> Peltier systems require more energy input than comparable compressor-based systems. However, while heat is emitted from Peltier systems, it is comparatively less than in compressor-based systems.
How is it possible to use more energy, but produce less heat? Are they comparing a tiny peltier cooling box to a full sized refrigerator?
EDIT: I guess you could argue it produces less heat if it moves less heat from inside the refrigerator, but that's not really an upside
Yea, that was weirdly worded sentence. Peltier cooling can probably be more efficient (not clear) when there's low temperature difference so if one wants slightly colder drinks than the room temperature there's an advantage. E: As someone here mentioned, wine fridge seems to be that perfect application.
It's hard to separate the heat generated by the equipment from the heat transferred from the conditioned space to the unconditioned space when talking about the heat produced. So I'm guessing, they're effectively saying the peltier system uses more energy while transferring less heat. It makes sense for some applications where noise and vibration are important or required capacity is low or compressors are otherwise unsuitable, but compression based refrigeration is usually the default.
Someone wanted to add an "on the bright side!" to a sentence which wanted to be "Peltier coolers are less efficient, and also less powerful: they will burn more electricity while moving less heat".
The usual observation is the Peltier system moves the heat about one millimeter, and then one needs a conventional cooler to remove the heat from that location.
You need a conventional cooler to remove the heat from the hot side of both. The only difference is that the compression one can move the heat further away.
But just because of the bigger surface area of the radiator. A conventional compressor with the same surface area of its radiator as the peltier system will perform the same (provided they are at the same temperature).
This is why heatsinks on CPUs have fins: to increase the surface area where the heatsink is in contact with the (hopefully cooler) surrounding air. The windings of a radiator are very much the same idea.
I think there is a difference in technical concepts that you are mixing up. What you are talking about is convection cooling. Both compressor based systems and peltier coolers need to transfer the removed heat to the environment. This is usually done with big surface areas and potentially fans to produce airflow.
In a traditional compressor based system this huge surface area is the radiator in the back and technically it also only "moves the heat" to the surface of that radiator. But give the size of that radiator it will be effective at transfering that heat to the surrounding air.
A peltier cooler with a smaller surface area needs a heat sink and or some form of active cooling (airflow, water etc).
I have a wine fridge that uses Peltier cooling and it feels like the perfect application, since the target temperature is not as low as a normal refrigerator, and quietness is important.
Do you have any recommendations? I live in a 1-bedroom apartment (so noise level is important) and have been looking for some quiet wine fridge (or perhaps a wine+canned beverage combo fridge). Any brands I should be looking at or avoiding? If I also want to store cans should I be avoiding Peltier coolers?
Just going off of general chemistry knowledge, vibration may be speeding up the chemical reactions that are happening in the wine (aging). Same as stirring.
EDIT> For a reaction A + B -> C, the more chances molecules of A and B have to get into contact, the faster the reaction could run since chemistry is just atoms bumping into atoms. Stirring or shaking can bring more unreacted A's and B's together.
If you live in Europe right now, beware that unless you're a big drinker, that wine fridge will be costing you more in electricity than the wine you put into it.
European energy is super expensive right now, fridges normally use quite a lot of energy, and peltier fridges use quite a lot more than normal fridges.
I agree. It should be around 3 times less efficient. My guess is that people that are into wine enough to buy a dedicated fridge probably plow to way more than 30 euro a month of wine.
> Everything sold in last fifteen years is A class or better for efficiency.
In case you are unaware: The European efficiency rating system is in fact percentile-based on some fictional reference appliance which is adjusted every few years [0].
A fridge with an "A" rating you bought 15 years ago would not get an A rating today.
(This sounds weird, but fights the problem that when the system was based on absolute values instead of percentiles, nearly fridges started being rated A, A+, A++ or A+++ to the point that the scale was very opaque to consumers.)
I wonder if a refrigerator that uses Peltier cooling could ever be practical by taking all of the space savings of getting rid of the compressor and using it to increase insulation or thermal mass, offsetting some of the relative inefficiencies of the underlying tech.
It doesn't appear this article will make it clear if such a thing is possible, since it lacks any hard numbers relating to the technologies it is comparing.
No. The efficiency is so terrible compared to what heat pumps can do it's ridiculous. Think a few percent vs 2-300%.
Peltier coolers have one thing going for them- they have no moving parts*. That's it. Which means they still have their niche! But everything else about them is awful compared to other cooling solutions.
*in practice most applications use fans for the heat sinks.
They're semi-practical today in niche applications. You can get a small peltier car cooler , and the inefficiencies aren't huge compared to the rest of the car's electric usage. They're not great though, and many are poorly made.
I wish there were better options for inexpensive, miniaturized heat pumps. I can think of all sorts of applications where it'd useful to have them. Plunge pools, hot tubs, trailers, greenhouses, dog houses, etc.
There are ultrasonic heat pumps (used in JWST for example, precisely because of vibraiton issue), but it will take time till it will be mass produced. No refrigerant needed either
Same thing different name. I said ultrasonic because that's the wavelength you need for it to be compact
The basic way of working it is basically
* lowering the pressure of gas causes it to cool, increasing it causes it to heat
* creating standing wave in a gas in pipe means the low pressure part will be cold, high pressure part will be hot, and the difference will equal to basically pressure induced thermal difference
sooo a fancy speaker in the tube. Already stuff that can be bought, just didn't get to the level required for consumer stuff.
Supposedly they can be pretty efficient too. Would also be basically noiseless, as the ultrasonics are limited to the cooling tube and easier to dampen anyway.
Most modern residential and commercial refrigerants aren't particularly flammable in air, they only score about a 1 on the NFPA 704 when under pressure. Also most technicians will "recover the charge" when possible during maintenance or decommissioning. There's been quite a bit of development in the equipment space in the past two decades.
On the other end, you can for some applications, use "R-290" which is just highly pure propane. Typically waist high commercial ice cream freezers will use a fully sealed R-290 system.
And finally there is also "ammonia phase change" refrigeration, which can be done with a compressor, but also a "generator" which is essentially just a propane or electrical boiler which creates the high pressure side of the system,
also with no moving parts. You will typically find these in highly portable and mobile applications, such as RVs.
The tubing isn't pinched shut, it's just pinched to restrict the flow. I believe it's to create a bottleneck where pressure can build up between the "hot side" and the "cold side" of the refrigerator. Before the pinch the compressor is driving gas into the pipe, where it meets resistance at the pinch point and compresses to liquid, shedding heat. Then after that point a narrow jet of compressed refrigerant squirts into a lower pressure part of the pipe and starts to evaporate, expanding and taking in thermal energy again.
Of course, the refrigerator already has some valve to do this, but the author mentions that it may not be designed for the pressure differential needed for propane, so you can make your own with a pinched tube.
To intentionally create a restriction, an orifice (auth uses the term expansion valve which is incorrect in this scenario).
Liquid refrigerant flow is restricted when it flows through the orifice. This causes the pressure to drop which causes the refrigerant to flash/boil. While the refrigerant is flashing/boiling, it is absorbing heat from the surrounding environment.
You mean overcharged... see how it's frosting all the way back to the compressor? That both wastes energy cooling the outside of the cabinet and can damage the compressor if liquid refrigerant enters it.
Ideally the charge amount should be such that the evaporator is fully frosted, but not much more than that.
It's worth mentioning the recent development of extremely miniaturized (soda can-sized) refrigeration compressors, such as those from Aspen Systems [1]; these allow for efficient cooling in small spaces that just can't be achieved with Peltier-based solutions.
An interesting project I've seen using one of them is this portable chiller suit to cool the human body on-the-go, intended for relief of Multiple Sclerosis symptoms [2].
For DIY/prototyping, there are a bunch of Chinese manufacturers now producing compressors of these sizes along with the appropriate driver board for the variable-speed AC motor [3]. I'm excited to see how this technology will be applied elsewhere in time.
I remember Adam Savage got something like these to keep from baking his noodle when he made the No Face costume for Comicon. I wondered when those sorts of things would show up for more general usage.
Isn't part of what makes these work also linear compressors? The old on/off models have gotten better over time, but when we were kids, some of them would practically cause brownouts on the electrical circuit. Too much startup current. So for electronics you need someone with a smoother transition.
> practically cause brownouts on the electrical circuit
That's because a compressor has an odd property... The motor requires substantially more starting torque than running torque. That's because the cylinder can be stuck on the compression cycle pushing against a cylinder of gas, and has no momentum to get it through. Also, the very first compression typically has far more gas in it (because the 'cold side' is actually warm before it starts).
If nothing were done to compensate, the motor would stall, overheat, and burn out.
Designers of fridge compressors have instead added a very high current starting coil, which will provide massive amounts of torque to get the thing going for a few seconds. Sometimes that might be 20 amps or more for a few seconds. When the motor is spinning fast enough, that coil gets disconnected.
The gas distribution part can be significantly compensated for by "unloading valves" of some kind—this is very common on piston air compressors—but with unloading, a major component of the starting torque is overcoming the angular momentum of the motor and compressor assembly. This is expressed as Locked Rotor Amps (LRA) on a compressor's spec sheet. There are devices that can limit the inrush current to about 25% of the listed LRA.
On most air conditioning compressors today, a scroll plate compressor is used, which doesn't have the unloading problem of the piston type of compressor.
>On most air conditioning compressors today, a scroll plate compressor is used, which doesn't have the unloading problem of the piston type of compressor.
Just replied to the same comment wondering if this has changed. I used to help my dad install residential and commercial HVAC systems and all of the compressors back then were piston type. I would tick him off by getting impatient when we were testing by calling for cool too soon after shutting it down and tripping the thermal beaker. (Come on dad it's hot up here!!! lol)
I imagine the scroll type are probably quieter as well and probably work nicely with inverter-based drives.
Yes, they do tend to be quieter, and as far as I know are the only kind on inverter drive AC compressors. It's also a continuous process without having peaks and valleys in the motor load which probably lends itself to work well with an inverter drive, but I'd have to think about it a little more to say if that's accurate.
I did go back and look that there are sometimes valves used in scroll AC compressors. Even with that the full compression cycle is much more gradual, and continuous happening over ~3 revolutions of the compressor instead of half a revolution for a piston so there's not a slug compressed gas to hold the compressor back in the same way as a piston compressor.
I did HVAC work in the late 90's till the late 00's. At that time, only the really expensive, high SEER rating units were variable speed. Cheaper 10-13 SEER systems had scrolls but they were fixed speed.
That starting coil can get stalled too, esp if head pressure is still very high. I did HVAC 30 years ago and back then you could reliably thermal-trip residential and commercial systems by simply turning them off then back on within 10-15 seconds. Motor would just hum for a few seconds then 'click. We'd always have to let it equalize for a minute or so before turning back on.
That's really fascinating, thanks for the aliexpress link! I'm tempted to get one for regulating the fermentation temperature of beer. (A while ago I saw a peltier cooled rod that dips into the fermenter, but I bet this would be a lot more efficient! - https://web.archive.org/web/20220625100157/https://brewjacke...)
Both peltiers (look for thermoelectric coolers on Amazon/Newegg) and phase change coolers are used by PC enthusiasts, but it’s not that common because regular liquid cooling is good enough for everyday use, and you don’t need to waterproof everything to protect against condensation when the cooler gets below the dew point. Not to mention mould...
After losing power in southeast Michigan for 3 days, a couple of weeks ago, I began wondering about low-power cooling. Something like a couple of solar panels, a battery of some kind, then a fan for an evaporator "swamp cooler" or something similar. I'd like to be able to cool a single room occupied by 2 adult humans and a couple of 75lb dogs.
How much power do those mini refrigeration compressors draw?
> Peltier systems require more energy input than comparable compressor-based systems. However, while heat is emitted from Peltier systems, it is comparatively less than in compressor-based systems. This means that there may be some savings in terms of air conditioning load in the area where the unit is housed, which could help offset the energy efficiency differential.
Either I am misunderstanding what is being said there, or the article author misunderstands how physics works.
Yeah, that also looks sus to me. According to my very limited understanding of how this works, compressor-based systems (actually more like phase-change systems) only have to pump heat energy from one place to another. Peltier systems have to do this while also fighting thermal conductivity from one side of the module to the other. Intuitively, the latter probably generates more waste heat...
But, compressor-based systems have to deal with friction (both mechanical and fluid). I'd still bet on their efficiency over Peltiers, though, unless it's a really small system.
The compressor motor converts a large portion of the electricity it consumes directly into waste heat which is why the motor warms up
much more than from the friction.
This is in excess of the heat being removed from the refrigerated area.
Air conditioners have the warm-running compressor outdoors, but refrigerated indoor appliances release that heat into the room along with the heat
transferred from inside the refrigerator. And then the AC encounters further inefficiency as it next removes this heat from the room.
Even the motor of a ceiling fan warms up a room (gradually) if there is no AC and windows/doors are not open adequately to let that heat escape.
Ceiling fans have big motors and are best outdoors like at a tiki bar.
This is true, compressor-based systems don't just magically move heat energy. But at least the hot/cold sides can be isolated much better than a Peltier.
At least for peltier cooling, as delta temperature across the peltier increases, it takes progressively more power to counteract conduction inside the peltier. This requires progressively more power and results in joule heating. Works great at low deltaT though.
A constant current driver can also be important - PWM results in increased joule heating. For 50% maximum constant current joule heating is (V/2)^2/R but when duty-cycling with 50% PWM, it's V^2/(2*R). Moving the scaling factor out from under the ^2 makes the savings more pronounced at lower deltaT as well. But constant current is more complex, especially with larger peltiers.
I cants really speak for Carnot cycle coolers, but they can often be effectively duty-cycled and can be placed remotely so the compressor waste heat can be vented into the room. Performance still degrades with dT though.
Do you endure the mid year heat like the vast majority of us? You could surrender and concede that environmental change is negatively affecting Australia's summers. And afterward, where will we as a whole be when winter comes? then visit https://ductscleaning.com.au/service/evaporative-coolor-clea...
86 comments
[ 1.9 ms ] story [ 132 ms ] threadHow is it possible to use more energy, but produce less heat? Are they comparing a tiny peltier cooling box to a full sized refrigerator?
EDIT: I guess you could argue it produces less heat if it moves less heat from inside the refrigerator, but that's not really an upside
I'm not aware of any point in the input/output temperature chart where peltier wins on efficiency.
This is why heatsinks on CPUs have fins: to increase the surface area where the heatsink is in contact with the (hopefully cooler) surrounding air. The windings of a radiator are very much the same idea.
In a traditional compressor based system this huge surface area is the radiator in the back and technically it also only "moves the heat" to the surface of that radiator. But give the size of that radiator it will be effective at transfering that heat to the surrounding air.
A peltier cooler with a smaller surface area needs a heat sink and or some form of active cooling (airflow, water etc).
https://blog.vinfolio.com/2016/09/13/does-vibration-ruin-win...
EDIT> For a reaction A + B -> C, the more chances molecules of A and B have to get into contact, the faster the reaction could run since chemistry is just atoms bumping into atoms. Stirring or shaking can bring more unreacted A's and B's together.
European energy is super expensive right now, fridges normally use quite a lot of energy, and peltier fridges use quite a lot more than normal fridges.
In case you are unaware: The European efficiency rating system is in fact percentile-based on some fictional reference appliance which is adjusted every few years [0].
A fridge with an "A" rating you bought 15 years ago would not get an A rating today.
(This sounds weird, but fights the problem that when the system was based on absolute values instead of percentiles, nearly fridges started being rated A, A+, A++ or A+++ to the point that the scale was very opaque to consumers.)
[0] https://en.wikipedia.org/wiki/European_Union_energy_label#Re...
Well, some of them do.
Unless the genius who designed your kitchen put the fridge directly next to the dishwasher, whose job it is to keep hot steam inside for a long time.
Yeah, don't put the box full of hot steam next to the box filled with cold stuff.
(I rent, so I can change nothing about this arrangement.)
It doesn't appear this article will make it clear if such a thing is possible, since it lacks any hard numbers relating to the technologies it is comparing.
Here's an example: https://www.amazon.com/dp/B07KZLJ7PB
Peltier coolers have one thing going for them- they have no moving parts*. That's it. Which means they still have their niche! But everything else about them is awful compared to other cooling solutions.
*in practice most applications use fans for the heat sinks.
The basic way of working it is basically
* lowering the pressure of gas causes it to cool, increasing it causes it to heat
* creating standing wave in a gas in pipe means the low pressure part will be cold, high pressure part will be hot, and the difference will equal to basically pressure induced thermal difference
sooo a fancy speaker in the tube. Already stuff that can be bought, just didn't get to the level required for consumer stuff.
Supposedly they can be pretty efficient too. Would also be basically noiseless, as the ultrasonics are limited to the cooling tube and easier to dampen anyway.
On the other end, you can for some applications, use "R-290" which is just highly pure propane. Typically waist high commercial ice cream freezers will use a fully sealed R-290 system.
And finally there is also "ammonia phase change" refrigeration, which can be done with a compressor, but also a "generator" which is essentially just a propane or electrical boiler which creates the high pressure side of the system, also with no moving parts. You will typically find these in highly portable and mobile applications, such as RVs.
Their other experiments are worth a read as well :)
The tubing isn't pinched shut, it's just pinched to restrict the flow. I believe it's to create a bottleneck where pressure can build up between the "hot side" and the "cold side" of the refrigerator. Before the pinch the compressor is driving gas into the pipe, where it meets resistance at the pinch point and compresses to liquid, shedding heat. Then after that point a narrow jet of compressed refrigerant squirts into a lower pressure part of the pipe and starts to evaporate, expanding and taking in thermal energy again.
Of course, the refrigerator already has some valve to do this, but the author mentions that it may not be designed for the pressure differential needed for propane, so you can make your own with a pinched tube.
Liquid refrigerant flow is restricted when it flows through the orifice. This causes the pressure to drop which causes the refrigerant to flash/boil. While the refrigerant is flashing/boiling, it is absorbing heat from the surrounding environment.
Ideally the charge amount should be such that the evaporator is fully frosted, but not much more than that.
An interesting project I've seen using one of them is this portable chiller suit to cool the human body on-the-go, intended for relief of Multiple Sclerosis symptoms [2].
For DIY/prototyping, there are a bunch of Chinese manufacturers now producing compressors of these sizes along with the appropriate driver board for the variable-speed AC motor [3]. I'm excited to see how this technology will be applied elsewhere in time.
[1] https://aspensystems.com/vapor-compression-refrigeration/
[2] https://hackaday.io/project/4729-bringing-cool-relief-to-mul...
[3] https://web.archive.org/web/20220916231432/https://www.aliex... (archived as of time posted for future readers)
Isn't part of what makes these work also linear compressors? The old on/off models have gotten better over time, but when we were kids, some of them would practically cause brownouts on the electrical circuit. Too much startup current. So for electronics you need someone with a smoother transition.
That's because a compressor has an odd property... The motor requires substantially more starting torque than running torque. That's because the cylinder can be stuck on the compression cycle pushing against a cylinder of gas, and has no momentum to get it through. Also, the very first compression typically has far more gas in it (because the 'cold side' is actually warm before it starts).
If nothing were done to compensate, the motor would stall, overheat, and burn out.
Designers of fridge compressors have instead added a very high current starting coil, which will provide massive amounts of torque to get the thing going for a few seconds. Sometimes that might be 20 amps or more for a few seconds. When the motor is spinning fast enough, that coil gets disconnected.
On most air conditioning compressors today, a scroll plate compressor is used, which doesn't have the unloading problem of the piston type of compressor.
Just replied to the same comment wondering if this has changed. I used to help my dad install residential and commercial HVAC systems and all of the compressors back then were piston type. I would tick him off by getting impatient when we were testing by calling for cool too soon after shutting it down and tripping the thermal beaker. (Come on dad it's hot up here!!! lol)
I imagine the scroll type are probably quieter as well and probably work nicely with inverter-based drives.
I did go back and look that there are sometimes valves used in scroll AC compressors. Even with that the full compression cycle is much more gradual, and continuous happening over ~3 revolutions of the compressor instead of half a revolution for a piston so there's not a slug compressed gas to hold the compressor back in the same way as a piston compressor.
This pages has some diagrams that shows the spiral plates and compression cycle pretty well. https://www.sciencedirect.com/topics/engineering/scroll-comp...
Not sure if that applies to newer systems.
> Vaseline is popular as it is a fairly easy application and can usually be removed by running the motherboard through a cycle in the dishwasher.
https://www.icc-usa.com/prepping-a-system-for-subzero-overcl...
People trying to break over clocking records just use liquid nitrogen: https://m.youtube.com/watch?v=WwJvHJ1hyto
How much power do those mini refrigeration compressors draw?
https://singledoorrefrigerated.com/red-bull-mini-fridge-for-...
There are now also some similarly sized compressors from other manufacturers as well, some that run on 12v. Lots of other options other than aspen.
Either I am misunderstanding what is being said there, or the article author misunderstands how physics works.
The self heating (for better or for worse) of a peltier is generated at the cooling chamber edge.
This is in excess of the heat being removed from the refrigerated area.
Air conditioners have the warm-running compressor outdoors, but refrigerated indoor appliances release that heat into the room along with the heat transferred from inside the refrigerator. And then the AC encounters further inefficiency as it next removes this heat from the room.
Even the motor of a ceiling fan warms up a room (gradually) if there is no AC and windows/doors are not open adequately to let that heat escape.
Ceiling fans have big motors and are best outdoors like at a tiki bar.
https://www.meerstetter.ch/customer-center/compendium/71-pel... has some nice plots and explanation of how performance varies with deltaT.
A constant current driver can also be important - PWM results in increased joule heating. For 50% maximum constant current joule heating is (V/2)^2/R but when duty-cycling with 50% PWM, it's V^2/(2*R). Moving the scaling factor out from under the ^2 makes the savings more pronounced at lower deltaT as well. But constant current is more complex, especially with larger peltiers.
I cants really speak for Carnot cycle coolers, but they can often be effectively duty-cycled and can be placed remotely so the compressor waste heat can be vented into the room. Performance still degrades with dT though.