PDA

View Full Version : help me balance my airflow



bacon
05-14-2005, 11:20 AM
ok guys here is the rig in its current config

http://img.photobucket.com/albums/v218/mikejbry/IMG_0720.jpg

This is how the air flow is at the minute

In:

Front: 2 x 80mm panaflow L1

Exhaust:

Rear: 2 x 80mm panaflow L1
Top: 3 x 120mm panalfow M1 @7v
OCZ 520 psu fan


Ok this is obviously unbalanced, i.e a lot of exhaust but not a lot pushing air in.

I dont like the idea of chucking air that has been heated by the rad into my case, i.e 200w of heat being chucked into the case, but will change the rad fans to intake if you suggest so


What if i use my rear fans as intakes as well?would balance it slightly more, and im sure not much of the psu fan heat would enter the case


After seeing nik's new shroud on his 120.3, im going to do the same and replace my gutted fans that do the same job.

Also anyone think that 3 panaflow L1's will be significantly quiet at 7v than my M1's at 7v?

final Q, my temps for cpu seem kinda high (ok i know mobo temps = crap but hey), i get to about 50C load with 1.65v through my winchester during prime at 2.8ghz. idles around 35C

my case temps are between 25 and 27C from a probe in hte centre of the case to teh right of the mobo

water spec:

storm G4
silverprop HL
AQX 50z 5/8" barbs
120.3 with 3x panaflow M1's
T piece 5/8" OD

cooling 3000 48 spaw and a 6600gt

i was hoping a airflow balance would improve my temps :)

Cheers guys

Ancient_1
05-14-2005, 04:38 PM
Personally I would have the 3 120s as the only fans drawing air into the case and use all the rest as exhaust. I believe that would give you the best overall temps with your setup. With the number of fans you have the extra 3~5° higher air temp of the air going through the rad and into the case will not hurt your case temps much but will help with your water temps.

=[PULSAR]=
05-14-2005, 04:59 PM
All the 80mm fans as intakes except the PSU one and the 120mm as exhausts you always want more flow for exhaust than intake.

EnJoY
05-14-2005, 05:53 PM
Incorrect, you always want to maintain positive pressure in a case as it keep more cool air in and helps keep dust out. Hence, you want more intake than exhaust if you can help it, or equal.

Holst
05-15-2005, 01:31 AM
My case is similar to yours and I also have more exhaust than intake.

I think the best option is to put a fan in the side pannel blowing in.

Posistion it over the graphics card to cool the memory down. I think im going to use a quiet 120mm fan in the side.

bacon
05-16-2005, 04:51 AM
My case is similar to yours and I also have more exhaust than intake.

I think the best option is to put a fan in the side pannel blowing in.

Posistion it over the graphics card to cool the memory down. I think im going to use a quiet 120mm fan in the side.

i would tho i dont fancy cutting a hole in the window panel, tho i realise this would be a very good idea with respect to my question, just not sure i wana cut the window yet



Ancient, i had considered that, what sort of temp diff do you think the extra heat from the rad would cause?

if it is only say 2-3C then ill happily reverse them

anyone sopinions of quietning the beast? 3 L1's be signif quieter than my M1's?

bacon
05-16-2005, 10:18 AM
also anyone know why my cpu temps are :banana::banana::banana::banana:e?

Ancient_1
05-16-2005, 11:13 AM
I Would guess 3° possibly more. Your case temp will rasir about the same but from my experience the components are cooled as well or better. I have 2 max3s and both are running P4s at about 3.4 and same vcore, my WCd board reads case temps the same to 3° higher but the PWM is almost 10° lower. But I have a little more airflow going through the WCd setup with 2 delta VHEs (push/pull) on my rad with a sanyo 120 in the top for exhaust along with 2 80mm sanyos (I like a slight neg pressure in my case, it seems to cool better even tho it does cause dust to collect at the joints) and only a single VHE for an intake with 3 80mm for exhaust on the air cooled machine.

On all my machines tho I filter all incoming air which does restrict the flow of my intake fans some, but also slightly quiets them.

I dont like the idea of not having all the fans in a case not helping draw air through the rad and if I didnt have enough exhaust fan power to cause a neg pressure in my case it would just be more pressure for my rad fans to overcome.

With the number of fans you have and the 120s pulling air in your case the higher temp of the air will have little to no effect on how well your components

nikhsub1
05-16-2005, 11:46 AM
What nikhsub1 would do: First, the rad needs outside air, not case air. I would have the rad up top as intake, not exhaust. The rad air wont be more than 1C hotter than non rad air, but the way you have it now you are just heating your water. Having the rad draw air in will also give the MB direct cooling of the fets, memory etc. I would keep the front fans as intake, and leave the others exhausting. IMO better to have postive pressure for dust reasons; more exhaust than intake will draw dust in from everywhere. To sum up: The only change I would make is to switch the rad from exhaust to intake, I bet your mobo temps go down as well as you water temps.

bacon
05-16-2005, 02:35 PM
ok im about to sort an order with phil (EPUK) for the shroud and gasket set for my 120.3

then while im at it, i will switch the rad fans to intake and see how my temps change

cheers for the advice


fans are going to be pushing now btw, make much diff?

nikhsub1
05-16-2005, 02:49 PM
Better to pull, but...

bacon
05-18-2005, 09:05 AM
ok cheers nik

will try with push i think, causes less problems my end


anyone have nay idea why my load and idle temps are 16-18C apart? on my cpu

i know mobo diodes are crap but must have some validity for idle/load difference?

Ancient_1
05-18-2005, 12:00 PM
I would guess a bad mount of your water block. Only time I have had more than about a 5° difference in load/idle was a disturbed mount of my block.

bacon
05-18-2005, 01:03 PM
I would guess a bad mount of your water block. Only time I have had more than about a 5° difference in load/idle was a disturbed mount of my block.

iv tried several mounts tho, my method of mounting is basically sliding the block along the threaded bolt and pressing it down, twisitng it a bit then tightening my thumbnuts pretty tight

always a perfect splat of AS5 in the centre (not very much)

not sure how it can vary, tho im willing to try another remount to see if things change.

cheers for the help

Cossey
05-18-2005, 01:11 PM
my san diego is the same tempwise (similar setup but with at 120.2 rad.) i tied checking the mounting and it was fine then i got the digithermometer out and checked my ambient and water temps.
idle
air at rad inlet (outside of case) 22.5oC
water temp (after rad before pump) 26.8oC
cpu (speedfan) 32oC

load (2hours of prime)
air 22.6oC
water 28.8oC
cpu 51oC

ive checked the block is fine im getting good contact and good flow. i worked out the cpu heat output roughly (2860/2200)^2*(1.60/1.40)^2*89 that equals 196w, i know the formula isnt quite right and tends to over estimate but still it should give an eye opener as to why are temps arent as good.
the old axps and xpms only put 100-120w out the new highly clocked 64s are putting alot more out so temps will rise accordingly. if you look over at the storm review on procooling and exprolate the power output results up then it would account for the much higher load temps.