The Planted Tank Forum banner
1 - 20 of 22 Posts

·
Registered
Joined
·
240 Posts
Discussion Starter · #1 ·
I built a DIY LED fixture that uses 27 Cree XP-Gs to light a 36" wide x 18" deep x 21" high tank. It hangs about 8" off the water surface and I run it with dimmable drivers at up to 1300 mA. (Pictures here). I'm wondering (a) how much PAR I'm getting at different dimming levels, and (b) whether I can increase PAR at depth by using optics.

I have a full set of 60 degree optics that I could install on the individual LEDs. But looking at them, I don't see how they could do much to increase light penetration into the tank. (The ones I have are these.) They are just white plastic cones with a rippled clear plastic lens at the end. The white plastic part absorbs light--it can't reflect or redirect it. So you waste any light that the LED emits outside the 60 degree cone. And the light that does emerge inside the 60 degree cone was already going in the way I wanted...right?

So what do LED optics do besides blocking light spillage?
 

·
Registered
Joined
·
357 Posts
optics help focus the light of the diodes. all cree LEDs have internal optics that spread at ~120 degrees, which is very wide and lends itself to less of the light beam effect and better color blending for multiple types of leds. It also provides good coverage at a lower height above the waters surface. if you have a very deep tank or need more intensity in a given area optics can deliver. typically though you need to raise you fixture higher to attain good coverage.
 

·
Registered
Joined
·
21,012 Posts
My guess for the PAR you are getting, with the light 27 inches from the substrate, is:
@350 mAmps - 120 micromols of PAR
@1 Amp - 290 micromols of PAR
@ 1.3 Amp - 350 micromols of PAR

That is a lot of LEDs for that size tank if they are all cool white and all on at one time.

With 60 degree optics those light intensities should increase by about 30%.
 

·
Registered
Joined
·
148 Posts
But looking at them, I don't see how they could do much to increase light penetration into the tank. (The ones I have are these.) They are just white plastic cones with a rippled clear plastic lens at the end. The white plastic part absorbs light--it can't reflect or redirect it. So you waste any light that the LED emits outside the 60 degree cone. And the light that does emerge inside the 60 degree cone was already going in the way I wanted...right?
I have those optics. While they just look like little bits of plastic there is more going on in there than you give them credit for. They don't simply attenuate the beam, they do actually refocus it into a smaller angle. You lose some light output due to absorption or bleed through the bottom of the lens (as with any optic) but I bet you're only losing 15% or so, perhaps less.

You can absolutely increase PAR at depth with optics. Furthermore, you get a more consistent PAR between the surface of the water and the substrate, which is a good thing.
 

·
Registered
Joined
·
240 Posts
Discussion Starter · #5 ·
My guess for the PAR you are getting, with the light 27 inches from the substrate, is:
@350 mAmps - 120 micromols of PAR
@1 Amp - 290 micromols of PAR
@ 1.3 Amp - 350 micromols of PAR

That is a lot of LEDs for that size tank if they are all cool white and all on at one time.

With 60 degree optics those light intensities should increase by about 30%.
Hoppy, I generally trust you on these things, but that just can't be right. With all 27 LEDs running at 350 mA, the light in the tank looks quite dim. I've been running this thing at about 7-800 mA for several weeks now and I have no reason to think it's too much light--if anything, I'm worried about the opposite. My glosso, for instance, is growing for the sky instead of spreading out. Ludwigia arcuata is growing beautifully and so are some crypts. There's just no sign of the scorching I would expect at those PAR numbers.

Do you mean that these are the PAR readings you would guess for the top of the tank?

Also, I think I'm using something like 10 cool whites, 8 warm whites, and 9 neutrals.
 

·
Registered
Joined
·
25 Posts
With an opticless fixture as high as the pictures you linked I would be running those led's at 1000ma.

My led fixture has the led's about 5" off the water (with out optics) your's looks like its about 18 inches?
With light and the Inverse-square law when you double the distance the light intensity is divided by 4 basically. I would lower the fixture if you need more light, if you must have it high up then add optics or crank up the led's to 1000-1300ma.
 

·
Registered
Joined
·
21,012 Posts
A LED light fixture like that one doesn't follow the inverse square reduction in light with distance. That is because the closer you get to the light, the fewer of the LEDs are contributing to the light at any one spot. If you had one of those fixtures 4 feet above the substrate, then I think you would get near the inverse square output. With the last LED light I made, which is similar to this one, I found that the intensity dropped directly proportional to the distance. I came up with a crude way to guess at the PAR you will get from such a light, using that data and data from other lights I have, getting this: http://www.plantedtank.net/forums/diy/136733-led-light-36-high-tank-18.html#post1443426 That is how I guessed at what PAR you are getting. Since I have no PAR measurements with XP-G LEDs it is just a guess.
 

·
Registered
Joined
·
122 Posts
The optics you have, and that most of use, are called TIR lenses, or Total Internal Reflection. Basically what that means is that the exterior profile of the lens body acts as a reflector would. When dealing with light, you can easily bouce it off a transparent surface up to a certain angle (this is where the Reflection part comes in). This works due to the differences in density between the two materials, typically PMMA acrylic for the lens, and air for the surrounding area. Typically, angles of incident under 45 degrees will just reflect, and continue on it's path. Angles greater than 45 degrees can pass through the surface. This is where efficiency loss come in. The white plastic housing is there to help reclaim some of the light that escapes the lens (white is a very good reflector).

Reducing the lens angle of the LED can significantly increase the intensity seen on the illuminated surface/object. It reduces the coverage area too, so keep that in mind. Not all lenses are created equal, so giving a broad "lens angle X will increase intensity by Y%" statement to using lenses is a little tough. I've seen some cheap lenses loose up to 20% output compared to a good lens.

As to why your high PAR LED fixture is not visibly bright to your eye is down to how LEDs produce light, and how your eye reacts to it. If you look at the spectral output of a cool white LED, and then at the sensitivity curve of the human eye, you will see that they are almost inverse of each other. LEDs produce the bulk of their light in the blue and red/orange regions of the visible spectrum. Your eye is much more sensitive to green light, and tapers off as you head more into the red and blue regions. As a result, an LED setup can and will be high PAR, but visibly dim compared to other light sources.

Anyway, a little long winded, but that's the explaination of what you are seeing.
 

·
Registered
Joined
·
240 Posts
Discussion Starter · #10 ·
As a result, an LED setup can and will be high PAR, but visibly dim compared to other light sources.
Thanks for the very informative post, Evil. I'm wondering what you would estimate for PAR levels for a fixture of the type I built:

* 36x18" tank footprint
* 27 XP-G LEDs (9 neutral, 10 cool, 8 warm white) driven at 1000 mA
* 27" from array to substrate
* either no optics or 60 degree optics.

If Hoppy is right that I'm in the neighborhood of 300umols, I'd better run home right now to dial down the dimmers--those light levels are appropriate for stony corals, not java fern...
 

·
Registered
Joined
·
1,985 Posts
They are just white plastic cones with a rippled clear plastic lens at the end. The white plastic part absorbs light--it can't reflect or redirect it. So you waste any light that the LED emits outside the 60 degree cone. And the light that does emerge inside the 60 degree cone was already going in the way I wanted...right?

So what do LED optics do besides blocking light spillage?
I'm curious as to how you reached these conclusions? No offense, but you are out to left field.

White is the best reflector. White reflects every colour colour of light, and thy is why it is white. As a result, it is very good at reflecting and redirecting the light. As a result of that, it focus all the light (well, most) outside of 60 degrees back into the cone. This adds to the light already within the 60 degrees, increasing the light output. The led doesn't put out any extra light, it just focuses it into a smaller area.
 

·
Registered
Joined
·
240 Posts
Discussion Starter · #12 ·
No offense, but you are out to left field.

When someone says "no offense" I usually assume they really mean "I know that you will be offended by what I am about to say, but I'll say it anyway."

What I should have said is that a white surface tends to scatter light rather than redirect it. So yes, white is a better reflector than, say, black. But if it were the "best reflector" for optical purposes, then car headlamps, telescope mirrors, and solar oven reflectors would be white. And yet they are not.

I suspect that our little optics are primarily little scatter chambers that bounce light coming from the LED outside the 60 degree around inside the cone space. A little might bounce a few times and make it out the lens, but most is absorbed as heat. This is just a suspicion. It could be tested pretty easily.
 

·
Registered
Joined
·
21,012 Posts
When someone says "no offense" I usually assume they really mean "I know that you will be offended by what I am about to say, but I'll say it anyway."

What I should have said is that a white surface tends to scatter light rather than redirect it. So yes, white is a better reflector than, say, black. But if it were the "best reflector" for optical purposes, then car headlamps, telescope mirrors, and solar oven reflectors would be white. And yet they are not.

I suspect that our little optics are primarily little scatter chambers that bounce light coming from the LED outside the 60 degree around inside the cone space. A little might bounce a few times and make it out the lens, but most is absorbed as heat. This is just a suspicion. It could be tested pretty easily.
I have tested an LED light with and without the lenses. There is no question that they do increase the PAR you get from the light.
 

·
Registered
Joined
·
1,121 Posts
Agreed with Hoppy. Adding optics can send PAR readings through the roof.

They do drop the total light output for the reasons that you are stating, but it is not a fatal flaw by any means at all.

In your case I would not add optics. I would, however, build the enclosure to block any light overspill. XPG's with no optics make wonderful cornea cookers.
 

·
Registered
Joined
·
122 Posts
I suspect that our little optics are primarily little scatter chambers that bounce light coming from the LED outside the 60 degree around inside the cone space. A little might bounce a few times and make it out the lens, but most is absorbed as heat. This is just a suspicion. It could be tested pretty easily.
It's a little more technical than being a "scatter chamber", but that's the basic principle. They way the light bounces around inside the lens is a little more deliberate than you make it out to be though. TIR lens efficiency is usually around 80% or higher (there are some crappy lenses that are worse than this). Some of the losses are due to absorbtion (even PMMA arcylic isn't 100% optically transparent, and as a result does absorb some light, turning it into heat), but the remainder of the losses are due to the light escaping the lens body due to too high of an angle of incident at certain points. The amount of light lost in a secondary 60 degree lens is pretty insignificant compared to the intensity increase produced by the narrower projection angle.
 

·
Registered
Joined
·
240 Posts
Discussion Starter · #16 ·
The amount of light lost in a secondary 60 degree lens is pretty insignificant compared to the intensity increase produced by the narrower projection angle.
Thanks again. I've moved from skepticism to planning to install these.

In your case I would not add optics. I would, however, build the enclosure to block any light overspill. XPG's with no optics make wonderful cornea cookers.
Redfish--you are way too late to get me to redesign this enclosure! It is done, and I'm all out of curly maple anyway. I think I will raise the fixture back up to 12" and try optics. Light spillage is definitely a problem otherwise.
 

Attachments

·
Registered
Joined
·
1,121 Posts
ROFL sorry about that.

But your pendant will do very well with 60's.


I do not recall what LEDs you are using, but if they are XPG's, expect to take at least 17 days to get them properly mounted. EVERY time I mount optics to an XPG that was already soldered, I end up having to desolder it, glue the optic in place with clear silicone or cheap white caulk, let it dry, an then resolder it.

It's worth it, but it's the proverbial PITA. Hopefully you will get lucky though lol.

If they are XRE's, they should snap right on, like a boss.
 

·
Registered
Joined
·
1,985 Posts
Thanks again. I've moved from skepticism to planning to install these.



Redfish--you are way too late to get me to redesign this enclosure! It is done, and I'm all out of curly maple anyway. I think I will raise the fixture back up to 12" and try optics. Light spillage is definitely a problem otherwise.
That enclosure is awesome. One of my favorites. I seriously didn't mean any offense, I just thought you were pretty far off-track. You've got it figured now though, adding the optics and raising the fixture wil decrease the spillage and make more uniform par levels from top to bottom. It's a better alternative to a low fixture without optics.
 
1 - 20 of 22 Posts
This is an older thread, you may not receive a response, and could be reviving an old thread. Please consider creating a new thread.
Top