under load..
Try to use a watt meter to see if there is drop before the shut down

 

I've lowered the clocks of GPU and did the stress test same as gaming and no shutdown, power brick is fine. GPU shutdowns PC even on battery so it's not power brick issue, it's GPU degeneration. Power brick is second generation - 180W, GTX 460M max is 100W + max 45W i7 + screen and other stuff there's still much power.

Regards

 

Any kind of GPU problems does not mean it will artifact. Artifacting mainly occurs when your RAM goes bad. Also, I am still putting my money on a bad voltage regulator. That is the only thing that explains why your laptop shuts down under heavy gaming and not when running stress tests or when you game but downclock your GPU. Clocks linearly increase power consumption. You reduced it by almost 35% by downclocking.

 

Hello,

I will check the voltage regulator too but the service manual isn't available on the internet, I will have to find it by myself on the MoBo. I'll also do the reballing if the voltage regulator will be fine. Thanks for the help guys.

Regards

 

How much are you going to end up spending on this? It might be worth getting a new laptop instead. This time, with a 3 year warranty extension.

 

Hello,

This laptop isn't typical notebook. It has 18.4 inch display and two hard drive bays + high-end components. I've checked and there's no other 18 inch laptop in 1500$. Only Alienware M18x which is available only from USA with delivery cost to Poland about 400$ (!)

There was Clevo X8100 and Clevo P180HM but then clevo stopped producing 18 inch notebooks, I've emailed them and they said that no new 18 inch model will be out.

This Toshiba is last generation of Toshiba 18 inch notebooks too, no new models, no parts on the market.

As You can see I just don't have a choice.

Regards

 

I'm just saying that a 17.3 notebook might be something to look at. I don't know what you are spending or what you will spend. Just saying that make that decision based off of the total money spent on your current machine vs total money you would spend on a 17.3 machine. All 17.3 today come with two drive bays. In fact, the only thing you'd miss out on is the screen. Think about it before you spend.

Good luck.

 

Interesting how there is load related issues and this particular card (GTX460m) has some strings attached. I suggest that you look for an upgrade with in the same power spec or less than your current card after all it is mxm hopefully not of a proprietary standard. GTX660m or better would be nice breath of fresh air gaming wise and likely to not have random issues such as this but Fermi era cards have vrm issues for both desktop and mobile.

 

OP mentioned twice that the MXM is modified by Toshiba.

 

Hello,

I will do the reballing on monday.

As for the new laptop i like big screen, very big. 17.3 is too small for me. I would like to buy even 19 inch but there's only Acer on the market so i bought 18.4

I don't know why poeple like these 15.6 miniatures. It looks so funny when someone is keeping his head 10cm from screen to have good view. Everyone have a car so i think weight isn't disadvantage.

And for the GTX 660M i said that MXM was modified by Toshiba but even if it wasn't i wouldn't buy 660M because performance is not much higher than 460M.

460M pipelines 192 x1(Fermi)
660M pipelines 192 x2(Kepler)

My 460M score on new drivers in 3dmark06 14.5k
660M score in 3dmark06 15.5k

And if you increase the voltage, then clocks in 460M it would be even faster than 660M because the newer one have only 128bit memory bus.

If I had normal MXM I would buy Radeon 6990M or GTX 675MX because they're cheaper now and I can improve the cooling, I have some electronic skills.

Regards

 

Yes there would be no point spending money on that upgrade.

You'll need materials or fluid mechanics skills to improve cooling. You cannot do much better (definitely not 460m to 6990m better) with simply adding some electronics. Just a FYI in case you are considering a much hotter GPU.

You cannot generalize seven billion individuals (or the major fraction of that number that own a laptop). There's several reasons people choose smaller laptops. Every Tuesdays and Thursdays, my computer architecture professor walks in with a 13 inch laptop. The desk in our classroom simply doesn't have space for something larger.

If you are frequently on the move and travelling by air, you cannot lug around a 17 inch laptop. Same goes for students. Some prefer purchasing a good desktop and a small laptop to take notes and do homework on campus.

So there's several reasons to own a smaller laptop. They don't apply to you, but that doesn't mean they do not apply to hundreds of millions of other people either.

Good luck with your efforts.

 

Ezio, have you had any luck or news on your rig? I've been having the same issue with my X505-Q882 for about a month now and haven't been able to narrow down the cause; there seem to be several cases of the same issue but none have reached any form of resolution - everyone wants to jump right to overheating as the cause, which I know for a fact isn't the case. I stuck mine outside in 20 degree weather and have cleaned the entire system out (and replaced both fans) and it still reboots. I replaced the AC adapter when I realized that it only reboots off of battery power, no change. I've run every test possible and every 'base' component seems to be running perfectly fine. My problems started after reformatting the machine - wanted to clean the drive up and I prefer starting fresh, but made the mistake of removing the recovery partition (I literally never use them) and I worry that there is some obscure power management driver I lost for good.

So far as I can tell, other X505 owners have replaced motherboards, hard drives, ram, graphics cards and the processors through Toshiba only to see the issue persist. It is driving me mad that I can't figure out what the problem is and the fact that there isn't even an error log to tell me why it's shutting off is really agitating. It's almost like the computer is reacting as if I'm holding the power button down manually.