Air Cooling > Water Cooling.

zoson · 02-16-2003, 05:44 PM

Ok, so the topic is mainly to get your attention. I know that water IS better, but I couldn't help but notice a few strange occurences.
On a consistent basis (every single cpu I have owned that has an internal diode ~ 6 cpu's) I have achieved higher overclocks with my watercooler (for those of you not familiar with my rig http://pictures.zoson.net) than with my aircooler (alpha pal8045 w/ 50cfm fan).
BUT also on a consistent basis the diode of my cpu has read lower temperatures than my watercooler, at the same voltage, and same CPU speed. I didn't think much of this really, until a friend of mine put a fan pointing at his waterblock and his idle temperatures dropped a good 11C (I can't remember how much his load temps dropped offhand), and his max overclock went up about 150mhz. He's using a chevette heatercore, with an eheim 1250 (317gph) and a Z4 waterblock on a 1700+ tbred-b JIUHB cpu in a completely open case (no side panels etc).
So my basic question is why would temperatures report lower with an air cooler, yet not overclock as high as my water rig, and why would pointing a fan at a waterblock drop temperatures so significantly?
Lastly, a question that has been touched on before, why are there no waterblocks with fins+fans (other than that pos reactor thing from cpufx) for more active cooling? If the ambient air in your case is cooler than the cpu/water won't having active cooling on your WB help significantly?
-Zoson

golovko · 02-16-2003, 05:58 PM

asuming your temp measuring devices are accurate and somewhat consistent, i would say that one reason (i dont have anything to back this up with) is that the rate at which they remove heat is not linear; while one may remove lower amounts of heat better, as the rate of heat production climbs the watercoolers abitility to remove heat exceeds that of the aircooler. again, i have nothing to back this up with, this is just a guess.

murray13 · 02-16-2003, 06:56 PM

IMHO and just a guess at that is that there is some hot air being trapped behind the proc in the socket with the water cooler. Adding a fan or using air cooling moves enough air around the socket to prevent this 'hot air' from building up.

winewood · 02-16-2003, 08:22 PM

My goodness wally. You may be on to something!
-Beeve

I was having the SAME problem. The only thing I can account for it the missing airflow around that area. This test can conclude if you are right. I have a fan.. a small 8 inch, and I will set it up, blowing into the case. Lets see if this improves anything..

winewood · 02-16-2003, 08:26 PM

Brainstorm.. Ok.. What if I just cut a hole in my case over the proc. Then mount a fan blowing on the cpu... hmm..
how would one power those 3 pin fans? I dont have anymore board attachments.. and cant find a power chord that looks like it will connect. Any ideas?

redleader · 02-17-2003, 02:21 AM

Your probe is inaccurate.

g.l.amour · 02-17-2003, 03:17 AM

i have had the same thing happen to me.

it has been said many times but alot of mainboards use under socket probe for displaying temps and diode for overheating protection. thats because the diode makes ppl nervous with the high readings.

and with an undersocket sensor, u most definetely will get a big effect blowing air around the socket.

cheerz

winewood · 02-18-2003, 09:11 PM

Red leader.. u saying that my probe is less accurate than the mobo thermistor??

02-16-2003, 05:44 PM	#1
zoson Cooling Savant Join Date: Nov 2001 Location: X Posts: 204	Air Cooling > Water Cooling. Ok, so the topic is mainly to get your attention. I know that water IS better, but I couldn't help but notice a few strange occurences. On a consistent basis (every single cpu I have owned that has an internal diode ~ 6 cpu's) I have achieved higher overclocks with my watercooler (for those of you not familiar with my rig http://pictures.zoson.net) than with my aircooler (alpha pal8045 w/ 50cfm fan). BUT also on a consistent basis the diode of my cpu has read lower temperatures than my watercooler, at the same voltage, and same CPU speed. I didn't think much of this really, until a friend of mine put a fan pointing at his waterblock and his idle temperatures dropped a good 11C (I can't remember how much his load temps dropped offhand), and his max overclock went up about 150mhz. He's using a chevette heatercore, with an eheim 1250 (317gph) and a Z4 waterblock on a 1700+ tbred-b JIUHB cpu in a completely open case (no side panels etc). So my basic question is why would temperatures report lower with an air cooler, yet not overclock as high as my water rig, and why would pointing a fan at a waterblock drop temperatures so significantly? Lastly, a question that has been touched on before, why are there no waterblocks with fins+fans (other than that pos reactor thing from cpufx) for more active cooling? If the ambient air in your case is cooler than the cpu/water won't having active cooling on your WB help significantly? -Zoson __________________ Core i7 930 @ 4254Hz 22x193 \| ASUS R3E 901 BIOS \| 12GB Mushkin Blackline 3x4GB @ 1547MHz 7-8-7-24-1N \| Lian Li PC-A10B 2x EVGA GTX275 SLI @ 720/1266/1566 \| 2x 80GB Intel X25-M G2 SSD Raid 0 \| 3x 1TB Samsung Spinpoint 7.2K Raid 5 \| Corsair 950TX H2O \| Enzotech Sapphire Rev.A \| EK-FB RE3 \| Swiftech - 2x MCW60/GTX275 Unisink, MCP655b, MCRes Micro.v2 \| 3x HWLabs SR1-120 あなたの神への弓 Time isn't wasted, when you're gettin' wasted.

02-16-2003, 08:22 PM	#4
winewood Cooling Savant Join Date: Nov 2002 Location: in my chair Posts: 574	My goodness wally. You may be on to something! -Beeve I was having the SAME problem. The only thing I can account for it the missing airflow around that area. This test can conclude if you are right. I have a fan.. a small 8 inch, and I will set it up, blowing into the case. Lets see if this improves anything.. __________________ -winewood-

02-16-2003, 08:26 PM	#5
winewood Cooling Savant Join Date: Nov 2002 Location: in my chair Posts: 574	Brainstorm.. Ok.. What if I just cut a hole in my case over the proc. Then mount a fan blowing on the cpu... hmm.. how would one power those 3 pin fans? I dont have anymore board attachments.. and cant find a power chord that looks like it will connect. Any ideas? __________________ -winewood-

02-17-2003, 03:17 AM	#7
g.l.amour Cooling Savant Join Date: Mar 2002 Location: on da case Posts: 933	i have had the same thing happen to me. it has been said many times but alot of mainboards use under socket probe for displaying temps and diode for overheating protection. thats because the diode makes ppl nervous with the high readings. and with an undersocket sensor, u most definetely will get a big effect blowing air around the socket. cheerz __________________ yo soy un tiburón

02-18-2003, 09:11 PM	#8
winewood Cooling Savant Join Date: Nov 2002 Location: in my chair Posts: 574	Red leader.. u saying that my probe is less accurate than the mobo thermistor?? __________________ -winewood-

02-16-2003, 05:58 PM	#2
golovko Cooling Savant Join Date: Mar 2002 Location: Purdue University, USA Posts: 141	asuming your temp measuring devices are accurate and somewhat consistent, i would say that one reason (i dont have anything to back this up with) is that the rate at which they remove heat is not linear; while one may remove lower amounts of heat better, as the rate of heat production climbs the watercoolers abitility to remove heat exceeds that of the aircooler. again, i have nothing to back this up with, this is just a guess.

02-16-2003, 06:56 PM	#3
murray13 Cooling Savant Join Date: Sep 2002 Location: Omaha, NE USA Posts: 216	IMHO and just a guess at that is that there is some hot air being trapped behind the proc in the socket with the water cooler. Adding a fan or using air cooling moves enough air around the socket to prevent this 'hot air' from building up.

02-17-2003, 02:21 AM	#6
redleader Thermophile Join Date: Jun 2001 Location: The deserts of Tucson, Az Posts: 1,264	Your probe is inaccurate.

Currently Active Users Viewing This Thread: 1 (0 members and 1 guests)