astroastro, If I understand correctly, then the GLH units use TRUE 1W dies BUT hooked up to a 3W LED CHIP. Is this what allows them to be overdriven to 1.5-2.0W with relative safety? I guess other manufacturers have them hooked up to 1W chips instead or do they have LESS than 1W diodes hooked up to 1W chips?
If this is correct, then what is the relationship and resulting difference between a 1 W die on a 3W chip and other "lesser" combinations?
This is a great question and I am glad somebody asked it. I can see by reading these forums that people are all over the map in understanding LED technology and LED technology as it relates to growers is totally new so there is bound to be a learning curve. I am no genius, a point my wife likes to make with great regularity, but I work with these LED's everyday so even for a crusty old HID grower like me something is eventually going to sink in. I don't know crap about much else but I sit around designing with HB LED's all day long- you are free to believe whatever/ whoever you want to believe- somebody who wants to sell you something based on inflated and misleading Watt information or an old analog designer fart with failing knees that has no dog in the fight.
The LED die that I refer to IS the LED chip- they are one and the same, just different terminology. I have no idea why or how the term 'die' came about when referring to the little semiconductor chips that form the base of all modern electronics- I guess electronic guys refer to them as 'dies' and also sometimes as 'chips' (usually after the silicon die have been packaged in the little black box that people are familiar with), and folks not so involved in electronic design refer to both of them primarily as 'chips' and very rarely get an opportunity to see the naked silicon die outside the black package- this gets things confused.
The first thing everybody that is interested in LED's should understand is that the 1mm^2 LED 'chip' is the backbone of the real high radiant power applications for LED. This would be the street lights, architectural illumination, wide area lighting, etc. This is the state of the LED industry at this time. Yes, there are larger chips that can handle higher power densities, but because of productions costs and relatively low demand for these parts, it would be incredibly expensive to start an LED design using these parts, availability problems would complicate the manufacture of the fixture, and well, just because it is there does not mean it would be a great idea to base a design off for a number of reasons.
OK, so far so good. The term '1W' die ('chip', I can't seem to get off my own terminology) seems to be offending a lot of people- it shouldn't. This is where the poisoning of the market has occurred by the marketing people. Anyway, the term 1W die is really just easy vernacular for engineers to use when referring to the basic 1mm^2 die. Theoretically (not realistically in a practical application) you could drive this same die with 10W of power if you wanted, if you could somehow figure out how to keep the die cool enough so that it does not burn up. And yes, if you really could keep the P-N junction (and the electrodes in a lateral, or planar type die die on sapphire) this cool, it would putting out an ass load of light. But this is not realistic, feasible, or practical.
So the term '1W die' is applied because for most applications using conventional cooling systems and materials, this is about the power level you can expect to drive the LED at without creating all the problems you will have with an LED regarding heat. Heat is the only real enemy you have with LED technology- nearly all the problems you will ever have with an LED can usually be traced back to thermal issues, and
any LED die is whatever Wattage you want it to be (well, not exactly, but you can kind of think of this way) assuming you can keep it cool enough- no easy task.
So- most true 1W die applications will be driving their LED's at power levels between 1 and 2W. The closer you run this die at the 2W current density is the riskier things get, and the customer is bearing this risk unless you really think most of these current crop of LED sellers are really going to be around to honor their '5 year warranty' 4 years down the road when radiant power has degraded to the point where efficient photosynthesis is no longer being sustained by the light source. Even 20 lb's of aluminum heat sink and 6 fans is not going to allow you to, responsibly, drive the die at the 3W marketing figure that the LED sellers post on their specification sheets. Also, it is helpful to understand why they have this '3W' data in the die manufacturer spec sheet- this shows another engineer what the V.I characteristics of the LED die are with regards to radiant flux- it is not intended to confer that you can take this die and and run it at steady state 3W power levels. You need to understand how the LED mfr is testing this, where they get this data from- they maintain a junction temperature of 25C (impossible in a real world application) and they pulse the die at these high current levels for like 10 milliseconds, long enough to get a radiant flux reading, but not long enough to destroy the die. In a real world application the die would in fact be smoking hot at the 3W current density level and radiant flux output would be significantly less than the data you see in the spec. Heat is the killer of efficient radiant flux generation as well as the life shortener (for lack of a better term) of the LED die.
I think that if the LED sellers would have been using all true 1mm^2 dies with enough bandwidth to fully cover the photosynthetic requirements of the plant there would never have been an LED controversy in the first place. These LED's put out A LOT of light very efficiently, they are the best solution available at this time. I think it is possible to basically cut to total Wattage applied to lighting about in half (approximately) and get similar results to HID, thus doubling your g/ W calculation. But, unfortunately, there will always be the people where this is not enough and will buy into the 10W of LED equals 1000W of HID argument made by the sellers- I mean you have to be realistic about your expectations. All the other benefits of LED would still apply- much less heat load in the grow room, not replacing bulbs every 6 months, etc. etc.
So anyway- for a variety of reasons related to cost, die availability, photosynthetic bandwidth requirements, etc., most of the current crop of Chinese based LED sellers are taking very small, fractional Watt LED's (really little 'chip' LED's), placing a bunch of them on a small metal core PCB, and calling them '3W' LED's. This is totally misleading and deceptive. Why- because pound for pound, Watt for Watt, these small chipetty chip LED's do not produce light as efficiently as their big 1mm^2 cousins-
not even close. You can put a bunch of them on that little MCPCB and still they will not perform as well- certainly not with regards to the amount of power you need to feed them vs. the amount of actual light you get out. This is related to the process and materials used by the makers of these dies, as well as the fact that these small surface area of the die itself is
really difficult to extract heat from.
Why are these smaller dies so prevalent and so cheap- because they are commonly used in huge, huge, huge quantity applications like cell phone backlights (for example). If you think that this category or class of LED die can grow good plants then please suspend you iPhone over your plants and let us know how that works out for you. The ask yourself, why have people had problems with cheap LED fixtures, why are the one's using the big dies so expensive, but they work, and why do the so-called Watt' ratings that the sellers are pushing
seem to have absolutely no basis in reality with the way the light actually performs?
The marketing and sales people have won- we are talking about the wrong thing in the wrong way to correlate the Watt rating with actual performance in the grow room. My big gripe with the lot of them is that they do not post photometric data- even the HID guys do this for you. The reason no one can figure out what is really working vs. what is not is because no one has any idea what the real spectral output of these LED fixtures is. "Well- I tested this fixture and it's junk and I tested that fixture and it's good"- my question is, why? It can only be told by knowing what the spectral or 'photometric' qualities of the fixture are- wavelength content and amplitude. Then everyone in the community would know what really works and what doesn't, instead of trial and error, guessing, and a bunch of talk about Watts that never seems have any bearing on real world results.
Here is the real deal- you can't patent a portion of the electromagnetic spectrum, any idiot like me with a spectroradiometer could easily figure this information out- so, for the normal guy out there why don't the sellers just post their real photometric data (like the HID makers do) and trust the customer to make a buying decision based on the real science of photosynthesis and their own growing experience? Because as long as they keep us distracted with the Watt debate, they don't have to address the short comings of their own designs.
Long post, sorry.
Some of the growers were in th middle of their and will be starting new threads using my LED's exclusively once the current grow is done so stay tuned in!
