opinions on diy COB led

dsmer

Well-Known Member
Hey guys! This is actually my first thread on here lol. I'm not a big forum poster but i like them for information and there are definitely some smart people on here! This may be long and I apologize lol

So heres my story. Until recently, like many others, iv'e relied solely on hps lights for flower for many years without issue. I currently run 2x600w and 400w in a 4x8 (give or take) space. Now although they work great, the price for running them is making me think alternatives. Currently where I live, the total cost of electricity went from .14cents/kwh to .24cents/kwh as soon as winter hit! So at the ~1400kwh i use every month (whole house) my bill went from $196 to $336! thats a ~$140 difference every month for the same usage. This medicine is for personal use (between 2 people) and I don't sell it so this is becoming an expensive hobby for me lol. Currently I removed the 400w and dimmed the 600's to 400 each for the time being. I also switched from a metal halide in veg to t5's which is actually working very well.

My goal is to replace one of my 600w hps lights with an equivalent cob led fixture, and it works well I will replace both. See another one of my problems is my a/c died last year and I wanna be prepared for when summer hits. This is bad because now I have to shell out $500+ on a new a/c plus @1200w 12hrs a day would be another $100 a month for electricity! Right now im air cooling my lights and it stays right at 75f with minimal elec usage. But only because its winter lol.

Which brings me to led's. I tried led's a few years ago for veg. It was a 200w california lightworks vegmaster. although it worked well, it didn't exactly "wow" me and I eventually went back to halide and sold the led. Until this point thats the only real world experience I have with led grow lights.

After reading thru a bunch of the diy led threads on here and gathering a bunch of info (thanks to all who spread there knowledge btw!) I think I'm gonna give it a go. I'm looking for some advise and some constructional criticism and hopefully these COB leds will be the future! Im excited for it.


Ok! I want to start building one of these things fairly soon but Im in no rush by any means. Heres a little background on my skillset. Im a welder/fabricator by trade and have been doing so for about 12 years. I also own my own equipment at home which is nice because i can make any style fixture I choose. Im also pretty handy with wiring. Ive done some home wiring (basic stuff) and alot on automotive applications. so Im fairly confident in basic elec knowledge though Im far from an expert. The funny thing is my last job i was actually a fabricator for a lighting company which made commercial and residential led and flouro lighting (mostly ornamental) fancy stuff for offices and schools etc. I built aluminum enclosures for them. unfortunately I left there before I got into led grow lighting which sucks because they had huge buying power from both cree and meanwell and had shelves FULL and i mean FULL of ballasts,drivers, led's etc. as well as tons of wiring. Oh well lol.

So id like to brainstorm some ideas on my next light. hopefully supra and others will chime in. bare in mind im not trying to "outdo" the hps lights, just want somewhat equivalent results with far less elec usage. My current basic idea for the first fixture is to use 6x cxa3070's being driven @1.4a with (2) HLG-185H-C1400A drivers powering 3 COB's wired in series each. This "should" give me about 325w draw (not including fans) but correct me if Im wrong. Does this sound good for a 4x4 footprint? If total draw from this light is under 400w and has the same results as 600w hps then I would save 200w each fixture I replace. for example: 400w saved total from replacing both hps lights = ~$35 a month or $420 a year! (i think thats a sign lol) thats not including having to replace hps bulbs every year (~$100 each) and a/c bills @~$100/month. so at first glance it looks like i would save around $1000 a year if you figure I use a/c for 4 months (or more sometimes) by switching to led. This of course is all theory in my book.

anybody have ideas on optimal fixture design? i would love to hear/see some ideas for building the ultimate fixture. I also work at a machine shop so I have access to mills, etc.

One last thing id like to throw in lol. I was thinking about the options of powering pc fans for active cooling. What about using thermoelectric generators to power the fans? maybe wedge a Peltier chip in between the COB and the heatsink. The heat from the led combined with the temp differential of the "cold" side of the heatsink should create a measurable voltage. maybe enough to power the fan, effectively using its own heat energy to increase efficiency. the fan would also be self controlling because the hotter the led got, the more voltage it would create thus spinning the fan faster. This would eliminate the need for external fan controllers and additional wiring+ increased efficiency (which is what where all after right?) of course this is only if the chip would create enough current, which im not 100% on.

again sorry for the long story lol but id love to hear some feedback! thanks for reading
 

SupraSPL

Well-Known Member
Sounds like a great design you have in mind, very efficient and cost efficient. For a 4X4 it would put you at 19.5 dissipation W / ft² which might be on the low side for optimal results. The buds on top would probably be fine but anything lower ones might be a bit too small and larfy than they would otherwise be. On the other hand the lower intensity is somewhat more efficient so overall you would get more meds for your wattage.

If you were to use passive cooling you could do a pair of 5.88" X 40" bars ($144+ shipping), for active cooling, a pair of 3.5" X 30" ($74+shipping+$5 for fans) or you could use (6) Alpine 11s ($60 shipped)

Good question about the peltiers, I wonder what the actual temp differential would be in they were inline with th thermal path. It might even be a higher differential if they were attached to the heatsink and had a small heatsink of their own on the other side. One thing for sure though, they are very inefficiency so if we need 2 watts for the fan we might need to harvest 10-20W from the Peltier which of course is not likely.
 

churchhaze

Well-Known Member
While this sounds like an interesting idea, would it not be like using energy captured by a solar panel to generate light? (when you can just install a glass window and let light in directly)

I believe you'd essentially be taking energy that would have been used for convection and redirecting it to power a fan.

One last thing id like to throw in lol. I was thinking about the options of powering pc fans for active cooling. What about using thermoelectric generators to power the fans? maybe wedge a Peltier chip in between the COB and the heatsink. The heat from the led combined with the temp differential of the "cold" side of the heatsink should create a measurable voltage. maybe enough to power the fan, effectively using its own heat energy to increase efficiency. the fan would also be self controlling because the hotter the led got, the more voltage it would create thus spinning the fan faster. This would eliminate the need for external fan controllers and additional wiring+ increased efficiency (which is what where all after right?) of course this is only if the chip would create enough current, which im not 100% on.
 

dsmer

Well-Known Member
While this sounds like a good idea, would it not be like using energy captured by a solar panel to generate light? (when you can just install a glass window and let light in directly through transparency)

I believe you'd essentially be taking energy that would have been used for convection and redirecting it to power a fan.

basically yes. The way I see it, The led generates heat as a byproduct from consuming electricity. this heat is wasted by drawing it into the heatsink and expelling it through the fan. So the heat energy you generated is essentially "wasted" while the usable energy is used in the form of light (obviously). i dont like to waste lol. I used to travel all over fixing boilers in renewable energy power plants and learned alot. ill use an analogy from that. Power plants burn a fuel (trash in the case of where i worked) to heat water to steam which is used to turn a turbine. that steam is the "energy" that was created in that process. and also alot of heat. that heat energy goes into cooling towers where the steam condenses back to water and reused again. when that steam gets sent to cool, essentially your wasting alot of that heat energy created. i know thats on a lot larger scale but remember your paying for that energy either way so whether its being used as light or heat i wanna use all of it to its potential lol. so although an led may not make enough heat to power a fan, a bunch of them used effectively may have enough usable power. i was thinking it may be enough to power a small water pump (for pc cooling. there very small and work very well) to liquid cool the heatsinks. same basic principle that the more heat created the more power the peltier will produce. again im just rambling and i may be completely wrong because i have no personal experience with TEC chips.

I also forgot to mention i grow scrog so i dont really have lower branches to worry about per say but i may wanna make this thing expandable so i can easily add more boards later. thanks so far for the advise
 

churchhaze

Well-Known Member
What i'm saying is that heat you're taking for the peltier is energy that would have been used for convection airflow. Gas doesn't move for free. When you take energy out of the smoke stacks, you reduce the speed of the gasses. You can only do that so much. In the case of a wood stove chimney (easy to grasp analogy), if you tried to steal every ounce of heat from the chimney, the inside would turn black and you'd end up with a smokey house... or worse... a chimney fire! (and the fire would go out!)

Basically you're suggesting the analogue of replacing convection in a chimney with a fan that sucks smoke out, powered by the heat you stole!!!

I know where you're coming from, but I think it's unlikely you will beat the airflow of natural convection using a fan powered by the same heat. On the other hand, I've never really looked into it beyond your question, so maybe you're on to something.

Natural convection (passive cooling) is self regulating anyway and already re-purposes the heat to produce airflow..
 
Last edited:

dsmer

Well-Known Member
What i'm saying is that heat you're taking for the peltier is energy that would have been used for convection airflow. Gas doesn't move for free. When you take energy out of the smoke stacks, you reduce the speed of the gasses. You can only do that so much. In the case of a wood stove chimney (easy to grasp analogy), if you tried to steal every ounce of heat from the chimney, the inside would turn black and you'd end up with a smokey house... or worse... a chimney fire!

Basically you're suggesting the analogue of replacing convection in a chimney with a fan that sucks smoke out, powered by the heat you stole!!!

I know where you're coming from, but I think it's unlikely you will beat the airflow of natural convection using a fan powered by the same heat. On the other hand, I've never really looked into it beyond your question, so maybe you're on to something.

Natural convection (passive cooling) is self regulating anyway.

I get what your saying and your most likely correct. I only got intrigued because i have seen some people using small heat sources to power TEG chips to power small pc fans. i may have to purchase a few cheap ones to experiment with but may turn into nothing. Im more interested in the actual led fixture design as efficient as possible.

anybody got any insight on placement of the COB module itself? instead of completely flat pointing straight down, maybe angle the boards in respect to the others to complement a complete coverage using the existing viewing angle of the led? so basically an led with a 115deg angle mounted flat pointed straight will cover 115deg but if tilted slightly along with others countering opposite directions "should" have a more complete coverage no? sorry if im being confusing lol
 

mahiluana

Well-Known Member
that heat energy goes into cooling towers
:peace: Have a look to the Coolmac-system.

- This is a watercooled led light + waterheater
- 300W lamp / 15h can prepare 80L of hot water.
- World record in low junction temp. and many other DIY ideas
- Up to 85% energy efficacy - another worlds record
- easy to copy @ incredible low cost
- possibility to produce in license. :fire:

Make yourself comfortable and start with my contents:

eg.



S6001959coolmac - Kopie.jpg
 
Top