DIY COB reflectors

I have been avoiding this issue and focusing on efficiency for starters. We have that pretty well sorted out along with heatsinks, so now I am wondering how we can maximize our yield with reflectors. This is going to become important as we (hopefully) start to see commercial lamps cranking lots of juice through Vero 29s. It is also important for those who are running that CXA3070 at 1.4A or higher.

If you have a CXA COB holder or a Vero, there are really nice aluminum reflectors that you can attach to the holders.

But these can add significant cost to large builds and availability of reflectors for the large COBs seems to be a problem. So for now, I am looking into makeshift reflectors. First thing is first, what kind of paint or surface is the most efficient? I cut holes in a bunch of yogurt cups and painted them with flat white, satin white, gloss white and chrome (Rustoleum brand paints). The angle of the yogurt cups is probably too steep for what we need, but since they are all the exact same size it will allow us to see what surfaces perform best. I will experiment with different shapes later on.



First test, Vero29 driven at 1.5A, cold pulse measurements from 15" away, directly on center

No reflector: 149 (100%)
Plain white plastic: 186 (125%)
Satin white (2 layers): 189 (127%)
Gloss white (2 layers): 189 (127%)
Chrome(2 layers): 178 (119%)
Flat white (1 layer): 201 (135%)
Flat white (2 layers): 206 (138%)
Flat white (3 layers): 206 (138%)
Flat white (2 layers, Walmart brand): 198 (133%)
Aluminum foil sloppy wrap job: 238 (160%)

Next I measure 8" off to the side. It was very interesting to see that a thicker layer of flat white actually improved reflection significantly. The sloppy foil wrap was a real surprise but will have to see how it performs off-angle.

Great job supra as always man

same height but 8" to the side:

No reflector: 79.9 (100%)
Plain white plastic: 101.5 (127%)
Satin white (2 layers): 102.1 (128%)
Gloss white 2 layers): 101.6 (127%)
Chrome(2 layers): 109.5 (137%)
Flat white (3 layers): 106.9 (134%)
Flat white (2 layers, Walmart brand): 102.3 (128%)
Aluminum foil, sloppy wrap job: 93.4 (117%)

Since the Rustolem "high heat" flat white beat the satin, gloss, plain plastic and Walmart flat white on and off angle, we can rule those four out.

The chrome performed surprisingly well off angle, but not enough to make up for its poor performance on angle.

The aluminum wrap results were chaotic, so I made sure to take the measurements with the reflector in the exact same rotational position. It performed very well on angle but poorly off angle. Defintely worth exploring more, but the flat white is more consistent and a lot easier to work with, so for now I will call Rustoleum High Heat flat white the winner (<10% titanium dioxide according to MSDS).


Krylon's version of the white high heat says 9% titanium dioxide.

Good work. Every lumen/watt counts and it was good to see all those comparisons. Nice

In terms of raw numbers there already, that's showing quite a gain from nothing to even *something* before you even take into account the angle of incidence etc... There is a lot more work to do, but as usual you're forging ahead with real numbers and results. Awesome!

When I started growing I read flat white is more reflective than mirror/polished/pebbeled metal. Always wondered why T5HO makers use mirror reflectors, if it would be worthwhile to paint those flat white. Your findings caused me to think about that again.

Also, I wonder if there isn't another efficiency factor: cleanability. Surfaces get dirty. If you can't wipe it clean won't it become the loser over time compared to satin or semi-gloss (which can be wiped)?

Very good point regarding the cleanability of flat white. It is a smooth finish, almost a satin so seems like it would be easy to wipe clean. Only one way to find out, smear it with some sticky earth worm castings I found under the leaves outside.


Rinsed and wiped clean


Tested again, it came out higher. Checked it 4 times, rotating each time and confirmed, consistently higher readings LOL
No reflector: 79.9 (100%)
Fresh paint, flat white: 106.9 (134%)
Soiled/cleaned flat white: 108.2 (135.5%)

Finally, a thread about reflectors!!

Supra, you mentioned that the aluminum foil appeared somewhat sloppy. While it provided a higher number at the direct-angle, it didn't do a whole lot off-angle. Could the latter result be due to how the aluminum was prepared and presented? Also, are you using baking aluminum wrap or legitimate reflective material specifically designed for indoor growing, such as mylar? I ask because I have a bunch of the latter stored in one my closets and wonder if it might worth pulling out for a future experiment.

Also maybe you could explain a little more about the numbers and figures your recording. While I've been absorbing as much information as one can take in during the past months that I've been a member of RIU, I'm not exactly familiar with the numbers you've recorded and there will surely be others out there scratching their skulls in ponderment. I get that a higher percentage is typically a good thing but I feel as if you sort of rushed into this research and sort of just threw up a bunch of data without creating a way for others to recognize or decode the math.

Pretend that most of us, if not all, have never gone as far as taking a calculus or physics class before.

randomlygrow said:

In terms of raw numbers there already, that's showing quite a gain from nothing to even *something* before you even take into account the angle of incidence etc... There is a lot more work to do, but as usual you're forging ahead with real numbers and results. Awesome!

Click to expand...

Yes for sure, huge gains at this distance! (15")

No reflector: 79.9 (100%)
piece of a brown paper egg carton: 90.7 (113.5%)

Id love a 13.5% increase in yield

SupraSPL said:

Only one way to find out, smear it with some sticky earth worm castings I found under the leaves outside.
....
Tested again, it came out higher. Checked it 4 times, rotating each time and confirmed, consistently higher readings LOL

Click to expand...

10 years from now people will be posting: "Dude, you gotta smear worm castings and fish emulsion on your lights. They'll grow better!"

This is gonna be huge, like feeding molasses to influence taste.

EDIT: We should get ahead of this right now. Start selling 3oz bottles of "Pigeon Poo-Wax" for $40, exclusive micro-fiber polishing cloth included free. "Nectar of the Pigeons is a proprietary blend of light enhancing chemicals. Ever wonder how those electric lines retain their like-new shine? It's the nectar! (For the first 200 callers we'll include a sample of Monkey Shine, an $80 value.)"

LOL @ AZ, EWC reflector polishing, I like it

AquariusPanta said:

Finally, a thread about reflectors!!

Supra, you mentioned that the aluminum foil appeared somewhat sloppy. While it provided a higher number at the direct-angle, it didn't do a whole lot off-angle. Could the latter result be due to how the aluminum was prepared and presented? Also, are you using baking aluminum wrap or legitimate reflective material specifically designed for indoor growing, such as mylar? I ask because I have a bunch of the latter stored in one my closets and wonder if it might worth pulling out for a future experiment.

Click to expand...

Yes it was standard kitchen aluminum foil and all crinkled up despite my efforts to keep it smooth (it was hilarious). We are taught not to use aluminum foil for that reason along with the fact that it reflects infrared heat. Mylar is great for flat surfaces although it also reflects infrared heat. With LED infrared is no longer a problem and I have not given up on aluminum just yet, but I suspect that the high heat white may be the ticket for DIY reflectors.

SupraSPL said:

LOL @ AZ, EWC reflector polishing, I like it



Yes it was standard kitchen aluminum foil and all crinkled up despite my efforts to keep it smooth. Traditionally we are taught not to use aluminum foil for that reason along with the fact that it reflects infrared heat. Mylar is great for flat surfaces although it also reflects infrared heat. With LED infrared is no longer a problem and I have not given up on aluminum just yet, but I suspect that the high heat white may be the ticket for DIY reflectors.

Click to expand...

I added another segment to the previous post .

Yeah I like the idea of just spraying an old container of something and using it after it dries as a COB reflector.

One could use a certain brand of reflective aluminum, one for growing purposes, and cut it a certain way and somehow adhere it to the container. The only issue would be heat and the material you would use to adhere, such as glue. Plus it's extra work to just cut out something and glue it and all. I suppose the only way you could persuade someone to go through all that hassle would be to establish some killer results/data showing that it would be worth one's time to go through the hassle (like polishing HS for hours on end, which you actually proved didn't do much).

AquariusPanta said:

maybe you could explain a little more about the numbers and figures your recording. While I've been absorbing as much information as one can take in during the past months that I've been a member of RIU, I'm not exactly familiar with the numbers you've recorded and there will surely be others out there scratching their skulls in ponderment. I get that a higher percentage is typically a good thing but I feel as if you sort of rushed into this research and sort of just threw up a bunch of data without creating a way for others to recognize or decode the math.

Pretend that most of us, if not all, have never gone as far as taking a calculus or physics class before.

Click to expand...

Good question AP. This test is very simple. The Vero is in a fixed position on a stand, not too close to any walls or floors. The lux meter is on a tripod in a fixed position, 15" away, directly in front of the COB. I click the driver on for just an instant to get the reading and click it back off. Then I wait a bit for the junction temp to cool back to ambient, grab a different reflector and get the next measurement. I each reflector very carefully, centered around the Vero. Every once in a while I will check the reading with no reflector, to make sure my baseline has not changed.

Any decent lux meter will have excellent precision and repeatability, but no accuracy whatsoever. If you imagine yourself at a rifle range, that means you could have very tight groups, but not necessarily on the bullseye.



In this case precision is very cheap ($20) but if we wanted to add accuracy to our measurements it would be about $5000 and in this case it would serve no purpose anyway. Another way of saying it, the actual amount of lux measured is most likely incorrect because the meter cannot see all colors equally so it is hopeless to try and translate into lumens or lux. But since we are only looking for relative comparison, this test can be reproduced by anyone with a $20 meter

Long story short, all we need to focus on to evaluate the paint, reflector shapes is the percentage data and the distance/spread.

Great idea, i test it soon

I've wondered about those round reflectors. Wouldn't they trap a lot of heat directly into the COB that would otherwise just flow to the sides? would be interesting to see some tests with and without.

Good point bondoman. I do not suspect that the reflectors would affect the junction temp very significantly, but even if it did, we have seen from the heatsink studies that the output of the COBs is only very slightly affected by changes in junction temp. In other words, it would be well worth it.

Aluminized & white roof paints would be a cool test too, very high reflectivity claims.............they stink though

great info supra........keep it up

I had two thoughts on this subject. Either a 3D printed cone out of white ABS or a scaled version of the sunsystem parabolic reflectors out of 1/16th inch aluminum painted flat white. If only I could find my tin snips.....

Nice project .

You might want to take good care of what kind of :
-plastics/polymers are being used,
-glues /adhesives /sealants used,
&
-paints / varnishes used .

Supra- Did you ever get a sample of the optic LED panel? I've been very curious as to how the lens effects the output.