block temp vs water/ambients

[SF101]

very simple opinion thread.

my situation is as follows atm.

I have a few questions about my cooling loop

I setup my loop mostly to lessen noise levels and I've accomplished this task I've went cold in more ways than one but never took the water plunge before this so did not know what to expect via cpu temps vs other temps.

i have a 3770k @ 4.5ghz @ 1.3v on core had issues going lower voltages maybe some degradation from going cold previously.

mobo - z77 oc formula - with the water cooled mosfets in use.

cpu is receiving water after the rez which is after the mosfets. i wanted the cpu to receive water after rads first but somehow i got it backwards i could change order if you think its necessary. to stabilize temps but ive read this is a minor factor anyways.

these parts are in play on this setup
http://www.xs-pc.com/products/radiat...-fan-radiator/

http://www.xs-pc.com/products/radiat...-fan-radiator/

http://www.xs-pc.com/products/waterb...asa-cpu-black/

now that you have some background here my question is if my temps seem normal too you.

18*C - air
24.2*C water measured with my fluke probe
25-32 idle temps depending on cores 2 cores run warmer than the other 2 this is measured in real temp.


80-87c full load temps "core dependent" prime 1024k's measured in real temp.
temp swing is almost instant water temps mabe climb 2-3 *C over 10-15 min at same air temps.

4.5@ 1.3v

i have 1250-1350 rpmish semi high static pressure fans in play and they dropped temps fairly well they are in pull - pulling from top and back of a 280 antec case and 2 fans pushing out the front of the case. so semi positive pressure into case.

my personal thoughts are the rasa block isnt being very efficient with the transfer of heat load.

but then i hear most blocks perform similarly.

opinions?

whats your thoughts on this?

 

[Boulard83]

Too much paste ?

And yes, the rasa bloc is not an "high end" bloc but i think it should keep a 3770K at decent temp.

Also, keep in mind that Ivy have paste between IHS and CPU die, they are not soldered like pretty much all previous CPU. This have an effect on high volts/clock OC, at some point the paste that intel use i just not good enough.

 

[SF101]

mx4 paste is used

 

[Conumdrum]

http://www.xs-pc.com/products/radiat...-fan-radiator/ Page not found

http://www.xs-pc.com/products/radiat...-fan-radiator/ same

Third link is prolly the same.

Loooks like a bad mount. Could be a better CPU block, more like that's just how it is.

6C Dt temps, good.

Got a bad CPU?, too much voltage, can't see the rad etc for the loop. Links broke.

Dunno yet. Need better data.

Temps change that fast? You have other issues. Bad CPU or a bad mount.

Lastly, those CPUs run HOTTT.

 

[manu2b]

I'd say bad mount, maybe not enough TIM, or not enough pressure on the CPU. Moooaarr strenght on the screw driver!

6C Delta is not bad.

 

[Conumdrum]

A lower Dt vs a high CPU temp ponts to a bad mount, a bad CPU block, a poor performing loop.

Y
You haven't run P95 small Ftts recording ambient room temps at idle or load yet. You needto run it for 30 min for temp to stabilize.

You measure water (DT temps, ar temps and CPU temps properly we might be able to help.


Then we will ask you you to do the same test at stock temps.

Then we will ask you to post pics of your CPU cooler pulled off the CPU and pics of the TIM on the CPU.

Then we want to know your exact fans on the rads, your pump, fittings, etc.

 

[irrykuxeh]

+1

Try remounting the CPU block. If that was the problem, then your core temps should go down (and be more even between cores) and your water temps may go up a bit.

 

[GTXJackBauer]

Not enough info imho. Need more info on the whole rig.

 

[SF101]

small ftps were nearly identical to what i ran at worst 1*C difference at times usually same.

i felt the mount was good a solid flat push in place mount as i've done with cold mounts but easier since less bs.

ill try again i have plenty of paste.

i used a good amount of paste and the rasa block only screws down so far then it stops .

ill try a remount but i was more woried about rasa cpu block not being that great.

but i guess it seams fair . since the water temps arent moving much but the full load deltas are.

 

[Conumdrum]

Pics.

 

[SF101]

k?

 

[GTXJackBauer]

Theres too many variables so we'll need to know what your rig setup is and loop etc. Fans, rads, blocks, pumps etc.

 

[SF101]

rads are rx120 and rx240 rads btw the other link was to the rasa block. got sent away for work today ill take a day and look at the cpu mount soon.

 

[SF101]

took the block off mount seemed great could hardly pull it off.
remounted with a bit less paste mx-4 , put some plastic washers on all 4 posts to give it a bit more tension.

reversed the flow so water from rad is hitting cpu now not sure how much diff that makes but i'm assuming a little.

temps dropped some saw a few degrees cooler idles - water in loop staying about the same 24*C , in house air temps is slightly warmer today 1-2*C not much diff there.

full load water temps dropped a lot tho especially the right away temps so it seems the block is working better with the added pressure i think that was my issue. nearly 10*C initial temp differences on all cores - the cores are now only seeing a delta between each other of 3-5 *C as well which is nice.

water running small fpus after 20min is about 25-25.5 stabilized in that range.

seeing temps of 75-80 on these small fpu's depending on core.

 

[Conumdrum]

Sounds better for sure. One last question, I forgot to ask, You have the water going into the in side of the block? If backwards you lose impingement water flow. I'm sure you got it right, but thought I would ask.

 

[SF101]

ya its in on the oval inlet i'm quite sure.

I wont be checking it tho lol. next time I touch that cpu block its going to be because i'm changing it for something beefier.

If this setup works well I'll probably expand it all to a quick link external water rack. and just add a 3x120 or a 4x120 to it and a gpu block but that's down the road a bit not sure if id need a second pump in loop for that setup its alot of water.

So far i like the one speed fan noise no winding up and down just a constant low hum so much quieter.

I used those cougar 9 blade fans just wanted to try them and see how the performed. they seem to move air well decent static presure ratings for their rpms. they are rated at 1250 but do 1350-1365 on my board.

i can hear the GPU fans over them from my giga windforce 3 7970