There’s nothing more pleasing on a hot day than an ice-cold beverage. While the vast majority of us have a fridge in the kitchen, sometimes it’s desirable to have a further fridge in the lab, games room, or workshop. To that end, you may find value in this ultra-cheap, low-cost DIY fridge build from [Handy_Bear].
Like many tiny fridge builds, this design eschews complex gas-cycle refrigeration techniques for simple Peltier modules. These are devices that have one cold side and one hot side, because they move heat when electricity is applied. This build uses a Peltier module fitted with a fan to better shift away heat from the hot side, improving the module’s cooling ability.
The “fridge” itself is assembled out of thick XPS insulation foam. A hot wire cutter was used to cut several slabs which were then assembled using hot glue. The Peltier module is installed on the back, at the top of the fridge. Thus, air which is cooled in this area will then travel down through the rest of the fridge’s cavity. [Handy_Bear] also goes over how to produce a working hinge and a gasket for the door, which helps with ease-of-use and efficiency. As a nice touch, a set of 12V LED lights are also installed inside, which light when the door is open. Just like the real thing!
The final build is noisy, slow to cool down, and it uses 60 watts of power to cool down just two regulation-sized sodas. Notably, you could fit two standard NATO smoke grenades in the same space, as they’re almost-identically sized (ask us how we know). However, smoke grenades don’t usually need to be refrigerated.
None of that means it isn’t fun though! Plus, [Handy_Bear] notes that adding a second Peltier would greatly aid the fridge’s ability to quickly chill your grenades sodas. You might even like to explore the use of special fan designs to make the fridge even quieter! Video after the break.
How many watts of power would a normal fridge take to cool down just two regulation-sized sodas?
African or European?
Sodas or fridge?
European sodas are the easiest to cool. African ones start at a higher ambient temperature!
A higher temperature differential would actually make them cool faster in relation!
The more important question, were there empty coconut shells involved?
Are they migratory coconut shells?
whst? i don’t know that.
AUUUUUUUGGGGGGGGGGGHHH!!
How do you know so much about *sodas*?
Peltier coolers that size are typically about 40 watts per second. These DIY cooler / chiller / fridges have been such a popular DIY project over the last decade you can get temperature control modules so the peltier wasting energy after its reached a target temperature.
What does “40 watts per second” mean?
Must be rate of change. 40 W in the first second, 80 in the second and after 1 hour, 144 kW :)
40 joules per second squared.
40 kilogram meters^2 per second^4
°C / cm³ (water) / s
40 watts per second…
Compressors are definitely more efficient than Peltiers. I don’t have exact measurements but it’s mostly due to the fact that compressors can cool much faster and only need to switch on intermittently. The Peltier has to run constantly and needs an external fan to circulate the air, and won’t cool to as low temperatures as a compressor can.
You typically only see Peltiers used in small spaces and for light duty, like for cup holders, iceless coolers, and ventilated seats in luxury cars. Stick to compressors if you want a refrigerator or freezer.
Do you happen to know where to source small compressor freezers, not counting just buying a minifridge?
alibaba or ebay
https://www.aliexpress.us/item/3256803019254951.html
Still might be faster/cheaper/easier to find a used minifridge being thrown out but this ain’t bad
From Hobbits?
There is more to the system efficiency than just the cold generation efficiency though – your typical fridge is full of so many litre of air that for most fridge design will easily flow out to be replaced by more ambient temp air every time the door is open.
Then you have the insulation value of the box itself to consider – smaller surface area is good for not losing heat compared to bigger, so with the right materials in the construction and perhaps thicker walls…
Then you have the lifecycle efficiency and scale – compressors do wear out, pretty quickly really if our new fridge is anything to go by, and they can leak their working fluid (whatever that happens to be, as there are so many choices still in use, even if now banned from sale), and a compressor can only really get so small before it stops working well – a peltier element array is relatively immortal and scales differently.
So a small custom fit fridge or even freezer with good insulation is probably not going to require the peltier elements to run continuously, and not having much air exchange every time the door is opened as the box is so small and most of that air is already displaced by the cans you are cooling – It might well be able to match or best the bigger compressor fridge in efficiency overall for some users. (Not saying this particular example and user will though – would need far more info to be able to calculate that sort of thing)
The heat contents of the air is minimal. It’s about 1 Joule per gram-Kelvin. Take 1 cubic meter of air and 20 degrees of heat, and that’s equal to 24 kJ or 24000 Watt-seconds. A good compressor with a CoP of 5 and say 100 Watts average input power will deal with that in 48 seconds.
The peltier element on the other hand has to run continuously at around 20-30% minimum power because it is not an insulator – quite the opposite. What you have there is a hole in the wall of your box where the peltier element sits, and big heat sinks on either side that draw ambient heat back in.
>smaller surface area is good for not losing heat compared to bigger
Smaller surface area implies smaller box. If you go by the ratio of surface-to-volume, the smaller box loses more heat per the amount of stuff it can contain. In other words, it’s better to store your cold stuff in one big box than two smaller ones.
Just reorienting the door to be on top, instead of the side, will stop most of the cold air spillage from the unit every time its door is opened.
Or keep the fridge stocked to keep the volume of free air to a minimum. Water bottles are good for thermal mass as well.
Why does the Peltier need to run constantly? Could it not be controlled by a thermostat as well? Don’t compressors also have external fans for both the warm and cold sides of the heat pump? My fridge does.
Compressors are still probably more efficient, but I’m not sure if that’s inherent to the design.
Because the peltier element is very good at conducting heat backwards when it is powered off. You have a big heatsink that is warmed up to 40-50 C and the instant you turn the system off, that heat will flow back in.
My 450 liter fridge (variable speed) pulls as little as 33w on low load.
Peltiers really suck power.
Cooling them in a regular fridge and then popping them in an insulated box is probably easier. We get frozen food delivered in polystyrene boxes. The boxes are about 1” thick polystyrene and keep raw meat frozen for about 24h.
Yep, the medicine I take for diabetes has to stay refrigerated, and a three month supply arrives in a nice small Styrofoam cooler, with cold packs surrounding the medicine boxes. The cooler is a bit better quality than the ones you can buy at the local supermarket, and I’ve repurposed them rather than throw them away. One is storage for a special dry food one of our cats requires, another is at work keeping a few cold sodas and water bottles by my desk so I don’t have to go to the break room to get something to drink. I think I’ll use the next one to build a little mini fridge like this for the home office to complement the one at work, but with the hinged door and Peltier cooler.
or using refreezable ice packs
Stupid insulin packs shipped ro the western arid lands like Reno usually have 4ea 8 x11″frozen packs and about a 2 cu ft interior sized Styrofoam insulation. Can’t recycle them and after keeping 5 or 6 just have to throw the whole pack minus the insulin in the trash
“Thus, air which is cooled in this area will then travel down through the rest of the fridge’s cavity.”
You missed the bit where there’s also a fan on the inside of the fridge?
No need for a fan, convection works with any temperature differential, positive or negative.
Hot air 👆, cold air 👇it just works!
Just a note – that white foam is EPS not XPS
Not a whiff of engineering or calculations there. Winging it on hope and wishful thinking.
Educational DIY project? Only if you learn something.
The mediocre amount of XPS insulation (25 mm thick, 0.25 m area), Rsi = 0.7, will leak 0.3 watts per degree below ambient.
That puny hot-side heatsink will have a temperature rise of about 0.8 C/W. With 60 watts being dumped into it, and pumping another few watts of actual heat, the hot side of the Peltier will be 50 degrees above room ambient.
We want to cold side to be well below ambient, so let’s aim for a delta-T of 70 degrees.
Squinting at a datasheet for a typical (decent) Peltier used in this application: https://peltiermodules.com/peltier.datasheet/TEC1-12706.pdf says that with 5 amps of drive you might expect to pump (maybe) 5 watts of heat at 70 degree differential. 5 watts on that smaller inside heatsink (1.4C/watt) will have a temperature rise of 7 C, so actual interior temperature could get to 13 degrees below ambient, at which point the enclosure is neatly leaking 5 watts of heat, so no net cooling power. Meanwhile the interior fan is also heating the place at the rate of 2-3 watts too.
So, OK, if the stars align, you could KEEP something at 10-12 C in a 23 C room, but how long would it take to COOL 2 cans from (say) 20C?
At a temperature differential of 50C, the Peltier can pump 20 watts, or 5 cal/s. Not bad, actually; it could cool two cans (700g) at an initial rate of about 0.4 degree per minute. However, that would slow down rapidly as it got cooler: you’d get the cans from 23 C to maybe 15 C in a half hour, then a demi-eternity to get the next 3 degrees.
With Peltiers you really need to remove the heat from that hot side, and not work near the max delta-T.
My DIY laser chiller uses eight of those modules on a water block. They are underdriven, drawing 240 W, plus 25 W for 4 fans, on a very large hot-side heatsink. That 8-module pack can comfortably keep the 40 W tube at 18 C in a 26 C ambient, but 4 modules would not.
I once calculated how much power you’d need to make two ice cubes in two minutes using some reasonable assumptions about pairing up peltiers for maximum practical CoP.
I got it down to something around 500 Watts.
50 grams of ice, from room temperature? So roughly 20 kJ of heat to remove.
In 120 seconds = 175 watts.
That’s mighty good CoP out of those Peltiers if you can do it with only 500 watts. It’s dumping that hot-side power that gets you. Maybe doable with a really good heat exchanger and a chilly dump location, like our 5 C incoming tap water temperature this time of year.
You have to run them at half the rated current. Assuming the chiller plate goes down to -10 C and the hot plate can be maintained at 40-50 C you might get a CoP of 0.3 – 0.4.
That’s not too difficult to maintain with a liquid loop cooler, like a car AC heater core that’s designed to dump 3-5 kW at full blast, but the whole system would be ridiculously large and loud.
What if you replaced the hot side heatsink & fan with – an AIO CPU cooler?
Or any (water based) liquid cooling system that includes a radiator, and multiple 120mm fans?
It would have been more efficient as an icebox, cool it with ice from your freezer. Still when you need a cold drink…
Okay so how do you know about the smoke grenades?
The episode of Top Gear where Clarkson reviews the Ford Fiesta: https://www.youtube.com/watch?v=BmEp9bBY5ds
How cold could something like this get? -1C?
To me, this is a wasted time for building a “two” cans refrigerator. You can buy one of these from Home Depot store or somewhere for cheaper price than what you built. Be honest, how much did you pay to make one for yourself?
The point is the exercise, not the utility of it. Those gadget mini fridges generally don’t even work…
I thought the typical use of Peltiers was cooling camera sensors and scientific use.
That was a reply to ‘sampleusername’.
. sigh
And in space.
By comparison, the compressor in a $200 electric ice chest consumes 60 watts or less while it runs, and for mild chilling it barely has to run. The COP is massively higher, and if you want you can reach negative 40 degrees with it even in a fairly warm ambient environment. They’re not even optimized for fully reducing air leaks, and the lids are often not well insulated, but they still work great.
Peltier cooling only works semi-efficiently if you minimize the temperature difference and you use the modules at less than rated power (unless you require that delta T). The last two digits of the 12706 peltier module are the amperage, and there are modules the same size going to at least 15, so that’s an easy physical change requiring just a voltage supply change. It’s worth it to keep the hot side as close to ambient as possible because for the same target temperature, every degree higher the hot side gets, the more power is required, and thus the hotter the hot side gets as the heatsink tries to dissipate that heat. Or in some scenarios, evaporate water to approach the dew point if it is lower and that is practical.
As for making a fridge, you need to seal the air from escaping so easily, along with keeping the heat conducted through the walls to a small fraction of what it takes to chill the contents or the air after opening the container. Then you might also consider running the numbers on radiative heat transfer, which may be significant even outside of sunlight, enough to deserve a surface coating. Think foils for low emission, and white for low absorbtion. Thicker foam is very worth it for reducing heat losses, but once you get past that (see Paul’s comment) you still have a bit of a wait just to chill the drink itself. The quick way to chill canned drinks is still to immerse them in something cold that’s more conductive than air, in enough quantity to not average out to lukewarm when putting a warm drink in. An ice chest with icy slush is an example.
Tech Ingredients has some well-done examples of peltier refrigeration, which may be more educational, though this is certainly still nice to see.
So a DIY Frigidaire EFMIS129-RED Mini Portable Compact Personal Fridge?
Think I’d rather spend the $30.
What if I use this one? https://www.electron.com/peltier-module-hebei-tec1-12730s-p466/
Is it going to be faster? How much power and which cooler do I have to use? Can somebody help me?