Light Emitting Diodes (LEDs), from an electrical as opposed to an optical perspective, can be operated using the same drive considerations as other diodes. Similar limiting properties must be taken into account as with standard electrical diodes including maximum reverse voltage and maximum forward current.

LEDs are preferably used in a circuit configuration that controls the current through the device rather than the voltage across the device since this will yield the most stable light output, which is critical in most applications. As with any diode, there are small variations in the junction voltage at a specific current due to unavoidable variations in the manufacturing process. The equations that describe the electrical behavior of an LED are the same as for any other diode and any simple electronics book will serve as a suitable reference.

The temperature dependent behavior of the diode forward voltage (Vf) at a specific drive current (I) implies that at higher drive conditions the electrical power dissipated in the device increases. Basically, power is dissipated in three areas in the device: the p-contact region, the pn junction region, and the n-contact region.

For properly designed devices the voltage drop across the resistances in the p and n regions is much smaller than the junction voltage and hence most power is dissipated in a very narrow volume of semiconductor material near the junction. This volume heats up very quickly, typically on the order of microseconds, and then heat flows out of the device by the lowest thermal impedance path(s) to the external thermal environment. Once thermal equilibrium is achieved the Vf, wavelength, and light output of the device stabilizes.

Overheating of the diode junction adversely affects the lamp performance for many reasons. First, the efficiency of an LED drops with increasing temperature, second, the lifetime of an LED is reduced at higher temperatures, and finally, the packaging material that surrounds the diode can be catastrophically damaged at high temperatures. This last effect can be very significant since it is inherently non-linear. The encapsulants used to package LED reach a glass transition temperature above which the plastic flows very easily, which can cause several deleterious effects in a lamp. To the end user of the LED lamp this implies that there is a maximum junction temperature, Tj, which should not be exceeded whether using DC or pulsed conditions.

There are two basic ways to control the current through a diode. In one case one or more LEDs in series are connected to a standard current supply which is set up to deliver the amount of current specified for a specific application. An advantage of this method is that it accommodates variation in the junction voltage of the LEDs at a specific current level. In the second case LEDs can be run in parallel with a single voltage supply. Normally a one or more series resistors are used to control the current through the diode.

The primary advantage of this scheme is that it’s simple and is readily used where a DC power source is already available (e.g. from a battery in a car). The primary difficulty is that the drive current through the diode will change with its Vf. A practical example will illustrate this.

Assume that standard off-the-shelf diodes have a voltage range of 1.90V to 2.2V at a current of 20mA. One design could have six diodes connected in series (i.e.no series resistor is used to limit the current), and they’re driven with a 12V battery. If the Vf values for each of the diodes is 2.0V, then exactly 20mA flows in the circuit. However, for diodes with Vf of 1.9V the drop at 20mA will be only 11.4V, which implies that they will draw more than the designed 20mA. For diodes with high values of Vf the situation is reversed and less than the design current flows.

The situation is made worse by the fact that in cars the allowed variation in voltage across the battery is very large. In general, it is impractical to drive LEDs with a voltage source without a current-limiting series resistor. The disadvantage of this is that excess power is dissipated in the series resistor which lowers the overall efficiency of the circuit. The conclusion is that each design needs to be optimized to a specific application, and either current or voltage drive circuits can be used with LEDs.

Reverse Voltage Considerations

An LED, like any diode, conducts current easily under forward bias, but blocks the current flow when reverse biased. However, since the devices are optimized for high light output their characteristics, as blocking diodes, are not very good. Typically a device will conduct a few microamps (uA)at –5V, but this leakage current becomes significant at more negative voltages. Therefore, using LEDs in an AC circuit (e.g. at 120VAC) where they both emit light and block current in the reverse direction is not normally recommended. Using LEDs under these conditions can lead to unpredictable performance and a significant reduction in the lifetime of the devices.