Precision M6300 overheat

Question

I recently upgraded to Core 2 Duo X9000... the temperatures went up (as expected, if compared to the previously used T7700). I was pretty confident this particular laptop is able to handle this... laid fracking $3000 back then... It really annoys me how it just shuts itself off now during work, where you, Dell, are supposed to help me get the bucks with this awesome tool. Yes, such useless things like games run fine and it never shuts down. I will explain further. So, I ran a temperature monitor, and I think I see the problem... The readings labeled 'THM' go up to 105 degrees Celsius during short multitasking system load spikes. In these particularly short and intensive spikes, the fans kick in waaay too late and lower their speeds too often in some false system load drop anticipation (for laughable half a second... what the?). Due to this, sometimes THM temp overshoots 105C shutoff threshold (from what I've read at Intel, that should be 'the limit' for motherboard and CPU, and THM is the motherboard, I presume?). Anyway, I've never seen THM go higher than 105C and the max reading is always at 105C, so the overheat is rather conclusive. THM temperature varies like crazy. From 60 to 105 in just few seconds and drops back to 50 in same eyeblink. This never happens, when the fans are always running at high speed. Aka, if the system load is constant, it never shuts off, so the Dell resource CD tests are useless, since they never test for varying multithreaded load. Dust cleaned and fans checked. Idle temps are: THM: 52C, CPU 46C, GPU 72C Maybe Dell can patch this up in the BIOS? Or can the fan speed adjustments be made less susceptive to temperature drops and more aggressive against rising temperature... ? This issue just feels so trivial.

gl1koz3 · Accepted Answer

I tried a program called CPUgenie. Set it to undervolt from 1.2V to 1.067V at 2.8GHz state. This effectively dropped CPU temperatures by 20 degrees Celsius and shutdowns stopped. Been stable for around 16 hours now. Tested with all sorts of things.

If the CPU was idle for a while and CPU usage suddenly spikes to 100%, the cooling fans are sometimes slow to respond. It waits until CPU gets to ~80 degrees Celsius (LOL!) and then kick in... and even not to the max speed. The cooling tubes are rather long in this laptop, thus cooling lag is BIG. Why is the threshold so high? And why is the cooling policy so lax in this workstation?

Even at this undervolt, the CPU temperature graph is somewhat like this:

time - CPU load - CPU temp - fan RPM (noise level)

0s - 0% - 50C - low / stopped,

1s - 100% - 86C - low,

4s - 100% - 88C - low,

8s - 100% - 90C - low / medium

20s - 100% - 94 - medium

30s - 100% - 94 - medium

40s - 100% - 96 - medium

120s - 100% - 96 - high

200s - 100% - 93 - high

600s - 100% - 91 - high

The irony of it... it never goes to max speed. (It sometimes does, but just by extra short periods of time during some apps loading or something). But when the cooling is really needed... like at 80-90 degrees, it feels flegmatic to respond. And this spike caused my laptop to ALWAYS reproducibly shut down after around 40 seconds before the undervolt, that dropped the temps by 10-20 degrees.

gl1koz3 · Answer

Not a single idea? I'm wondering if normal use temps THM 57, CPU 50, GPU 73, HDD 44 all Celsius is even considerable as 'OK'.

Locked Topics – Laptops General

Was this post helpful?