Water cooled COBs

sixstring2112

Well-Known Member
seasonal energy efficiency rating (SEER). This is the total seasonal cooling output in Btus divided by the watt-hours of electricity consumption.

mitsubishi 4 ton minislpit =22 seer

chillking 4 ton highest rating= 18seer
so @CobKits can you point me to some data sheets that tell us how water cooling is 70% efficient ? and i would need that to heat the room as well,without looking to hillbilly lol

theres talk of a 24 seer diaken about to hit the stores.

theres pros and cons to everything.if i was growing in water i might consider it,or a basement of a house that already has heat.i understand the water transfers heat better than air part.but the units as a whole are much closer in running costs than some would have you believe
 

CobKits

Well-Known Member
well for one thing you dont need a chiller to operate water cooling effectively. remove the compression/condensation cycle and just use a radiator to cool the water down. or in a dry climate an open system with evaporative cooling is good for a substantial temperature drop. even if the water hitting your waterblocks is as warm as 30C there still is a lot of room to suck the heat off your cobs

secondly pulling the heat directly off the chips is fundamentally a lot more efficient than radiating all that heat into your space and then using air cooling to reduce the temperature

thirdly if you take the energy savings from not using inefficient compressor/condensers and use that money every month buy more chips and be more efficient i think you will come out ahead in the long run, it increases the efficiency of the two steps above

my .02....but its premised on a simple affordable reliable water cooled system
 

sixstring2112

Well-Known Member
theres a point where efficiency meets stupidity imo.and for me that point is mixing water and electrical components in a fixture 2 feet from my face every day lol.i would think everyone would be triple fusing everything just for safety?
sure is cool to see you guys experiment with the stuff though,just not my gig
 

CobKits

Well-Known Member
understood. i kind of think of it as water is grounded, just like all the heatsinks, etc. but it does add complexity for sure, esp at your scale
 

sixstring2112

Well-Known Member
i was looking at setting up a water loop that went outside and came back in and went through a few of those icebox style radiators last year.just pump the water kinda slow and let our cold ass weather cool the water.but i never did it.i pretty much just pull the cold air into my shop and just dump it out at the ceiling and it cools my whole op once we see daytime highs under 45f
 

PSUAGRO.

Well-Known Member
theres a point where efficiency meets stupidity imo.and for me that point is mixing water and electrical components in a fixture 2 feet from my face every day lol.i would think everyone would be triple fusing everything just for safety?
sure is cool to see you guys experiment with the stuff though,just not my gig
I agree, watercooling is not worth it as minisplits/leds are rapidly gaining efficiency

yeah, mitsus are the shit..............dead fucking reliable with a surge protector on the junction box, only lost one due to a storm and learning the hard way about surges.

their pushing above 30 SEER now(hyper heating inverter), even a waterloop is looking to be not worth it in the near future ........sorry ttstick
 

Shugglet

Well-Known Member
This, while true, only goes so far. The fact that water is a far more thermally dense fluid than air can be taken advantage of in several ways inside the chiller to improve efficiency, this is what I meant by the two are different.

Two AC units aren't cheaper than one chiller, either to buy or operate. Further, one chiller can automatically handle two bloom rooms on a flip. At small scale, so can minisplits- but only water chilling scales up. Since I'm modeling heat flows of larger facilities, it makes sense to use the tech. At a smaller scale than this, just get a minisplit.
Remember tty that all of this was started in your response to me saying I could DIY a smaller version of your setup for far cheaper and make it just as reliable. Then you started going off about how "ACs and Chillers werent the same". Its important to remember we were discussing their reliability at this time. (you seem to forget often what it exactly is you are discussing, or trying to discuss I should say). How does the fact that one transfers heat to water and one transfers it to air effect them from a mechanical standpoint?

And look your second paragraph, you completely either forget or ignore that the 2 AC reference was again, to my original point that I could build a DIY chiller out of them NOT run them as 2 regular AC units. But I shouldnt have to say that because someone who has "vulcan logic" should be able to follow a simple damn conversation without CONSTANTLY twisting what the other person said to try to prove your point. You honestly amaze me bro.

Either youre too dumb to follow. And if thats the case I apologize. Or youre simply a hypocritical disingenuous prick with a massive superiority complex. No offense.

We generally have cool temps in the UK other than summer days (which are the only time we consistently get above 22C) I'd love to have an underfloor heating setup.

2kW running 24/7 (2 staggered areas running 12h each) would heat up a kitchen and bathroom floor nicely. That's my longterm goal; seems more energy efficient than any other solution when thinking about the bigger picture.
Hope you didnt miss the point I made about the simple adjustment you can make to your setup without adding anything yet still make it more efficient :)
 

sixstring2112

Well-Known Member
how efficient are those as heaters, say relative to a resitive electric?
Well if its a grow room you really only need heat for 50% or less of the day.the lights heat the room half the time.its tough to say but they have a ton of info on google.where im at natural gas is cheap as hell but im not sure i want to go that route and have all the associated ducting that goes with it.
 

Shugglet

Well-Known Member
Yeah no. The thing is, you keeps telling everybody to go for a chiller and I'm asking how much power it uses to see if that makes sense. That is really the only thing I have been asking.
I think the problem your having is the fact that his setup covers all his cooling needs and does it rather efficiently.

Asking him how much power his uses is kinda pointless because his cooling needs dont align with yours. Given the same amount of BTUs, one should be able to build a more efficient setup with a water cooling design vs an air cooling design. Now of course this completely ignores startup costs, and a hefty startup cost it would be. Probably NOT worth it to retrofit an existing setup I would think, but if you were designing a new setup it would be worth a look IMO.

Tty is just terrible at explaining points imo so I hope this helps.
 

benbud89

Well-Known Member
So... here's version 3.

As some of you know I made a big purchase of 250x Citizen CLU046-1212 COBs. It was always my intention to run the chinese COBs for a couple of months to make sure the water cooling worked before investing in some more expensive/efficient COBs. I had intended on getting started in about a month's time but the opportunity came up to get the citizens for cheap so I took it earlier than expected.

View attachment 3767994

My intention was to have units of 600w drivers with 30 COBs each. 3x units per 8x4 tent. 90 COBs per tent.

At the moment it's pretty hot during the day so I'll be running only 2x 30 COBs until mid September.

So here's version 3. Each unit took about 2 hours to upgrade from the version 2 units I posted at the start of this thread... which isn't too bad :)

View attachment 3767991

I used aluminium angles to minimise the flexing I previously had going on.

The old COBs were removed with a razor blade, they came off pretty easily. The new COBs were stuck directly to the water blocks with Arctic Silver Thermal Advesive after a cleaning with isopropyl alcohol.

Here's a pic of one of my tents currently:

View attachment 3767992

COB surface temp is 46C, water temp out of the radiator 30C, water temp going in to the radiator 38C, floor temp 26C, canopy temp 30C.

I know it looks like some COBs are much hotter than others but really it's just a difference of a degree or two which shows up dramatically due to camera sensitivity. Likely due to differences in how well the thermal interface was applied and maybe minor differences in COB voltage.

View attachment 3767998

As I said I'll be adding another unit in the middle in about a month for 1800w top lighting. Then I'll add another 320w of side lighting per tent... that's the plan anyway!
I think your stuff is really, really cool. Are they still run at about 800mA, and if so, what does that end up meaning in terms of efficiency and lumens? I really like your thread, thanks for sharing it all, cant wait for some more pictures with plants in them!
 

loftygoals

Well-Known Member
I think your stuff is really, really cool. Are they still run at about 800mA, and if so, what does that end up meaning in terms of efficiency and lumens? I really like your thread, thanks for sharing it all, cant wait for some more pictures with plants in them!
Thanks :)

They're running at 33.5V and 620mA each so just over 20W.

In terms of efficiency these are supposed to be 119 lm/w (90 CRI so the PAR is probably higher than the lumens would suggest) at 37.3W and 85C. I've estimated I'm getting around 133lm/w by running them at 20W and water cooling them to a surface temp of 46C. TBH this is all theoretical and it doesn't really matter to me what extra efficiency I'm getting or could get. Why?

1. Without water cooling I'd not be able to run this many watts in such a small area in my loft. I can't get a big AC unit in my loft. So comparisons to air cooling at the same wattage are moot (for my situation).
2. I couldn't afford more efficient COBs at the moment. 250 of these Citizen chips cost me £550 delivered. If I went with 120 CXB3950s instead I'd be looking at paying at least £3200 on the grey market. The extra efficiency and power saving (probably around 10% going from 55% to 65%) would take a hell of a long time to offset that upfront cost.

My lighting energy costs £1660 per year. If I needed 15% less energy (1-(55/65)) to get the same results I'd save £249 a year... (3200-550)/249= 10.5 years for the CXBs to pay for themselves with efficiency savings over the cost of buying the Citizens.

I'd love to say I'll still be using the same setup in 10 years. Realistically that ain't gonna happen.

Edit: Just edited my power savings calc which I think was a bit off.
 
Last edited:

resinhead

Well-Known Member
Lofty. Your awesome for doing this! Forgive me for not reading everything before dropping my input.

Just that your water should be treated to prevent corrosion. Propylene glycol I believe is what is currently used in Solar technology water heaters. Also you should avoid dissimilar metals in your design to prevent electralosis.

Awesome work again! Just like your first lofty goal.
 

resinhead

Well-Known Member
Also I was thinking about how some packaged AC units Incorporate a "sling ring" to use condensate water to mist the evaporator to further aid in removing latent heat.

Have you considered doing something like that? Or maybe just dropping your radiator in ice water?
 

JorgeGonzales

Well-Known Member
In terms of efficiency these are supposed to be 119 lm/w (90 CRI so the PAR is probably higher than the lumens would suggest) at 37.3W and 85C. I've estimated I'm getting around 133lm/w by running them at 20W and water cooling them to a surface temp of 46C.
133 lm/W is spot on, however you arrived at that number, and as you said the efficiency is much higher than that would imply because it's 90CRi. So you are correct on all counts, and somewhere in the high 40s for efficiency (at least).

30%+ more efficient than the best HPS has to offer, nice job.
 
Last edited:
Top