Peltier device questions.

[negative37dBA]

Hi all. Wanted to gets some people experiences with Peltier cooling devices. Just got 1 preassembled and pretty sure its setup backwards. Have read conflicting advice as to reversing polarity... imagine that conflicting advice on the net... When I hook it up properly as in Red is positive the heatsink gets ice cold and the water block gets hot...lol. Raised my water temp 3 degrees in 10 minutes or less. Not the effect I am looking for. I reversed polarity and it seemed to do nothing to water temp. The heat sink did get insanely hot despite it having 2 fans installed on it. They run so fast they sound like mini jet fucking engines. So loud and the heat sink so hot I would never trust it that way.
Thoughts? It is a 140 watt unit that uses 2 chips. If I can get it to cool enough and dissipate the heat correctly I may be able to use it. It is for keeping a small, 3.5 gallon, cloner cooler. Thanks for any input you may have.
Peace, negative.

 

[Dr.Mantis]

Peltier chips are very inefficient wattage wise. Additionally, it takes a ton of energy to change the temperature of water. The heat capacity of water is about 10x of iron for example. Meaning, changing the temp of water will take 10x the amount of energy vs iron of the same mass. To change the temp of 5 gallons of water by one degree c takes about 75kJ.

I’ve never used peltier chips to cool a reservoir, but my guess is it’s not the way to go. Is this a thing others do?

 

[Ca++]

75kj is about 8 minutes work for a 140w unit. We can't consider how cold it will be in an hour though, as this isn't really a relevant. viewpoint. We are not trying to cool the water.
We make the tank up at the right temperature, then need to combat heat gain. This means efficient pumps and insulation jackets are on the menu. Then we have to start looking at K values. A mini fridge is often a 50w device, and will lower the contents 20c below ambient. Running the cooler longer makes no difference, as with a 20c differential across the fridges walls, 50w is coming in through them. The greater the difference between inside and out, the more heat will come through. Hence the insulation is such a critical part. The cold line back to the tank needs some thought. Just like any fridge motor.

It's really not worth doing the maths for most of us. Just trying it is so much easier. 140w isn't nothing.
If we look at a domestic fridge, it might have 1000w of cooling. It can get down near freezing even if it's 50c outside. That is good insulation, and the fridge can do this quite quickly after shutting the door. Most of the time, they are not even running. This might mean making a cardboard box some 50mm bigger then the res. Then bagging the res, putting it in the box, and spray foaming the gap. The bag was so you can take the res out still, rather than foaming it in permanently.


The thermo electric coolers do tend to work in either direction. So the heatsink should of received the 140w of power consumption, and 140w it pulled from the cold side. My 30w units gets condensation on it quickly.

I guess it's 12v ? you can run them on less, or go pwm to control them. The pwm boards with solid state relays are like $20 to regulate the cooling. PC use 12v pwm based cooling for fans, for a few dollars, and the ssr is just a few more. You might do it for 10. Then it's only as hot as it needs to be.

 

[negative37dBA]

Thanks for the thought guys. Might do a couple test setups to see if it will be worth it. I really need just about 3 degrees of cooling in less than 5 gallons and only in the summer. I normally clone in the winter but I am forced to do so now because of situations. Going to run it for a bit in a sample bucket to see if I can dial it in.
Thanks again. negative.
It is 12v...another question. Is it really going to pull 10amps like it says? That has to be different with 12v cause there is no way the wires could handle 10amps 120v. Must be 10 times less.

 

[Dr.Mantis]

75kj is about 8 minutes work for a 140w unit. We can't consider how cold it will be in an hour though, as this isn't really a relevant. viewpoint. We are not trying to cool the water.
We make the tank up at the right temperature, then need to combat heat gain. This means efficient pumps and insulation jackets are on the menu. Then we have to start looking at K values. A mini fridge is often a 50w device, and will lower the contents 20c below ambient. Running the cooler longer makes no difference, as with a 20c differential across the fridges walls, 50w is coming in through them. The greater the difference between inside and out, the more heat will come through. Hence the insulation is such a critical part. The cold line back to the tank needs some thought. Just like any fridge motor.

It's really not worth doing the maths for most of us. Just trying it is so much easier. 140w isn't nothing.
If we look at a domestic fridge, it might have 1000w of cooling. It can get down near freezing even if it's 50c outside. That is good insulation, and the fridge can do this quite quickly after shutting the door. Most of the time, they are not even running. This might mean making a cardboard box some 50mm bigger then the res. Then bagging the res, putting it in the box, and spray foaming the gap. The bag was so you can take the res out still, rather than foaming it in permanently.


The thermo electric coolers do tend to work in either direction. So the heatsink should of received the 140w of power consumption, and 140w it pulled from the cold side. My 30w units gets condensation on it quickly.

I guess it's 12v ? you can run them on less, or go pwm to control them. The pwm boards with solid state relays are like $20 to regulate the cooling. PC use 12v pwm based cooling for fans, for a few dollars, and the ssr is just a few more. You might do it for 10. Then it's only as hot as it needs to be.

He mentioned he wanted to use it as a cooler on a cloner? Your right that 140w isn’t nothing, but getting high efficiency from a TE cooling system isn’t easy, so it’s not going to be 100% of 140w, probably much lower unless some big steps are taken. Additionally, if the heat sinks is directly on the backside of the chip without some sort of exchange mechanism it’s all going to be dumped around the chip, further reducing efficiency.

From a practicality standpoint, to me it seems easier to salvage a compressor and coil from a junk fridge and use that. I guess it depends how involved one wants to get.

 

[Ca++]

W=VxI buddy. I would think more than 10 amps. 10 is just 120w

You could forget about cooling the res, and just cool the feed to the pump?

Fridge could be used in it's entirety. Buy a coil of soft copper, and just put it inside. Use it as a table to put the cloner on

I like the idea of the TEC. I reckon it can be made to work. If the 10 amps is causing a stumble, you can buy an old ATX supply for a tenner. The sticker tells you what the 12v lines will give, and 10 is quite low. IIRC there is only one green wire, an you touch it on the case to bring the power on.

 

[negative37dBA]

Just did a test on it for 7 minutes. I guess it is built wrong cause everything gets hot...water block and heatsink both. Raised a 4 gallon bucket 4 degrees in 7 minutes..opposite effect of what it is supposed to do..wow. Heat sink is right on the chips and water block is directly on the other side of the chips. If I hook it up the other way, reverse polarity, then the heat sink gets ice cold and condensates but the block heats up...I mean wtf?
Dont think this is going to work. Yes it is on a cloner or going to be..the cloner is only 3.5 gallon, has no rez. Just a bucket with a small pump.
Peace, negative.

 

[Ca++]

Put your bucket in a bigger bucket, and fill the gap with expanding foam. Bag both, if you don't want the foam to stick.

I presume you have two TEC's in this unit, and you are switching the power around on both? If you just switched the power around on one, it would behave as you describe.

$5 no bids, 15 amps at 12v https://www.ebay.com/itm/155652319388

I'm really interested in how this turns out, and feel sure it's just teething troubles

 

[negative37dBA]

Did some more playing with it last night. Pulled it all apart. I have some MUCH better heat sinks that are larger as well. Going to try to use them and a bigger 140mm dual fan setup to help remove the heat. The ones they used were like 50mm May be able to get this functioning. I will keep you updated. Thinking of using a old cooler I have with a removable top for a cloner. Already insulated. Maybe a Igloo water cooler...like you see on jobsites and in work trucks.
Getting to be a bit more than I wanted to do because I so rarely have issues. Could be a fun project and learn some in the process tho.
Peace, negative.

 

[Ca++]

If you have an old cooler, that would be a great start. Most electric one's have the cooling bit in the lid, so condensation drips away from it. Which is perhaps less use, as it's cooling the air, not the water. Just the base of a normal one would be good though.

Have you any sign board? The corrugated stuff could be layed up like ply. Giving you an insulated deck, with a bit of rigidity. If you cut the holes in each ply before laying them, you could make the holes different sizes. Bigger on the top sheet, then those below. Forming a recess for your foams.

We can read the cooling specs of coolers online. Insulated well, you won't need 140w to cool a few degrees. Most of these things are 50w and still it's too much cooling.
The 12v power wires to the TECs could take a 12v motor speed controller of the typical pwm style. $10 board. As I think this will be quite a lot of cooling.

25L box here, with 50w TEC, says it will hold at 5c (I think that means 15c below ambient)

Your lid might not be as sealed, but your not looking for beer temperatures. I think your project weighs up well. I might copy it...

 

[negative37dBA]

I could always just drop down the power and use only 1 chip. Just scale it down a bit. Was just out in my shed eyeing up old coolers I have...lol.
Peace, negative.

 

[Ca++]

I could always just drop down the power and use only 1 chip. Just scale it down a bit. Was just out in my shed eyeing up old coolers I have...lol.
Peace, negative.

Sounds like progress

Any old fuel hose out there? I have some high pressure stuff, that's only about 5mm bore, yet 20mm external. Big thick walls, that might offer some insulation. While the rubber used should be quite resistant. Insulating the hose isn't easy while prototyping.

Did you get any of them ~3 DC pumps? Thread around here somewhere. Lots of pressure to get the flow going, but little power use, which is just heat. It could stay as a submersible, but work inline after priming. I'm constantly thinking I could of got a smaller one, which isn't normal after buying a pump lol

 

[negative37dBA]

Set it up again. When I tore it down they had assembled it backwards...hot side to the water block...what retards from China.... sometimes I think it is on purpose. I used thermal tape and 2, 4 foot long heat sinks I have from a LED light build on the hot side. I also used the heat sinks that came with it and thermal grease on the cold side of the water block to help even more. 2 x 140mm fans blowing over it now. Set my timer for 20 minutes and we will see what happens. Non insulated lines but they are very short and the bucket is insulated with the silver foam bubble stuff...cant remember the name. Like reflectix or something like that. It is in a 72 degree room and using a very small active aqua sub pump

 

[negative37dBA]

Ok. So now I am getting slightly frustrated and more confused. It took the temp UP by a degree in 20 minutes. I know I have this hooked up right. I am very experienced at DIY. Among many other things built my own 630 watt PLC LED light from scratch. It just is not working. The big sinks got pretty hot and the cold side ones actually felt a little cool but the temp probe tells the story. A good temp probe to, one from the short path days, so down to a tenth of a degree.
Watched a couple videos and this is exactly how I see it used minus some fancy controllers to keep temp at a certain level...I dont need that. Going to shelf this for now, maybe come back to it later. Cant see why what I see in the video I cant recreate. That said...the guy only got 3 degrees of cooling from 12 gallons and he used a $217 dollar controller...not worth it. This was with 1, 40mm chip...I have 2, 50mm.
Thanks for all the comments and help.
Peace, negative.

 

[Ca++]

When you come back to this, try with one chip. When you know that ones right, swap to the other. Get each working individually, before pairing them up. Perhaps one is faulty, or even has it's wires the wrong way around. Only running them individually will see.

 

[Hiddenjems]

What problems are you having? My cloners work best between 70-75*. I keep the air in my rooms at 70-72*. That keeps my cloners in the right range with constantly running pumps in them.

Ever since I started using southern ag garden friendly fungicide, I haven’t had a single issue with rot.

 

[Ca++]

Most people with these run lower temps to avoid stuff growing in them. The thing is.. we are trying to grow stuff in them. The warmer you can get them (say 77f/25c) the faster our roots grow. So we are at odds with our own needs, running them cool. Things like bleach are probably the real answer. There are people running these at decent temperatures, when using additives to stop the bio-life. Another option is UV filtering, but it doesn't play well with the feed.

Southern AG have done a couple of useful fungicides. They have been bio products, and shouldn't be over-dosed. People have found the product itself can be a problem, if too much is used.

 

[Hiddenjems]

Most people with these run lower temps to avoid stuff growing in them. The thing is.. we are trying to grow stuff in them. The warmer you can get them (say 77f/25c) the faster our roots grow. So we are at odds with our own needs, running them cool. Things like bleach are probably the real answer. There are people running these at decent temperatures, when using additives to stop the bio-life. Another option is UV filtering, but it doesn't play well with the feed.

Southern AG have done a couple of useful fungicides. They have been bio products, and shouldn't be over-dosed. People have found the product itself can be a problem, if too much is used.

It’s so concentrated I typically run 1-2cc’s in a 50 gallon res every 5 days.


I used bleach for years. The only problem I had was keeping the correct ppm of chlorine. It can change really fast with Rez conditions. The southern ag keeps the roots healthy across a wide range of conditions, and it doesn’t get used up if there is an issue.

The best solution I’ve seen on an industrial scale is chlorine tanks tied to sensors and a dosing solenoid. This eliminates and salt issues from using sodium hypochlorite.

 

[negative37dBA]

What problems are you having? My cloners work best between 70-75*. I keep the air in my rooms at 70-72*. That keeps my cloners in the right range with constantly running pumps in them.

Ever since I started using southern ag garden friendly fungicide, I haven’t had a single issue with rot.

I am having no real issues. Just through 1 week of cloning here in 100 degree temps...gets hot in the dez. Cloner stays at 75 all the time so it is going fine. Was just trying to see if a small investment could bring it down a couple degrees. Also thought of using this in the summer in my Flood and Drain rez.
I have had great success cloning for several years now. Simple spray cloner in a 5 gallon bucket. I use my own home made clear rez so that keeps the water fine.
Peace, negative.

 

[Hiddenjems]

If my cloner gets under around 71ish degrees it takes about a week longer for my clones to be ready.