Peltier based cooling box - Refrigeration test [Part 6/6]

In this video I test the previously built cooling box with a minimal load. I put 2 bottles of water (2x2 l) in the box and ran it with the same specifications as previously. Both Peltiers were running at 6 A. The cooling was slow. Based on my calculations, the water would reach 4°C after 9.5 hours. So, maybe it is possible to cool something with this overnight. My suggestion is to add 3-4 of this water bottle, cool it down to roughly 4°C, then add the things that you want to cool. The already cold water would act as a buffer, so it would be easier for the Peltiers to keep the system cold. But for better performance, the number of Peltier coolers should be increased.
Link to my website: curiousscienti...
Please don't forget to subscribe!
If you want to support my work, please consider buying the parts using my affiliate links:
curiousscienti...

[@andpe161]

Since the cold heatsink quickly reaches 3 C, the bottleneck is with the heat transfer from heatsink to the inside air, or between inside air and the bottles. Both cases are improved by increased internal airflow. If you wish to investigate closer, try cooling a loose heatsink, or check how long time a regular fridge needs to cool the bottles.

[@CuriousScientist]

Yes, I mentioned this several times that one of the bottlenecks is the cold side heatsink. Unfortunately, this was the largest available heatsink, so I used this for testing.

[@vladimirorlov413]

Your videos are amazing and they have so much useful information. I’m using 2 peltier units to cool a small vivarium for plants that require low night temperatures (55F). I’ve read online that a conventional thermostats will quickly damage the peltiers. Have you thought about creating a video about making a temperature controller for a peltier?

[@CuriousScientist]

Hi! I already made one and shared all the resources for it. :) kzbin.info/www/bejne/jpuncpqCetionrc

[@gerardmoangodelrio7334]

hey,how did it go with that project?I wanted to do something similar with a terrarium trying to go from 29 to 23 druing the day.If you suceed with the project,could you give me a little bit more information? Thank you!

[@oberhartkids9319]

Thanks for the videos. You definitely put in a lot of time. I enjoy all numbers you are taking to show the results. One opinion from me, what about having a smaller box like a small can cooler, then filling it with a couple litres of water and running those water heating with tubing into the water? Say you had initial ice/cold water of 17C and wanted to maintain this temp in the cooler while running the cold water then out to a radiator?

[@CuriousScientist]

Thanks! Maintaining the temperature would still require the same power as on the "hot end" you would still introduce the same amount of heat. With a pre-cooled water you just delay the effect.

[@oberhartkids9319]

@@CuriousScientist thanks. I think I have a plan, that's had a lot of changes in my head. Hopefully I can share it when the parts arrive, or next summer. You and others real, raw data i think is quite helpful to anyone interested in this. This is why I really enjoy KZbin.

[@ivanilarionov1893]

I will write here too, because I saw something interesting. You are trying to cool down 4 liters of water from 23 to 4 degrees, which is dT = 19. The needed energy should be 4x19x4.2/3600 = 0.088 kWh. If your estimated time is correct, then 0.089/9.5 = 0.01kW. So only 10 wats are used to cool the water. But your Peltier are supposed to provide much higher input. So there here are my assumptions: 1 your prediction is wrong - unlikely. 2. Your box has very big losses - unlikely. And 3 - the peltiers transfer much lower Q than they are supposed to at this current. I think this is likely, because the small cold sink limits it.

[@CuriousScientist]

Thanks! Yes, the cold side heat sink is definitely too small. I don't remember in which video in this series but I show the effect of a fan blowing the air on the heatsink inside the box (cold side heatsink). It leads to a better thermal coupling between the environment that we want to cool down and the cold side of the Peltier. BTW, in a few months there will be some drastic changes which will allow me to work on Peltier projects again. I will pick up some old measurements and experiments and try to improve them based on my experience gained since those old videos. I also have some ideas for future projects.

[@paulpease1788]

So about 5C from room temperature is not what is desired but what about maintaining temperature of 2 bottles taken directly from a refrigerator to go to the beach or camping? What if you used an "infinite" heatsink sized maybe 25cm x 25cm x 5cm. Wouldn't that allow more of the heat transfer from cold side to hot side? Also, my heatsink fins are much closer to each other. There are calculators to determine ultimate heatsink size and fin/pin separation. Refrigerators are used to cool things down but a box that keeps things cold would be of equal use for outdoors. Especially if it also kept things above freezing in winter months

[@CuriousScientist]

Larger heatsink on the cold side would definitely be better for better heat transfer! This was the largest heatsink that I could find at this moment, so that's why I used it. I guess, it would be easier to maintain something which is already cold, because the Peltier only have to work against the heat leakage and it does not have to create further temperature difference (supposing that the goal temperature is the temperature of the things we put in the box).

[@Jeffa67]

I have a 26litre esky that's internal tub is all metal. Could this have the cold side heatsink connect to it and be more effective ? Or even connect the cold side of the peltier direct to the tub ? Any thoughts thanks? Great information 🙂

[@CuriousScientist]

@@Jeffa67 Wow, I learned a new word today. I did not know that these cooler boxes are also called "esky". Supposing that the whole internal surface is metal and you directly attach the cold side of the Peltiers to it, it would improve the thermal connection between the Peltier and the volume to be cooled. This was one of my issue, my heat sink on the cold side of the Peltiers were too small for this purpose. But in your case, if you have the whole interior cladded with a metal sheet which is in a good thermal connection with the Peltiers' cold side, you should get good results. Also, you can put the things in the box in a way that they directly touch the wall of the esky and therefore the thermal connection becomes better than through air.

[@Jeffa67]

@@CuriousScientist hello from the land down under. Lol esky is an Australian term and I forgot about our local terms. Thanks for the reply and again the post. Also may I suggest a number system on you fridge builds as you have with the peltier videos. I was trying to follow the sequence and was unsure about the timeline. Watching this on a smartphone with the screen limitations. Thanks again, Jeff

[@CuriousScientist]

@@Jeffa67 Haha, cheers! I will look into this numbering strategy and try to come up with something. I will try to add some intelligent coding or something to the titles. Probably during this weekend, because I will have some free time.

[@wayneo7220]

I'm enjoying these peltier videos. They are a gold mine. Do you think adding a desiccant bag would help lower the temps since dry air cools faster?

[@CuriousScientist]

Thank you! Well, I don't know to be honest. I haven't tested it, but according to the physics it should help. The question is if it would help a lot or not? The amount of air is so little in this box, so I think that there are better ways to improve the cooling. To make this box really work, I would use a "stupid amount" of Peltier coolers (8-10, maybe more) and also large heatsink with a fan inside the box. But with large number of coolers comes large electricity bill and large amount of heat to dissipate.

[@wayneo7220]

@@CuriousScientist Considering the size of the cooler, I thought it did quite well.

[@CuriousScientist]

It did OK, but could have been better. I will make a new iteration of this box once I am finished with my other projects. I will use more Peltiers and a beefier power supply to speed up the process and to increase the cooling capacity.

[@wayneo7220]

@@CuriousScientist I recall a company about 20 years ago developed a commercial air conditioning system that incorporated a "desiccant fan" to remove humidity before being cooled. The fan had a proprietary desiccant coating on the blades. Not sure how it would have been "recharged" once loaded with moisture though.

[@CuriousScientist]

Oh, if you heard about these things, then you would be very interested in this video if you haven't seen it yet: kzbin.info/www/bejne/iJDKZaGKacZlbrc Actually, Tech Ingredients is an amazingly good channel, I really recommend you to watch them! Very good videography, high-tech solutions and extreme amount of knowledge!

[@Angel_Padilla]

I think it will take more than 9.5 hours. Because as the water temperature drops there will be a loss of performance in the transfer of heat between the cold face and the water. I love the videos because I'm starting a pretty similar project, a 12-24v DC and 220v camping fridge. Starting from a commercial ice chest, model igloo marine 36, I will build the top side with peltier 12a x2, Arduino, voltage control, performance calculation, automatic light with door opening, integrated LCD and carbon fiber cover. I will try to make a video of the process.

[@CuriousScientist]

Nice project! I was also thinking about converting a commercial cooling box/ice chest into a cooling box, but when I checked the price of the boxes, I gave up on the idea. Also, building from scratch was more fun to me. Regarding the Arduino, what would you use it for? How are you planning to realize the cooling control? Do you want to turn the Peltier on and off based on some temperatures, or you want to keep it on most of the time and just control the current/voltage? You project sounds really interesting, it would be nice to see the details. I also agree with the cooling time, it might take longer. It would be fun to try it, but I am a bit concerned about my electricity bill. If I can suggest something for you, use as big as possible heatsink on the cold side too!

[@andpe161]

Do you have a way to control the voltage? Just using PWM would not be energy-efficient. PWM control of mosfet, followed by a big filter could work -- in effect a custom switched regulator.

[@CuriousScientist]

@@andpe161 PWM is not the best solution. I started to work on a "better" one. Keep checking back to my channel for more information. As I progress with the project, I will upload more and more new videos. It will be a microprocessor-controlled power supply.

[@Angel_Padilla]

@@CuriousScientist I have been studying for several months the power required to build the 36-liter refrigerator. My goal is also to reach 4 degrees but a little more difficult because in Spain the climate can easily reach 28 degrees 30 degrees outside the refrigerator. That's why I'm going to do it on a marine igloo whose insulation I hope is much better. I already bought 90% of the materials, including the igloo marine. The design of everything I have almost complete, but it will still take a few months to finish the project because I have to finish another and my work does not leave me much free time. My outlay has already been about 180 euros approx. I hope it cooled enough. For this reason I want to see the test of your project until it reaches 4 degrees, I hope that at those temperatures the insulation is sufficient and the cooling curve does not turn horizontal before arriving. In my project, the arduino will control the voltage that reaches the peltiers so that when the desired temperature is reached, the consumption drops, but without turning off the peltiers completely. I will have to see the voltage it needs to maintain the temperature. In addition, it will send the most optimal voltage for the best performance, I have already seen your tests and 12 volts is not always the most optimal. The ardino will also control when it is plugged in at 12 volts to limit the current drawn to 15 A maximum so as not to damage the electrical system of the vehicle. For 220V plugged no limit amps. For the interior heatsink I have a large aluminum grill weighing 500 grams. With a delta fan of 5000 rpm, for the hot side I have two cpu coolers of 140 W each. For the hot side I know that a larger one would be needed, but I will see the possibility of the arudino controlling the voltage of the peltier when the cpu cooler gets too hot or even replacing the peltiers with the 8A model. I have no problem with the cost of electricity because the project is to make long trips in the vehicle or camping where you only pay for the connection. I will record videos and make a summary that I would like to share with you.

[@Angel_Padilla]

@@andpe161 Good, for that I have two open channels, what you say about the mosfet plus the filter or directly buy 2 step-down modules of 15 A each, these purchased modules already have a perfectly balanced filter. As I read in other tests, the peltier loses performance with pulsed signals. I already did tests with mosfet and filter but I do not have great knowledge of electronics and it is difficult to configure a filter that works well in different loads. This ends up causing power supply problems. I already tried with different frequencies. The option that I will surely install will be a stepdown to which I will make a small modification, I will replace the analog potentiometer with another digital potentiometer that will be controlled by Arduino. This is the only remaining part of the design but I need to do more advanced testing with an oscilloscope.

[@wayneo7220]

I noticed the CPU coolers were mounted horizontally, but heat pipes need to be mounted vertically for the most efficiency. This is because the evaporator end needs to be below the condenser end for the best evaporation, since it is aided by gravity. This chart shows how the inclination angle affect the power output of the cooler. So your results would improve considerably if the cooler were mounted vertically. celsiainc.com/resources/calculators/heat-pipe-calculator/

[@CuriousScientist]

It really depends on the type of the heatpipes. The inner structure can differ and the cooling can be affected by the inclination differently. For example, most of the computer coolers are mounted sideways and they perform really well. Laptop coolers are also horizontal. Also, for this kind of cooler construction it was not possible to mount the coolers differently.

[@wayneo7220]

Horizontal heat pipe coolers work, just not as efficiently as vertical units. That was my point. As you can see on the chart there is a ~47-48% drop in efficiency between 90 degree and 0 degree inclinations of heat pipes.

[@andrewtitcombe8378]

Vapour refrigeration system cold room fin spacing is wider than a freezer room evaporator fins. I assume this is because as the heat transfer efficiency at a lower temperature becomes lower and therefore needs a larger surface area between the the cooling surface and the air passing over it. (my guess, but will study this further )

[@DeepBlueDreamer1]

Ok so I’m talking cold air blowing A/c after this…. Let’s say you have that setup in a highly insulated box, basically a freezer. Then you have it filled with ice and a closed system antifreeze saltwater blower with ice blocks inside the box. What the hell happens next?

[@CuriousScientist]

I don't really understand your question. Could you explain it better please?

[@formalfighter]

If the internal fans are placed near the top of the box, would it create a convection current inside the box? (As cool air sinks and warm air rises)...

[@CuriousScientist]

I am not sure if there would be a convection current. But I think the best way is to just place the fan(s) on the heat sink. More direct heat exchange and more airflow across the fins.

[@formalfighter]

@@CuriousScientist Yes, that I agree, but placing the fin near the top would help? Also, in your opinion - can a Tec1-12706 module Peltier module cool a box of say 10 liter volume to 4 degree Celsius?

[@CuriousScientist]

One single Peltier will not really work. You need multiple of them as they have low cooling power especially when you want to maintain low temperatures. The heat sink can be anywhere in the boy, it is a small volume anyway.

[@formalfighter]

@@CuriousScientist Thank you for the reply man. I was thinking as you cooled 64L with two 10A Peltiers, so by comparison 10L is a much smaller volume, hence one Peltier... Currently I am waiting for my tools which hard to get by due to the lockdown situation at my country... so can't do the physical experiment yet. :(

[@CuriousScientist]

Have you watched the video properly? There were also 2 bottles (4 liters in total) of water inside the box which acted as a "thermal load", something that the Peltier coolers should cool down. You could see that their temperature could not even reach 15°C. I guess, you do not only want to cool down 10 liters of air but you also want to cool something inside the box. Cooling air and cooling water are two very different things

[@johnyoung7691]

What was the air temperature inside the fridge?

[@CuriousScientist]

Watch the videos, the answer is there.

[@johnyoung7691]

@@CuriousScientist Good on ya mate 👏👏. I'll spend the next hour watching all of the videos to find the answer to my question that you could of answered in similar effort to replying to my comment.

[@CuriousScientist]

@@johnyoung7691 If you go to the very end of the video or go to my website (link is in the description), there is literally a HUGE chart which shows the time vs. temperature data measured at six different points in the setup... Sorry, I cannot make the data even more easily available... 🤷‍♂

[@spacejunk51]

and now that you've the condenser and evaporator setup you can test this efficiently. Minus the fact that water and aluminum is not a good thermal conductor in todays standard. I bet you'll need a lot less power this time.

[@CuriousScientist]

I won't build another system.

[@andrewtitcombe8378]

It's a pity you never ran the test longer with 1 litre of water until the temperature of the water never got any colder. And then let everything reach room temperature again and the did the same experiment with 2 litres of water. take the energy used form the first experiment from the second. Then you are left with the amount of energy used to cool 1 litre of water. Do this calculation for every degree of change through the temperature range. And the inefficiency graph can be produced across the temperature rage of the Peltier system. Also the water bottle temperature change is a curve but over a long time period. As the temperature difference between the surfaces get closer together. This is why the cold side temperature curves to almost flat as the temperature difference across the cool side heat sink becomes more similar to the external part of of the heat sink that is in contact to the air flowing over it. speeding up the air flow will help but will still flatten out over time as the cool surface and air become similar in temperature. The point that things can't get any colder is when the coolant can not extract any more heat due to the system inefficacies and heat being absorbed through the insulation is greater than the cooling capacity at that system temperature.

[@CuriousScientist]

Unfortunately, my electricity bill doesn't allow me to run these tests all day long.

[@andrewtitcombe8378]

@@CuriousScientist I understand. I am looking for data on these Peltier's and your water cooling heat side seems to be the more efficient way rather than air cooled. And your project came closest to a set up that could give some actual efficiency figures with your data logging etc on KZbin. I am building a cool side water circuit for chilling a insulated water bath for chilling a 5l beer keg. Hot side is air cooling into a beer fermentation insulated compartment as it is situated in a unheated building outside in the UK. using two temperature controllers . one controls the temperature of the ready to drink beer keg cool side water circuit. while dumping the hot side into the fermentation chamber The other controller adds an additional 30watt PTC heater if extra heating is needed or extract fans to bring in fresh cool air to remove heat from the chamber if the added heat from the hot side Peltier is higher than the chamber set point. I need water bath to chill 5l beer plus 2l of circulated water over a time period in the region of 24 hrs from 22C to 8C. I need enough heating from the hot side to maintain 20C in the fermentation chamber . while the night time temperatures fall to 0C and days are 15C summer is 25C to nights of 16C. I may have to go down the refrigerant system route. But as it is such small scale was hoping to use the Peltier method Yours was the best video on KZbin to help me come close to knowing my idea has a chance of working across the annual temperature ranges . Once i have it built i will data log via an esp32 and DS1820 probes for cool side water temperature Keg temperature Fermentation temperature Fermentation chamber temperature and external air temperature to the Thingspeak graphing server to remote monitor and record values . Basically a system is inefficient if the wasted energy is thrown away. But by using it to warm a fermentation chamber then the energy is not wasted. Sorry for my long reply. But your video gives me hope.

[@CuriousScientist]

Nice approach! I think 5+2 liters might be okay for a properly sized Peltier cooling system. I would not try it with only one, but let's say 3-4 Peltier coolers at least. 4x TEC12708 or 12710 would be nice to try. If you can make use of the "insane amount" of waste heat, then yeah, it is better from that perspective that you can use a larger portion of the electricity that you are paying for. But of course the cooling efficiency is still bad. Look around in my Peltier-related playlist as I have some videos on water cooling with the cold side of the Peltier cooler...etc. I also made a completely automatic microcontroller-based Peltier controller. Check this out: kzbin.info/www/bejne/jpuncpqCetionrc