A Gardener’s Journey Into DIY COB LED Lighting

Chip on board or COB LEDIn an earlier post, we alluded to the fact that were going to try LED’s this year.  The shipment has come in and after a bit of DIY ingenuity, we have our new DIY chip on board (COB) LED lighting online!  In my initial testing, I can say that I am fully impressed and have high hopes for this technology.

Chip on Board (COB) LED Basics:

I spent a fair bit of time reviewing modern LED technology and was intrigued by modern full spectrum COB (or chip on board) LED’s available from companies like Bridgelux, Cree and Citizen.  Unlike the common grow LED’s that emit a blue/red/purple color, these COB’s are designed to emit a more natural colored light.  Say what you will, I know that plants have learned to live with full spectrum lighting from the sun, and these “blurple” LED lights never made sense to me.

A chip on board, or COB LED is essentially a manufactured chip that has many dozens of smaller, low-wattage LED’s.  It typically has two inputs, positive and negative, that power all the individual LED’s.  The chips are commonly rated to 100 watts, but most times running them at a lower wattage will be more electrically efficient.  A coating covers the LED’s, which also helps the manufacturer control the general light spectrum that is emitted from the LED.

This is the spectral output curve of the COB LED's I purchased

This is the spectral output curve of the COB LED’s that we purchased. You can see that it offers a wide spectrum of light.

Full spectrum essentially means that the light source will produce most frequencies of visible light.  (Everything from blues, greens, yellows and reds.)  This means the light will be more natural, like the sun or other common sources of lighting.  COB’s are measured in degrees of kelvin, which indicates how warm/red or cool/blue the light will generally be.  This is not exclusive in full spectrum lighting, so a warmer (e.g. 3500K) LED will still produce some blue light and a cool (say 6500K) LED will still produce some orange/red light.  The rating just indicates what the chip will produce “most” of.  In general, warmer lights are better for flowering/fruiting whereas cooler lights are better for vegetative growth cycles.

Most intriguing to me, COB’s are highly efficient.  What this means to the gardener in common terms is more growing power AND less electrical usage.  After some initial testing, I found that I can get the same lighting to my current T5 fixture with about half the power usage.  We’re living in one of the most expensive areas to buy electricity in America, at 20 cents per kilowatt-hour, so every watt matters.  The light will literally pay for itself in less than 2 years with year round use.

From what I read, these COB LED’s are in wide-spread use across the cannabis growing industry, but little information was publicly available for common plants and vegetables.  Even though we’re growing different plants entirely, I feel this industry does a fair job at evaluating indoor garden light technologies and their users give honest, straight forward reviews whether a light is good or not.  Everyone, and I mean everyone, raves about COB’s for growing marijuana.  When it comes to growing lights, plants use the same light spectrum whether it’s a cabbage or cannabis plant.  What works for one should work for another, except for a couple horticultural exceptions.

The DIY COB LED Kit:

We ordered our DIY COB LED as a kit from a small company called Timber Grow Lights.

I decided to take a plunge and purchased a kit from Timber Grow Lights to see how they’d do for growing garden starts.  While there are commercial COB systems available, there is much money to be saved from DIY based kits.  Basically, you supply the light’s frame and everything else is included in the DIY kit.  I liked Timber’s offering because they put “finishing touches” like heat-shrinking connections and offering everything that is needed in a single kit.  I decided this DIY effort was fully within my means as a hobbyist fabricator.

The kit that I purchased was designed to grow cannabis and it came with 3500K COB’s, better for this plant’s flowering characteristic. I decided that I would buy some different COB’s that would be more suitable for garden starts.  I ventured on over to COBKits.com and found they had Citizen COB’s in 6500K, which would be more suitable for our purposes of early seedings and vegetative growth.  In hindsight, I could have saved some money, but I also now have another four COB’s that I can put into production in the future.

In all, the kit from Timber included the following:

  • (4) Citizen 1212 COB’s
  • (4) COB Holders
  • (4) Light reflectors
  • (4) Heatsinks
  • (1) LED Driver with custom connections for power, LED’s & a dimming potentiometer
  • All the necessary wiring & hardware (except frame, see below)
  • Cool stickers & nice envelope with the invoice and welcome letter
  • Cost:  $299 + $20 shipping to Alaska

Beyond this, I had to supply a few things to get everything put together:

  • 3/4″x3/4″x1/8″ aluminum frame
  • 6-32 machine screws & nuts
  • Size 6 washers
  • Total cost:  $12

The extra COB’s I purchased were $59, but this wasn’t entirely necessary and more of a personal preference for future flexibility.

A brief note on costs.  I know I could have done the kit cheaper than I did if I designed and ordered everything from individual vendors.  In Alaska, though, shipping costs are often a major factor and ordering from 5-6 different vendors will usually eat any savings you may have had.  After calculating things, the best deal for me was the kit from Timber.  This also allowed me to get up and running in a few days instead of waiting 2-3+ weeks for Amazon and other places to actually get here.  I want to grow plants, not dink around waiting for my project to finally arrive.

Building The DIY COB LED From A Kit:

Assembling the DIY COB LED kit was fairly straight forward, requiring only minimal tools and basic fabrication skills.The DIY process wasn’t terribly difficult, but being my first one, there’s a number of things I would do differently.  I think when I have my full shop online, I might re-make the frame with better precision.  For now, though, my cobbled together effort will do.  Overall, without any guidance, it took me about five hours to assemble the frame and wire up the light.  (Spoiler, my second DIY COB light took much less time.)  The only tools required were a drill, screw driver, adjustable wrench, hex bit and a multi-meter

My first test didn’t work, so I had to break out the multimeter.  I figured out one of my connections between the COB and COB holder was not making good contact.  Easily fixed and I was off to the races.  This was actually a good experience as I learned several things about how COB’s work and how to diagnose the system if things go awry later.

After I made sure the default kit worked, swapping in the other LED’s I purchased was super easy.  I just had to unscrew a couple of hex screws, clean the heat sink with alcohol, insert the COB into the holder and apply new thermal protection.  (Similar to what you’d do when installing a CPU in a computer.)  The replacement LED’s were considerably more blue, which is perfect for my particular application.  I’m sure the 3500K LED’s would have been fine as well, as these still produce blue spectrum light, but I’m going with what I know about indoor horticulture.

I was impressed that other than aluminum angle bracket and frame hardware, I didn’t need anything else that wasn’t in the kit from Timber.  It was practically plug and play as far as DIY kits go.

First Thoughts On The DIY COB LED:

With all that out-of-the-way, my first thoughts are that this new light is stupendous!

First, this light is ridiculously powerful.  I don’t have means of measuring PAR, but I can measure lux.  I can say the output of the COB’s at 18″ is easily twice that of a 6-bulb T5HO light right beneath the bulbs.  I’m seeing 15,000 lux under humidity domes 22″ away.  I’m glad I took a minute to measure the output, this light is a potential plant killer if not used carefully.

The DIY COB LED is now in production and helping a few of our plants that suffered over the winter.We hung it at around 24″ from the the grow table’s surface.  We are getting very acceptable coverage across 10×20 trays.  I used a lux meter on my phone to dial in the lights to the desired level, for which I used my T5’s measurement as reference.  One of the benefits of the Timber kit is that they include a potentiometer which allows dimming of the LED.  This allowed me to dial in the light’s strength to a level I was comfortable with, 75% offered me 22,000 lux at 24″.  I dialed in the LED’s just a bit above the T5’s output, partly because I can and because it reduced energy usage by almost 50 watts.

The cobs raised the room temperature around three degrees Fahrenheit.  (Not bad.)  The heatsinks get warm to the touch, but not much warmer than my T5 fixture.  If I had these in a tent, I wouldn’t be severely concerned about cooling requirements like I would with an HID light.  They would likely benefit from air circulation, though, just as with the T5’s.

Visually, the light from the 6500k COB LED’s is similar to the T5’s output.  Interestingly, though, it seems more “precise” from a visual perspective.  Perhaps a little less warm than the 6500K bulbs we have in our T5 fixture, but definitely the same color range.  I might order some 4000K and 5700K chips to experiment with, but this would likely be for an experimental build in the future.

The buy in cost for the LED’s is higher than the average lighting, but the electrical savings make it worth it in the long run.  For comparison, I paid around $220 for my current 6-bulb T5HO’s and bulbs run me about $10 each to replace.  (I replace 1-3 a year now.)  Given that I should see a 10 year run out of these LED’s and lower operational costs, the LED makes a lot of sense financially.

The Jury Is Still Out On LED COB’s:

For now, the DIY COB is just nursing a couple plants that didn't survive the winter too well and a couple house plant cuttings!
What remains to be seen, and will prove the value of the COB’s to me, is how they will grow plants.  I only have four hours on them at the time of this writing.  I do have a lot of confidence and got exactly what I wanted, so I have high expectations.  I applied my knowledge of indoor horticulture and built what I thought would be the best for my application.

At this point, I expect that the beginning of April will be when we can really start to put this light to work.  For now, we’re helping out a few freshly rooted cuttings and some plants that suffered terribly from our long, dark winter.  We should have seedlings under the light for the next wave of planting, scheduled for the first week of April.

I know the light will grow marijuana well, but that’s not what I’m growing.  I need to see performance across a variety of vegetables and that’s what this experiment was all about.  These will see tomatoes, cabbage, broccoli, Brussels sprouts, green onions, leeks and virtually every other kind of vegetable we can grow here in the sub arctic!  I expect as good, and probably better, results than my T5’s, but will only know that in a month or so.

What I hope to do here is split up different veggies across the T5 light and the LED fixture to try and get a performance comparisons.  While plant growth does matter, electrical savings is king here in the north and that’s why I went this route.  Still, if I don’t see what I expect to see as far as growth, I might need to explore further and develop a better solution.

I will update the blog with results from the light and give ongoing thoughts about it.  I would say by May I will have a fully developed opinion on these COB LED’s and know whether I think it was the right investment.

3 Year Update:

We’ve got three garden seasons of artificial days and nights under our belt using the DIY COB’s. These lights have become vital to our garden. See our three year usage update here!

0 comments… add one

Leave a Reply

Your email address will not be published. Required fields are marked *