Homemade Peltier cooler for small DWC setup?

[Shafto]

Anyone ever do this? A peltier on a low voltage DC circuit and a thermostat to turn it on/off would be a great low power solution I'd think.

Put a heatsink and a PC fan on the hot side, silicone the cold side into the wall of your resevior, or even just on the side through a thin plastic wall would work.

http://cgi.ebay.ca/TEC1-12706-91-2W...271?pt=LH_DefaultDomain_0&hash=item3a5ff6c627

Simple solid state devices. Tmax is 67C, which means that if you can keep the hot side near ambient with a heatsink and fan, the cool side will go 67 degrees cooler than ambient. Not of course with all that warmer water around it, but it will provide nearly 50W of solid state reliable cheap cooling for a few dollars. Just need a beafy 12V supply on hand.

I'm going to give it go, just thought I'd check to see if anyone else has done so before me. I haven't seen it yet.

 

[LastWaltz]

I have used them before to cool co2 flux measurement chambers and they were sooo easy to work with. I say go for and please post some pictures and some temp difference results, if it goes well I'll be taking apart some old field research equipment lol.

 

[dansbuds]

Yes please let us know the results of this . temps in individual buckets are my reason for not going with DWC this time .... i have no place for a rez cuz of being stealth . if these work i could put one on each one of my buckets .... 3 gallon would probably cool nicely . hmmmmmmm

 

[David762]

What a great idea! I think I will set up something too.

What a great idea! I think I will set up something too.

I think you are onto something here. I have seen surplus Peltier devices with heat sinks attached, just begging to go into some project or other. Using a few of these as a substitute for a swamp cooler or (bloody expensive) split AC unit in a grow room may be just the thing to knock the temp down by 20 degrees.

Anyone ever do this? A peltier on a low voltage DC circuit and a thermostat to turn it on/off would be a great low power solution I'd think.

Put a heatsink and a PC fan on the hot side, silicone the cold side into the wall of your resevior, or even just on the side through a thin plastic wall would work.

http://cgi.ebay.ca/TEC1-12706-91-2W...271?pt=LH_DefaultDomain_0&hash=item3a5ff6c627

Simple solid state devices. Tmax is 67C, which means that if you can keep the hot side near ambient with a heatsink and fan, the cool side will go 67 degrees cooler than ambient. Not of course with all that warmer water around it, but it will provide nearly 50W of solid state reliable cheap cooling for a few dollars. Just need a beafy 12V supply on hand.

I'm going to give it go, just thought I'd check to see if anyone else has done so before me. I haven't seen it yet.

 

[supermanlives]

i have been wondering too. i have a setup from a cooler,heater for cars.

 

[geopolitical]

They sell a commercial salt water cooling rig for small (2-10 gallon) tanks that's based off a peltier called an Iceprobe.

 

[Shafto]

Thanks geopolitical!

That's exactly what I was thinking of and confirms my design idea.

I'll be doing mine on the cheap. Probably build my own temp controller with a microprocessor and either a thermocouple or a digital temp sensor. I have most of the parts lying around. $120 +$60 is too steep!

 

[hazydreams]

a while back i was considering using a peltier cooler with a stainless stell wort plate chiller.


drawing, if some one pulled this off with like 500 watts of peltier coolers i think it would work awesome!

 

[StealthDragon]

nice idea. I have a 12vdc car "cooler" with a peltier that I've been thinking about converting to a micro watercooling thingie..or cooling a small rez with it.

I'm interested in what you work out.

 

[durka_durka420]

best way would be to mount it to a cpu liquid cooling system and then have a stainless coil full of some of the pc coolant would work great dont use copper with any contact inside resivour as it will make the copper release bad shit.......you can also use a dissassembled water cooler (older one or newer one has to have a compressor) and put the water chilling probe in the resivour....if the tubes are copper plastidip works wonders

 

[Shafto]

Stainless is a terrible thermal conductor.

Anodized aluminum is the best choice for corrosion resistance and thermal transfer, well, aside from pure diamond, silver plated copper would be best, but that's a little extreme.

As long as there's flow through the resevoir you won't need to use any kind of block. In doing so you would lose some of your cooling to the air. The moving water will transfer the cold away from the peltier better than just about anything. Best off just to submerge the cold end.

I can't wait to try it out.

 

[calibob]

peltier elements are ususally meant to lower dew point of gasses not liquids. Typically it takes two 10 inch glass impingers to lower a sample that contains less than 25 % moisture by volume to a dew point of 41 F at a flow rate of 8 liters per minute. The power supply is usually 24 vdc but relatively high current. Also the heat sink usually requires a cooling fan. So if you can keep the flow rate down and use glass impingers mounted in aluminum blocks you stand a chance. I don't know if you will need proportionate control of current flow to the element but you might. Good luck, it's worth a shot.

 

[Shafto]

I suppose they are usually used in your field to do that.

Thermocouples have many uses, from measuring tempature, creating electricity, creating cold, and creating heat. They aren't usually used for anything. They're used for a ridiculous amount of things.

I bought my mom a peltier effect wine chiller a couple years ago.


Using any type of glass would only be beneficial with a caustic liquid of some sort, or something that can't be contaminated, otherwise it will just hinder thermal transfer.

It's a simple device with only a simple approach required. Hot end outside the res with a heatsink, cold end in the res.

 

[calibob]

The wine cooler chiller works because the liquid being chilled is not moving. Once you put the liquid in motion it becomes alot harder to lower temps especially if you are running it through a pump. Glass tubing coated with a thermal transfer compound and mounted in an aluminum block with the element mounted on it is an excellent method of cooling. Glass does not spread heat evenly like aluminum but glass in a chilled aluminum block works great, like a wine bottle.
Most Peltier elements I have seen have been flat devices with a hot side and a cool. How are you planning to put the cold side in the res without screwing up the element. By the way if you need sources of components to try (surplus) I can steer you in the right direction.
All this is based on the assumption you have a fair size res to chill. I try and stay away from metals because they corrode and when they do the thermal transfer goes away. Most ferts contain salts so unless your running organics you will affect the metals.

 

[Shafto]

Anodized aluminum will not be affected.

Using glass will add thermal resistance, and will make the system much less efficient. Moving the liquid also makes the cooling more efficient.

Moving the liquid over the cool element ensures there is a higher thermal delta which ensures more thermal transfer.

Thermodynamics aren't foreign to me, much of my profession revolves around designing heatsinks.

Did you look at the link posted? It will be quite easy to implement. Cool the hot side, extent the cold side into the res with a thermal conductor (aluminum) with enough surface area for proper thermal transfer into the fliuid (don't need much surface area with fluid, especially moving fluid, unlike air).

 

[calibob]

I'm pretty sure moving liquid through a pump head raises temp, unless you cool the pump head. Now that might be the place to cool, the pump head. The delta by the way is dependent on the differential temp, what will happen is instead of cooling the water you will be in effect warming the transfer plate. A well known chiller company out of Reno tried Peltier for cooling Lobster tanks and decided to go with refrigerant instead. For less than 100 bucks a refrigerant unit can be configured. But like everything, it all works on paper. Once you get it up and running please update with results, I honestly am interested.

 

[Shafto]

How much the pump raises the temperature of the liquid is completely indepenent from this device's operation.

It will raise the temperature to pump liquid, but that has nothing to do with that I said.

Moving a medium over a hot or cold surface will improve heat transfer. This is why you put a PC fan on a heatsink, you move the air over the heatsink, just as you should move the water over the "coldsink" in the res. You can rely on natural convection currents created by the temperature difference, but this is not as effective. It doesn't matter that the fan causes friction and slightly warms the air above ambient passing through it, the efficiency of thermal transfer for your given hot or cold surface area goes way up.

Warming the transfer plate is the same thing as cooling the liquid. Both are happening at the same time and are the same thing. Glass half full or half empty if you will.

Thermodynamics are not always easy to grasp. I am well aware of what a delta in temperature is "by the way" please don't speak down to me. If you do not move the medium over the surface gradient layers form lowering the delta between a layer of "insulation" created in the higher thermal resistance medium, as compared to the surface. This lower delta causes less heat transfer, just like a lower electrical voltage poses less potential for conductance.

If you are interested you should do some more research on thermodynamics.

 

[Killy]

Any news on the use of Peltier for cooling air in grow box.

Can TEC be used with two CPU coolers attached to cold and warm side, to cool down CPU cooler on the cold side.

 

[padyakol]

can any of you guys post pictures of your successful DIY chiller/reservoirs? I am currently doing one now and will post pics soon. I know this is an old thread and i am hoping you guys are still growing. I'm a newbie and i will try out NFT. will the aluminium react to the tds and ph of water? this is my only concern. i got a mini fridge peltier cooler that i am adapting to my 6L reservoir.

 

[Maddox]

Any news in here? Trying to figure out an easy way to cool out a 8 gal res, checking out and still the iceprobe seems to be the only viable option, and the price aint gone much down.