A Curved Shaped LED Light

skyled said:

What JMD says isn't incorrect
First, 1watt (luminous power) of 550nm green give 683lm
1 watt of 450nm blue give 31lm
the human eye is not a plant... and "lumens are a way to quantify in relation how well the human eye sees it" according to JMD,
I agree
Then a 150w CFL is a 100w light, that's why it's useless to compare a 150w HPS to a 150w CFL
Moreover according to what I said, you can't compare two lights with lumens if their spectrums are different

People use HPS because it's really cheap and it works
Moreover the spectrum choice is not easy when you purchase a LED unit

Is it too hard to search "lumens" in google? I don't think so
The next time you will say "that's incorrect" you should take the time to search a little


The absorption is not the only important point,
the colour ratio is more important than absorption
as instance, the big quantity of green in the HPS spectrum induce shade avoidance symptoms,

Click to expand...

My god, Yes I understand the definition of "lumens" and It's relation to human eye sight. JMD is a smart cat no doubt, but saying LUMENS have NO RELEVANCE in indoor horticulture IS incorrect. Most growers still go by lm/per sq ft when determining light requirements why??Yes it's an old and dated form of measurement but still prevalent in the MJ industry.

I have been a big proponent in the use of 660nm as a flower booster(still am), and our discussions have been endless on the perfect led ratio for cannabis, IT WILL NEVER HAPPEN IMO. I feel we need to start thinking of standardizing horticultural leds like hid systems. A warm white (200lm+ per watt) panel perfect for flower(high peak @ 660nm) and a cool/neutral/outdoor white mix(200lm per watt) panel for veg. That's it , no gimmicks(14+ wavelenghts/etc.)

150w cfl draws 150w, so it's valid(the spd's are close too).........didn't say equivalent too http://www.amazon.com/150-Watt-Spectrum-Warranty-FE-IIIB-150W-50K/dp/B002ZPE8EW/ref=sr_1_7?s=hi&ie=UTF8&qid=1377708098&sr=1-7&keywords=150w+cfl and they are cheaper than HPS systems................still no greenhouse uses them HPS is KING and will continue to be until leds hit above 200 lm per watt white/reliability is improved/ passive cooling/ easily consumer servicable(like changing a bulb)

Oh and HPS flowers FANTASTICALLY, my highest yields ever , if you can cool it it's a great option indoors(CMH too).............but I don't wanna troll the led section with HID propaganda.

Listen we can sit here all day and study PAR curves/charts/peaks&valleys/nms/emerson effect/quantum theory/string bean studies lol/etc...GROWERS WANT RESULTS........When I worked in the crappy PA wine industry the Head viticulturist (chaddsford) and I would sit there and talk for hours during breaks about using NEW techniques/products to increase our field efficiency and maybe yields per acre......In the end MOST where never used due to his worry(job) about going under a certain tonnage per acre.... It's all about the end results!! what's the point of switching to all led farm if you yield loss is around 15-20%, the cooling savings are easily mitigated by the sale of extra product(MJ)!! High powered/efficient white leds are what's gonna change the industry IMO.......


<<<seriously, these are the yields we want, Pin's HID grows are legendary...show me one led "PAR tailored" spectrum/panel with these results....... not even the CLW ss 800 5w leds high watt panel(650w) produces anything like these 2 400w cmh bulbs. makes you wonder doesn't it?




"the colour ratio is more important than absorption"..........<< I'm not even gonna comment on that

I see,
lumens could be significant or not, it depend on what kind of light you compare
sometimes you don't know if your HPS draw more lumens than others because its radiometric efficiency is higher or if its spectrum is more green ( as instance)
lm/sq ft could be used only with HID I think, it's a bad means to measure but whitout radiometric efficiency, people couldn't use anything else

HPS do a great job, I agree . (I prefer CMH, the spectrum is pretty )
But it's not caused by lumens, If you have said to me, HPS draw 40% (35% maybe?) that's why it works fine OK but not lumens.
Or explain me in what the 200lm/w is important


When I said colour ratio is more important than absorption, The two are linked ( sometimes I think ) so it's not really true

And if you want to make a panel 200lm/w with high peak at 660nm, you've got two solution
-an excellent radiometric output 60%?
-a high peak at 550nm or near to

skyled said:

Yes, I did it.
I don't understand, what do you mean?

Click to expand...

http://www.heliospectra.com/sites/www.heliospectra.com/files/field_page_attachments/what_light_do_plants_need_2012-10-05.pdf


I am still a "all about par" guy, cause the par is the amount of energy in the light. A study like that shows that the whole spectrum is still very well used...even if not exactly in the peak zones.

but doesnt that like say green only makes sense when u made sure red and blue is like fully covered, as long as thats not completly true any "full spectrum/white/green" is a waste (except peaks of those in r/b) because r b is used more efficient? How to tell when that point is reached, when its rather better to implement more white or green in a fixture then more r and b?

I am a little odd in my reason for being in this hobby. I am not in this for max yields, but more for making things the most efficient possible so I can then build closets/lights for others locally as a side business (and there is A LOT of side business to be had if I can prove it works, which is where I am at now, the proof of concept phase). In the end I believe for most people it all comes down to $. If I can show someone who is spending $2000+ a month in electricity how to lower their bill by 75% while still maintaining good growth I think I can sell the idea to a lot of people. Even if the yield is only 75% of what they are getting with MH, they are still putting more $ in their pocket in electrical savings. Ideally I would like to get the spectrum more refined because I know LEDs can match MH in yield while using at a minimum 50% less electricity, but my goal is to do it while using 25% the electricity that MH grows use. My next grow I will be removing a few 660's and adding 505nm cyans to my DIY light to see how that works out.

In terms of importance for me I place them in this order, 1. Spectrum 2. Power Consumption and 3. PAR (Intensity). I understand things require a certain level of intensity to grow (PAR), but I don't make it a priority to think some magic # on a meter is all that matters and I should focus on reaching that number at all cost. Since my quest is different than most peoples, most will not agree with my priorities. For me I feel they are important, just like with my aquariums for which I had the same goal 6+ years ago to go from MH/T5HO run tanks to LED in my quest to spend the least amount of $ on electricity while still having identical results. To me I believe I was successful in that hobby, going from over 1300w on a 75G aquarium down to under 300w while keeping the aquarium/corals growing the same. I see no reason at all why I can't accomplish the same thing in this hobby, use a fraction of the power and have results that are identical to MH setups. If I can do it on a hobby that shoots for double the light intensity that this hobby uses I see no reason I can't do the same in this hobby.

Others don't care to get that involved with saving $ or efficiency, and that is fine, everyone has their own reasons for being in a hobby. If you don't care about being the most efficient, even 100% white LEDs will work. Until someone proves to me how 'yellow' in high intensity is critical in the over all growth of plants I see no reason to bombard a plant with extremely high levels of yellow light and I still think it is just wasted electricity. I do believe in 'full spectrum' but I also believe the yellow spectrum, while beneficial, only needs a very small amount to satisfy a plants needs. That is why out of 96 LEDs on my DIY light, I only used 6 warm whites to fill in the spectral gaps. The one spectrum I am lacking in intensity that I see plants could use more of is Cyan (505nm). Most other spectrums in the visual range I feel I have covered more than adequately for plants requirements, but I admit I am guessing at that just from past experiments I have done, and looking at my 'first grow with LEDs' wouldn't even convince me to switch. I know I have made quite a few mistakes on the first grow so just by making a few minor changes and not doing the same mistakes twice I feel the second grow will easily out perform my first grow at least 4 fold without using any more electricity. Only time will tell.

^^^^Are you saying that you can build 300w led panels that are capable of producing the same as 1200+w HID (2-600w hps!!!!!)????? because I want one

Yes we've been told that 500-600nm is "wasted" light for half a decade now, I believed it too from my studies and led's prospects at targeting these par peaks specifically; was exciting and still is. BUT we have to come to a realization that we WONT find the perfect mix for cannabis(won't even go into different strain requirements) and I know everyone thinks I'm just using a "shotgun" method promoting VERY high lumen per watt full spec white. We need to embrace leds strengths, being under a constant technical evolution and major efficacy gains every six months. White leds are being developed and improved faster due to it's general purpose applications=== $$$$$$$ to be made.

You will still save huge electrical/cooling costs (more lumens per watt and rising quickly) with ALL white/few 660nm and still get a nice working environment where pest/disease are easily identifiable====plus the bees might find the flowers too.

200lm per watt is already available (cree mk-r) but cost $$$$$$$$ ...........soon these emitters will get even better for less money, then watch out HID.

I don't buy into the whole 'lm per watt' stuff. That is a great way to measure efficiency for general household lighting, but doesn't work for spectrum specific lighting. Blue light has a very low lm per watt ratio, but extremely high in energy output, called HEV light. http://en.wikipedia.org/wiki/High-energy_visible_light What that means is even though it is very poor in lumens and lighting up a room, it actually provides more energy for photosynthesis than a 130lm per watt white LED in some applications.

To really use PAR or lumen per watt to make comparisons, they need to be the same type of light output (ie. MH, T5HO, and CFL should only be compared to white LEDs, once you try and compare them to colored LEDs everything falls apart)

Yes, IMO it is very easy to build a 300w LED light that is capable of producing the same output for growing things as 1200w of HID. My 140w LED light on my aquarium puts out more light than 1200w of HID. I actually believe the light I made now, that uses 215w, will be comparable in 'plant' output to 1800w of HID once I swap a few LEDs out and finally put the optics on the LEDs. The lenses were back ordered when I made the light and came a month into the grow so we decided to wait until the second grow to take down the light to add all the lenses. We have had a few other issues to learn about and resolve that were more important for our first grow (ventilation, nutrients, fighting PM).

PSUAGRO. said:

Sorry but that's incorrect............I understand that you look at graphs and see those chlorophyll a/b peaks and hps is FAR from being optimal, but then why is it still used by 95% of greenhouses/professionals?? The bees don't even like the color of HPS, and Dutch growers still use it!!! Because a 1000w gavita HPS will still out-yield anything.

Click to expand...

I might not have explained myself properly. What I meant was that lumen is only a useful measure of light if you want to know how bright it seems to you.
If lumen is all you want, then go for a ton of 555 nm LEDs - at this wavelenght the sensitivity of the human eye is at its peak, and you will get most lumen.

Lumen has no relation to how well it works for a plant. For plants can 1000 lumen beat 2000 lumen, if the spectrum is right.

That's why lumen is useless - which was my point Hope it makes sense now.

High lumens per Watt AND high CRI = awesomeness. Take CCT into account.

I win. Cya nerds.

tags420 said:

http://www.heliospectra.com/sites/www.heliospectra.com/files/field_page_attachments/what_light_do_plants_need_2012-10-05.pdf


I am still a "all about par" guy, cause the par is the amount of energy in the light. A study like that shows that the whole spectrum is still very well used...even if not exactly in the peak
zones.

Click to expand...

I agree with this, but it depends of the proportions.
What I mean is : does a poor quantity in orange-yellow range cause grow trouble ?

PSUAGRO. said:

^^^^Are you saying that you can build 300w led panels that are capable of producing the same as 1200+w HID (2-600w hps!!!!!)????? because I want one

Yes we've been told that 500-600nm is "wasted" light for half a decade now, I believed it too from my studies and led's prospects at targeting these par peaks specifically; was exciting and still is. BUT we have to come to a realization that we WONT find the perfect mix for cannabis(won't even go into different strain requirements) and I know everyone thinks I'm just using a "shotgun" method promoting VERY high lumen per watt full spec white. We need to embrace leds strengths, being under a constant technical evolution and major efficacy gains every six months. White leds are being developed and improved faster due to it's general purpose applications=== $$$$$$$ to be made.

You will still save huge electrical/cooling costs (more lumens per watt and rising quickly) with ALL white/few 660nm and still get a nice working environment where pest/disease are easily identifiable====plus the bees might find the flowers too.

200lm per watt is already available (cree mk-r) but cost $$$$$$$$ ...........soon these emitters will get even better for less money, then watch out HID.

Click to expand...

The MK-R (that i will use in my panels) might be 200lm/w but don't peak at 660nm, as you said, you have to add some 660nm led .
So your panel don't draw 200lm/w.
Moreover Those wich draw 200lm/w are the CW, not the WW
and the MK-R give aproximately 325 lm/w at 100% output (CW and WW only), so they need to be 60% efficiency to draw 200lm/w, You must be at 25°C with the best of the best bins to reach 60% ( for the CW, it's impossible for the WW)

CaliJoe said:

I don't buy into the whole 'lm per watt' stuff. That is a great way to measure efficiency for general household lighting, but doesn't work for spectrum specific lighting. Blue light has a very low lm per watt ratio, but extremely high in energy output, called HEV light. http://en.wikipedia.org/wiki/High-energy_visible_light What that means is even though it is very poor in lumens and lighting up a room, it actually provides more energy for photosynthesis than a 130lm per watt white LED in some applications.

To really use PAR or lumen per watt to make comparisons, they need to be the same type of light output (ie. MH, T5HO, and CFL should only be compared to white LEDs, once you try and compare them to colored LEDs everything falls apart)

Yes, IMO it is very easy to build a 300w LED light that is capable of producing the same output for growing things as 1200w of HID. My 140w LED light on my aquarium puts out more light than 1200w of HID. I actually believe the light I made now, that uses 215w, will be comparable in 'plant' output to 1800w of HID once I swap a few LEDs out and finally put the optics on the LEDs. The lenses were back ordered when I made the light and came a month into the grow so we decided to wait until the second grow to take down the light to add all the lenses. We have had a few other issues to learn about and resolve that were more important for our first grow (ventilation, nutrients, fighting PM).

Click to expand...

Can you explain to us what do you use to compare ? When you say "comparable in plant output to 1800w of HID"
I can believe that your 215w is comparable to 800w HID, but 1800w I have some doubts . But you may be right..

lax123 said:

but doesnt that like say green only makes sense when u made sure red and blue is like fully covered, as long as thats not completly true any "full spectrum/white/green" is a waste (except peaks of those in r/b) because r b is used more efficient? How to tell when that point is reached, when its rather better to implement more white or green in a fixture then more r and b?

Click to expand...

Indeed green light could have positive effects on growth, but not in any proportions.
As I said a big amount of green compared to Red and Blue cause shade avoidment symptoms.
And I don't think that a MJ plant wich grow under a canopy yeld fine

True. If you notice the same bulb lumens change as the color rendition changes. The warmer the color the lower the lumens, but the benefit of those lower lumens is greater

CaliJoe said:

I don't buy into the whole 'lm per watt' stuff. That is a great way to measure efficiency for general household lighting, but doesn't work for spectrum specific lighting. Blue light has a very low lm per watt ratio, but extremely high in energy output, called HEV light. http://en.wikipedia.org/wiki/High-energy_visible_light What that means is even though it is very poor in lumens and lighting up a room, it actually provides more energy for photosynthesis than a 130lm per watt white LED in some applications.

To really use PAR or lumen per watt to make comparisons, they need to be the same type of light output (ie. MH, T5HO, and CFL should only be compared to white LEDs, once you try and compare them to colored LEDs everything falls apart)

Yes, IMO it is very easy to build a 300w LED light that is capable of producing the same output for growing things as 1200w of HID. My 140w LED light on my aquarium puts out more light than 1200w of HID. I actually believe the light I made now, that uses 215w, will be comparable in 'plant' output to 1800w of HID once I swap a few LEDs out and finally put the optics on the LEDs. The lenses were back ordered when I made the light and came a month into the grow so we decided to wait until the second grow to take down the light to add all the lenses. We have had a few other issues to learn about and resolve that were more important for our first grow (ventilation, nutrients, fighting PM).

Click to expand...

Until such time that there is a plant sensitivity curve STANDARD that may be referenced for manufacturers to overlay there lamps specific spectral radiometric output in the 3 peak absorption points and the areas within those points, such as the DIN 5031-10 curve the debate will continue. I'll give credit to California Light Works who publishes this factoid on page two of their spec sheets in the asterisk at the bottom of the page.

http://www.californialightworks.com/info/specs-solarflare.pdf

This recent article by a major LED grow lamp manufacturer also illustrates that the lack of a standard sensitivity curve coupled with the way the light is being measured in the field is misleading growers into believing one lamp may outperform another based on the way the information is being interpreted.
http://www.gardenandgreenhouse.net/index.php?option=com_content&view=article&id=1604:efficient-gardening-stumping-light-meters-with-efficiency&catid=138:july-august-2013&Itemid=18

To show a grow lamps output in a single numeric value prior to there being a standardized sensitivity curve is never going to be relevant. A better solution to the current state of affairs in how mfg's publish radiometric output data would be to show the usable PAR power in watts that is available in these regions is a better solution than what is currently being employed.

Inda-Gro publishes each of their lamp outputs in the V-C-F watts/region format. This format allows for an honest interpretation of how a given lamp may be expected to perform if the grower knows the sensitivity curve of the species that is being grown. The universities that purchase their lights for their own use have requested spectral distribution graphs and that the radiometric output value be submitted in this format.

http://www.inda-gro.com/products.html

PetFlora said:

The warmer the color the lower the lumens, but the benefit of those lower lumens is greater

Click to expand...

I agree and disagree,
The warmer the color the lower the radiometric efficiency,
however Lm/w (luminous watt I mean) could be higher in warm white than cool white, it's the case for XP-G2 as instance :
near to 310lm/w at 100% output Cool white
near to 330lm/w at 100% output Warm white
based on cree datasheet (graphs)

Sky read this paper and you'll get a better understanding in why lumens/watt, uMole, kelvin, lux, footcandle is so unimportant in relationship to how way plants absorb light.

http://www.inda-gro.com/pdf/MeasuringPlantLight.pdf

The other major issue facing mfg's is the lack of an industry standard for a plant sensitivity curve that mfg's can design lamps to meet the absorbance regions of a specific plant. Earlier I referenced a German standard DIN 5031-10. While that is often pointed to as the reference in a specific lamps spectral radiometric output calibration. Even this DIN standard has not been widely accepted as a broad enough to satisfy a generalized sensitivity curve for a majority of terrestrial plant species. That being said the standard may ultimately end up being issued broken down into species types or even perhaps at a minimum of vegetative or flowering plants to be of real use to the gardener having to select the best lamp for their application.

While spectral distribution graphs are a nice graphic depiction of the lamps output the Y axis is typically shown as a unit of relevant intensity value and not an actual power produced by the lamp that fall within that region. I like the ability to see the actual PAR watts/region (this could also be listed as a PPF uMole/Region value) which shows how much usable power can be found in these regions. I think most mfg's will adopt this 3 part strategy unless that is they don't produce balanced radiant power within these regions which would make them less willing to show their lamps output in that format.

First let me say this thread has been an interesting read! I have been using LED's for about 3 years now and always get at least 1-GPW, My record however is 1.45-GPW and my only tricks are LED's + DWC for the grow method and I use multiple smaller wattage panels spread out vs higher wattage panels. Out of curiosity, how often does hid result in 1 gram per watt?

I believe spread to be a very important aspect with LED lighting and with my own experiments showing 4 - 50w led lights will out yield 1 - 300w led light by 40% all while using 1/3 the watts, I imagine the same could be done with HID though keep in mind we are upping the plant count when using the multiple smaller wattage light approach.

skyled said:

Can you explain to us what do you use to compare ? When you say "comparable in plant output to 1800w of HID"
I can believe that your 215w is comparable to 800w HID, but 1800w I have some doubts . But you may be right..

Click to expand...

I understand the skepticism. If I didn't know what I do about light/photosynthesis, I would be in the skeptic camp as well reading my statement.

I admit, this is completely unscientific as we don't know the exact numbers for the plant we are growing, so I am using the standard absorption graphs. I do have a PAR meter and both a 'poor mans' Spectrometer and lab grade spectrometer (at work) to break down light into spectrum/intensity in a crude manner. What I mean when I say a 215w LED matching the output of 1800w of HID comes down to PUR, not PAR. My 215w with optics should put out as much PUR as 1800w. Without optics it already beats 800w of HID in PUR output. PUR = Photosynthetic USABLE Radiation while PAR is Photosynthetic Active Radiation.

Anyone interested in making their own spectrometer cheaply, here is a link. I have the 'VHS box' type at home and it does work well for the cost.
http://publiclab.org/wiki/video-spectrometer-construction

chazbolin said:

Sky read this paper and you'll get a better understanding in why lumens/watt, uMole, kelvin, lux, footcandle is so unimportant in relationship to how way plants absorb light.

http://www.inda-gro.com/pdf/MeasuringPlantLight.pdf

Click to expand...

First, I love the end of this text
Then, I agree , I don't use lm/w Kelvin ... they are useless

I proceed like that:
-I make my spectrum, considering absorption and colours ratio ( after long enquiries )
-I search what kind of LED could result in this spectrum, and define the leds ratio
-I choose among these leds, those wich are the best radiometric efficiency

And that's all
I don't need lumens or other bullshit

CaliJoe said:

I understand the skepticism. If I didn't know what I do about light/photosynthesis, I would be in the skeptic camp as well reading my statement.

I admit, this is completely unscientific as we don't know the exact numbers for the plant we are growing, so I am using the standard absorption graphs. I do have a PAR meter and both a 'poor mans' Spectrometer and lab grade spectrometer (at work) to break down light into spectrum/intensity in a crude manner. What I mean when I say a 215w LED matching the output of 1800w of HID comes down to PUR, not PAR. My 215w with optics should put out as much PUR as 1800w. Without optics it already beats 800w of HID in PUR output. PUR = Photosynthetic USABLE Radiation while PAR is Photosynthetic Active Radiation.

Anyone interested in making their own spectrometer cheaply, here is a link. I have the 'VHS box' type at home and it does work well for the cost.
http://publiclab.org/wiki/video-spectrometer-construction

Click to expand...

I see, if you have measured this, I believe you . However your 215w LED match the 1800w HID PUR but the lighted area is smaller with leds right?
The link is Very interresting

Hey Cali! You make a valid point as to the difference between PAR and PUR in that unlike PAR, PUR is a level of actual plant absorbance that would again be species specific. Once again therein lies the problem in these sensitivity curves have to be based on plant species, native origin and phase of development. Once these sensitivity curves can be standardized than it is a relatively simple mathematical exercise to develop a single number value perhaps something like 'PAR Lumen' that would weight the the lamps overall radiometric output based on region and the relative importance of energy being radiated within those regions.

Until such time if I can tell how much radiant power the lamp produces as a watts per the 'net photosynthetic action regions' I can make a much better assessment of where the energy the lamp produces is being outputted when comparing different lamps and technologies.

In the datasheet there are the spectrums of each led
y axe : relative power(or energy) distribution
X axe : wavelength
I think that you know this, but why is it not good enough for you? with this base you could calculate what you want

Par lumen is a kink as black white

The reason why it will never happen is, because there is and probably never will(in my lifetime) be a definitive answer on what plants like best exactly. Even you say there is too much discrepancies and differences from plant to plant. You want a plant version of lumens but the human eye requirements(555nm bell curve) are a lot simpler and accepted than the multi-peaks and unknown importance of certain nm's for plants.

It is a two part system(intensity + spectrum) and cannot be put together to make a single measurement that will be anymore actuate than lumens. PAR is what matters for yield spectrum supplies the quality of growth IMO...and the way to make sure your par matters is to don't be an idiot and buy a poor spectrum...cause though there may be no ideal/perfectly accepted growth spectrum yet, nearly everyone is 100X better than hps and hit the generally accepted peaks.

A plant can only photosynthesis the photons being put out(no matter the nm's a photon is a photon...blue is more energetic but still just one photon). And PAR tells you how many photons there are equally. And unless you are buying HPS, every other lighting source manufacture uses a substantially better spectrums then hps. The reason why they never yield as much or what they are said to, is because there is not enough light energy(par) to grow to the goal. No matter how good the spectrum is.
At IG you guys know all about daily umol value...well if it is not meet than you will not yield. But you can meet that goal with a less ideal spectrum(hps) but still get the growth. And with that said a lower daily umol could be established once more spectral efficient lighting is accepted, but till then umols are still the focus.

I actually kinda like the way indagro has broken down to V,C,F. but a plant light sensor will not happen. The down side to the v,c,f, is it's just numerical way of looking at a spectral chart...you still have to know what peaks you want for certain crops...and the light could just be in that region and not on a peak. But way better than a plant light sensor for sure. I would actually like it in PAR/region and then I will accept as a good tool. Not all watts are the same as far as output. So actually you still need to look at more than just that.

So I still don't get how when a company shows you a graph of a perfect/great growth spectrum and it does 1500umol...how is that not enough or inaccurate info to make a decision. When I see a spectral graph the first thing I look at is the red and blue to see how much is in those(and what exact nm's). And then take into account the rest of of the light being put out as additional befits to complete the whole spectral process or wasted light, if too much.

And not to get off to another thing but I would really like full unit makers like led's and IG to show more foot print graphs to show where all the light is going.

Skyled...I think you are looking at the wrong numbers. The xpg2 doesn't do anything close to 300lm/w at full power. LED's are most efficient at about 350mA-700mA (2-2.5w) and the higher you drive them close to full power the lm/w drops more and more.
http://pct.cree.com/dt/index.html
Use this to see what each cree chip actually puts out at different drive currents and actual operating temps.

EDIT:5 new post while I wrote all this shit