cheap ebay DIY LED

[Nexogen]

Very cool naoki, Bravo! However some objections; LED device let's say efficient is only CXB3070 3000K, 90 CRI. I have not seen any protection if the fan breaks down.

You can probably use something like:
2013newestseller New Dc12v Heat Cool Temp Thermostat Temperature Control Switch Thermometer -50-110°c - Household Thermostats - Amazon.com
KSD301 250V 10A 176F 80C NC Thermostat Temperature Control Switch 5pcs - Programmable Household Thermostats - Amazon.com

[naoki]

Nexogen, are you saying 90 CRI is better than 80 CRI? Or are you saying that there are other efficient LEDs? If later, I agree, there are other efficient ones other than Cree and Bridgelux. For example, XF-3535L I posted earlier in this thread is pretty good. Even these cheap chinese COB LEDs, if you run the "100W" COB at 5-10W, you can gain the efficiency. But for the relatively easily available and affordable model, these two (CXB and Vero) are difficult to beat.

With regard to CRI, here are the numbers from the data sheet:

With CXB3070, 3000K at 1900mA 85C

80 CRI AD bin: 9000 lumen
90 CRI Z2 bin: 7390 lumen

Both of them are the top bins for each category.

Now, we don't care about lumen, so we need to convert these to PAR (micro mol/s) after considering the difference in the emission spectra.

Some other people have digitized the spectra, and based on his data, I can make following calculation:

Cree 80CRI 3000K has 0.01424752 micromol/lumen
Cree 90CRI 3000K has 0.01633887 micromol/lumen

80 CRI AD bin: 128.2277 micromol/s = 9000 lumen * 0.01424752 micromol/lumen
90 CRI Z2 bin: 120.7442 micromol/s = 7390 lumen * 0.01633887 micromol/lumen

So 80CRI gives 6% more PAR. Generally, lower CRI gives more efficiency for photosynthesis. An easy way to understand this is like this: to get a higher CRI, they generally need to use "thicker" or "more" phosphors, so there is more loss when the blue light hit the phosphor.

I'll probably post about efficiency a bit more when I get some time (not for a while), but similar calculation tells that the CXB3070 low bin (AB) 80CRI 3000K is "theoretically" getting about 1.96 micro mol/J at 1.4A (assuming 50C Tj), which is better than BML (1.66 micromol/J) and beyond the best HPS and LED (both around 1.7 micromol/J) measured in this paper (table 3). Well, I admit that it is a bit unfair to compare the theoretical value against the empirical measurements.

Thanks to the second link to the thermostat. This looks interesting and cheap. In addition to the thermostat you linked (the first link), this kind of cheap one can be used:
10 Pcs Bimetal Temperature Control Switch Thermostat 90C N.C KSD9700 - Electrical Outlet Switches - Amazon.com

Basically attach this to the heatsink, and hook it up to the AC side of the driver. I have this kind of temperature cut off switch, but I haven't used them yet. Even though the CPU fans have a long life, I have experienced a couple sudden fan failures in the computer in the last 25 years or so. So it is a good thing to have the heat protection if you worry about it.

Here is a related topic:
Cooling high-power LEDs: The four myths about active vs. passive methods | EDN

[Nexogen]

I refer to efficiency in Absorption Spectrum as shown in the chart above it seems 3000K CRI 90 is best. Either method used by me is to create photosynthesis absorption spectra and "white" use to have a "Domestic" light.

---------- Post added at 08:59 PM ---------- Previous post was at 08:49 PM ----------

After doing some experiments I found that Bimetal Temperature Control Switch is not good contact in the first few seconds when reappear ON, so use it only as a last protection, the electronic one is very good.

[naoki]

Are you talking about the fact that red light is more efficient in the action spectrum of photosynthesis?

I calculated the PAR based on the spectral power distribution (SPD) curve which is similar to what you linked. If I digitize the Action spectrum of PS, then I can calculate the Yield photo flux (YPF), which basically incorporate the action spectrum. Then I agree the difference will become smaller.

One thing you may or may not know is that when you are looking at the SPD, Y-axis is relative energy (watt), but photosynthesis is based on the number of photons, so you have to convert the SPD to number of photons. Blue photons have higher energy than red photons, so the blue peak becomes smaller in the plot with # of photons on Y-axis.

---------- Post added at 05:14 PM ---------- Previous post was at 05:07 PM ----------

Originally Posted by Nexogen

[/COLOR]After doing some experiments I found that Bimetal Temperature Control Switch is not good contact in the first few seconds when reappear ON, so use it only as a last protection, the electronic one is very good.

Yes, it should go on and off, right? I wonder if turning on and off the driver at the AC side quickly damages the driver or not. It wasn't recommended to cut off the DC side because CC driver try to compensate for infinite resistance before the protection kicks in. I'm assuming that it can handle it, but I'm not sure. There are a couple other (more elaborate) heat protection which people do in rollitup.org.

[Nexogen]

I use high voltage output drivers and few COB in serial. However, if interrupt line, driver enter in protection.
I have a bimetal (80 degrees Celsius) with COB series and electronic (50 degrees Celsius) on the power 120V to the driver.

---------- Post added at 09:36 PM ---------- Previous post was at 09:33 PM ----------

And yes, in my experience red light is more important in the action spectrum of photosynthesis.

---------- Post added at 09:43 PM ---------- Previous post was at 09:36 PM ----------

In my opinion, in the picture with 4 COB and orchids need at least 100W 620-660nm Red plus 50W Blue.

---------- Post added at 09:48 PM ---------- Previous post was at 09:43 PM ----------

The light is too high. I think you need to use reflector coup and lens

---------- Post added at 09:57 PM ---------- Previous post was at 09:48 PM ----------

believe me this one is extra
2013newestseller New Dc12v Heat Cool Temp Thermostat Temperature Control Switch Thermometer -50-110°c - Household Thermostats - Amazon.com

[Nexogen]

A type of driver used is expensive but very good (price as of 4 drivers 50W).
HLG-185H-C1050B Mean Well LED Power Supplies
http://www.google.ca/url?sa=t&rct=j&...06674449,d.dmo

[naoki]

Originally Posted by Nexogen

And yes, in my experience red light is more important in the action spectrum of photosynthesis.

McCree 1972 has the measurement (and frequently used for YPF calculation). blue peak is about 70% of the red peak. Inada 1976 got the similar results. But you should note this is monochromatic response. The data are from crop plants, and there is no example of action spectra from shade plants.

Originally Posted by Nexogen

In my opinion, in the picture with 4 COB and orchids need at least 100W 620-660nm Red plus 50W Blue.

For my plants, it is actually perfect. At the leaf level, I get around 150 micromol/m^2/s. The taller ones get a bit too much. Williams (1983) measured photosynthetic rate of Paph insigne, and it started to saturate around 100 micromol/m^2/s.

Originally Posted by Nexogen

[/COLOR]The light is too high. I think you need to use reflector coup and lens[COLOR="Silver"]

If I put it closer, the leaves start to bleach. Indeed, some of the large Phrags in the back goes too close to the light, and they occasionally have bleached leaves.

The problem of COBs is that they are more of point sources, so they create strong shadow effects. You can see the high contrasts in the photo. So I'm not using the reflector (so the stray light can reach to the lower leaves). Also, in the grow tent, it is surrounded by reflective materials, and I'm growing the plants on the side walls, too.

Originally Posted by Nexogen

A type of driver used is expensive but very good (price as of 4 drivers 50W).
HLG-185H-C1050B Mean Well LED Power Supplies
http://www.google.ca/url?sa=t&rct=j&...06674449,d.dmo

Yes, HLG-H-C series is popular because of the high efficiency. With 1050, you should be able to use 5 CX(A/B)3070 to get the max AC/DC efficiency.

So is yours HLG + 4x CXB3070 CRI90 at 1050mA? That is pretty nice.

[Nexogen]

Shrink written above me, not because I do not know but I could not explain.
Spectrum of chlorophyll absorption curve (about) is shown below.
The figure clearly shows what colors are absorbed.

[Nexogen]

What makes me believe "White" SSL (Solid State LED) is not enough to meet the full requirements photosynthesis is clearly seen in the figure below.

http://www.google.ca/url?sa=t&rct=j&...7IttE8yiLeXFPA

PS: Sorry to those who visit the site, I go very hard with the English language...

[naoki]

Originally Posted by Nexogen

Spectrum of chlorophyll absorption curve (about) is shown below.
The figure clearly shows what colors are absorbed. The problem is the "penetration" of light and as is well known wavelength shorter penetrate more easily so with lower losses on the route of the light from source to the plant. That says it all why red should be greater than the amount of radiated light source.

I'm trying to follow your logic, but I'm still not sure what penetration you are talking about here. By any chance, are you talking about activating the deep chloroplasts by green light? If so, in a high light situation, blue and red light get absorbed by chlorophylls at the upper layer of cells, and they don't reach to the lower level (within a leaf). But at the upper layer of the cells, the photosynthesis become saturated and the overall photosynthetic rate become not limited by the amount of light, so additional red and blue lights are wasted. If you think about how photosystem antenna complex (the place where pigments absorb light) have many many pigments, but the following PS reaction occurs only by the energy reached to the reaction center (P680/P700), you can understand this. Many chlrophylls are excited, but PS ETC can't catch up. So the absorbed red and blue light get released as heat (wasted). On the other hand, green light can penetrate deeper into the leaf because the top level chloroplasts absorbs less of these. So under the high light condition, adding green light increase the PS rate than adding additional blue or red light.

Green light drives leaf photosynthesis more efficiently than red light in strong white light: revisiting the enigmatic question of why leaves are g... - PubMed - NCBI

But I have a feeling that you are misunderstanding the relevance of Chlorophyll absorption spectra. This is a common (and amazingly wide spread) misunderstanding as others and I have pointed out here. I guess that the textbook I use for undergrad intro plant class is a bit ambiguous, so my students are frequently confused, too.

Absorption by chlorophyll is just a part of photosynthesis. There are already papers, which describe the overall efficiency of different wavelength on photosynthesis, which incorporates absorption, energy transfer among pigments, how different wavelength reaches to the chloroplast etc. As I mentioned McCree and Inada are the classic papers.

McCree, K.J. 1972. The action spectrum, absorptance, and quantum yield of photosynthesis in crop plants. Agricultural Meteorology 9: 192-216
Inada, K. A. 1976. Action spectra for photosynthesis in higher plant. Plant & Cell Physiology 17: 355-365.
These are not available online, but if you are interested, you can PM your email.

Introduction in this poster has the action spectrum (relative quantum efficiency, RQE) derived from McCree.
http://cpl.usu.edu/files/publication...b__2576523.pdf

Or if you go here, you can get the entire book of:
Plant Growth Chamber Handbook
Edited by R.W. Langhans and T.W. Tibbitts
North Central Regional Research Publication No. 340
Iowa Agriculture and Home Economics Experiment Station Special Report No. 99
Chapter 1 contains the basic information.

Another thing I should note is that understanding the requirement for photosynthesis is relatively easy; they can use any light within the PAR as long as you have enough of it. There are slight efficiency difference, but it doesn't matter (too much) if you use pure red, blue or green light. Photomorphogenesis part (how light influence the morphology and biological functions) is the place where color spectrum become more relevant.