in a sealed room, 1000W of IR, or HID, or LED, or a toaster, will produce the same amount of heat. thats physics
all light has a radiant heat component to it, not just IR. But now youre not talking a sealed room when you are removing heat. removing heat from a cob can be as easy as a ducted HID fixture
old ducted reflectors are almost free on craigslist, etc. not too difficult to mod these to hold led heatsinks which vent the back while physically isolating it to leave the LES exposed. you can plate the front with AL or get teh proper heatsink widths a simple variation on this is to take a 4" pipe/duct and section a portion of it in half, and mount it tight to fit to the top of a bar heatsink. another variation would be a 4-6" pvc pipe with a series of tees that fit over the top 75% of a pin fin and drew the tinest amount of air up and out. i can sketch some of these out if this doesnt make sense
watercooling is another emerging option
many people have had trouble maintaining the temps tehy are used to when switiching to cobs. your climate is a big factor here of course
if youre gonna go thru that much trouble you can water cool the cobs directly cpu-cooler style
i might be confusing natural day/night and plant day/night cycle in your post, but 2 things jump out at me:
1. heatload should be minimal during the natural day when lights are off and you dont have to worry about CO2 which would be off. conventional AC in the space would have a light load
2. many people believe (myself included) that higher ambient temps are useful with LED relative to HID which increases leaf temps with all of the IR. typically 5 deg warmer seems to work well (85 w led+co2 vs say 80 with HID+co2)
3. some have claimed positive benefits to plant growth patterns when there is a 'negative temp differential' (i.e. slightly warmer in plant dark cycle). search forums for this for more infor
efficient lighting saves you money every hour you run it
mini split costs you money every time you run it. if you can reduce or eliminate heat load you'll come out ahead in the long run and pay for your cobs in a year or two (and have many more years of light out of them).
the pvc pipe with the downward facing tees with passive heatsinks partially sticking up into the ducts is the easiest design i can think of and takes advantage of the wide variety of 'bolt together' predrilled pinfins. literally the smallest fan you can fit to the duct to create any amount of draft should be all you need. or alternatively a small 4" inline could easily draft a 5x10 or 10 x 10 room.
if you really want sealed to save CO2 its hard to beat water cooling which is complex and expensive but co2 aint cheap either
some general rules of thumb (depending on efficiency of leds i.e how hard driven):
-1W of HID is replaceable by 0.6-0.7W of LED in most cases, substantially reducing overall heat loads
-re: leaf temp target, slightly higher ambient temps are generally acceptable with LEDs relative to HID
-due to the even nature of multiple cob fixtures heat/light stress on tops of plants is greatly diminished
-there are quite a few ways to be 'completely sealed' or 'partially sealed' (i.e reduced venting/ room air exchange relative to HID
-if you can figure out a system (modified sealed hoods and /or water cooling) with no air exchange the savings in CO2 should really help justify the complexity of the system over time
Thank you for all the info, really gives me a lot to think about.
I like the idea of 4" inline duct pulling outside air, with an old cpu cooler, mounted inside the ducting, and then sealed in order to only expose the LES, for each COB, and then exhausting the heat back outside, to create a closed loop system that would allow for a sealed room. This has pretty much the direction I have been steering, so I'm glad you all agree on the concept. I think you are right, water cooling would be better used directly on the chips, like water cpu coolers, but I feel like the ducted system should be sufficient, since it should not be that important how cold the air (or water) really is that runs thru the ducts, but rather that the heat is evacuated period, before it radiates into the room.
Is anyone using a ducted system like this? and if so, is it for the sealed room reason, or just extra cooling? I am trying to plan ahead for co2. It might be a little while before I can afford that after the adoption of COB led.
The real question I need to fully understand, is whether or not they can stack up against a supplemented DE setup. Which I have a feeling, might be long shot, but the spectrum looks real nice on the 3500k's, much like the my two spectral charts on top of each other, but the lack of any sort of UVA, is what makes me question it's viability as a complete replacement. I know there are other ways of getting this but they are not ideal sources, like flourescent's, plasma's, etc. But hopefully soon to come UVA/B Led's!! But I figure I can always add those to my array at a later time, when they are available
I will probably test maybe 3 chips, and a meanwell driver, to try out in a closet, and see how nice of a plant they really grow. All of the questions/concerns I had made previously are all subject to that one question.
So anyone whose got bud porn, GR/W numbers, GR/meter squared, GR/umol, etc, all that info is useful, but please include the number of cob's and the kelvin/bin numbers, and spacing/optics, so I can get an idea of where you all are coming from, and where I might be going with this.
edit: ooh, and anyone who has test results, even better!
I am looking specifically for people using CXA/B 3590, setups, but if you think you've just got the killer mix of led's in your setup that work better than this platform, and want to share how you came up with it, and why, please do.
Thanks again!
