THE FIELD EFFECT
TRANSISTOR
A Necessary Device for the Modern IC
by Dan Shanefield
he commonly used bipolar transistor — in which elec-
_
DRAIN
Metal
Ox.
N
P
N
(I) P
N
Ox.
Silicon
_
D
Ð 5K
Pot.
_
+
SOURCE GATE
Metal Metal
O
+
G
+
S
SYMBOL
Figure 1. Simplified cross section of a JFET, with an operating
circuit. It is N-channel, depletion mode, and normally on.
The symbol is at the right-hand side of the figure.
G
D
S
A
B
C
D
Figure 2. An N-channel JFET wired to be a constant
current self-regulating device, with the symbol shown next to it
on the left. The other two symbols, to the right, are for constant
current sources that include power supplies such as batteries.
Ttrons or holes pass through two PN semiconductor
junctions — is essentially a current amplifying device.
Although voltage can be amplified indirectly if the “
common emitter” or “common collector” wiring configurations
are used, it still is true that a small amount of input current
must always flow into the transistor’s base region for
control purposes. (The reasons for these facts are
explained in the book referred to at the end of this article.)
Another type of semiconductor device, the field effect
transistor or “FET,” is not as familiar to many electronics
hobby enthusiasts, possibly because it is easily damaged by
misuse. The FET amplifies voltage directly, and the current
needed for control is so low that it cannot be measured with
common instruments. This transistor was actually the first
type of semiconductor amplifier predicted theoretically at Bell
Labs, back in the 1950s, but it was not developed into a practical device until after the bipolar type had become popular.
However, FETs have now become the most common type,
with tens of millions of them in each microprocessor IC chip.
With such a huge number of transistors operating in a
single chip, we certainly don’t want much current to be
required for the control of each one — the battery power
would be used up fast, and a lot of heat would have to be
removed. Also, there are many other applications where
super-low input current is desirable. An obvious example is
in the first stage of an accurate voltmeter, where we don’t
wish to cause any new voltage drops by draining current
out of the circuit being studied.
Still another advantage of the FET, probably less important, is the fact that its input versus output characteristics
are similar to those of vacuum tubes. Because tubes have
been used since about 1910, we have a great deal of experience with them, and some designers feel more comfortable with FETs than with bipolar devices, especially in audio
amplifiers. (Whether or not this is truly an advantage borders on emotional factors as much as scientific ones. Some
readers might recognize the author of the present article as
an early partisan on one side of this intensely debated issue,
so we won’t discuss it any further here!) At any rate, the FET
responds entirely to voltage at the controlling electrode, and
this can be used to throttle fairly large amounts of output
JULY 2005
71
