I have been screwing around with water cooling for some time
now. I originally built the "beehive" because I knew there had to be
a more efficient design than the massive blocks of copper with a few holes
drilled in them. Since then my "beehive" water sink has held its own against
the best hollow and drilled blocks available. Even so, I had designed it for
testing on a peltier and knew it could be as efficient but scaled down for
the FCPGA format.
In my quest for efficiency I blatantly ripped off the swirling
motion of the water from the "Tide pool" (the BEST aluminum block design)
and added several types of internal
sinks to try and match the performance of the "beehive" on the FCPGA,
and even tried a silver base, but could
never get the temps to equal those of the larger version.
I also have been screwing with some air coolers. One of my prototypes used
a heavy stranded aluminum cable as the pins for the sink . I did this to try
and locate the bottom of all the pins as close as possible to the CPU die,
unfortunately it didn't work all that well and I moved onto other ideas. Pretty,
ain't she?
So,
while thinking about the water block and looking at the cable, BINGO! - I
could make the pins in the water block longer by twisting them into a cable!.
Because its round and relatively tight wound it wouldn't inhibit the swirling
effect and it would have much more surface than a solid rod. Also I
had noticed the top of the beehive ran very cool ,the copper was too thin
to transfer the heat to the top of the cap .With this new design I would be
able to solder the "core" to the cap also ,adding to the effective surface
area of the sink. Here you can see the twisted pins and the hole in the cap
where they will pass through.
Here it is finished, you can see the pins coming through the top. Also, notice
how small it is next to my Celeron2 key chain.
Here you can see how it completely misses the feet on the Duron. (and the
silly putty holding the clip on :)
I am proud to say that it keeps my clocked up Duron the same
temp under load as the much larger beehive (35c at 28c ambient), and
bests my silver bottom version by 2c under load. This seems to be as far as
I can go with metal based water cooling, so I will now get into designs involving
far more exotic materials
(more on this later) and try to shrink the size even further while maintaining
the same performance.
Dr Surlyjoe
