mcbeen.com

I developed a 3D cad model tonight of the part I need, and got some quick estimates. Unfortunately, the costs were absurdly outside my price range. Alternatively, I have found a source for a replacement hunk of copper for the tail end of the water block and it only costs about 5 bucks. We’ll see if it comes through. In the mean time, I have fashioned a temporary cover plate for the water block that allows me to use it to cool the memory and processor of the graphics card.

It is by no means complete, but I wants to test to see how well an alternative would function in case the power module section can not be modified. So far the unit has run under full loads for hours with no leaks. I went ahead and attached it to my card this evening to see how it functioned. I’m getting excellent core temps, although my system water temp has increased to about 27C at idle.

I think tomorrow I will go ahead and make a final diagram of the cooling system, and reroute some of the water flow. The tap lines I installed at the beginning to cool my DSL modem do not run in series with my main cooling loop for the graphics card and cpu. They are small 1/4″ lines so they do not take much away from the main loop, but I’m tempted to redirect more cooling anyways. I would estimate at least a 3rd of the water flow bypasses the radiator, which would never be a good idea. Pinching them off has no effect on system temps however, so maybe I was right to begin with, and it doesn’t really matter. Here’s the idea… The water coming from the CPU and Video card go directly to the radiator. The water going to the modem comes right off the pump, and goes back into the reservoir. The idea behind it is there’s little heat being generated relatively by the modem, so pumping the exhaust water through the radiator is a waste of energy. The bad part of this idea, is that I’m killing off possibly a 3rd of my flow and pressure through my main heat sources and rejection. Testing does not bear this out, but i’m still concerned and curious.