Water cooled cobs

[Randomblame]

You can see most of the wiring back on page one of this thread, in the second group of pictures. You're right, 5th cob in series with 4 others in 2S2P means a Vf of around 104. Was just wishful/hopeful there might be enough slack in the numbers that the driver could handle it. Just sheer morbid curiosity to play with it, after finally finding the spare CPU cooler. But nope, like nfhiggs said, the driver's in limp mode.

In other news, after removing the 5th cob and returning to normal, tried a lux meter app on my Nexus 7 tablet (because for some reason lux meter apps on my HTC m9 max out at 8000lux) and measured 30k lux at ~10" directly under a cob.

I'm happy with this light. It's not the most cost effective way to deliver 170watts of light but for an engineer it's pure mental masturbation.

Forget those apps, they are not calibrated and differ a lot from sensor to sensor. My OP6 maxed out at 34999lx while my old S2 can show values up to 14klx.
But there is a high difference when I use my cheap lux meter(max. 200klx, 12$, e3ay).

With such an app you can at most check how the light distribution is, but about the actual brightness it says nothing.
I'm pretty sure these cheapo lux-meters are also not calibrated, at least mine seems to show lower values.
I've compared my readings with a friends PAR-meter readings and found out the real PPFD is ~100-120μMol higher than my converted estimations using the lux numbers and the calculated conversion factor.
So in the end using lux is a bit of a gamble because it depends on the sensor and if it's calibrated or not.

Cheap handheld PAR meters are available from ~150$ to up to +2 grand. A cheap 12$ lux-meter can be used to estimate PAR numbers, but most of the time it's more guesswork.

But it is useful to even out coverage when you use more than one LED and to determine how homogeneous the light distribution is. Gardeners have used lux meters for over 50 years and they still do. PPFD is just a number but it depends on how well the light is distributed and therefor a lux-meter is as good as any PAR-meter.

 

[ANC]

I've seen a local rig made with single row strips... Simple 20mm square tubing with those square plastic plugs you get for them, epoxied in. and then barbed connectors put into those...

 

[shimbob]

I've seen a local rig made with single row strips... Simple 20mm square tubing with those square plastic plugs you get for them, epoxied in. and then barbed connectors put into those...

I like the square tubing idea. I'm planning my next light, 1280 watts of F- serie strips, but for now to get it up and running asap I'll use Al angle bars. The eventual construction of a square tubing water cooled frame will definitely keep me preoccupied.

Speaking of the Nexus 7 Lux meter, I tried it on the CIrrus light. Thing just went bonkers, the number fluctuated nonstop, 45, 300, 32000, 8000, 300... just all over the place nonstop, never settled.

 

[mahiluana]

The eventual construction of a square tubing water cooled frame will definitely keep me preoccupied.

yeah brother just do it
today i will finish mine -

105 x 55cm / "wireless soldered !" 7 band spec. / 500W for 1,5m² / ~ € 300,-

as a recommendation !!! don`t chill the water with add. electricity

--- try to use the produced hot water in (shower, kitchen, washing mashine, radiators in wintertime....) --- to be a master of light... and maybe world-champion of energy-efficiency

(sorry - at the moment it`s me to claim this position)

 

[mahiluana]

to get our own subforum?

--- too much silence - toooo slow
i know it`s arrogant to claim oneself a "worldchampion".

on the other side:
everybody could use the above design with better (more efficient) chips and drivers
and claim this title for him (her).

 

[ANC]

Snow cooled giraffe

 

[mahiluana]

Snow cooled giraffe

slowly cooked apes

 

[shimbob]

Hard at work vegging the new clones.
This will be our first round with the cobs from the very beginning of the round, looking forward to dat 1/2" nodal spacing!

 

[hour]

I'd like to tap some 80/20 extrusions that come with the hole that goes all the way through. Tapping and using some proper elbow fitting with teflon tape would be sweet, but the tubing diameter would be like 1/4" ID at biggest... I think the hole that goes through the extrusion is smaller than that even.

Anyone tried that, or water cooling using something similar to fish tank airline tubing? The only thing holding me back from even using 40mm square cheapo water blocks like OP is the tubing bulk. 80/20 would be easy AF to hang and easy to tap, and its profile might lend to some natural cooling anyway. I have a 4 cob 1212 200w setup that I may try this on. Two-2ft sections of 80/20 with two cobs on each, plumbed together, and covering a 30x30" tent. Could even drill this without springing a leak if you got a different profile extrusion. Just not sure if the hole diameter, and max gph of that size, would be adequate. I'm thinking.. probably, with a fan blowing around, two cobs per 2ft section, and water cooling.

 

[mahiluana]

Anyone tried that, or water cooling using something similar

aluminium square tubes - closed with flanges is cheapest and best...

 

[hour]

aluminium square tubes - closed with flanges is cheapest and best...

So I buy the aluminum tube with a flange like eztube's offerings? Then plug the end with a plastic cap, epoxied, and drill the plastic cap for a water fitting?

Hell, something like this might work without water cooling even, if each 2 foot section is only cooling two cobs running at ~50w... and a little fan blowing on the track. This is a super-flanged extrusion from eztube.

 

[mahiluana]

Then plug the end with a plastic cap, epoxied, and drill the plastic cap for a water fitting?

no plastic - no expoxy
i use aluminium and rubber to close the tube -

 

[hour]

So if your cut tube ends aren't perfectly square you'll leak? Seems kind of janky and a lot of extra bulk

 

[mahiluana]

Seems kind of janky

plastic and expoxy is janky !

a screwed flange with a rubber sealing is the professional way to close a tube in pipeline constructions.

the lamp fixture in above pic ^ is a worldrecord holder in energy efficiency, because it is watercooled
and can rewin 60% of the inverted electrical power as hot water in my boiler.

with this profil you will dissipate most of the heat to the air - only a small part of the heat will go to the water tube in the center.

to have a max. of heat input to the water -
i even recommend this:

 

[hour]

Is that pic pointing out the waterproof driver or the thing above (not sure what that is)

I see the point of trying to reclaim as much heat as possible for max efficiency, but I'm happy if I can find a cheaper alternative to pin heatsinks, a cleaner approach than active cooling through fans, and a tolerable temperature inside the tent. If the extrusion is transferring heat to the air that's not very conducive to getting heat out of the tent, so I guess I should probably whip one of these up and see exactly what water cooling through such a narrow passage can accomplish.

I'm okay with still running my same exhaust fan at the same speed - which has been keeping up even during this summer heat, but it'd be super cool to never wire up dozens of PC fans again. Water Cooling Lite edition

 

[1212ham]

If the extrusion is water cooled, it should be cool enough to not transfer heat to the air.

I cool 200 watts with about 3-5 gallons in a plastic storage bin, located in an adjacent room. I maintain 84°F air temperature in the tent with an Inkbird temperature controller, it runs a $6 submersible pump in the reservoir, no radiator. An Inkbird humidity control operates the extraction fan.

The only thing holding me back from even using 40mm square cheapo water blocks like OP is the tubing bulk.

I think heat transfer will be an issue, that's a long, narrow heat path to the COB. Given my experiences, I'd do it like @shimbob.

 

[ANC]

So I buy the aluminum tube with a flange like eztube's offerings? Then plug the end with a plastic cap, epoxied, and drill the plastic cap for a water fitting?

Hell, something like this might work without water cooling even, if each 2 foot section is only cooling two cobs running at ~50w... and a little fan blowing on the track. This is a super-flanged extrusion from eztube.

It might work without water cooling, but considering the efficiency loss when heating your geyser with bought electricity over just using the waste heat from your lights, I'd still do the effort to make a water cooled rig.

Honestly, I'd just get plain square tubing to save money, and spend the savings on a more reliable pump.

 

[nfhiggs]

Honestly, I'd just get plain square tubing to save money, and spend the savings on a more reliable pump.

Word. The 1010 profiles are pretty pricey compared to the Quickframe stuff they make.

 

[hour]

Doesn't seem terribly expensive if you build the remainder of the frame out of cheaper aluminum, $27 for a 4 foot section of this stuff.

The profile I initially posted wasn't really the one I was looking for, but one I found on google images since you can't copy image on 8020's website. Screen shot and uploaded this which might be better.

The larger hole diameter is 0.262 inches. Wonder if one could tap both of the larger holes and find a barbed fitting for both, or an elbow for one and a T for the other so one line could supply both runs through the extrusion.

This one is 1.5" wide and 3/4" tall, thicker aluminum, and the water would flow right above the cob on both sides.

Also no texture or fins so less heat transfer to the air I'd guess

 

[nfhiggs]

Doesn't seem terribly expensive if you build the remainder of the frame out of cheaper aluminum, $27 for a 4 foot section of this stuff.

The profile I initially posted wasn't really the one I was looking for, but one I found on google images since you can't copy image on 8020's website. Screen shot and uploaded this which might be better.

The larger hole diameter is 0.262 inches. Wonder if one could tap both of the larger holes and find a barbed fitting for both, or an elbow for one and a T for the other so one line could supply both runs through the extrusion.

This one is 1.5" wide and 3/4" tall, thicker aluminum, and the water would flow right above the cob on both sides.

Also no texture or fins so less heat transfer to the air I'd guess

Problem with that kind of profile is there is much less surface area in contact with your water. Plain square tubing maximizes that. The EZtube you showed above would work much better.