Well I went to the junkyard and grabbed a heater core out of a Chevy S10. I ran a little toilet bowl cleaner through the heater core and flushed it out. Then I filled it with vinegar and let it set for a half hour and flushed it again. I use a hose outside with a cut off end to flush it. The water flows through it VERY well. I did have to pick some pine needles out of a few of the fins
I chased it down with a wire wheel on my cordless drill and the battery died, I have 4 other batteries but they are all dead as well lol. So while I'm waiting for them to charge I figured I'd post an image. I set two 120mm fans next to it for size reference. I'll post some more as I progress. It will be inside of an enclosure/shroud of some sort and attached to the rear of the case.
I soldered on some fittings with 1/2" pipe threads on em. So now I can put whatever I want, for barbs, on it. In this case I twisted on a couple brass 1/2" hose barbs.
I set the new radiator up quickly and for quick testing/reference. For the moment it has two 120mm fans just resting up against it and held in place with zip ties. They are old crappy fans that are barely good for anything other than just moving a little air, but they spin and are the only 120mm fans that I have at this moment. This setup is with no shroud and the radiator outside of the case.
I'll be building a housing/shroud here soon.
After my system warms up, the temps are as follows with an ambient of 77F.
30c idle (warmest core the other 3 are 1c - 2c cooler.
50c Full load, all cores, Prime95, large FFTs after an hour, the other cores are only 1 degree cooler and fluctuate a bit, up to -3c difference of the warmest core. Edit: I should have continued testing on Prime95 the temps were rising at this point and they probably would have went higher.
This is a substantial drop from the other radiator and is bound to get better once I have better fans and a shroud. I'll post back when more work is done.
I also have an E5200 and I will be testing how well this works with it as well, so I have some dual core results in the mix. I do not plan on testing the E5200 until I have the shroud built.
Well it's back to the drawing board for the water block. The thing is that this doesn't cool better than the AC7 Pro. I decided to lap it as well and it's temps dropped to nearly the same as the water cooling setup, perhaps a degree or 2 warmer. So I'm gonna design something new. I must say though, the water cooling does a much better job when ambient temperatures are lower. The air cooling doesn't react the same as the water in lower room temps.
I might get a degree or 2 cooler if the room temp is 5 degrees cooler. With the water cooling I get nearly 10 degrees cooler in the same situation (with all cores at 100% in both situations). Which seems to lead to the idea that a better shroud would help some, but I want to try a different water block design anyway. First I am going to try 5/16 tubing instead of 1/4. This alone will nearly double the water flow. Plus I'll only have to use 3 pieces of tubing which will give me more surface contact area.
I've already built 1 mock up and figured out that I am going to need a tubing bender because the heavier tubing is much more rigid. Not to mention that I sorta messed up the first attempt while soldering the tubing together; heavier tubing needs more heat and my old jig for holding everything together needs some adjustment.
^You know I was thinking of making my own block during Spring break (from two weeks from today). Also adding more surface (esp via fins) should surely help. May be we could work some designs out for water blocks? Another thing I noticed was that the heat pipes can transfer heat just as well as water or if not better due to the fluid used inside it.
Let me make a correction here, the water cooling is indeed better, I may have jumped the gun a little bit in my last post there. I didn't fully test the AC7 after reinstalling it. Though my initial design needs a tweak and I think using the 5/16" copper lines would help a bit. I am also going to order some flat stock (pure copper) and try a few things with it as well. I spose I could fire up a CAD program and draw up what I was thinking....
^Oh one more thing, can you tell me where you buy copper from? I usually buy copper ingots from Ebay for my projects (just received 2 1/2 lbs ingots 2 days ago). 1/2 LBS (3x4x.5 in) and 1 (4x5x.5) LBS. Cost about $16 with shipping.
Note: Not all copper ingots are the same size even if the weight is the same. Contact seller for size.
My design will my somewhat similar to what I have now. I will use a 1.5" X 1.25" X 3/8" piece of copper. Drill 4, 1/4" holes through the 1.5" side. Then solder copper tubing on each side. So there will be 8 pieces of tubing total, sticking out of the copper block. I will then dimple the tubing close to the block at the 4 inlet points to create turbulence before the water enters the block.
The block will resemble this..
The copper tubing will only be into the block about 1/4" and soldered into place. I may dimple the end of the tubing where it goes into the block instead, I haven't decided yet. In fact I will probably just make the piece of copper as long as the retention module is wide. I can remove the little plastic pieces from it that way...
Well I am having fun tinkering and perhaps I could come up with something that will sell. It is surprisingly cheap to make a water block, even out of nothing but copper.
In case anyone was wondering. The watercooling setup that I currently have is nearly 7c cooler than the AC7 pro at idle and about 9c cooler full load. My initial thoughts were way off base because I had turned the system on from a totally cold boot with the AC7 installed and temps looked very low; after I gave the system time to warm up things changed a little.