Processor Heat

Okay so I have an np2090 with a T7300 processor, and I've been doing some apophysis renders (pure cpu) and found the weirdest thing while monitoring my temps:

When I tilt the computer 90 degrees so the back of the screen is on the table, the temperatures skyrocket to 85-86C, and the fan goes to 100%. When I have the computer upright, tilted the other way, or upside down, the temperatures go back to reading 65C within two or three seconds.

What's wrong/should I be worried?

 

Are you blocking the air vents in the back?

 

No I am not

 

You're not blocking the vents - but are you restricting the air-flow through them? In other words, if you tilt the laptop over in the same position you described, but hold it in mid-air in that position instead of letting it sit on the table, do you still get the temp increases?

 

my laptop get a higher temps when its directly on the table.. its normal.. but 86degres is a little high i think...

maybe you should invest on a notebookcooler

 

if you are blocking either the intake (on the bottom) or the exhaust (on the back), you will lose fan cooling. i have an fl-90, and putting it 90 degrees on a table would certainly block the rear exhaust... put your left hand back behind the computer and block the exhaust vent- i'll bet you get a similar increase in temperature that goes away when you move your hand off the vent.

this applies only if you DON'T have a 9 cell... with a 9 cell the back wouldn't be flush so it wouldn't block the air. woot!

 

I'd recommend a notebook cooler either way, especially for intensive stuff. Really helps with extending the life of your hardware.

 

Sounds like a real problem. I`m idling 35-40 C for the CPU,max temp I saw was 80C on a stress test with Orthos (damn that thing tortured my CPU). Either something is not properly placed inside or the cooling is faulty. 60 C idling is BAD !

 

Wait something's being misunderstood...

The vents are NOT being blocked, this is a matter of the orientation of the laptop ONLY. If I pick up the laptop the temps skyrocket in a matter or two or three seconds to 80+C when I hold it so the back of the screen is facing downward. If I hold it in the air any other way (vents unblocked) the temps quickly (two or three seconds) run down to 60C.

This is all under CPU load-- I'm idling at 45 in a warm room.

 

why are u tipping ur laptop? why not just leave it in the position that it doesnt get hot?

 

All he's saying is that it's not supposed to happen. Either position he may want to chose the temperatures should remain the same. My belief is that something is not sitting right inside his laptop.

 

Ditto. Either the temp sensor is drunk or there`s something loose in the fan department.(and I don`t mean no ball game)

 

Thanks Fade to Black/eleron, you hit it spot on.
So finally, is there anything I should/can do?

 

Call the company you bought the laptop from, especially if it's still under warranty. It sounds to me like the heatsink/fan on top of the cpu are loose, so they sit in place and dissipate some heat when the laptop is in proper orientation, but move out of place when you tip the laptop. You'll probably have to send it in so they can take it apart and re-seat the components(if this is the problem), but going without it for a week or two is probably better than a) voiding your warranty or b) making your heat problem worse.

 

i suppose your heat sink or fan could be loose. have you considered just opening it up and taking a look? it would not be hard to see.

here is a link to the ifl-90 disassembly guide...
http://www.asimobile.com/notebook_files/fl90/IFL90_ch5-Disassembly Guide-intel.pdf

this will make it easy for you. you could open it up and see if anything obvious is going on. word. i am, of course, not
responsible if you break anything, don't go opening your laptop if you don't feel comfortable doing it.

 

Ok. I just asked about the on-the-desk as opposed to up-in-the-air because even if the rear vent wasn't blocked, if it was in close proximity to the desk-top surface the airflow could be sufficiently restricted to cause heat build-up; however, since you've confirmed that this happens even up-in-the-air it cannot be a restriction caused by an external object.

Another question - does the fan make any funny noises, either when it comes on, or when it runs at either low speed or high speed? I'm thinking of something whiney, like a bad bearing. Also, do you hear any changes in the fan noise (e.g., a whine, or even a change in the fan speed) when you tip it over on its back? It's possible that the fan has a bad bearing, but that the bearing isn't loaded on the bad spot until it's tipped on its back.

Also, can you actually feel the system getting hotter? Does the air coming out the back feel hotter, or do the keyboard or the case/palmrests get noticeably hotter? It's possible that there's a loose connection somewhere that causes a false high-temp reading when the notebook is put on its back.

Another possibility is that something inside the laptop (such as a loose or overly long cable, a stray piece of paper - maybe a label that's come loose, or the like) is shifting around when you put it on its back and is restricting the air-flow within the notebook.

Along that line of thinking, it could even be a dust-bunny shifting around and blocking air-flow when the notebook is on its back. Have you been cleaning the vents and fan as per the guide here: http://forum.notebookreview.com/showpost.php?p=1439598&postcount=1

Also, it could be a bad connector on the wiring from the fan unit to the motherboard.

Lastly, and worstly, it could be a flaw in the motherboard or the printed circuitry, in which case the only fix is replacement.

For now, you should get the disassembly guide for the notebook (000111 has posted a link to them), open it up, make sure the vents and the fan, and the interior generally, are clean and clear of dust or other blockages, and then check for any loose wiring, connectors, or etc.

 

000111: I just took the large plate off the bottom, wiggled some stuff around and checked the screw tightness, everything seems in order. The fan blades have the thinnest coating of dust ever, but nothing's in the heat sinks or anywhere else.

Shyster:
Wow thanks for the detailed reply!
Fan seems and sounds fine, it does speed up when it reads the 80C, probably because the BIOS is programmed to do that. I think there's just something wrong with the sensor or sensor placement.

The heat of the laptop itself feels the same (it takes two seconds for the temps to jump to 80 when held particularly, the CPU load remains the same, etc.). I doubt it's an actual heat jump.

I did and no blockages. Looks clean to me.
I guess the question now is whether or not my CPU actually hits 80+C. The cooling seems fine, and oriented normally it reports about 66C under sustained full load on a desk, and 45C idling, which seems more reasonable.

In any case I don't think it's worth sending it in.

Thanks for the links, they were interesting and helpful!

 

hi guys; is it normal for my np2090's t7300 cpu to go up to 88 degrees Celsius when under stress? i've been playing around with undervolting and all i could get is 79 degrees at 1.150 VID without running into a nice BSOD. aren't these temperatures too high?

 

daemmon, I merged your thread with this one since it deals with the same topic.

 

Daemmon , those temps are a bit concerning. Check for dust clogs, clean the fans , keep the notebook on a solid surface, raise the back so that the air under the laptop will properly flow and you might want to check the temps again after that. If there`s no change, I would contact your reseller. If temps will drop, I think you`ll already have your answer.

 

Daemmon, it's kind of unnerving that you're having a similar problem. I haven't tried undervolting yet.

Out of curiosity what happens at a sustained 88C? Isn't the max operating temp of the t7300 100C?

 

form what i can figure out it goes into throttling, because as soon as it reaches 88.0 it starts going down to 86 and then rising again to 88.0, never exceeding this value

 

It does sound like it may be a sensor problem. Which of the temp monitoring utilities have you been running? It might be interesting to run a comprehensive monitoring utility, such as cpu-z or hwmonitor (available from here: http://www.cpuid.com/hwmonitor.php ) and then take the system through the steps of putting it on its back and then returning it to a normal position while the utility is running. If the utility manages to read cpu core temps from a different location than the monitor you're currently using, that could be proof that the flaw lies in the circuitry your current monitor uses.

On the other hand, even if another monitoring utility reads the same temperature info (and thus cannot distinguish a real temp change from a spurious one), it should still be reading other variables such as core voltages, frequencies, and throttling - if the monitored temp hits a high, but none of the other variables show a correlating change, that would suggest that the core is not implementing its on-die temperature reduction measures, which would in turn suggest that the temperature readings you're getting are, in fact, spurious, and that the flaw lies somewhere in the circuitry that is supposed to read off the core temps.

Finally, if you're really adventurous, and have a bit of spare money and time on your hands, and you've gotten tired of re-sorting your sock-drawer , you could open the system up enough so that you can see the cpu and its attached heatsink, start the system up, and then use an infra-red thermometer to read the temperature of the cpu/heatsink system directly while you move the system onto its back and then back upright. If the temperature you read this way doesn't show changes that correlate with the reported temps in your monitor, then that would indicate again that the temperature monitoring circuitry is faulty.

Long-term though, I would think about sending the system in for evaluation, particularly if it's still under warranty. Even if the problem is just a faulty temp sensor, it may eventually worsen to the point where you cannot rely on it at all in any position, in which case you could run the risk of not catching a real thermal event that ends up cooking your cpu.

 

ok, i've opened up my laptop and it seems the heat pipe had accumulated quite a bit of dust; i cleaned it and now the temperature under stress reaches a maximum of 71 degress undervolted (that is in 15 minutes of orthos stress prime), as opposed to 80 degrees before the cleaning, and rises quite slowly. Also, when idle but with winamp, yahoo messenger, google talk, utorrent, nod32 and a few other low demanding applications running, the temperature varies between 51-52 degrees. perhaps a better cleaning (i did not unmount the heat pipe, nor the fan) would have yielded a better result.