I'm a science teacher and we are going to explore different design/effects of grow-light setups this year in class. I've gotten some great ideas for this project from reading from this stackexchange site, thanks! I also see problems I hadn't even considered, double-thanks :-)
For cost and brightness and versatility purposes I'll be buying the ~$10/5m strips of 5630 SMD LED's. But, there are issues of efficiency (current-controlling resistors wasting power) and unrealized brightness needing to be address and also here.
First, in the question exploring why there was such a huge loss of energy with LED's that are energy efficient, it was explained that the blue and white LED strips might be up to 80% efficient, but red LED strips be only 45% efficient. Now, it's not the LED's 'fault', as I understood that discussion, but the circuit design. The strips (regrettably) used a consistent circuit design, regardless of the forward voltage of the LED's being used. The blue and white LED's have a 3-3.6 forward voltage and with 3 in a series meaning they will need anywhere from 9 to 10.8 of those volts, but the red LED's have 2-2.2 Vf so will only need 6 to 6.6 of those Volts. And, if one is using a 12 V power supply, there's a lot of 'pressure' that needs to be reduced with resistance. And as I understand it, the resistors 'waste' that energy. SO... I'm wondering, if I can match my power supply more closely to the demands of what each 3-LED module needs and then adjust the resistance to what it needs to be, I could substantially increase the efficiency of the strips. (I know, it's crazy, but remember, I have many students who would be happy to solder resistors in parallel on these strips!!!--a learning opportunity.)
What do you think about this idea? Lots of these sites where I buy don't give the Vf specs of the individual 5630 LED's so how could I find that out?
Relating to that question, how ought I decide at what current to drive these LED's (so as to be able to calculate the total resistance each module will need)? The discussion about why an LED strip was using less power than expected got me wondering if I should be driving the strips at a higher voltage than they recommend, for example 14.5 Volts to get the hoped-for brightness as given in that discussion. I know heat sinks and LED life relate to this, I'd really like to hear people's thoughts and recommendations on this, too.
For those people interested, from discussions w/plant people I've had at the U of MN about this project, I'm going to suggest to my students to start out with a 3:2:1 ratio of red:warm white:blue LED's in their initial designs of plant-grow systems (though that is the ratio of lumens so depending on the comments on the preceding questions, this may not be the ratio of lights, themselves).
EDIT to include new questions from comments: Do I need the data sheet to determine what the highest light value I can drive the LED at and still have it last a long life? Or, can I driving it up until it reaches some rule-of-thumb temp.
Also, It seems that the closer one can match the voltage of the power supply to the sum of the Vf of the LED's, the smaller the resistor and therefore the less power wasted. Is that true? What margin of error should one calculate for? (BTW yes, each of the modules has 3 LED's and 1 resistor in parallel.)