Issues of M17

Update. Problem 2 is fixed, read below.


Update on problem1.
-Tried vista32, same temperature.
-Tried new ATI driver, same.
-Other M17 owner with exact same spec with me, but way lower temperature.



I finally find out the cause of the issue of the heat myself. Yup, the AW TS was no help at all on this issue, only wasting my time and make me look like an idiot when I was trying to convince them it's not the ERROR. Finally I find the cause and call them again. Kevin from Roswell understand the issue and the cause after 1/2 hr on the phone (which is way faster than other techs). I will continue to update and REVIEW their support service, which would help the potential AW buyers on making decision.


Hey, guys. Some of you already know the problems that I found with my new M17. So I just want to start a new thread so that (1.) I can keep documenting the process for my own use, (2.) for other new owners to confirm my findings, and (3.) hopefully, I can find the solution.

Problems:

1. HOT GPUs. The picture says all. All 3D games would make the gpu go to 100+ in few minutes. The GPU doesn't slow down at this insane temp.
Didn't try other drivers yet, it's the latest. but the point is the fan is not cooling them at all. The GPUs are at 50 degree on idle 2D. and the fans are trying to cool it 24/7, but can't. The air coming out from the fan are not HOT. Check post 7 for more findings.

http://img98.imageshack.us/img98/3618/hwmonitorprohightemptv9.jpg

2. Fake Dead (Bright) Pixel after driver loaded, and I can't get rid of it.
Right out of box, it's there. It's on top of every window/application.
After trying to fix the problem by messing with many driver tool and app and many reboots, the pixel moves to another place, but still there. Sometimes if a driver tool refreshes the screen and take the background and colour bit settings away, the pixel is gone. Once I get the background and colour back, it comes back at the same place. Refresh method cannot be reproduced 100%.

http://img56.imageshack.us/img56/7258/pixelbd7.jpg

Issue 3:
Paint peel off. No, it's not the wear and tear theory.
This happens to my second M17, but not first M17, and I treat both M17 the same as a baby.
On this second M17, only the top palm rest paint is off. Other parts that are even sharper doesn't peel off.
I had my first M17 for 1 month - no peel.
I had my second M17 for 1 week - start to peel.







Talked to Roswell. He has to talk to others to find out the problem. He says it's a C-panel, which requires to take out motherboard to replace, so I would have to send it in to get it repair.

Issue 4:
Closing the lid doesn't make it to go to sleep/shutdown/hibernate in fresh vista and fresh windows 7.
Yeah, I did do the settings, just in case you suggest.
Seems like some connector is off.
Roswell will give me a call tomorrow about it.

UPDATE: on issue3
Roswell told me that they couldn't tell from the pictures if the paint is peeled off by itself or by damage done by me. So I will have to send back the laptop to let them inspect to see. And if they think the peel is done by me, then they will have to charge me.
I told them there was no scratch or dent or any damage to the laptop, the paint is just peeled off by itself, but there is no way I can prove it. It's all at their discretion.
On top of that, before they gave me such answer, I told them I wouldn't send this back for 3 weeks to repair anyways, I just want to report to them that there are such problems. I guess they didn't take that as in good will. I told them I would ask for repair if the problem gets worse or the peel gets very big.

 

Problem 2 is solved.

http://4help.alienware.com/cgi-bin/...TImcF9zZWFyY2hfdGV4dD1tMTc*&p_li=&p_topview=1

The cause of problem 1 is found by myself, not by the AW TS.
Since I spent minimum of 2 hrs on the phone with AW TS and seeing that I am getting no where, no matter how many times I phone in. So, I took the chance and open up the laptop and find the fault on their part. I am not going to discuss the specific fault, because this is probably an individual case. They promised to get me full refund on this meltdown system, and have a new system for me after days of bargaining. Will update further when I finally get the money back, because they ask me to buy 2nd system and return the first back.

Update Dec 10, 2008:
I just got my new M17 replacement, the test record of the new one is 11157 in 3dmarks06 while the old one was 10861. Both are installed by the same guy.
Assembler Initials: RR
Tester Initials: AI
Will test further for stability and temperature.

 

phew, 100°C is too hot, definitely. have you tried different GPU drivers? also, does the white dot only appear on the wallpaper, or on other things as well?

 

didn't try other drivers, it's the latest. but the point is the fan is not cooling it at all. the GPUs are at 50 degree on idle 2D. and the fans are trying to cool it 24/7, but can't.

the dot is on top of everything.

 

ok. can you feel hot air coming out the back? come to think of it, can you even hear the GPU cooling fans spin&blow?

 

Is only one GPU being loaded in that image, or are they both under load, if they are, one is still 50*C lower than the other.

How warm is the air coming out of it?

 

Base on my findings and tests on the HOT GPUs, here is the possible problem.

1. The dude, who install the video cards and heatsink and fans, screwed up and didn't put thermal pad in between, or the fans are install backward or something. It should be thermal related issue of the installation rather than the GPU, because the chance of BOTH GPUs have same problem is small.

Or

2. Vista64 screws up the sensor driver for the temperature for these GPUs. A Vista32 might solve the issue.

 

Yes, i mention many times before the fan are spinning HARD. only warm air coming out, which brings me to post#7.

 

i'm leaning towards 1). the other option would be more BIOS-related (as seen with the m15x and the x36 BIOS), and other users of this NB haven't had this issue.

 

The screen is just an example of over 5+ different games tested.
In this example, only 1 gpu is loaded because of atitool is in windowed mode.

 

can you feel hot air coming out the back? can you hear the fans spin up when the GPU is stressed?

 

Installing a fan backwards wouldn't work, they're only designed to fit in one way in most notebooks.

The thermal paste could be the issue, too much or too little could affect it.

And Vista 64 could also be the problem. If you have a spare 32 bit version or XP lying around maybe installing one of those would be a good choice.

 

Yeah, but I would want AW to replace the GPUs too because they have been at over 100 degree for many times through out the tests, probably more than 30 minutes in total. They will die soon.

 

Have you called Roswell yet and asked them about these temps?

 

Does anyone know what the thermal limits are for the Radeon cards?

For Reference, the 8700m's in one of my M9750's hits 100 degrees c regularly and Alienware says that is normal for those cards and not to worry unless they reach 110 degrees c. Since the 3870's and 8700's run similar clock speeds perhaps you are in the safe range for those cards. The only way to be sure is to call TS and see what they say.

 

They'll probably take 100*C fine, although running at 100*C isn't going to be good for anything.

 

Most likely it's not ok, I reach 110 degree in less than 3 minutes of CoD4.

If I play longer, it might go over.

 

Does roswell read forum?

Can I send him this thread?

 

The Roswell Crew is a customer service group you can email them at [email protected] or you can just call them up and explain it.

 

I am on the phone with them, and they seem they are from the States.
Will update soon.

 

dude maybe your hwmonitor is not calibrated well. If your card went to 110C it would fry in just seconds.

 

Roswell says the temperature reading is not correct.
and there is no program right now can read these cards correctly.


I wonder why the OCZ guys can .

 

Maybe check it with Rivatuner and Everest.

 

rivatuner doesn't work, i will try everest.

roswell says they tried everything.

 

Everest shows the same.

Another thing is , u see the different CPU temp reading from HWmonitor and everest?

 

Hmm considering thats 3 reading a hot temp, i have a feeling it is probably that warm.

CPU temp will probably be different because Everest uses a different T-Junction it should be at 100 for C2D CPUs i guess Everest is wrong. Core Temp and HWmonitor can be trusted, they read the same temp on ym desktop.

 

The worst case is that, I use it for few months and request a change on the graphic card, because other users do not have this high temperature.

 

Why worry about the high temperatures? As long as you don't see any glitches in your games or benchmarks, that means the GPU isn't damaged. As long as it isn't damaged by the heat, then it's fine.

 

Because I had 2 video cards that died on me RIGHT after warranty.
And I don't this to happen to me again, or at least lesser chance.

 

But afair these are NOT normal temperatures for GPUs like this, besides when GPU is skyrocketing to 110 deg. Celsius it might affect everything around it...

I completly understand you Fusion, hope you`ll solve the problem soon.

 

thx man... it's taking me hours/days to test everything... man.

 

I still wonder one thing. In my good old Toshi A100-002 with GeForce Go 7600 GT everytime when temperature of GPU or CPU is too high system is shuting down with a blink of an eye. I had it when my fan was too dusty and did not manage to cool everyting down properly.

Does M17 have this kind of sensor? Maybe it`s broken. Or maybe really temp readings are wrong and that`s why sensor doesn`t do anything? Do you feel very high temp on surface of laptop anywhere? Above average?

EDIT:


Besides, I think (hope?) that AW is making some stress test for laptops before shipping so they would notice so abnormal behaviour of GPU temperatures. It`s not likely something was broken during the shipping.

 

the surface of the laptop is not hot, and the air blowing out is not hot.

So it maybe the reading is wrong, my Mike's M17 with same spec reading is way lower.

So it can be no thermal pad is applied, so the heat is in the GPU core, so no heat is getting out and to the surface of the laptop.

 

It's generally built into the drivers which then tells the OS that something is wrong. Cards like this can usually take temps of 100*C but it's not good.

 

Ok, I install Vista 32, the temperature still go over 90 easily. I really feel uncomfortable about it, I will request them to send someone to exchange the GPU and the Heatsink for me, if that doesn't help, then I will request to change the system for me. I really don't want my invest to go down the drain 1 year later like my M1710. Am I asking too much?

 

fat chance. AW rarely (if ever?) sends out repair guys. you either have to do it yourself, or send it in.

 

new driver, still same.


Trying vista 32 , still the same temperature.

 

What game are you running?

 

Any game would result the same, I tried a lot.

In that scnshot, i just used ATITools.

 

Update: I find out that the graphic card does shut down itself at 110 degree for brief second.
Going to call AW.

 

First off, let me say that it's terrible that your brand new M17 is doing this **** to you. Mine arrives this week, and I'm praying for a solid setup. This is coming out of left field here, but have you thought of just opening it up and checking things out? Could be that there's a screw lose on a heatsink, or maybe the technician was preoccupied and slapped the heatsinks down with no thermal paste. Again, left field here. I don't even know what type of tamper seals AW puts on their stuff. My E1705 had nothing, but my 7950GTX had a tamper sticker over one of the heatsink screws. I have removed the stickers in the past with a blow-dryer. Good luck.

 

I don't know if the AW think I am voiding the warranty for doing that.

 

I am so pissed off because I have to spend DAYS to test it, and HOURS to convince there is something wrong which they don't seem to believe 110 degree is a problem.
Now, I finally find out the reason for the heat problem, but I am not going to disclose their fault until they give me a solution.

 

OK, I talked to Kevin in Roswell, he finally believes my findings, and acknowledge the issue. Now, I have to wait till Monday for solution/compensation.

The pictures above says a lot, but I will wait for him to give a call later how are they going to compensate/solve for these DAYS that I try to resolve the problem their guy created.

 

After my own temporary fix for them. here is the result of the heat.



No higher than 70 degree.

 

So what was the reason?

 

he relised not to use his laptop in the **** oven!

 

It's their human error. I will post the reason, solution, their compensation and support service review all together after they call me back later.

 

too much thermal paste perhaps.