Water cooled COB Build/pic heavy

And what the fuck is up with these BJB holders being so finicky and the thermal pads.....these things are fuckin terribly delicate. definitely going back to thermal paste. These things have cost me 3 days of chasing open circuits!!!:cuss:
:cuss::cuss:out of 24 cobs hooked up, 6 failed to light first time install. It took removing them, and manually adjusting the lil leg on the bjb holder to line up properly with the chip. @CobKits
 
i like the paste myself

ive put together hundreds and hundreds of bjbs and have never had one not light up. have never looked twice at a holder to see if the contacts were bent before installing and they always seem to work for me. lemme know if your method is any different than this:

1. clip chip into holder via the two "tits" on the holder clipping into the holes on the chip. chip will be solidly mounted in holder and will take force to remove
2. clean chip and heatsink with alcohol and let dry
3. apply paste (or pad) to chip
4. mount chip on heatsink with M3x8 screws
5. insert 18 GA or smaller solid wire
 
Last edited:
This probably sounds stupid but how effective would it be to run multiple high power cobs that are run through with the water that fills the hydroponic reservoir? Heat the water, cool the chips. I had this in the back of my mind for several years.
 
Who knows how effective this would be?!
You get a bit more light at lower Tj. from the COB's but on the other hand there a pumps and fans needed to keep water temps down... If you manage to make the profits higher than the sum of the additional consumers, then it would be profitable.
Mabye with a big passive radiator and only a additional pump..
I've seen a 650w gaming PC passively cooled with a big converted oil radiator for room-heating. He removes the heating element and insert a 240l pump instead. There is a youtube video but I don't have the link.
There are also other big passive PC cooling radiators like this one below.

The Alphacool Cape Cora HF442 + HF642 are scaleable to nearly every cooling capacity up tp 650w and more with really good ∆T for a passive radiator and this modell is from 2013...
You can connect a 442 and a 642 together to get a 1042 radiator or add another 442 for a 1442 unit.
Only limiting factor is flow and pressure but with a good pump I see no reason for a fail.
There are also vids on youtube about this radiators.
 

Attachments

  • Alphacool Cape Cora HF642.png
    Alphacool Cape Cora HF642.png
    1.6 MB · Views: 17
biostudent said:
This probably sounds stupid but how effective would it be to run multiple high power cobs that are run through with the water that fills the hydroponic reservoir? Heat the water, cool the chips. I had this in the back of my mind for several years.
Click to expand...
Bad move for DWC. Might work for ebb not flood. Great for aquaponics.
 
Godfather420 said:
So @ttystikk, what would you do differently on your second room. We have different styles but environmental control is the same. It's seems from reading that you had a lil bit of hard time keeping the temp up. What would you do differently to fix that?
Click to expand...
I'm considering a rework of my water cooling system to pass warm water through the modules.
 
ttystikk said:
Bad move for DWC. Might work for ebb not flood. Great for aquaponics.
Click to expand...
How's that?
I was under the impression that ~20 gallons of water could easily cool up to a 500W worth of cobs without heating up the water too much (since specific heat capacity of water is very high) to degrade nutrients.
 
biostudent said:
How's that?
I was under the impression that ~20 gallons of water could easily cool up to a 500W worth of cobs without heating up the water too much (since specific heat capacity of water is very high) to degrade nutrients.
Click to expand...
DWC needs cool water or roots get the slime.

Aquaponics needs warm water- at least for warm water species of fish. There's a thread here someplace about a guy who did that.
 
biostudent said:
How's that?
I was under the impression that ~20 gallons of water could easily cool up to a 500W worth of cobs without heating up the water too much (since specific heat capacity of water is very high) to degrade nutrients.
Click to expand...
20 gal of water would be downright hot water if it was taking ~250W of pure heat off of cobs for 12 hours at a time
 
biostudent said:
I'd like to try this one day. Maybe have the water pumped around to cooler areas.
Click to expand...
As someone who's already done it, it's my opinion that is a solution increasingly in search of a problem. As LED lighting gets cheaper and more efficient, this will be less and less necessary.
 
@Godfather420 and @ttystikk this is one the last true threads I ❤️! An exchange of ideas and information, without any name calling and belittling! Please, keep this discussion going. With my next grow room I would like to implement a smaller scale of what you guys are doing (not 90,000 btus ). Here in Oregon we fight the humidity year round so all this information is quite useful.
 
CobKits said:
i like the paste myself

ive put together hundreds and hundreds of bjbs and have never had one not light up. have never looked twice at a holder to see if the contacts were bent before installing and they always seem to work for me. lemme know if your method is any different than this:

1. clip chip into holder via the two "tits" on the holder clipping into the holes on the chip. chip will be solidly mounted in holder and will take force to remove
2. clean chip and heatsink with alcohol and let dry
3. apply paste (or pad) to chip
4. mount chip on heatsink with M3x8 screws
5. insert 18 GA or smaller solid wire
Click to expand...

Number 5 is the only change I made. I went with tinned stranded 18gau

IMG_4806.JPG
As you can see in the pic, where the little metal finger of the BJB holder strikes the cob, its not exactly on the money.

IMG_4808.JPG
In this pic you'll notice with the wire inserted it is cocked to the left.... Could have something to do with the tinning process and it not being circular...We figured a work around.

The next hurdle was diagnosing the shitty relays we got. So check this out, we actually got clones of clones. FUTEC makes a knock off of Crydom Relays. It's a 5 vs 35$ issue. Since the FUTEC's became so popular, apparently other chinese manufacturers started cloning them. They didn't seem to be sensitive enough to pick up the 3.26v from the pcb.

Oh yeah!!!!!!!! Braking news... In about two weeks, we should be ready to start sending out pcb's for for controllability of these mean wells.... Timer, dimming, temp sensing, fan control.... all that shit.
 
Randomblame said:
Who knows how effective this would be?!
You get a bit more light at lower Tj. from the COB's but on the other hand there a pumps and fans needed to keep water temps down... If you manage to make the profits higher than the sum of the additional consumers, then it would be profitable.
Mabye with a big passive radiator and only a additional pump..
I've seen a 650w gaming PC passively cooled with a big converted oil radiator for room-heating. He removes the heating element and insert a 240l pump instead. There is a youtube video but I don't have the link.
There are also other big passive PC cooling radiators like this one below.

The Alphacool Cape Cora HF442 + HF642 are scaleable to nearly every cooling capacity up tp 650w and more with really good ∆T for a passive radiator and this modell is from 2013...
You can connect a 442 and a 642 together to get a 1042 radiator or add another 442 for a 1442 unit.
Only limiting factor is flow and pressure but with a good pump I see no reason for a fail.
There are also vids on youtube about this radiators.
Click to expand...

If you'll scroll back a few pages, you'll see my BTC MINER that was 50 650w basic miners all passively cooled with two car radiators and a 24" fan.
 
You must log in or register to reply here.
Back
Top