Field-effect transistor

Powerful N-type field effect transistor

A field (unipolar) transistor is a semiconductor device whose principle of operation is based on controlling the electrical resistance of a conductive channel by a transverse electric field created by the voltage applied to the gate.

The region from which the charge carriers go into the channel is called the source , the region into which they enter from the channel is called the drain , the electrode to which the control voltage is applied is called the gate .

History of field effect transistors

Field effect transistor circuit

In 1953, Dyke and Ross proposed and implemented the design of a field effect transistor - with a control pn junction .

For the first time, the idea of regulating the flow of main carriers by an electric field in an insulated gate transistor was proposed by Lilienfeld in 1926-1928. However, difficulties in putting this idea into practice made it possible to create the first working device only in 1960. In 1966, Carver Mead improved this design by shunting the electrodes of such a device with a Schottky diode .

In 1977, James McCallagham of Bell Labs found that using field effect transistors can significantly increase the performance of existing computing systems.

Classification of field effect transistors

Types of field-effect transistors and their designation on circuit diagrams

Field-effect transistors are classified into devices with a pn-junction control and with an insulated gate, the so-called MIS ("metal-dielectric-semiconductor") - transistors, which are also called MOS ("metal-oxide-semiconductor") - transistors, the latter being divided into integrated channel transistors and induced channel devices.

The main parameters of field-effect transistors are: input resistance, internal resistance of the transistor, also called output resistance, slope of the gate-gate characteristic, cut-off voltage, and some others.

Pn junction transistors

Fig. 1. Design of a field effect transistor with a control pn junction and an n-type channel
a) with a shutter on the side of the substrate;
b) with diffusion gate.

The drain-gate characteristic (left) and the family of drain characteristics (right) of a field-effect transistor with a gate in the form of a pn junction and an n-type channel.

V_{GS}

{\ displaystyle V_ {GS}}

- gate-drain voltage;

V_{DS}

{\ displaystyle V_ {DS}}

- drain-source voltage;

I_{DS}

{\ displaystyle I_ {DS}}

- drain or source current;

V_{P}

{\ displaystyle V_ {P}}

- shutter voltage lock, or cut-off voltage.

A field-effect transistor with a control pn junction ^[1] (JFET) is a field-effect transistor in which a semiconductor plate, for example, p-type (Fig. 1), has electrodes (source and drain) at opposite ends, with which it is turned on into a controlled circuit. The control circuit is connected to the third electrode (gate) and is formed by a region with a different type of conductivity, in the example in the figure - n-type.

A constant bias source included in the input circuit generates a reverse (blocking) voltage on a single pn junction. An amplified signal source is also included in the input circuit. When the input voltage changes, the reverse voltage at the pn junction changes, and therefore the thickness of the depleted layer changes , i.e., the cross-sectional area of the region in the crystal through which the flow of the main charge carriers passes changes. This area is called the channel.

Field effect transistor electrodes are called:

source ( English source ) - an electrode from which the main charge carriers enter the channel;
drain ( English drain ) - an electrode through which the main charge carriers leave the channel;
shutter ( English gate ) - an electrode that serves to control the cross section of the channel.

The channel semiconductor type can be either n- or p-type. By the type of channel conductivity, field-effect transistors with an n-channel and a p-channel are distinguished. The polarities of the bias voltages supplied to the electrodes of transistors with n and p channels are opposite.

The current and voltage at the load, connected in series to the channel of the field-effect transistor and the power source, are controlled by changing the input voltage , as a result of which the reverse voltage at the pn junction changes, which leads to a change in the thickness of the blocking (depleted) layer. With some blocking voltage $V_{P}$ ${\ displaystyle V_ {P}}$ the cross-sectional area of the channel becomes equal to zero and the current through the channel of the transistor becomes very small.

Since the reverse current of the pn junction is very small, in static mode or at low operating frequencies, the power taken from the signal source is negligible. At high frequencies, the current taken from the signal source can be significant and goes to recharge the input capacitance of the transistor.

Thus, the field-effect transistor according to the principle of current control is similar to an electric vacuum lamp - a triode , but in view of the drain-source current-voltage characteristics it is close to an electric vacuum pentode . With this analogy, the source in the field effect transistor is similar to the cathode of a vacuum triode, the gate to the grid, and the drain to the anode. There are also differences, for example:

there is no cathode in the transistor that requires heating;
any of the source and drain functions can be performed by any of these electrodes;
there are field effect transistors with both an n-channel and a p-channel, which is used in the production of complementary pairs of transistors.

A field-effect transistor differs from a bipolar transistor, firstly, in the principle of operation: in a bipolar transistor, the output signal is controlled by the input current, and in the field-effect transistor by the input voltage or electric field. Secondly, field-effect transistors have significantly larger input resistances , which is associated with the reverse bias of the gate pn junction in the considered type of field-effect transistors. Third, field-effect transistors have a low noise level (especially at low frequencies) compared to bipolar transistors, since field-effect transistors do not inject minor charge carriers and the channel of the field-effect transistor can be performed inside a semiconductor crystal. Carrier recombination processes in the pn junction and at the base of a bipolar transistor , as well as generation and recombination processes on the surface of a semiconductor crystal, generate low-frequency noise.

Insulated Gate Transistors (MOS Transistors)

Fig. 2. The device of the field effect transistor with an isolated gate.
a) - with an induced channel, b) - with an integrated channel

An insulated gate field effect transistor (MOSFET) is a field effect transistor whose gate is electrically isolated from the channel by a dielectric layer.

In a crystal of a semiconductor with a relatively high resistivity, which is called a substrate, two heavily doped regions with the opposite conductivity type relative to the substrate are created. Metal electrodes are applied to these areas - source and drain. The distance between heavily doped areas of the source and sink can be less than a micron. The surface of the semiconductor crystal between the source and drain is covered with a thin layer (of the order of 0.1 μm) of the dielectric . Since silicon is usually the initial semiconductor for field-effect transistors, a silicon dioxide layer of SiO ₂ grown on the surface of a silicon crystal by high-temperature oxidation is used as an insulator. A metal electrode is applied to the dielectric layer - a gate. It turns out a structure consisting of metal, dielectric and semiconductor. Therefore, insulated gate field effect transistors are often referred to as MOS transistors.

The input resistance of the MOS transistors can reach 10 ¹⁰ ... 10 ¹⁴ Ohms (for field-effect transistors with a control pn junction 10 ⁷ ... 10 ⁹ ), which is an advantage in the construction of high-precision devices.

There are two types of MOS transistors: with an induced channel and with an integrated channel.

In MOS transistors with an induced channel (Fig. 2a), the conducting channel between the heavily doped source and drain regions is absent and, therefore, a noticeable drain current appears only at a certain polarity and at a certain gate voltage relative to the source, which is called the threshold voltage ( U _Zipor ).

In MOS transistors with an integrated channel (Fig. 2b), an inverse layer exists at the gate surface at zero voltage at the gate relative to the source, with a gate channel connecting the source to the drain.

Depicted in fig. 2 structures of insulated gate field effect transistors have a substrate with n-type electrical conductivity. Therefore, heavily doped regions under the source and sink, as well as the induced and built-in channel, have p-type electrical conductivity. If similar transistors are created on a substrate with p-type conductivity, then the channel will have n-type conductivity.

Channel Induced MOSFETs

When the gate voltage is zero relative to the source, and when the voltage is applied to the drain, the drain current is negligible. It represents the reverse current of the pn junction between the substrate and the heavily doped drain region. At a negative potential at the gate (for the structure shown in Fig. _2a ), as a result of the penetration of the electric field through the dielectric layer into the semiconductor at low gate voltages (smaller than U _Cp ), a field effect depleted by the main carriers occurs under the gate and region of space charge, consisting of ionized uncompensated impurity atoms. At gate voltages greater than U _Zpor , an inverse layer arises at the surface of the semiconductor under the gate, which is a p-type channel connecting the source to the drain. The thickness and cross section of the channel will change with the voltage on the gate, respectively, the drain current will also change, that is, the current in the load circuit and a relatively powerful power source. This is how the drain current is controlled in an insulated gate field-effect transistor and with an induced channel.

Due to the fact that the gate is separated from the substrate by a dielectric layer, the current in the gate circuit is negligible, the power consumed from the signal source in the gate circuit is small, and is necessary to control a relatively large drain current. Thus, an MIS transistor with an induced channel can amplify electromagnetic oscillations in voltage and power.

The principle of power amplification in MOS transistors can be considered from the point of view of charge carriers transferring the energy of a constant electric field (energy of a power source in the output circuit) to an alternating electric field. In the MIS transistor, before the channel appeared, almost the entire voltage of the power source in the drain circuit fell on the semiconductor between the source and the drain, creating a relatively large constant component of the electric field strength. Under the action of a gate voltage, a channel appears in the semiconductor under the gate, through which charge carriers, holes, move from the source to the drain. Holes moving in the direction of the constant component of the electric field are accelerated by this field and their energy increases due to the energy of the power source in the drain circuit. Simultaneously with the emergence of the channel and the appearance of mobile charge carriers in it, the voltage at the drain decreases, that is, the instantaneous value of the variable component of the electric field in the channel is directed opposite to the constant component. Therefore, the holes are inhibited by an alternating electric field, giving him part of their energy.

MOSFETs with integrated channel

Fig. 3. Output static characteristics (a) and drain-gate characteristic (b) of an MOS transistor with an integrated channel.

In this scheme, an MIS transistor with an isolated gate and an induced channel is used as a nonlinear element.

Due to the presence of an integrated channel in such an MIS transistor (Fig. 2b), when the voltage is applied to the drain, the drain current is significant even at zero gate voltage (Fig. 3b). The cross section and channel conductivity will change when the gate voltage of both negative and positive polarity changes. Thus, an MOS transistor with an integrated channel can operate in two modes: in the enrichment mode and in the depletion mode of the channel by charge carriers. This feature of MOS transistors with an integrated channel is also reflected in the displacement of the output static characteristics when the gate voltage and its polarity change (Fig. 3).

The static transmission characteristics (Fig. 3b) go out from the point on the _x- axis corresponding to the cut-off voltage U _UI , that is, the voltage between the gate and the source of a MOS transistor with a built-in channel operating in the depletion mode, at which the drain current reaches a predetermined low value .

Calculation formulas $I_{c}$ ${\ displaystyle I_ {c}}$ depending on voltage U _ZI

1. The transistor is closed $U_{3u}<U_{nop}$ ${\ displaystyle U_ {3u} <U_ {nop}}$

I_{c}=0

{\ displaystyle I_ {c} = 0}

The threshold value of the voltage of the MOS transistor $U_{nop}=1.5B$ ${\ displaystyle U_ {nop} = 1.5B}$

2. Parabolic section. $U_{3u}>U_{nop}$ ${\ displaystyle U_ {3u}> U_ {nop}}$

I_{c}=K_{n}[(U_{3u}-U_{nop})U_{cu}-{\frac {U_{cu}^{2}}{2}}]

{\ displaystyle I_ {c} = K_ {n} [(U_ {3u} -U_ {nop}) U_ {cu} - {\ frac {U_ {cu} ^ {2}} {2}}]}

$K_{n}$ ${\ displaystyle K_ {n}}$ - specific steepness of the transfer characteristic of the transistor.

3. Further increase $U_{3u}$ ${\ displaystyle U_ {3u}}$ leads to a transition to a gentle level.

I_{c}={\frac {K_{n}}{2}}[U_{3u}-U_{nop}]^{2}

{\ displaystyle I_ {c} = {\ frac {K_ {n}} {2}} [U_ {3u} -U_ {nop}] ^ {2}}

- Hovstein equation.

TIR structures for special purposes

In structures of the metal-nitride-oxide-semiconductor (MNOS) type, the dielectric under the gate is two-layer: a layer of SiO ₂ oxide and a thick layer of Si ₃ N ₄ nitride. Between the layers, electron traps are formed, which, when a positive voltage MNOS structure (28–30 V) is applied to the gate, capture electrons tunneling through a thin layer of SiO ₂ . The resulting negatively charged ions increase the threshold voltage, and their charge can be stored for up to several years in the absence of power, since the SiO ₂ layer prevents charge leakage. When a large negative voltage (28 ... 30 V) is applied to the gate, the accumulated charge dissolves, which significantly reduces the threshold voltage.

Metal-oxide-semiconductor (MOS) structures with a floating gate and avalanche injection ( LIZMOS ) have a gate made of polycrystalline silicon, isolated from other parts of the structure. An avalanche breakdown of the pn junction of the substrate and the drain or source, to which a high voltage is applied, allows the electrons to penetrate the gate through the oxide layer, as a result of which a negative charge appears on it. The insulating properties of a dielectric allow this charge to be preserved for decades. The electric charge is removed from the shutter by ionizing ultraviolet irradiation with quartz lamps, while the photocurrent allows the electrons to recombine with holes.

Subsequently, structures of storage dual-gate storage field effect transistors were developed. A gate built into the dielectric is used to store a charge that determines the state of the device, and an external (normal) gate controlled by bipolar pulses to introduce or remove charge on the built-in (internal) gate. So there were cells, and then flash memory chips, which have gained great popularity these days and made a noticeable competition to hard drives in computers.

To implement ultra-large integrated circuits (VLSI), ultra-miniature field microtransistors have been created. They are made using nanotechnology with a geometric resolution of less than 100 nm. For such devices, the thickness of the gate dielectric reaches several atomic layers. Various, including three-gate structures are used. Devices work in micropower mode. In modern Intel microprocessors, the number of devices ranges from tens of millions to 2 billion. The latest field microtransistors are made on stressed silicon, have a metal gate and use a new patented material for a gate insulator based on hafnium compounds ^[2] .

In the last quarter century, powerful field-effect transistors, mainly MOS-type, have developed rapidly. They consist of many low-power structures or structures with a branched shutter configuration. Such RF and microwave devices were first created in the USSR by the specialists of the Pulsar research institute V. V. Bachurin (silicon devices) and V. Ya. Vaksemburg (gallium arsenide devices) The study of their impulse properties was performed by a scientific school of prof. Dyakonova V.P. (Smolensk branch of MPEI). Это открыло область разработки мощных ключевых (импульсных) полевых транзисторов со специальными структурами, имеющих высокие рабочие напряжения и токи (раздельно до 500—1000 В и 50-100 А). Такие приборы нередко управляются малыми (до 5 В) напряжениями, имеют малое сопротивление в открытом состоянии (до 0,01 Ом) у сильноточных приборов, высокую крутизну и малые (в единицы-десятки нс) времена переключения. У них отсутствует явление накопления носителей в структуре и явление насыщения, присущее биполярным транзисторам. Благодаря этому мощные полевые транзисторы успешно вытесняют мощные биполярные транзисторы в области силовой электроники малой и средней мощности ^[3] ^[4] .

За рубежом в последние десятилетия стремительно развивается технология транзисторов на высокоподвижных электронах (ТВПЭ) , которые широко используются в СВЧ устройствах связи и радионаблюдения. На основе ТВПЭ создаются как гибридные, так и монолитные микроволновые интегральные схемы . В основе действия ТВПЭ лежит управление каналом с помощью двумерного электронного газа , область которого создаётся под контактом затвора благодаря применению гетероперехода и очень тонкого диэлектрического слоя — спейсера ^[5] .

Схемы включения полевых транзисторов

Полевой транзистор в каскаде усиления сигнала можно включать по одной из трех основных схем: с общим истоком (ОИ), общим стоком (ОС) и общим затвором (ОЗ).

Схема включения полевого транзистора с управляющим pn-переходом с общим истоком

Схема включения полевого транзистора с управляющим pn-переходом с общим стоком

Схема включения полевого транзистора с управляющим pn-переходом с общим затвором

На практике в усилительных каскадах чаще всего применяется схема с ОИ, аналогичная схеме на биполярном транзисторе с общим эмиттером (ОЭ). Каскад с общим истоком даёт большое усиление по мощности. Но, с другой стороны, этот каскад наиболее низкочастотный из-за вредного влияния эффекта Миллера и существенной входной ёмкости затвор-исток ( С _зи ).

Схема с ОЗ аналогична схеме с общей базой (ОБ). В этой схеме ток стока равен току истока, поэтому она не даёт усиления по току, и усиление по мощности в ней во много раз меньше, чем в схеме ОИ. Каскад ОЗ обладает низким входным сопротивлением, в связи с чем он имеет специфическое практическое применение в усилительной технике. Преимущество такого включения — практически полное подавление эффекта Миллера, что позволяет увеличить максимальную частоту усиления и такие каскады часто применяются при усилении СВЧ .

Каскад с ОС аналогичен каскаду с общим коллектором (ОК) для биполярного транзистора — эмиттерным повторителем . Такой каскад часто называют истоковым повторителем . Коэффициент усиления по напряжению в этой схеме всегда немного меньше 1, а коэффициент усиления по мощности занимает промежуточное значение между ОЗ и ОИ. Преимущество этого каскада — очень низкая входная паразитная ёмкость и его часто используют в качестве буферного разделительного каскада между высокоомным источником сигнала, например, пьезодатчиком и последующими каскадами усиления. По широкополосным свойствам этот каскад также занимает промежуточное положение между ОЗ и ОИ.

Области применения полевых транзисторов

КМОП-структуры , строящиеся из комплементарной пары полевых транзисторов с каналами разного (p- и n-) типа, широко используются в цифровых и аналоговых интегральных схемах .

За счёт того, что полевые транзисторы управляются полем (величиной напряжения приложенного к затвору), а не током, протекающим через базу (как в биполярных транзисторах), полевые транзисторы потребляют значительно меньше энергии, что особенно актуально в схемах ждущих и следящих устройств, а также в схемах малого потребления и энергосбережения (реализация спящих режимов).

Выдающиеся примеры устройств, построенных на полевых транзисторах, — наручные электронные часы и пульт дистанционного управления для телевизора. За счёт применения КМОП-структур эти устройства могут работать до нескольких лет от одного миниатюрного источника питания — батарейки или аккумулятора, потому что практически не потребляют энергии.

В настоящее время полевые транзисторы находят всё более широкое применение в различных радиоустройствах, где с успехом заменяют биполярные. Их применение в радиопередающих устройствах позволяет увеличить частоту несущего сигнала, обеспечивая такие устройства высокой помехоустойчивостью. Обладая низким сопротивлением в открытом состоянии, находят применение в оконечных каскадах усилителей мощности звуковых частот высокой мощности ( Hi-Fi ), где с успехом заменяют биполярные транзисторы и электронные лампы. Биполярные транзисторы с изолированным затвором ( IGBT ) — приборы, сочетающие биполярные и полевые транзисторы, — находят применение в устройствах большой мощности, например в устройствах плавного пуска , где успешно вытесняют тиристоры .

Notes

↑ И. П. Жеребцов . Основы электроники. Ed. 5-е. — Л., 1989. — С. 114.
↑ Дьяконов, 2004 .
↑ Бачурин, Ваксембург, Дьяконов и др., 1994 .
↑ Дьяконов, Максимчук, Ремнев, Смердов, 2002 .
↑ Li, 2006 .

Literature

Дьяконов В. П. Intel. Новейшие информационные технологии. Достижения и люди. — М. : СОЛОН-Пресс, 2004. — 416 с. — ISBN 5980031499 .
Бачурин В. В., Ваксембург В. Я., Дьяконов В. П. и др. Схемотехника устройств на мощных полевых транзисторах: Справочник / Дьяконов В. П.. — М. : Радио и связь, 1994. — 280 с.
Дьяконов В. П., Максимчук А. А., Ремнев А. М., Смердов В. Ю. Энциклопедия устройств на полевых транзисторах / Дьяконов В. П.. — М. : СОЛОН-Р, 2002. — 512 с.
Li, Sheng S. Semiconductor Physical Electronics. — Second Edition. — Springer, 2006. — 708 с. — ISBN 978-0-387-28893-2 .

[1] И. П. Жеребцов . Основы электроники. Ed. 5-е. — Л., 1989. — С. 114.

[_d4c21d45addf72e5-2] Дьяконов, 2004 .

[_165c304d2120f65c-3] Бачурин, Ваксембург, Дьяконов и др., 1994 .

[_f97954bc79e67a21-4] Дьяконов, Максимчук, Ремнев, Смердов, 2002 .

[_224de1957d51145a-5] Li, 2006 .