Building LED lights from facts, no theories

[McCheez]

hmmm... red and blue... Cop lights might work...

Let's get some Maui Wowie and try it. We could call it Hawaii 5-0!!!:rastasmoke:

McCheez Reviewed by McCheez on 07-03-2009. Building LED lights from facts, no theories I was going to post this at the Perfect LED Grow Light thread, but as some of what im going to post was posted 2 years ago on the stickied thread about LEDs and people still continue developing lights from wrong ideas, i think a thread about this topic is largelly needed. The main problem is related to efficacy of spectrums. When the firsts LED experiments at Overgrow, we work on the hypothesis that blue and red light are more effective. It was an appealing hypothesis that promises large Rating: 5

[Bubbleblower]

A group of grower are performing experiments with standarized measurements in order to find the optimal spectrums (different according to irradiance used?) for cannabis. If you want to join us, you are welcome. Stadistical analysis of experiments may strongly fasten the research, so sharing our results benefits all.

Hey knna, I'd like to join you and may be I can interest others also.

I only heard the clock tick and don't know what time it is , but I have some ideas on how measurements should be taken and better spectrums can be found that I would like to discuss.

[knna]

Hi, Bubbleblower ,

glad to find people interested.

Light measurement is the most complicated part without expensive equipment. What I do is calculate how many micromols of photons equals to a given number of lm for a given spectrum. I obtain the spectrum of LEDs used using an spectroscope, a camera and some sofware.

Having the amount of uE emitted, at least we can know the average light density in the grow. In order to obtain irradiance data, I use the same procedure than to calculate the absolute uE, so it leads to have a correction factor between uE and lux (measured with a light meter). Only problem is only works for each separated spectrum, so I need to do a measurement for each color used and then add up all them. I use a grid to build a irradiance map for a given distance from lamps.

Consuption of LEDs alone and full system is measured using a voltmeter and amperimeter.

About yields, I suggest to record as many info as possible. Both fresh and dry weight, and unprocessed yield and manicured one. I suggested to weight all the plant weigth (without roots) is fresh, later manicure and weight stems and manicure's rest (better if separately). After drying, report the bud's dry weight and stems and manicure's rest dry weight. I know its too much work weighting, but I think its necessary to eliminate the large differences induced by how each grower process its yield. With such data for each grower, at least for one time, it would be possible to normalize weight reports in a meanglifull way.

Im really interested on any suggestion on how to standarize measurements.

[Bubbleblower]

Im really interested on any suggestion on how to standarize measurements.

I know so little that I have many.

To find better spectrums and PPFD levels of course we need photons per nm.
Some use SPD's to convert and others spectrometers to measure the irradiance and they get a -normalized- spectrum, the amount of photons per nm and total irradiance per m2 or so.

But one bulb could give its energy per cm2 on 10 times more surface than another so the sweetspot should be taken into account..
Also the normalized spectrum would be very inaccurate, because different wavelengths loose their energy in a different pace. That effect is big.
So spectrums change depending on the distance from the bulb and we are not really giving the spectrum we think or the amount of photons.

May be we can use 3-dimensional top down measurements and normalize the spectrum once more for every wavelength/distance from the bulb.
That will give a better indication of the real amount of photons and give us like an average spectrum.
We could take measurements with 10cm increments horizontal and vertical.

[knna]

Yeah, thats why I say to calculate both the "theoretical" average irradiance, meaning the absolute photon emission divided by the surface area of the grow (footage) and or/ volume area.

And aditionaly take measurements using a standard light meter (a 20-30$ one) and convert it to photons/m2 using an Excel sheet. This is a true, actual measurement of the light reaching leaves (the main measurement is taken at top of canopy level).

There is no differential absortion of different wavelenghts in air (when talking about 1, 2 or 3 meters as most, and mostly distance is way below 1m). Another thing is how light quality changes below canopy due the differential transmision of leaves, but I always claim to use undercanopy lighting as the most efficient one, that apart of it, largely avoids this problem. So irradiance measurement using LEDs is usefull both at top canopy and undercanopy.

It would be nice people has spectrometer or spectrorradiometers, but I wouldnt expect it. Its necessary to find ways of measurent that any grower interested may do it if he has the will, having a limited equiptment, as a standard light meter. My purpose is to build a database of the spectrums of LEDs most used, so all people has to do is measure with a light meter and then convert lux or fc into uE/m2.

[Bubbleblower]

Yeah, thats why I say to calculate both the "theoretical" average irradiance, meaning the absolute photon emission divided by the surface area of the grow (footage) and or/ volume area
And aditionaly take measurements using a standard light meter (a 20-30$ one) and convert it to photons/m2 using an Excel sheet. This is a true, actual measurement of the light reaching leaves .

That should give very interesting -and surprising- results.

There is no differential absortion of different wavelenghts in air (when talking about 1, 2 or 3 meters as most, and mostly distance is way below 1m). .

May be I put it wrong. I noticed for example a much bigger difference in mW/m2 output relative to distance (below 1 meter) in UVB as in longer wavelengths. Also I saw this graph on page 7 which is very interesting.

Another thing is how light quality changes below canopy due the differential transmision of leaves, but I always claim to use undercanopy lighting as the most efficient one, that apart of it, largely avoids this problem.

I like the idea of intra/under canopy lighting a lot, it was next on my list.
Since no more than 15% (?) of the light penetrates through the canopy this should boost yield per m2 (I obviously like m3 better) and only little light is needed to get significant results.

Its necessary to find ways of measurent that any grower interested may do it if he has the will, having a limited equiptment, as a standard light meter.

Lux at certain distances from the bulb, both horizontal and vertical would be easy. I think many people are willing to take these measurements if they understand why that is important and how these data will be used. Everybody -on forums like these- should know where they can find and send this information.
We should get them with the program.
I will ask in a few forums that have thousands of growreports and do many tests.

Let me go further down my list.
Just check this pic and you understand the potential implications for cannabis.

[knna]

May be I put it wrong. I noticed for example a much bigger difference in mW/m2 output relative to distance (below 1 meter) in UVB as in longer wavelengths. Also I saw this graph on page 7 which is very interesting.

Very useful article, thanks.

But that graph refers to how angle of viewing affect the spectrum. And its not related to how light transmit, but to the LED source itself. This difference is caused in white LEDs by how the phosphor layer is added. LED manufacturers had improved it strongly on the last years, but is imperfect yet.

Standard LEDs dont emit almost any UV, and less specifically UVB.

I like the idea of intra/under canopy lighting a lot, it was next on my list.
Since no more than 15% (?) of the light penetrates through the canopy this should boost yield per m2 (I obviously like m3 better) and only little light is needed to get significant results.

Purdue University is working hard on this, and in general all researchers involved on the development of grow areas for NASA missions.

Planophills plant's species strongly benefit for intracanopy lighting. Average improvement in yield is being 35% for the same amount of light used, but some species has increased it by 100%, doubling yield with same amount of light.

Percentages of IC lighting generally varies from 25 to 50% of the total light.

Due undercanopy areas are mostly underlit, and many times severely underlit, very little light (compared to the amounts of light used overcanopy) transform lower leaves from sinks to sources of energy and resources. Bottom leaves arnt discarted, as happen when there is only top lighting, due the arey are a ballast for the plant. This mean undercanopy lighting achieve a way better use of plants resources and allows to use optimal irradiances for all the leaves of the plant, strongly improving photosynthetic efficacy.

We are getting excelent bottom buds using very low irradiances undercanopy, below half of used overcanopy (on uE/cubic feet basis).

I believe Ive linked it before, but check here the Excel spreadsheet

[Bubbleblower]

Very useful article, thanks.

But that graph refers to how angle of viewing affect the spectrum. And its not related to how light transmit, but to the LED source itself.

Unfortunately there is a lot I don't understand correctly, also because I got much disinformation. May be I did something wrong, I compared a light-distance chart of 5 HPS bulbs with measurements of UV output from 8 other lamps.
uW/m2 was -relatively- lost faster than foot-candles. The graph seemed to confirm what I read in all the why plants are green articles.

Standard LEDs dont emit almost any UV, and less specifically UVB.

For my UVB I use Philips medical grade UVB broadband. It seems somehow the most "extreme" wavelengths within extended PAR are most important to the plant.

Planophills plant's species strongly benefit for intracanopy lighting. Average improvement in yield is being 35% for the same amount of light used, but some species has increased it by 100%, doubling yield with same amount of light. Percentages of IC lighting generally varies from 25 to 50% of the total light.
Due undercanopy areas are mostly underlit, and many times severely underlit, very little light (compared to the amounts of light used overcanopy) transform lower leaves from sinks to sources of energy and resources. Bottom leaves arnt discarted, as happen when there is only top lighting, due the arey are a ballast for the plant. This mean undercanopy lighting achieve a way better use of plants resources and allows to use optimal irradiances for all the leaves of the plant, strongly improving photosynthetic efficacy.

We are getting excelent bottom buds using very low irradiances undercanopy, below half of used overcanopy (on uE/cubic feet basis).

Great. So what I need for my room is at least 8 rings of about 25 watt of leds in the correct combination.
It is for 1.4m2 where I have 2 induction -electrodeless- lamps of 300 watt* plus the 2 x 40 watt UVB (and 2 x 50 watt HPS but only as signaling light).
My idea is also to "finish" my spectrum with red leds most of all.

I believe Ive linked it before, but check here the Excel spreadsheet

Darn, homework :thumbsup:

* Here is a testreport of the 6500K version of it with an accurate SPD. I should still get this for my 2700K.

[Bubbleblower]

I believe Ive linked it before, but check here the Excel spreadsheet

Mmm, I still don't get the privilege to access that page, do you have it somewhere else?

[oldmac]

Hey Bubbleblower,

Just saw the spectrum graph you got there, I would like to point out to you to be very carefull, it may or maynot be accurate. The chinese have a tendency to lift spectrum graphs from american companies, and thier diodes don't actually meet those graphs.

The first large DIY LED light I built was going to be based on reds and white LEDs, since the whites (I had looked at) had plenty of blue spike (driving the phosphor) and then covered green, yellow and some orange/red. When finished the light sucked, it lacked the needed blue. I took a white diode sent it to someone to spectrographed it's output was a lot different. The blue spike was there but way to the left missing the needed wavelenght. The rest never reached orange/red. After looking around I found an identical spectorgraph, like the one they sent me, for a luxeon white. I have not trusted the chinese since, I thought they would saved me money but in the long run a costly lesson.