Coleman Thermoelectric Cooler

[Andrew Leigh]

Hi,

my friend was throwing away an older model Coleman Thermoelectric Cooler which he said did not work very well. I took it from him to see if I could fix it for camping so here goes with a couple of questions;

1) Firstly, the contact surface on the plug is corroded to to point that only after vigorous jiggling did the fan come on. I assume this poor contact forms a high resistance joint and impedes current flow?

2) In his attempt to fix the cooler he has used copious quatities of the white thermal transfer paste, can one use too much paste?

3) The peltier device is about 40mm x 40mm seems very small for a 32 quart fridge. Can one tell the "cooling effect" from the amount of current it draws?

4) Have peltiers become more efficient or are they limited by their chemistry?

5) Years ago I got a sample of a grey thermally conductive flexible sheet material made by Courtaulds Aerospace. Intended use was to replace the paste as it was dry and easy to work with, spec's claimed it was more efficient but is was way more expensive. Anyone familiar with the material?


Cheers
Andrew

 

[Sceadwian]

1) Significantly.
2) Yes, thermal paste is actually a good insulator. Both surfaces need to be flat, apply a modest layer of paste to both sides, and then scrape off the excess with a straight edge (like a razorblade). Just remember the thermal paste serves one and only one purpose and that's to eliminate air gaps.
3) The element probably is small those little fridges aren't generally designed very well. Not sure how to go about determine it's cooling power though. Keep in mind the biggest limiting factor is ambient temperature and keeping the hot side of the element as cool as possible, otherwise you can't pump heat from the cold side.
4) Peltiers efficiency is usually pretty low. They've gotten better but not dramatically so, and again they're still highly limited by their ambient air conditions.
5) No idea what that stuff is, I find it on mosfets frequently, wish I knew what it was myself, and where to get sheets of it.

 

[Andrew Leigh]

Hi

Excellent thanks for the rapid reply, will assemble again as per your recommendations and hard wire the peltier for test purposes.

Cheers
Andrew

Would that grey material be better than paste for the peltier?

 

[Andrew Leigh]

Hi,

well it did not take but 5 min for me to feel the evaporator cooling so methinks it was purely poor connections and too much thermal paste.

I cannot control how well the unit performs but am confident that it will perform as well as it did when originally bought.

Thanks again
Andrew

 

[Sceadwian]

Something as simple as adding a fan to the internal and external heat sinks could help performance a bit if it doesn't already have them.

 

[Andrew Leigh]

It has both fans thanks, will see how the temperature progresses during the day.

 

[Andrew Leigh]

Hi,

cleaned the plug contacts, fridge got down to 13 deg C should be 3~5 deg C. Plug still getting hot.

Took the plug out of the circuit and connected directly to the Peltier with the thermal cutout in series. Fridge got down to 6.1 deg C but this was overnight at slightly lower ambients.

Comments read state that Peltiers are innefficient and that the Δt is small. How much would a second Peltier help, to my mind it would certainly bring the temperature down more rapidly but if the cooling ability is 30 dec C below ambient this will be maintained and the fridge would no get cooler?

Will check on the fridge pefromance today, the next step will be to check if the thermal cutout is actually working and if the Peltier is in fact able to cool further.

Cheers
Andrew

 

[colin55]

Each peltier module has a temperature difference. Putting them in series increases the overall temperature drop (rise). The unit I have, has a temperature differential of 20 degrees C. Keeping the hot side as cool as possible is very important. This creates the limiting factor - the factor as to the temperature of the low side of the module.
It is posible that some of the individual "diodes" will fail due to temperature-rise and this will reduce the effectiveness of the unit.
The actual temperature of the hot side and cold side will be created relatively quickly (less than 10 minutes) and this will determine the number of series units in the array.

 

[Andrew Leigh]

I could not see any identifying feature on the peltier I have but it is drawing 4.2A at 12V so that makes it a 50W unit.

Based on a couple of measurements the Δt is between 15 an 17.5°C. I assume the design was also for 20°C but there are inefficiencies / dead "diodes".

Be that as it may the real requirement is to drop 30°C below ambient. The day ambient in the Kruger National Park is 40°C plus in summer. In an aircon car this would be less, probably 30 odd and this would allow the cooler to get to the desired temperature.

I can buy locally the following unit https://www.electro-tech-online.com/custompdfs/2009/11/DT12-8.pdf but I am unable to make head or tail of the temperature differential based on the data sheet, any ideas? They are claiming 66°C in nitrogen?

Cheers
Andrew

 

[colin55]

The dead diodes will not make any difference to the temperature-differential capabilities of the module. The dead diodes simply reduce the amount of energy that can be transferred from one side of the module to the other.
You need to add an evaporative cooler to the hot side. This is a few layers of water-soaked material with a fan blowing over the surface. The material needs to be kept water-soaked. Place a dish under the material and it will "wick-up" as it gets evaporated.
This will lower the hot side by 10 degrees C

Let's be more specific about the temperature-differential capability.
The peltier module in your cooler will reduce the temperature by 20 degrees C. This means that if the outside temperature is 40 degress, the internal temperature of the cooler will be 20 degrees.
The 20 degree differential is not the difference between the hot and cold sides of the module. The actual difference between the hot and cold sides of the module may be as high as 60 degrees.
This means you will get much better performance if you reduce the temperature of the hot side.

 

[Andrew Leigh]

Hi Colin,

looks like I can get a 70W unit of similar spec's (size and cooling). Think I will buy this as a temporary measure and then replace the existing unit. This should give me a further 40% "power". The cooler will probably only suffice as a salad cooler but thats OK will take some pressure of my main camping fridge.

When I get back from vacation will have to do some machine work and will then use the 50W which will stay on permanently. The 70W will then be controlled in order to top up the temperature.

Any thoughts

Regards
Andrew

 

[colin55]

You are not going to gain anything by buying a larger module.
The temperature differential of all these modules is about 20 degrees C and you have to reduce the temperature of the hot side to reduce the temperature of the cold side.

 

[Andrew Leigh]

Hi Colin

That what I thought you might say, ΔT is what it is. However, surely the rate of cooling increases as will the recovery time after the cooler has been opened.

Going to be difficult to get the "wicking" organised due to space. Might be able to weave some proper "wick" into the base of the heatsink (back and forth) and have both ends dangle in water. Having said which, the heatsink barely gets warm when placing a finger behind the location of the peltier so it is not like the fan and heat sink are not working.

Would it help diagnostically if I were to provide measured temperatures say of the both heat sinks and the air over the heatsink?

Cheers
Andrew

 

[tcmtech]

Possibly using a two layer peltier cooler system could give you considerable gains however you will run into a power consumed and size of heat sinks issue at some point.

The better modern peltiers are about 35% efficient at heat transfer.
To make a two layer system work the second layer has to move all of the heat the first layer produces plus what it pumped also. For that you would need three times the capacity on your second layer Peltier cooler units.
And a much larger heat sink in order for them to work efficiently.
Also your power consumption jumps by a factor of four as well.

So if you can swing the added space and power consumption you can get your cooler to work faster and more effectively but not so efficiently.

 

[smanches]

5) Years ago I got a sample of a grey thermally conductive flexible sheet material made by Courtaulds Aerospace. Intended use was to replace the paste as it was dry and easy to work with, spec's claimed it was more efficient but is was way more expensive. Anyone familiar with the material?

Are you talking about Mica sheets? I know they are used to electrically insulate component tabs from heatsinks (like a TO220 mosfet tab that is connected to the source pin.) It's mostly rigid, but can be bent some. Good thermal conductivity, but a good electrical insulator.

 

[Andrew Leigh]

Are you talking about Mica sheets? I know they are used to electrically insulate component tabs from heatsinks (like a TO220 mosfet tab that is connected to the source pin.) It's mostly rigid, but can be bent some. Good thermal conductivity, but a good electrical insulator.

No,

this was a glass cloth substrate impregnated with a conducting silicone resin. It is very flexible.

Andrew

 

[Sceadwian]

Are you sure about than Andrew?

 

[Andrew Leigh]

Certainly the material I have feels like it, will put a flame to the corner of a piece and check if the substrate is glass, the odour should betray the base resin.

 

[Sceadwian]

The only ones I've ever seen look like silicon or teflon sheets, I have no idea what the material is but that's what it feels like, as long as it's thin the main goal isn't to transfer heat but to avoid insulating as well as air would, so even if the material is a crappy heat conductor it's going to be many times better than the air bubbles, it's just there to fill those tiny little gaps and avoid spot heating, the packaging/heatsink materials conduction properties will take care of the rest.

 

[k7elp60]

I have a portable cooler, I don't remember the brand but as I recall they will only cool about 30 to 40 degrees cooler than ambient temperature, and the peilter junctions does take a while to get the cooler cool. When I decide to use mine for a trip, I plug it into the AC line(with an adapter) 24 hours before I want to use it. Then it is cool and keeps the contents cool also.
If a reverse the power plug it will warm up. I don't ever recall using keep anything warm.