'Redline's LED - DiY - Build' aka"GROW LIGHT"

[chrisman1547]

redline,
I'm very interested in building my own array. I recently bought a procyon and now I want to start building my own. I have similar questions of others. such as what exactly do I need to buy and how to exactly design the array. is there any books you recommend to read or certain websites to go to. Any help would be greatly appreciated.

[redline]

There are no books since this is the very beginning of serious homebuilding and experimentation. To be honest I don't think there is any advantage at this time to using LEDs, except for a strong desire to experiment. While they are now working, they have yet to come close to out producing HIDs ( think however, that is very close to happening with SNStealth).

LED arrays can be quite difficult to build and get working right. You need to have at least a basic electrical/electronics background and be able to do light mechanical and electronics construction. I have been spending some time on the "perfect LED" string. One of the guys is having some problems, even though he sounds like he has a pretty good grasp of electronic construction.
You might want to check out the entire string and jump on it with any questions.

There is scattered info on the internet, but you have to have some background to understand it.

[chrisman1547]

Redline,
I have a basic understanding of electricity I took 3 physics classes in college and played around with car stereos a lot, but I need to brush up on my soldering and schematic symbols and ohm's law, its been awhile. I wasn't too clear in my last post. also, thanks for the link to the other thread.
I started to read through the "perfect LED", but its rather long.
Thanks

[redline]

That's good enough, you got the basics. I can work you through it.

First decision: Scrog or regular grow ? This will make a big difference in the construction techniques.
I lean heavily towards SCROG or plants trained with a flat growing surface.

If you go SCROG you build your lights where they are spread out. If you are growing taller plants, you go with a "Brick" type design like Procyn where you got a lot of wattage in a small area.

[Projman]

I have a new design on the way that will address the major issues that I see with using LEDs for grow lights.

I don't think the leading commercial configurations are heading in the right direction. They are still using "HID thinking" trying to compact the lights and power into a small area and then hang them a couple of feet above the plants to get a bigger footprint.

Current LEDs simply do not have the penetrating power to do this. Compare 3 watts of light originating from a single point vs. 400 watts from a single point. Plus everyone is still trying to grow 2 foot to 3 foot plants with LEDS.

My current design is a modular- plug and play system. It is a 2ft x2.5 ft aluminum frame. It is designed to cover 5 to 6 square foot growing area.
It will use 16 to 20 small arrays. Each module will be individually adjustable for height and angle. Plus I will be able to change out light configurations for veg and bud.

This will allow me to closely follow the top contour of a SCROG and be able to get the lights within 2" to 6 inch" It will also allow me to make easy changes to accomodate growth.

That sounds fantastically logical. Shift the grow light paradigm. Is this new direction on paper yet? Very exciting.

[Projman]

One of the issues with wider flat panels vs. the traditional high output fixed source is that wider panels will make some plants farther from certain LEDs. Not so bad if you have all 450 and 660 LEDs but if you start sprinkling in a few far reds in the panel, some buds won't see the light.

Unless you just went with many more bulbs or the sprinkled LEDs were simply higher output to reach better / broader.

[redline]

Yeah, I got it on paper and got all the materials, started doing the metal work, just have been really short on time.
The problem with all the commericial panels is they use 5mm LEDs which are next to useless, you are far better off with CFLS.

My system is not actually a panel, but a system of small brick type units using high power LEds like the K2s and LEDengins. Each module is between 12 to 15 watts and can be individually adjusted for height and angle plus can be placed precisely at the desired location in the 2' x 2.5' grid, so you can fine tune it to the growth contour.

My only problem left to solve is the cooling. Individual fans on each heatsink are power hogs using about 3 watts per fan...plus they also complicate the construction. I am trying to work on a unit type cooling system design that only uses 1 to 3 fans to cool the 12 lamp modules.

[Projman]

Hmmm... Coolong is big. I've read that the LEDs run more "true" color, efficient and lower wattage when cool.

On a ScrOG, would it be necessary to have all of that height and angle variability? If you're training the bud and have them all the same height within 2"-3" , and the ScrOG is flat as a pancake, seems that a really flat panel would work. Assuming you have the higher intensity 5W+ LEDs that have an active projection range of 12" maybe?

I haven't grown anything yet and obviously have a limited grasp. That's why I'm asking. Hope that's OK...

[redline]

It is all still experimental, so no one really knows what is going to work best.

At least for me, it is very hard to hold a scrog grow within 2 to 3 inches. It is best to stop training before 2d week of budding and there will still be some uneven growth after that.
Getting LEDs to really work, especially without using a focusing lens, you should be within 1 to 3 inches of the top. . Making the lights adjustable is not that big a deal, so I just went ahead and incorporated it into the design.

[Projman]

Sounds great. Obviously flexability is more powerful, and I certainly see your point about hovering 1-3" over the top of the ScrOG

I see most builds are using a single LED for Blue and one for red. Consider those to be the primary wavelengths. Physics et al had done a lot of work looking into a range of secondary wavelengths. That seems very reasonable, but hard to incorporate a sprinkling of these secondary and expect them to hit all pats of the plant.

I wonder about a round panel that simply rotated?