Water Cooling CXA 3070

SSGrower

Well-Known Member
The only thing Id change about either of these two designs posted so far is that I would use thermal epoxy to afix the chips to a flatened and smoothed piece of square stock. Most drivers, even "cheap" chinese ones have built in protections that would prevent sufficient sparking to ignite ordinary combustable materials. Additionally the common hlg meanwell drivers are class 1 div 1 so they are incapabale in all tested scenarios including malfunctions of generating a spark sufficient to ignite even a flamable atmosphere. I sleep well at night with my basement made modified cpu cooler.
 

Wunderwi

Active Member
The only thing Id change about either of these two designs posted so far is that I would use thermal epoxy to afix the chips to a flatened and smoothed piece of square stock. Most drivers, even "cheap" chinese ones have built in protections that would prevent sufficient sparking to ignite ordinary combustable materials. Additionally the common hlg meanwell drivers are class 1 div 1 so they are incapabale in all tested scenarios including malfunctions of generating a spark sufficient to ignite even a flamable atmosphere. I sleep well at night with my basement made modified cpu cooler.
Problem I found with epoxy is hard to mount reflectors. I have angle welded to each side of my tube at cob locations. Allows for drilling and taping without compromising the tube wall.
 

SSGrower

Well-Known Member
Didnt realize you were using holders, saw 2 pics one with zipties one with screws. Mechanical soundness aside for either of those two, getting a good mating surface between the cob and "heatsink" is a part of good design, the holders do a good job of that and si ce you use reflector makes sense to me.
 

QuinndianaJones

Active Member
Yeah, I used thermal paste and zip ties so I could easily change out wiring or cobs given I learned to solder for this project and some of the solders were shite.
My design definitely has a few issues but honestly leaking isn't one. I did rigorous testing before even switching it on, and the use it's gone through more than satisfied me of its safety. I was and continue to be much more worried about the fire starting potential of an hps bulb at full power than of a fire started by water cooling 25-50% powered cobs running off meanwells.
 

posativek

New Member
Okay, so I've done some research on the efficiency of water cooling vs forced air cooling and it seems to me that water cooling is the way to go. I understand active air cooling is usually effective but my decision to go water is based on two things, I already have a few high gph pumps and other equipment used to create a nice water cooling loop, and I want to come as close to the max luminous flux possible with a 1400ma driver, which requires a Tc of as close to 25c as possible, at as low a cost as possible.
My plan is to mount my CXA3070 chips onto a one inch square aluminum tube with capped ends and intake/outtake barbs at either end. My feeling, backed by some research, is that with an increase in barb size I can increase water flow and boost cooling capabilities even though turbulence would be low in a tube. I understand that by using aluminum I limit my ability to use other more conductive metals such as copper in a radiator but this is mainly for proof of concept since I can't find anyone who actually tried it this way and copper tubing like that is prohibitively expensive at the moment if it can even be found.
I will be building a 5 foot length with 5 cobs running off a Meanwell 240h 1400a to test this out within the next week once I can get to the metal shop for some tubing. Any advice in the meantime would be greatly appreciated.
Thanks y'all!

I don't have time to read the thread so apologies if unwanted info. I just saw a guy use a thin sheet of copper attached to a tube of copper underneath and use this as a supposedly very efficient, water cooled, heat sink.
 

QuinndianaJones

Active Member
I don't have time to read the thread so apologies if unwanted info. I just saw a guy use a thin sheet of copper attached to a tube of copper underneath and use this as a supposedly very efficient, water cooled, heat sink.
Yeah, I thought about going copper, but the price was kinda high and I plan on adding a radiator to the reservoir and a copper radiator is pretty pricey.
 

Wunderwi

Active Member
Yeah, I thought about going copper, but the price was kinda high and I plan on adding a radiator to the reservoir and a copper radiator is pretty pricey.
Do what I did. Like $45 use a heat exchanger for a outdoor wood boiler. Have to add a small amount of antifreeze to stop the corrosion from having aluminum and copper together.
 

Attachments

SSGrower

Well-Known Member
Do what I did. Like $45 use a heat exchanger for a outdoor wood boiler. Have to add a small amount of antifreeze to stop the corrosion from having aluminum and copper together.
Did you source that new? Did you see any cylindrical ones when looking?

Yeah, I used thermal paste and zip ties so I could easily change out wiring or cobs given I learned to solder for this project and some of the solders were shite.
My design definitely has a few issues but honestly leaking isn't one. I did rigorous testing before even switching it on, and the use it's gone through more than satisfied me of its safety. I was and continue to be much more worried about the fire starting potential of an hps bulb at full power than of a fire started by water cooling 25-50% powered cobs running off meanwells.
If you check the backs of the chips and there is a discoloration in the paste, please reconsider my post

Didnt realize you were using holders, saw 2 pics one with zipties one with screws. Mechanical soundness aside for either of those two, getting a good mating surface between the cob and "heatsink" is a part of good design, the holders do a good job of that and si ce you use reflector makes sense to me.
 

QuinndianaJones

Active Member
@SSGrower yeah I'm pretty sure I'm gonna do the thermal epoxy next, that and amend the way it's hanging so weight is more evenly distributed. Other than that this has been an excellent investment imo.
 

Wunderwi

Active Member
@SSGrower yeah I'm pretty sure I'm gonna do the thermal epoxy next, that and amend the way it's hanging so weight is more evenly distributed. Other than that this has been an excellent investment imo.
Just a suggestion. Not sure how wide your light is. You could look at 2 strips of aluminum 1" x 1/4" x light width. Drill some holes and use chain to support it. Run the strips under the aluminum tubes.
 

QuinndianaJones

Active Member
Just a suggestion. Not sure how wide your light is. You could look at 2 strips of aluminum 1" x 1/4" x light width. Drill some holes and use chain to support it. Run the strips under the aluminum tubes.
Do you think quarter inch would be thick enough to avoid flexing? With water in it the light is pretty heavy.
 

mahiluana

Well-Known Member
I used thermal paste and zip ties
have a lock to a solution with minimal heat resistance and Tc

S6001989.JPG

the lamp is able to store 60% (180 W) of the electrical input (~320 W) into a heatexchanger.
Every day it prepares ~ 80L of hot water, which I can use in the shower and kitchen to have
savings in my oil-central-heating-bill.

S6001959coolmac - Kopie.jpg
S6002017vert.jpg
S6001991.JPG S6001993.JPG
 

mahiluana

Well-Known Member
engineered by a professional
your house burning down
Safety should always be number 1
(: you can always install a smoke detector with integrated fire sprinkler system - to increase the mess o_O

Watercooled led-light at the end is a great idea if you chose to harvest the produced heat,
which is still ~75% of total input power.

Light-Heat-Cogeneration could also be a very powerfull tool, to combat global warm up.

Tubes, fittings and hoses are not "high-tec" - but yes - a bit of professional control and overview:sleep:? on your own work, is always a good idea.
 

QuinndianaJones

Active Member
Makes sense, I'll probably just grab some prefab L channel and hanging wire.
(: you can always install a smoke detector with integrated fire sprinkler system - to increase the mess o_O

Watercooled led-light at the end is a great idea if you chose to harvest the produced heat,
which is still ~75% of total input power.

Light-Heat-Cogeneration could also be a very powerfull tool, to combat global warm up.

Tubes, fittings and hoses are not "high-tec" - but yes - a bit of professional control and overview:sleep:? on your own work, is always a good idea.
I use the heat to heat a floor in the winter but I'm still researching how to recapture the heat as energy to use for cooling the room in the summer.
The setup I have is more efficient than 25/75 light/heat, it's closer to 35-45/55-65 with the under driven cooled cxa3070 cobs and meanwell drivers. Nitpicky I know but that 10-20% cost me a bit. However that's still a lot of energy to use, between 275-550 watts and I'd love to recapture it.
I keep meaning to buy an automatic extinguisher to mount next to the drivers and wiring but I forget :-?
 
Top