Exploring Below Ambient Water Cooling

[4Ryan6]

I have a 4 or 5 gal cooler from Canadian tire and really impressed how cold it keeps things, so I’m thinking maybe I could cool flowing water. While searching out info on the Pelteir cooler I ended up here.
I want to cool the water in my boats live well down to about 40 degrees or less from about 70 degree lake water. It holds about 4 gal of water and I have 30 amps at 12 volts to play with from my outboards charging system. I can cool the hot side with lake water, insulate the tank and run lake water through the cold side into the fish tank. Cool fresh lake water running into the live well would be just awesome.
There are some discouraging opinions here. Is this idea worth trying out?

Thanx much
Mikey

You might want to start here > http://www.tetech.com/index.html the website has a heat load calculator to give you some idea of what it would take to accomplish your goal.

At the above website you can buy fully functioning units already assembled, I did not go that route because I was not sure in the beginning the Peltier cooling would be successful and additionally most of their units require higher than 12v operating voltage.

I imagine the live well has fresh lake water entering then an outflow line at the 4 gallon mark that allows water to leave the tank always keeping fresh water inside.

Cooling the hot side with lake water is a feasible idea simply because you'll be porting that heated water back to the lake, so you will not be heating your pump and would have no need for a radiator.

I have not run any tests to see exactly how low I could get the water temperature in the Reservoir, but I do know 50F or 10C is doable with my setup running 4 gallons of water in the reservoir.

Here's the unknown, the variable ratio to the Peltiers cooling capability to counter the heat generated by the CPU which fluctuates and increases it's heat output under load, vs cooling the constant temperature inputted from the lake, it may very well work better with your proposal than my setup.

I'm presently at this point in question mode myself, what works better?

One large high wattage/amperage peltier like I'm presently running or a multitude of low wattage/amperage peltiers running in series, that's letting the cat out of the bag as where I'm presently considering my next experimentation route, basically to see if I can achieve the same results on lower power draws.

Most are not even considering the peltier I'm presently running a 50mm x 50mm is size wise 20x the peltiers that were used when the first direct on the CPU cooling took place, and the larger the peltier the more power it takes to run it.

There are a massive amount of peltier sizes and load ranges available, far more than I had a clue about when I first started this cooling adventure!

What you have to pay close attention to is the voltages they run on, because that is widespread different also.

But back to your idea of cooling your live well with a peltier setup, I think it's definitely possible and could possibly be something other fishermen would be interested in also!

This website has a lot to look over as far as doing it yourself > http://www.shop.customthermoelectric.com/main.sc
maybe it would help me to help you if you could toss out some questions and ideas of how you may proceed after looking over the 2 websites.

Ryan

 

[4Ryan6]

Present setup picture update.

This is a good side view of the Peltier sandwich, the Thermalright TRUE is on a cold plate I made from aluminum bar stock, 2" x 5/16", cut square 2" x 2" sitting between the TRUE and the hot side of the Peltier.

The Swiftech Apogee XT rev2 water block is upside down contacting the cold side of the peltier, the return from the CPU is going through the TEC assembly before dumping into the Reservoir Cooler, picking up the cold from the peltier.

Here is a Larger View of that picture.

Details of assembly.

There's only 1 gallon of water in the cooler, don't ask me what that white is on the temperature sensing line, it must be camera flash?

The above is a full view of the tubing routing.

 

[4Ryan6]

The pumps output line is split into 2 flow outputs, one goes straight to the Diverter valve for TEC cooling only, the second line goes through the 360 radiator then to the Diverter valve, allowing cooling by traditional radiator only.

The Diverter valve allows which cooling line to run, either the traditional radiator cooling line which is above ambient cooling, or the TEC cooling line which is below ambient cooling.

All of the CPU loop lines are insulated, the GPU loop is completely independent of the CPU loop and is solely cooled by the 240 radiator and there is no need to insulate that loop.

Threw this pic in with the Tom's Hardware Hammer for the heck of it!

CPU load temperature using the 360 Radiator loop cooling is 49c.

CPU load temperature using the TEC loop cooling is 35c with a water temperature around 13~15c.

I'm presently running the TEC loop, my water temperature while posting this is 10.2c, ambient room temperature is 22.6c. CPU idle temperature is 13c, CPU load temperature at a 10.2c water temperature is 33c.

That's why I said earlier this cooling solution is the best of both worlds, if I'm gaming or running benchmark or stress testing, I'll run the TEC loop, if I'm surfing the net and such as that, I'll run the rad loop.

Ryan

 

[mikolrayne]

Hey Ryan,
Excellent work, as usual!
I really like the diverter valve part.
Your temps are pretty cool too. Pun intended, of course.

Hmm, anyway theoretically speaking, it's correct to assume that if we add a rad to a TEC loop, the rad will be trying to raise the water temps to ambient temps right?

 

[4Ryan6]

Hey Ryan,
Excellent work, as usual!
I really like the diverter valve part.
Your temps are pretty cool too. Pun intended, of course.

Hmm, anyway theoretically speaking, it's correct to assume that if we add a rad to a TEC loop, the rad will be trying to raise the water temps to ambient temps right?

It's not theoretical, that is exactly what happens, the radiator has a reverse effect on the chilled water, it warms it, so under those circumstances you have 2 types of cooling working against each other.

That's what the valve is for, to isolate each cooling loop from each other.

All the radiator has to cool with is ambient room temperature air, if the ambient air is much warmer than the water running in the system, the radiator warms it instead of cooling it.

I discovered that way back in the beginning of this thread when I was using ice in the cooler, and ran a radiator side loop and opened both lines at the same time.

The radiator started warming the water in the cooler with ice in it, and wasn't messing around about doing it either.

The peltier yields below ambient temperatures and you are paying electrical wise to produce those temps, so you want to preserve the cold as much as you possibly can.

That's the reason for the insulated cooler/reservoir, all the tubing being insulated, and definitely for the 2 isolated cooling loops.

 

[mikefix]

Hay Ryan

Thanx for the support and taking care of my email . I have't checked the sites you gave me yet but will do this soon. I have prolems getting around this forum, about an hour to reply to you and hope this is right. When I order my tecs I'll be ordering one around 400 w and 4 around 100 w ,thinking ,with 4 tecs I could make the surface area greater on the cooler side, then with 1 400 w and give the water flow alittle more time to cool for the same amount of current.
You mentioned putting a multitude array of tecs in series, wouldn't that drop the voltage down across each one equally ,then giving you less performance

Thanx
Mikey

.

 

[4Ryan6]

Hay Ryan

Thanx for the support and taking care of my email . I have't checked the sites you gave me yet but will do this soon. I have prolems getting around this forum, about an hour to reply to you and hope this is right. When I order my tecs I'll be ordering one around 400 w and 4 around 100 w ,thinking ,with 4 tecs I could make the surface area greater on the cooler side, then with 1 400 w and give the water flow alittle more time to cool for the same amount of current.
You mentioned putting a multitude array of tecs in series, wouldn't that drop the voltage down across each one equally ,then giving you less performance

Thanx
Mikey

.

Hi Mikey, When I mentioned multiple TECs in series I wasn't referring to their wiring, just a reference to multiple TECs, they would need to be wired in parallel.

 

[toolmaker_03]

Well I would like to add some insight to this. The TEC pad itself is an electrical component and there for is controllable by the voltage going to it, of course at some point the voltage will drop so low that the TEC will not activate at all. The lower the voltage the lower the wattage being used and less heat will be produced, what I am getting at is you could feed your power through a 50 AMP regulated speed controller. This would be controlled by an ardino micro controller; a simple PWM program can control how much voltage to give TEC’s and an analog temp program can control that by the temp received by the ardino. In other words you could set a hysteresis of let’s say 5C to a temp boundary of 13C to 18C and have a delay between readings of something like 2 to 3 minutes. example let’s say that the temp in the room is at 12C (its winter) the system would be off, once the temp reading on the ardino hit’s above 18C the speed controller would allow a 7V constant through to the TEC, and each time it takes a reading that is above 18C it would increase to voltage by let’s say 2V, until it reached its max at 15V. The system would hold at 15V until the temp reading is below 13C, and each time it got a reading of less than 13C it would drop the voltage by 2V again until the system gets below 7V at with point it will shut down the voltage all together. While I find your mechanical approach to the Tec interesting I think that this might work as well, but this will add a little extra cost to the project, and that is a factor to be considered as well.

 

[4Ryan6]

Well I would like to add some insight to this. The TEC pad itself is an electrical component and there for is controllable by the voltage going to it, of course at some point the voltage will drop so low that the TEC will not activate at all. The lower the voltage the lower the wattage being used and less heat will be produced, what I am getting at is you could feed your power through a 50 AMP regulated speed controller. This would be controlled by an ardino micro controller; a simple PWM program can control how much voltage to give TEC’s and an analog temp program can control that by the temp received by the ardino. In other words you could set a hysteresis of let’s say 5C to a temp boundary of 13C to 18C and have a delay between readings of something like 2 to 3 minutes. example let’s say that the temp in the room is at 12C (its winter) the system would be off, once the temp reading on the ardino hit’s above 18C the speed controller would allow a 7V constant through to the TEC, and each time it takes a reading that is above 18C it would increase to voltage by let’s say 2V, until it reached its max at 15V. The system would hold at 15V until the temp reading is below 13C, and each time it got a reading of less than 13C it would drop the voltage by 2V again until the system gets below 7V at with point it will shut down the voltage all together. While I find your mechanical approach to the Tec interesting I think that this might work as well, but this will add a little extra cost to the project, and that is a factor to be considered as well.

Thinking and hypothesizing ideas is great, but thinking something or idea will work, and it actually working, are completely different.

There are a lot of ideas I thought would work and found out the hard way they wouldn't, there is a solid fact regarding a peltier that you are obviously unaware of, that needs to be factored into your theory.

That fact is if the hot side isn't powered with enough voltage to maintain it's heat to the point the cold side starts to affect the temperature of the hot side then the peltier performance seriously degrades.

The peltier hot side has to get hot to produce the cold effect of the other side, that's a fact!

Using voltage regulation is a great idea as far as thinking you'll be saving electricity, or attempting to maintain a certain constant temperature level, unfortunately the peltier will not cooperate with your idea once the voltage drops low enough for the cold side of the peltier to affect the hot sides temperature.

It will literally stall out and stop functioning.

While I find your mechanical approach to the Tec interesting

While my mechanical approach to you may be interesting as you say, it actually works!

Now if you want to put your ideas to actual testing facts at your expense with some hard tested proof, and pictures to go along with it be my guest.

But how about do it in your own thread because from the very beginning of this thread you have inputted some serious left field ideas, and I would appreciate it if you started, and did it in your own thread!

Additionally I have attempted to keep the pictures posted in this thread very manageable to other THGF users that may have slower internet connections, in your other threads you show no concern for that, thus the reason for the suggestion to start your own thread with your actual tested ideas, and facts, and pictures, not just theories.

Thank You Ryan

 

[mmaatt747]

Getting what you want from the TEC is a balancing act, you have to let the hot side get hot, for the cold side to get cold.

I know that sounds crazy but that's how it works

Crazy? It actually makes my head hurt, hehe. I keep wanting to rationalize that if you cooled down the hot side it would in turn cool down the cold side as well.

Did you know if you can externally apply the hot and cold to a peltier plate it's opposite reaction is to produce electricity?

I've heard the auto industry is trying to apply this to electrical systems in cars. By using the extreme heat from the exhaust headers in conjuction with the much cooler engine coolant, they can generate electricity. Internal combustion engines generally lose somewhere in the neighborhood of 35%-45% of their energy to waste heat and this would do away with the extra load on the engine from the alternator.

Anyway, I've really enjoyed this thread. I've only ever used ambient air cooling in my computer builds and this is really fun to read.

 

[4Ryan6]

Crazy? It actually makes my head hurt, hehe. I keep wanting to rationalize that if you cooled down the hot side it would in turn cool down the cold side as well.

When I first started with the peltier my first hot side cooling was using the ice chilled water to cool the hot side, that went no where as the chilled side immediately overrode the hot side, which at first made me think I had bought a bad peltier.

Then I switched to the Thermalright air cooler bringing my cold on the hot side to above ambient temperature levels and the peltier performed well, it was after that searching for information on the net that I discovered it was a common occurrence.

Believe it or not the Thermalright TRUE cools so good when I had dual fans running on it, it could actually impede the hot side performance, I'm only running one cooling fan on the TRUE now, and the hot side/cold side balance is as close to perfect as you could ask for.

Anyway, I've really enjoyed this thread. I've only ever used ambient air cooling in my computer builds and this is really fun to read.

Thank You!

Sorry it took so long to respond! Ryan

 

[4Ryan6]

Something new to share with you all, I've been making statements regarding the peculiarities that come with peltier cooling, for instance the lets say for reference purposes the hot/cold reality.

The peltier is a simple device you run electricity through it and the reaction between the two dissimilar metals results in one side getting hot and the other getting cold, if you don't cool the hot side it gets too hot, overheats, and burns up, simple as that.

Depending on the voltage the load temperature of the peltier, (and load temperature is when voltage is applied period!), and it varies from spec'd model to spec'd model.

It would seem the best solution would be to keep the hot side as cold as possible, but that's not the case, you have to allow the hot side to get hot for the cold side to get cold, but since we're only talking a approximately 1/8th inch thick plating or 4mm, that's a very close distance to play the hot/cold balancing act.

Simple discovery the hot side can override the cold sides effectiveness but news flash the opposite can also happen, the cold side getting too cold, can override the hot sides effectiveness, What?

So you're after the balancing act of hot and cold to get what you're after.

My latest experimentation has been to swap out various CFM rated 120mm fans on the heat sink cooler cooling the hot side in an attempt to control the cold side output.

I now have the perfect balance which has allowed,> I gamed for 3 hours and 15 minutes yesterday and during that time period my water temperature not only held my full load gaming, but actually dropped the water temperature in the cooler by 5.2c during the 3hr and 15min of gaming, (running in SLI with the CPU overclocked to 4500mhz and hyper threading enabled!).

So I am no longer heating the water with a gaming load, my setup has now overcome that!

My setup has been below ambient since the thread was reopened, using the very same peltier I started this TEC cooling adventure with!

Through much experimentation I have discovered the perfect amount of CFM fan air flow for the heat pipe heat sink I am using, however any heat sink vs air flow would have to be discovered for what ever heat sink you may use if you go this cooling route.

I have also discovered the maximum cooling efficiency of the heat sink itself, (and believe it or not), it's setting the fan up as a pulling setup on the heat sink, pulling eliminates any fan motor heat added and the dead air space the fan motor housing creates, and allows full unobstructed air flow being pulled through the cooler.

My tests have fully confirmed that, but I'm sure there are those of you that disagree, seeing as how you may as I do have a heat sink with no heat pipes in the fan motor body area, it still makes a difference so test it yourself.

As a push fan setup, (which single fan air coolers are suggested being setup!), the fan motor body still creates a dead air space, there's no getting around that with the present fan designs, I don't care if it is a heat pipe heat sink or a radiator, it still happens.

Ryan

 

[4745454b]

You're saying pull works better? At least for this setup? Whats on the other end of the heatsink? Wall, open room, window?

 

[rubix_1011]

Less than 8w of power the fan is drawing is making a difference in the overall cooling performance? Seems like that wouldn't be enough draw/output to make a difference considering the overall amount of heat watts being dissipated.

What about using a shroud to eliminate the dead spot?

 

[4Ryan6]

You're saying pull works better? At least for this setup? Whats on the other end of the heatsink? Wall, open room, window?

Open room of ambient air.

 

[4Ryan6]

Less than 8w of power the fan is drawing is making a difference in the overall cooling performance? Seems like that wouldn't be enough draw/output to make a difference considering the overall amount of heat watts being dissipated.

What about using a shroud to eliminate the dead spot?

Since the Thermalright TRUE heat pipe cooler can actually totally handle the total heat load created by the peltier, the only purpose of the cooling fan is to control how hot, the hot side of the peltier is allowed to get.

With the fan setup pulling air through the heat sink the dead spot is already eliminated so a shroud is not necessary in this setup configuration.

 

[cbrunnem]

Since the Thermalright TRUE heat pipe cooler can actually totally handle the total heat load created by the peltier, the only purpose of the cooling fan is to control how hot, the hot side of the peltier is allowed to get.

With the fan setup pulling air through the heat sink the dead spot is already eliminated so a shroud is not necessary in this setup configuration.

i definitely can understand why a pulling fan is better. unless you mean better for another reason other then cooler temps. looking at it from a fluids sense when the fan is pulling the air is much less disturbed. more laminar flow but when you push it becomes more disturbed and has less ability to absorb heat due to lower cfm. its like a big box fan, air goes in smooth but comes out very disturbed.

 

[4Ryan6]

i definitely can understand why a pulling fan is better. unless you mean better for another reason other then cooler temps. looking at it from a fluids sense when the fan is pulling the air is much less disturbed. more laminar flow but when you push it becomes more disturbed and has less ability to absorb heat due to lower cfm. its like a big box fan, air goes in smooth but comes out very disturbed.

Yes, much smoother airflow!

 

[4Ryan6]

Gamed some last night and the water temperature drop during gaming is consistent!

Which is much better than I had even hoped for!

If the gaming load could be equaled and just held the water temperature at a constant, I would have been totally satisfied, but to game and still be able to drop the water temperature makes me ecstatically happy!

Started up this morning the water temperature was 18.4c which is 4c below ambient room temperature as of this posting the water temperature is 14.0c a 4.4c water temperature drop in 1 1/2 hr. at 8.4c below the start up ambient.

However it was 30f outside this morning so the heat came on in combination of the TEC heat the ambient room temperature is now 25c which means the actual water temperature presently is 11c below ambient and dropping.

So this update is for those that are interested! Ryan

 

[Motopsychojdn]

You definitely seem to have got to grips with the setup dude, I'm sure greater degrees of control will come with time
now get that patio/porch screen finished
Moto

 

[4Ryan6]

You definitely seem to have got to grips with the setup dude, I'm sure greater degrees of control will come with time
now get that patio/porch screen finished
Moto

Hey Moto!

What's left to do with the porch is minor details, we've been enjoying using it now for about a month!

As far as the cooling, I do have some future further testing probably using lower power drawing peltier or peltiers in combination to drop the electrical use as much as possible and see if the same cooling or better can be achieved.

You know it's never over! :lol:

So hows things on the home front? Ryan

 

[4745454b]

Was going to ask if the winter temps are hurting you yet...

At this point you need to worry about condensation? Are you going to drop the voltage to the TEC or increase the fan speed to cool the hot side more?

 

[4Ryan6]

Was going to ask if the winter temps are hurting you yet...

At this point you need to worry about condensation? Are you going to drop the voltage to the TEC or increase the fan speed to cool the hot side more?

Weather wise lately we don't seem to have much Fall season anymore we go straight from hot to cold it's 35.6f here in South Carolina according to our outside temperature gauge? Weird.

Condensation is not a problem until the water temperature goes below 10c, since adding more cooler insulation, (More on that below), I'm averaging running in the 14c area plus or minus 2c.

As far as dropping the voltage no the voltage is constant.

And regarding fan speed it's presently exactly where it needs to be with this setup.

One thing I have not shared is the cooler itself side walls has excellent insulation but the lid did not, it was actually hollow, so I drilled holes in the underside and injected expanding foam into the lid.

Now the cooler maintains a lower water temperature when the machine is not running, so when the machine is started up the peltier does not have to drop the temperature as much.

So the fine tune balancing of the fan speed was to arrive at a plus or minus constant.

Since this is working so well I do intend further testing and studying in the future regarding other lower wattage rated peltiers, to see if this cooling solution can produce the same results with a lower amperage draw, meaning less electrical power consumption.

It would be nice if I could get this cooling solution down to one machine power supply instead of having to run 2 of them, that's a future goal.

 

[4745454b]

What I meant was if the temp drops even while gaming, you should eventually reach a point where condensation is an issue yes? So to stop it from forming you would need to either lower the voltage so the TEC has less to work with, or cool the hot side more so the Peltier effect is less.

Same over here. In Northern Cali where I am we go from high 90s to rain and 60s in a month. Who needs fall?

 

[4Ryan6]

What I meant was if the temp drops even while gaming, you should eventually reach a point where condensation is an issue yes? So to stop it from forming you would need to either lower the voltage so the TEC has less to work with, or cool the hot side more so the Peltier effect is less.

Same over here. In Northern Cali where I am we go from high 90s to rain and 60s in a month. Who needs fall?

I see what you're asking now, when I see the water temperature in the cooler starting to go below 10c, I simply cut off the power to the 2nd power supply running the peltier until the water temperature in the cooler comes back up to at least the 15c range, then turn the power supply back on.

I monitor my water temperature in the cooler with a digital thermal probe wired temperature gauge, and occasionally look to see where the water temperature actually is.

It's not a real expensive model, you can spend some serious dough and get one that you could set a temperature alarm to sound at a certain temperature, but it's display reading is right next to my monitor and easy to occasionally look at.

My water temperature presently is 12.0c so I'm fine as far as condensation goes, but if I stayed online another hour, I would temporarily turn off the 2nd power supply.