Since thermoelectric cooling systems are most often compared to conventional systems, perhaps the best way to show the differences in the two refrigeration methods is to describe the systems themselves.

A conventional cooling system contains three fundamental parts - the evaporator, compressor and condenser. The evaporator or cold section is the part where the pressurized refrigerant is allowed to expand, boil and evaporate. During this change of state from liquid to gas, energy (heat) is absorbed. The compressor acts as the refrigerant pump and recompresses the gas to a liquid. The condenser expels the heat absorbed at the evaporator plus the heat produced during compression, into the environment or ambient.

A thermoelectric has analogous parts. At the cold junction, energy (heat) is absorbed by electrons as they pass from a low energy level in the p-type semiconductor element, to a higher energy level in the n-type semiconductor element. The power supply provides the energy to move the electrons through the system. At the hot junction, energy is expelled to a heat sink as electrons move from a high energy level element (n-type) to a lower energy level element (p-type).

Thermoelectric Coolers are heat pumps – solid state devices without moving parts, fluids or gasses. The basic laws of thermodynamics apply to these devices just as they do to conventional heat pumps, absorption refrigerators and other devices involving the transfer of heat energy.

An analogy often used to help comprehend a T.E. cooling system is that of a standard thermocouple used to measure temperature. Thermocouples of this type are made by connecting two wires of dissimilar metal, typically copper/constantan, in such a manner so that two junctions are formed. One junction is kept at some reference temperature, while the other is attached to the object being measured. The system is used when the circuit is opened at some point and the generated voltage is measured. Reversing this train of thought, imagine a pair of fixed junctions into which electrical energy is applied causing one junction to become cold while the other becomes hot.

Figure 1: Cross Section of a typical TE Couple

Thermoelectric cooling couples (Fig. 1) are made from two elements of semiconductor, primarily Bismuth Telluride, heavily doped to create either an excess (n-type) or deficiency (p-type) of electrons. Heat absorbed at the cold junction is pumped to the hot junction at a rate proportional to current passing through the circuit and the number of couples.

Figure 2: Typical TE Module Assembly

In practical use, couples are combined in a module (Fig. 2) where they are connected electrically in series, and thermally in parallel. Normally a module is the smallest component commercially available.

Modules are available in a great variety of sizes, shapes, operating currents, operating voltages and ranges of heat pumping capacity. The present trend, however, is toward a larger number of couples operating at lower currents. The user can select the quantity, size or capacity of the module to fit the exact requirement without paying for excess power.

There is usually a "need" to use thermoelectrics instead of other forms of cooling. The "need" may be a special consideration of size, space, weight, reliability and environmental conditions such as operating in a vacuum. If none of these is a requirement, then other forms of cooling should be considered and in fact are probably desirable.

Once it has been decided that thermoelectrics are to be considered, the next problem is to select the thermoelectric(s) that will satisfy the particular set of requirements. Three specific system parameters must be determined before device selection can begin.

These are:

• T_C Cold Surface Temperature
• T_H Hot Surface Temperature
• Q_C The amount of heat to be absorbed at the Cold Surface of the T.E.

In most cases the cold surface temperature is usually given as part of the problem - that is to say that some object(s) is to be cooled to some temperature. Generally, if the object to be cooled is in direct intimate contact with the cold surface of the thermoelectric, the desired temperature of the object can be considered the temperature of the cold surface of the T.E. (T_C). There are situations where the object to be cooled is not in intimate contact with the cold surface of the T.E., such as volume cooling where a heat exchanger is required on the cold surface of the T.E. When this type of system is employed the cold surface of the T.E. (T_C) may need to be several degrees colder than the ultimate desired object temperature.

The Hot Surface Temperature is defined by two major parameters:

1) The temperature of the ambient environment to which the heat is being rejected.
2) The efficiency of the heat exchanger that is between the hot surface of the T.E. and the ambient.

These two temperatures (T_C & T_H) and the difference between them (DT) are very important parameters and therefore must be accurately determined if the design is to operate as desired. Figure 3 represents a typical temperature profile across a thermoelectric system.

Figure 3: Typical Temperature Relationship in a TEC

The third and often most difficult parameter to accurately quantify is the amount of heat to be removed or absorbed by the cold surface of the T.E. All thermal loads to the T.E. must be considered. These thermal loads include, but are not limited to, the active or I²R heat load from electronic devices and conduction through any object in contact with both the cold surface and any warmer temperature (i.e. electrical leads, insulation, air or gas surrounding objects, mechanical fasteners, etc.). In some cases radiant heat effects must also be considered.

Single stage thermoelectric devices are capable of producing a "no load" temperature differential of approximately 67°C. Temperature differentials greater than this can be achieved by stacking one thermoelectric on top of another. This practice is often referred to as Cascading. The design of a cascaded device is much more complex than that of a single stage device, and is beyond the scope of these notes. Should a cascaded device be required, design assistance can be provided by Melcor personnel.

Once the three basic parameters have been quantified, the selection process for a particular module or group of modules may begin. Some common heat transfer equations are attached for help in quantifying Q_C & T_H.

There are many different modules or sets of modules that could be used for any specific application. One additional criteria that is often used to pick the "best" module(s) is Coefficient of Performance (C.O.P.). C.O.P. is defined as the heat absorbed at the cold junction, divided by the input power (Q_C / P). The maximum C.O.P. case has the advantages of minimum input power and therefore, minimum total heat to be rejected by the heat exchanger (Q_H = Q_C + P). These advantages come at a cost, which in this case is the additional or larger T.E. device required to operate at C.O.P. maximum. It naturally follows that the major advantage of the minimum C.O.P. case is the lowest initial cost.

Power supply and temperature control are additional items that must be considered for a successful T.E. system. A thermoelectric device is a D.C. device. Any A.C. component on the D.C. is detrimental. Degradation due to ripple can be approximated by:

DT / DTmax = 1 / (1+N²), where N is % current ripple.
Melcor recommends no more than a 10% ripple.

Temperature control can be generally considered in two groups: Open Loop and Closed Loop, or manual and automatic. Regardless of method, the easiest device parameter to detect and measure is temperature. Therefore, the cold junction (or hot junction in heating mode) is used as a basis of control. The controlled temperature is compared to some reference temperature, usually the ambient or opposite face of the T.E..

In the Open Loop method, an operator adjusts the power supply to reduce the error to zero. The Closed Loop accomplishes this task electronically. The various control circuits are too numerous, complex and constantly being upgraded to try to discuss in this text. There are several manufacturers of control circuits and systems that are better equipped to give expert counsel in this specific area. Suffice it to say that the degree of control, and consequent cost, varies considerably with the application.