Need advice/comments on my refrigerated case idea

Chrome · Post by **Chrome** » Wed Jan 31, 2001 3:19 am

From what I have seen, most people that want case and/or processor cooling tend to follow the school of thought that there must be a lot of airflow through the case with a sometimes active cooling of the relevant components, such as the use of water or peltier cooling or a combination of both on components such as the processor or video card.

What has always surpised me is that there is very little effort to actively cool the entire case of the computer, with the possible exception of submersing the motherboard, expansion cards and processor in a nonconductive fluid such as mineral oil.

What I propose to do is a compromise between the two; to seperate the motherboard and mounted components, (such as processor and expansion cards) from the drives and power supply and actively cool them using surplus cooling units from 12v portable refridgerators.

I figured I could do this with prodigious use of cardboard and insulating materials such as urethane foam and/or RTB to make an airtight and insulated environment, along with a tray of silica gel to absorb moisture. The drives and power supply would still be air cooled in the upper half of the case, and an additional power supply or two would be used to run the peltier coolers and the various fans located throughout the case, and these power supply(s) would also be located in the upper half of the case.

Now, however ambitious this poject would seem, I realize that there are many pitfalls involved, and I probably can't predict or perceive any where near all of them. Here is what i have anticipated so far:

1. Condensation. In a cooled environment there is the danger of condensation occurring and making the motherboard and any of components connected to it go zzzzaaappp. I figure I can basically eliminate this by making the motherboard environment airtight,as well as insulating it, and using silica gel beads to absorb any resident moisture. Note that earlier I mentioned a 'tray' of silica beads. Periodically, I would remove the beads without opnening the case proper, stick them in my oven, and dry them out. Also, since the refridgeration units are not within millimeters of components, I'm not too worried about condensation on components at startup, and since they are not in direct contact with components, its okay if they drip a bit. Hopefully though, the silica gel beads will suck up most of the moisture. In any case, I operate my computer in a relatively stable environment, so I don't expect my pc to be xposed to extremes in humidity.

2.Stability under overclocking. This is not my ultimate goal. While it would be nice to overclock, and I may well do so if conditions permit, that is not why I am doing this. My ultimate goal is stability. If I can get a few extra MHz out of my components, so much the better. But if I can gain a lil extra stability by having everything run 5-6 deg. C cooler, than I am all for it. I intend to use some crude ducting, so hopefully the cpu and video card will have some extra cooling.

3. Case to hold all of this. I intend to get a single width server case to hold all of this, so space is not much of a problem.

4.Will it work? I don't know. I've asked people I know, and they seem to think my ideas are somewhat solid, but they, like me, don't know of anyone who has approached case cooling in the manner I propose.

I actively welcome any holes anyone can shoot in this, as I have beat it to death between myself and my real world contacts.

smaier69 · Post by **smaier69** » Wed Jan 31, 2001 8:30 pm

well, in my humble opinion, the premise of the solution does sound viable. i think your number one concern would be condensation. almost anything else would be of a lesser concern (IMHO, of course) since pretty much every component inside the case would be kept at a lower than spec temp range (compared to normal aspiration). also bear in mind that its not necessarily the current relative humidity thats the problem.... its the availability of any humidity over time - giving the water time to collect - thats the real problem.

what i would try if i were in the same position as you is this....

there is something called "comnformal coating" thats used in the electronics industry to coat electrical components and PCBs to provide a non-electricaly conductive and waterproof coating for said chips/boards/etc. some high-end cooling solutions such as peltiers can also generate condensation, and applying an even coating over the components to be protected is a known safeguard.

i originally read about this some time ago at the Swiftech site ( www.swiftnets.com ), but its really geared for their product line now, but i managed to find a little read at benchtest.com related to this. see.... http://www.benchtest.com/condensation2.html

------------------
"Those who fail to recognize the past are condemned to relive it"

-Jim Jones

YARDofSTUF · Post by **YARDofSTUF** » Wed Jan 31, 2001 9:46 pm

I am waiting for a working method, hopefully my 7th or 8th pc will be AC cooled... so make it work! lol good luck man, i think its a great idea, and one that is most likely possible now or within the year.

------------------

cy88 · Post by **cy88** » Wed Jan 31, 2001 10:13 pm

I can't give any comment on this cuz i donno much about cooling...but all i can say it sounds like it's a great idea. Just give it a try, u never know until u tried....good luck man!

------------------

Gaming-Module · Post by **Gaming-Module** » Tue Feb 13, 2001 9:29 pm

Why not test your theories on a crappy system, not for overclocking or system performance, but rather to see if your countermeasures are effective enough to warrant usage with substantially more expensive equipment.

Just set up an old 133 system under said conditions before you start refrigerating 1500 dollars worth of computer components.

TonyT · Post by **TonyT** » Tue Feb 13, 2001 11:01 pm

It would be a lot easier to just run a 3 or 4" duct from an AC vent on a portable AC unit or from a vent line in the house wall. And use a stock cpu fan and an add-on exhaust fan in the case to pull the cool air through the case & then out of it w/ the moisture in the air. A thermostat could be placed in the case if a portable unit is used.

It seems that the main issue with your plan is the issue of condensation. Most of it will occur around the compressor/condenser unit itself, but some moisture will end up in the air that is cooled. Keep in mind that you should also plan on methods for compensating for the natural amount of humidity that already exists in air, as it will vary with weather conditions, even indoors.

The reason I brought up the idea of an AC unit is because the main thrust here is a cooler environment. AC units produce much less humidity than refrigerator units. And they are easier to regulate via a thermostat. They are quieter too. Refrigerator units are not designed to operate for extended periods of time. Whereas an AC unit is made to run & run & run.

LOL - My computer is in a cabinet I made next to my desk. It has a door but no back. On the wall behind the cabinet is the jack for the cable line from the pole. I made the hole large enough so a tiny bit of outside cold air gets sucked through by the fan in the case in the wintertime. Sometimes my tootsies get cold though!

mjfmlv · Post by **mjfmlv** » Wed Feb 14, 2001 12:05 am

Check out the Vapochill case at NFSExtreme.com, ATX tower with onboard refridgeration unit!