stupid question on cheapest heatsink ever (flat plate)

[CobKits]

so lets be all caveman and take a 1/4" piece of solid aluminum bar which should be cheaper than a heatsink

if you had a 1 square foot piece of this, is it safe to assume that you radiate heat from top and bottom and so you count 2 square feet of surface area (negating edge thickness)

1 sq ft = 930 cm^2
2 sq ft = 1860 cm^2

used passively thats 1860/120 = 15.5 heat watts that can be dissipated
with fans blowing on both sides thats 1860/40 = 46.5 W

amirite? or do we only count the top (im thinking passive... on a large sheet where the heat could get 'trapped' underneath)

 

[420masterofdon]

on swiss grower.ch forum there is somebody using an old traffic sign as a heat sink try asking in the "english section"

 

[MeGaKiLlErMaN]

so lets be all caveman and take a 1/4" piece of solid aluminum bar which should be cheaper than a heatsink

if you had a 1 square foot piece of this, is it safe to assume that you radiate heat from top and bottom and so you count 2 square feet of surface area (negating edge thickness)

1 sq ft = 930 cm^2
2 sq ft = 1860 cm^2

used passively thats 1860/120 = 15.5 heat watts that can be dissipated
with fans blowing on both sides thats 1860/40 = 46.5 W

amirite? or do we only count the top (im thinking passive... on a large sheet where the heat could get 'trapped' underneath)

Do you have a link to the specific bar? I would like to take a look at it. Also do you have a link to your grows with monleds (Blurple)? I would like to see a well documented run.

 

[wietefras]

I'd be afraid that indeed the heat would become trapped underneath. I have the same problem with fluorescent lights in a small cabin. Those big reflectors really trap the heat too.

Maybe making some cuts through the plate might help?

 

[MeGaKiLlErMaN]

I'd be afraid that indeed the heat would become trapped underneath. I have the same problem with fluorescent lights in a small cabin. Those big reflectors really trap the heat too.

Maybe making some cuts through the plate might help?

That would also slightly increase surface area by the inside edges of the holes right?

 

[wietefras]

@MeGaKiLlErMaN , Yeah, although the cut would decrease the top and bottom surface, but I'd assume that is compensated by the inside edges.

 

[PSUAGRO.]

nothing new, stick a small fan on it and you have the hans panel cooling method............driver being remote makes this happen

http://hanspanel.com/index.php/ledgrow-led-panel-70-w

primitive but "functional"

 

[CobKits]

well i was thinking of it as take a 5x5 tent and cut yourself a 4x4 sheet of 1/4" aluminum (3 lb/SF so not ridiculous). you now have an unlimited template workspace to put rows of cobs, monos, far reds, etc etc. normal airflow circulation in there blowing on it, maybe a fan up top above it blowing across. put the money into chips instead of thermal management. use (4) smaller, cheaper 28mm 10 W cobs on 6" centers vs 40W 38mm cobs on 12" centers, which would result in more even heat transfer. heck you could even snake a ghetto water cool line across the top without any direct physical contact to the chips

 

[PSUAGRO.]

well i was thinking of it as take a 5x5 tent and cut yourself a 4x4 sheet of 1/4" aluminum (3 lb/SF so not ridiculous). you now have an unlimited template workspace to put rows of cobs, monos, far reds, etc etc. normal airflow circulation in there blowing on it, maybe a fan up top above it blowing across. put the money into chips instead of thermal management. use (4) smaller, cheaper 28mm 10 W cobs on 6" centers vs 40W 38mm cobs on 12" centers, which would result in more even heat transfer. heck you could even snake a ghetto water cool line across the top without any direct physical contact to the chips

^^^^^^^^^^Now that is new

 

[MeGaKiLlErMaN]

well i was thinking of it as take a 5x5 tent and cut yourself a 4x4 sheet of 1/4" aluminum (3 lb/SF so not ridiculous). you now have an unlimited template workspace to put rows of cobs, monos, far reds, etc etc. normal airflow circulation in there blowing on it, maybe a fan up top above it blowing across. put the money into chips instead of thermal management. use (4) smaller, cheaper 28mm 10 W cobs on 6" centers vs 40W 38mm cobs on 12" centers, which would result in more even heat transfer. heck you could even snake a ghetto water cool line across the top without any direct physical contact to the chips

There's cheap$12 100W water-cooling CPU blocks on newegg, they're the teal/light blue ones

That was where I was going with the flat metal! Birds of a feather on this one!

http://www.newegg.com/Product/Product.aspx?Item=9SIA3XT1R66839&ignorebbr=1

 

[Rayne]

I do not see why one could not build a COB based LED array using just a piece of sheet metal as the heatsink and the mounting surface. A lot of the old model MarsHydro units were built with a fan blowing air down and across the, sheet metal, LED circuit board.

 

[JorgeGonzales]

Courtesy of @mauricem00

 

[CobKits]

There's cheap$12 100W water-cooling CPU blocks on newegg, they're the teal/light blue ones

That was where I was going with the flat metal! Birds of a feather on this one!

http://www.newegg.com/Product/Product.aspx?Item=9SIA3XT1R66839&ignorebbr=1

exactly. even cheaper on ebayor alibaba. you could just chuck em up there with some thermal paste and go. eliminates the conundrum of mounting cob to waterblock. the secret is yeah we are are using a secondary medium between cob and waterblock but because out heat load is so low and so distributed it works in this case. you may not see the overall effect of cob case temp at 25C but at that low current it would still be awesome and with 1/4" plate you could have blind holes so the cob mounting screws dont penetrate all the way thru, and locate the waterblocks right on top of the cob locations

 

[CobKits]

@JorgeGonzales , what is the thickness of that material (or is it irrelevant)? looks like a 6"x6" is closest to 14x14mm and at 10 cob watts @ 60-70% eff. id be 3-4W of heat which is less than 10C temp rise. is that Tc or Tsink (or are they considered equivalent in this case?)

 

[MeGaKiLlErMaN]

exactly. even cheaper on ebayor alibaba. you could just chuck em up there with some thermal paste and go. eliminates the conundrum of mounting cob to waterblock. the secret is yeah we are are using a secondary medium between cob and waterblock but because out heat load is so low and so distributed it works in this case. you may not see the overall effect of cob case temp at 25C but at that low current it would still be awesome and with 1/4" plate you could have blind holes so the cob mounting screws dont penetrate all the way thru, and locate the waterblocks right on top of the cob locations

Exactly! Did you have a link for the metal youre talking about? Im looking on onlinemetals and a 48x48inx.025" sheet costs $350

 

[CobKits]

ill buy it locally. i think ill get a 12 x 48 x 1/4" and a 12 x 48 x 3/16" to prototype

1/8" might be presumptious

 

[JorgeGonzales]

@JorgeGonzales , what is the thickness of that material (or is it irrelevant)? looks like a 6"x6" is closest to 14x14mm and at 10 cob watts @

Good question, no idea. Maybe @mauricem00 can check his engineering book.

It definitely matters because it affects spreading resistance, which is complicated math I don't understand, but the the thicker the material the better the thermal spreading resistance.

Google gave me this:


Spreading resistance matters less with lots of diodes spread out evenly, and matters more you are using a relatively small amount of large heat sources spread far apart, like COBs, because you want the heat to spread to the entire surface efficiently.

 

[MeGaKiLlErMaN]

ill buy it locally. i think ill get a 12 x 48 x 1/4" and a 12 x 48 x 3/16" to prototype

1/8" might be presumptious

Hey what ever gets it done, let me know how that all goes as in even using 4-5" bars that would be great

 

[mauricem00]

Good question, no idea. Maybe @mauricem00 can check his engineering book.

It definitely matters because it affects spreading resistance, which is complicated math I don't understand, but the the thicker the material the better the thermal spreading resistance.

Google gave me this:
View attachment 3778506

Spreading resistance matters less with lots of diodes spread out evenly, and matters more you are using a relatively small amount of large heat sources spread far apart, like COBs, because you want the heat to spread to the entire surface efficiently.

in the graph I posted the thickness is 2mm and the material is aluminum. about the same as the aluminum PC boards used in many commercial LED grow lights.this PC board is the only heatsink used in 150 watt UFO grow light but in that application it has a cooling fan to lower TC

 

[MeGaKiLlErMaN]

in the graph I posted the thickness is 2mm and the material is aluminum. about the same as the aluminum PC boards used in many commercial LED grow lights.this PC board is the only heatsink used in 150 watt UFO grow light but in that application it has a cooling fan to lower TC

Nice, I'm not sure how thick I want to go for a large structure 3.18mm seems like a safe bet, obviously the thinner the better.

https://www.onlinemetals.com/merchant.cfm?id=76&step=2&top_cat=60&showunits=mm