CPU temps, what have i done wrong?

Right, finally got my ek block today and have completed my cpu loop which includes the following:

Pump: DDC 18w with xspc res top
Rad: XSPC RX360
Fans: 6x yate loons(push/pull)
block: ek supreme hf with plate number 2 installed

Now ive got the system working fine so decided to check the temps and what a shock :P. Heres what core temp is showing:

Idle -
http://i255.photobucket.com/albums/h...2/cputemp1.jpg

Load-
http://i255.photobucket.com/albums/h...putempload.jpg

Its odd, ive had a wc setup with just a d4 pump/res and a polarflo block and a generic 240mm rad, the temps with that setup was maxing at just 35 or 40C. Now with a huge overhaul and the top end WC items out it seems odd how i cant achieve this with my fans running max!

Now the first idea was air bubbles but this loop was great, got rid of air bubbles within 10mins (ive ran it for a few hours leak testing) and my older loop had loads but performed better still haha. Next, perhaps its my cpu block not making good contact? Ive screwed it as much as possible though so im not sure. EK supplied mx-2 paste and ive never had problems applying paste so it shouldnt be that....

Any other suggestions? Im working tomorrow so wont be able to have a good look so would appreciate ideas/things to check for sunday :D.

I would redo the paste. Oh and take pics I want to see it! :)

Do you have a way of measuring the coolant temps? Is there anything between the two surfaces other than TIM, e.g. plastic film on the base of the waterblock?

nah i made sure i removed the sticker :D, i doubt there could have been anything inbetween other than the tim however it will be on the list to redo when im off.

And Jay, trust me you dont want to see pics.... my building project for a case was put on hold for awhile, finally got back to it today but i wanted to use this gear so shoved it in my current case. It looks terrible, had no way of mounting it so ive used brown tape to secure it hahaha.

I will get pics shortly just for the lols though :P.

edit: for your viewing pleasure :L.
(poor quality, from phone not my canon 40D :(

http://i255.photobucket.com/albums/h...2/IMAG0085.jpg
http://i255.photobucket.com/albums/h...2/IMAG0088.jpg
http://i255.photobucket.com/albums/h...2/IMAG0089.jpg
http://i255.photobucket.com/albums/h...2/IMAG0093.jpg

The cables are everywhere but thats because i just wanted to get the WC loop done and head straight to building my new case :P, also the 3x yate loons make it even more messy since they have 2x molex each!.

Oh and just my luck with the ek block... i buy the only version that doesnt accept 2x 1/2" compression fittings! Had to find a normal barb to replace one of them >.<, very annoying as the other versions accept them fine :(.

right well ive reapplied the thermal paste as i had some time spare, minimal change in temp (talkin about 1C). Im stumped, ive upgraded my bios as it may have been the dreaded stuck/out of sync temperature probes from when these chips were released, however i havent really noticed it before.

I will have a look for a temperature probe, i may have a thermometer here somewhere and will just shove it in the fill port. However i would like to get one of these inline temp things as that would make it a lot easier :P. Any suggestions?

stick the sides on so you get airflow, if the motherboards hot the chips will idle hot.

will try that now,

Dipped my finger in as i couldnt find a probe, the water was pretty darn cold :P. So im thinking its the ek block failing as the top of it is stone cold.

Will report back with side panel on.

edit: no change in temp, even added another fan on the NB heatsink to help >.<.

i see its not running high flow if you used plate 2, you should have started with 3 to get more flow through the block, i use flow meters with my loops and koolance 5 way block that comes with a temp probe as it doubles as my drain port.

the top of the blocks never going to get hotter than the water as thats not getting chance to get the heat from the base plate, thermal dynamics 101 there. the waters moving through takng the heat from the base off to the rad, if the block was failing to work the top would be warmer as the waters not getting a chance to remove the heat.. i assume the lines in the plate run with the cores ? if not it will up the temp a few degrees.

did you use to much paste ?

are the 100% load temps to high ?

i do not worry about idle temps if my load temps are high, cores idle around 40 on my i7 at 4ghz and peak at 80 under the intel burn test

oh yes your fingers not much use when its all running below body temp :)

yeah sorry dunno why i said that, im not a complete dope :D. Plate 2 was the best in between in terms of flow rate and performance so i decided to go with that, the stock plate is shown as the worse so thats why i didnt leave it in there. I believe they are seated correctly, i redid it a couple of times as i kept re reading the instructions haha.

Didnt use much paste, never do but when i was taking off the previous paste from earlier it was all fine and evenly spread. The load temps are about 46/7 C which to be honest is quite high for my rig wouldnt you think? This is an e8400, a relatively low TDP chip (65w i think) and its not even overclocked! Its been undervolted for ages. And even more annoying is that my cheapy WC setup performed on par if not better.

Will take the block apart if i get time just to double check though.

edit: Just looked at the instructions and it shows that the plates are ment to be vertical and the cores should be horizontal (so not in line), are these instructions wrong then? :P


gonz0 is your WC setup including a cpu/motherboard and your 2x 480's? Or is it just the cpu?

was the test matching your system ? low restriction cpu block, rad and 18w pump in the loop ?

i went for a lower restriction block (koolance) 350 to 360 and got a total change of sod all temp wise lol

just £40 out of pocket after flogging taktak the 350.

got any good old arctic silver to try ?

the system im using is exactly the same as the one in my profile bar the change in cooling and the memory is at stock now (reduce power :P).

I will have a dig around for my as5 but i lost it a bit ago so had to rely on ek supplying the mx2 :P

edit: Oh dear.... ive just realised ive put the stock plate back in by mistake! Well i guess that explains a 1-1.5C difference, now wheres the other few C's coming from haha.

You may get better results from running the loop in this order.

pump - rad - cpu - res.

maybe a degree or 2 but thats all

I would also consider there is still an airlock in the cpu block.
Lay it on its side and get the res above the rad.

hmm, i cant get the res above the rad as the tubing would kink and come off due to the angle its at :P. I think you maybe onto something here though, but i think its in the rad as i can hear it trickle water every so often which is what i remember my old WC loop did and that had air bubbles in it.

Any other way to remove them?