Intresting performance boost

Ok, back when I first bought my WC rig I enough tubing, I have an Apex Kit and I bought 1/2" and 3/4" tubing, but I didn't buy enough. I then proceeded to buy more tubing but I accidently got the 5/8" OD tygon stuff (Please, do yourself a favor and don't ever get this) Though it had alot of kinking problems and such I had no money to spend $20 more on tubing. So I just lived with it for a few months. My temps were 32-33 for my video card and around 30 for cpu (in bios, never checked it load) Well I finally bought some more 3/4" OD tubing and replaced all the kinking stuff, and it was looking great. I check my video card and cpu and the video card is running 37c and the cpu is running 30c, I'm thinking, WTF? I could see in the resevoir that the flow rate was much much faster then before without the kinks, I tried a few things to no avail, and my gpu temp is still 5 more then it normally was with the worse tubing. I decided to top off my resevoir and Voila! I'm getting 31 on the video card and 27 on the cpu again. My resevoir wasnt empty or anything, just a little over half way, and topping it off took off 5-6 degree alone on the gpu and a few on the cpu. So I would probably recommend this to any WC'ers that have a resevoir, but does anyone have any sort of explaination? Just something of intrest I found.

Chances are that the low water level in your reservoir allowed the system to suck air in to the line, therefore increasing temps.

Well that's just the thing, it was well over the intake and the way it was set up, there is no way air could have gotten into the system. If the water level was low enough to take in air I would have filled it up sooner.

My guess is that obvious surface motion in the res increases the friction head, making the res much more restrictive. In larger systems, engineers recommend that any pool or reservoir be deep enough that the pump doesn't cause the surface to churn. I don't see why that would be any different on our scale.

Not using load temps to monitor changes.......

I dont beleive anything until I see load temps.

ls7corvete said:

Not using load temps to monitor changes.......

I dont beleive anything until I see load temps.

Click to expand...

whys that?

At full load, equipment always generates about the same amount of heat. "Idle" is less consistant because the CPU is always doing something but the load can vary.

also motherboards/cpus all idle differntly. For instance a pentium M and any new chip underlocks/undervolts during idle. My athlon XP doesnt. So my idle vs load shouldnt be jurasticly different. Currently around 4c. A amd 64 idle vs load should be larger. This does depend on the radiator if the jump is larger.

Thats odd my res isnt topped off, its close though. And im getting really good temps.
My 651 3.4ghz stock gets 22C idle, and at 4.76ghz I get 26C idle.
So if I top it off I might get 18C idle!