DIY COB reflectors

stardustsailor

Well-Known Member
Simulation of an DIY COB reflector ....

1)First of all a good photo of the reflector ...rfd.jpg

2) Then ,use the " engauge digitizer " to "digitize " the reflector profile ...
reflector.JPG

3) use the COBs IES file supplied by the manufacturer
vero naked.JPG

and the x,z ( or y ) values from the digitizer on the free (locked )edition of "shape "..
I'm sure you'll figure out how ...

And ....

refxsim1.JPG

refxsim2.JPG

refxsim4.JPG


Not so good those reflectors (the particular ones ..)
refxsim3.JPG

Way high concentration of luminus intensity ..


The shape can export an IES file of the reflector ....
For reference ..Ledil Angelina manufacturers IES (on " IES viewer " freeware )
ledil angelina.JPG

And my DIY reflector ...
ies diy.JPG


Quite weird pattern ...Lots of power at the center ..
:-x...
Better luck ,next time ...
Ain't so easy to make a good reflector ...
 

stardustsailor

Well-Known Member
I've tried a hypothetical conical shaped -reflector ...
Like the ones Supra is using ...
Much better ,I've to admit ...

The reflector "profile " ...
reflectex.JPG

The Shape ,takes that "profile " and "spins" it a full circle..
There's your reflector ...
So this one has these dimensions in 3D :
-32 mm diameter of -base
-70 mm diameter of ' shield lips ' (aperture )
-30 mm height .
Capture.JPG



Place VERO / CXA series cob product IES file ..For light source data
Capture2.JPG
Edit and import
Capture3.JPG


In the next tab ,place manually one -by-one (no copy/paste allowed )Capture5.JPG reflector x,y ( " z" ) values Set : |Cartesian profile |
Capture6.JPG

And hit calculate..
Very easy to learn software ,plenty of (hidden somewhat ) settings ..
refwsh1.JPG

Working space set at 1m height ..
refwsh3.JPG

Follow these steps to extract an new IES file of the reflector ..
Capture4.JPG

Download freeware "IES viewer " and view the render ...
ies exp.JPG
 
Last edited:

nogod_

Well-Known Member
All greek to me...
bongsmilie

I've tried a hypothetical conical shaped -reflector ...
Like the ones Supra is using ...
Much better ,I've to admit ...

The reflector "profile " ...
View attachment 3316746

The Shape ,takes that "profile " and "spins" it a full circle..
There's your reflector ...
So this one has these dimensions in 3D :
-32 mm diameter of -base
-70 mm diameter of ' shield lips ' (aperture )
-30 mm height .
View attachment 3316751



Place VERO / CXA series cob product IES file ..For light source data
View attachment 3316752
Edit and import
View attachment 3316753


In the next tab ,place manually one -by-one (no copy/paste allowed )View attachment 3316766 reflector x,y ( " z" ) values Set : |Cartesian profile |
View attachment 3316768

And hit calculate..
Very easy to learn software ,plenty of (hidden somewhat ) settings ..
View attachment 3316747

Working space set at 1m height ..
View attachment 3316749

Follow these steps to extract an new IES file of the reflector ..
View attachment 3316757

Download freeware "IES viewer " and view the render ...
View attachment 3316758
 

Positivity

Well-Known Member
Theres always the reflector adapter from ledil for 3070s and maybe various similar/smaller types too. Less than a buck and it tightens the beam up a little. You could probably mount without the ideal holder if you wanted to...it has its own screw mounts.

Leaks a little light but i like it. Even when I'm not using my reflectors i like having them there for a little extra focus.

Cheap option..might even be able to silicone it on at the screw spots

image.jpgimage.jpg
 
Last edited:

nogod_

Well-Known Member
I have a question regarding penetration:

I have seen countless times that folks use reflectors and lenses to increase penetration. This doesnt really make sense to me. In order to reach deeper into the canopy dont you want the photons hitting your canopy at a variety of angles rather than straight down ie. perpendicular?

In other words, doesnt adding more emitters not only improve the evenness of your coverage but also your penetration because you have a better chance at hitting the gaps between the topmost leaves?
 

Positivity

Well-Known Member
Lots of ways to look at it..

Ideally you'd have them spaced perfectly over the canopy at just the right height.

In a tent the light gets reflected mostly so its not really needed.

But if someone wanted to run a simple setup of one light hid style, in an open area, the reflectors would give a nice and tight footprint. probably more suitable for tree style growing.
 

stardustsailor

Well-Known Member
In a trial not overpopulate the LED section with new threads for the same topic,I think Supra will allow me to keep posting long posts about reflectors ...
Her's some further thoughts and research from a dedicated DIY designer (yes,that's me,myself and I ,all together ... :P..)
Huge "EGO" issues,I'm trying to get over with ...
Anyway ....

Reflector basic thoughts ....
SHIELDING ANGLE.JPG

Now,that it's been understood what the "shielding angle " of a reflector is ...
Let's move from a bulb to a COB ...

reflector b.jpg

Say the COB is a VERO29 with a LES diameter of 29.2 mm .

On top there's a cylinder- shaped reflector ..
(Cylindrical it's the name ? )

Dimensions of the cylindrical reflector are :
Height: ~ 38 smthng mm
Diameter (inner ) 49.8 mm ,outer diameter is ~ "nobody gives a $hit right now " ..

Now ..
The Vero29 has a " Full Width Half Maximum " angle of 120 ° ...


So ,80-90% of radiant power is " within" a cone of emitted light ,with a (full ) cone angle of 120°.
The rest 10-20 % of radiant power is emitted from 120° to 180° ,360° around the light source (the COB in our case ) .

So ,for which reasons reflectors could or should be used in a grow site ,using COB solid state light ?
Mainly for three reasons ...

1) To 'restrict " the light within certain limits and / or "projected" shape / pattern .

2) To aid homogenize the light power-wise.
Make power distribution of incident light more even.
Via light diffusion ,of course .

3) Make use of that 50% of light power ,otherwise irradiated from 120° up to 180°.


Info bites :
An useful calculator :
http://store.marinebeam.com/beam-angle.html

Field angle : The angle of a light cone where the irradiation / illuminace is 10% of maximum .
So...Back in the example ....
From the COB -reflector unit ,direct light exits the combo,at a cone with 78.5°.
( ! COB not speculated to be dimension-less "point source" ,as for measuring the FWHM angle ! )

Total Shielding angle of reflector is 92.5°.
FWHM angle - Slielding angle = 27.5°
/2 = 13.75° of light from the Half width half max ( 60°) is being reflected ,at each "side" ..

So ,some % of light within the 80-90% power of FWHM cone ,will be reflected ,at the upper part of the reflectors
sides or "profiles" ...(red colored part )

The rest light ,from the part " 10-20 % @ 120°-180° " of COBs irradiated cone ,
will be reflected from the lower part of the profile ,towards the base and from the base itself ...

Those are some basic thoughts ...

Now ,some more complex ,to follow .......
 
Last edited:

stardustsailor

Well-Known Member
So ...
Once again ....
"Why I should use a reflector with my COBs ? "

1) To 'restrict " the light within certain limits and / or "projected" shape / pattern .
Easily ,becomes understood ,that the shielding angle of the reflector plays a crucial role here,
as also the reflector overall shape itself .
A side note : It will be wise the reflector to reflect as less as possible of the
FWHM light cone .A good starting point is 80-90° total cone angle for circular light patterns .

2) To aid homogenize the light power-wise.
Make power distribution of incident light more even.
Via light diffusion ,of course .

Here ,things become complex .....White paint or aluminium/metallised polymer ?
Later on,more thoughts about the subject ..


3) Make use of that rest 10-20 % of light power ,otherwise irradiated from 120° up to 180°.
The lower part (towards the light source ) of the reflector it is the case ...

Info bit :
http://ledil.fi/sites/default/files/Documents/Technical/Articles/Article_1.pdf
 
Last edited:

alesh

Well-Known Member
...
The Vero29 has a " Full Width Half Maximum " angle of 120 °

So ,50% of radiant power is " within" a cone of emitted light ,with a (full ) cone angle of 120°.
The rest 50% of radiant power is emitted from 120° to 180° ,360° around the light source (the COB in our case ) .
...
Interesting post, but this statement is not true. It's 50% of maximum value, not 50% of (total) radiant power.
For Vero29, about 80-90% of radiant power is within 120° cone.
 

stardustsailor

Well-Known Member
1) To 'restrict " the light within certain limits and / or "projected" shape / pattern .

&

3) Make use of that 50% of light power ,otherwise irradiated from 120° up to 180°

So ,it's the physical shape and dimensions ,playing probably the most crucial role here ....

Shape,it's a reflector designing software ,but amongst others it lacks plenty of features ..
One of these -a great one- is the type of reflection ..
Specular or diffuse ?
Shape ,only calculates with specular reflection routines and only accounts for the reflectivity ,
which can be adjusted. (% of light reflected vs absorbed by reflector materials )

So ..The cylindrical reflector simulated at Shape ..
Shape only simulates the reflected by the reflector rays .It does not show the direct -"unreflected" - light rays .
Simulated is "the profile " of the reflector ..Say that your seeing an X-ray of a reflectors half.
Red dot is the light source .


cylinder1.JPG

Extreme reflections of light ,both under and over the Half ...(ok you can have Full if the whole reflector was shown ) Width of Half Maximum power angle ,towards the ' opposite ' side ..

Whgere the other part of the reflector is located ,to reflect some of the rays ...
As whole ....
cylinder2.JPG
Actually a cylinder-shaped reflector turns out to be a very bad idea ...
It actually spreads the light ,rather than "restricting " it ....

Most Cobs have a FWHM angle of 120°
That translates ,that if one of them is hanged 'bare' ,at 1 meter height from ground ,
the circle of 80-90% of total power ,(the base of the 120° cone ) has a diameter of 3.46 meters ...
calc.JPG



That is way much ...Hanging a COB at one meter ,one should/would expoect a growing area of 1 square meter ..
1 x1 m ..
A bare COB will have much of it's light to be reflected by side-walls of site ,if hanged bare ,at 1 m height ...

If FWHM was 80° ?

Then the 80-90% power circle would had a diameter of 1.68 m ..
Much-much-much better ...

So ..

Regarding purchasing small circular aperture reflectors,those the ready made kind, from various brands,
a COB LED grower should choose within the 70°-90° range ,for most of cases .
 

stardustsailor

Well-Known Member
So ,still in the "shape " issue ...
A bowl shaped reflector .. ( a weird synthesis between cylinder and hemisphere,of a profile )

bowl1.JPG

bowl2.JPG

bowl3.JPG

Much light ray concentration points ..
That ,apart from creating "glare " reflections ,
also may cause the air to heat up ,inside the reflector cavity ,
as the air is relatively stagnant at that area ....


A cone shaped reflector ...
cone1.JPG

cone2.JPG
 

stardustsailor

Well-Known Member
So ..One way of using reflectors with COBs is the small individual ,
cup-shaped reflectors with cylindrical aperture .

-Those reflectors with a FWHM angle rated at 70°-90° ,seem to be the best for most cases .

-From possible shapes ,the conical one seems to be much superior,regarding specular reflected light distribution,
than cylindrical or bowl-shaped ..No glares ,no light concentrations ,no possible reflector cavity hot -spot issues ...
 

stardustsailor

Well-Known Member
Which leaves the " 2) To aid homogenize the light power-wise.
Make power distribution of incident light more even.
Via light diffusion ,of course .
" part for last ....

It has been proven that most plants benefit from diffused evenly distributed light ,both
power wise and spectrum wise ...

http://www.ncbi.nlm.nih.gov/pubmed/18028265
http://aob.oxfordjournals.org/content/early/2014/04/29/aob.mcu071
http://www.actahort.org/books/957/957_4.htm

Regarding the power -wise part ...

An attribute of a COB reflector should be the diffusion of light ...
http://en.wikipedia.org/wiki/Diffuse_reflection

With as less "cost" regarding light absorption ,possible .

Two ways-options open up here ...
mat.JPG


White Paint Reflective Surface.
A flat (matte) /satin finished white layer is almost 100% diffusive regarding quality of reflection ..
But ,it presents ,some losses ,reflection -quantity -wise , due to light absorption from paint material.
Those losses can be as high as 20% ,in some cases ,
though at average range somewhere between 5-10 % .

Notes :
-Myself have measured temperatures reaching up to 90° ,10 cm in front of a CXA3070 LES (@2100 mA ).
Any possible paint used on DIY or commercial reflectors ,better be the "high temperature grade " kind .
It should be rated " 150° temperature " ,at least.

-Possible VOC outgassing from paint .

Aluminium Reflective Surface.
Shape software simulates the reflections ,as if the reflector has a reflective surface ,
resembling the one flat mirror-polished aluminium has ...
While the light losses from absorption is usually at the 2-5% range ,the diffusion is rather poor ...
A flat mirror polished aluminium surface is almost 100% specular reflective ..

Still,this type of surface seems to be the most favourable option of LED dedicated reflector manufacturers .
But ...
They seldom ,if not at all ,utilise flat mirror-polished aluminium /metalised poilymer surfaces ..
Either it is satin finished (brushed & anodised -transparent protective layer )
or it has some sort of "patterns" ,in order to diffuse the light .
(Honey-comb ,diamond,hammertone,etc .
More kinda like a "digital" version of the "analog" white paint diffusion....
By adding "texture" to flat mirror -finish .Multiple tiny mirrors ..)

More sophisticated solutions ,of that type , include more complex reflective / refractive surface designs , based on Fresnel phenomena ,where multiple "micro-scales" at the reflecting surface ,create a very even light distribution ,as light rays are "guided" with high precision ,power-wise .
(Utilising a design like that a reflector or lens can "disperse " concentrated light rays -say at the middle point of a LES ,for a lens -or on the contrary "concentrate " diffused light -at the edges of the same lens )
.A wide range of light power distribution patterns and shapes ,can be achieved with this way ,
while the reflector remains compact enough .
As a DIY option ,I think that is beyond reach ,to back-engineer ,copy,design and finally build .
fresnel vs spherical.jpg
 
Last edited:

stardustsailor

Well-Known Member
So,proceeding with DIY reflectors project for COB those 'guides' should be kept in mind :

-Shielding angle should be around the 70°-90° range .
-Overall shape should be as conical as possible .
-If white paint is utilised ,it should be of "High Temperature" grade ,flat or satin finish .
 
Top