I need to come up up a way to refridgerate a small (24"W,14"D,18"H) case for displaying pies. I need to hold the internal temperature at about 39F degrees. My plan was to put a few peltier junctions across the top, cold side heat sinked to the inside of the box, hot side heat sinked to the room. Usual fans et al across the top. I have enough micro experience to write the code that turns on one/more based on the a temperature sensor, and to run the relays, but candidly, I have two areas where I am going to wander.
1) How many do I need ? The junctions I have in mind are 1.58"x1.73". The normal room temperature is 75. The web site lists the following
deltaT=79deg,
Imax=8.1A,
Vmax=16.1V,
Thot=50deg C,
Qmax=80.6W
How can I tell from this how many I should use ? Trial and error is my backup plan, but I would like to know.
2) Power Supply - One can wander for days on the right power supply. Anyone got a recommendation on a modestly inexpensive power supply ? I'm presuming (Based on above) that I can run anything from 12V to 16. but with amps being 8-5a*n-junctions. They will all be controlled independently, so running all at once should be a rare. (Pies are cold going in, so I don't need blast chiller speeds).
Any additional thoughts would be appreciated. I need to keep the cost low but not absurdly so.
Thanks
Peltier Project
Re: Peltier Project
The first issue is heat gain through your enclosure. The coolingwdicken wrote: My plan was to put a few peltier junctions across the top, cold side heat sinked to the inside of the box, hot side heat sinked to the room.
1) How many do I need ?
2) Power Supply
devices are there to remove heat that leaks into the enclosure.
Adding fresh content at a higher temp will also place additional
loading on the cooling system.
The air inside the chamber should also be stirred by a fan,
as the main cooling mechanism is conduction. Adding low-e
glass or reflective materials will reduce the heating from IR
(sunlight) sources.
Commercial cool displays use mechanical refrigeration. This
may be the most practical solution, as Peltier devices are only
5 - 10% efficient, compared with mechanical pumps at 40-60%.
The Peltier devices use multiple junctions, which have a
very low potential difference each. A module has multiple
junctions in series, to make the total voltage practical.
Once you know the total number of junctions required for
maximum cooling you will know the voltage and current
required to power the entire array.
Operation at lower power (when maintenance cooling
is required) is just a case of modulating the DC power
supply, and this can be made more efficient with PWM
control, and further refined with feedback control from a
temp sensor.
A popular voltage will open the door to using a commercial
power supply, readily available for surplus vendors.
(12V, 24V, 28V, 48V are common).
Some years ago portable Peltier based coolers that plug
into the car DC socket (aka cig lighter) were introduced.
These appear to have faded from popularity, possibly
due to initial cost or a lack of performance in the real world.
Re: Peltier Project
You're better off with direct contact between the pies and peltier devices. Consider a design where the shelves in the case are directly cooled from the cold side of the device and the hot side is connected to a heatsink with heat pipes that conduct the heat outside the case to the evaporators (heat sink fins). Conversely the hot heat sink can be hidden in the base and cooled by fans and only the lower shelf is directly cooled.
Cooling from the top simplifies construction but decreases efficiency (since you have to cool all the air before the air can cool the pies). These devices consume significant amounts of DC current. The cost of a power supply plus the cost of electricity for a year could in fact out price a conventional refrigeration unit (like a glass door wine cooler). These are not the LEDs of the refrigeration world.
I've been doing a few experiments myself. With a 2" square device with a AL block about 3" square and 1/3" thick on the cold side and a large finned heatsink on the hot side. This setup eventually stabilizes at about 8C but once the hot side gets warm, the cold temp creeps up. I am getting a Delta T of about <~20C until I cool the hot side with a fan and it goes back to 8C. I'm using an H bridge, a 12V, 9A power supply and a standard temp controller (with PWM capability).
You might do better to start with the cold plate in a 1.5 CF mini fridge, cut it out of that box and install it in your case.
As for the case, double pane glass would be best but cooling the shelves directly will reduce heat gain where you need it (at the pie)
Note these devices are popular in cars because a high current DC supply is already there and free to use. AC systems require a rather expensive high current DC supply (a similar issue comes up when one tries to use a car amplifier in the home on AC power). The peltier device itself is cheap, just look at eBay. To design this efficiently, you will need to learn to calculate heat flow across a heat sink. A peltier does not make cold, it just pumps the heat one way, it must be used with appropriate heat sinks and thermal mass to ensure stable and efficient operation.
Pretend you have an IC chip that dissipated 80W of power and you want to keep it at room temp, now you need a heatsink that removes 80W/25C. Take a look at the CPU coolers, they are very efficient. Also look at overclocking websites, they have lots of ideas for using TEC (Thermo Electric Coolers)
Cooling from the top simplifies construction but decreases efficiency (since you have to cool all the air before the air can cool the pies). These devices consume significant amounts of DC current. The cost of a power supply plus the cost of electricity for a year could in fact out price a conventional refrigeration unit (like a glass door wine cooler). These are not the LEDs of the refrigeration world.
I've been doing a few experiments myself. With a 2" square device with a AL block about 3" square and 1/3" thick on the cold side and a large finned heatsink on the hot side. This setup eventually stabilizes at about 8C but once the hot side gets warm, the cold temp creeps up. I am getting a Delta T of about <~20C until I cool the hot side with a fan and it goes back to 8C. I'm using an H bridge, a 12V, 9A power supply and a standard temp controller (with PWM capability).
You might do better to start with the cold plate in a 1.5 CF mini fridge, cut it out of that box and install it in your case.
As for the case, double pane glass would be best but cooling the shelves directly will reduce heat gain where you need it (at the pie)
Note these devices are popular in cars because a high current DC supply is already there and free to use. AC systems require a rather expensive high current DC supply (a similar issue comes up when one tries to use a car amplifier in the home on AC power). The peltier device itself is cheap, just look at eBay. To design this efficiently, you will need to learn to calculate heat flow across a heat sink. A peltier does not make cold, it just pumps the heat one way, it must be used with appropriate heat sinks and thermal mass to ensure stable and efficient operation.
128W of electricity removes 80W of heat. With Thot=50C you need a rather large heatsink and you could burn youself on it so it needs to be physically isolated and fan cooleddeltaT=79deg,
Imax=8.1A,
Vmax=16.1V,
Thot=50deg C,
Qmax=80.6W
Pretend you have an IC chip that dissipated 80W of power and you want to keep it at room temp, now you need a heatsink that removes 80W/25C. Take a look at the CPU coolers, they are very efficient. Also look at overclocking websites, they have lots of ideas for using TEC (Thermo Electric Coolers)
Re: Peltier Project
Hi,
For your delta T and your enclosure size this could require as many as 25 of your chosen
devices. Obviously this is a lot and will require quite a bit of power too.
The trick then is to insulate the top, bottom, sides, and back with thick insulation where
ever there will be no P devices. This reduces the number required to 2 devices if they
work the way they should, or use 4 just to make sure.
These devices require heat sinks on *both* sides in order to both keep them cool *and*
also distribute the cooler air into the chamber, along with fans on *both* sides.
The devices should be mounted so that the cold heatsink is inside the chamber while
the hot heatsink is outside the chamber, meaning they will mount through a cutout in
the top or back.
Doubling the front panel plexiglass would help keep the cool inside. When the panel
is doubled, keep a small space between panes as this will increase the effective
insulation property of the panel. This is done by using thin strips of material
at the edges to maintain some spacing.
It is probably a good idea to build it so that you can get it apart for cleaning.
There is also the possibility of liquid cooling by running a liquid through a pipe and
use a radiator inside the chamber. The other side of the system could be in a
freezer with a heat exchanger.
NOTES
1. Without an inside fan the chamber may develop hot spots which could damage the pies
over time, as well as prevent the P devices from working efficiently.
2. Without an outside fan the devices may overheat.
3. You may have to check local laws whenever you are dealing with devices that
hold food in storage for sale to the general public.
4. A good idea when dealing with something as variable as this would be to build a
small scaled down version and test it. Perhaps build a single (one) pie unit first that
only holds one single pie and go from there. You can always use this in another
location later once you get the big version up and running. From there you can
scale up to whatever size you need.
For your delta T and your enclosure size this could require as many as 25 of your chosen
devices. Obviously this is a lot and will require quite a bit of power too.
The trick then is to insulate the top, bottom, sides, and back with thick insulation where
ever there will be no P devices. This reduces the number required to 2 devices if they
work the way they should, or use 4 just to make sure.
These devices require heat sinks on *both* sides in order to both keep them cool *and*
also distribute the cooler air into the chamber, along with fans on *both* sides.
The devices should be mounted so that the cold heatsink is inside the chamber while
the hot heatsink is outside the chamber, meaning they will mount through a cutout in
the top or back.
Doubling the front panel plexiglass would help keep the cool inside. When the panel
is doubled, keep a small space between panes as this will increase the effective
insulation property of the panel. This is done by using thin strips of material
at the edges to maintain some spacing.
It is probably a good idea to build it so that you can get it apart for cleaning.
There is also the possibility of liquid cooling by running a liquid through a pipe and
use a radiator inside the chamber. The other side of the system could be in a
freezer with a heat exchanger.
NOTES
1. Without an inside fan the chamber may develop hot spots which could damage the pies
over time, as well as prevent the P devices from working efficiently.
2. Without an outside fan the devices may overheat.
3. You may have to check local laws whenever you are dealing with devices that
hold food in storage for sale to the general public.
4. A good idea when dealing with something as variable as this would be to build a
small scaled down version and test it. Perhaps build a single (one) pie unit first that
only holds one single pie and go from there. You can always use this in another
location later once you get the big version up and running. From there you can
scale up to whatever size you need.
LEDs vs Bulbs, LEDs are winning.
-
- Posts: 458
- Joined: Thu Feb 05, 2004 1:01 am
- Location: Minneapolis, MN
- Contact:
Re: Peltier Project
The company I work for has a subassembly in a larger machine that is similar in size and must be kept at around 10 deg. celsius. We use three large (about 3"x3") Peltiers, but I don't remember how much current they use. I do remember that it's a lot and I vaguely remember writing the code to shut them down if they get too hot (fans can fail and then everything goes south rapidly). The walls are about 1" thick and filled with insulation.
You'll also need to circulate the inside air or you'll get steep temperature gradients.
How are you going to handle condensation? Pies have a lot of water content that will evaporate and condense on the cold walls and may end up making the pies soggy from being dripped on.
My suggestion would be to purchase a small refrigeration unit, or use a mini fridge and duct cool air into the pie box. Either of these will have a dehumidifier to pull moisture out of the air.
You'll also need to circulate the inside air or you'll get steep temperature gradients.
How are you going to handle condensation? Pies have a lot of water content that will evaporate and condense on the cold walls and may end up making the pies soggy from being dripped on.
My suggestion would be to purchase a small refrigeration unit, or use a mini fridge and duct cool air into the pie box. Either of these will have a dehumidifier to pull moisture out of the air.
Re: Peltier Project
Rather than a milti-pie case, you might find it easier to make a platform with a glass dome (bell jar) over it. One per pie.
Put the hot side and heatsink with fan and P/S in the base and cool the plate the pie sits on directly. Condensation should not be a problem if you can sell the pie in a day. Store the leftovers in a regular fridge overnight unless you also have a cold case.
Since 128W of electricity removes 80W of heat the hot side heatsink needs to be 60% larger than the cold side (or 60% more efficient). The CPU coolers made for PCs are extremely efficient. Using 3" peltier devices per CPU cooler on the hot side and all pressed against a large Al plate on the cold side.
Another post today reminded me that your completed project should have both a high limit thermostat on the hot heatsink and a thermal fuse rated 10C above that so that the system can shut down instead of melt down in the event of a fault. This is in addition to the thermocouple you need on the cold side to regulate temp. Thermostats and thermal fuses both work in series with the main DC supply so that power is cutoff if it gets too hot. A thermal fuse should cut all power while a high limit should cut power to the device only.
BTW, I googled "Pie Cooler" and there are lots of them on all the auction sites and craigs list. Seems the economy has put a few bakers out of business.
Put the hot side and heatsink with fan and P/S in the base and cool the plate the pie sits on directly. Condensation should not be a problem if you can sell the pie in a day. Store the leftovers in a regular fridge overnight unless you also have a cold case.
Since 128W of electricity removes 80W of heat the hot side heatsink needs to be 60% larger than the cold side (or 60% more efficient). The CPU coolers made for PCs are extremely efficient. Using 3" peltier devices per CPU cooler on the hot side and all pressed against a large Al plate on the cold side.
Another post today reminded me that your completed project should have both a high limit thermostat on the hot heatsink and a thermal fuse rated 10C above that so that the system can shut down instead of melt down in the event of a fault. This is in addition to the thermocouple you need on the cold side to regulate temp. Thermostats and thermal fuses both work in series with the main DC supply so that power is cutoff if it gets too hot. A thermal fuse should cut all power while a high limit should cut power to the device only.
BTW, I googled "Pie Cooler" and there are lots of them on all the auction sites and craigs list. Seems the economy has put a few bakers out of business.
Re: Peltier Project
Interesting ideas all.
I guess you all have convinced me that it is just not worth doing, or at least not the way I wanted to do it. I'll do a bit more research, but the one pie per case idea sounded interesting.
Thank you all for input. I appreciate the thought you all put into this.
I guess you all have convinced me that it is just not worth doing, or at least not the way I wanted to do it. I'll do a bit more research, but the one pie per case idea sounded interesting.
Thank you all for input. I appreciate the thought you all put into this.
-
- Posts: 1917
- Joined: Fri Aug 22, 2003 1:01 am
- Location: Izmir, Turkiye; from Rochester, NY
- Contact:
Re: Peltier Project
peltier would be perfect for a pie transporter though. Run off 12V supply to precool the box, put pie in, unplug from power supply and carry to vehicle, plug into 12V in vehicle.
Dale Y
Re: Peltier Project
How is that different or better than a low tech cooler withdyarker wrote:peltier would be perfect for a pie transporter though. Run off 12V supply to precool the box, put pie in, unplug from power supply and carry to vehicle, plug into 12V in vehicle.
freezer bricks (whatever they call those gel bags), or
even a block of ice?
Re: Peltier Project
Aside from not needing any ice or another freezer, the TEC can potentially run for days and can maintain more than one set temp (32F). Extended storage on ice will freeze the pie (or melt the ice), a cooler could be set above freezing. A well designed TEC cooler, can reach operating temp very quickly. (if you start with water or a warm gel pack, it could take hours)
For short trips, there is little practical difference.
The world loves gadgets, even when a low tech solution is readily available.
For short trips, there is little practical difference.
The world loves gadgets, even when a low tech solution is readily available.
Who is online
Users browsing this forum: No registered users and 20 guests