Building LED lights from facts, no theories

[oldmac]

This is something I mentioned before on another LED thread, maybe you have some thoughts about.....

I would like to see a multi color high power (10watts?) LED, (LedEngin?)
that would have instead of Red Amber Green Blue two Reds, one 660 one 627, one Blue and a forth color (cyan?) all together in one LED.

A complete grow light in one 10w LED.

I know that a lot of gains have been made with high power whites, just because that's what is marketable. Horticulture is just a small niche market, especially in comparisons to general lighting needs. Somehow we need to do better then just get trickle down technology.

[RackitMan]

Looks like there may very well be a new(er) player: induction fluorescents. The watt/lumen outperforms the best T5s and lamp life is 5 to 10 times as long. My big concern is they are very likely way too expensive for the hobbyist. Perhaps in a few years this technology will become the default and prices will drop accordingly. I requested a quote.

I have no affiliation, but here is their ad copy:

Everlast Induction Lighting

Unmatched 100,000 hour rated bulb life!
Replace Metal Halides 1-for-1
Reduces energy costs up to 60%
Lowest lifetime cost of any light source
5/10 Zero Maintenance Warranty Standard
6/10 Zero Maintenance Conditioned Spaces Warranty
Excellent Lumen Maintenance, 70% for 10 yrs
Universal voltage standard 120/208/220/277
Unlimited On/off cycles with Instant Re-Strike
Operates 800° cooler than HID sources
Balanced 5000 kelvin 90+ CRI output
Low temperature starts to (-30c)
Dimming options for all fixtures How EverLast® Technology Works

Profiled heavy-duty extruded aluminum housing and crystal clear tempered glass lens. Features an anodized aluminum reflector and gas-tight silicon rubber seal made for wet locations. Ideal for retail, manufacturing, outdoors, wet or dusty locations. Flexible mounting brackets allow for this fixture to be suspended or wall mounted with angled adjustment. Available in 100, 150, 200, and 400 watt versions. 100 watt emergency ballasted for power outages is also available.

[redline]

OldMac,
Here are a couple of pics of what I am working currently on, whenever I can find the time. It is an anti-brick design. I have moderately high hopes for it but I am really having difficulty finding time to work on it.
The metal work is very labor intensive. My next step is to make the control panel.

[knna]

Redline,

what material have you used on those heatsinks? Are they of extruded anodized aluminium? What size are they? How much watts of LED are you going to install on each?

I too think that brick design is the worst possible way of doing LED grow lamps.

Right now i ordered a batch of extruded heatsink bars (13.4cm wide, about 5 1/2") to do LED modules that may be placed either on top, using an frame like yours, or hanged between plants.

RackitMan, induction fluorescents was released at the beggining of the decade. It was a very promisory technology, but manufacturers had been unable to offer it at competitive prices. And they have a reliability problem with the induction units: while the bulb itself may last very long (60-100Kh), the inductor units often fail between 25-30Kh. So it nevers gets a decent share on sales, and prices continues high. And with 150 lm/w LED expected on 1-2 years, nobody is going to invest more on induction fluorescents development.

[oldmac]

RackitMan,

Yeah I saw those induction tubes a few years back. I discounted thier use because they where only available in the same light range as MH, not really good for blooming. And a good producing bloom bulb is what we need to get to, we already know that in terms of dgrees Kelvin we would like a color temp of about 2,500 for MJ, that is why I was excited about finding 2,700K T5 tubes finally. ( Instead of the usual 3,000K)

Damn knna, I'm pissed at you! You have gotten me reading spec charts and science papers again, and going crazy in my head how to build a perfect LED light to meet my coming needs. Really fell back into the too much "theory" and too much thinking. But late yesterday I got some aluminum channel to build with, and started to layout my T5s when it struck me.

I need to produce LED strips of about 4', using Red only. Maybe a combo of Cree and some 660 from somewhere else. Each 4' strip could be made of aluminum extrusion, with a heatsink profile. My current design for the T5s is 4 tubes at VHO, all I really need to do is sanwhich them with 5 strips of LED, hopefully about 100watts per strip, depending on the ability to cool them.

That just made me think that I wish I had the two Ti SmartBars instead of the SmartLamp. The bars would fit the design of using 4' T5s better and would have a few more watts of LED. Oh well, let me see what I get with what I've got.

I actually have a 600 digi ballast and bulb, and was going to use it over one of my 4x4 trays and use the LED /now to be hy-bred over the other. Try to get direct comparisons. I think tho, because of some recent money set backs I'm going to have to wait a bit to build some more LED based lights.

Right now me and my partner (I'm slowly passing the baton to someone younger then I) are setting up a "quick and dirty" grow set-up in our new location in order to regroup some money losses, so that we have the money to finish off my dream grow set-up room(s). Plenty of room to work and grow, and systems that produce. He is a frustrated dirt farmer, does hydro to waste in pots of soilless mix. Setting up and 8'x9' room, with 48 sq' of grow space and 2- 1000w and 1- 600w HPS (was mine). Once he's underway, I'll set up a table to hold my 2- 4x4 aero/fog grow trays and get started with at least 1 LED/Ti SmartLamp hy-bred.

Oh, in running designs in my head, I thought about the Reflex tubes you mentioned. I think it could be good to do away with reflectors for the T5s in a LED strip design hy-bred....it could give more space to the heatsink aluminum. I googled them and can't find anything about them. Could you steer me to more info on them. Thnx knna. I'm not really pissed at you BTW, you just got me thinkin' and thinkin' about some of this stuff again.

[oldmac]

OldMac,
Here are a couple of pics of what I am working currently on, whenever I can find the time. It is an anti-brick design. I have moderately high hopes for it but I am really having difficulty finding time to work on it.
The metal work is very labor intensive. My next step is to make the control panel.

Redline,

I reallly like that design, each brick being height adjustable and angle adjustable. Seems it could be the answer to grow taller plants with LEDs.

When I built my LED strips last year, the Red Cree's had a Vf of 2.7v and all where within a tenth of that. So a string of 10 made an easy descion to use 28v power supply. I remember you worked with some LedEngin 660nm reds, do you recall what they actually measured? and how consistent where they?
I'm thinking, just build bigger strips (48" vs 24") with more power, and a mix of reds.

[redline]

I cannot remember exact number. But every single one of them was within .1 volt of the min (some were even outside of specs) I had designed for "typical" and had to end up squeezing in an extra LED.
I was a little dissapointed, since this would indicate they were a low bin unit, and will place you low on the watts/current curve.
I really haven't had a chance to do a decent comparison test on the 660nms.

[veggii]

hey mac thx i'm still very sick and things are looking very bad, have you seen these t5 bulbs they are the fullest spectrum i've seen
a color index 96cri !! very curious too see how well they would grow plants
BlueMaxââ??¢ Full Spectrum Fluorescent Lights

there about $11 ea or they were ,anyhow gotta go good luck

oh and 2700k hell yea!!



..

[oldmac]

Veggi;

The T5's I used for the 2' hy-bred set-up where full spectrum blubs with a very high CRI also, don't remember exactly what it was now.

I do remember they where made by Hagen and cost 19.99 per bulb (and I needed 8!)

I'm going to replace them soon with the 2700K, and see how it effects things.

I was originally hoping to end my indoor and greenhouse grow after this year, and move everything to the new location. It's very difficult to harvest the 8x16 greenhouse because of the odor, trying to dry and cure that much material. Could not let my kids or grandkids or anyone else for that matter, near my house for a few weeks. This year I only got to harvest some top colas from some of the seed mothers, then everything got stolen from the greenhouse one night. Most of the money I live on....gone. Worst part they poisined my outside dog to get to the greenhouse. Major hurt comming to three of the town's crack-heads who did this.

Anyway, my small indoor op keeps putting out enough MJ for me and all of the patients I take care of. And a couple of fast and dirty grows and me and my partner can recoup some money and build my dream indoor set up, hopefully before I pass on.

[redline]

Veggi... Sad to hear you are not doing well, hope you get better soon.

KNNA.. Each module will be about 15 watts. 12 modules total. No 660nms in the small modules, but I will be adding 2 strips 2.3 feet longstrips that will be 660nm. I will toss them in about a month before harvest.
My goal is to stay under average of 40 watt per square foot.

My primary goal is to replace a 500 watt CFL array that is used for Vegging only. If I can save 300 watts during the 6 week veg cycle, that will be a huge savings since they are running 18 hours a day.

To be honest I don't have high hopes of replacing my HPS during flowering. I am running about 50 watts per square foot for my HPS. So I will be comparing it against 40 watts per square foot LED. I think that is the best we can possibly hope for. I think the claims about 150 watt LED replacing 300 watt HPS are grossly optimistic. I have yet to see LED compete on a watt to watt basis against HPS. If I can get 40 watts LED to compete against 50 watts HPS during flowering, I will be overjoyed.

Heat sinks are extruded aluminum. I buy them from a guy on EBay who cuts larger slabs into any sizes you want. Very convenient and cheapest source I have found. I am going to try and run ducted area cooling over the top of the heat sinks, but have drilled holes to mount fans on top of each one if need be. Fans are little power hogs.

OldMac, Regarding your Red strips: Has anyone come out with a 660nm besides LED engin? Check out Luxeon K2s or the IIIs as opposed to Cree's.
They have a much wider bandwith and do a pretty fair job of delivering photon juice at 660nm. The Crees have a razor sharp bandwith around 620nm. I haven't had a chance to compare efficiencies, but imagine they are both very close.

In the past the LUXs have also been cheaper and available in better BINS. I have not worked with Crees, but from what I have read, the bare emitters are not supposed to be hand soldered. Given a choice, I will always work with an emitter instead of a star to eliminate an extra thermal barrier and save a few bucks.

I am now working on cutting the front panel to accomodate panel meters and switches. I also have to finish up two power strips with Cinch Jones plugs. The metal working is what really bogs me down, since I am working with basic tools. Lots of drilling, grinding, filing, tapping, cutting, blood,swearing etc.