What are the differences between SCR, IGBT, and MOSFET transistors for an inverter circuit? Which on

These are all completely different devices and used in different applications. All can be used for switching power on and off. SCRs (Silicon Controlled Rectifiers) are nonlinear and cannot be used in linear amplifiers. SCRs also can only be turned on with the control terminal. An SCR will remain conducting until the switched current goes to zero, which greatly limits their use.

All these devices are DC, they can be used in AC circuits but require more components and design time. For many AC circuits a triac is used.

IGBTs, MOSFETs and BJTs all can both increase and decrease the over the output current over a wide range, about a factor of a million, with the control terminal.

SCRs

An SCR is a four layer semiconductor, PNPN where the outside P is the anode. This is a positive feedback circuit, one current is supplied through the gate terminal it is multiplied be the two BJTs and fed back into the gate terminal.

A PNPN structure is often accidentally constructed in modern IC layout and causes what is called “Latchup” where at a certain supply voltage enough leakage current gets into the internal PN layers to trigger the current multiplication. Latchup can draw large amounts of current and typically results in interconnect being melted.

The dopant levels and distances between the SCR layers are not very critical, you only need an overall gain through both transistors of more than one for it to work. SCRs have been in production since the late 1950s. SCRs can handle very large currents with a low “on” voltage at high current and reasonably fast switching time. For the highest power circuits they can be switched on at moderate voltage, limiting the power dissipated during switching. They always turn off at zero current where they do not dissipate power during the switch time.

SCRs are used in AC circuits, one set for the positive excursions and one for the negative, to control power factor and provide speed or power control. Because it acts as a switch when voltage is present it generates current spikes and voltage changes that can cause interference and harmonics. They are also used in pulse width AC power regulation and motor speed controls.

BJT

Bipolar Junction Transistor, BJTs, are a three layer semiconductor, PNP or NPN. The doping values of the three are much more critical than in an SCR and high gain (beta) transistors took longer than SCRs to be optimized. A BJT has a pretty constant ratio of base current to collector current over a very wide range.

These transistor are used in all types of amplification and were the primary three terminal semiconductor from the 1960s through the 1980s. They are still used as discrete devices and in integrated circuits. The basic BJT was improved with heterojunction, using a different semiconductor for the emitter and base, to allow higher gain and higher frequency response. The most common heterojunction now is silicon-silicon germanium for the emitter and base respectively.

BJT construction uses a highly doped emitter, moderate doped base and a low doped collector. This arrangement allows high current gain and high emitter to collector voltage.

BJTs can be low current or high current capability by scaling the area and allowing high heat dissipation in the package.

A BJT has a low base to emitter impedance and a high collector to emitter impedance. You can use it to design amplifiers with high or low input and output impedance in different arrangements, common emitter, common base and common collector.

Because it requires base current for operation there is always at least a low, it can be microamps, current required on the base for operation.

BJTs are called minority carrier because the collector current actually passed through the base of the device as a minority current in the base.

MOSFET

MOSFETs, Metal Oxide Semiconductor Field Effect Transistors, have a P or N type connection between the source and drain. The amount of carriers in this area, and the current, is a strong function of the gate to source voltage.

Unlike BJTs there is little difference between source and drain. Some FETs can be flipped and still work well. Most FETs have the bulk semiconductor connected to the source, and the bulk is the opposite type from the channel. A NFET is usually built on P type material. As a result most FETs have a diode in parallel with the source and drain which will be forward biased if the device is flipped source to drain.

The current capacity is proportional to the width of the channel, which can be a meter or more on a modern power device.

As the schematic symbol implies the gate is a capacitor and no DC current will flow. During switching or other voltage changes on the gate there are charge and discharge currents flowing into and out of the gate.

The beauty of these devices is you can make logic gates with only leakage currents when the gates are static. This behavior is what makes possible the billion gate ICs that are in production now. For bipolar transistor logic there is always base currents flowing which results in large currents on a large IC even when static.

DC impedance on the FET gate is very high, leakage currents a nanoamp or less. These can also be used in the arrangements of BJTs to result in differing input and output impedance.

IGBT

The Insulated Gate Bipolar Junction Transistor is actually a FET and a BJT together, combining the positive features of both for high power devices. The FET input provides a high impedance input and the BJT can provide a very low voltage collector emitter for low power dissipation when switched on.

I put the terminal names in quotes because the “Emitter” is clearly the PNP collector. It is labelled emitter for the complete structure because it acts similar to the emitter on an BJT.

Here is the schematic symbol.

The way it works is the FET supplies all the base current for the PNP transistor. This device provides the high input impedance and a power gain stage with the low “on” voltage of a BJT in a single device.

IGBTs are primarily used as power switching devices. These circuits uses their low loss characteristics to provide high efficiency, low drive power systems.

IGBTs have moderate frequency response, they are not used for RF circuits. Switching supplies and motor controllers run at the tens to hundreds of KHz.