The Planted Tank Forum - View Single Post - DIY led lighting - have I got it right before purchasing
 View Single Post
01-06-2013, 02:10 AM   #14
Build My LED

Join Date: Nov 2012
Location: Austin, Texas
Posts: 68

Quote:
 Originally Posted by ced281 I think you're right about the generic LEDs being less efficient at converting watts to lumens than the crees. But I think you can partially work around that by buying the "high power" LEDs which are supposedly rated at 160+ lumens. Here are 3W LEDs from China which generate 170-190 lumens. [Ebay Link Removed] If you look at the specs for the Cree XP-G's their 5W LEDs generate 260 lumens at 700mA. http://www.cree.com/~/media/Files/Cr...g/XLampXPG.pdf Anyone know how to calculate the lumens per watts for these? Will a 5W Cree really run at 5W if you're only running at 700mA instead of it's max 1.5A? From back in my high school days, P=IV so Power = Current * Voltage. I'm assuming that the voltage doesn't change for these guys so if you decrease current the power (wattage) should decrease proportionally. > So the power consumption of a 5W Cree running at 700mA is closer to 2.3W. > Which means the lumens per watt is ~113 > Whereas the lumens per watt of these "high-power" Chinese LEDs are ~60 (180lumens/3W) > Which means the efficiency of the Crees is almost 2x! Does that make sense??? Some of the limitations of light penetration by lower wavelength colors can be somewhat dealt with by using more focused beams (who knows how much it can be avoided though). Having greens and yellows can be argued to be "wasted" light because they don't aid in photosynthesis, but the more green and yellow light you send into the aquarium, the more those colors will shine off your plants and into your eyes. I would be careful about having too many red likghts. I read somewhere (I think in the lighting sticky) that algae is better at utilizing red than blue light, so having high levels of red light without proper plant load could be conducive to algae blooms, etc. More food for thought! Let us know what you're planning. If I end up having more free time I might start up a thread documenting my DIY lighting for my DoAqua 90P tank. I just got my makersled fixtures and am waiting on my LEDs from China =]
Hey guys/gals. I wanted to chime in here on a few of these topics, as there is so much misinformation concerning LEDs and spectrum related to plant growth on the Internet. Before I started Build My LED, I managed a horticulture lighting division for an LED manufacturing company. We competed against all of the global LED companies (i.e. CREE, Philips, Bridgelux, etc.), so I have some direct experience in this segment. Having said that, I am not a planted tank expert I do know lighting and LEDs, so together, we should be able to help move the ball forward.

Concerning Lumens per watt. First of all, you need to compare apples to apples when comparing LEDs. LED manufacturers don't have a consistent method to publishing these numbers, so you need to dig into the spec sheets. Most companies publish this metric by flashing the LEDs with a 20 millisecond pulse of electricity. The LEDs are ‘cold’, so they are very efficient during the test. If you would test that same LED 10 seconds later, it would produce a lower Lumen number, since LEDs produce less light as they heat up. If you would test that same LED after it has been sitting over an aquarium for three hours, you would have an even lower number. While on this topic, it is important to note there can be significant losses with the optical and electrical systems in any lighting system. By the time you consider all three levels of losses (thermal, optical and electrical), the overall fixture efficiency is nowhere near the values published for the bare LED.

Concerning the green/yellow light, wasted energy issue. In summary, this is the worst myth on the Internet concerning horticulture lighting and photobiology. As long as photons between 400 and 700nm are absorbed, they are useful for photosynthesis. There are no wasted wavelengths in this band of light (electromagnetic radiation). Here is how the myth is usually propogated: “Plants are green, so they are reflecting the green light back to your eye, so the plants are not using it. Hence, green light is wasted energy. “ Scientifically, this is absolutely wrong. Green light is very useful to plants, it just doesn’t get absorbed by the chloroplasts as efficiently as blue and red light. Hence, the plant appears green. However, it is not reflecting all of the green light, and green light even has some advantages over red and blue light. Green light (and far red light) can penetrate deeper into the plant canopy, so spectra with green light usually outperform spectra that only contain red and blue light. In fact, NASA published a paper concerning this issue in 2004, and the green-enhanced light actually grew 45% more biomass than the red and blue spectra.

Finally, Kelvin is basically a useless metric for comparing different light sources, as it does not define a single color. In other words, you can buy a 5700K light from 20 different companies, and they will all look different, even though they are all correctly labeled 5700K. Google the 1931 CIE color chromaticity diagram, and you will see vertical black lines near the center of the chart. These vertical lines define the specific color temperatures (i.e. 5700K), and you will notice how they stretch from the pink region into the green region. That is why you will see so much variation while comparing lights based on the Kelvin scale. Alternatively, I recommend you look at the actual spectrum of light, as this is what is used to calculate the Kelvin temperature. By looking at the spectrum power distribution (SPD) chart, you will be in a much better position to compare various light sources for horticulture applications.

I hope this helps ‘shed some light’ on the subject on LEDs and horticulture lighting. This wasn’t meant to be a comprehensive dissertation on the above topics, but I wanted to chime to add my two cents to the discussion :-)

Nick

Last edited by Build My LED; 01-06-2013 at 02:14 AM.. Reason: spelling