Watercooling run by

[JamesM]

Basically I have the following..

OCZ HydroFlow HF-MK1 CPU Waterblock
XSPC RS240 120mm Radiator
XSPC Dual 750 Bay Reservoir & Pump

Which I have setup in the following way

pump/res --> cpu --> rad --> pump/res

I'm running a 8400 at 4GHz and with orthos running i'm getting 62C and 52C. Does this seem right? I thought it was a bit high personally? The CPU is lapped and totally flat so its not the heatsink on the CPU, why is one core so high compared to the other? Now I will point out the only place for the rad was at the top of the case so its extracting the warm case air as well, will this effect temps massively?

Couple of other things. The case is CM690 with a front and side case fan (plus the two on the rad) all running at 7v

 

[Conan257]

You have to remember there are many variables with water cooling...

Your pump isn't biggest in the world so doesn't shift a massive amount of water... Your CPU block isn't one of the greatest either...

Core temps can differ through various reasons, even if one is working harder than the other.... Compare these temps to your main CPU temperature (this may be 10c lower)...

 

[JamesM]

Right running orthos for a while and I end up with core 0 48 and core 1 at 37

Still not sure why there is the large difference between the temps?

 

[MikeTimbers]

Your block may be flat and if you lapped the IHS on the i7 that would be flat but you can't affect the quality of the thermal connection between the IHS and the die. In my experience all multi-core chips show quite large differences between the cores.

Here's my tip du jour:

Ignore it.

 

[fornowagain]

First thing is to understand where the temps come from, each core has a DTS (sensor) value in a register (MSR) on the chip. The value is a negative the moves towards a value called Tjmax. So when the DTS is at zero, the temps are at Tjmax. What all the programs do is subtract that value from an assumed Tjmax value, a generic value. Thing is, seems that the Tjmax values are factory set, can't be read directly and vary for each core. So if you have the wrong Tjmax value, you'll have the wrong temp. Also the DTS sensors aren't linear, they're designed to be accurate near their activation point called TCC (as it approached Tjmax), so at idle the values can be even further out. Add to that the solder bonding for each core may not be equal, the IHS won't be flat, the HSF/block mount may not be great.. et al

Now Realtemp allows you adjust the assumed Tjmax and idle offset values for each core. Have a read up on Unclewebbs posts on XS, you'll get the idea. But you can calibrate some of it out if you wish.

http://www.xtremesystems.org/forums/showpost.php?p=3533707&postcount=2958

"The digital thermal sensor (DTS) accuracy is in the order of -5°C +10°C around 90°C ; it deteriorates to ±10°C at 50°C. The DTS temperature reading saturates at some temperature below 50°C. Any DTS reading below 50°C should be considered to indicate only a temperature below 50°C and not a specific temperature. External thermal sensor with “BJT” model is required to read thermal diode temperature."

http://www.tomshardware.co.uk/intel-dts-specs,news-29460.html

 

[JamesM]

right so basically, if I go according to the lowest temp then I should be ok?

Left orthos running last night at 4.2GHz and it finished no worries and one core hit 60C max and nothing above.