7V or 12V only a 1C difference?

Dutch Guy · 25 Apr 2006 at 08:20

Yesterday was (finally) a warmer day so I could do some testing with my new system.

The CPU temp is normally 36C but as the ambient temp was 25C instead of 20C the CPU temp was a little higher at 39C

I run the 120mm fan on my rad at 7V but last night thought I would see what the difference was at 12V (besides the noise) and was surprised to see that it only made a 1C difference :confused:

So what does this mean, my rad is too small, my fan is underpowered or the sensor is crap?

Drazic · 25 Apr 2006 at 09:14

Possibly ya rad but depends what it is and same for the fan. Also what hose size and block and pump you got

I can run my rig in my sig pasasivly and it only goes up by 5c thats how good my rad and pump is

Jimbo Mahoney · 25 Apr 2006 at 10:26

What CPU is it? (re: thermistor)

What rad is it?

What pump / block do you have?

It's an interesting question - the only thing I can think of right now is that it infers airflow is not the limiting factor in your cooling system.

As to what else it means, I'm not sure.

I'm doing a quick test with my system, as I know my radiator is NOT the limiting factor

(It's an actual car radiator, not a heatercore, and has three times the surface area of a 120.3). Will post back in a sec.

Jimbo Mahoney · 25 Apr 2006 at 10:40

OK, here are the results for my setup (CPU is under load 24/7):

Ambient = 20.6'C

5v - water = 24.9'C, CPU = 40'C

12v -water = 22.7'C, CPU = 38'C

The fans are 4 x 120mm Akasa Amber, each pushing 44cfm @ 12v.

So, at 5v each, they do a total of ~ 73cfm, at 12v ~ 176cfm.

Also, I dug around for an accuracy of the diode in the AMD chips - for the 754 package, AMD say +/- 7'C :eek:

Dutch Guy · 25 Apr 2006 at 10:48

CPU: San Diego 3700+ @ 2750Mhz 1.525V

Radiator: Aluminium heatercore 120mm:

Waterblock: Innovatek rev3:

Maybe the small diameter tubing (8mm inner, 10mm outer) stops me from getting better performance?
Or maybe the radiator can't get rid of more heat due to the limited surface area?

[edit]saw your reply, looks like you get a smal difference as well, strange that this wasn't noticed by me in the past as I had watercooling for years now [/edit]

Jimbo Mahoney · 25 Apr 2006 at 16:46

What pump? (ideally flowrate and head pressure).

It could be the combination of low head pressure pump + small diameter tubing = low flowrate = less efficient heat exchange at interfaces (rad/air and cpu/block).

EDIT - I make your CPU 100 watts output under load. Delta T in both cases is ~15'C.

15/100 = 0.15 'C/W

Seems very reasonable for that watercooling system. The limiting factor would be the pump, tubing and waterblock IMO.

R B Customs · 25 Apr 2006 at 17:15

do you have a spacer / shroud on your rad...
if the fan is butted up close to the fins this MAY be why your only gettign 1C