DIY - LED
Posted 26 March 2017 - 12:07 AM
mate how long do you estimate these will run for 15 or 20 years?
Should go 10 years.
Cree COB are rated for 50,000 hours.
Mean well drivers warranty for 7 years.
Posted 27 March 2017 - 05:09 PM
Hey everyone, been a while :) I know KCMO has a killer COB thread he's been working on, and I didn't want to clutter up his thread. That thread is a treasure trove of good information, and some grower skill.
I am gearing up to get a 4x4 going and I am still extremely disappointed in off the shelf LED offerings. This frustration is leading me down the DIY path.
This thread will be about that journey.
So, to be honest - my interest in COB is not very high. They do well, plenty of people are using them with great results- but the efficiency is not that great, and they act like a single point light source. Passive heat sinks need to be huge, and even active heat sinks tend to be pretty bulky.
My intuition tells me the high power single diode approach has still not been mastered - and they become cumbersome projects for the DIY crowd. You have complete control of the photon spread, and the photon density can be equally high as a COB with equal or better efficiency.
Also, there is a new hybrid approach in the game. The proprietary name is "Quantum Boards". https://horticulture.../quantum-boards
These are basically a compromise from the ease of DIY of a COB, but photo distribution similar to a more traditional "blurple" panel approach.
Unfortunately these are out of stock right now, and apparently when they come back in stock the LEDs will be from a higher BIN... lower forward voltage with a higher output.... combining to provide a 10% efficiency boost (4% from lower voltage, 6% from better output). I am sure these will also sell out quickly and be out of stock again.
Unless I can get a hold of some of these QBs, the plan will be to design a more traditional panel. Even if I do get some QBs I would want to add additional 660, 730, and maybe even 840 diodes - because flowering plants love these.
Any who, just wanted to say hi to everyone and get something started. I bought some cheapo panels for veg lighting while I get rolling.
your plan with the QB is pretty neat. growmau5 has implemented something similar in his newest setup.
new generation COB efficiency is probably a bit better than the QB on the system level. the circuit boards
that chips are mounted to seem to cause quite a bit of efficiency loss if the chips drop from 207 lumens
per watt nominal to 150 lumens per watt actual. i can keep my old gen bridgelux vero 18s under 40-45C
most of the time, and they make 15-16 watts of heat per chip. the new ones should run 10-12 watts of
heat per chip if drove similarly, and make 160-180 lumens per watt, depending on if i use vero 18 gen 7
SE or vero 29 gen 7 SE. also really digging the new push connectors that the vero gen 7 SE, no COB
mounts, no special molex connectors, just poke the wire in and done.
remember that 50,000 hours isn't a projected lifespan, it's L70- meaning approximately what point is the
old COB emitting 70% of the lumens of when it was new.
you did well by getting the 3500k, Cree 3500k is pretty similar to Bridgelux 4,000k, which has been my
favorite spectrum so far.
IMO, go with three separate fixtures, that way you can adjust their height independently or add/remove
one more thing, you might want to consider less light in terms of watts per square foot. your current situation
looks like it will be 37.5w per sq ft (600w/16 sq ft), which is a little higher than the recommended max of
35w/sq ft. i just ran 300w at approximately 52% efficiency in a 2x4, and it was less return for g/w than 200w
in the same space. also required more cooling, and i had to reconfigure my entire shared air exchange system
as a result, huge PITA. definitely would need at least a 400 cfm fan to exchange the heat from your project.
- mate0x likes this
Posted 27 March 2017 - 08:28 PM
Obviously the advances in COB have swayed my original opinion :D
Yea I will be running a good size fan, and my basement is truly underground - so the temps are always low, even with the heat on. In the past I needed to run an oil filled space heater to keep the temps up with my old LED panel. I will be running 400cfm anyway, just for air exchange reasons. I like to do 3x per minute.
Thanks for the tip on the L70, I realized it was not a "point of death" but didn't know what the % cut off was.
I will keep it modular - currently deciding wether or not to mount the drivers on the frame, or wire them up remotely. Added weight and heat, vs running multiple wires outside the tent. I don't have a great place to mount the drivers to. I might have to make a panel out of some of the extra aluminum I have and mount that to the wall. I plan on putting some eye hooks on the frames and use rope ratchets.
growmau5 was recommending 20 PAR watts per square foot, this light only achieves 19 in a true 4x4. The frame I built is only a 42"x42" an the direct footprint underneath from 16-24" should make this on point with that 20 PAR watts. Hopefully not too much light.
I will measure this unit at operating temperature with a kill-a-watt and a PAR meter, see how it's looking. I was really concerned with even coverage - and would sacrifice some efficiency for that coverage.
Hows thangs in general man? I read you are moving to Co?
- kcmoxtractor likes this