Application

The bipolar transistor along with the MOSFET is the most popular transistors in the commercial electronic market. While MOSFETs are more common in high-density circuits, bipolar transistors enable high-speed circuits because of their high transconductance. A good example is the BiCMOS circuit where MOSFETs provide the density and bipolar transistors provide the current drive for large capacitive loads. Another advantage of the bipolar transistor is that the threshold for turn-on is less sensitive to process variation. In bipolar memory applications, soft-error from a-particle is more tolerable. Bipolar transistors are also used in power applications where high voltage (400 V) and high current density are required. The limitations of a bipolar transistor compared to a MOSFET are intrinsic speed constraints due to charge storage, low input impedance, high power dissipation, more complex processing, and large area. For example, in bipolar memory circuits, the chip size is typically 16 times that of MOS technology for the same memory count.

A bipolar transistor can be used as a temperature sensor. It is an improvement over a p-n diode temperature sensor because in a diode, other current components such as recombination current are less predictable and have different temperature dependence. In a bipolar transistor, the collector current is a pure diffusion current from the base-emitter junction while the recombination current only goes to the base.

Related devices Heterojunction Bipolar Transistor

To improve injection efficiency, the emitter of a bipolar transistor needs to have a very high doping level. Beyond a critical doping level, the emitter energy gap decreases and, as a result, the injection efficiency drops. The heterojunction bipolar transistor (HBT) was proposed by Shockley in 1951, and was analyzed in more detail by Kroemer in 1957. Practical HBTs started to be built in the mid-1970s with the emergence of LPCVD, MBE and MOCVD technologies. They incorporate a heterojunction in the emitter-base junction, with the emitter having a larger energy gap.

The base region can be doped more heavily to reduce the base resistance. The emitter can also be doped more lightly to reduce the emitter capacitance. The injected electrons have higher energy and higher velocity to reduce the base transit time, resulting in an intrinsically faster device. An additional heterojunction at the collector, sometimes called double-heterojunction bipolar transistor (DHBT), minimizes minority carriers injected into the base from the collector in the saturation mode. HBTs are usually realized in compound semiconductors due to the availability of heterojunctions of minimal lattice mismatch. HBT on Si bipolar has been studied using SIPOS (semi-insulating polycrystalline-Si) as the emitter.

Darlington Amplifier

A Darlington amplifier consists of two connected cascaded bipolar transistors. The emitter of one transistor is fed to the base of the other transistor. The drawback of a Darlington amplifier is that a larger voltage CE is required for the output.

Tunneling-Emitter Bipolar Transistor

The emitter injection inefficiency can be improved in a tunneling-emitter bipolar transistor by inserting a tunneling barrier near the emitter-base junction. Since tunneling is related to the product of the effective mass and the barrier height the injection efficiency is improved.

CHAPTER 19

SILICON-CONTROLED RECTIFIER

The SCR (silicon-controlled rectifier or semiconductor-controlled rectifier) is the parent and still the main member of the thyristor family. It is the workhorse of high-power electronics and is sometimes simply referred to as a thyristor. (A thyristor is loosely defined as a device having a four-layer p-n-p-n structure, leading to bistable behavior.) The first p-n-p-n structure was described as a bipolar transistor with a p-n hook-collector by Shockley in 1950. The current-gain mechanism of this hook-collector transistor was further analyzed by Ebers in 1952. The switching characteristics of a two-terminal p-n-p-n structure were first explored by Moll et al. in 1956. Subsequently, the control of switching using a third terminal was examined by Mackintosh and by Aldrich and Holonyak in 1958.

Since the 1960s, the SCR has significant commercial value in the power electronic industry. The SCR is the basis for many specialized but closely related devices.

An SCR has a four-layer p-n-p-n structure. The outermost n- and p-terminals are called cathode and anode, respectively, and the contact to the p-base is the gate (or cathode gate). The top view of the device usually has a circular shape, with the gate located at the center. Notice that the n-base layer is much thicker, from a few microns to a few hundred microns, has a much lower doping density ( ), and is designed to support a large blocking voltage. The impurities usually are incorporated by diffusion into lightly doped starting material, and the two p-regions (( ) can be diffused at the same time. SCRs with high-power capability are discrete devices. They are mounted on pedestals with a good heat sink to dissipate generated heat. The anode is usually bonded onto the package since the gate terminal is near the cathode and needs to be connected separately. As the name implies, SCRs are made of silicon because of its good thermal conductivity, high-voltage and high-current capability, and more-mature technology.