reldnahkram: (Default)
[personal profile] reldnahkram
It seems the computer case I want doesn't exist. Given my needs, I think microATX will be fine for a motherboard, and thus I can get a smaller mATX case. But I also want 120 mm fans, and they don't make small cases with big fans. Well, there's like one or two out there, but they're ugly as sin and expensive. I guess I'll just settle for a bigger case when it comes time to build (which will give me room to do stupid things like experiement with water cooling in the future) or get a small case and drill out the fan holes to 120 mm myself.

Date: 2006-01-26 04:17 am (UTC)
From: [identity profile] tirerim.livejournal.com
Clearly, you need one of these, except they don't exist yet.

Date: 2006-01-26 05:02 am (UTC)
uncleamos: (Default)
From: [personal profile] uncleamos
Whoa, cool.

Date: 2006-01-26 10:57 pm (UTC)
From: [identity profile] reldnahkram.livejournal.com
The point of this exercise is not only to get a quiet computer, but to do it myself. Sure I can by some off-the-shelf system, but I want to put the components I want into it, and then have it be quiet. Also, the Totemo uses a picoBTX motherboard, which locks me into an Intel processor, and AMD right now is better in terms of power usage and performance.

Date: 2006-01-26 06:27 am (UTC)
From: [identity profile] sildra.livejournal.com
Water cooling's more trouble than it's worth. Before you know it, you've flooded the lab three times. A bit of advice, though: a blowtorch solves everything. "When all else fails, use a blowtorch" is the wrong way of seeing things--the best way is to just use a blowtorch before anything has a chance to go wrong.

Date: 2006-01-26 07:04 am (UTC)
From: [identity profile] tirerim.livejournal.com
Good point. In lieu of a blowtorch, I suggest going directly to jet engine cooling.

Date: 2006-01-26 02:07 pm (UTC)
From: [identity profile] sildra.livejournal.com
Oh, I meant blowtorches fix everything water cooling related. Or, at least, that's been my experience.

Date: 2006-01-26 10:53 pm (UTC)
From: [identity profile] reldnahkram.livejournal.com
But that's not very efficient for indoor use or for consistantly cooling a computer, nor do I think it could actually provide enough cooling for a computer processor.

Date: 2006-01-26 10:49 pm (UTC)
From: [identity profile] reldnahkram.livejournal.com
I don't know what kind of water cooling you're talking about, but by no means would it use enough water to flood a lab. And most water cooling rigs I've seen use flexible pipe that would probably melt if blowtorched. I think it'd be extremely cool to do water cooling with neatly arranged rigid pipe attached to the sides of the case, either on the inside or the outside, but I'll start with the flexible stuff, because it's easier.

And if not water cooling, then what? Immerse the whole thing in mineral oil? Because that's easier than water cooling. And besides, this is fun, it's not fun if there aren't a few extra problems to solve :)

Date: 2006-01-27 12:06 am (UTC)
From: [identity profile] sildra.livejournal.com
Oh, I use flexible plastic tubing too, for the parts where the water cooling is actually near the electronics (except in one case where I have a big grounded copper plate for my circuit, rather than a circuit board, and I soldered a copper pipe to the bottom of the plate). The blowtorch was for attaching the valves and stuff to the water coming out of the wall. And the water coming out of the wall was what was sufficient to flood the lab. Three times.

Date: 2006-01-27 12:10 am (UTC)
From: [identity profile] reldnahkram.livejournal.com
Water coming out of the wall?

Standard computer watercooling uses a closed system, using an aquarium pump for pressure and a small radiator with a fan for thermal dissipation. Copper waterblocks are commercially available for most standard processor types, video cards, and the occasional other computer part.

Date: 2006-01-27 12:35 am (UTC)
From: [identity profile] sildra.livejournal.com
Well, I couldn't use a fan for thermal dissipation for two reasons.

1) It might have been insufficient (unless there was a really big fan).
2) It would have been way too loud. Our heat exchanger is only barely quiet enough as it is, and it uses the chilled wall water for heat dissipation.

The water actually going to the electronics is a closed system--it all has to be distilled water so as not to muck up the heat exchanger.

Date: 2006-01-27 12:47 am (UTC)
From: [identity profile] reldnahkram.livejournal.com
Hmm...how much thermal energy are you trying to dissipate? Super-extreme computer folks use things based on the Peltier effect to cool things even more, but that's just absurd.

Date: 2006-01-27 12:52 am (UTC)
From: (Anonymous)
DISTILLED WATER = GOOD
DISTILLED WATER = GOOD
DISTILLED WATER = GOOD

Three liquid cooling mottos to live by...

Date: 2006-01-27 12:53 am (UTC)
From: [identity profile] sildra.livejournal.com
Only 1 or 2 kW (that's 1-2 kW of heat, not 1-2 kW going into the system). Our system is definitely overkill. We run it at 45 A, but at fairly low voltage. The reasons we have it are, as I said above, air cooling would be too loud. And also overheating would be a catastrophic (weeks of work and at least $1000) failure.

Date: 2006-01-27 12:58 am (UTC)
From: [identity profile] reldnahkram.livejournal.com
Ah, there's the difference. Most computer processors are under 100 W.

Profile

reldnahkram: (Default)
reldnahkram

September 2016

S M T W T F S
    123
45678910
11121314151617
18192021 222324
252627282930 

Most Popular Tags

Style Credit

Expand Cut Tags

No cut tags
Page generated Apr. 11th, 2026 08:54 pm
Powered by Dreamwidth Studios