Optic Lighting Vero COBs updated

Right without the lense in place there is a lot of light emitted at high lateral angles, about 170 degrees total. That lateral light would not make it to the canopy unless the COB were closer, but at 16" much of it is lost. Even with a reflective wall, the wall reflects in all directions so much of it is lost that way as well. So the lense is grabbing most of that light and redirecting into the "canopy" in a 90 degree beam. From what I can tell, about 2% of the light is lost in the lens itdself, 4% is reflected laterally off the backside of the lens and another 4% is reflected from the other side of the lens (not sure how that works) but it looks like we lose about 10% of the light, which is by far the lesser of two evils at this distance.

Lateral view of the emission pattern of a bare Vero 29


Apparently if the lense was made of low iron glass and anti reflective coated on both sides, we would only lose 1% instead of 10%, but the AR coating is not robust. I wonder if it could hold up under such an intense pounding of light 12 hours a day?

Well,for one more time I will adopt the role of devil's advocate ..

None ,of the available Chip-On-Board L.E.D. arrays is actually designed to bear such a type of lens..
I won't be talking about light losses -mainly because of the glass thickness***- ,but in this case what matters is
the service life of the diodes of the L.E.D. C.O.B. array and phosphor degradation ( Color temperature shifts to higher Kelvin values i.e from 3000K slowly degrading to 3500K and then to 4000K and so on ...)

Most of these type of lenses will have some openings left for outgassing V.O.C.s,
so VoCs also are not a serious issue.

These type of thick glass lenses ,will cause a tremendous increasement of heat inside the phospor doped ,silicone light emitting surface .That phenomenon will eventually degrade phosphor particles and the COB will gradually shift into emitting cooler light temperature (=towards blue ) .

Due to elevated heat tempeprature ,the diodes are "suffering" themselves also..
Some might even get fried .
(Ask Psuagro,he knows-he has experienced that kind of shit ).
Of course all these won't happen the very first day ,week or month ..
It takes some time -of operation- for these things to start showing ...

These lenses are very -very bad idea ,at least regarding my point of view.
Very bad idea.
COBs should have a minimum of 50 mm "empty space " / "air gap" ,before anything is installed in front of their LES.

The right approach into forming a "tight beam" when utilising COBs ,is the one used by Ledil and other renown optic brands .And that is the use of a reflector along with a -thin- glass/PMMA lens ,covering the reflector .That is the-only- suggested way .

Cheers.




***
The words "opacity" and "opaque" are often used as colloquial terms for objects or media with the properties described above. However, there is also a specific, quantitative definition of "opacity", used in astronomy, plasma physics, and other fields, given here.

In this use, "opacity" is another term for the mass attenuation coefficient (or, depending on context, mass absorption coefficient, the difference is described here) at a particular frequency of electromagnetic radiation.

More specifically, if a beam of light with frequency travels through a medium with opacity and mass density , both constant, then the intensity will be reduced with distance x according to the formula


where

• x is the distance the light has traveled through the medium
• 
  is the intensity of light remaining at distance x
• 
  is the initial intensity of light, at
  

For a given medium at a given frequency, the opacity has a numerical value that may range between 0 and infinity, with units of length2/mass.

http://en.wikipedia.org/wiki/Opacity_(optics)

I ordered 2 to try out..lol..

What i'm expecting is the Tc measurement point will show much higher. I'll see if i can take a few measurements with and without. The temp sensors are already there so it should be a easy project

Might be able to test it through a range of currents..

Kinda moot though as i think suitable reflectors will be available eventually

Positivity said:

I ordered 2 to try out..lol..

What i'm expecting is the Tc measurement point will show much higher. I'll see if i can take a few measurements with and without. The temp sensors are already there so it should be a easy project

Might be able to test it through a range of currents..

Kinda moot though as i think suitable reflectors will be available eventually

Click to expand...

From "CXA design guide" :
(...)
CXA Lens Material Considerations
Polymers, i.e., plastics and polymethyl methacrylate (PMMA), and glass are the most common materials used for optical
lenses. Although glass typically has better optical properties than plastic, glass is used less frequently because it is
heavier, more expensive and more fragile than plastic. The light absorption, reflection and transmission properties of
plastics can vary considerably, even within the same class of material, e.g., polycarbonate. Cree recommends the use
of optical grade plastics for lenses used with CXA LEDs to ensure good optical efficiency and long‑term reliability. The
use of non-optical grade plastics is to be avoided. This includes materials used as a luminous opening, i.e., a window, in
a luminaire. The flammability rating of a polymer material should also be taken into consideration when designing and
specifying optical components. UL 94 can be helpful in providing guidance.18
CXA LEDs transmit no significant infrared (IR) light, but, as do all high‑powered light sources, do transmit significant
photonic energy that, if absorbed by the lens material, can cause the material to heat up. The focusing effect of the lens material can cause the lens to reach a temperature higher than the Tc of the CXA LED producing the light.
(...)



-But the lens is made out of glass ,so what's to worry about ?
-Yup ,but the lens is also very-very close to the LES of the COB .They are also of a certain thickness and mass...of glass . Do the rest of "math" by yourselves...

..Heat from the diodes underneath ,heat from the glass lens on top
and the poor LES in between ...
A big fat heatburger ....

Cheers.

Can it be fixed [ the lense problem] and can the 50mm gap to the LES be overcome with active cooling....the benefits of light reflection seem to be there, rather simply too, although there is an obvious drawback...becoming mucho caliente....
..So is it another thrown in the drawer concept or one worth refining as an option? that part seems unclear.....

will the focusing affect always trump what we are trying to accomplish, making it moot? regardless of densities, materials, angles that are chosen?

@stardustsailor , I was definitely concerned about that possibility so I took the time to check temp droop with and without the lens attached. With lens = 2.1% without lens = 2%

The blow through style of heatsink in this case actually blows a nice stream of air blowing past the LES and lens. It would be interesting to see how temp droop would be affected if I blocked that airflow.

Also worth mentioning, the lens seems to strongly outperform my best reflectors. The lens shifts the color temp ~500K warmer, so that is not a direct comparison for my luxmeter, but the increase in performance is stronger than I would expect from a spectral shift. I think the problem with the reflector is that it is scattering some of the light rather than guiding it to the target?

I still use reflectors where appropriate, they are a LOT better than nothing.

Lifespan and efficiency loss aside, wouldn't such tight spots actually make it harder to evenly cover a canopy? You could make a well defined venn diagram with 2 of those vero 29s + lenses. It's cool, but is it really beneficial?

Yes you can see from the pic and from the lux measurements that because of overlap we will not get even coverage with this setup at 16". And 16" may be too close I havent tested that yet, just an intensity estimation (28W/ft²). I noticed as I move the lamp further from the target, the circles expanded and combined until they matched up into 1 circle. So what it comes down to is using the right lens angle for your particular distance to canopy and distance between COBs. Maybe a 120° lens would be a better fit for the Vero18 setup. Sounds ridiculous since the Vero is 120° to begin with, but now I realize it is more like a 170 in practice.

So for the 2.1A Vero29 version, each COB will be dissipating 80W and I believe is using the same 90° lens, so the distance to canopy will be larger and maybe we would get slightly more even coverage.

That said, it is a good problem to be working on. Better that the photons are there causing un-eveness rather than ending up with a much lower intensity everywhere in the canopy. Reminds me of the saying "you cant go broke paying taxes", (not that I am endorsing taxation).

110-120° is the " Full Width at Half Maximum " emission angle of most COBS ,without any additional optics ...
That means that most of the photons emitted are within the 120° projected light cone .

Of course there is some light emitted in the angle range over 60° ,from center .
The total amount of photons emitted in the 60-90° region -from center-equals less than 10% altogether.
( Some of those "60-90° region " photons ,via reflection onto a "reflective wall " do reach the leaf canopy .)

Utilising a thick glass lens ,yes it will "concentrate " all/most of the emitted photons ,into a well defined "beam" .
Still the light losses are about 10-15% -if not more( ...And those photons never will be reaching the leaf canopy .. )

So ....Using that kind of lenses is like ...
Making a great round hole in the surface of water ...
..
Nothing to gain from ...Only lose light energy ,in favor of "shape" ...
Plus risking the phosphor & diodes performance and service life.

Maybe in the near future ,a 90° (not a FWHM angle in this case ) reflector( uncovered ,unlensed ) ,
would be a nice addition ,when these reflectors will be massively available and easy to find .
Until then ...
Personally , I prefer my COBS unlensed .
My last "one cent" on the issue.

Cheers.

SDS, I agree completely with your logic and I thought the same thing until I saw the numbers on my luxmeter. I encourage you to grab a cheap $3 lens and luxmeter and I think you will see what I mean. It puts a lot more of the light where we want it and because your lamp style has a long distance to canopy, it will benefit most of all (although I am sure it will work awesome either way )

stardustsailor said:

110-120° is the " Full Width at Half Maximum " emission angle of most COBS ,without any additional optics ...
That means that most of the photons emitted are within the 120° projected light cone .

Of course there is some light emitted in the angle range over 60° ,from center .
The total amount of photons emitted in the 60-90° region -from center-equals less than 10% altogether.
( Some of those "60-90° region " photons ,via reflection onto a "reflective wall " do reach the leaf canopy .)

Utilising a thick glass lens ,yes it will "concentrate " all/most of the emitted photons ,into a well defined "beam" .
Still the light losses are about 10-15% -if not more( ...And those photons never will be reaching the leaf canopy .. )

So ....Using that kind of lenses is like ...
Making a great round hole in the surface of water ...
..
Nothing to gain from ...Only lose light energy ,in favor of "shape" ...
Plus risking the phosphor & diodes performance and service life.

Maybe in the near future ,a 90° (not a FWHM angle in this case ) reflector( uncovered ,unlensed ) ,
would be a nice addition ,when these reflectors will be massively available and easy to find .
Until then ...
Personally , I prefer my COBS unlensed .
My last "one cent" on the issue.

Cheers.

Click to expand...

Just like you say. Here's my calculation.

I just did a quick measurement with a bare Vero 29. Directly in front of the COB I get 14000 lux. I used a ruler to maintain the same distance and moved the sensor 45 degrees off center, the reading was 7000 lux. I moved it 80-85 degrees off center and the reading was 1400 lux. Not sure if that helps but there is quite a lot of measureable light outside the 120 degree cone.

SupraSPL said:

I just did a quick measurement with a bare Vero 29. Directly in front of the COB I get 14000 lux. I used a ruler to maintain the same distance and moved the sensor 45 degrees off center, the reading was 7000 lux. I moved it 80-85 degrees off center and the reading was 1400 lux. Not sure if that helps but there is quite a lot of measureable light outside the 120 degree cone.

Click to expand...

Yes, very helpful. It matches the data from the spec sheet pretty closely.
It seems like there's a lot of light outside the 120° cone but the opposite is actually true -- it's just about 13%.
edit: I can hardly imagine that 120° lens losses would be (much) smaller

the guy who does all the measurements for Amare. I directed him to this thread. He said in a few days. He will start testing with and without lenses. Using a db meter (from what I said about sound levels) and an infrared pyrometer. They're testing with cxa 2025's though.

I would still stick with reflectors over secondary lenses until blux/cree cobs get a more focused primary angle option? seriously, what's the hold up on this? They are designed for retrofitting recessed lighting, don't get the 120+ degree used.

I agree with SDS^^.....seen it myself with the innoled fixture, yes being cooled PASSIVELY increased the failure rate drastically

what about a short reflector to "catch & bounce" the stray side output, leaving most of the light unobstructed.......don't like covering any led, temp droop differences aside i think what's actually occurring under the secondary lens can't be an optimal environment IMO..

This is very easy to check because temp droop is a direct reflection of junction temp. You can monitor the output and the Vf to see if adding the lens has any effect. Let the lamp run until temp stable then add the lens and watch what happens to the lux output and Vf. So if there is no significant temp droop or Vf change, there is no significant change in junction temp or reliability. We run our COBs a lot cooler than they were designed to run, I expect a near zero failure rate on CXAs and Veros in our diy grow lamps. I ran a Vero 10 at 2X the maximum current and it was no problem, these things are built tough (except when it comes to overvoltage).

Generic COBs on the other hand seem to be hit or miss. So far out of 3 100W COBs I have received, 2 were damaged to the point of uselessness. One of them was damaged in a way that would be hard to notice unless you were checking the output because it lit up but put out only half as much light as the other. One of the 20W COBs died within seconds when I ran it at 40W.

I'll be a guinea pig for trying a couple lenses long term for starters, I'm praying for the best.

This topic is old news for the benefits of lenes vs reflectors.
@earserhead used to have a great write up in the a51 blogs in their old site. Can you post that if you have it in your archives?

I was amazed with the results of bare cobs. It l blew the doors off most of what I knew about penetration and power. Showing me that PPF seems to be of the most importance, assuming no ridiculous light spill. Not PPFD and distance from the source.
I also think that the bare cobs are closer to a disused light...specially when using many cons for spread. Lenes counter act that a little.
Like SDS and alesh said, near all the photon will reach the canopy even bare. But a nice ~90* reflector will get the few missing down to the canopy.
I am not on the lens train. But would love to see what you guinea pigs pull in the end vs bare or reflectors.

yea...i've looked at them enough..may as well try it out.

Thanks for the link @medicinehuman, i'll be testing the same one you posted

Are cobs like people in that they'll insist they don't need corrective lenses until they finally get them?