Convection radiator design

[Khledar]

Actually, the surface causes a change in both emission and absorption or radiative heat. "Black bodies" are good at emission and absorption of radiative heat, not one or the other. I'm not sure what the "first principles" are, because they don't seem to agree with the laws of thermodynamics.

Radiation is fairly counter-intuitive in some ways - compared to conduction and convection, radiation is considerably more complex.

Finally - found a link that's not too cryptic:
http://www.phys.virginia.edu/classes...radiation.html
And some sample emissivity values:
http://www.engineeringtoolbox.com/36_447.html

However, I still don't think the radiative energy is going to make much of a difference in a radiator application (ironic they're named the same).

Originally Posted by #Rotor

so let me get this right.... a black object will radiate energy better than say a white or shiny chrome object... Is it not the other way around... and in "other way around"... I'm referring to a black object absorbing radiated energy ( light for instance) better. I can not see how the color can have any effect on an object's ability to radiate energy.

from first principles, the color of any object is but nothing more than a reference to it's ability to not absorb a particular set of frequencies in the EM spectrum. Thus an object to appear black in color, needs to absorb close to 100% of the energy in the complete visible spectrum that falls on it.

[lolito_fr]

nice fan you got there bladerunner
what power is that thing rated for??

[Butcher]

And I thought my 172mm fan was big.

[AntiBling]

Exactly. heh heh I think I've got fan envy.

[BladeRunner]

Well I thought if I must be forced to use a proper rad & fan, lets be really silly about it

The Kenlowe fan will be used on a car at some point, (as that's what it's designed for, additional cooling if towing, or for custom cars... big V8's in smaller bays etc).

If you want a fan like that, a cheaper option is your local car dismantlers. Just about every front wheel drive car has 1 or two fans like that on the Rad, (but they may not look quite as sexy as the kenlowe)

It's supposed to be 12v but I've been running it at 3v, 5v & 7v for a month and it has no start up problems at any of those voltages, nor has it appeared to affect it. It uses 2.4 amps @ 3v, 3.2 amps @ 4v and 4amps @ 5v. While it couldn't ever be classed as near silent, Its pretty quiet @ 3v under the desk... just a soft woosh... quite soothing actually.....wouldn't want it right next to my ear though I'm going to put it and the rad in the loft soon, (was 6C up there last night)

[iroc409]

good thread, i spent a couple hours tonight while doing some other things pondering this very subject. i've got a pc60, and i love the case. i'm personally the type that prefers a rather subdued look to things, more of a 'sleeper' look, and i'm not much one for 'rice-like' eye candy. i prefer more the eye candy like what bladerunner produces, eye candy in craftsmanship. 'elegance' i suppose.

anyways, i was trying to think how to put something together for my lian li. i really don't like the idea of cutting up the case. i was planning on using a bim2, but i just don't think that would be enough heat dissipation for a usual cpu/gpu/nb setup. i'd have to go with a bigger radiator, and to make it work right i'd have to open up the front. i wouldn't have a problem cutting up the inner chassis, but i really don't want to rip a giant hole in the front facia. i think that would run the lines of the case.

so, that means a radiator outside. but, the thought of having fans whirring outside of the case and a fairly ugly traditional radiator hanging out makes me cringe, not to mention the noise it adds.

i've seen the design like this ( http://de.msnusers.com/_Secure/0SgA...452648696800048 ) before, and it's quite intriging (SP). i think if finished properly and done carefully enough, it could still have a fairly unobtrusive and elegant look to it, and still function properly.

i guess i'd like to know as well if something like that can blow off enough heat. i've also been thinking about the possibility of using a more block-style radiator. something that in appearance looks perhaps more like a water block (sort of). think kinda like those older coolermaster aluminum hdd cooler jobs, but with water running through and larger fins. maybe using a scaled-up fin design like those small passive heatsinks used for nb or sb cooling (comes with the antec chipset kit). the inside would possibly be like old-skool maze blocks (not necessarily danger den's, but just the older wc blocks in general with the internal "maze" like designs. then, the radiator would be flat on the back and firmly attached to the side panel of the case. since the case is all aluminum, i would think perhaps with a design like this, you might get extra heat dissipation from the case itself (i don't see why we can't use these kinds of cases more for heat dissipation. if a flat piece of al or cu works for memory and stuff, why not with a whole case?).

the second idea would be a bitch to have machined... i certainly don't have the ability to do it, and i'm sure it would cost quite a bit to have it made. so it isn't something i'd probably just go out and try, unless someone could at least partially validate my idea. i'm not really convinced that it'd work well, but it could. keep in mind it would be fairly large... 1/2 to 3/4 the size of the side panel of a mid tower.

imo, it would have to be built out of aluminum, and finished nicely (preferrably to match the case finish as much as possible), to match up and maintain a elegant and unobtrusive look.

the more i think about it, the finned tube design would probably look nicer, but the block design would maybe be more efficient? i dunno, i'm just pondering at this point. i might call up some metal shops around and see if they have any finned tubing like that, i didn't think that would be readily available. a setup like that i would think could also be very low restriction, which may or may not be a blessing.

i'm also looking for a reasonably low-maintenance system, which would be another plus for this setup (less mechanical parts), amoung other small things i've been planning to add to a new system. but this isn't really the thread for that.

let me know what you think. maybe i'm all wet (tee hee).


edit:

also, for me free airflow isn't a huge issue. my machines sit under my desk, but it's a fairly large desk and it's one of them glass jobs with a tube frame, so there's no real airflow restriction from the furniture (aside from the obvious desk surface above the machine... about 18" or so).

also, on the fans, you can pick up some reasonably priced new ones in all sorts of sizes from jeg's (www.jegs.com). i think they start around $30 or so and go up from there... anywhere from i believe 10" to 16". but they'll pull several amps @ 12v, and are quite noisy (think of the fan you hear in the summer from a late model fwd when the a/c is on).

[Butcher]

Originally Posted by BladeRunner

It's supposed to be 12v but I've been running it at 3v, 5v & 7v for a month and it has no start up problems at any of those voltages, nor has it appeared to affect it. It uses 2.4 amps @ 3v, 3.2 amps @ 4v and 4amps @ 5v. While it couldn't ever be classed as near silent, Its pretty quiet @ 3v under the desk... just a soft woosh... quite soothing actually.....wouldn't want it right next to my ear though I'm going to put it and the rad in the loft soon, (was 6C up there last night)

I'm not aware of any fans that have issues with insufficient voltage. Generally not providing enough power to a motor just makes it run slower.

Iroc409: your link is broken...

[8-Ball]

Originally Posted by Butcher

I'm not aware of any fans that have issues with insufficient voltage. Generally not providing enough power to a motor just makes it run slower.

I understand, that they will simply run slower. The point I was asking about, was that I was sure that some motors (in various applications) struggle to START with lower than spec voltage, but once started, can be lowered significantly.

8-ball

[Butcher]

Originally Posted by 8-Ball

I understand, that they will simply run slower. The point I was asking about, was that I was sure that some motors (in various applications) struggle to START with lower than spec voltage, but once started, can be lowered significantly.

8-ball

Yes, my raditor fan has a startup voltage of around 6.75V (it's a 24V fan), but it's stall voltage while running is about 5.9V. I run it on a temp sensor based on water temps which means it typically runs at 7-11V.

[WAJ_UK]

just thought I would add a few emissivity values that might be of interest all at 300K

Polished copper:0.03
Polished aluminium:0.03
anodised aluminium:0.84
Black paint (parsons):0.98
white acrylic paint:0.90
white zinz oxide paint:0.93

so you see colour isn't really all that important. maybe the best solution would be to dip it in an anodising bath

[lolito_fr]

hmmm. WAJ_UK, thats weird ??? Need to check those figures.
you got me asking myself questions...

[8-Ball]

The reason the alumium and copper figures are so low is because they are polished.

Polished surfaces have low emmission and absorption values.

A VERY thin layer of anodising would be good, since the conductivity of alumina ain't that great.

8-ball

[iroc409]

the link was one of the images from the 2nd post or whatever from first page, the passive aluminum tube radiator (finned tubes).

i've been thinking about my idea a little more, and i think i've come up with a solution i can live with. build the finned tube radiator on top of the case instead of the side, and use the 80mm blowhole built into the top of the case. granted, it's not a lot of airflow, but would probably greatly improve the heat transfer.

if the pipes were mounted widthwise, it would be much prettier to plumb, but lengthwise could manage the airflow considerably better, but the plumbing would be a lot nicer the other way. also, there would be more tubes in parrallel, so the water flow would probably be less that way.

eh?


oh yeah, anyone online that might sell that finned aluminum stock?

and you still could have a fairly nice "block" style rad built for the top, and build it to use the air a little better (and could be made to match well), but the machining would be insane methinks for that kinda work. so it's probably not a very viable solution, unless someone here wants to attempt it on their own equipment.

[iroc409]

moo!

this is what i'm talking about... but possibly custom building it, if i can find that tube. finned tube i can find, but not that particular style...

http://www.highspeedpc.com/Merchant2...=InnovaKonvect

[lolito_fr]

I like your idea iroc, however I've realised there's a much easier way for humble diy-ers like me to make something that works. In fact redleader might be chuckling when he reads this

So yesterday, got itchy and started making a proof-of-concept small-scale prototype radiator, for testing purposes. I never got round to doing any tests because I got diverted from barking up the wrong tree(?) by Groth's link(nice one!) for a review of the Xice passive cooling kit, which got me slightly excited. Even more so when I saw what lay inside that £200 box!
So of course, I don't have £200 (to spare) but I do happen to have 7.5m of ¼" black micro irrigation pipe in my garage. Isn't life sweet?

I also happen to have a 10ohm power resistor and a variable power supply, as well as a couple of thermometers. So after spending a couple of hours taking temp measurements, I now know that I need approx 30m of pipe to keep a 100W load at less than 10°c above ambient.
Hopefully this will be enough to keep CPU, GPU and HDD cool (warm?) in summer, at full load.
(tests reveal that with 7.5m of pipe, dT =10°C @ 25W and 17°C @ 45W)

So at this stage I'm wondering how I can pack 30m of this pipe in a small a footprint as possible??? Obviously there needs to be some space between the loops for the air to move around…
Also I can't just use one single 30m line, because the flow rate would be non existent. So that means working out how many lengths to put in parallel…more Excel fun ahead: )

[pelle76]

OT: About fans not starting at low voltages. By adding a capacitor to a simple Fancontroller circuit, the fan will get an extra boost at startup so that it starts spinning.

[iroc409]

Originally Posted by lolito_fr

I like your idea iroc, however I've realised there's a much easier way for humble diy-ers like me to make something that works. In fact redleader might be chuckling when he reads this

So yesterday, got itchy and started making a proof-of-concept small-scale prototype radiator, for testing purposes. I never got round to doing any tests because I got diverted from barking up the wrong tree(?) by Groth's link(nice one!) for a review of the Xice passive cooling kit, which got me slightly excited. Even more so when I saw what lay inside that £200 box!
So of course, I don't have £200 (to spare) but I do happen to have 7.5m of ¼" black micro irrigation pipe in my garage. Isn't life sweet?

I also happen to have a 10ohm power resistor and a variable power supply, as well as a couple of thermometers. So after spending a couple of hours taking temp measurements, I now know that I need approx 30m of pipe to keep a 100W load at less than 10°c above ambient.
Hopefully this will be enough to keep CPU, GPU and HDD cool (warm?) in summer, at full load.
(tests reveal that with 7.5m of pipe, dT =10°C @ 25W and 17°C @ 45W)

So at this stage I'm wondering how I can pack 30m of this pipe in a small a footprint as possible??? Obviously there needs to be some space between the loops for the air to move around…
Also I can't just use one single 30m line, because the flow rate would be non existent. So that means working out how many lengths to put in parallel…more Excel fun ahead: )

hrm, some interesting ideas. what about using looped/spiraled tubing a certain size, but adding several of these spirals .5-1" apart for parallel flow action? then you'd have several smaller spirals, which you might be able to pack in quite a bit in a small space. just a thought.

unfortunately for me, i don't really have a lot of resources readily available for DIY stuff. so i'm kindof between a rock in a hard place of either buying it premade, or having someone build it for me. both can as we see be fairly costly

if i get some extra time later today i'm going to see what i can find from some local machine shops.

[Butcher]

Originally Posted by pelle76

OT: About fans not starting at low voltages. By adding a capacitor to a simple Fancontroller circuit, the fan will get an extra boost at startup so that it starts spinning.

For that sort of circuit you'd want a higher rated transistor. The BC337 is only rated at 625mW, while even a relatively small 80mm fan can pull up to about 1W. Also the transistor would probably need a small heatsink.
If low speed running is desired you also might want to look into PWM fan controllers, as that allows the motor to run at much lower speeds with high torque. PWM does tend to make motors run relatively hot and it can make them buzz.

[AntiBling]

I took a different if less elegant approach to get the fans to start every time. I use a relay for each fan with the common side to the fan, n.o. side to a pot to adjust the slow speed voltage and the n.c. side to 12V directly. The relay is triggered by a make-on-delay timer that energizes the relay 2 seconds after power up. ( the timer is adjustable but 2 seconds is enough to be sure they are spinning ) So on start the thing whirs to life in a huge rush of air, then CLICK! and it quiets down. Sounds kinda cool... it may be crude but its been in service that way for about a year with no problems. If the relay ever fails to fire, the fans will just stay on 12V.

[V12|V12]

Guys,
This whole thing about painting it black and it will magically radiate heat better is a farse... WHen I was working at OSU.edu, while attending classes, I was involved with there college Formula-1 competition. And we did some testing and what not and we anonized the engine black and other misc temp sensitive parts, but it wasn't just some cheap paint that you get at Home Depot or the like, it's either anonizing or SPECIAL conductive paint! The stuff that you get at regular stores will actually insulate temps! The whole thing about black radiating heat better is true IF 1) The paint is designed for it, or it's an anonizing process, and 2) the coated surface radiating the heat is NOT, I repeat NOT exposed to any heat radiating surface, or else it will absorb MORE heat that the benefits of the coating! Hence the engine block and parts are not exposed to the sun or any heat emitting surface! Relatively speaking of course, there always something emitting heat or reflecting it back...

Just think about it, once you coat something with a store bought paint, which is composed of a color pigment and some sort of epoxy/glue lacquer to make it stick, you have formed another layer which will insulate heat. And the color is only going to effect it, IF there is another source of heat output near it. IE paint 2 boxes; 1 black, one 1white, now put something in the boxes that gives off heat. Take the boxes and put them in the sun, the white box internal temp will be cooler than the black, because it's reflecting some of the extra outside heat load being introduced. Now take the boxes out of the sun and into a dim lit room. The black one is not going to be any cooler than the white one, based on using the same paint type and internal heat source. Paint it black or red or blue etc... it's going to insulate temp, but by how much? NOT A WHOLE LOT! Unless you're caking the layers on and using some mighty thick primer, it's NOT going to hurt temps with in a noticeable amount, I'd say do it if you like, it's not like all of a sudden your temps are going to rise 1C cause you added a THIN spray paint layer. Someone want to try a test on that? Hell just get 2 copper pipes, run a pumped water loop with them, paint one black or whatever color and leave the other one bare. See what temps the pipes are after a good 10min. I'm going to bet the painted one doesn't show a concerning temp diff...

[AntiBling]

http://www.chembio.uoguelph.ca/educm...p/blackbod.htm

Like V12 said, dont worry about it. The effect is too small to care about if all you are doing is cooling a computer. I doubt you could even measure it outside a physics lab.

[lolito_fr]

Well I'm not worried
Interesting feedback anyway.

Iroc, did you get anywhere with the finned tubing?

I bought 50m of the plastic pipe stuff for 10€, which is really dirt cheap. I think I know how I'm going to put it all together, and hopefully it won't be in the least bit painful
The pic gives an idea of how the pipe will be coiled, although the frame will be nothing like the one shown.
The end result will be quite slim, roughly 8cm, and will be mounted on the side of the PC case. It will be encased in some kind of grill for looks and protection in transport.
ATM it looks like I'll be using 10 or more parallel loops, approx 3m each, interconnected with copper manifolds.
Maybe this weekend it will take shape…

[Pears0]

As it happens rotor, its right in both cases. matt black is both the best at absorbing and givin off raidiated heat. Life trys to keep things balanced. So if it absorbs somethin easily, it also tends to give the said object easily. Take a sponge and a rock. Rock will absorbe water very very slow. But ull b dambed if u can squeese water out of a rock. And the opposite is true with a sponge. Its just how life tends to work.

Edit: If ur confused, i was to n thought the end of page 1 was te end of the thread, lol!

[V12|V12]

Yeah gloss paint is not good for radiating heat, since it has a smoother finish and doesn't have the air pocket like pits of a matte surface/paint... I like matte, all that shiny stuff get's smudged with finger prints and dust shows, esp on black! Like my case...

[joesgarage11]

I've had some experience with passive radiators in my water cooling setup. Been running two 10' lengths of 3/4" copper tube connected in series with no fan on them for about 6 months. The CPU temps stay 20-25 degrees C above ambient. The tubing rad seems to reach a state of temperature equilibrium after running the computer for about 30 minutes.

My first attempt at passive had 4 shorter lengths of 3/4" tubing with fins soldered on, but the fins didn't help the cooling much. BTW, my rad is painted black. Not for extra cooling, but it just looks better that way.