Self-made LED panel general discussion

Hi everyone!

In order to keep things tidy and not to pollute MajorCoco's journal, I thought it might be a good idea to have a thread to discuss on DIY LED growlights.
So, a few subjects to debate:

• White vs blue LEDs to cover the 450ish nanometers part of the spectrum

Most commercial panels use mostly red & blue light to cover the highest photosynthetic rates, but some recent DIYers replaced the blues with whites with good success (example here:https://www.opengrow.com/topic/44537-50w-high-tech-diy-led-lamp-test/); the blue & red combo seems to still have the record for yield per watt (http://www.ledgrow.eu/)
What do you guys think ? Personally, I really like the idea of white LEDs, part of this being because it is easier to see the plants under white light than it is under red light

• LED brands

The standard in the industry seem to be the China made bridgelux chips. What are they really worth against Cree's, Philips' or Osram's ?
Some eBay sellers advertise that the output luminous flux can be as high as 50lm at 660nm for the 1w chips. A little bit of math means that they can output 1200mW. If the test is done at 350mA, that means the LEDs can actually output 1.55W of light per watt consumed. This is obviously not possible, but even at 700mA, that would mean 780mW output per consumed watt; where the highest quality bins of Osram Oslon SSL only output 520mW/W (Luxeon Rebel : 460mW/W). Who's to be trusted ? How do the Chinese LEDs compare to Osrams and Philips ?

• Cooling

Cooling numerous high power LEDs without having to spend a lot of money can be a tricky challenge. I thought about computer cooling; some graphics cards now dissipate around 250W of heat and can be efficiently cooled via -rather cheap- heatsinks. The best part about them is that they are easy to find and all types of performance are available. The only real downside I found is that the contact surface is usually rather small; it seems to be better suited for single very high power chips than for arrays of LEDs.

As I plan to build my own panel, I would be glad to discuss those subjects

See ya around!

PS: please accept my apologies for my poor English, I'm not a native speaker

English is good and I'm on board with the topic. For those that know me, I'm a keen bean on building my own panels. Just ordered some multichip pcbs looking at making a grouped version for side lighting.

Thank you

I've been looking for info on white LEDs and I've come up with some conclusions. Some people say that it's better to use whites instead of reds or blues because whites have better efficiency. This does not appear to be true. There is something very confusing when comparing LED chips, especially when doing it in the perspective to build the best grow light: white LEDs have more lumens per watt than reds and blues. The problem is that the lumen is not a good unit to compare lights when the colors are different because it is a unit measuring the intensity of light as the human eye can see it.
So I think that looking at the radiant flux is much more relevant as it is closer to the actual amount of photons that are emitted by the LEDs. And guess what I found out... When comparing radiant flux of different LED colors, the white LEDs are not more efficient than the blue ones or the red ones. Most manufacturers don't provide the radiant flux of their white LEDs, so I had to do a bit of math to figure it out when only having the luminous flux (lumens) available. For the Osram Oslon SSL Pure white LED, for a luminous flux of 119 lumens, I found a radiant flux of about 433mW (at 350mA). The LED has a forward voltage of 3.2v which means that the radiant efficacy is around 387mW/W. The Oslon SSL Hyper Red (660nm, obviously from the same manufacturer) is rated for 386mW/W and the Oslon SSL Deep Blue (450nm) is rated for 460mW/W.
This has led me to believe that white LEDs are not more efficient than red or blue LEDs. In fact the most efficient LEDs are the blue ones which is no big surprise as the chemical used in most modern LEDs seem to emit blue light (I am no expert in that domain though). Unfortunately, the plants do not have use for large amounts of blue lights so efficiency alone doesn't mean much.

In conclusion to this short post, I would say that in my opinion, nothing really justifies the use of white LEDs instead of blue or reds; aside from covering a larger spectrum (which no one really has really proven that it is useful) they seem to have some of the drawbacks of HID lights (emitting light that cannot be used by plants) and not take full profit of the capability of LEDs to emit in very specific wavelengths.

As I am expert in none of the above subjects and only used basic logic to come to these conclusions, I'd be happy to be criticized by better advised individuals

Haha your findings I would agree with. It's not about lumens but par. Par is what plants eat to live on. The rest is just for us to take a better look

I think that PAR is watts per square meter of irradiance within the 400nm-700nm range so comparing radiant flux (watts) and comparing PAR is basically the same thing. In both cases, it is definitely very different from lumens.

Good info patrik...........+rep......I see many growers/manufacturers adding cool/warm white leds now, even full white, who knows what's the right spectrum anymore. It's gonna take some experimental grows(that are being done here/other forums) to show us what truly works IMO....

Experimenting surely is the best way to compare; there might always be a gap between theory and practice. But so far the best results I've seen - pure efficiency wise - have been made by "red & blue" lights. Well, anyway I do not consider that as a solid proof because most of the LED grow lights out there are "red & blue" and thus, with more people using them it is more likely to find the best results. This is why I still keep an eye on grows made with white LEDs, it is not unlikely that a good surprise could happen.

Well i'm doing I test with a panel you may know, it will be started as soon as my friend drops off his "freebies" from Attitude that he doesn't want...... ha .......i'm quite fond of this little panel.......the show will start soon

FROM KNNA THREAD, SOMETHING IMPORTANT ABOUT THE PHOTONS

Photosynthesis is tightly linked with total amount of photons absorbed. This concept is the base of all, and it should be clear for any grower. So im going to analyze it deeper:

-Amount of photons. Not of watts, or lm. Plants use photons, so the number of photons is the essential figure to consider. The more the photons which reach the plant, the better (up to a limit).

Its important to note that same energy (for example 1 watt) of blue (450nm) have 33% less photons than of red ones (670nm) (450/670=0.67 : as noted before, energy carried by a photon is inversely proportional to its wl) if we take the amount of red photons as base. If we take the amount of blue photons as reference, then 1 watt of red ones carries 49%, near half, more photons (670/450=149). So very often, producing as more red photons possible is the most effective way of using artificial light for growing plants (if the efficiency of producing 1 watt of each are similar).

This is very good to know, thank you for the input. Reading this really supports the idea that using red LEDs for growing is the way to go.
I read some of Knna's topic but they were old ( 2007-2008 ). Back then, white and blue LEDs were a lot more efficient than red ones, so using whites instead of reds made sense. Now, as I found in one of the above posts, red LEDs are as efficient as whites and just a little bit less than blues so using whites now makes a lot less sense.

Now if we decide to compare whites to blues (to cover the left hand side of the spectrum), the blues are more powerful in pure radiant output but according to what you just wrote, the whites can still emit more photons because they cover a spectrum which is longer in terms of average wavelength. This, however, does not necessarily mean that they will be more efficient towards the plants because a lot of the "warm" part of white LEDs is in the 550 nm range, where the photosynthetic action is the lowest.

Anyway, I cannot tell whether blue + red or white + red would outperform the other but I think if one outperforms the other it would be by a very small difference. I should definitely try myself but even if I actually try to compare both solutions, it would take an insane amount of attempts to have a reliable result (damn you, statistics). So unless a group of passionate people with high means decide to do the comparison, I guess we will never know for sure
There is another fact to consider which has been mentioned by knna and which is that the results may vary from a strain to another... All in all: very complicated to have one truth.

I haven't read through all of this yet but wanted to toss in that Bridgelux are American made in CA and CREE are the Chinese made LEDs - American company but they moved their manufacturing overseas. Or at least that's the last I heard.

When it comes to comparing light when it comes to plants it's a bit tricky; Neither PAR nor Lumen are going to give you the whole picture.

Say you have 100w of 450nm light. All of the 100w of light will be PAR but that doesn't mean all 100w of the light will be used by the plants. You see plants have different chlorophyll and caratenoids for different spectrum and they have a cap - a maximum limit of how much light in a particular spectrum they can absorb at any particular time. With a narrow narrow spectrum you very quickly hit the maximum absorption levels of those few matching photosynthetic pigments, leaving the rest of the pigments with nothing to do meaning the surplus light is wasted.

White light is not narrow spectrum and does not have these problems. Additionally white LEDs are much more efficient when it comes to flux/w which helps offset that they are a slightly lower % PAR compared to total radiation. And although white LEDs have a broad spectrum the majority of it is still in the PAR range unlike MH and HPS. Aside from the small bit of green (which is still technically PAR light according to a meter) the majority of light emitted by white LEDs is usable.

Well, going over the plant's absorption cap can easily be avoided by distributing the light more evenly (instead of using a small fixture, we should use a large one; about the size of the growing surface). Moreover, while it was true a few years ago, I don't think that white LEDs are much more efficient than reds or blues anymore, as I stated before. If you want detail about my calculations, I can give you my full procedure in a latter post. Basically, I use the two info available on white LEDs specsheets: total luminous flux (in lumen) and relative spectral irradiance curve. I also have the ISO-10527-2007 specification at my disposal in order to have accurate radiant flux@wl => luminous flux conversions.
The results I found make sense but it is true that I cannot be sure. I wish the manufacturers provided the total radiant output of their LEDs because I believe it is a primordial value when it comes to electronic efficiency.

I think you should probably go by the numbers released by the manufactures which do state a pretty significant difference in efficiency. I'll upload the tables direct from CREE and Bridgelux once they fix this new forum bug. I can't post any pictures till they fix that.

I would indeed be very interested to see that, can you tell me where I can find this info ? Or maybe you could upload the images to http://imgur.com/ and paste the links here, it might work.

Won't let me attach a linked picture... lame. Most of the info can be found direclty from CREEs website as well as the BridgeLux website. Looking at the new XRE they are deffinately getting closer. Much much better than their older diodes.

Here's the specs for CREEs XREs. http://www.cree.com/~/media/Files/Cree/LED Components and Modules/XLamp/Data and Binning/XLamp7090XRE.pdf

Here are the XPEs which contain the red - http://www.cree.com/~/media/Files/Cree/LED Components and Modules/XLamp/Data and Binning/XLampXPE.pdf

100-122 for the white and 45-73 for the red.

There is no mention of radiant efficiency and the figures you are quoting are those of the luminous flux, which is not relevant in the grow lighting context. As I mentioned before, the luminous flux is the intensity of light as perceived by the human eye. Basically it is the radiant flux (power output, in watts) multiplied by a constant of 683 lm/w and then multiplied by luminous efficiency of the emitted wavelength which is given by the following black curve:


What that curve means is if we have two light emitters with the same power output -let's say 500 milliwatts- and the same efficiency, but one emitting at 555nm and the other one emitting at 625nm, the first one would be rated at 541.5 lumens and the other one at only 219.2 lumens. What I'm trying to say is that more lumens doesn't mean a higher efficiency.
So, with white LEDs, the lumen output is high because there is a lot of radiance in the 525-600nm range but that does not mean that they are more efficient; in terms of PAR (watt per m²) or in terms of photons emitted (cf an earlier post about photons).

I swear those one of those pages is in mW flux - but even that isn't going to help you much more than lumen rating when most all light energy for both colors is PAR. You're looking for total umol? Still won't help you compare them in the context of plants... We need someonr to compare PAR and lumen readings for all of these so we can't just adjust the provided tables by the appropriate ratio - but even then we're still no closer to mathimatically settling the argument. lol

Ah, I see what you're saying. Very true and great point. I must admit, although I understand it better than some lighting isn't my field of study, plant biology is.

If they equivalencies are equal, and I see how they might be, I still think the plant would be able to use a far larger portion of the light from the white compared to red provided enough light was being offered. In minimal lighting the Red might perform better, but only in minimal lighting.

What do you think explains the blue LEDs emitting less lumen than the reds? Shouldn't it be the other way around? And green blows the two away while being in the middle... nvm, that's likely just the voltage difference between the colors.