New Posts  All Forums:Forum Nav:

m1710 and heat issues - Page 2

post #21 of 63
Thread Starter 
Room temp is around 17°C, and the cpu and gpu is around 70-79°C on browsing and text editing. 88°C on 1080p films and 99 after a 3dmark.

Using NBF MobileForce 195.62 so no PowerMizer.

Guessing as mentioned opening it up and cleaning etc is the only way. Did get out the dust when the gpu fan was replaced last year by a tech.

But the symptoms have come back, stuttering hd films and programs that stop responding etc before it freezes up sometimes.
post #22 of 63
Quote:
Originally Posted by Versalius02 View Post
Hi all Thought I'd post in here instead of making a new thread.

I've had my M1710 since late August 2007 (bought new) and am now on my third 7950 GTX. Over the past few weeks I've watched the max temp during gaming sessions increase to the point where an hour long session of Oblivion last night pushed it to 102 C, maintaining at least 98 C for a majority of the time ingame. This is a game I used to max no problem on this sytem and not get up over 75 C. I began seeing artifacts last night on some of the textures, so I quit the game and looked at the temp history (as recorded by Rivatuner, which I left running in the background) to see the max was 102 with the card consistently at 98 and above for the entire span observed. My idle temp as I'm temping is 60 C when it used to be around 54 C. I've cleaned out my fans several times and have tried raising the back of the laptop up off the desk with erasers (this helped back when I first got the system) to no avail. Should I just be calling Dell to get things rolling now? I've been putting it off since I've got finals over the next couple weeks and don't really have time to deal with Dell, but if the system goes nuclear on me in that time...

Thanks for any advice.
Ummm one thing that happened to me when updating drivers is that the fans would sometimes fail to go to full. Your idles are on par with mine, but I never hit that high, the fans do go on though. It MIGHT be you have a leaky or bad heatpipe, but I doubt that very much. I prefer using drivers for this card that are in the years of which games this card was MADE to play, aka dx9 to very early dx10. Those have worked the best for me...although I think the bios controls the fans, it's always possible. I'm also running the a07 bios for the m1710 (same as the 2007 e1705 bios, same laptop), which might help you. It sounds like a fan thing, if the drivers don't help you and your still in warranty call dell and try to get an alienware :P.

If you're not, then I suggest taking the WHOLE laptop apart to clean everything, not just blowing air into it, thats really, really not good enough. There's guides on this site to help you, plenty of forums and dell even has online manuals. If THAT doesn't help, look for a new heatsink assembly on ebay for cheap (getting one for the m6300 precision 3600m gpu would be better than your current one), and putting arctic silver 5 thermal compound on the gpu and memory, it helps temperatures by in my experience a good 6 C. Using older drivers with coolbits (or any drivers with everything enabled) allows you to use powermiser which is an amazing tool for keeping temperature in check.
Hate to ask but are you overclocking? maybe you should consider underclocking the gpu a bit AND cpu (using RMclock utility, or smth similar like notebook hardware control), so as to limit frames by roughly 10% and it'll save you a fair bit in temperature.
Otherwise, it sounds like either your thermal paste has become shitty, or your fans are biting the dust...get it? haha.
As in the fans are either dying or too dirty still...just in case I'm dealing with a less than 50 iq guy.
post #23 of 63
Quote:
Originally Posted by bache View Post
Room temp is around 17°C, and the cpu and gpu is around 70-79°C on browsing and text editing. 88°C on 1080p films and 99 after a 3dmark.

Using NBF MobileForce 195.62 so no PowerMizer.

Guessing as mentioned opening it up and cleaning etc is the only way. Did get out the dust when the gpu fan was replaced last year by a tech.

But the symptoms have come back, stuttering hd films and programs that stop responding etc before it freezes up sometimes.
Wow awesome ambient temperatures....therefore....
This really sounds like bios throttling to me. the stuttering mainly. check your cpu usage when it happens with ctraltdel, then select view and select kernel times. If your cpu is 100% and that little red line is close to or AT the top, then it's definately throttling. (unless your cpu is from the 60s)
I've heard alot about ppl using windows 7 with the new drivers 150+ and the cards running a lot hotter than usual. I suppose thats fine if you have a 9000+ series card that can turn off when not used, but ours dont have that ability, so getting rid of powermizer is uber suckage. You sure it's gone too? there might be some advanced settings you haven't turned on, google around about it.

Unfortunately it's throttling to save itself from exploding, the only way to fix is to find why it is. Now is the cpu always hot too or just the gpu? cause thats a whole different thing if BOTH are acting super hot. Can you feel the fans at all? have anyway to force them on full like with the i8k fan control?

This is all to find out what exactly is the issue, but really, I think it's just dust. But who knows! want to make absolute sure.
Mine overheats every 4 months from dust buildup, and I gradually lower my clocks throughout the year until I finally clean it.
It builds up really fast, mostly cause the air tends to blow through the inner (towards center of notebook) part of the heatsink the most, so thats where all the heat exchange is, thats where all the air goes, thats where all the dust goes, and once its blocked, things get really bad fast.

Go here.
http://www.notebookforums.com/thread232373.html
post #24 of 63
To both of you really try to get powermizer working, it's an amazing help for idle temperatures, I go down about 10 degrees with it enabled.
post #25 of 63
i use this to make sure its enable
post #26 of 63
Thread Starter 
Thanks, going to try it.

As ive read that stripping it down is the only way. But after watching the tech taking it apart last time it doesn't look fun..

Did a test of 20-30 min in 3dmark and reached 96°C in a cool room on raised stands.
LL
post #27 of 63
Quote:
Originally Posted by matchbox2022 View Post
Ummm one thing that happened to me when updating drivers is that the fans would sometimes fail to go to full. Your idles are on par with mine, but I never hit that high, the fans do go on though. It MIGHT be you have a leaky or bad heatpipe, but I doubt that very much. I prefer using drivers for this card that are in the years of which games this card was MADE to play, aka dx9 to very early dx10. Those have worked the best for me...although I think the bios controls the fans, it's always possible. I'm also running the a07 bios for the m1710 (same as the 2007 e1705 bios, same laptop), which might help you. It sounds like a fan thing, if the drivers don't help you and your still in warranty call dell and try to get an alienware :P.

If you're not, then I suggest taking the WHOLE laptop apart to clean everything, not just blowing air into it, thats really, really not good enough. There's guides on this site to help you, plenty of forums and dell even has online manuals. If THAT doesn't help, look for a new heatsink assembly on ebay for cheap (getting one for the m6300 precision 3600m gpu would be better than your current one), and putting arctic silver 5 thermal compound on the gpu and memory, it helps temperatures by in my experience a good 6 C. Using older drivers with coolbits (or any drivers with everything enabled) allows you to use powermiser which is an amazing tool for keeping temperature in check.
Hate to ask but are you overclocking? maybe you should consider underclocking the gpu a bit AND cpu (using RMclock utility, or smth similar like notebook hardware control), so as to limit frames by roughly 10% and it'll save you a fair bit in temperature.
Otherwise, it sounds like either your thermal paste has become shitty, or your fans are biting the dust...get it? haha.
As in the fans are either dying or too dirty still...just in case I'm dealing with a less than 50 iq guy.
Thanks for the reply.

I've been running the A07 BIOS since it was released, coupled with the Dox 182.46 drivers (best performance so far). I've heard the fans kick into high gear even just watching 720p movies, it's just they do it so late that by then they can't seem to make a dent in the temps. My M1710 is on its third GPU in its life span with this one installed about 13 months ago. During the first about 10 months the temps never got above 85 C (except when I tried the Batman: Arkham Asylum demo - my fault for thinking it could handle it at 1200p), it's just in the past month or so that I've seen the temps steadily rise until I notice my hands getting cooked on the keyboard.

Here's a screenshot of HWMonitor after 30 minutes of Oblivion at slightly reduced settings (1680x1050 instead of my normal 1920x1200). After an hour of Oblivion (which I used to run maxed out no problem), the system shut itself down and wouldn't start back up for a couple minutes as it cooled off. Upon startup, the screen displayed a message basically saying it had heated up to solar levels. So far, no artifacting outside of games though. The first time my GPU fried, I was watching a SD video when half of the screen became completely discolored, in my case mostly greenish hues. I just hope that doesn't happen for at least a week as I start finals tomorrow.

Edit: Forgot to mention a couple things. I'm still under warranty until late August (3 yr Complete Care) and Powermizer is definitely working as according to Rivatuner both the core and memory clocks are at zero right now with 2D settings at 200/300 and the 3D settings I have at the stock 575/600.
LL
post #28 of 63
yes well its a dell... i have noticed that myself, that as the gpu fails the temp will progressively start to reach higher and higher max temp until it finally fail for good
post #29 of 63
Quote:
Originally Posted by ataxy View Post
yes well its a dell... i have noticed that myself, that as the gpu fails the temp will progressively start to reach higher and higher max temp until it finally fail for good
Do you know why that is? realistically theres no reason for it too I thought.

Anyone else around that might know? I don't know where zzpulp has been these days, he seems to really know this stuff well.
post #30 of 63
Quote:
Originally Posted by bache View Post
Thanks, going to try it.

As ive read that stripping it down is the only way. But after watching the tech taking it apart last time it doesn't look fun..

Did a test of 20-30 min in 3dmark and reached 96°C in a cool room on raised stands.
Given that your cpu AND gpu are reaching rediculous levels, I'm really thinking it's dust, taking it apart and using compressed air and isopropyl alcohol is the only good way.

It actually is fun, the first time ever it's a little odd, but it's the same with sex, then you can't wait to do it again.

Sex is a little better in my opinion though.
post #31 of 63
Quote:
Originally Posted by Versalius02 View Post
Thanks for the reply.


Edit: Forgot to mention a couple things. I'm still under warranty until late August (3 yr Complete Care) and Powermizer is definitely working as according to Rivatuner both the core and memory clocks are at zero right now with 2D settings at 200/300 and the 3D settings I have at the stock 575/600.
Your temperatures for the gpu are waayyyyy to friggin high. To the point it won't gradually kill it, it will kill it quickly if you don't get this settled. 104C is shutdown temperature for this card according to its bios.

If you use rivatuner and newer drivers, use rivatuner to downclock the card to 200/300 ALL the time (including 3d mode) and set the 2d mode even lower, it can go down to 100/80 fine, even 50/60 is stable (but you get some errors if trying to play anygames, not synchronzing with the screen correctly), any lower and you get shutdowns.
But you need to do this in the meantime until you solve it, above 95 is bad, hitting threshold is unacceptable if you want it to keep running.

Interestingly your cpu seems fine for temperature, which makes me think either your gpu fan isn't doing as well as it could (aka. dying), your heatsink assembly is somehow damaged in someway, or the most likely thing, DUST! and a LOT of dust. I bet you'll find a carpet in there. You should really take it apart and put AS5 on everything, you can buy arctic silver 5 for just about 15 bucks. We can help you with using it on here too if you need help.
Definately go open up your laptop and get this dealt with, if it's dust and you've never done this, it's an afternoon thing. Really it isn't that hard to do.
http://support.dell.com/support/topi...dhs&~ck=anavml

That will bring you right to the place you need to go to find your lappys service manual. The hardest part is the hinge cover, and thats just fear out of breaking it, which you can't.
post #32 of 63
Quote:
Originally Posted by matchbox2022 View Post
Do you know why that is? realistically theres no reason for it too I thought.

Anyone else around that might know? I don't know where zzpulp has been these days, he seems to really know this stuff well.
well a bit like any over stressed circuitry it overheat until it blows

as for is temp well i am not surprised the max i have reached on mine and this is with a clean laptop is 90c-95c playing L4D and 100c-105c trying Batman arkham asylum but i also remember when i got it playing assassins creed and temp not going higher then 80c full res and detail to the max that it was playable.

anyway this laptop is badly conceived the fan on the right side should be as big as the one on the left side , as i have said the heatsink should have been made of copper and mostly the stupid and i am sorry but the stupid engineer had to go and sandwich the gpu in between the keyboard and motherboard with no air coming straight to it, in my point of view its as useless as blocking the air vent, but at least they tought about putting colored led for the vent .

honestly if it would be a car lots of people would have died
post #33 of 63
true dat, they need more airflow to the gpu, don't get why they didn't use 2 equal sized fans. I've got a semi working setup right now where ambient air is being drawn past the video card, but it's not quite working right and don't have enough time to do much about it until august. Well...who knows. But yeah why would it get hotter during time even with cleaning? and I do mean proper cleaning.
post #34 of 63
as i said failing circuitry tend to overheat as it fails
post #35 of 63
Thread Starter 
Took the plunge and opened the 1710, cleaned fans and the cpu\gpu. Didnt have any arctic silver so couldnt try that.

Attached an image of temp after startup with only browsing.
LL
post #36 of 63
yep still pretty high, are you using i8kfangui?
post #37 of 63
Thread Starter 
no i8kfangui, ive read the thread getting it workin on win 7 . http://cdiefer.proboards.com/index.c...lay&thread=664

But not tried that yet. The cores are around 50-55ºc, and gpu 53 ºc on light browsing.

The cleaning seems to have lowered overall temp by 10-14 ºc
post #38 of 63
so dust was part of your problem

now install i8k and you should be able to shed another 10c with proper setting
post #39 of 63
your cpu is strangely a wee bit high, I think that's just cause you have it sticking to the highest multiplier when browsing and highest available voltage. (Dell does that by default to keep bsod from happening from undervolting)

Don't know if it works on win 7 yet, but RMClock utility is pretty much the best thing you can use to configure your cpu of all time. And of course it's free.

Your gpu is right on with mine when I'm not using powermiser. sounds like it's running much better. When I use powermiser I drop about 4 C. Right now with fans on full idling, I'm running my gpu at 42C, and when I turn on powermizer that'll go to 39C.

Really though, your gpu at 53ish is perfectly normal for this laptop, so long as it's not hitting 100C at any point it is fine.
If you want to lower your case temps a little bit, give rmclock utility a try.
RMclock and powermizer have the added bonuses of not having to have your fans on full all the time as with i8kfangui...although whoever made that is amazing.

Interesting to note, I tested my lappy with i8k, and on full it works well, but according to speedfan, when I manually set the speed in the bios to full, I get 4600rpm out of the fans. i8k is giving me close to 4000. That requires having overclocking ability though.
post #40 of 63
Thread Starter 
well, the first thing i noticed was that the left fan is no longer on full. So its much much quieter, actually able to play 720p video while using word with no fans disrupting. After 5 hours is 55 c at idle, and gpu at 56c.

So opening it up helped a lot. Next step is maybe apply arctic 5. ( thats taking out the left heat sink ?) As someone said this would not affect the warranty?

Anyways, it has been opened twice before by techs, changed lcd and fan.

Got some months left on nbd service and consumer warranty.
New Posts  All Forums:Forum Nav:
  Return Home