A voltage reference is an electronic device (circuit or component) that produces a fixed (constant) voltage irrespective of the loading on the device, power supply variation and temperature. It is also known as a voltage source, but in the strict sense of the term, a voltage reference often sits at the heart of a voltage source.
The distinction between a voltage reference and a voltage source is, however, rather blurred especially as electronic devices continue to improve in terms of tolerance and stability.
Voltage references are used in ADCs and DACs to specify the input or output voltage ranges.
The most common voltage reference circuit used in integrated circuits is the bandgap voltage reference. A bandgap-based reference (commonly just called a ‘bandgap’) uses analog circuits to add a multiple of the voltage difference between two bipolar junctions biased at different current densities to the voltage developed across a diode. The diode voltage has a negative temperature coefficient (i.e. it decreases with increasing temperature), and the junction voltage difference has a positive temperature coefficient. When added in the proportion required to make these coefficients cancel out, the resultant constant value is a voltage equal to the bandgap voltage of the semiconductor. In silicon, this is approximately 1.25V. Buried zener references can provide even lower noise levels, but require higher operating voltages which are not available in many battery-operated devices.
The Design of Band-Gap Reference Circuits
The band-gap reference has been a popular analog circuit for many years. In 1971, Robert Widlar introduced the LM113, the first band-gap reference.1 It used conventional junction-isolated bipolar-IC technology to make a stable low-voltage (1.220 V) reference. This type of reference became popular as a stable voltage reference for low-voltage circuits, such as in 5-volt data acquisition systems where zener diodes are not suitable. Band-gaps are also used in digital ICs such as ECL, to provide a local bias that is not adversely affected by ambient noises or transients.
The principle of the band-gap circuit is well known and will be mentioned here in the briefest terms. The circuit relies on two groups of transistors running at different emitter current densities. The rich transistor will typically run at 10 times the density of the lean ones, and a factor of 10 will cause a 60 millivolt delta between the base-emitter voltages of the two groups. This delta voltage is usually amplified by a factor of about 10 and added to a Vbe voltage. The total of these two voltages adds up to 1.25 volts, typically, and that is approximately the band-gap of silicon.
More details at http://www.national.com/rap/Application/0,1570,24,00.html
Voltage Regulators and References