Building LED lights from facts, no theories

[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.