New high power CREE LEDs

My home-designed LED light using soley warm white CREE LEDs CXA2011 clearly outperformed my early HTG 120w tri-color LED lamp.

No, I don't have a light meter or comparison grows, but as a longtime stoner (Woodstock) and former Silicon Valley engineer, I know a wee bit about growing and hi-tech.

I am excited and getting set to upgrade hy home light (still works perfectly and grows a potent personal stash), I am ready to upgrade to CREE's newest LEDs, the CXA2520 and 2530 (for architectural use).

Don't expect to see these in any consumer lamp, though I am convinced they are superior to red and blue LEDs for many reasons which I may go into later.

Bottom line: My current LEDs get 78 LPW when driven at half power. These new ones get 110-112 LPW for the same power; a 43% improvement and with a greater PAR than a comparable HPS.

The downside? The cost for buying the latest technology. Figure $28 for a 40w LED plus another $40 for the driver and another $20 for the heatsink. Add in the mounting and power cord, etc. and you are looking at $100 for a 40w LED.

Is it worth it? To me because I live in the desert now and have a tiny, heat-sensitive grow space that I cannot cool.

The good news, of course, will be the price reduction in coming years, and the here and now challenge to HID.

 

No, they will not be coming in specific colors, but are available in 5 different Kelvin ratings: 5K, 4K. 3.5K, 3K, and 2.7K.

Currently I use the 3K LEDs for my entire grow as they contain enough blue for seedlings and veg and sufficient red, deep red and far red for proper flowering. One could switch from cool to warm for various stages of growth, but I don't see the need.

BTW, the cost I gave you was for the raw components to build. A fully manufactured fixture would be way higher.

One main reason (besides cost) you will not see these is because of the paradigm that has already been set of using mainly red and blue narrow spectrum, fairly narrow angle component LEDs. In the early days (like maybe 6-7 years ago) I was a huge proponent of white LEDs and by-and-large was frowned upon. Now, almost all mfgs have added some white LEDs to their lamps.

I believe that plants respond best to full-spectra and evenly mixed light. When you have red and blue narrow angle LEDs, they are like little spotlights of color (dots) hitting the leaves and it is impossible to get a consistent and balanced mix.

These CREE LEDs (no, I am not affiliated with them) are arrays of thousands of tiny LEDS on a die about the size of a quarter with a wide angle 115 degree spread. While they need a seriously large heatsink, but the heat is directed away from the plants, thus they can be placed very close to your buds and leaves without damage.

Many LED lamps will bleach your plants if too close as they are almost like mini-lasers. You won't get that affect with these because of the aforementioned even light spectra mixing.

If I had the funding behind me, I am near-certain I could come up with the best lamp on the market based on my research and the results I got from my home lamp. My current design has much room for improvement and I got 1 gpw. I think using my current LEDs I could get another 20% with a properly designed reflector (I am using white painted plywood rather than specular aluminum.)

Now if I added these with 43% more output, I believe I could easily approach 1.5 gpw.

One more interesting factoid is that I drive mine at about 55% (22W on a 40W LED) for longevity, but with proper cooling they can safely be driven at 90% (can't do that with most lamps), but you would need to switch to a different (and more expensive) driver.

 

A few people said they are not that impressed with 'only' 112 LPW, but here is the deal:

This is twice that of a CFL and CFLs have scattered unfocused light, it is probably more like four times.

This is 40% more output than T5s and PLLs.

What about HID? In raw numbers this is comparable to MH, CMH and low-powered HPS.

High powered HPS boasts some pretty impressive numbers (around 150 LPW), but much of that light is in the less efficient yellow range with very low blue and not much deep and far red.

The spectra output of these has a decent curve in the blue and a large curve in the red with minimal amounts of green and moderate yellow - a much better match.

The 5K version of the CREE array has 10% more lumens than the 3K version I will bet getting, so why not get the 5K? Because lumens is a measure mostly of green and yellow light and that is not what we want for our beloved plants. The point being: don't get too caught up in lumens.

 

Thanks, broseph. Mostly I stopped posting because real knowledge falls on deaf ears and I am rarely into yelling. Most people are comfortable going along with 'common' knowledge which is often far from it.


The engineer in me really wants to build this thing. My first lamp works well enough, but looks like a kludge as it was built with plywood and in three stages. Because of my cramped, hot grow space, I am limited to about 180 watts max.

This baby would give me about 40% more output from the same wattage plus another 15-20% with a better reflector and would blow commercial units away, but I have been doing the math. With a fabricated steel case and reflector and all of the cutting edge parts, I might be looking at a cost around $400+ to make it look semi-professional.

This is no biggee if I could find a way to bring it to market, but alas, I am engineer and not a salesman; a Wozniak without a Jobs.

It would be putting out about 16,000+ lumens at around 150 watts. Doesn't sound like much, but the spectrum is mostly in the correct photosynthetic range and reflective losses will be minimal unlike all non-LED light sources. I would guess it be four time better than CFL; twice as good as T5 and would easily surpass a low power HID by some 10-15%, but probably not a high power one.

Figure most current LED lamps are putting out 30-65 LPW; a 110 LPW is an enormous improvement.

*sigh*

Might give me another ounce every four months over my current setup, so what to do? What to do?

I do have a whole garden full of Peruvian Torch that I might sell to fund this project...

 

It will have one 6K LED and 4 each 3K LEDs.


So basically it will have three settings:

1. 34W of cool light for seedlings and early veg.

2. 102W - 34W of cool and 68W of warm for mid/late veg / early flowering.

3. 170W - 34W of cool and 136W of warm for mid/late flowering.

 

As far as I know, I am the only one to grow with the previous generation as they are not red or blue nor designed for horticulture and are very expensive

I got great results with a mediocre design. This design will be vastly improved and with 43% more base photonic output.

 

Will be a while. The LED holders from MOLEX are not even out yet. I can build it without them, but they are solderless plug and play so I can easily swap in replacements if need be.

I may start on fabricating the housing while I am waiting. Grow log won't be started for at least 3-4 months.

Because I cannot currently cool my grow space it runs 82-86 degrees during the day. In addition to greater light ouput, I project I should be able to drop the daytime temps about 3-5 dgrees on average which will also help yield.

Also I will be using HUGE heatsinks so no fans will be necessary.

 

BTW, I may be selling the components to my current lamp (too bulky and expensive to ship as is) at about 2/3 of my cost with 97.5% of the life left (1344 hours out of 50,000 hours) to any parties interested.

4 each LED holders, LEDs, heatsinks and power supplies/drivers for a total of 160 watts of LEDs drawing 88 actual watts. They are driven at 55% of maximum current for long life.

I will let you guys know as I get closer to my new build.

Purchase price will include technical support from me, but as they are 'plug 'n play' will probably not be necessary. Perfect for a cabinet or small closet grow.

 

OK, am taking the plunge. Numbers are rounded up to include shipping. Some of the non-specific parts are guesses.


5 each CXA2520 LEDs @ $20 = $100 (the higher powered CXA25230s are $30 each)

5 each mondo 50w heatsinks (slight overkill) - $75

$150 watt driver/supply - $70

Miscellaneous hardware: screws, hangers, eyebolts $25

Power cord $5

Thermal conducting adhesive creme (can't wait on the LED holders) - $5

Blank steel sheetmetal 24" * 24" - $25

White and black paint - $10

Fabrication - cut, punch, bend, fold, weld steel case ? No quotes yet


Output will be about 16,500 lumens in the correct PAR range @ about 155 watts.

 

I have grown very well under my current DIY lamp given the over-hot temps of up to 87F daytime. Part of the problem in addition to very poor cooling, was that I only used 88 watts of the CREE LEDs and supplemented with some CFLs and a few less effficient red and blue 10 watt LEDs.

Even so when I grew short, bushy plants I was getting close to 1GPW and much lower yields with tall, spindly varieties as would be expected - and this with horrible reflectors.

The next time I want to attempt with zero supplementation, more efficient LEDs at almost twice the power and a decent reflector.

Now a reflector for LEDs is not nearly as important as with omnidirectional lights sources such as CFLs, HID and fluoros, but still these are 115 degree angle LEDs and I want to recapture and redirect more of that lost side light. Yes, I have reflective walls, but that is no substitute for a proper reflector.

Many commercial lamps are going to narrower and narrower angle LEDs or wide angle with focusing lens. Even though LEDs run cool on the side facing your plants (you cannot place your hand on the heatsinks - ouch!), they can still easily burn them with such intense light and have 'hot and cold' spots on your leaves. Because of the wide dispersion on mine from four light sources, there is very little shadowing and the plants can grow practially right into the lights with no bleaching.

Those excited about a grow log will have to wait a while (at least four months) and it will not tell you very much without a proper comparison.

I am going to contact the local hydroponic store and see if I can do some PAR readings compared to other lamps of similar wattage, including HID, T5 and LED. This may be far more useful information than a grow log done under poor conditions as my closet will still be too hot (albeit cooler) and will still have poor ventilation.

As my situation does not allow me to modfiy the area, this will not change under my current living conditions.

 

Just noticed my e-mail from DIGI-KEY, a parts distributor:

Last week you contacted us regarding Molex's CXA25 LED Holder for CREE. Sorry it took so long for my response, but I was finally able to speak to the SSL guy at Molex and they are waiting for the tooling to be complete in their mfg plants.

As I stated, the holders add expense and are not mandatory, but then no thermal goop or soldering is necessary and changing an LED is as simple as taking out and putting back in two hold down screws.

Connecting stiff solid core wires is as simple as pushing them in the socket and releasing with a pin.

This makes for a very clean appearance and IMO is worth it. Thay are roughly $4.50 a piece.

Just not sure how long the wait will be.