DiY LED - Cree CXA3070
- Thread starter SupraSPL
- Start date
Doer
Well-Known Member
For my London mates, remember. Mind the Gap. 
TIM = Thermal Interface Material
http://techtransfer.universityofcalifornia.edu/NCD/22858.html
TIM = Thermal Interface Material
http://techtransfer.universityofcalifornia.edu/NCD/22858.html
beppe75
Well-Known Member
I was thinking to use 2 alluminium L profiles (front of each -other) and three straight alluminium panels to carry the cobs 2 on the outside panels and one in the middle... smth similar to bizfactory's frame ( hope you are ok biz with me sharing this.)
thanks for the reply Dloomis
bizfactory
Well-Known Member
no probs! means I did an okay job!
beppe75
Well-Known Member
yes indeed, it actually inspired my project along with some other great ideas in this forum.
wirat
Well-Known Member
Yup, what I mean is can you use the Kapton tape to hold the COB on just as if it were masking tape. No screws no drilling, etc. I have been trolling around and I see lots of photos posted where it appears that is all that's holding them on. (that and whatever adhesion the heatsink paste offers). I heard some of the heatsink paste becomes like a glue.While we wait for his answers, here is everything about Kapton Tape. I have been thinking about it, too.
You get some elongation as the temp. rises but not much.
http://www.dupont.com/content/dam/assets/products-and-services/membranes-films/assets/DEC-Kapton-summary-of-properties.pdf
Some of the properties such as tensile strength will degrade to zero over time, in heat. It doesn't like UV exposure, or water.
But, I am not sure what you mean. Just use it like masking tape? Or is it double sided between the COB and the sink? The problem is not attachment, you know?
The problem is to create a thermally efficient bond between the junction and the heat sink. Not an easy problem. To see the problem you need imagination as it happens on the microscopic level.
Supra when through enormous work to polish his heat sink. Why? He made the surface as gap free smooth as possible. It shines, it reflects, it looks like 4 wires when there are only 2.
I saw a prototype of our latest processor yesterday. It is polished as fine as a mirror in a telescope. Why? Same thing. It is what the polished side the heat sink clamps to.
So, heat can only transfer directly molecule to molecule. If a molecule of aluminum is in DIRECT contact with molecule of COB bottom surface, heat will flow easiest. But, if there is the most sub-microscopic gap in smoothness there, it is an AIR GAP. <GASP>
This is why use expensive silver based gap filling compounds, because the gaps we are talking about are not visible to the eye.
So, can Kapton apply enough pressure to keep the gap filler happy? That's the main question.
That's why we use screws, polish work, and gap filler.
Doer
Well-Known Member
I am reading in the design guide you are not suppose to use screws, EVER.
http://www.google.com/url?sa=t&rct=j&q=&esrc=s&source=web&cd=1&ved=0CC0QFjAA&url=http://www.cree.com/xlamp_app_notes/cxa_design_guide&ei=9qMNVarCO4KfoQTnhYGIBQ&usg=AFQjCNE6JRMsvNiA8LZTq3oiS7sYrEC7jw&sig2=_wMNMFnGUD232W9OyNYECg&bvm=bv.88528373,d.cGU
wirat
Well-Known Member
Well, gald you mentioned that, I did just see Luxeon Thermally Conductive Adhesive Tape for High Power LEDs, but I can't recall seeing anyone use it here, so I thought it must not be good. There is a Youtube of the guy applying it and it looks as easy as sticking one side on the heatsink and then pressing the LED on. It is double-sided tape that attaches the LED as well as acts as the thermal bond.
http://www.luxeonstar.com/thermally-conductive-adhesive-tape-for-high-power-leds
Last edited:
Doer
Well-Known Member
Seriously, that is no way to look at DIY as far as I am concerned. There is no way for anyone to be completely knowledgeable with all the new tech coming on line.
Doer
Well-Known Member
Using latex gloves for those greasy fingers is highly recommended.
Dloomis514
Well-Known Member
I used Arctic Silver ADHESIVE, the stuff stick like glue, worked great. Also used thermal paste with kapton tape, but the COB feel off. Was glad I used adhesive.
Dloomis514
Well-Known Member
Oh and btw, Supra doesn't polish anymore, no benefit for the effort expended was his conclusion.
Doer
Well-Known Member
See? I think the Kapton will not last in the Grow room environment.
Doer
Well-Known Member
Well, there is. But, is the Cree polished like a mirror on the back? You need both these surfaces incredibly smooth to make a mating surface with very tight tolerance.
I read once that is the Hubble telescope mirror is so smooth that if it was the size of the USA the highest hill would be less that 1 mm.
You have to grind, let's say. Extrusions are extremely, not smooth. Flat? Yes.
What I will do is use a belt sander starting at 120 grit, lightly just to knock down the tops. (men, I have done some much fine auto and airplane painting and sanding it is in my blood.)
The trick is to go lightly and understand what you are trying to do.
So, for 120 grit I will stop before the lows get too scratched. Very light, a couple of passes. Then switch to 220, then 360. You don't. need to polish like that. I do remember Supra said he pulled a Doer at first and went over board with the polishing.
beppe75
Well-Known Member
Anyone has some feedback for my build... i want to order the parts but i need some (good) feedback from you guys to see if i've choosen the right path... so i'm making my questions once again...
my project consists on :
5xCXB3070-0000-000N0HAB30G-ND run by a single HLG-185H-C1050B
5x Arctic alpine 64 GT Rev.2, 80mm max cooling power 70w
----i also do not know if this switches: ( http://www.amazon.it/dp/B00EJ64MR8/ref=wl_it_dp_o_pd_nS_ttl?_encoding=UTF8&colid=1VAMHG6WXZDWD&coliid=ICII2EMKPCZVP)
are ok for my build in terms ov current and other details???
----what do you think will be the rendiment of this lamp in a 36''x36''x72'' ds secret jardin tent...???
----what do you think in terms of efficiency?
----hould i go with the project or should i wait for better cobs ???? i think this are ok though?
i plan to add some 2 more cobs in the near future or some deep blue immediately if you reccomend that...
so to have a better w/sft ratio than the 21.1 i get with the current setup...
Any help appreciated... i don't want to miss the remaining cobs or drivers... i've noticed they end up pretty quickly.
Thank u guys again...
Rock' n' Roll ya all.
my project consists on :
5xCXB3070-0000-000N0HAB30G-ND run by a single HLG-185H-C1050B
5x Arctic alpine 64 GT Rev.2, 80mm max cooling power 70w
----i also do not know if this switches: ( http://www.amazon.it/dp/B00EJ64MR8/ref=wl_it_dp_o_pd_nS_ttl?_encoding=UTF8&colid=1VAMHG6WXZDWD&coliid=ICII2EMKPCZVP)
are ok for my build in terms ov current and other details???
----what do you think will be the rendiment of this lamp in a 36''x36''x72'' ds secret jardin tent...???
----what do you think in terms of efficiency?
----hould i go with the project or should i wait for better cobs ???? i think this are ok though?
i plan to add some 2 more cobs in the near future or some deep blue immediately if you reccomend that...
so to have a better w/sft ratio than the 21.1 i get with the current setup...
Any help appreciated... i don't want to miss the remaining cobs or drivers... i've noticed they end up pretty quickly.
Thank u guys again...
Rock' n' Roll ya all.
Doer
Well-Known Member
more from the design guide about polishing a heat sink
Cree recommends that the heat sink used with a CX family LED has an average roughness value (Ra) less than 10 μm.
Figure 17 shows typical roughness values resulting from various manufacturing processes.2 Once a heat sink is manufactured, finishing the heat sink by polishing or milling, for example, is important to achieve a smooth, flat surface. For comparison purposes, Table 4 contains size measurements for several grit sizes in several standards systems.
(an extrusion will slump in the middle as it cools)
The quick way to check the flatness of a heat sink is to use a razorblade as a straight edge and touch the edge to the heat sink. Look for any gaps between the razorblade edge and heat sink. Figure 12 shows the procedure.
Figure 12: Checking heat sink flatness, left: a gap below the razorblade edge, right: no gap below the razorblade edge.
Figure 15 shows the relative lumen output from a CXA2540 LeD connected to the same Cree heat sink and operated at various power levels with various TIms. The lumen output percentages are calculated relative to the output with thermal grease as the TIm. Note that a thermal pad allows the LED to operate at nearly the same lumen level as does thermal grease, but with the other TIms the lumen level is initially lower than with thermal grease and decreases as the power increases.
-----------------------
So, for me, obviously at 50w anything but those top two TIMs result in 5-10% illumination drop. So, this is really the most important thing, next to keeping the amps lower.
Cree recommends that the heat sink used with a CX family LED has an average roughness value (Ra) less than 10 μm.
Figure 17 shows typical roughness values resulting from various manufacturing processes.2 Once a heat sink is manufactured, finishing the heat sink by polishing or milling, for example, is important to achieve a smooth, flat surface. For comparison purposes, Table 4 contains size measurements for several grit sizes in several standards systems.
(an extrusion will slump in the middle as it cools)
The quick way to check the flatness of a heat sink is to use a razorblade as a straight edge and touch the edge to the heat sink. Look for any gaps between the razorblade edge and heat sink. Figure 12 shows the procedure.
Figure 12: Checking heat sink flatness, left: a gap below the razorblade edge, right: no gap below the razorblade edge.
Figure 15 shows the relative lumen output from a CXA2540 LeD connected to the same Cree heat sink and operated at various power levels with various TIms. The lumen output percentages are calculated relative to the output with thermal grease as the TIm. Note that a thermal pad allows the LED to operate at nearly the same lumen level as does thermal grease, but with the other TIms the lumen level is initially lower than with thermal grease and decreases as the power increases.
-----------------------
So, for me, obviously at 50w anything but those top two TIMs result in 5-10% illumination drop. So, this is really the most important thing, next to keeping the amps lower.
Doer
Well-Known Member
The thing is and what Supra is saying I think, when do you know you are at 10um? Any abrasive at all can take you way beyond.
In fact it says, polishing begins at .8 μm.
I found something that makes sense here. Compare to belt sander grits.
This show that 80 grit on a belt sander is already below 10 Ra. So, as @SupraSPL said, no need for more than that. Duly Noted.
In fact it says, polishing begins at .8 μm.
I found something that makes sense here. Compare to belt sander grits.
This show that 80 grit on a belt sander is already below 10 Ra. So, as @SupraSPL said, no need for more than that. Duly Noted.
Abiqua
Well-Known Member
I would bet low thermal impedance, like the the rest of the "layer" types....I have looked at a few....even bought some from Digikey in the past, the 3m mounting tape....but all have shitty impedance compared to paste, which sucks, because applying paste manually imho, is one of the variables that needs to be eliminated....
