The above pictures show my older pump, which has been replaced with a larger version.
Room: 79 F
HD: 82 F
CPU: 84 F
Case: 93 F
GPU: 110 F
The cooler the room temp, the more efficient the cooling system becomes. (i.e. core temps will equal room temps if room drops to 68 F)
Max CPU temp with AH loaded, 103 F with case increasing to 100 F and GPU up to 125 F if room is below 80 F.
CPU was 2.1GHz, now 3.0GHz
FSB now @ 1400MHz (was 1066MHz)
Memory under clocked to 700MHz
GPU overclocked to 635MHz, from Factory OC of 565MHz
Cooling provided by Thermaltake. Seems this brand has a very bad image when it comes to making liquid cooling parts. I've modified the radiator by placing 2 fans, using a push/pull design maximizing airflow through the radiator, while also using an 80mm fan on the side of the case to provide cooler outside case air to enter the radiator enclosed in the case. I use a 120mm fan to cool Hard drives and I also use 2 small fans to cool the memory. I've added green lights to the resorvoir to allow me to easily see the coolant level in the reservoir. This setup requires coolant to be added every 6 months to maintain a high level of coolant, which will minimize the amount of air bubbles collecting in the pump, which will cause pump failure. I plan on creating to allow a quicker way to add coolant, because the current process takes me 30-40 minutes. I'd like it to just take <5 minutes.
When my last pump stopped working, the temperatures began to rise over a period of 1 hour. Although my CPU temperaure never reached the thermal shutdown limit, it was my GPU that was the hotest link. It heated the coolant which then broke through the seal and exploded through the waterblock of the GPU. Steam engulfed the entire case with making a loud sizzling noise, much like an overheated car.
Then sparkes occurred from the pump and it's the wires fried adding a great smell to the room. My system then shut down. The coolant was not conductive, so no damage to the motherboard or the cards below the video card were damaged. After inspecting the GPU waterblock, I found that the screws just needed to be tightened and after replacing the pump, all was well. Only damage encountered was to the video card, as one DVI port no longer functions with all colors, so I use the other DVI port. My PC was essentially a brick until a replacement pump, which no local store stocked, arrived 4 days later.
In conclusion, Ruler2, try to think about the initial cost, and the cost to maintain the system, both in downtime and annual costs. Ensure you have a BIOS, driver or some small application that can monitor CPU/GPU temps and perform a system shutdown automatically. Have fun with finding creative ways to improve whatever kit you buy, or if you buy the parts individually. Also with lower temps, you can decrease voltages to CPU with allows even a great reduction in temperatures. I also use my vid driver to only enable overclocking when certain apps are running which means my GPU temps drop even more when doing everyday chores with my PC. Lots to learn and experiment with once you go to liquid cooling.
