Improvised A/C idea

hey doodles.

had this idea boncing around for nearing a year now, figured get some outside input on it.
its a bit of a "why would you do that" sort of thing, which im aware of.
my guess is it will probably only work a smidge, it would be interesting to know how much or how little it works

so basically, a peltier device is a small plate that runs a 12v current through. one side of the plate generates heat the other side generates cold, and you direct that heat or cold to what you want to cool or heat.

to my unsterstanding its less efficient at cooling but still does a good job on small things, like water coolers, car fridges and cooling CPU's.

so here is the basic run down of a potential way to use it as a air cooler (if its even effective)

basically, attach pelteir to a section of fan box with cold sink sticking into fan box to transfer cool temps to passing air. then insulate the other side and run a water plate to pump the heat away from the hot side to a small radiator on the front of the car.


Considerations i thought of when thinking it up (incase it interested anyone to know the thought process)
1. keeping cold side as short as possible so no cold loss in getting the cold to the fan box. hot side can be as long as it wants as more surface area to dissipate heat, and the more water in the lines the more thermal capacity the hot side has.
2. the faster you can cool the hot temps the more effective the cold side will be as it pulls the cool from there so if the hot side is kept cool there is more cold "resource" to use.
3. using a cold sink in the fan box will make for a faster cooling of the cold sink but will have less cold thermal capacity. for more thermal capacity but slower cool down you could use a radiator like the hot side which will have fluid which has a much larger thermal capacity.



thoughts, suggestions, inputs, tell me im crazy?

for those that will ask, I have no idea when or if i will do this its just something i've been thinking through

To save compressor weight and drag on the engine?

Sounds cool man. Do they make them in big sizes? Or could you run a few in parallel?

My only issue would be no like drying out of air and thus ineffective demisting at times, but if you're willing to try it in qld where you need AC more than I do, why the hell not!

Seems a bit out there, but hey that kinda adds to the coolness (ignore the pun) of the idea. I'm guessing to have a better idea of answering the "Is it worth it?" Side of things, you'd need to have two areas of investigation; what Peltier device you're using, it's current draw to produce differing levels of 'coldness' and then how well that removes heat from the cooler box, and secondly how well the cooler box can be used to pump cold air into the cabin. First bit I feel is something that would be very difficult to simulate.
The second however, you could use any way of cooling your cold sink and while it stays cold for a period of time, test that in your fan box and optimise it for use in the cabin. With this you could trial different sink temperatures, cabin fan speeds and generally the practical side of how the cooled air would be vented. If you get a null result from this ie the fan box massively inefficient, air doesn't flow well through entirety of cabin, then you can call it a day, knowing that no matter how much effort went into it, the practical side of things was the cause of it being unsuccessful.
However if you get good results, well then Peltier device here you come

It should work well as a demistifier, as it still condenses air as a traditional AC system does. Only questions I can pose are that perhaps you might have an issue with ice build up on the device surface? But would that even change the efficiency of the system? And if so, what would you put in place to over come it?

Just my thoughts though, aha.

Hugh-CE said:

To save compressor weight and drag on the engine?
Sounds cool man. Do they make them in big sizes? Or could you run a few in parallel?
My only issue would be no like drying out of air and thus ineffective demisting at times, but if you're willing to try it in qld where you need AC more than I do, why the hell not!

Click to expand...

nah not to save weight or drag, just cos car doesnt have A/c or even loom for it and i CBF'd putting a pump in and loom until the 6a12 is in and stable
the plates come in 1 size 40x40mm, can run a few but they do chew energy, one or 2 would be good enough I wouldnt want to use more.
Im going to look into how compressor A/c de humidifies the air see if something similar can be done

My $0.02 while on my phone, is I understand the car's AC is so so much more efficient in regards to effort:cooling ability/coldness, but I haven't read up enough about it.
You would need a highly efficient heat/cold sink as well for temperature transfer, and a sealed enclosure to ensure it doesn't loose its temperature.

Going on these same thoughts, might not be too bad an idea to have temp transfer tubes or something to have seat cooled (can get heating elements into the seats relatively easy enough so thats not required )

donki said:

My $0.02 while on my phone, is I understand the car's AC is so so much more efficient in regards to effort:cooling ability/coldness, but I haven't read up enough about it.
You would need a highly efficient heat/cold sink as well for temperature transfer, and a sealed enclosure to ensure it doesn't loose its temperature.

Going on these same thoughts, might not be too bad an idea to have temp transfer tubes or something to have seat cooled (can get heating elements into the seats relatively easy enough so thats not required )

Click to expand...

Agreed it would be very inefficient compared to a traditional crank driven compressor a/c. Thats the nagative for sure. Positives are you could leave it on for an entire journey and it wouldnt be sapping power or efficientcy of fuel.
I was planning on insulating the whole cold side really well. And hot side till its past the firewall. To try reduce cold loss as much as possible.

As for cooling seat. I was under the impression the heaters were a resistive wire like a electric stove. How would you make up a cooled seat. Cooling tubes would be more bulky and hard to sit on?

Pretty sure the cooled seats in Lexus GS's is done via the air from the A/C system, fed through a moulded plastic pathway into the underside of the seat, then through the perforated suede/leather. But I could be wrong. Seems like an interesting idea but I can't see it being effective on a large scale. I've traveled with a portable fridge before and it was only reliably capable of bringing the temperature inside 20-30* below current ambient temp. Which sounds alright. But that was and enclosed, insulated, stationary air capsule.

On the subject of fuel efficiency/power drag. Wouldn't the alternator have to work harder to support the peltier? Which could lead to premature alt. failure or increased fuel use anyway? Not too sure on current draw of one of these things. This whole paragraph is abit of a guess.

Is this something you are considering because of the 6A's space constraints and annoying hose routing?

On phone again; think along the lines of thin, wide, low pressure water tubes, akin to a solar pool heater that looks like a waterbed for snakes lol. Overall, its starting to become a lot more effort than its worth

Its a nifty idea however I dont think it would be very effective. Reasons being.

Peltier cooling is very inefficient, like 10-15% Carnot, its main purpose is for precise temperature control over a small range, for things like cooling electronics etc, this is because it can be controlled electronically to be very precise, whilst being solid state and quiet.

Vapor compression (auto AC) is much more efficient, i think up around 40-60% Carnot. much better at dumping large amounts of heat outside.

By the time you get a TE pad large enough to do any effective cooling inside the car, you need a large radiator outside the car, and basically you end up with a setup very similar to the factory setup but with less capabilities.
If you want something more space efficient for the 6A12 setup, just get some custom lines made up.

the reasoning was never to conserve space for the 6a12, but moreso an exploration of an alternative method
once 6a is in and stable it will be getting real A/C for sure

I have decided that its not suited to full on A/C, but I did want to have a play with one for a fun little project so i'll think of something else practical

Maybe try a cooler drink holder for a can or something

Simply they cannot provide anywhere near the cooling of your stock ac unit. Even with 4 panels installed I doubt you would notice ay change in air temp. This with a 24 amp load on the alternator.
Also the hot side would be a *mum* to trasfer the heat outside. They are in cheap coolers for a reason.

Install one in a sealed box and see how long it takes to cool the air inside and you will see what I mean.