High W/C temps

Sayso

Sayso

Sayso

Soldato
Joined
30 Sep 2008
Posts
3,451
Location
Chelmsford
My set up is a Q9550 at 1.37Vcore FSB and NB voltages well below maximum, a Silverstone T07 with extra front fan intake.
water cooling is a PA120.3 with Yate loon 2200s (set at about 1000) in push configuration, a XSPC 750 pump/res combo and a Dtek Fuzion 2 with quad splitter.

idle is about 5C above ambient but load temps are 20C+ above idle when running small FFT in prime95 (usually mid 50s depending on ambient temps).

now i know these temps are perfectly reasonable but ive seen posts on here with people cooling there GPU and NB and still getting less of a delta under load, what could i do to improve my temps......

i have reseated my block 3 times, what is the best way to orientate it?

i have a new pump arriving tomorrow and im going to shorten my looping ( i left it long for testing puposes)

my loop goes pump/res > cpu block> RAD

i cleaned the RAD and bled it before operation.

any ideas?


edit. would the orientation of the quad core insert make a difference as it come with no instruction i just put it in as i though it should go?

also when i take it off lod the idle temps sit 5-10C higher than they were.
 
Last edited:
Just have a quick gander at this thread, where it turned out the air was being recirculated when trying to cool the rad. Lot's of different things to try in here :)

Click Me :)
 
Happy i have read that thread in fact i have a couple of posts in there concerning this issue :)

a little bit of a different set up, i have a T07 with the RAD pushed up against the bottom grill, even with sides off temps are the same so its not an air circulation problem (though i could knock a few degrees off with different placement of my case)
i wouldn't say its pushing in cold air as my room isnt cold, it can get upto 25C if the babys in here, ive taken that into account when looking at idle temps i would expect around 10C higher than ambient on the CPU with this set up which is what im getting, im puzzled about the difference between idle and load and level of load temps which is idle temps + upto 30C.
 
I idle at around 38 at load at 72C on IBT :eek: so that's worse than you. Not sure what to suggest.
 
Yeah those temps arent right for a mildly OCd 45nm quad, especially with a 120.3. Feel the rad under full load, see how warm it is. If it doesnt feel like 45C or whatever the CPU is at, it's probably a contact problem. How is the flow rate? Check the water in the res, see how much it is rippling and stuff.
 
I'd put money on there being air in the radiator. That bleed screw is there for a reason. PA120.3's can be very hard to bleed indeed. Your radiator is sitting on it's side? Little things like the inlet being at the bottom and the outlet at the top can make the difference between an easy bleed and a real beggar of a job. Try lying the case on it's side or propping it up on end and see if a big gulp of air doesn't pop out.

If it's not the radiator then the recommended orientation for the block is to have the inlet/outlet arranged North/South with the outlet at the top so air comes out naturally. It's possibly also worth checking that you haven't got the inlets and outlets mixed up. I had that and the performance was poor.

I'm assuming your loop is silent and you can't see any bubbles flowing around in the pipes and that your pump/res is full to the top and not sucking any air?
 
Walter do you think my temps are reasonable?

Under IBT Q6600 3.7GHz-1.52vcore at 74C

120.3 cooling NB>mosfets/PWM>CPU

Thanks
 
Happy said:
Walter do you think my temps are reasonable?

Under IBT Q6600 3.7GHz-1.52vcore at 74C

120.3 cooling NB>mosfets/PWM>CPU

Thanks
Click to expand...

Sorry, but yes - I think they're quite good really for the amount of heat you're dissipating.

My last Q6600 system was running 3.65GHz with 1.55V and it was loading at 60C with nothing else in the loop, and that was using a TFC 360mm and 6 120mm fans in a push-pull configuration.

I find the extra fans do make a worthwhile difference and I use the Arctic Cooling 12025PWM fans as you can run 6 of those off a single motherboard PWM header. You should only really run 5 off one header, but it seems to work fine with 6.
 
Many thanks, I thought at worse the temps would be 'ok', but seems it's fine. Bit silly really but you see others worrying about theirs then you worry about yours :o :p
 
WJA96 said:
I'd put money on there being air in the radiator. That bleed screw is there for a reason. PA120.3's can be very hard to bleed indeed. Your radiator is sitting on it's side? Little things like the inlet being at the bottom and the outlet at the top can make the difference between an easy bleed and a real beggar of a job. Try lying the case on it's side or propping it up on end and see if a big gulp of air doesn't pop out.

If it's not the radiator then the recommended orientation for the block is to have the inlet/outlet arranged North/South with the outlet at the top so air comes out naturally. It's possibly also worth checking that you haven't got the inlets and outlets mixed up. I had that and the performance was poor.

I'm assuming your loop is silent and you can't see any bubbles flowing around in the pipes and that your pump/res is full to the top and not sucking any air?
Click to expand...

will try to bleed the RAD again perhaps i didn't get all the air out but everything else you mentioned is fine, fitted a new pump today and got the quad nozzle around the right way so have seen a drop, also ran with sides off with fans on full which as you can imagine made a little bit (about 5C) difference, going to post pics of my rig in watercooling pic thread have a look and see if im being a total nob (noob) and im making an obvious mistake :(
 
Sayso said:
will try to bleed the RAD again perhaps i didn't get all the air out but everything else you mentioned is fine, fitted a new pump today and got the quad nozzle around the right way so have seen a drop, also ran with sides off with fans on full which as you can imagine made a little bit (about 5C) difference, going to post pics of my rig in watercooling pic thread have a look and see if im being a total nob (noob) and im making an obvious mistake :(
Click to expand...

Is it running in the pictures? The little bit of foam in the res looks like it is, but you never can tell.

I'm not sure about the visuals on the black coolant (:rolleyes:at self), and the right-angle connector on the reservoir is going to slow up the flow a bit but the rest looks about right. It is quite a big loop considering it's just cooling the CPU.

The only other thing I would say is it looks like the PSU fan will be fighting the third fan on the rad for air, but that's probably more a PSU issue than a radiator issue.
 
it had only just started running in the pics and the third fan does overlap a bit,
the right angle is coming from the rad.
it is a big loop but with the 18W DDC shouldn't be a problem, by moving the pc about i had about an 50ml gap appear so it looks like there was some air in the loop
 
Sayso said:
it had only just started running in the pics and the third fan does overlap a bit,
the right angle is coming from the rad.
it is a big loop but with the 18W DDC shouldn't be a problem, by moving the pc about i had about an 50ml gap appear so it looks like there was some air in the loop
Click to expand...

Did you move it about much? My guess is if you stand the case on end, so that the barbs on the rad are to the top then you'll get a lot of air out of it. Or turn it the other way with the bleed screw open and listen for hissing.

That would give you issues with the res/pump combination though, so I would stand the PC on the edge of a table and take the rad out so that the barbs are to the top and run the pump. I also find that unless the reservoir is absolutely full, you can just constantly re-inject air into the loop. That said, I won't use anything else than those res/top combos.
 
ive had the case on each side, and had it rested on its front with RAD bleed screw open and didnt get any hissing or bubbles, im having a bit f a problem taking the RAD out due to super strong velcro i applied doooooooh!!
 
Sayso said:
Q9550 at 1.37Vcore..............(usually mid 50s depending on ambient temps).
Click to expand...

Sounds ok to me, although I guess it could be higher than you'd expect with a 120.3 just for the CPU.

Sayso said:
also when i take it off lod the idle temps sit 5-10C higher than they were.
Click to expand...

That suggests that the airflow through your rad isn't quite right. Looking at your pics it looks like the rad fans and the PSU fan are competing against each other for air flow. Also, I can't see how the air flow works in that bottom chamber of the case with the sides on. Does it have a grill on both sides?

I wouldn't of thought you'd have too much trouble with bleeding the W/C system with that set up as long as it's flowing into the bottom of the rad and out the top.
 
yeh it has a grill on both sides.
one fan isnt working to full potential due to the PSU fan, have tryed everything i can think of...i suppose its just going to have to be a bit warm and silent :)
 
Oh yeah, sorry I didn't see it had been mentioned already :o. What about turning the fans around or putting them on the other side of the rad so that all of the fans in that section of the case are working in the same direction?

Maybe you've just got a slightly hotter Q9550. I would have thought that 1.37vcore is quite high for a Q9550?
 
You must log in or register to reply here.
Back
Top Bottom