Heatsinks for DIY LED lamps

J

J-Icky

Well-Known Member
Ok so I was originally gonna use the arctic alpine 11 plus for my build but its becoming harder to find online and the ones I can find are more than I want to pay. I need 6, as I have 6 Vero 18s that i want to run at 1050mah, would the alpine 64 gt rev 2 work or the alipine 11 gt rev 2. The 64 gt rev 2 says its rated for 70w and the 11 gt rev 2 says 95w and I would only be pushing roughly 30w's on each one. I'm not sure how the CPU wattage is comparable to the wattage we use but I would think as long as they have the surface area to hold the cobs either of these should be more than enough right?
 
A

array

Member
J-Icky said:
Ok so I was originally gonna use the arctic alpine 11 plus for my build but its becoming harder to find online and the ones I can find are more than I want to pay. I need 6, as I have 6 Vero 18s that i want to run at 1050mah, would the alpine 64 gt rev 2 work or the alipine 11 gt rev 2. The 64 gt rev 2 says its rated for 70w and the 11 gt rev 2 says 95w and I would only be pushing roughly 30w's on each one. I'm not sure how the CPU wattage is comparable to the wattage we use but I would think as long as they have the surface area to hold the cobs either of these should be more than enough right?
Click to expand...
The arctic 11 plus showed up on ebay tonight at $15 + free shipping. Just grabbed two.
 
littlejacob

littlejacob

Well-Known Member
Bonjour
6 artic gonna cost 90 $!!! I believe it is less expensive to buy a hs and a 120mm fan...jmo!
Have a great day ★
 
nevergoodenuf

nevergoodenuf

Well-Known Member
Can someone do the math on a heat sink for me? I plan on using the hlg 320-36b at 95% efficient and 6 Vero 29s. I will be cooling this with fans but I want it to survive passive. it will be 40" long, but not sure how wide. The COBs will be 2, 2, and 2, with the heat sink bent down the center 5* or so. Will 2 heat sinks be cheaper or one piece. Would a 10" x 40" heat sink cool 320 watts?
 
salmonetin

salmonetin

Well-Known Member
...on other way... ...again my pov... on the pic... normally the fins are oriented parallel to large side... for pasive or semipasive way...or active way too...


...but with this fin orientation we loose... on large rectangular heatsinks...my pov its go with the central example...for rectangular heatsinks... ...similar to various single square heatsinks positioned in file.... one square heatsink for cob led... more fins per cob for my pov...;)




...depend on heatsink orientation too...



...yeah i know we normally use the horizontal heatsink position... 70%...

...for my pov if we use the fins oriented parallel to the short side... we gain on large heatsinks...for passive... or semipasive way...

...on this way we can place the cobs in the same axis...in the center...but its only my inexpert pov...;)

..thanks Wilson... maybe im too :joint::bigjoint:...then...

...e function desactived...thanks sanjuan...;)

:peace:

Saludos
 
Last edited:
disengaged

disengaged

Member
I don't have real spec's on this heatsink but ran across a pair at a garage sale. Wish I knew what it was for, if the mounting slots built into the design could be used in some way.
It is 300mm long (12 inches). The thickness of the heaviest part of the heatsink is about 3.5mm (0.15 inch). Weight is 535 g.
As far as a fan goes... there would be a 120mm fan quietly blowing air over it but not a raging noisy huge amount of air flow.

Guesses on if it could handle a couple of COB's? Haven't decided on what to use. Maybe one or two Vero 29's or CXA3070's running around 1050mA to 1400mA. Simply trying to light up a little cabinet. I don't have it in my current budget to buy 4 COB's at this time, going to need to start with two... which is enough, just not ideal spreading of light sources.
 

Attachments

B

benbud89

Well-Known Member
SupraSPL said:
its saul good man, a great thing to double check our maths, I am notorious for "careless mistakes" always rushing through too much math since 1st grade LOL. Anyway, looks like your surface area is very close to my figure. What is the formula in Box F4 of your first sheet?


Regarding PPFD, looks like you used AB bin and I used AD. Here is where my math came from, assuming CXB3070 3000K AD bin at 1.4A.

50W X 8 = 400 dissipation W
400 * .48 efficiency = 192 PAR W and 208W of heat
192 * .9 (lens/reflecter/scatter losses) = 172.8 PAR W / 12ft² = 14.4 PAR W/ft²
14.4 * 4.66 * 10.7 = 718 PPFD

208W of heat * 120cm²/heat W = 24960 cm² of heatsink surface area for passive cooling

He has 72" of width so if we take 6" of each end we get about 60" that should be covered by the heatsink, depending how close to the edges the COBs are mounted. The 8.46" X 62" is 24634cm², a bit less than the recommendation but should work well.

A pair of 10"X28" is 26,261 cm², slightly more than the recommendation but offers good spread and good opportunity for staggering. He could definitely tweak it a bit and with 26" pieces, The 10" is a bit cheaper/cm² and if it were me I would prefer the system in 2 pieces for adjustability and it works well as far as wiring, one driver and 4 COBs per heatsink. Then again if you have an even canopy, one heatsink might be a great way to go.
Click to expand...
So it is that simple when calculating the watts going from electricity to heat? Just multiply the total watt usage by the .%efficiency of the COB? If I go with a 52% COB wattage, then 48% is heat I have to account for, correct? Then it is a little easier to get close to the cm2 for passive cooling. So far I had been doing my possible surface area on account of the total wattage, which is a whole lot less now, if I understand this correctly. Not to bad. 120cm2/w, does that maintain a steady temperate in my heatsink, or what kind of rise should I expect in my heatsink? Thanks as always, Supra. So thankful for these RIU posts everybody, yourself including, are making. It's a real treat.
 
coolbreez1

coolbreez1

Well-Known Member
The alpine was good because it was only $10 at first, now it appears there has been a run of them and the prices are rising. I think as littlejacob said, cheaper to buy a heat sink and 120mm fan at that price point, and you are getting more bang for your buck.
 
SupraSPL

SupraSPL

Well-Known Member
benbud89 said:
So it is that simple when calculating the watts going from electricity to heat? Just multiply the total watt usage by the .%efficiency of the COB? If I go with a 52% COB wattage, then 48% is heat I have to account for, correct? Then it is a little easier to get close to the cm2 for passive cooling. So far I had been doing my possible surface area on account of the total wattage, which is a whole lot less now, if I understand this correctly. Not to bad. 120cm2/w, does that maintain a steady temperate in my heatsink, or what kind of rise should I expect in my heatsink? Thanks as always, Supra. So thankful for these RIU posts everybody, yourself including, are making. It's a real treat.
Click to expand...
That is correct, we used to recommend surface area/dissipation W but since there is such a wide range of efficiency now and because we have a better idea what % our efficiency actually is, it is possible to use surface area/heat W. General guideline, 40cm²/ W of heat for active cooling and 120cm²/W of heat for passive cooling, assuming the profile shape is well suited to the application.

The heatsinks will reach a stable temperature, probably within 30 minutes and will maintain it indefinitely. It can be tricky to measure the heatsink temp rise for a variety of reasons but I try to measure in the hottest place I can find, usually in between the fins behind the COB or possibly on the front of the heatsink right next to the COB. The photon output can interfere with that measurement and can be blocked with a piece of foil during the reading. You can try to use the Tc measurement point with a thermocouple but the same problems occurs, photon output can affect the reading. With these recommendations I am aiming for heatsink temp of 35C or less, assuming ambient temp of 25C. This will vary depending on active vs passive, heatsink configuration, drive current and air circulation in the grow space.

I think it makes sense for DIYers to try and keep temp droop under 4% when the system is running at max power, but that will also depend on your drive current. If you are willing to accept a higher temp droop, you can use less heatsink surface area, it will not harm the COBs significantly.
 
disengaged

disengaged

Member
The alpine were $7.50 when 4 or more were purchased from newegg a month ago but that has changed. How many home PC builders were really buying 4 or more? It was the only one I found that day with discounts for multiple purchases :) I think they know their market.

As far as the heatsinks I posted in #1149, there is a 50mm / 2 inch flat so a reflector would need to be used which cleared the fins which get in the way... and maybe cut some fins away. I can't remember where I read this, but anodized black may be much more effective at dissipating heat on the cooling fins which have a slight serration to them. The flat surface could use some polishing. Considering Vero 18's currently because of the fit, need to see if reflectors and the COB holders they attach to will clear. Will be running at 1050mA.
Probably a waste of aluminum to use them for red / blue panels.
 
littlejacob

littlejacob

Well-Known Member
Bonjour
I am happy to be back here...I almost died saturday...I was in a middle of a storm...maybe you heard about it on tv...I know it is nothing compared to a twister/tornado...but it was scary to see a car floating in front of me...trees rocs, metal pieces, refrigerator....CRAZY! And no net for 2 days...I am lucky!
Is anyone know what temperature oir hands can handle? To know approximately the temps of my hs and driver (this one is way hoter than my hs!)
Have a great day ★
 
disengaged

disengaged

Member
sounds like a crazy storm littlejacob, glad you are ok and hope it didn't stress your drivers!

I'm still running with the idea of using these and found more on ebay - searching "300mm heatsink".

Any guess on if they could handle two COB's each? 1050mA, either Vero 18 (6 total) or 29 (4 plus some supplemental)

They have that 50mm flat which could hold most COB's. There is that "ledge" which I could cut off so it wouldn't block light, or maybe there is a reflector in stock somewhere. Seems the cost of adding a reflector bumps it up a bunch. Are lenses needed for the 2x3.1' footprint with 45" between top of growing media and light hanging bar?

I played around with fans and the only one that made any sense would be mounting a 90mm as I mocked up in the second photo. I would run those fans at reduced voltage to keep the noise down.

They are "upside down" from how they would be used. I could not both cut off the section and mount a fan as shown since I would be cutting off the section I am mounting the fan to.

Sorry for lighting, I was using a low CRI LED shop light and a flash :)

heatsink3.jpg heatsink2.jpg heatsink3.jpg heatsink2.jpg heatsink3.jpg
 
robincnn

robincnn

Well-Known Member
I try to measure the heatsink temperature behind the COB between the fins as well. Within 10mm-15mm from behind cob center.
If the heatsink is thick the only accurate way to measure is to drill a hole so that the thermocouple can get close enough.
upload_2015-10-9_16-2-22.png
Source AN 30 bridgelux

Heat from LES was a big problem when i was measuring the case temperature. With Vero 29's I saw that the C13584_CLAMP-VERO29 was very effective in securing the thermocouple wire and isolating it from LES heat.

My estimate is M1 will exceed 50C even at 20 heat watt.
 
Stephenj37826

Stephenj37826

Well-Known Member
robincnn said:
I try to measure the heatsink temperature behind the COB between the fins as well. Within 10mm-15mm from behind cob center.
If the heatsink is thick the only accurate way to measure is to drill a hole so that the thermocouple can get close enough.
View attachment 3517926
Source AN 30 bridgelux

Heat from LES was a big problem when i was measuring the case temperature. With Vero 29's I saw that the C13584_CLAMP-VERO29 was very effective in securing the thermocouple wire and isolating it from LES heat.

My estimate is M1 will exceed 50C even at 20 heat watt.
Click to expand...
I was thinking so too but wonder how it would do with a decent breeze?
 
You must log in or register to reply here.
Top