These temps ok?

[JamesM]

Right firstly what i'm cooling

I5 2500k at 4.5GHz with 1.36v on load (might be able to further tweak this but not by much)

Then the loop

EK-Supremacy - CPU Waterblock - Nickel Acetal
EK DCP 4.0 Pump (12v) + res combo
XSPC EX120 120mm Radiator
XSPC EX240 Crossflow Radiator

3 Scythe 1850s at full speed
Pump is running off a fan header and runs at 1290rpm instead of 1930rpm but this is adjusted with temps, so higher the temp the quicker the pump runs.

Temps on full load (LinX) after a few mins average low 60s (61, 64, 62, 58)

Does that seem normal for the CPU used and the voltages?

I want to add the 670 to my loop in the future and want to make sure everything looks fairly normal before I do. I know the pump isn't the greatest but for the price i'm happy with it.

I run the pump slower to keep the vibrations and noise down, is it better to run it full speed or does it not make much difference?

Due to case size I can't really alter the radiator sizes as I don't want to start cutting stuff up. I could fit a top 140mm radiator to help but from what i've read it doesn't appear to be worth it. I choose the crossover radiator on front (intake) purely because it means the tubing is neater for me, the 120mm radiator is on the rear as an extrator. Is it better having this as an intake too and leaving the top 140mm as the sole extractor?

 

[pgi947]

Temps look fine, especially considering the ltx block.

Remember synthetic stress will generate far more heat (10+c) then gaming will.

 

[JamesM]

Woops my bad, got the wrong CPU block.

It's actually the EK-Supremacy - CPU Waterblock - Nickel Acetal which i'm using, copy and paste error

I just thought for some reason water was the miracle cooler and I'd be seeing temps of 40 and 50 load at those speeds/voltages rather than low 60's.

70 is the max safe 24/7 temp for Sandy Bridge isn't it?

 

[andybird123]

definitely try running the rear rad as an intake - pushing hot air over a rad negates the point of having it

I get 50's on a 3930k @ 4.5 1.42v, with an LTX block, so it must be possible to get lower than 60's on yours even in this weather

 

[JamesM]

That's what I thought. Does using different thermal paste make much of a difference? Would running push and pull help?

Just worried adding a 670 to the loop is going to increase temps too much.

I think when I come to add the 670, i'm going to take the loop apart and dissemble the CPU block to make sure it's clean as even though I flushed the radiators and everything before installing

 

[andybird123]

what paste are you using now?
there's not very much difference between most of them to be honest

do you have any way of checking what your water temp is?

push pull might make a few degrees difference, but it would be more effective to add more radiator if you can

have you also checked that you are definitely using the inlet and outlet port on the CPU block the right way round (maybe a stupid question but it has been known to happen) the one near the middle should be inlet and the one nearer the edge is the outlet

you could also try running the pump on full just as a test - if you're going to change the fan on the rad to intake then change one thing at a time so you know which makes the bigger difference

 

[Sir SpankalotUK]

thermal paste, most are excellent now but I think M4 gets the general thumbs up, you'll want to achieve the thinnest all over coating possible, so use the smallest amount possible. I use a rubber glove and manually spread it around and try to seat the block and move it in a circular fashion to get an even spread of thermal paste. If you achieve a really good coating, it should be quite hard to attempt to lift the block straight off.

Metal conducts heat well, thermal paste not so much, so limiting the amount in the way helps massively. Then make sure you have no kinks in the tubing and away you go.

 

[spixelspixel]

My 2500k had a max core temp of 62 with 1.36v after intel burn test so temps seem right

 

[andybird123]

thermal paste, most are excellent now but I think M4 gets the general thumbs up, you'll want to achieve the thinnest all over coating possible, so use the smallest amount possible. I use a rubber glove and manually spread it around and try to seat the block and move it in a circular fashion to get an even spread of thermal paste. If you achieve a really good coating, it should be quite hard to attempt to lift the block straight off.

Metal conducts heat well, thermal paste not so much, so limiting the amount in the way helps massively. Then make sure you have no kinks in the tubing and away you go.

wtf!?

I use a grain of rice or 2 sized amount in the centre of the CPU and let the mounting pressure do the rest... if you are applying enough that you need to spread it around manually then you are using too much

also never heard of M4 paste
unless you mean MX-4?

which is ok, as is PK-1, ChillFactor 3, even good old AS5 still holds it's own in most tests

 

[Sir SpankalotUK]

yes I did mean MX-4, so thanks for that correction.

And yes that is about the amount of thermal paste I use (try to use 1 grain of rice equivalent), I know some people use the method of spreading it around by using the mounting pressure, but if you screw the mounting plate down unevenly you could wind up with uneven distribution of the paste, which will affect cooling. As with most things, theres always two ways of doing things, 1 may not be better than the other.

 

[andybird123]

if you are mounting your blocks unevenly then you are going to get worse temps regardless of how much paste you apply and should be a pretty clear indication that you need to remount

I tend to follow the instructions with the block on mounting but most tell you do do it crosswise with both hands at the same time and alternate frequently exactly so that you do get an even mount