Overheating problem!

[Xenonvector927]

Hehe, bet that got your attention...

At Christmas last year I put together a build with an E2200 (OC to 2.6 on stock) and the cores idle around 29C (Ambient around 24C). I have an ACPI THRM temp of a 70,71, and sometimes 72C. This is what is detected by the BIOS as the CPU temp. Now I now that Intel chips shut down when overheating above 80C or so, but my question is, because my mobo (P5N-MX cheapo board) detects the CPU temp as a constant 70C or so, if my fan failed one day and my CPU overheated, am I correct in saying that it would not shut itself down and fry itself and the mobo? BTW this is a known issue with this board - definitely not the actual temp.
BIOS updates haven't helped - I am nervous as when I upgrade in a few years this will be decommisioned as a HTPC and it will most likely be stashed in a cabinet or something, so I wouldn't notice if the fan stopped working. HELP!
Thanks, guys.

 

[exit2dos]

If you do put it in a cabinet, I would think about taking the back panel off the cabinet for better ventilation, as well as, getting a better heatsink/fan.

As far as burning your CPU, your CPU should throttle (get slower) if it gets too hot. If the throttling doesn't cool the CPU down enough (such as in your fan dying example), then your system will shutdown to prevent the CPU from frying.

 

[Xenonvector927]

If you do put it in a cabinet, I would think about taking the back panel off the cabinet for better ventilation, as well as, getting a better heatsink/fan.

As far as burning your CPU, your CPU should throttle (get slower) if it gets too hot. If the throttling doesn't cool the CPU down enough (such as in your fan dying example), then your system will shutdown to prevent the CPU from frying.

Thanks, but my question is that because the CPU temp sensor is stuck on 71C, if my CPU did overheat, wouldn't it not be detected because the temperature reading would still be on 71C, which is not going to kill the chip, and hence it wouldn't shut down and just continue overheating? Or is there something inside the chip itself? Thanks again

 

[SirCrono]

I belive, but could be wrong, that the chip itself shuts down if temps climb too much.

If I remember correctly, even if you disable all bios settings regarding thermal throttling the chip itself reads it's own temp from the same diods that coretemp or realtemp use and it's hard wired to shut down in the event of what intel defined as "catastrophic cooling failure", this is a feature that comes from the old PIII days.

Also, I once ran my e6300 to 120ºC (as reported by coretemp before the shutdown occured) and it didn't get damaged or anything (just so you can get some peace of mind), after that I checked my HSF and it was loose (pushpins kinda suck).

If you are interested in a beefier reading, http://download.intel.com/design/mobile/datashts/31407804.pdf, I recommend page 89.

 

[Xenonvector927]

I belive, but could be wrong, that the chip itself shuts down if temps climb too much.

If I remember correctly, even if you disable all bios settings regarding thermal throttling the chip itself reads it's own temp from the same diods that coretemp or realtemp use and it's hard wired to shut down in the event of what intel defined as "catastrophic cooling failure", this is a feature that comes from the old PIII days.

Also, I once ran my e6300 to 120ºC (as reported by coretemp before the shutdown occured) and it didn't get damaged or anything (just so you can get some peace of mind), after that I checked my HSF and it was loose (pushpins kinda suck).

If you are interested in a beefier reading, http://download.intel.com/design/mobile/datashts/31407804.pdf, I recommend page 89.

Cool, thanks for that, and the link. Case closed.