DiY LED - Cree CXA3070

bicit

bicit

Well-Known Member
SupraSPL said:
Here is what I have for the 2.079" profile:
View attachment 3266048

Other than surface area, the other things to take into consideration are distance between fins and baseplate thickness. A large distance between fins makes for better passive cooling ability. Thick baseplate allows the heat to spread to the fins faster which is very helpful if your heat load is concentrated in one area, example COBs. Thicker baseplate adds weight and material to the heatsink so it increases the cost without increasing the surface area. That can make it harder to compare the cost of one heatsink to another.
Click to expand...
Can I see the equation that you used to calculate the perim value? I'm having a hard time getting the same numbers. I think I'm just setting it up wrong.

Guod- Are the vero 10's really only 30% efficient at 350ma?

I should clarify, I'm considering a semi-passive design similar to the T-lights Mr.Flux built but with a pair of efficient 140mm fans blowing on the heatsinks while still being able to function in the event of a fan failure. Would the 1.813 profile still be inadequate? Would the 2.079 function any better under these circumstances?

Supra- Is your 110 cm^2/w figure for raw aluminum?
 
SupraSPL

SupraSPL

Well-Known Member
bicit said:
Can I see the equation that you used to calculate the perim value? I'm having a hard time getting the same numbers. I think I'm just setting it up wrong.

Guod- Are the vero 10's really only 30% efficient at 350ma?

I should clarify, I'm considering a semi-passive design similar to the T-lights Mr.Flux built but with a pair of efficient 140mm fans blowing on the heatsinks while still being able to function in the event of a fan failure. Would the 1.813 profile still be inadequate? Would the 2.079 function any better under these circumstances?

Supra- Is your 110 cm^2/w figure for raw aluminum?
Click to expand...

perimeter=(N2*(N4*2))+(N5*2)+(N6*2)
N2 = number of fins
N4 = fin height
N5 = riser height
N6 = width (cross section)

Vero10 3000K @ 350mA = 33.3% efficient

You raise a good question bicit, I do not know if heatsink USA heatsinks are pure aluminum or an alloy. I will check on that.
 
C

churchhaze

Well-Known Member
Random thought...

Would it be possible to use a calorimeter to determine efficiency if we could somehow isolate the entire heatsink inside the calorimeter?

My idea is that if you know the amount of electric power you're using, and you know how much heat is being dissipated, isn't the difference the power of light emitted?
 
SupraSPL

SupraSPL

Well-Known Member
Yes it would be possible. I have done very basic tests based on that to compare LEDs. It works great for qualitative comparison but my method wasnt accurate enough for quantitative. I think Mr Flux has done this type of test.
 
bicit

bicit

Well-Known Member
SupraSPL said:
perimeter=(N2*(N4*2))+(N5*2)+(N6*2)
N2 = number of fins
N4 = fin height
N5 = riser height
N6 = width (cross section)

Vero10 3000K @ 350mA = 33.3% efficient

You raise a good question bicit, I do not know if heatsink USA heatsinks are pure aluminum or an alloy. I will check on that.
Click to expand...
Looks like they uses 6063 T5 and 6063 T6 aluminum which are alloys. However they are basically identical for our purposes.

What I meant to ask if it was for unfinished aluminum? Because anodizing can improve passive cooling performance 3-8% from what I've read.
 
Last edited:
stardustsailor

stardustsailor

Well-Known Member
bicit said:
Looks like they uses 6063 T5 and 6063 T6 aluminum which are alloys. However they are basically identical for our purposes.

What I meant to ask if it was for unfinished aluminum? Because anodizing can improve passive cooling performance 3-8% from what I've read.
Click to expand...

Heatsinks mainly convect heat with air ...
( just a type of ' conduction ' between different states { gas-liquid-solid } of matter -)
That convection can be static (passive cooling ) or forced (active cooling ) ...

If the heatsink is anodised black ,now some heat is radiated from heatsink ...
(Quite more than a none anodised heatsink.)
But ...
8% is way much of a percentage and maybe applies for low levels of heat convected ,
thus heat radiation becomes significant ...

In led array applications ,where some serious amounts - several Watts - of heat power have to be dissipated ,
then the real percentage of heat radiated from a black anodised heatsink is at 1% level or even lower,at 99% of occasions ...

Anodising a heatsink ,serves really , if the heat dissipated remains at low / moderate levels ...


Cheers.
 
Last edited:
bicit

bicit

Well-Known Member
epicfail said:
I had my heatsinks anodized for protection and durability but choose black to help dissipate heat better. They will last through years of upgrades if that's what I decide to do with them.
Click to expand...
From the looks of it, while black is the most efficient color, it really makes no difference. Could add some creative flare do a nice technicolor acid wash finish. Anodizing is pretty straight forward.
 
Dloomis514

Dloomis514

Well-Known Member
Gonna pull the trigger on a 3070 @ 3000K, any of the high bin parts (Z2 or above) out there.

I see
http://www.mouser.com/ProductDetail/Cree-Inc/CXA3070-0000-000N00Z230F/?qs=/ha2pyFadugY1sJGAfyUsZl4AheJW3WukO5VbBRSZ5E=
for $40.

Looking at
http://www.amazon.com/Intocircuit-Adapter-Charger-Benq-Monitors/dp/B0023Y9EQC/ref=sr_1_1?ie=UTF8&qid=1412357876&sr=8-1&keywords=5a+12v+power+supply
to the run the fan.

http://www.ebay.com/itm/50-W-Watt-High-Power-LED-Driver-AC90V-265V-50-60HZ-Waterproof-/311054139120 to run the COB

Any ideas which of the attached pictures would be the best heat sink? I am thinking the middle one, which was good for a quad core.

Thanks.
 

Attachments

S

Skaumannen

Well-Known Member
If you know what model cpu the coolers were used with you can look up the TDP. But for 50-60 watt I think all of them will do the trick. I am using an old cooler for a Amd 940 cpu, and it runs cool to the touch with 21-22 Celsius ambient temp. Running at 1400ma or 52w.
 
Dloomis514

Dloomis514

Well-Known Member
Skaumannen said:
If you know what model cpu the coolers were used with you can look up the TDP. But for 50-60 watt I think all of them will do the trick. I am using an old cooler for a Amd 940 cpu, and it runs cool to the touch with 21-22 Celsius ambient temp. Running at 1400ma or 52w.
Click to expand...
That cooler was from a Q8400 with 95W of TDP, looks like that one is good to go for one COB 3070 @ 3000k
 
C

CannaBare

Well-Known Member
Do we know if Chromaticity matters in the way we should purchase our LEDs?

I noticed A while ago they were offered in 2-step or 4-step and purchased 4-step Z2 bins. Would we notice a difference from 2 to 4?
 
Abiqua

Abiqua

Well-Known Member
Dloomis514 said:
Gonna pull the trigger on a 3070 @ 3000K, any of the high bin parts (Z2 or above) out there.

I see
http://www.mouser.com/ProductDetail/Cree-Inc/CXA3070-0000-000N00Z230F/?qs=/ha2pyFadugY1sJGAfyUsZl4AheJW3WukO5VbBRSZ5E=
for $40.

Looking at
http://www.amazon.com/Intocircuit-Adapter-Charger-Benq-Monitors/dp/B0023Y9EQC/ref=sr_1_1?ie=UTF8&qid=1412357876&sr=8-1&keywords=5a 12v power supply
to the run the fan.

http://www.ebay.com/itm/50-W-Watt-High-Power-LED-Driver-AC90V-265V-50-60HZ-Waterproof-/311054139120 to run the COB

Any ideas which of the attached pictures would be the best heat sink? I am thinking the middle one, which was good for a quad core.

Thanks.
Click to expand...
You should look at Arrow.com. Believe I saw the Z2's for $33 and in stock. Also saw the Z4 3500k for $36 but don't quote me :) :peace:
 
B

BuddyLeeCU

Member
Can someone post links for drivers for the cxa30xx and 37xx?

Having a hell of a time navigating the digikey and arrow selection and eBay is a mess.
 
You must log in or register to reply here.
Top