LED grow light with adjustable wavelength ratio

[Karl H.]

Hello all,

i just wanted to show you my new homemade LED grow light.

I use fourteen different wavelengths in my grow light. On my current LED Panel four of the fourteen wavelengths are fixed, as kind of a "background radiation". The intensity of the other ten wavelengths can be adjusted, via the LED driver i built for my grow light.

This way you are able to create whatever mixture you want within the fixed wavelengths.

But enough sayed (I hope you can understand what i tried to say. My English isn´t that good)

Here are some pictures, of the LED panel, and some random wavelength ratios.

five Stage LED driver:

LED panel:

random wavelength ratios

Some feedback would be nice. :)

greetings

Karl H.

Edited August 16, 2013 by Karl H.

[Zlatokrt]

Looks good, i have been thinking about similar thing in the past, but my electronic skills are not good enough to built it (mainly the driver). :-)

What is the average input/output?

[Karl H.]

Hallo,

thanks for the feedback. :)

The divers have an efficiency of approximately 92-96%, depending on the environmental conditions.

Compared to a Philips 840 T5 fluorescent tube, i´m able to save nearly 80% of Energy, with the same area illuminated, and the same or even better growth an coloration. But to get such a high energy reduction, the driver has to be tuned very precisely.

greetings

Karl H.

[mobile]

How are you calculating the efficiency compared to 840 fluorescents? Are you making the assumption that the wavelengths of the LEDs are 100% efficient for plant growth?

[Karl H.]

Hello,

no i´m not making the assumption that the wavelengths are 100% efficient for plant growth. They are probably not 100% efficient.

To figure out the efficiency compared to the 840 flourescent tube i made an experiment.

I saw out different Plants (mainly sundews) on glass wool. The seedlings all got the same amount of fertilizer. One group was illuminated with my LEDs an one group was illuminated with the fluorescent tubes. The input power of both the LEDs and the tubes were monitored constantly.

First i weigh out the seedlings every two days under the flourescent tubes. With this data i was able to determine the biomass implementation per time division on a given input power for the photon source.

Now i tuned my LED light, to get the same biomass implementation per time division, as with the flourescent tubes.

At some point i reached the same biomass implementation with my LEDs as with the flourescent tubes. Now i started, to tune the LEDs so that the seedling also get the same coloration as under the flourescent tubes. I did this by measuring the pigment densitiy with a standart camera an an avaluation software. (by simply comparing the color saturation on different parts of the plants)

After 6 months of tuning measuring an comparisons, i got the same growth results of the seedling, under the two lamps.

(Of course i sow out new seeds every time i changed a value on my LEDs. As you can imagine it took a very long time to compare the two lamps)

As i now got the same growth results with the flourescent tube and the LEDs, i compared the input power. The result was a 80% energy drop of the LEDs compared to the input power of the 840 flourescent light fixture.

The result certainly isn´t 100% exact. But i was able to minimise the measurement inaccuracy by using high highly accurate laboratory scales. So i think i got pretty close, to the actual efficiency value.

greetings

Karl H.

[mantrid]

How many seedling were tested? you can improve the validity of your results by using a larger group of seedlings to eliminate growth variations determined by genetical differences

[Karl H.]

Hello,

every time i changed a value i sow out 50 seeds. All seeds were from the same plant.

greetings

Karl h.

[pmatil]

Very nice. I'm thinking of designing an adjustable brightness plant light with LED modules. I'll maybe use DX constant current drivers for each LED or use one larger one or design one myself as I'm an electronics engineer.

Edited August 22, 2013 by pmatil

[Karl H.]

Thank you. :)

Making the PCB´s on your own is probably the best, and cheapest way. I also made all PCBs by myself. The driver costs only 2€ of materials to make it.