DIY COB reflectors

SupraSPL

Well-Known Member
Optic lamp should be on its way :) Hopefully in a few days. I will try to test the effect of the lens using a test similar to the reflector test.

Good link SDS. I have looked through those before (probably on Fasttech) but didnt see anything with a large opening that seemed like it would perform as well as a DIY. Might be worth a shot though.
 

SupraSPL

Well-Known Member
Was looking over the MSDS and realized that the Rustoleum flat white (7790) has 15% titanium dioxide while the high heat version (7751) is only 10%. The professional semi gloss white (239108 ) and standard gloss white (7792) that I used also have 15% though. Maybe the non high heat flat white can outperform them.

Also noticed that chrome (7718 ) I used has 5% aluminum flake.
The "aluminum" version (7715) of that same line of paints has 10% aluminum flake.
Gloss Fleet White (V2196838 ) has 15% titanium dioxide and 5% barium sulphate
Leak seal Aluminum (267972) has 5% aluminum flake 5% barium sulphate
 
Last edited:

Mechmike

Well-Known Member
Was looking over the MSDS and realized that the Rustoleum flat white (7790) has 15% titanium dioxide while the high heat version (7751) is only 10%. The professional semi gloss white (239108 ) and standard gloss white (7792) that I used also has 15% though. But maybe the non high heat flat white can outperform them.

Also noticed that chrome (7718 ) I used has 5% aluminum flake. The "aluminum" version (7715) of that same line of paints has 10% aluminum flake.
Maybe this would work as a top coat over the white on the inside of the reflector cup. http://www.amazon.com/Rust-Oleum-214944-Reflective-10-Ounce-Spray/dp/B000LNVC1M/ref=sr_1_1?ie=UTF8&qid=1418831252&sr=8-1&keywords=reflective+spray+paint
 

SupraSPL

Well-Known Member
I visited HD and Lowes, found the 7715 (10% aluminum) and the 7790 (15% titanium oxide). Also grabbed a can of "mirror effect" 267727 just out of curiosity. It is frigid cold and rainy though, hard to get the paints to dry so will test them tomorrow.
 
Last edited:

FranJan

Well-Known Member
Supra have you taken any temp readings from the COB with and without the reflectors? Reflectors are supposed to increase heat so I think seeing if there is a significant enough increase in heat from the reflectors would be something one would want to know though I guess if you use enough aluminum in the paint it would increase the COB/lamp's innate ability to deal with heat, right :)? I saw you were saying that some COBs you were playing with can maintain their output under higher heat so maybe if there is an increase in heat it would be meaningless anyway but it might be interesting to see what the temps are after 12 hours or so of being on.
Kudos Supra and thanks for sharing all your work. Stay safe bro!
 

SupraSPL

Well-Known Member
FJ, funny you should mention that. I was messing around with a Vero29 tonight, running at 2.23A and I had the idea of watching the Vf to see if it changes when I put the reflector on. I only monitored it for about 30 seconds but it is very sensitive and there was no change in Vf, therefore no change in junction temp. I will test that a lot more thoroughly when I get more time. If there did turn out to be a slightly higher temp, it seems it will be minimal and as you pointed out, would have very little effect on output.
 

Positivity

Well-Known Member
Great stuff supra as usual. Wish i could help more but I'm stumped myself. I always end up at a dead end looking for the exact reflector parts I want.

Seems theres a lot of hesitation stocking the required parts that have been developed. Whether its because of cob tech moving quickly and not having set compatibility standards in place. Who knows...I'm sure sooner or later it wont be hard to get the parts we want at a good price

But..for now...what your doing is one of the best options. Looking for alternatives..
 

SupraSPL

Well-Known Member
So I am back at it. I setup up an 18" X 18" test area on a flat wall to make it easier to quickly see the difference in performance in the center, mid point and outer corner.
In order from best to worst:

Rustoleum flat white "High Heat" 7751 (10% TiO², 5% Talc) (satiny/flat, slightly off white to the eye, easy to get a smooth finish)
Color Place flat white 259865 (5% TiO²) (satiny/flat, slightly off white to the eye)
Rustoleum flat white 7790 (15% TiO², 15% Talc) (very flat, very white, rough texture)
Rustoleum aluminum "Metallic Finish" 7715 (10% Aluminum flake) ( PITA to get a smooth finish)
Rustoleum semi gloss white 239108 (15% TiO², 10% Talc) ( PITA to get a smooth finish)
Rustoleum gloss white 7792830 (15% TiO², 5% BaSO⁴) (visibly white) ( PITA to get a smooth finish)

So although the the 15% flat white is truly flat, like chunks of eggshell. It diffuses the light too much. The flat white heat paint is more like a satin. Same with the Walmart color place flat white, it is satiny smooth. I tracked down the MSDS, turns out it is made by Rustoleum and it has 5% TiO². It also significantly outperforms the 15% TiO² flat white.

So that raises the question, why did the 10% TiO² flat white heat paint outperform the 15%TiO² semi-gloss white? I think it was because the semi gloss has extra resin added.
 

SupraSPL

Well-Known Member
Found one that might be good to try, Rustoleum painters Touch, 2X Ultra Cover satin blossom white, paint + primer (249843). It is 20% TiO², 5% Talc
blossom.jpg
 

SupraSPL

Well-Known Member
Since I had the wall test area set up, I did a quick experiment on reflector shapes. Not surprisingly, the yogurt cup is not the ideal shape (walls are too steep). The reflector on the right is from a 2 liter soda bottle.

Both of these reflectors were painted with 2 or more layers of Rustoleum flat white "High Heat" 7751 (10% TiO², 5% Talc). The one on the right strongly outperformed, everywhere in the 18X18"" test area.
DSC07819a.jpg

As far as the actual numbers, from a distance of 18" from a Vero29, the 2L reflector increased the lux in the center on the 18X18" box from 15400 to 22600 (+147%). Halfway to the edge, increased from 11500 to 15800 (+137%) and at the outside corner from 5700 to 7900 (+138%)
 
Last edited:

az2000

Well-Known Member
What if you used the plastic cone as the skeleton, line it with tin-foil (inside and outside) with enough excess at the base to form large, flat tabs that could be clamped onto the heat sink (or kapton tapped). Then the reflector would contribute a little as surface area to remove heat from the heatsink.

Mash the tinfoil as flat as possible, paint the inside white. The rough surface would be like a pebbled surface giving more diffusion. The outer surface would remain unpainted to act more like a radiator?
 

churchhaze

Well-Known Member
A silver mirror finish is also white, and it reflects more than a flat white paint. If it wasn't white, mirrors would filter certain wavelengths more than others.

Why don't they make mirrors out of polished white paint?

Flat white might reflect "better" if better to you means dispersing the light more evenly.

(and what's so bad about aluminum? looks the best at blocking UV-B!)

 
Last edited:

az2000

Well-Known Member
A silver mirror finish is also white, and it reflects more than a flat white paint. If it wasn't white, mirrors would filter certain wavelengths more than others.

Why don't they make mirrors out of polished white paint?
That sounds logical, but are there any tests proving it? I posted earlier that when I got into growing I read where various surfaces were tested. Flat white beat every surface including polished mirror.

It doesn't sound intuitive. But, I imagine there's a difference between reflecting an exact image versus lumens generally. That there would be a cost involved in the former.

A white wall simply looks brighter to me than a mirror hanging on that wall. So, it seems there's something more happening than the mirror being better at reflecting images (images being light, ergo the mirror must reflect light better).
 

churchhaze

Well-Known Member
If you angle mirror perfectly, you can easily make it look dark, yet if you line up the angles perfectly, you can reflect a bright spot perfectly into someones eye. In that case, what you see is brighter than a white wall. The mirror finish does not disperse the light evenly, but then again, neither does the COB emitting the light. I thought the reason we wanted reflectors was to salvage the negligible portion of the output distribution pattern.

If you reflect the sun into your eyes using a mirror, does it look bright?

A white wall simply looks brighter to me than a mirror hanging on that wall. So, it seems there's something more happening than the mirror being better at reflecting images (images being light, ergo the mirror must reflect light better).
 
Top