I recently purchased a cheap (~$25) 36W thermoelectric cup holder designed to heat or cool a soda can sized object. It works very quickly for cooling and draws about 3A at 12V. For heating, it heats about 4 times slower and draws only about 0.7A. The fan only runs when it is in cooling mode. According to the vendor, it is not unusual for it to heat slower than it cools, and it is normal for the fan not to run in heating mode.

The lack of fan and lower current in heating mode leads me to believe that either it is not using the TEC for heating mode, or that for some reason it only runs the heating mode at 1/4 the TEC capacity.

Is there any reason why it would be advantageous to add a separate resistance heater or to run the TEC at 1/4 capacity without the fan for heating purposes? It seems that it would be best to have it heat something as quickly as it cools it by just changing the direction of the current in heating mode.

Eventually I will probably take it apart to see what it is really doing.

  • \$\begingroup\$ Link to spec??? \$\endgroup\$ Oct 2, 2018 at 3:22
  • \$\begingroup\$ @Tony There really isn't a spec that I could find. There is general info at alibaba.com/product-detail/… \$\endgroup\$
    – crj11
    Oct 2, 2018 at 3:31
  • 1
    \$\begingroup\$ We can only speculate, but maybe they want to avoid the heat sink getting cold enough to cause condensation, which could damage the device or your car, depending how it's designed. \$\endgroup\$
    – The Photon
    Oct 2, 2018 at 3:33
  • 2
    \$\begingroup\$ If you look at the diagram for the very inefficient heat engine that is a TEC, you'll see that in cooling mode it will draw perhaps 6 watts of heat out of the drink, but if the TEC is reversed, it will push 40 watts of heat in. This might be why it is designed to run at a much lower current when in heating mode. \$\endgroup\$
    – tomnexus
    Oct 2, 2018 at 4:29

2 Answers 2


It takes more energy to get from room temperature to hot drink temperature than it does to get from room temperature to cold drink temperature, so given the same thermal transfer rate, heating a drink will take longer to begin with.

Peltiers are pretty cool things. The difference between performance in heating and cooling mode likely has to do with several factors. Peltiers have fairly strongly defined optimal operating currents.

At too low a current, the thermal conduction losses of the peltier will dominate, determined by thermal conductivity and temperature differential along with convection as the interior grid of the peltier is likely not insulated.

At too high a current, the heat the peltier produces due to \$I^2R\$ losses through the little bismuth cubes will dominate.

Used as a cooling element, a peltier is a relatively inferior device in terms of efficiency but gets used because of the lack of moving parts, tiny form factor and electrical versatility.

Look at the way it operates in cooling mode. The heat is pumped out of the cold side of the peltier and into the hot side heatsink. The cooler you keep the hot side of the peltier, the better your efficiency will be. The hotter the heat sink gets, the more difference there will be between its temperature and ambient, and the more heat will be shed into a given volume of air. Note that all losses in the system will contribute to higher heatsink temperature. In this mode there is great value to moving air across the heatsink.

Used as a heating element, a peltier has the very special property of offering better than 100% electrical heating efficiency. You don't have to pay for all of the watts of heat with watts of electricity, but it does have to come from somewhere. I'm guessing this is where they've gone with having the device work slower in heating mode. They are avoiding the efficiency losses with using the fan (temperature differential between heatsink and ambient will be smaller in heating mode than in cooling mode and all thermal losses in the system will decrease the temperature differential and therefore decrease the value of using the fan. With a resistive heating element added, you would be able to trade off efficiency to heat the liquid more quickly.

As a final thought, in heating mode, the colder the cold side becomes, the greater the temperature differential over the peltier becomes without the liquid inside the device having to change temperature at all. The total heating ability of the peltier alone is limited to its own losses plus available external thermal energy. They are not incredibly durable devices, being composed of hard ceramic and somewhat brittle metal cubes and they are not immune to the thermal stresses from the differentials they create over themselves. The current the device operates at in heating mode may minimize these stresses, and using a resistive heater would allow the stresses to be minimized by having the heat that otherwise must be produced by losses that could be stressing the peltier produced elsewhere. On the other hand if the heatsink was shielded, you could attempt to drive it far enough below ambient to make using the fan worthwhile to speed up a portion of the heating sequence at a still-higher-than-100% electrical efficiency at risk of stressing the peltier. If you are going to use high temperature differentials, good practice is to cut grooves partway through the ceramic plates in a grid so that if the plate cracks it will crack along these lines. The grooves should be aligned between bismuth cubes and as much as possible not cross copper plates on the side being cut.

  • 1
    \$\begingroup\$ "Peltiers are pretty cool things" -- that's a pretty one-sided opinion. Looking at it from the other end, they're hot. (Sorry. Couldn't resist. Ignore me and proceed). \$\endgroup\$
    – TimWescott
    Oct 2, 2018 at 15:07

When comparing cold vs heating performance if your input power is not the same, the results will not be the same. Insulation and heat conductance or resistance Rja also matters greatly for each thermal layer too. So without your thermal and electrical design specs and results, your results are sub-optimal and abnormal.

  • Although, I don’t know, but if you are keen, research how Mercedes SUV’s optimized their design for hot + cold containers in the front console for coffee and cold beverages, perhaps in patent searches.

The Peltier Effect Modules PEM or coolers wafers (PEC) are about 50% efficient in heat transfer so if you apply 40 watts of heat from P=VI input you get about 20 watts of cooling effect.

The is due to an effect like back EMF in motors where as the heat flow begins it draws less current, so a 20% or more voltage is required to regulate constant P. So a current sensor alone is not an exact indicator of power.

The critical design factors are the thermal resistance [‘C/W] the thermal mass of the heat sink and cups for hot and cold side and the amount of thermal heat loss outside the cup holders. This optimizes the thermal power transfer to the target. In order to achieve maximum efficiency, this low thermal resistance must be like CPU’s ~0.1’C/W and must never be operated without a heat sink.

If there is no cup of liquid to absorb the heat to improve the heat transfer dual heat sinks and dual high velocity fans on either are needed side to create the large temperature gradient.

When this is designed correctly a 50’C difference can be centred around a room temp of 23’C and thus achieve sub-zero temperature and hoist beverage temperatures above 50’C.

However the time duration to transfer this power to the liquids depends on the mass available to absorb the difference on each side.

Keep in mind that if your hot side does not rise as fast as the cold side reduces, either your input power is not the same for each condition or you thermal resistance on the cold or hot sides is not low enough.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.