CPU overheating w/ AIO cooler?

[Alasdair]

HI,

 

So the issue is that when i run a stress test, like real bench, that uses 100% of my CPU, it almost immediately thermal throttles (going as low as 55-60% usage after 30 mins) and basically goes straight to 100 degrees (using a liqmax 240 CPU cooler).

The reason why i'm not sure about this is that, when my computer blue screens (whole other series of issues) it then boots to give me a CPU over temperature issue, but when i boot into bios, the CPU is roughly at 50-60 degrees, not the 100.

To boot, when i use realtemp to monitor the temperatures during the stress test, as soon as the test is stopped, the temperatures shown go between 100/98 degrees and 70-80 degrees for all cores randomly. It's as though the sensors aren't sure if the CPU cores are running near the MAX temp for the CPU or if they are at the usual temperature range for CPU's running this CPU cooler (should specify, during the stress test, the temps are more or less pinned at 100).

 

So now i have described my problem (hopefully you can understand what i mean), is there any way to fix it?

would it be a good idea to just remount the cooler and possibly re-seat the CPU as well?

I'm just not sure as, while i run these tests/play games to test the computer, it randomly blue screens on games after about 5-15mins (reliably), giving an error with nothing to do with CPU over temperature, yet when i boot, it gives the over temperature warning. As for the tests, when I put a heavy enough load on the CPU and GPU, it more or less does the same thing as the games, even though the GPU never went above 80 degrees (though when it is CPU only stress test, it doesn't blue screen, just thermal throttles).

 

At this point, this has been going on for several months now and I have only just gotten round to actually doing something about it, but I'm not sure what to do.

 

Any help would be appreciated.

 

MY computer is as follows:

 

CPU - i7 6700k

mobo- Asus GENE 8

ram - 32gb 

GPU - gtx 1070 EVGA FTW

PSU - RM 850w

CPU cooler - Liqmax 240

 

[NelizMastr]

Sounds to me the cooler isn't making proper contact or the pump is dead. Did you apply thermal paste correctly? Did you remove the plastic film from the water block when you put it on the CPU? All fan cables (including the one for the pump) connected to the motherboard correctly?

 

It should be near ambient temps at idle, give or take 5 degrees, not 30.

[FuzzyLogic]

Make sure the pump is plugged in and working, and try reseating the block.

[Alasdair]

26 minutes ago, NelizMastr said:

Sounds to me the cooler isn't making proper contact or the pump is dead. Did you apply thermal paste correctly? Did you remove the plastic film from the water block when you put it on the CPU? All fan cables (including the one for the pump) connected to the motherboard correctly?

 

It should be near ambient temps at idle, give or take 5 degrees, not 30.

I think the pump is working, cause when i boot the system up after leaving it for a while i can hear the liquid flowing in the tubing right at the very start.

i did remove the plastic film 

I will check the fan cables and pump cable when i reseat the block and see where that gets me

[Alasdair]

23 minutes ago, thegreengamers said:

Make sure the pump is plugged in and working, and try reseating the block.

okay, i will check everything, reseat the block and see where i end up

[Alasdair]

1 hour ago, Alasdair said:

I think the pump is working, cause when i boot the system up after leaving it for a while i can hear the liquid flowing in the tubing right at the very start.

i did remove the plastic film 

I will check the fan cables and pump cable when i reseat the block and see where that gets me

 

1 hour ago, Alasdair said:

okay, i will check everything, reseat the block and see where i end up

Okay, i have re-seated the CPU (just in case) and reseated the water block of the AIO, and applied a new bit of arctic silver 5 thermal paste (it's a little old but i think it is still good).

i also checked all of the fan headers, and yes everything is plugged in and working fine.

 

So first off, the ambient was a little better than before (before it was around 30 now its closer to 20) , and since re-seating the CPU the computer seems to just run a bit faster than before (but it's most likely my imagination)

 

The results so far for the testing i'm doing are good. Currently i am running Furmark and a stress test with CPU-z to test the full system load (or my attempt at it). It's been running for 18 mins, and so far the GPU hasn't gone above 80 degrees and is at 100% load (totally fine as expected). The CPU is also doing okay, it’s doing much better than before. Currently it is sitting around 75 degrees on each core, though it is dipping down to 68 and up to around 85, but the test with CPU-z is only at 50% of the CPU (any ideas of any better test to run?).

 

So next up i will trying playing games on it and see where that gets me (if it blue screens or not) but so far the overheating issue (touch wood) has been solved just by re-seating the block for the AIO.

 

So I thank you for the help you gave so far with fixing the over heating issue!

 

[Maticks]

Some motherboard vendors were very agreessive on their boards during the 6700k release with auto voltage. If you’ve over clocked the chip on stock voltage it might not be stock.

 

i did this with my 6700k and had it running in th 90s. When you stress test it check hwinfo for the cpu voltage. Mine was at 1.49volts at auto and changed fixed mine to 1.36v and it’s stable cool and overclocked.

msi and gigabyte I heard we’re doing it. Msi released a bios update to fix it later. But not everyone applies thoughts.

 

sounds like you believe the pump is good that might be the only other cause if thermal pasted is applied correctly.

[Alasdair]

52 minutes ago, Maticks said:

Some motherboard vendors were very agreessive on their boards during the 6700k release with auto voltage. If you’ve over clocked the chip on stock voltage it might not be stock.

 

i did this with my 6700k and had it running in th 90s. When you stress test it check hwinfo for the cpu voltage. Mine was at 1.49volts at auto and changed fixed mine to 1.36v and it’s stable cool and overclocked.

msi and gigabyte I heard we’re doing it. Msi released a bios update to fix it later. But not everyone applies thoughts.

 

sounds like you believe the pump is good that might be the only other cause if thermal pasted is applied correctly.

I've ran a few tests with it since i changed the thermal paste, and well it keeps doing the same CPU over temperature issue as before.

 

I did watch the CPU voltage when benching it again, and it stuck around 1.472v but again went straight up to 100 degrees as soon as i booted up planet side 2 and even just couterstrike ( a game which i would consider really not intensive at all) 

 

At this point i really just don't know what to do with it, since it's obviously not an issue with the cooler, since both fans are working and the pump is working (i can feel the coolant flowing through the tubing it i touch/squeeze the tubing a little).

 

Other than taking the cooler off and putting on way more thermal paste than i normally have been, I just don't know what to do.

 

The only additional information i can supply is that it keeps spitting out an IRQL_not_less_or_equal error on the blue screen, and that it seems to still be overheating.

 

If anyone has any ideas it would be much appreciated 

[Alasdair]

Okay, i think from looking on line for the usual i7 6700k voltages, the stock is around 1.2v, which means that in my case, the cpu is running like 0.2 volts higher than stock, which might be the cause for the over heating, especially since i haven't overclocked my CPU, it's still running at the stock of whatever i got it at a couple years ago now.

 

So at this point i'm just a little unsure as to how i would go about undervolting the CPU in my bios (really unsure how to navigate the bios).

 

If any one could help me with using the bios, would be much appreciated.

[Alasdair]

okay, so I've done a bit of messing about a testing with the computer.

I can confirm  that the cooler is perfectly fine, but i found the issue, the CPU is kicking out a crapload of heat. Like the water that is going to the radiator of the cooler is noticeably warm, then the returning water is cold to the touch (so the lss, the cooler is fine)

 

However i then tried to undervolt the CPU to around the 1.35/1.39v region, but i can't find the LLC value to change it and as a result, whenever i boot into windows, it crashes (though i don't get a CPU over temperature error so i suppose that's good)

i found this post (which is where i got the idea from) on someone who had a similar issue to me, but they seem to have far greater knowledge on how to deal with this kind of thing than i do.n

https://rog.asus.com/forum/showthread.php?90828-Is-there-a-way-to-lower-Core-Voltage

 

If anyone could pssibly help me out in how to bring the voltage of the CPU down to an actual reasonable value (bare in mind that the stock voltage for these chips i think is like 1.2v), other than that i'm really not sure what to do about this, rather stuck atm.

 

Any help is much appreciated  

[Maticks]

1.47 is very high, mine is delidded and cant go over 1.41 with an overclock before it overheats.

 

did you accidently flick the ln2 switvh on your motherboard you mentioned it was a rog asus.

it has one those. Those voltages are really high 1.5 can kill the chip if its not cooled enough.

 

try factory defaulting the bios. LLC leave on default its different for every board. Get it running at stock clocks first. Lookup forums for best setting for flat LLC.

 

In australia i would take that board and cpu back to the vendor i bought it from and ask for a new cpu and different motherboard. Youve probably caused cpu degredation with all that voltage. Doesnt sound like its your fault maybe the motherboard.

[Laxmantim]

Can you post a picture of your BIOS overclocking settings? You may have accidentally clicked and changed something because 1.47 is extraordinarily high especially if you are running stock speeds and not overclocking also there was an issue around the time the 8700K came out with asus boards where the default setting had the cpu cranked WAY up compared to other mobo manufacturers, maybe something similar happened with your board?