-
Heat dissipation
I haven't posted here in a bit, but got a question. After my tax return gets back I am going to have a local machine shop make one of these out of copper.
http://img.techpowerup.org/120127/Capture210.jpg
My loop is currently 25% Dex-Cool 75% water and has been tested to -10C circulating hot side would be pure distilled water while the cold side would be the dex-cool obviously. I have a MCR420, MCR320 and MCR220 stack plumbed into the loop with multiple 320/220's sitting on the shelf depending on how this goes. Circulation provided by a D5 (@D4) and D4. That would be for the TEC side obviously the CPU/GPU side will be a maze 4 GPU and I can't remember what the hell I have on the CPU doesn't matter. Got a MCP350+top for the pump. Fans on the rads are Delta 250+ CFM 120x32mm in a push/pull.
What should I shoot for wattage wise on the TEC's that block will fit 9 40x40mm TEC's. I have the powersupplies to run pretty much whatever (a big pyramid and multiple meanwells). I don't pay for electricity so I don't really care about the power draw. I am shooting for very cold and I am going to get this to be an all in case minus the powersupplies just for giggles.
-
40mm tec's are pretty small and hard to find in any real level of potency. I'd suggest moving to 50mm's, and poke around at old style TEC block design. The pathway system is a bit flawed. Mostly being that it's restrictive, doesn't provide alot of surface area, and with big enough heat sources (aka TEC's) will have a decently large differential across it on the hot side. (Which can be good if done right, but you'll want more of a pin matrix).
I'd move to 50mm's and go 4-6 long to make a nice stack that'll fit a bit easier then a 6-8 inch wide monster package.
-
Good to know will go back to my CAD drawings of a maze4 like design