Radiator Making

BroKilla93 · Post by **BroKilla93** » Thu Dec 07, 2006 8:57 pm

Ben, or who ever out there knows how to make a radiator i would much appreciate it. I have to make one for my Xbox 360 laptop. I would like it if someone could tell me the parts and show me how to make it (with pictures) or just let me buy a slightly (.5 or 4 inch of the hight).

Post by **gannon** » Thu Dec 07, 2006 9:30 pm

As far as I see it, there are 2 ways that are economical for a DIYer.
#1: take a block of aluminum and a router (or other cutting device) and carve out the excess metal so you have the fins / internal water routes.
#2: take a sheet of aluminum and bend flaps to create fins. You can then drill holes and route copper tubing through them and secure.

I recommend #2 since it's easier, cheaper, and probably more effective

Not sure of the exact dimmensions needed and the amount of heat dissipation it needs to put up with.

BroKilla93 · Post by **BroKilla93** » Thu Dec 07, 2006 9:35 pm

gannon wrote:As far as I see it, there are 2 ways that are economical for a DIYer.
#1: take a block of aluminum and a router (or other cutting device) and carve out the excess metal so you have the fins / internal water routes.
#2: take a sheet of aluminum and bend flaps to create fins. You can then drill holes and route copper tubing through them and secure.

I recommend #2 since it's easier, cheaper, and probably more effective
Not sure of the exact dimensions needed and the amount of heat dissipation it needs to put up with.

Thanks, but I dont see how to do it. Could you tell me everything I need. Then tell me how to do a bit (who im a kidding, a lot) more in depth. Could you show me a picture or good drawing or something showing me the steps.
Thanks im kinda slow at this stuff.

Advertisement: Get FireFox it spell checks

Post by **gannon** » Thu Dec 07, 2006 9:49 pm

(I have firefox, I just don't use the SP function much)

Anyways, (crappy gimp drawing

)

#1: Sheet aluminum
#2: Cut aluminum
#3: Stack pieces, clamp in place, drill holes
#4: Route copper tube through the holes (make sure it's a tight fit, seal with a good thermal conductor)

BroKilla93 · Post by **BroKilla93** » Thu Dec 07, 2006 9:58 pm

Oh your spelling was fine. Im just in love with FireFox it kicks -blam- against Internet Explorer.

I think i get it. How thick should the allumium be and also why does it work. (should i use a fan also or is that dumb.) Do you know of a site that tells you how to do it or at least shows me how it works, cause normally when i know that i can do the impossible (if it works!!!) Also i was planning on making is smaller, (do i need fins simmaliar to a heat sink??.)

See i resited the urge to spell check.

Also thanks for helping me. I see your the only one who did *gets 1 million dollar check and considers giving it to Gannon*

Also do you own that website http://gannon.tk/store.php
cause i could help out and give it shine.

Post by **gannon** » Thu Dec 07, 2006 10:09 pm

Yeah, my website, it's kinda got tossed in the corner while I do more work on the new http://portablesofdoom.org/ and a few other sites. I even have a brand new code/design done for it. Just never got around to putting it all up

Anyways, 1/8" should work for the aluminum. It works because the hot air/liquid (fairly sure vegetable oil would be the best/safest) inside the copper tubing transfers it's heat (through the copper which has great thermal properties) to the aluminum (which has slightly lesser thermal properties but is cheaper/stronger IIRC) which then with the heightened surface area lets the air absorb the heat more easily.
Oh, and I recommended the oil because incase of a leak it wouldn't destroy the electronics since it's dialectric.
If you want to use a fan it'd help. I'd use a fan on the inside of the case pulling air in (through a dust filter) straight to the radiator, with holes on the oposite side of the case to aleviate pressure/let air out.

BroKilla93 · Post by **BroKilla93** » Thu Dec 07, 2006 10:14 pm

does the stuff here ruin electronics, and also is the oil a good coolant or just something that wont ruin the electronics (there was this guy who filled his computer up with the oil it was cool
http://www.koolance.com/shop/product_in ... cts_id=265

Post by **gannon** » Thu Dec 07, 2006 10:36 pm

Not sure since they don't say what the coolant is. I wouldn't be surprised if it was just colored water though

timmeh87 · Post by **timmeh87** » Thu Dec 07, 2006 11:17 pm

so for a coolant you are looking for two things: high heat capacity (the amount of heat one unit of coolant can carry away) and low conductivity.

so heres a small list i compiled for you with heat capacities listed in joules per kilogram per degree kelvin.

conductive
--------------

water - 4181
mercury - 139

non-conductive
-------------------

ethanol (alcohol) - 2440
ethylene glycol (antifreeze) - 2450
glycerin - ~2000
vegetable oil - 1670
air - ~1000

unfortunately, heat capacities are by weight, not by volume. so a Kg mercury holds 3.3% as much heat as water, but also occupies 7.4% as much space. (making water still superior as a coolant).

the number for antifreeze is also for PURE ethylene glycol, most commercial antifreeze is something like a 40% solution in water.

anyways. id say veg oil or water. cheapest and safest.

or else, look into a product called "florinert"

http://en.wikipedia.org/wiki/Fluorinert

Post by **gannon** » Thu Dec 07, 2006 11:40 pm

Given those numbers and some other info:

Vegetable Oil:
Density: .894 Kg/L
Conductance: 1670 J/Kg/K (Or J*K/Kg)
Dielectric: Yes

Water:
Density: .932 Kg/L
Conductance: 4181 J/Kg/K (Or J*K/Kg)
Dielectric: No

Since this is portable, let's say you have .5L of liquid, so:
(.894 Kg/L) * (.5 L) * (1670 J*K/Kg) = 746.49 J*K with Vegetable Oil
(.932 Kg/L) * (.5 L) * (4181 J*K/Kg) = 1948.346 J*K with Water
So if you risk breaking the electrics you get a 261% increase in cooling. As long as you build it right and fully test it before hand that might be the way to go. Although it's better safe than sorry.

timmeh87 · Post by **timmeh87** » Fri Dec 08, 2006 12:02 am

isnt the density of (pure) water 1? isnt that like some sort of fundamental fact? (did you do a conversion from imperial or something?)

sorry to be a nitpicker, but i think you also meant 'capacity' instead of 'conductance'. and the units are more correctly J/(Kg*K)

(since university ive become anal with my oder of operations).

anyways. another thing to consider is the actual thermal conductance. something might hold lots of heat but be unable to absorb it very quickly. if i had to guess id say water has a higher heat conductance too.

im pretty sure aluminum has more capacity than copper (about twice as much per kg) but less conductance (almost half), which is why copper is preferred as heat sink material.

some useful numbers here:
http://en.wikipedia.org/wiki/Thermal_conductivity

Post by **gannon** » Fri Dec 08, 2006 12:31 am

timmeh87 wrote:isnt the density of (pure) water 1?

No, just really close. The newer unit is g/cm^3 where water is even closer to 1.

The maximum density of pure water at a pressure of one standard atmosphere is 999.861kg/m3; this occurs at a temperature of about 3.98 Â°C (277.13 K).

timmeh87 wrote:sorry to be a nitpicker, but i think you also meant 'capacity' instead of 'conductance'.

Perhaps, I meant it as the amount of heat it could conduct and disipate.

timmeh87 wrote:and the units are more correctly J/(Kg*K) (since university ive become anal with my oder of operations).

Then why did you say:

timmeh87 wrote:listed in joules per kilogram per degree kelvin.

Which is J / (Kg / K) ?

BroKilla93 · Post by **BroKilla93** » Fri Dec 08, 2006 3:53 pm

gannon and timmeh thanks for helping me out. So all you need to build a radiator is aluminum, copper, and copper fins. I got a wee bit confused on that part again, maybe (pretty pls) you could make one for me and i could buy it, or you could give me a more in depth version of how to make (for someone who flunked that class but, got an easy A in electronics)

Oh i understand joules per kg/l. Something like that (yeah right, but now i have an idea of how to do it)

So am i correct in saying that by getting aluminum and drilling holes in it (see gannon's wonderful illustration) and sticking copper pipes in and bending them so they go back and forth about 6 times that the heat will magically be transfered. Also do i need coper or aluminum fins and/ or a aluminum cover.

timmeh87 · Post by **timmeh87** » Fri Dec 08, 2006 4:01 pm

well i was talking in g/cm3... using the definition of a liter to be 1000cm3 (or 1m3) and the definition of a kg to be 1000g, then i have proven that 1000*kg/l, 1000*kg/m3, and g/cm3 are in fact the same units.
so.. water shouldn't be 'even closer' with g/cm3 over kg/l

not that its a real big issue anyways. your math remains valid.

also, when i said 'joules per kilogram per degree kelvin' i was just. well i guess i wasn't thinking about it very hard. i assure you that the units from my last post are the correct ones. sorry for any confusion.

BroKilla93 · Post by **BroKilla93** » Fri Dec 08, 2006 4:15 pm

Look two post up i just edited it...