Quote Originally Posted by PatRaceTin View Post
Great project

How about cooling performance compare to normal WC Rads?
Thanks for the nice comments guys!

It idles at around 30C-ish with an i750 at 4.2-4.3 using 1.43vcore and with the overclocked 4850 GPU in the loop. Don't have a thermometer unfortunately so don't know what the ambient is. At load in Intel Burn Test it stabilises at low 60's celcius IIRC. Gaming gets to low 50s max, Orthos low to mid 50s. Need to get some temp probes and do some proper testing really as the extent of performance testing so far is pretty much leaving IBT running hundreds of loops, come back in an hour or two, have a look at temps. Play games for a few hours or so, look at temp readout occasionally. The trouble is, the HK3.0 keeps getting gunked up with rubbish. Should stay clear now hopefully.

In terms of flow and pressure loss, with an XSPC edge, 2m of 1/2" tubing and using a 10w DDC with alphacool top with the outlet going into a 5 gallon container, it gave a flowrate of 1.87gpm. Martins flow estimator gives a figure of around 2.1gpm for an xspc edge, 2m of tubing, 10w DDC with top and a pa120.3, for comparison. Think it works out as restrictive as about 3 120.3-type radiators. Running an 18w DDC doesn't seem to make much difference. I'll do some proper testing when I have a controller for the pump.

As it's fanless it relies much more on radiative cooling to get rid of heat than normal pc water cooling radiators. When doing stability-testing runs or gaming I can't really feel any air rising through it, just a warm glow. Might just be that the air's coming off a large surface so very slow airspeed, but sitting it on its side (so the fins are horizontal) only has a small effect of CPU temps as far as I can tell - I'd guess a few degrees.


The copper is still reasonably unoxidised, and still has burnt tacky flux from rosin-core solder used to solder the fins on where the bead-blastier didn't get to every nook and cranny between the fins and pipes. The emissivity of it (how good it is at radiating heat) is probably still pretty low. A perfect radiator has an emissivity of 1 (termed a black-body radiator). Polished copper has an emissivity in the region of about 0.05. Fully oxidised copper about 0.6 iirc.


So the heat transfer can be improved by coating the fins with a thin, high-emissivity coating. Paint (of all colours) is great for this (emissivity 0.92-0.95), but also tends to be quite thick so acts as a thermal blanket and lessens heat transfer by conduction to the air. I'm trying to find a finish that'll be a great thermal radiator and also very thin. So it's probably going to get electroplated at some stage when I work out what would be best to electroplate it with. The difference should be pretty big - If the fins in my project were mirror-polished Copper they would radiate very little - 600W less heat compared to if they were painted, at 40C for the fin temperature and an ambient of 20 C(assuming the heat transfers okay to the air and doesn't simply continue to pass through and get absorbed by the fins next to them).

If anyone has any ideas for electroplating copper heatsinks I'm all ears!