Is my T61P finally dying?

[Akilae]

So I've had my T61P since 2008. It has been sent back once for a motherboard replacement (2010 I think?), and has since been running okay. Not well, but okay. So the sequence of events is as follows:

1) Post motherboard replacement, the laptop ran a tad hot, and the fan didn't quite rev as it used to (actually, wouldn't rev up at all...), leading to heat-related shutdowns when gaming. I installed TPFanControl, which fixed things somewhat. The GPU would peak at 90s, but the fan would then rev up and cool it back down to 80s. Idle temp sits around mid 60s in winter, low 70s in summer.

2) A few days ago, I ran a quick game session, then set the laptop aside for the night. The next evening when starting up the game, the laptop hard locked.

3) I ran a few other games/programs to test. CPU-intensive programs run okay. Anything that called up more advanced GPU functions resulted in one of the following right as the screen shifted from splash screens to something more graphically intensive:
3a) Hard lock with screen frozen.
3b) Hard lock, graphical glitches, then blank screen with power still on.
3c) BSOD linking back to display drivers.

4) I decided to update the display drivers to official nvidia drivers. Things ran well for a while, then I got hard locks again. This time, the BSOD would be this:

Hardware malfunction
Call your hardware vendor for support
NMI: Parity Check / Memory Parity Error
The system has halted

I looked up this error, and it's usually tied to ECC memory. Except I don't have ECC memory installed.

5) I noticed my fan has also stopped revving up as much. When starting up games, the top left hand corner of the laptop is warm to touch, but conversely the bottom right hand corner is hot.

So is there anything I can do to determine what's wrong with the laptop? Does this sound like something that a heatsink/fan replacement would fix? Should I start shopping for a backup computer/laptop ASAP?

[RealBlackStuff]

Sounds to me in the first place like overheating.
Remove palmrest, keyboard and KB bezel, then remove the fan.
Clean out all dust and debris. If you use a blower or canned air or similar, stick a toothpick or similar through the fan to stop it from over-revving.
Clean up old gunk on CPU, GPU and cooler with Isopropyl Alcohol 91% or better.
Apply fresh, thin layer of AS5 on CPU and GPU, and reassemble. Don't forget fan power cable.
Try it with loose keyboard before you put the rest back together.

[Akilae]

Should I be worried about the thermal pad breaking? Instructions I've looked up have been careful to say things like "Beware about the thermal pad being broken! Don't use thermal paste! You'll need a new thermal pad!" I don't want to pull off the heatsink only to be stuck with nothing to replace the thermal pad if it breaks.

[JeffCullen]

edit: deleted due to my bad!

[RealBlackStuff]

Careful here:
some machines have only thermal paste (on CPU only, or on both CPU + GPU), and some have both paste (always on the CPU) and a pad (always on the GPU).
And to make things interesting: some also put either a pad or paste on the Southbridge chip.

And just in case: I did not recommend to replace a pad with paste...

If there's a pad, try to save it as best you can, and do NOT try to clean it.

[JeffCullen]

Must have had old desktop CPU heatsinks in my head where the stock pads were rubbish and AS5 was preferable... thanks for the correction and clarification!

[Akilae]

So basically I won't know what condition (or if there even is a thermal pad) unless I pry off the heat sink assembly... ideally I'd like to be prepared in case the thermal pad does not come off intact. Is it possible to use an aftermarket thermal pad? Would anybody happen to know what size/thickness is used in the T61P?

[Cigarguy]

A thermal pad was applied at the factory on the GPU and chipset. Unless you are very unlucky you will not damage the thermal pad when pulling the HSF up. Unlike thermal paste, chances are good it's still fairly flexible and is in good shape. I've done a lot of teardown and reapplication of thermal paste and have not damage or destroy any thermal pad yet. When you remove the HSF, the pad will be stuck to the heatsink, leave it there, don't play with it and don't remove it.

[Akilae]

Well that was interesting.

I disassembled the laptop and cleaned the fan/heat sink assembly. I had run out of the good stuff (Arctic Silver), so I went with the only thing I had: a tube of thermal paste that had come with my roommate's Zalman heat sink. Blew out a lots of dust, scoured the chips and the heat sink clean of residue, reapplied, and screwed everything back on.

Now the GPU idles at 60 according to TPFanControl, and 66 according to CPUID Hardware Monitor, compared to high 60s and low 70s before (CPUID was telling me high 70s...). Under the same game stress test I had used earlier, it now peaks at 83, compared to 95 before.

I'm still not sure if this will cure the laptop of its hard locks and BSODs. Only time will tell. For now I'm already happy that temperatures are more under control.

[Cigarguy]

The temps after the work is more reasonable. It should decrease a by a few more degrees once the paste have time to settle in a bit. I think you just had a case of the old paste drying out, no longer is effect and causing the temps to soar.

[WalkGood]

I've had problems with BSOD and "Memory Parity Error" message. Kept getting worse.

Problem is solved now. It sounds strange, but updating your Wireless LAN card driver will most likely fix the memory parity error.

I went from having the BSOD prob;em 5-plus times a day, to none at all in 2 weeks since updating the driver for the wireless LAN card. Give it a try. Couldn't hurt.

[jedisurfer1]

your t61 seems to be running very hot. I would also undervolt your cpu with rmclock or throttlestop. It shares the heatsink with the gpu and chipset so less heat here is a good thing. I set tpfancontrol at 1 or 3 always on. And tpfancontrol says I hover at cpu low 40s idle, 37* gpu (cpuid hardware monitor says it's 45*). Mine's been running rock solid since 2010 when I got the motherboard replaced from lenovo with a known good video chip. It has been the most stable laptop I've owned and has been problem free since the replacement. It has run almost 2 years without a bsod or crash, all while running 3 virtual machines.

Edit: I've found drive 195.62 for windows 7 x64 (get the inf from laptopvideo2go.com) to be extremely stable for vmware). I also don't ever game so video performance isn't an issue with me.

[TuuS]

Just to clarify, the discrete T61 and t61p uses thermal paste on CPU and GPU and a pad on the chipset. The Integrated Intel boards use thermal paste on chipset/gpu (both in one chip) and cpu and NO pad T60 and T60p discrete units use pad on GPU and chipset and thermal paste on cpu, however the pad on the gpu was a bad idea and I would recommend replacing it with thermal paste as long as you have a good fit of heatsink to gpu. The pads often get cooked, I've seen them fall apart in pieces, even turn brown/black like they were burned. I prefer to use the T61p heatsink on T60/p discrete models, it requires a notch be cut in the retainer clip to clear the larger heatsink, but the benefits are substantial.

Here is a picture of the mod...

http://www.e30tuner.com/assist/frankenp ... _hs010.jpg

[Akilae]

Okay, so the hard locks are still ongoing.

I don't think this is heat related anymore. I ran a session of Minecraft and watched the temperature on the GPU climb up to 78 degrees, then hold steady around there. A few seconds later the laptop hard locked, then threw up what I can only describe as a garbled BSOD with the Parity Check error. Garbled because the screen was obscured by thin red-ish lines. Reboot, try again, and it hard locks again, this time just frozen with the sound stuttering.

Not sure what else I can try...

[ajkula66]

Not sure what else I can try...

Remove the wireless card and see whether the machine behaves differently.

Good luck.

[WalkGood]

Remove the wireless card and see whether the machine behaves differently.

Good luck.

Or at least update the wireless card driver.

[Akilae]

I'm using the most up-to-date version of Microsoft's divers though, not Lenovo's. Does that make a terribly large difference? AFAIK, Lenovo's came with Access Connections, and I greatly prefer Microsoft's setup...

[ajkula66]

Remove the wireless card and see whether the machine behaves differently.

Good luck.

[Cigarguy]

Use HWMonitor and report back on CPU temp. If it's hitting 90+ deg C then I can see this happening. Also try what Ajkula66 recommend.

[WalkGood]

I'm using the most up-to-date version of Microsoft's divers though, not Lenovo's. Does that make a terribly large difference? AFAIK, Lenovo's came with Access Connections, and I greatly prefer Microsoft's setup...

You could try Lenovo's for a few weeks. See if things improve. If there is no difference with your error then roll back to the MS driver u like.