Cool setup. I'll have to read more of the thread later. I'm doing a setup with mid range LED's in a grow tent. I would like to know for sure if filling in with white does anything. What is the spectral sensitivity of a par meter? What is it using to guess what a plant wants?
-
If you need help identifying a pepper, disease, or plant issue, please post in Identification.
My Hydro Habanero's: DWC with DIY LED's
- Thread starter Ictus
- Start date
Well the sensitivity depends on the meter brand and type. Lux meters are basically worthless as they try to approximate how the human eye perceives brightness.
A PAR meter on the other hand is tailored to measure PAR - Photosynthetically Active Radiation. The means ~400-700nm, which also happens to be the visible spectrum (400 being blue almost UV, 700 being far red). So PAR meters do not "guess" what a plant wants; they KNOW which range of lighting affects photosynthesis.
That said, the spectral response of any given PAR meter is far from perfect. Apogee's:
Here's how to intepret that. The red line is some hyothetical light source that emits even spectrum from 400-700, with abrupt cutoffs on either side. Obviously there is no such light source, so this graph is just to demonstrate how accurate the meter would read those spectra.
The green line is what a typical sensor would read from that hyothetical perfect PAR source. The blue line is the results apogee gets with some kind of filter that's built into the sensor.
So you can see that the Apogee meter will UNDERESTIMATE blue, and OVERESTIMATE red.
Since there is no such thing as that perfectly even spectrum, and most lights have a bit of a noisy spectrum, the errors tend to just average eachother out. So the spectral response is considerd 'good enough'. To make things even more complicated, the meter has both "Sun" and "Electric" modes, and those basically internally bias the spectral response. The manufacturer says that "electric" is better for LEDs, but I have done quite a bit of researching this myself and it appears that by "LEDs" they mean mostly the red/blue only grow kind. Folks over at Advanced Aquarist did some very scientific testing and they determined that the sunlight mode was actually more accurate for most white LEDs and most high quality broad spectrum lamps of all types.
not sure if this is the original article I read on that, but it's also very informative: http://www.advancedaquarist.com/2005/7/review
On my setup, if I read in "electric" mode, the results are about 7-10% lower (so 1000 becomes 900). But since my LEDs are pretty mixed in spectrum, I think "sun" is more accurate. Where you could get a BIG error is if you measured the PAR from pure Red LED. Those have a very narrow range somewhere around 650nm. So the apogee meter would give that LED a false high reading. On the other extreme, a pure Blue LED would give a false low reading. I don't think "elecric" mode is going to help all that much in either case.
Either way, the mode doesn't change the results THAT much to really care. 1000 vs 900 PAR really doesn't matter, since we probably can't be that accurate in the first place. We just want to get into the ballpark with lighting intensity.
There's a sensor by Licor thats slightly more accurate (the link above has the info on it).
The attractiveness of pure Blue/Red LEDs is because that's where chlorophyll A and B absorb light. But there is also research suggesting that spectra outside of those ideal ranges could also be very important to the plants somehow. I've seen some LED plant tests in pure red/blue that apparently had enough intensity.... but plants were just lying down and looking odd.
There's even FURTHER research saying that small amounts of UV could be important too. Even white LED's will not give you that, you'd need UV LEDs. Personally I would never mess with a UV LED, especially if there are optics involved, because of just how easily it is to blind yourself, or lose your night vision for several hours (blue LEDs will do that to you!).
So LEDs can never be perfect, but for me, white LEDs are the better option. I could ramble on more but it feels like I already wrote alot
A PAR meter on the other hand is tailored to measure PAR - Photosynthetically Active Radiation. The means ~400-700nm, which also happens to be the visible spectrum (400 being blue almost UV, 700 being far red). So PAR meters do not "guess" what a plant wants; they KNOW which range of lighting affects photosynthesis.
That said, the spectral response of any given PAR meter is far from perfect. Apogee's:
Here's how to intepret that. The red line is some hyothetical light source that emits even spectrum from 400-700, with abrupt cutoffs on either side. Obviously there is no such light source, so this graph is just to demonstrate how accurate the meter would read those spectra.
The green line is what a typical sensor would read from that hyothetical perfect PAR source. The blue line is the results apogee gets with some kind of filter that's built into the sensor.
So you can see that the Apogee meter will UNDERESTIMATE blue, and OVERESTIMATE red.
Since there is no such thing as that perfectly even spectrum, and most lights have a bit of a noisy spectrum, the errors tend to just average eachother out. So the spectral response is considerd 'good enough'. To make things even more complicated, the meter has both "Sun" and "Electric" modes, and those basically internally bias the spectral response. The manufacturer says that "electric" is better for LEDs, but I have done quite a bit of researching this myself and it appears that by "LEDs" they mean mostly the red/blue only grow kind. Folks over at Advanced Aquarist did some very scientific testing and they determined that the sunlight mode was actually more accurate for most white LEDs and most high quality broad spectrum lamps of all types.
not sure if this is the original article I read on that, but it's also very informative: http://www.advancedaquarist.com/2005/7/review
On my setup, if I read in "electric" mode, the results are about 7-10% lower (so 1000 becomes 900). But since my LEDs are pretty mixed in spectrum, I think "sun" is more accurate. Where you could get a BIG error is if you measured the PAR from pure Red LED. Those have a very narrow range somewhere around 650nm. So the apogee meter would give that LED a false high reading. On the other extreme, a pure Blue LED would give a false low reading. I don't think "elecric" mode is going to help all that much in either case.
Either way, the mode doesn't change the results THAT much to really care. 1000 vs 900 PAR really doesn't matter, since we probably can't be that accurate in the first place. We just want to get into the ballpark with lighting intensity.
There's a sensor by Licor thats slightly more accurate (the link above has the info on it).
The attractiveness of pure Blue/Red LEDs is because that's where chlorophyll A and B absorb light. But there is also research suggesting that spectra outside of those ideal ranges could also be very important to the plants somehow. I've seen some LED plant tests in pure red/blue that apparently had enough intensity.... but plants were just lying down and looking odd.
There's even FURTHER research saying that small amounts of UV could be important too. Even white LED's will not give you that, you'd need UV LEDs. Personally I would never mess with a UV LED, especially if there are optics involved, because of just how easily it is to blind yourself, or lose your night vision for several hours (blue LEDs will do that to you!).
So LEDs can never be perfect, but for me, white LEDs are the better option. I could ramble on more but it feels like I already wrote alot
The only real drawback to white LEDs is they emit some green. Green is the least effective color inside the PAR range.
But the amount of green they emit is very small, so I personally think white LED's match the sun's spectrum ALOT better than some blue/red grow lights.
Here's the sun's spectrum:
High in blue then steadily tapering downwards.
Here is the spetral response of a few different types of Cree LEDs:
Even if you are doing only red/blue LEDs, you should make sure and find out as much about the spectrum they give as possible. Sometimes the band they give is VERY narrow, like only 440-450. Ideally you want a wider range. That way you then have two hills instead of two sky scrapers on the graph. I don't have the research in front of me to back up the why, but apparently giving the plants too much light in only a single spectrum can do bad things to the way the plant gets rid of oxygen molecules or something to that effect. Different things besides just chlorophyll A and B also respond at different wavelengths and they prefer to be in a certain ratio with eachother, as I understand it.
But the amount of green they emit is very small, so I personally think white LED's match the sun's spectrum ALOT better than some blue/red grow lights.
Here's the sun's spectrum:
High in blue then steadily tapering downwards.
Here is the spetral response of a few different types of Cree LEDs:
Even if you are doing only red/blue LEDs, you should make sure and find out as much about the spectrum they give as possible. Sometimes the band they give is VERY narrow, like only 440-450. Ideally you want a wider range. That way you then have two hills instead of two sky scrapers on the graph. I don't have the research in front of me to back up the why, but apparently giving the plants too much light in only a single spectrum can do bad things to the way the plant gets rid of oxygen molecules or something to that effect. Different things besides just chlorophyll A and B also respond at different wavelengths and they prefer to be in a certain ratio with eachother, as I understand it.
Is the reef tank still set up and going then? If so this may in fact be part of your problem if there is one. Have you thought about the salts and other trace elements in the humidity from the tank and how it would effect the plants in the same room.
Right, so the PAR sensor is picking up green which is not what the plant needs, so while the band-gap is in the photo reactive region for plants it still doesn't have a perfect bias. Unless Im missing something and the sensor can break up the light into spectral bands. Sunlight isn't really a perfect source for plants. Just because its in the solar spectrum doesn't mean plants need it (as you mentioned green). You have me a bit worried about my lights since they all have UV diodes. Im also baffled as to how LED's produce continuous spectra white light...
dont white led's have a layer of phosphorescent material kind of like florescent bulbs?. i think this is how they emit mixed spectrum's.
Yeah I just wouldn't have expected broad spectrum. I would have expected banded, granted heavily banded.. I'll have to look deeper into it.
Yes, all white LEDs are actually blue LEDs that have a phosphor coating. Since blue is the highest energy state for the photons, it means the phosphors can absorb some energy and step down the light energy in the form of color shift. It's only possible to reduce energy, so you can't have a red light that makes white light using phosphors.
There is some energy loss of course but it's not much.
It's obvious when you look up again at the Cree white light spectrums. The big peak on the left is the Blue LED, and the hump to the right is the phosphors. The warmth could be intepreted as peak-to-hump ratio
physicsman, you are correct that all PAR sensors will measure green light to be just as effective as red or blue. AFAIK, none of them bias this out. However, there are a whole lot of folks who question how wise it is to emit the green light entirely.
There are two quick points I'd like you to consider before assuming that "blue" and "red" (which are totally arbitrary constructs of language in terms of this discussion) are the only things necessary for plants.
1) Plants do absorb SOME green light even in the major pigments:
Notice how each red and blue zone actually has TWO separate peeks. The point I alluded to in my last post, was that 90% of the "average" red/blue LEDs have such a narrow band of light that they have no chance in hell of hitting both of those peak absorption zones. You simply fail by trying too hard, unless you get it EXACTLY right!
2) There are many minor pigments besides just A and B and some of them even have their peak absorption inside the green range. Not all plants have them all apparently, and some plants can pick and choose when to deploy various pigments. Here's just a few:
So it's a bit simplistic to limit ourselves to just A and B. Plants have adapted to make the BEST possible use of the suns full spectrum. Even if the plants seem to reflect most of the green (hence they appear green), we have to wonder WHY that is the case. Maybe they need its presence in order to regulate other pigments by complex interactions we barely comprehend.
Plant Pigments that absorb light
Until we understand EXACTLY why that is, we shouldn't exclude the green light. We should err on the side of caution and at least provide SOME green light!
*edit*
I called this a 3rd point but I reduced it to more of a footnote:
The Cree LEDs have a pretty big dip in intensity between Blue and Orange. Yes they emit a good amount of green light, but very little cyan. The warm ones put out more orange and red than they do green. And the cool ones put out WAY mo blue than green. So the white LEDs aren't wasting as much light in the green range as you might assume.
There is some energy loss of course but it's not much.
It's obvious when you look up again at the Cree white light spectrums. The big peak on the left is the Blue LED, and the hump to the right is the phosphors. The warmth could be intepreted as peak-to-hump ratio
physicsman, you are correct that all PAR sensors will measure green light to be just as effective as red or blue. AFAIK, none of them bias this out. However, there are a whole lot of folks who question how wise it is to emit the green light entirely.
There are two quick points I'd like you to consider before assuming that "blue" and "red" (which are totally arbitrary constructs of language in terms of this discussion) are the only things necessary for plants.
1) Plants do absorb SOME green light even in the major pigments:
Notice how each red and blue zone actually has TWO separate peeks. The point I alluded to in my last post, was that 90% of the "average" red/blue LEDs have such a narrow band of light that they have no chance in hell of hitting both of those peak absorption zones. You simply fail by trying too hard, unless you get it EXACTLY right!
2) There are many minor pigments besides just A and B and some of them even have their peak absorption inside the green range. Not all plants have them all apparently, and some plants can pick and choose when to deploy various pigments. Here's just a few:
So it's a bit simplistic to limit ourselves to just A and B. Plants have adapted to make the BEST possible use of the suns full spectrum. Even if the plants seem to reflect most of the green (hence they appear green), we have to wonder WHY that is the case. Maybe they need its presence in order to regulate other pigments by complex interactions we barely comprehend.
Plant Pigments that absorb light
Until we understand EXACTLY why that is, we shouldn't exclude the green light. We should err on the side of caution and at least provide SOME green light!
*edit*
I called this a 3rd point but I reduced it to more of a footnote:
The Cree LEDs have a pretty big dip in intensity between Blue and Orange. Yes they emit a good amount of green light, but very little cyan. The warm ones put out more orange and red than they do green. And the cool ones put out WAY mo blue than green. So the white LEDs aren't wasting as much light in the green range as you might assume.
onefowl1,
Yes I have thought about it just didn't meantion it, since I felt I was already rambling on about way too many possible factors
But now that you mention it, the salt/humidity in my garage is a PITA. I have to take extra caution keeping my drill press and table saw from rusting out from it. I slacked last year and had to spend a long night just me and some naval jelly. Was Not sexy.
The tank will be up and running until December unfortunately. The guy I am selling it to is having his house built currently and is even having a special nook built for the tank. I will be moving out of my house right around then also.
Just for fun, here's a full tank shot of my 220gallon from about a year ago when it was thriving, before a not-so-sudden alkalinity crash wyped out the SPS and I sold most of the big LPS:
I'm really going to miss this tank but not the labor and overly hard to reach neatly tucked away electrical nonsense I put together. I built this "fishroom" so I know a thing or two about going way overboard on control
PM,
what kind of fixture are you using? If it has UV leds I doubt they are very bright for safety reasons. I don't mean to alarm you, as a typical metal halide bulb can put out a decent amount of UV (especialyl HQI) and people don't go blind. I'm worried about the 3w UV leds, as those are REALLY tiny and REALLY bright and a concentrated UV beam from them - especially with optics - could be very dangerous. But I doubt your fixture would have that.
I can tell you from personal experience you shoud wear sunglasses anytime you are goofing around with optics on any LEDs, until you get the fixture mounted. I've had dark spots where I looked into a blue LED optic that lasted a few hours.
Does it have optics?
what kind of fixture are you using? If it has UV leds I doubt they are very bright for safety reasons. I don't mean to alarm you, as a typical metal halide bulb can put out a decent amount of UV (especialyl HQI) and people don't go blind. I'm worried about the 3w UV leds, as those are REALLY tiny and REALLY bright and a concentrated UV beam from them - especially with optics - could be very dangerous. But I doubt your fixture would have that.
I can tell you from personal experience you shoud wear sunglasses anytime you are goofing around with optics on any LEDs, until you get the fixture mounted. I've had dark spots where I looked into a blue LED optic that lasted a few hours.
Does it have optics?
Good info. I have looked up most of this before though not the same depth (I didn't look into carotenoid's). Right now Im using this on some Butch T's that I just started. We'll see how it goes. I have a small one that keeps a Ghost Pepper going in my house (it is very very bushy) Even though it says '6 band' its a bit of a misnomer since it has only one of each of the 425, 380, 12k and 6k. That might be enough but I think they just use one since they're expensive and this is a mid level light. The angles on the lenses are great I can get my hands 1.5" apart without casting much of a shadow. If if could afford it I would rock this one.
Nice, it is using the 3w Cree LEDs. I definitely would avoid looking into the UV LEDs... but I did just learn something I'll paraphrase here for you.
Apparently the danger from the UV LEDs comes from the fact that they won't correctly dialate the pupil to block out the light. That LED looks like it's basically a near-UV or even a far-violet LED, so it's also putting out plenty of light in the visible spectrum... not to mention all the other LEDs which would quickly cause you to look away or close your eyes
so the most dangerous situation would be testing a single pure UV LED, since you could look right at it and your eye could get way too might light and not react quickly enough.
I'd love to see what kind of PAR your fixture puts out. It's probably way more than mine even though it has no optics. No optics also equals more spread or better coverage. Everything is a trade off
That 2nd link you posted is BEASTLY lookin. But before spending that kind of money, I'd certainly want to know exactly what Bin # the manufacturer is using. I have a suspicion they would use cheaper bins which might help make the fixture cheaper. Wiring that many LEDs by hand would be no joke, but you could probably produce a fixture that puts out way more light doing it yourself. I'm still not sure if I'm up to that size of an LED build yet.
Apparently the danger from the UV LEDs comes from the fact that they won't correctly dialate the pupil to block out the light. That LED looks like it's basically a near-UV or even a far-violet LED, so it's also putting out plenty of light in the visible spectrum... not to mention all the other LEDs which would quickly cause you to look away or close your eyes
so the most dangerous situation would be testing a single pure UV LED, since you could look right at it and your eye could get way too might light and not react quickly enough.
I'd love to see what kind of PAR your fixture puts out. It's probably way more than mine even though it has no optics. No optics also equals more spread or better coverage. Everything is a trade off
That 2nd link you posted is BEASTLY lookin. But before spending that kind of money, I'd certainly want to know exactly what Bin # the manufacturer is using. I have a suspicion they would use cheaper bins which might help make the fixture cheaper. Wiring that many LEDs by hand would be no joke, but you could probably produce a fixture that puts out way more light doing it yourself. I'm still not sure if I'm up to that size of an LED build yet.
I don't know how geeky you are but I'll delve into my thoughts about the white light. The spectrum of atoms and molecules are dictated by the number of atomic transitions of a particular species and in molecules there is a complex interplay between the atoms in the molecule that allow for many many more spectral lines. If you look at a compact fluorescent the majority of the light is discretely banded along the emission lines of mercury (naturally) and there is a broad overlapped phosphorescent spectrum. The only effect that know of that would blur these lines is Doppler broadening which I guess means that the the LED's are really hot in a small area.
whats that laptop for?
also that controller thing next to it... specifically the peristaltic pumps it uses, do you have any problems with them? i see those on ebay, they are cheep and i plan on using two of them with a ph controller.
It has optics. They are built onto the LED chip. They range from a 120 degree lens to 60 degrees. The UV LED is pure UV and looks like a dead chip (i.e. no visible light).
That laptop was there to interface with the controller, which could control and monitor ph and a few other things.
The dosing pumps were used for three different things:
1) Auto topoff of RO water. Very important to keep a constant salinity level.
2) pH Control via Sodium Carbonate additions.
3) Ca blind dosing via timed Calcium additions
I ran into trouble when during a particularly stressful week of crunch at work, I neglected to refill my 1 gallon Alkalinity and Calcium jugs. So the "awesome failproof automated system" was dosing air and sadly wondering why the ph was never increasing.
With so many big stone corals, they can suck down alkalinity like crazy!! By the time I remembered to check the tanks, it was too late, necrosis had set in for the SPS.
PM, that's some crazy stuff Definitely interested in all things lighting and electronics but I have to admit some of the atomic spectra stuff can go over my head until I really spend some time with it.
Hmm wonder what kind of Fringe device is required to generate a Doppler broadening field? What about the prismatic effect? Obviously that's a terrible idea because of transmission losses and you probably couldn't really direct and control the light without some crazy 4d looking schematic.
The dosing pumps were used for three different things:
1) Auto topoff of RO water. Very important to keep a constant salinity level.
2) pH Control via Sodium Carbonate additions.
3) Ca blind dosing via timed Calcium additions
I ran into trouble when during a particularly stressful week of crunch at work, I neglected to refill my 1 gallon Alkalinity and Calcium jugs. So the "awesome failproof automated system" was dosing air and sadly wondering why the ph was never increasing.
With so many big stone corals, they can suck down alkalinity like crazy!! By the time I remembered to check the tanks, it was too late, necrosis had set in for the SPS.
PM, that's some crazy stuff
Definitely interested in all things lighting and electronics but I have to admit some of the atomic spectra stuff can go over my head until I really spend some time with it. Hmm wonder what kind of Fringe device is required to generate a Doppler broadening field? What about the prismatic effect? Obviously that's a terrible idea because of transmission losses and you probably couldn't really direct and control the light without some crazy 4d looking schematic.
yea i mean the pumps themselves, no doubt the controller is good, but do the pumps them self work well.
this is what im referring to.
http://www.ebay.com/itm/Aquarium-Peristaltic-dosing-Pump-Head-motor-/280736334672?pt=LH_DefaultDomain_0&hash=item415d308350
im going to wire them into a 12vdc power supply and let my ph controller control them with its internal relays.
edit: actually from that profilux website, it looks like it might not be the same pump.
Doppler broadening is the broadening of spectral lines do the temperature of the stuff emitting the light. Temperature translates to atoms and molecules jiggling around. The hotter something is the faster the atoms are bumping around. In analog to sound waves when the atom/molecule is moving toward you and emits a photon the wavelength is decreased and the color is bluer and if its moving away from you the color is redder. Since the atoms/molecules are going in random directions you see doppler shifting in both directions ergo a broad peak instead of a sharp one. And, since the largest Doppler shift is coming from the fast atoms/molecules and they only move fast if they are hot you can tell how hot something is by how broad the spectral peak is. Judging by the spectral distribution of some of these CREE 3W's I'd say that they are cooking (in a very small volume).
Yea I've never had any problems, they do what they're told. I've never really heard of anybody else having a problem with paristaltic pumps either as long as you don't over work them. Far more common are issues with whatever controls the pump... be it a float switch or pH controller.
PM,
What are the 6 'violet' lights in the photo? I don't see any blue or violet mentioned... just UV on those stats.
Sorry I saw those 6 and just assumed it was the "UV" the specs referred to. Now I'm wondering what those are...
I see it only specifically says the HP White LEDs are Crees. I wonder who makes the other LEDs.
Keep in mind that all Cree LEDs already have a 120degre spread, and that's with the acrylic dome all of them have. You said some are 60 degrees some are 120... this sounds odd to me since the cree dies themselves have a 120degree optic. maybe the non-cree leds have 60deg optics and by 120 they just mean the default Cree spread, or do the Cree HP White LEDs also have the optics?
.. hmmm I'd love to see some nice clear closeup pictures... those ebay ones are blurry
PM,
What are the 6 'violet' lights in the photo? I don't see any blue or violet mentioned... just UV on those stats.
Sorry I saw those 6 and just assumed it was the "UV" the specs referred to. Now I'm wondering what those are...
I see it only specifically says the HP White LEDs are Crees. I wonder who makes the other LEDs.
Keep in mind that all Cree LEDs already have a 120degre spread, and that's with the acrylic dome all of them have. You said some are 60 degrees some are 120... this sounds odd to me since the cree dies themselves have a 120degree optic. maybe the non-cree leds have 60deg optics and by 120 they just mean the default Cree spread, or do the Cree HP White LEDs also have the optics?
.. hmmm I'd love to see some nice clear closeup pictures... those ebay ones are blurry