I don't mean to be rude Iwamoto, but you sir are the ghetto mod king, that is a wild looking beast, and I like it, haha. You are a little quirky, I get that now, I also think you are young, but you aren't really giving the components you have the best chance of performing well, but your thread is entertaining to read. I also looked at your absolutely ludicrous implementation of TEC's into your loop, did you seriously hook some 12706's between a couple of aluminium plates and put a single waterblock on the cold side and no heatsink on the hotside, I can't quite make sense of what you did because you don;t explain it anywhere and I can't understand most of what you say anywhere. I just look at the photos ans cringe, the one with the AIO waterblock clamped to the TECs and bulldog clips mounting it all together. I see a couple of waterblocks that appear to be on the TECs later, did you seperatre the hot and cold side water cooling loops on them? regardless you are using some 12706 TEC's which don;t have anywhere near the wattage to make much of a difference at all except in idle. EDIT: I think I can see 2 water blocks on the cold side in one photo, can;t see what is cooling the hotside, then I see you have another aluminium waterblock with some tecs but can't see what is cold side and what is cooling hotside, maybe nothing from all I can tell. Either way you have zero authority to say anything about TEC cooling until you come up with something that makes more sense than whatever it is you have going on, it's clear that your inadequate implementation is what has turned you against TEC's when you are just butchering thee whole idea
I've already seen that monster TEC, the heatload from it alone is monstrous, there is another guy trying one out right now with a really nice custom made 62mm waterblock that should actually do quite well, but that power consumption scares me haha, but he should be able to get some pretty low cold plate temps, if he can get a handle on the hotside heatload anyway.
lzf, yeah I think the wavy channels are good for the hotside, you want as much coolant in contact with the copper as possible and some nice turbulence to break up the laminar flow through the block which means more coolant comes into contact with the copper surface and it isn't too restrictive, just have to make sure all parts of the block are getting equal flow, might need to make it dual pass which maybe you said already, so that it only flows through half the channels one way then through the other half the other way, but means the second pass might be a degree or two hotter but it is acceptable. EDIT: is that one long channel snaking through the hotside, at 12x3mm thats probably correct, but say 3 smaller channels would be better, like flow going through 3 x 3mm wide by 4mm deep channels, more surface area than 1 12x3mm channel but same flow characteristic. lot of drilling though
For the cold side you just want the fluid in the block for as long as possible with the least amount of restriction possible, so a straight channel but maybe three passes in an s shape, I don't think you need to be worried about it freezing, its actually pretty hard to freeze it if its flowing I think. But you are on the right track, I established that different block designs were required for hot and cold sides a while back but most people don't bother. Main principle for hot side is turbulence and surface area, for cold side low restriction and long coolant path through the block to give the water a chance to chill, even splitting the flow up three ways so it is really slow flow might help.Edited by LiamG6 - 11/7/15 at 9:51pm