Usually, this is a three-terminal resistor where the center connection is manipulable. If two terminals are used (middle and end), it acts as a variable resistor. If all three are used, it acts as a variable voltage divider. A common application for such devices is as a volume control in audio equipment.
Types of potentiometers
Typically, a potentiometer is constructed as a flat circular graphite contact to generate resistance, with a wiper on an axle connected to the center terminal. This wiper typically travels just under one revolution around the contact. However, multiturn potentiometers also exist, where the contact may be helical and the wiper may move 10, 20, or more turns. In addition to graphite, other materials may be used for the resistive element, such as resistance wire, carbon particles in plastic or a ceramic/metal mixture. One popular form of rotary potentiometer is called a string pot. It is a multi-turn potentiometer with an attached reel of wire turning against a spring. It's very convenient for measuring movement.
A potentiometer may also be linear, to provide a slider control instead of a dial control. In this context, the word linear describes the geometry of the resistive element - a rectangular strip, not an annulus as in a rotary potentiometer.
A rheostat is essentially a potentiometer, but is usually much larger, designed to handle much higher voltage and current. Typically these are constructed as a resistive wire wrapped around a toroidal core with the wiper moving over the upper surface of the toroid, sliding from one turn of the wire to the next.
Potentiometers are available with both linear and logarithmic resistance-position characteristics. In this sense, linear describes the electrical response of the device, not the geometry of the resistive element. A linear pot ("type B") has a resistive element of constant cross-section, resulting in a device where the resistance between the wiper and one end terminal is proportional to the distance between them. A log pot ("type A") has a resistive element that either tapers in a controlled way from one end to the other, or is made from a material whose resistivity varies from one end to the other. This results in a device where the distance between the wiper and one end terminal is proportional to the logarithm of the resistance between them. The log pot is used as the volume control in audio amplifiers, where it is also called an "audio pot", because the amplitude response of the human ear is also logarithmic. It ensures that, on a volume control marked 0 to 10, for example, a setting of 5 sounds half as loud as a setting of 10. There is also an anti-log pot or reverse audio taper ("type C") which is simply the reverse of a log pot. It is almost always used in a ganged configuration with a log pot, for instance, in an audio balance control.
Theory of operation
A simple potential divider setup consists of a circuit with a cell and a resistor. The components of another circuit are connected at one end to a terminal of the cell, and the other end to a sliding contact which can vary resistance in parallel to the second circuit, thereby varying the potential difference across the components in the second circuit.
The resistance arising from the resistance in the potential divider in parallel with the resistance in the second circuit, resulting from its components, is RA. The total resistance, RB, is the resistance not involved in the second circuit plus RA. The ratio of the RA to RB is proportional to the potential supplied to the components in the second circuit.
One of the advantages of the potential divider compared to the variable resistor is that whereas variable resistors have a maximum resistance where some current will always flow, dividers are able to vary the current from maximum to practically zero. Zero voltage occurs in the second circuit when the contact is moved so the resistor is not involved in it. Now the components in the second circuit can be considered as in parallel with a conductor of no resistance. As the total resistance (RT) is equal to the sum of the individual inverse of the resistances in parallel:
1/RT = 1/R1 + 1/R2
RT = (R1 * R2) / (R1 + R2)
As R2 = 0:
RT = (R1 * 0) / R1
The total resistance is shown to be zero. As V = ir, the voltage is also zero. Of course this will never actually be achieved as there will always be a small resistance in the wires.
Historical meaning of 'potentiometer'
The original potentiometer is a type of bridge circuit for measuring voltages. The target voltage is connected across a piece of resistance wire, and a standard electrochemical cell of known voltage is then shorted across a variable-length section of the resistance wire using a sliding contact. The contact is moved until no current flows into or out of the standard cell, as indicated by a galvanometer in series with the cell. We then know that the voltage across the selected section of wire is equal to that of the cell. It is then just a matter of calculating the unknown voltage from the cell voltage and the fraction of the length of the resistance wire that was shorted across the cell. The galvanometer does not need to be calibrated, as its only function is to read zero.