Field effect transistors are constructed similarly to transistors and diodes, but they have specific characteristics that pose unique challenges to electricians. David Herres explains.

In the past, I wrote about transistors. Field effect transistors (FETs) are a further refinement, operating on the same basic concepts – doped semiconductor regions that acquire different characteristics depending upon the bias provided by an external source of electrical energy.

Recall that in a traditional diode or transistor, the p-n or n-p junction conducts or does not conduct when the electrons and holes (absence of electrons) are repelled or attracted by the polarity of the bias, so that they migrate toward or away from the junction. They are charge carriers and they must be present at the junction for current to flow through the output circuit.

A similar activity takes place within the FET, but the structure and electrical characteristics are different. Specifically, the large output responds to small changes in the voltage applied at the input. A very low level electrical charge is present here and the current flow is minute. What this means is that the input impedance is very high. Unlike the traditional transistor, the input circuit sees a very slight load so that it is not appreciably altered by its connection to the FET.

This has important implications and is highly advantageous in many applications. For example, the front end (input) circuit of a high-impedance voltmeter can use an FET so that the circuit under test will be electrically isolated and unaffected by the process of taking the measurement. To see how this is possible, we’ll take a look at the internal structure of an FET and see how it works on a sub-atomic level.

The FET principle parts have picturesque and highly descriptive names. The source corresponds to the transistor’s base. The gate and drain are similar to the base and collector, respectively, in the traditional transistor.

The source and drain are connected by a channel and it is here that the action takes place in response to the bias supplied at the input. The channel in the output circuit behaves as a resistance, which varies in response to the bias applied at the input. The current, in accordance with Kirchhoff’s Current Law, is the same at source and drain terminals and anywhere along the channel.

The shape and resulting conductivity of the channel vary in response to the electric charge placed on the gate and source terminals. The traditional transistor is a current device, whereas the FET is a voltage device. The idea is that the voltage at the gate terminal modulates the actual physical dimensions of the channel, thereby determining its conductivity (the reciprocal of its resistance). The output current responds, conforming to Ohm’s law.

The gate is insulated from the channel (including source and drain) by means of a very thin non-conductive layer, and that is why there is almost no current flow at the input, so that the device is essentially invisible to the circuit that is connected to the input. However, the gate to channel functions as a capacitor, placing an upper limit on the frequency it can handle. At higher frequencies the capacitive reactance drops off so that the FET is no longer a high-impedance device and eventually the link is shorted out, shutting down the preceding state and possibly destroying it.

FETs come in two varieties, junction field effect transistors (JFETs) and metal oxide semiconductor field effect transistors (MOSFETs), or insulated gate field effect transistors (IGFETs). Either of these can exist in p-channel or n-channel types, making for a total of four varieties.

The MOSFET operates in a similar manner to the JFET, but it is constructed differently and its electrical characteristics differ significantly. The gate is insulated from the channel by an insulating layer, originally metal oxide, which is deposited during the manufacturing process. The oxide layer could be contaminated by sodium ions, which are inevitable in all surroundings. These ions cause changes in the electrical characteristics of the device, which are unacceptable. The solution has been found to cover the insulating medium with a protective layer of silicon nitride, which blocks the harmful sodium ion infiltration.

MOSFET is a not altogether accurate name for this device, because metal oxide is no longer in general use as the insulating medium. IGFET is more accurate, although among technicians the term MOSFET persists.

MOSFETs have become the most widely used of the FETs and indeed of the entire transistor family. They are used in digital circuits in computers where they perform high speed switching operations, and in analogue applications as well, such as in high-end audio MOSFET amplifiers. They are capable of handling relatively high amounts of power and are seen in the variable frequency drives that permit speed control of three-phase induction motors.

If the defect in a piece of malfunctioning equipment has been isolated to a circuit board that has one or more MOSFETs, and a preliminary visual examination does not reveal a burnt component, and if ohmmeter measurements do not indicate an open resistor, capacitor or inductor, the time has come to look at the status of the semiconductors.

Enormous numbers of MOSFETs are embedded within integrated circuits (ICs), which will be discussed in a subsequent article. Of course it is not possible to open up an IC in order to test, much less replace, internal MOSFETs. But often these devices exist as discrete components. To perform fully definitive tests requires expensive test gear, but fairly accurate determinations may be made with your multimeter.

Most modern multimeters have a diode test function. This applies between 3V and 4V to the device under test. You need a second meter to measure this voltage. In the rare instances where it is lower, the test cannot be made. In order to perform the test, the MOSFET must be removed from the circuit. These devices are damaged by heat and static electricity. When desoldering or soldering back in place, use the smallest iron that will do the job and do not apply heat longer than necessary. A heatsink, which looks like an alligator clip with smooth jaws, is useful for intercepting the heat that would otherwise travel from the solder joint along the lead to the semiconductor device. Alternately, needle nose pliers would work.

As for static charge, this may be carried by an ungrounded tool or the human body, if you touch the terminals. A grounding bracelet, made for the purpose, keeps your body at ground potential. If you don’t have a grounding bracelet, touch a well-grounded surface at frequent intervals to keep a charge from developing.

To test the MOSFET, connect the meter’s negative probe to the source. Then touch the meter’s positive probe to the gate. This charges the MOSFETs gate capacitance and the charge will remain for quite some time. Shift the positive meter probe to the drain. If you get a low reading, so far so good.

Maintaining the positive probe’s connection to the drain, touch your finger simultaneously to the source and gate. If you touch all three terminals, that works fine. This will discharge the gate and the meter will swing high, meaning that the output circuit is not conducting. The MOSFET is good.

This preliminary, non-dynamic test is not as all-inclusive as a complete operating profile provided by laboratory-grade equipment, but should suffice most of the time.