Water cooled COBs

wietefras

Well-Known Member
True, you can go higher than say 50%, but it's relatively expensive. Generally you won't make that extra investment back on power savings, so for most people there is no use for going above 50% right now.

I calculated that the price of the COBs (PAR per $) doubles if I wanted to go from 50% to 53.4%. I pay around a euro per watt per year so it would take me 7 years to get back that extra investment in power savings (if I have the flowering light on for 12 hours every day of the year).

BTW If you want to go for high efficiency (ie above 50%), the Citizen COBs aren't very good. I think then you would better (cheaper) off with Cree COBs. Citizens are cheaper when you push them a bit harder. Although I estimated that on the version 5 Citizen's. Perhaps the version 6 COBs improved enough to raise that bar a little.
 

MeGaKiLlErMaN

Well-Known Member
True, you can go higher than say 50%, but it's relatively expensive. Generally you won't make that extra investment back on power savings, so for most people there is no use for going above 50% right now.

I calculated that the price of the COBs (PAR per $) doubles if I wanted to go from 50% to 53.4%. I pay around a euro per watt per year so it would take me 7 years to get back that extra investment in power savings (if I have the flowering light on for 12 hours every day of the year).

BTW If you want to go for high efficiency (ie above 50%), the Citizen COBs aren't very good. I think then you would better (cheaper) off with Cree COBs. Citizens are cheaper when you push them a bit harder. Although I estimated that on the version 5 Citizen's. Perhaps the version 6 COBs improved enough to raise that bar a little.
I'm going with the luminous $19 cobs, might see if the crees prform the same but it's hard to say. Can't beat $19 for the performance of an 1818-1836
 

wietefras

Well-Known Member
So then say 4 years to earn back the extra investment?

Personally I'd rather spend as little as possible on COBs now. As they seem to be improving in efficiency and dropping in price at quite a high rate.
 

MeGaKiLlErMaN

Well-Known Member
So then say 4 years to earn back the extra investment?

Personally I'd rather spend as little as possible on COBs now. As they seem to be improving in efficiency and dropping in price at quite a high rate.
Agreed, to me $1400 for a light that will last for over 5 years without any maintenance is worth it. If it performs like my other light I'll get a much better yield. My goal of to replicate the summer sun. More of a yield per area than a g/W
 

wietefras

Well-Known Member
Agreed, to me $1400 for a light that will last for over 5 years without any maintenance is worth it. If it performs like my other light I'll get a much better yield. My goal of to replicate the summer sun. More of a yield per area than a g/W
You won't get a better yield per area though. You only save small amount of electricity.
 

MeGaKiLlErMaN

Well-Known Member
You won't get a better yield per area though. You only save small amount of electricity.
That's the same cost of a DE 1000W hps with bulb changes every 6 months. $1400 is spent in 5 years either way... But the cobs will last far past that. That same light should have a PPFD around 1500-1700 in a 4x4 and a de hps has 750PPFD in the same space by itself... I'll do a heat test once I have them running to see how fast the room heats up. After the harvest I just had I can say that the cobs produced way more bud total for less watts. So yes there is a good increase in yield com paired to hps. Numbers will follow.
 

mahiluana

Well-Known Member
What's the point of that claim?
I mostly talk about watercooling, because I want people to use the heat energy created
by led light.
I don`t care so much whether it is 70% or 40% heat - generated inside a led chip.
It is important for me, that i`m able to suck 60-70% of this heat energy at the backside of the led platine and store it into water, - to use it.

!!! You can do that with all brands of hp led chips with a bigger platine.

So my max. to rise efficiency is between 70% of 70% and 60% of 40% - so let`s say 24-49%.
And it really doesn`t matter if you use cheap chinise or more efficient quality chips.
Driving them @ low current will not lower their total heat production significantly.

This way it is easy to convert a cheap chinise epistar cob with 100lm/w into incredible 280lm/w

Uuuuuuuhps - sorry :eyesmoke: see it a bit relative

and don´t forget, that also epistar chips gain efficiency by driving them @ low current.

So let`s make the bill: ~

100lm/w @ 1,5A = 4 x ~140lm/w @ 375mA

As i can store 55% of my lamp power(295W) in my cooling water (160W heat energy)
the bill changes into:

140lm / 0,45w + 0,55w heat energy to use

- sorry 140lm/0,45w = > 280lm/w

There is no tricky fake information
- you only have to accept your lighting system to be a waterheater.



It`s hard to say if mounting cheap chinise cobs and producing more heat would be better than
sucking less heat and more light - because it depends a bit on the money you have to pay
for hot water normally and how much you need.

In my case with the savings in hot water - i can pay the electricity for my lamp.
Every day - every year - this means important savings of money and CO2


300W panel + heatexchanger cost me 400,-EURO - 8 month to compense the money by energy savings + a free 500-600g package of weed.

+++

Low junction-temp. moves efficiency as you all know.
The first barrier for the heat normally is the rel. low thermal conductivity of grease or glue between chip and heatsink.

Coolmac patented systems connect the chip directly to the cooling water and move the junction towards the heat with a > doubled platine`s surface directly under LES.

25°C of the cooling water in the coolmac give much better results than 25°C of an aluminium heat sink. 25°C coolmac "simulates" ~10°C of a common heat sink.

My power meter shows up to 55,5W - pretty much for a 50W cob 33V@1,5A

Next panel to proof will be built with meanwell&cree and citis - (: promised.










 

wietefras

Well-Known Member
OK I'll agree that the COBs you are using there are only going to be 25% to 30% efficient. Those drivers maybe 80%? All that "free" heat will be nice to create a big bucket of luke warm water. Always handy to have that.
 

mahiluana

Well-Known Member
"free" heat
:fire: Growing cannabis indoors uses a lot of power and if you want to keep track of it you can could log that on the Arduino too. ;)

--- must be brainstorming to think to all controll posibilities of a grow.
But what i need urgently is a waterproof temperature sensor and valves to process the heat.

In the past it was me or my girl friend to tell me when to take a shower.
- Now it`s my heatexchanger. :cuss:o_O

When my 80L boiler is filled up with heat (36°C) i use a second circuit that rise humidity and cool down by a room fontain. Next winter - for shure i will have extra-radiators to heat the flat.
:peace::leaf:
 

Shugglet

Well-Known Member
25°C of the cooling water in the coolmac give much better results than 25°C of an aluminium heat sink. 25°C coolmac "simulates" ~10°C of a common heat sink.
This is silly.

I also have doubts about improved efficiency by cutting holes in the bar to have the water in direct contact with the cob. The downsides of that seem to outweigh the potential gains imo.

Im assuming you use those bars to heat your grow room too? Im just wondering if your system heats up to 36c, those bars are effectively radiating heat into your grow, so if youre also cooling your room youre just battling yourself there. Also considering how you like to save the heat I would maybe consider insulating those bars, or using smaller water blocks to transfer the heat.

On second thought, I would definitely insulate or downsize to water blocks and get a radiator to tap into the heated water supply. Much more efficient design. I know you like efficiency ;)
 
Last edited:

mahiluana

Well-Known Member
This is silly.
:eyesmoke: I think it`s clever, easy and cheap. :idea:
Heat conductivity of aluminium is really good (235) - thermal grease much less (4-12)
and the surprise water only (0,55).
The difference is that heat transfer of an aluminium heatsink happens at the end of the fins or let`s say surface of alu. and air. You need a big surface - and weight of your sink must be related to the amount of heat you have to sink.
In the coolmac system the fin and heat transfer is placed at the bottom of the chip platine - so the way out of the system is much shorter.

The alu. heatsink "start" to work when room temp get lower than heatsink temp. (~30°C)
the area directly under your chip will be much higher (?) because there are cm" of heat jam to the surface.

Coolmac make ways much shorter - digging into the chip and lower junction temp.
to a possible 1°C (maybe someone owns a trout pond and want to make some extra kilos
before it gets icy.)

Don`t tell me about condensation, because at the end i don`t even need aluminium or stainless steel for the tubes -
i`m shure you can fix it also with plastic square tubes and sealed leds under a hole.
Should be much cheaper and better than expensive aluminium and grease.:bigjoint:
 

l0wbob2016

Well-Known Member
Very nice, need to say that u got some tricky ideas. Also using the warm water from the COB's is a very nice idea if u got the space and the possebility to set it up for an easy use.

Making that hole in the heatsink to let it cool better is an idea i also played with a lot cause u dont have heattransfer between different materials until the heat gets into the water. I hope it doesnt leak :D

That watercooling sink u made by urself i think, is very nice and if i am not wrong u can do it in any size u want right?

keep the work going
 

ttystikk

Well-Known Member
:eyesmoke: I think it`s clever, easy and cheap. :idea:
Heat conductivity of aluminium is really good (235) - thermal grease much less (4-12)
and the surprise water only (0,55).
The difference is that heat transfer of an aluminium heatsink happens at the end of the fins or let`s say surface of alu. and air. You need a big surface - and weight of your sink must be related to the amount of heat you have to sink.
In the coolmac system the fin and heat transfer is placed at the bottom of the chip platine - so the way out of the system is much shorter.

The alu. heatsink "start" to work when room temp get lower than heatsink temp. (~30°C)
the area directly under your chip will be much higher (?) because there are cm" of heat jam to the surface.

Coolmac make ways much shorter - digging into the chip and lower junction temp.
to a possible 1°C (maybe someone owns a trout pond and want to make some extra kilos
before it gets icy.)

Don`t tell me about condensation, because at the end i don`t even need aluminium or stainless steel for the tubes -
i`m shure you can fix it also with plastic square tubes and sealed leds under a hole.
Should be much cheaper and better than expensive aluminium and grease.:bigjoint:
Condensation was a big issue for me.
 

1212ham

Well-Known Member
the one-of-these-days clu058 water-cooled build. 1 cob per square foot at about 30W will be well over 65% efficient when water cooled. citi tool says pushing 200 lm/W under those conditions. 10 W of heat dissipated over that fat ass 38 x38 mm aluminum chip is cake.

note how the BJB mounting holes fall outside of the waterblock for easy mounting

View attachment 3768053
What temperatures are assumed for 65% efficiency? What might be typical case temps for citi 1212's on pin heat sinks at something like 25 and 50 watts?
I have a design for cooling blocks that will put coolant in direct contact with citi 1212s etc. and I'm just wondering what efficiency gains might be. I just decided I like my design enough that I have to make it, time for cad drawing. How do I properly measure case temps to get some real data? Solder a thermocouple to the cathode?
 

ttystikk

Well-Known Member
What temperatures are assumed for 65% efficiency? What might be typical case temps for citi 1212's on pin heat sinks at something like 25 and 50 watts?
I have a design for cooling blocks that will put coolant in direct contact with citi 1212s etc. and I'm just wondering what efficiency gains might be. I just decided I like my design enough that I have to make it, time for cad drawing. How do I properly measure case temps to get some real data? Solder a thermocouple to the cathode?
How do you plan to seal the coolant so it doesn't leak between chip and mounting block?
 

mahiluana

Well-Known Member
Condensation was a big issue for me.
:peace: Maybe insulate carefully all contacts(lack, silikon) and just let it happen is the easy way.
Waterproofed case to put drivers and contacts inside and sealed lenses to protect the phosphors is more professional.
With condensation on the aluminium you will have more corrosion.
I found a cheap, more or less ecological product made out of sheep wool against corrosion.
They say it protects even against salty water in ship tanks for 20 years.
It´s also perfect for leather and has a bit greasy consistence.
I use it in- and outside the tube - everywhere as a finish.
If you have real waterdropping maybe that: https://www.amazon.de/Antikondensat-Vlies-Schwitzwasser-20m-24m²/dp/B005UY4DLI is a solution - but i never tried.
:eyesmoke:
 

ttystikk

Well-Known Member
:peace: Maybe insulate carefully all contacts(lack, silikon) and just let it happen is the easy way.
Waterproofed case to put drivers and contacts inside and sealed lenses to protect the phosphors is more professional.
With condensation on the aluminium you will have more corrosion.
I found a cheap, more or less ecological product made out of sheep wool against corrosion.
They say it protects even against salty water in ship tanks for 20 years.
It´s also perfect for leather and has a bit greasy consistence.
I use it in- and outside the tube - everywhere as a finish.
If you have real waterdropping maybe that: https://www.amazon.de/Antikondensat-Vlies-Schwitzwasser-20m-24m²/dp/B005UY4DLI is a solution - but i never tried.
:eyesmoke:
Condensation was a problem and also a symptom. The solution to my mind is a different approach. I'm still trying to set up the test rig and we'll see how it works out.
 

mahiluana

Well-Known Member
any size u want right?
:leaf:-THANKS for the flowers -
The wattage is limited at 1800W per lamp system in order to respect your fuse. (~230V)
For shipping and packing it would be the best not to exceed 1,5m tubes.
Just plan a raster for your chips and growroom (quantity of plants is important)
and you can place the chips exactly where you want.
You can hang several single bars in a row or paralell, always connected by a water hose.
The Coolmac is available with al brands of driver and chips - with heat exchanger or without.
:fire: cheerio
 

mahiluana

Well-Known Member
space and the possebility to set it up
I guess many growers have their grow room close to the place where they live.
Many will take a hot shower from time to time:hump:

My waterboiler(heat exchanger) still hangs close to my grow and the way to my bathroom is about 8m.
I connected 2 garden hoses (cold/in + hot/out) to the boiler and take them with me when i take a shower.
I screw off the showerhead of my previous armature and connect cold in/hose. That`s it.

This way you don`t need any changes in plumbing.- Everybody can fit a garden hose - ? :hump:

Since i roll in and out my rollin hoses every day - i feel like a real rollitup firefighter in his daily fight against the lamp fire of my grow room - but also against the planets "big fire" called climate change. If i would live in my own flat - this would be a great reason to do a more fixed installation.

Previous my hot water was prepared by central oil heating.

In the future i will be able to reduce my costs for heating and hotwater down to ~50%.
More or less the amount, that i need to pay for the electricity bill of my grow lamp.

So growing indoor can become a costneutral matter ! :clap::clap::clap::clap::clap::clap::clap::clap:

The whole system is very simple to copy and adapt to your needs, easy to mount and understand.

I think - i`m not the only one, to have the possebility to watch his hot water clock coming down.

:leaf::hug::fire::hug::leaf:
 
Top