Anyone used these LED lights?

[naoki]

Jacob, as you said, CXB3590 can be driven soft to get amazingly high PAR efficiency. I'm driving it at moderate current, 1.4A ea, and 4 of them are excellent for 4x4' (120cm x 120cm) area for orchids in a grow tent.

I'm not completely sure about the price in Europe, but I would go with Citizen CLU048 instead of CXB. There will be a new model of CXB coming out soon, and the price of CXB is reduced. But you'll get much better deal from CLU048 (at least in the US) even after the discounts of Cree.

If your box is smaller, and if you care about the efficiency/heat issue, you might want to consider mid-power strips. For example, Samsung H-series, which is supposed to be out any time now, is quite amazing (almost 190lm/W). But it is a bit pricey.

Katrina, the spectra of LED is quite different between florescent and LEDs. It probably do influence the aspects of physiology, but there isn't enough research in orchids. There is a study that in phalaenopsis, bluer light (e.g. 5000K) causes more production of some chemicals used for defense against pathogens. But other influence of spectra on orchid physiology is not tested well. For photosynthesis, there isn't much effects between 3000-5000K. I stick with anything between 3000-5000K.

For photosynthesis, I think 6500K T5HO is a myth, too. From my PPFD measurement, amount of light available for plants isn't so different between 3000-6500K. Because of human eyes, higher K seems brighter, and they are probably thinking "daylight" spectrum is good for plants. But 6500K fluorescent light is not similar to the real daylight.

[katrina]

Thanks naoki! I, unfortunately, bought my stock of T5 HO bulbs last year (I buy by the case) so I'll use those up this winter and will make the switch next time. I've been using the 6500s and they get replaced every year so I'll make the adjustment next time. Too bad I didn't ask these questions a long time ago but I've gotten good growth and bloom to date...so it'll only get better in the future.

Again, your knowledge and expertise is greatly appreciated.

BTW - because my head was spinning from all the LED info and because I didn't want to spend the $$ required to get what I thought would be the ideal LED set up...I just went w/another T5 HO unit. I figure I'll take the time to learn more and wait for the prices to come down and I'll do the switch over down the line.

[Ray]

Naoki, I think there is more to the 6500K T5HO story, as there can be different spectra that mimic sunlight. A "big box store" lamp and a good plant light might both be rated at that color temperature, but may have entire ro different spectra.

[naoki]

You are right, Ray, the 6500K can be achieved by completely different spectra. But most people are using relatively cheap, commonly available fluorescent bulbs, right? Other fancy bulbs which aquatic plant people use seem to have quite different spectra (from different phosphor, maybe), but they are quite expensive. So I was mostly thinking about the cheap bulbs. I measured one of those purple fluorescent bulbs for plants (people used to call it Gro-lux etc.), and they do have more PAR.

Also, you may be right that there could be some other reasons why some people believes 6500K fluorescent bulbs are better for orchids (or plants). I haven't found any convincing evidence (especially for orchids) so far, but I'd like to know if there is some data to support the claim.

Photosynthesis isn't the everything, and the different spectra could influence the shape/physiology of orchids. The later aspect is difficult to assess since there isn't much data yet. So looking at the photosynthesis part (approximating with PAR), here is a rough relationship (or the lack of relationship) between K and PAR. Planted tank people from South Africa has compiled emission spectra data from many different types of light. They digitized them, and calculated some relevant specification. I extracted fluorescent bulbs, and plotted PAR/W against corrected color temp (K). PAR/W is amount of photons usable for plants (between 400-700nm) for a given watt of electricity (the unit of y-axis is micromol/J). A higher value means better efficiency for plants.



You have to be careful with the X axis, which is the "calculated K". The K value is calculated from the published emission spectra, so they are different from what the company call it. I don't know why, but they probably used slightly different method of calculation, or digitizing error? For example, warm white (3000K) is around 4500-5000K in the plot. 6500K bulbs appears around 8000K. Basically there is nothing special about 6500K.

This data set contains both fancy bulbs and common bulbs, so you might think that this would obscure the pattern. I looked at a single brand (Osram), and higher K might have slightly more PAR, but it is very small advantage.

A final note is that PAR/W in here isn't the system wide efficacy (i.e. energy loss in ballast is not considered). Modern DIY white COB can achieve >2.0 micromol/J, and some can push it to 2.5 range. But many cheap LEDs are just a hair above those fluorescent bulbs.

[Jacob Reitsma]

Thanks everyone for being a great source of information about LED!

My grow box is 70*80cm and 100cm high and my new LED plan is 3*CLU048 3500k 90cri and 3*CLU048 4000k 97cri. They both have a reasonable lm/w of 125 and 116 and they will give me 26.700 lumen wich is plenty I guess? I will also install a dimmer.
For cooling I was thinking about Arctic Alpine 11 GT Rev. 2, they are cheap and only take 1.5w each at full power( I will install a dimmer on them also).


Feel free to give me feedback because it's my first LED fixture and also my first grow box.

Greets!

[naoki]

Jacob, 26700 lumen for the space could be a bit too much. Actually half of that is probably enough. You could use dimmerble driver.

If you want to maximize PAR, you want to go with the lower CRI. There is a confidential/internal document from Citizen floating around. According to the doc, here are the values comparing PPF/W (micromol/J) of different CRI and K.

3000K 70CRI 2.08
3000K 80CRI 2.12
3000K 90CRI 1.94

3500K 70CRI 1.96

4000K 70CRI 2.09
4000K 80CRI 2.10
4000K 90CRI 1.98

5000K 70CRI 2.21
5000K 80CRI 2.17

These values are from CLU058-1825C4 (generation 5), driven at 2.25A, Tc (case temperature)=25C (pulsed). So real values (non-pulsed with higher Tc) will be lower than this.

In addition to PPF, they calculated something what they made up, called A-PPF (absorption PPF), which basically ignores 510-610nm (green). Then higher CRI has more photons in the red and blue regions. However, as Ray has pointed out, this is bogus because absorption spectrum doesn't tell you the complete story of photosynthetic efficiency (i.e. the difference between photosynthetic "action" spectrum vs chlorophyll "absorption" spectrum).

If you are going with bunch of smaller ones (e.g. CLU048-1212), another option is using a bigger heatsink like Nexogen's. I don't know the source of heatsinks in Europe, but in the US, we can get cheap ones from HeatsinkUSA. For me, the shipping was too expensive, so I use CPU heatsinks (many are from recycled). I made one for my friend with a similar heatsink, Arctic Alpine 11 Plus, and it is decent and quiet. One thing I didn't like about it was that it uses a proprietary fan.

[estación seca]

Think about individual, tiny chunks of light - photons - as crumbs of food you throw at your plant. Some crumbs are more nutritious than others. Even the most nutritious crumbs have to be eaten in sufficient quantity for the plant to survive. The plant needs still more of the less nutritious crumbs, if that's all that is available.

When we describe light for plants, we care about how many crumbs we give, and how nutritious they are.

The number of photons is measured in photon flux or lumens. Our eye perceives more photons as "brighter."

The "nutritive value" of the photons depends on their color, or wavelength, which words mean almost the same thing. We want to give plants sufficient numbers of photons of the correct color. Photons have differing amounts of energy. Our eye sees this as color. Lowest-energy visible light is red. Progressing up the spectrum, photons carry more energy: (lowest) red > orange > yellow > green > blue > indigo > violet (highest.) There is light we cannot see, of lower energy (infra-red) and of higher energy (ultra-violet, to X-rays and beyond) than visible light.

Chlorophyll traps photons. It uses this energy to split water into oxygen and hydrogen atoms. The oxygen atoms combine to form oxygen gas, which floats away into the atmosphere. The plant attaches the hydrogen atoms to carbon dioxide, in a process that forms sugar. The plant can combine sugar to form starch or wood. The plant can also chemically eat the sugar to grow, and use the energy to form almost all the other chemicals the plant needs. (The remainder have to come from the soil.)

Chlorophyll is best at catching red and blue photons. It reflects green photons, which is why plants appear green. Red and blue crumbs / photons are useful for plants. Other colors are less useful. Green is just about useless.

So, for growing plants, we care about the number of photons we provide, and the color of the photons.

Photon flux, the number of photons flying past a given point per unit of time, is measured by lumens.

The "color temperature" of a lamp, reported on the box in degrees Kelvin (K), refers to how the light color appears to a human eye. It is an aesthetic measurement. Bright bluish white? Cooler white? Yellowish? Reddish? Color temperature, K, was not devised for figuring out which fluorescent lamps are good for growing plants. So it is best to not think about K when thinking about plant lights. More on this later.

The amount of light useful to a plant is reported with the PAR value, Photosynthetically Active Radiation. (Radiation means the emission of energy, in the form of photons or heat.) PAR describes only the light plants can use, not the total amount of light emitted by the lamp. The total amount of light emitted by a lamp is measured in lumens. A lamp might have a high number of lumens (be very bright) but have a low PAR if the light emitted is not at the wavelengths used by chlorophyll. An example of such a lamp would be the "10,000K" LED lights sold for saltwater reef tanks. Many aquarists, including me, have found we cannot even grow algae with such a light over a freshwater tank.

PAR takes into account the color/wavelength of the light emitted, and the number of plant-useful photons emitted by the lamp. As a result of depending on more than one measured variable, PAR is a calculated number, and is not something that can be measured with a simple meter, like an old 35mm film camera light meter. There are electronics inside doing calculations. To calculate PAR one has to measure the emission spectrum of a lamp, and the photon flux at each frequency.

Plant growers care about PAR. Lamp manufacturers mostly don't care about PAR. Almost all their products are designed and sold to provide light for humans to see. As a result, lamps sold for home lighting aren't designed to grow plants, and aren't generally tested for PAR. Manufacturers of lamps sold for growing plants care about PAR. They select chemicals (phosphors) inside the tube to emit light of the proper color/wavelength for plants.

Different phosphors inside the fluorescent lamp emit different colors of light when electricity is applied to them. Our brains perceive white light when there is a mixture of colors. Fluorescent tubes contain mixtures of phosphors, chosen so the light emitted by the chemicals, viewed together, looks white to us. 6500K light is supposed to look a certain way: bright, daylight white. It is possible to use different mixtures of phosphors in the tube in different combinations to arrive at 6500K light. Some of these combinations will emit light that is better for plants to grow. So, not all 6500K fluorescent lamps will provide equivalent light for plants.

Years ago, plant hobbyists figured out, mostly by trial and error, that readily-available fluorescent tubes, intended for home lighting, gave the best results with their plants when they had a K rating of about 6500. So we have that number in our heads.

But - it is possible to make a fluorescent lamp with a different color to human eyes (different K) that could be even better for plants than a 6500K lamp. What used to be sold as a Gro-Light looked purplish, and certainly was not 6500K.

Measuring color temperature in K is mainly for fluorescent lamps. It isn't used with LED lights because their technology is different. Almost all LEDs emit many fewer colors, or wavelengths, of light than do most white fluorescent lamps. So it can be simpler just to describe the few colors/wavelengths of light emitted by LED lamps. Growers would want to choose LEDs emitting light in colors plants can use, and not choose LEDs emitting other colors (except for aesthetic reasons.) So color temperature, K, is not a good measurement for comparing different lamps for growing plants.

The person buying the electricity wants the lamp to be as efficient as possible, meaning to emit photons of just the right colors, and waste as little electricity as possible in heat formation, or emitting photons of less desirable colors. In other words, the gardener views efficiency as the highest amount of photosynthetically active radiation emitted for the amount of electricity provided.

The lumen is also a poor measure for comparing different lamps. The reported lumens measure all the photons emitted, including ones not useful to plants. A green fluorescent lamp of X lumens will be much less useful for plants than a LED array composed of specific pink and blue colors, although the LED might have only a fraction of the lumens of the green light.

And, photons not used by chlorophyll may be absorbed by other pigments in the leaves, and be converted to heat, burning the plant. So we do not want a lamp with lots of lumens but not much of the light usable by chlorophyll.

To compare lamps you need to know the different colors represented in the light emitted, and the amount of photons of each color emitted. In other words, the only good comparator is PAR. The other measurements are not useful. If you have two lamps that you know emit light at colors useful to plants, the lumen may be a useful approximation. But PAR will always provide a better comparison.

When there were few choices of lamps for growing plants, gardeners figured out that 6500K fluorescent lights were OK, and brighter to the eye was better. But now that we have LED lamps emitting specific wavelengths/colors of light, we can be more picky. Discussions now center around how many photons of the correct color can be emitted per lamp; how much electricity is wasted heating the lamp, rather than producing photons; and how much the lamps cost per useful photons emitted. It is still very confusing for hobbyists, partly because it involves physics, but also because few manufacturers care about growing plants. That market is so much smaller than lighting for human vision that few companies are focusing on producing great plant lamps.

It seems probably that, before long, LEDs designed for plants will be easier and cheaper to use than any alternatives. By that, I mean you will be able to buy something at a store or the Internet that was designed to grow plants, open the box, plug it in, and grow your plants; and, the LEDs will be cheaper to buy, and use less electricity. You will want to pitch your old fluorescent setup, even though it still works. But we aren't there yet.

So, if your electricity is really expensive, or you don't have any light setup now, or you enjoy putting electronics together, definitely go with LEDs. You will need to do a fair amount of reading. The rest of us will probably stick with what we have been using a while longer, or until lamps for our technology stop being manufactured.

[naoki]

Jacob, I got a bit of time to check specs of CLU048, and here are a couple options I would try for your size.

Here is an option of high-efficiency, dimmable driver:
data sheet

1x HLG-120H-C1050A or B (A or B are different dimmer option) + 4x Generation5 CLU048-1212C4-353M2K1 (Generation 5, 3500K, CRI80, data sheet).

About $50 for the driver, and $13 x 4 for the COB in the US.

At full power (Tc=50C), each COB get Vf=35V, 36.8W, 5803lm. The PAR efficiency is respectable 2.27 micromol/J (pretty close to expensive Fluence grow light). But I'm sure this is too much light (I'm assuming that your wall is reflective). So you can dim down it to 700mA, then you get 2.41micomol/J. Note that this combination uses serial connections, so you have to deal with high voltage, so be careful.

I think this is better than 2x CLU048-1825 because you'll get more even coverage with 4x smaller ones (1212). BTW, the 2nd number in the model number represents the number of dies. so 1825 has 18 diodes in series and 25 loops in parallel (total of 18 * 25 diodes per COB). So 18xx requires higher Vf (about 50V).

---------- Post added at 04:12 PM ---------- Previous post was at 03:23 PM ----------

Good summary, Estacion. But a couple comments/correction if you don't mind.

Originally Posted by estación seca

The "nutritive value" of the photons depends on their color, or wavelength, which words mean almost the same thing. We want to give plants sufficient numbers of photons of the correct color. Photons have differing amounts of energy. Our eye sees this as color. Lowest-energy visible light is red. Progressing up the spectrum, photons carry more energy: (lowest) red > orange > yellow > green > blue > indigo > violet (highest.) There is light we cannot see, of lower energy (infra-red) and of higher energy (ultra-violet, to X-rays and beyond) than visible light.

Good analogy, but bringing in energy level of each photon could cause misunderstanding. What you said is correct, but it is irrelevant for photosynthesis (PS). PS is the quantum process. So the number of photons is relevant. To be honest, this is still a bit puzzling to me.

Originally Posted by estación seca

Chlorophyll is best at catching red and blue photons. It reflects green photons, which is why plants appear green. Red and blue crumbs / photons are useful for plants. Other colors are less useful. Green is just about useless.

This kind of statements cause the wide-spread misunderstanding of the meaning of the absorption spectrum. Absorption is simply a part of photosynthetic efficiency. So do not focus on the chlorophyll absorption spectrum. Photosynthetic action spectrum (aka Photosynthetic Relative Quantum Efficiency, RQE or McCree's curve) is a bit more informative (even though there are pit-falls). PLEASE take a look at the red graph next to Yield Photon Flux section in this wikipedia, and compare it with chlorophyll ABSORPTION spectrum (top of the page). Yes red is most efficient, but green is nearly as efficient as blue. This curve is probably different for orchids, but it is much better representation than Chl. absorption curve.

Originally Posted by estación seca

Color temperature, K, was not devised for figuring out which fluorescent lamps are good for growing plants. So it is best to not think about K when thinking about plant lights. More on this later.

If you are trying to squeeze out the performance, K does influence. Yes, K doesn't tell the whole story at all. However, for decent light, the manufacture publishes the emission spectra for different K. From this, you can get lots of information out of it (e.g. calculate PPF, or get the ration of R/FR etc). Different K influences the physiology of plants. It is not well studied for orchids, so it is rather dangerous to generalize from other plant studies. But it is not irrelevant.

Originally Posted by estación seca

The amount of light useful to a plant is reported with the PAR value, Photosynthetically Active Radiation. (Radiation means the emission of energy, in the form of photons or heat.) PAR describes only the light plants can use, not the total amount of light emitted by the lamp. The total amount of light emitted by a lamp is measured in lumens. A lamp might have a high number of lumens (be very bright) but have a low PAR if the light emitted is not at the wavelengths used by chlorophyll. An example of such a lamp would be the "10,000K" LED lights sold for saltwater reef tanks. Many aquarists, including me, have found we cannot even grow algae with such a light over a freshwater tank.

Correct, but PPF of PAR is also an approximation. If you know the emission spectra, we can convert lumen to PPF. And many of us has been doing it for a while.

Originally Posted by estación seca

PAR takes into account the color/wavelength of the light emitted, and the number of plant-useful photons emitted by the lamp. As a result of depending on more than one measured variable, PAR is a calculated number, and is not something that can be measured with a simple meter, like an old 35mm film camera light meter. There are electronics inside doing calculations. To calculate PAR one has to measure the emission spectrum of a lamp, and the photon flux at each frequency.

It would be best if we can measure the spectrum with a spectrophotometer, but these are still too expensive. Most of us measures PPFD of PAR with quantum PAR meter. Some cheap ones don't have good response curve, but if you go up to Apogee SQ-500 level or Li-Cor (mine), you can easily measure PPFD. This will give the density at a point, and it is not same as the total output (PPF), which requires a integrative sphere. The difference is similar to lumen vs lux (or fc).

Originally Posted by estación seca

Plant growers care about PAR. Lamp manufacturers mostly don't care about PAR. Almost all their products are designed and sold to provide light for humans to see. As a result, lamps sold for home lighting aren't designed to grow plants, and aren't generally tested for PAR. Manufacturers of lamps sold for growing plants care about PAR. They select chemicals (phosphors) inside the tube to emit light of the proper color/wavelength for plants.

You are correct, they usually don't publish PAR data. But as mentioned above, we can easily calculate PAR from the digitized data sheet.

Originally Posted by estación seca

Measuring color temperature in K is mainly for fluorescent lamps. It isn't used with LED lights because their technology is different. Almost all LEDs emit many fewer colors, or wavelengths, of light than do most white fluorescent lamps. So it can be simpler just to describe the few colors/wavelengths of light emitted by LED lamps. Growers would want to choose LEDs emitting light in colors plants can use, and not choose LEDs emitting other colors (except for aesthetic reasons.) So color temperature, K, is not a good measurement for comparing different lamps for growing plants.

I think you are talking about monochromatic LEDs. K is still used for white LEDs. At this point, white LEDs can achieve higher PAR efficiency at a lower cost. That's why many newer grow light is based on white + red.

The person buying the electricity wants the lamp to be as efficient as possible, meaning to emit photons of just the right colors, and waste as little electricity as possible in heat formation, or emitting photons of less desirable colors. In other words, the gardener views efficiency as the highest amount of photosynthetically active radiation emitted for the amount of electricity provided.

Originally Posted by estación seca

The lumen is also a poor measure for comparing different lamps. The reported lumens measure all the photons emitted, including ones not useful to plants. A green fluorescent lamp of X lumens will be much less useful for plants than a LED array composed of specific pink and blue colors, although the LED might have only a fraction of the lumens of the green light.

As mentioned, with known spectra, we can convert it to PAR. If you are comparing 4000K LED against 4000K LED of another model, lumen comparison can give you the approximate result.

Originally Posted by estación seca

And, photons not used by chlorophyll may be absorbed by other pigments in the leaves, and be converted to heat, burning the plant. So we do not want a lamp with lots of lumens but not much of the light usable by chlorophyll.

Chlorophylls aren't the only pigments which captures the light and pass the energy. You might want to learn about how Photosystems work in chloroplasts.

Originally Posted by estación seca

It seems probably that, before long, LEDs designed for plants will be easier and cheaper to use than any alternatives. By that, I mean you will be able to buy something at a store or the Internet that was designed to grow plants, open the box, plug it in, and grow your plants; and, the LEDs will be cheaper to buy, and use less electricity. You will want to pitch your old fluorescent setup, even though it still works. But we aren't there yet.

There is no doubt that plant optimized LEDs will appear (there are conferences targeting this topic). But the tricky part is that optimum spectra for lettuce is different from orchids. Also, the optimum can change depends on the developmental stage.

[estación seca]

The detailed information is useful to have available. It is far over the heads of the vast majority of orchidists winding up on this page. They just want to know what light to buy. It will be meaningless to them. They will not read about electrical engineering, electronics assembly, Maxwell's laws, chlorophyll biochemistry, nor photosynthesis energetics.

I simplified my explanation for the great majority of orchid hobbyists, so they could grasp the overall issues.

These hobbyists want to buy something ready-to-go, and know how to compare it to what they've been using. I wrote to explain why, for them, the first of these can be done, but the second, for the average grower, is out of reach. I also wrote to explain how a hobbyist might think about what might be a good lamp for plants.

There is a clear market for products these orchid hobbyists want to buy, but the market may not be large enough to attract manufacturers' attention. My impression right now is that suitable LED orchid growing lights cost far more than most hobbyists are willing to spend. Future cost savings from decreased electrical consumption is not any more on the minds of most people than is compounding interest in pension accounts.

There is also a market space for people to do the testing manufacturers won't do, and report the results in something most hobbyists can understand. Hobbyist want something that looks like this:

If you used this .... Then buy this ....
One CFL 6500K 150 Watt equivalent......
<a particular model number of LED lamp from a particular manufacturer>
A bank of 4 x 48" T5 HO 6500 K...
<a particular model number of LED lamp from a particular manufacturer>

And so forth and so on.

[Lamda]

@naoki and @estación seca, I think you are both contributing very valid points.

I have been reading across reef/dart frog and orchid boards to try to get a better understanding of the current LED tech available as I have recently built a vivarium with dart frogs but also grow orchids and have planted many in the enclosure.

I did want to point out that early in this thread, @stonedragonfarms noted using reef LED systems successfully in growing and blooming orchids. This is exciting for me. Why? I keep reading all these posts on how bad these reef lights are for regular plants - 10-12K Kelvin not good! So I shouldn't even look at these, right? Especially since reef lights are usually $$$$.

After a lot of reading and searching, many discussions on plant grow LEDs revolve around "efficiency". It's true, you will get the best "efficiency" if you can get a fixture that outputs only the spectrum that the plants need for full growth. The blues, reds and even a little green - depending on the plant being grown. This makes sense - we're only expending energy generating the wavelengths/intensities needed for growth and this also reduces building cost of the light.

BUT it seems many growers forget that we like to enjoy our orchids as well - usually where they are being grown and flowered. And that magneta looking red/blue light just doesn't cut it unless it's for a dedicated grow room. So efficiency isn't the ONLY metric - for me at least.

This brings me back to the reef lights - which are traditionally very expensive. That is, unless you look into these "black box reef LED" fixtures. You can get one fullfilled by Amazon for $90-140USD. These have variable dimmers on different spectrums. The blues can be dialed down or completely turned off and they also have a range of 6500K and 14000K "full spectrum" white LEDs built in. They usually thrown in a few red and green LED as well. Note - I'm not quoting the wavelengths - they can vary a little from unit to unit. The buyer should check to make sure there's the appropriate mix.

Here's an example of one with many buyer's reviews:
https://www.amazon.com/Galaxyhydro-5...s=aquarium+led

While these aren't the "most efficient" fixtures one can buy, they allow some customization of the spectrum at low cost through the dimmers, and for the more adventurous DIYers, direct swapping out of individual LEDs. I just purchased one that has 55X3W LEDs. From what I can tell, this will be more than enough for my 18" cube vivarium. (They are not driven at full 3W for better efficiency, so it's not a true 165W.)

I'll lose available "wattage" from the blue LEDs that I'll be turning down low with the dimmer - so definitely need to buy a higher "wattage" fixture than needed, but at $140 for a wifi controlled fixture with built in sun rise, sunset and moonlight timer, it seems to be the best bang for my dollar for a light that will look good to my eyes and hopefully also grow orchids well.

I would love to hear more from @stonedragonfarms and their experiences.