7811fx overheating issue and initial troubleshooting done

hey all. i've had my 7811fx since feb of this year. before coming to this forum, i've had to send in my laptop back to the manufacturer as it was freezing while playing a game. there was an issue with the video card and the motherboard was replaced. now i'm deployed in iraq and would rather not send it back for repairs.

so far i've updated the video card driver using 186.03 off of nvidia's website, all other drivers are straight off of gateway's website that never seem to get updated. i have updated the bios to 9c.17. i used the sticky guide to disassemble the laptop and reapplied arctic silver 5 (it's all i have here) to the cpu, northbridge, gpu, and the gpu ram.

i disassembled the laptop yesterday and i'm still seeing roughly the same temperatures for the acpi, cpu, and gpu. the heatsinks and fans are clean and free of dust. i don't experience the audible pulsing fan for the gpu. at idle the temps for acpi, cpu, gpu are 42C, 29/22C, and 36C. while playing games it's 65C, 55/47C, 97C. sometimes the laptop will freeze with the screen going black and the audio goes in a loop. i have to do a hard reset after this happens.

i'm pretty sure it's not the hard drive. before i got my zalman nc2000 cooling pad the hard drive temp would go as high as 63C. it idles at 40C and under load it's 45C.

according to arctic silver's application instructions it states that it takes up to 200 hours and several thermal cycles to achieve maximum particle to particle thermal conduction and for the heatsink to CPU interface to reach maximum conductivity. so should i wait over a week to see if things get better?

all the hardware on the laptop is stock
p8400, 4GB ram, 9800m 512MB, 200GB hdd
vista 64 sp2

 

If I read this right your GPU is running a bit hot and just a few degrees from "slow protect".

You stated that laptop has been totally pulled down with AS applied to GPU, CPU, others.

If you have no heat damaged parts, I would venture that you have bad contact or a damaged heat pipe on the GPU? GPU fan not running at high speed?

You can use "nVidia Monitor" to write temp event log to a text file that can be read after the game or crash. This runs great in the back ground during game.

Hopefully someone, with a like problem, will chime in with their fix.

Stay Safe!

 

i thought that there might have been a bad contact after receiving my computer with a new motherboard from the manufacturer. the gpu fan works fairly hard when i'm playing a game. it'll keep my coffee warm for a good while if i leave it near the exhaust port.

i'll download the system tools and reply back with the report. it's going to take a couple of hours. i got a 128kb wireless connection at the premium price of $85/month. i love iraq.

 

Yeah the GPU temp is really odd... are you running an insane overclock by anychance?

If not theres something not making contact to let the heat escape as well as its supposed to.

Even under extreme stress testing my GPU never went over 89 degrees... well except the time it hit 11 million degrees, but that was a weird error in HWMoniter, or i was able to harness my computers sheer awesomeness to create a mini star in my livingroom

 

nothing is overclocked. i suspected a bad contact between the gpu and the heatsink, which is why i disassembled and reapplied TIM.

i have experienced tzs0 hitting 98C playing cod5 and reduced frame rates. but since updating the bios to 9c.17 the issue hasn't reoccurred, it just freezes. i might revert back to 9c.10 or even 9c.08 and just deal with the slow frame rates.

 

If you do revert, you know that RMClock will solve the TZS0 crash right?

 

i thought undervolting prevents tzs0 from reaching 98C. either way, the gpu will still hit 97C.

 

as per hydra's request, i've installed nvidia system monitor and recorded temps for when the system crashed. the first number represents the status prior to the crash and the second number represents the last entry in the log

GPU/GPU FB usage: 39% 35%
GPU/GPU usage: 84% 78%
GPU/GPU Shader Clk: 1500MHz 1500MHz
GPU/GPU Memory Clk: 799MHz 0MHz
GPU/GPU Core Clk: 600MHz 0MHz
GPU/GPU Temp: 91C 255C
CPU/CPU2 Tj Temp: 48C 49C
CPU/CPU1 Tj Temp: 57C 52C

i don't understand why it froze. i thought the down-throttle point for the 9800m is 105-110C. i'm going to revert to 9c.10 and see if the ACPI will reach 98C.

 

you have a fault somewhere... theres no way your GPU ever got to 255 degrees C



Yeah start with new BIOS and we can everytually try everything until the cause of the problem shows. I dont really trust the .17 BIOS. Seems everyone is jumping in on them and its causing a few issues. But who knows, i never flash BIOS unless there is something it actually improves or fixes (and with the gateway BIOS they never do anything worth the flash)

 

Ain't that the truth

AlphaQ99, no guarantees your chip will just throttle back cleanly. The last reading of 255C may have been a random death bit before system locked?

The 91C was before you started stress test?

My old paint may hit 84C during a game when last checked...

 

the 91C was observed during gameplay.

ok, so i'm back on 9c.10. i disassembled once again to ensure the TIM got a good even spread. looking back at the log, the gpu reached 101C and held it for 45 sec before it locked up and i had to do a hard reset.

 

OK, bad GPU or heat pipe NFG. You said that the air will keep your coffee warm? Sounds like the heat pipe is good?

It's sounding like mainboard was not changed out? I would start a ticket, asap, stating overheating continues after last "repair". This is why I hate soldered on GPU's so got a three year just for that reason.

It sounds like you've done your part, maybe ask for a model upgrade if you can stand the wait.

 

yeah, it's just not the answer i wanted to hear. the computer was working great up until a couple of months ago. i do appreciate the quick replies and the good suggestions. and hopefully my issue helps some other people that are having the same issues. but i think i'm going to wait until this deployment is done. i'd rather not go without having a computer. it's my main source of entertainment.

 

Yea I dont like those temps, can you downclock while gaming? That would lower performance a bit but it would lower the temps as well.

Try keeping the clocks low for now, thats about all that can be done, those temps are OK, but much higher then that it starts to get dangerous. Have it sent back and have a MB swap performed ASAP, which could be a while seeing as your deployed.

Best of luck.

 

i've re-applied copious amounts of arctic silver 5. i have downclocked the gpu by 90% using nvidia system tools. i didn't like the performance i was getting. i changed the settings back to the default clocks and just lowered in-game settings.

on a side note, how many 7811fx users play their games at 1920x1200? if my gpu goes over 85C i usually play at the next lowest resolution and adjust other settings (AA, anistropic filtering, etc.) my goal is to play games around 60fps and make sure my gpu doesn't get fried.

 

There is something wrong with these graphics chips, I am thinking.... I was playing Warrior Within on the highest settings on a P-7811FX from Gateway, and all of a sudden.... my computer locks. I thought "Okay..... memory leak issue!" and restarted my computer and after the SAME EXACT AMOUNT OF TIME.... locked up again.
This time, I checked the air coming out the side of the computer...... burnt my finger, LITERALLY. I download GPU-Z, have it logging the GPU temperatures.... before it fails, it is only at 90 Celsius, which is...... low for a failure point.

Try WALL-E from Disney, which is supposed to be LIGHT on the GPU... same thing at highest settings and even when I set the settings down to 800*600, the GPU is STILL near overheating.

Try Ice Age 3..... no problems, the air is so cool coming from the GPU and CPU that it feels ROOM TEMPERATURE!

Something is either wrong with some games or something is wrong with the chips for these things....... I recently went back and played Bio-Shock with the highest settings (1920*1200, everything highest except anistropic filtering).... no problems. It ran hot (85 Celsius), but never hard locked up on me.
Crysis....... 1280*1024, no problems, with almost everything on 'High' settings.

I"m beginning to think that it is some games that don't like the NVidia graphics cards or have incompatibilities with NVidia graphics cards of somekind.

Unfortunately, I cannot send my computer back to Gateway for a checkup, because it is 'out of warranty period'. I've also checked for dust.... none there now and it still overheats. Checked for broken fan on GPU... nope, fan is working fine. Check to see if the GPU is 'unseating' or if the heat pipe from Northbridge to GPU is loose.... nope.
I'm not willing to try and reapply thermal paste myself, because I didn't do it right one time on another laptop I owned, and busted it. If it's a laptop I fix at work.... que sera sera, there is usually a warranty thing on it, but not on my own computers.

 

Christopher - the GPUs on most of the 7811-FX's are definitely faulty.

If you're getting lockups not related to heat issues then this fix I posted might work for you:

http://forum.notebookreview.com/showpost.php?p=5404773&postcount=36

 

Thank you for this fix. I'm actually beginning to think it is not the GPU that is faulty but the drivers for notebooks from NVidia. Why? Because I was 'brave' and went into System Settings and FORCED the 'stock speeds' that the GPU was supposed to be running at according to Gateway through "Enable Driver-Level hardware overclocking" in RivaTuner (notice: I allowed RivaTuner to detect the stock settings)...... and the GPU stayed cooler than anything and I got some more 'speed' out of my processor which I wasn't getting with the 'Performance 3D' setting that you suggested (notice to everyone: THIS MAY BE DANGEROUS IF YOU DON'T KNOW WHAT YOU ARE DOING!).

I'm willing to bet my liver, spleen and heart, to be blunt..... that the drivers are OVERCLOCKING the GPU somehow or shutting down some 'protection' and that was the reason behind the black-screening.
I'm getting low 70C temperatures now with a big boost in performance, and the people who I talked with on Tom's say that is the usual temp they get when they are running desktop GPU's that are NVidia 9800-series.
Heck, I was even able to run Quantum of Solace's Demo at 1920*1200 with EVERYTHING but Anistropy turned up as high as it would go..... and the GPU STILL stayed cool.

Sure, it was slower than anything (8-15 fps when nothing much was happening onscreen) but it proved that this GPU is capable of MUCH MORE than I was previously getting out of it. Some of the people I talked with were shocked that I even got that at that high of a resolution at those settings without frying my GPU (it stayed at a little over 75C).

I have now installed RivaTuner on all my systems with discrete graphics, including my parent's gaming computer (my mom is a BIG 3D exploration/point-and-click game fan) and I'm getting more performance out of them as well, and they are running ATI graphics.

 

Agree, this sounds more of a driver issue if Rivas's "Force constant performance level" cures a lot of your problems.

There has been a huge amount of variables when reading the overheating posts. Lots of press on defective NVidia GPU's but I lost track on this issue. Then we may have bad cooling systems due to lack of airflow to thermal pads NFG.

Certainly we play bug proof games that would never cause a lock up? I lost count restarting Witcher and a few others. In any case, when logging temps using Nvidia monitor during actual game play, I never exceeded 84C with or without random lock ups.

Slow protect is 100C, maybe we got slow protect "crash" at < 90C? Just so many variables as I repeat myself. I would love to see 10 of exact same machines running same game with drivers,but that is another tangent.

Good find nomios, I could not do much with earlier versions of rivatuner, I'll download and record temps and playthru when time allows.

 

I am writing this post to help anyone with gateway fx 78 series notebook overheating issues.

I started experiencing laggy graphics after owning my gateway p7805u for about one year. GPU temperatures reached 80-90 degrees at idle and rose to 110+ degrees while playing games.

I was able to get my temperatures down to a max of 70 degrees under load and idle at aorund 50 degrees. Some steps require that you open your laptop so proceed with caution and only follow the steps if you know what you are doing, research the topic thoroughly, and get someone to help you if you are unsure.

Step1
Obvious, but can make a big difference - get a can of compressed air and clean all the dust out of the exhaust/intake ports located on the left side and rear of laptop (underneath draws air in, then it is expelled out the side/back) If there is a lot of dust it will dramatically affect airflow and cause higher than desired temperatures.

Step2
This involves opening the case and applying thermal paste to both the CPU and GPU. I used arctic silver 5 which is available at any radioshack for $10.

1. Open the panel on the bottom of your computer - this requires removing 5 screws with a small phillips head screwdriver.

2. After removing the panel you will see a cooling fan on the side of the laptop and the cpu and gpu is located to the side of it. To remove the heatsink from the cpu and gpu you will need to unscrew the 4 small screws around the proccessor and the one screw located on the side of the gpu.

3. After loosening these 5 screws you will be able to lift the heatsink straight up out of the laptop. ( it is all attached with a copper pipe that runs to a heatsink next to the fan - this is all one piece and lifts out together. Dont force it, it comes out quite easily once screws are loosened. The screws will stay attached to the heatsink and this is great because you wont lose them)

4. You will notice that the gpu uses a thermal tape (horrible) and the cpu has some cheap paste on it - wipe/peel both of these off of the cpu and gpu and the heatsink ( I used some alchohol pads, you can also purchase products designed specifically for this, also at Radioshack)

5. Let the surfaces dry, and then apply a SMALL amount of heat paste in either a dot the size of a bb or a small line on the gpu and cpu ( the gpu has a slightly larger surface area). You do not want an excess amount of paste - only enough to cover the polished copper surfaces after the heatsink has been refastened (really about the size of a bb dot or an uncooked grain of rice line).

6. Replace the heatsink (it drops straight down to replace and straight up to remove) and tighten the screws - you should do this a little at a time alternating between the screws until they are all tight in about 3 passes.

7. That is all there is to it - if you are unsure if you applied the correct amount of paste you can again unscrew the heatsink and lift it straight out - if there is any clumping on the heatsink around where the edges of the cpu or gpu raised copper surface ends, you applied a bit too much. The object is to apply as little as possible and cover the entire surface of the cpu/gpu. Depending on the heat paste you use, for instance arctic silver 5 is a bit thicker at first, it may not cover the entire surface immediately as this paste is designed to spread under heat and eventually dry out and provide maximum cooling after about 200 hours use - you will notice a HUGE difference right away though. If you applied too much paste clean it off and start over.

These two steps alone allowed me to run games on higher graphics settings than i ever had before without coming close to even the idle temperatures I experienced before.

Step3
This is just making sure your laptop is not sitting on a flat surface - it needs to be propped up somehow to allow air to flow underneath the laptop. Laptop cooling pads can be a good option although 99% of them ignore basic laws of physics and since air will find the path of least resistance do absolutely nothing to help in that regard. A good one is something like the cooler master infinite cool http://www.coolermaster.com/product.php?product_id=2581 that draws in air and forces it to the bottom of the laptop, letting the least amount of air escape from the sides/bottom of the cooling pad. I should however note that cooling pads will for the most part have very little effect other than to raise the laptop off of a flat surface and allow better airflow underneath it.

Following these simple steps should solve overheating issues I have seen so many post about and experienced myself. I would like to end by saying gateway support is horrible and they should just be honest and tell you that you are on your own. And if done properly, none of this should void your warranty. If you are unsure about the steps I outlined you can find more information in insane detail by searching for it. This post is mostly just to let people know what worked for me.

 

here, i wrote this for people but i guess no one read it, let me know if it helped

http://forum.notebookreview.com/gat...fx-goes-blackscreen-crashes-while-gaming.html