Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 1

CHAPTER V- Micro Wave Devices are a form of electromagnetic radiation with  ranging from 1m to 1mm (or)  between 300 MHz (0.3 GHz) and 300 GHz. This includes both UHF and EHF (millimeter waves). Applications  Microwaves are extensively used for point-to-point telecommunications. Microwaves are especially suitable for this use since they are more easily focused into narrower beams than radio waves.  Microwaves are used in long distance communications like TV and telephone communications.  They are extensively used in very long distances communications such as spacecraft communication and between ground stations and communications satellites.  Their higher frequencies allow broad bandwidth , high data transmission rates, and smaller antenna sizes (since antenna size is inversely proportional to frequency).  Microwaves are also employed in ovens and in radar technology.

Solid State Microwave Devices As the operating frequency increases, the stray reactance, device capacitance and transit time as well as cable losses are increasing. For example at microwave frequencies, a thin film resistor looks like a complex RLC network with distributed LC values and a different R value. Ordinary electronic components behave differently at microwave frequencies and hence special devices were developed to use in microwave region. Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 2

MW Solid state devices are classified into different categories as MW Bipolar Transistors - MW Heterojunction Transistor (HBT) MW FET such as High Mobility Transistors (HEMT) Microwave Special MW Diodes The MW special diodes are classified into different groups based on their operating principle  Quantum Mechanical Tunnel effect - Tunnel  Transfer Electron Device(TED) – Characterised by bulk effect of -  Avalanche and Transit time effects – IMPATT and TRAPAT  Parametric Excitation effect – Parametric

Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 3

Tunnel Diode (Or ESAKI diode) Invented by Leo Esaki in 1958. A is associated with quantum mechanical tunnelling phenomena. Quantum tunnelling is the quantum-mechanical effect of transitioning through a classically-forbidden energy state. (ie) an electron or charge carrier pass through a potential barrier without having sufficient energy to do so.

A tunnel diode is a PN junction that exhibits between two values of forward voltage. Both P and N sides are degenerately doped (very heavily doped) with impurities several 1000 times that of a typical PN junction diode (1019 to 1020 impurity atoms /cm3). Because of heavy doping the depletion region is very thin and the tunnel distance ‘d’ is very small (5 to 10 nm). Also the heavy doping causes, the Fermilevel within the valence band on P side and within the Conduction band on N side. The amount of degeneracy qVp and qVn is typically 50mev to 200 mev. When there is no voltage (V=0) is applied to the tunnel diode, the diode is in the thermal equilibrium. Under this unbiased condition, the probability of electrons going from states in the conduction band on N-side to states in the valance band on the P-side and in opposite directions are the same. Therefore the net tunneling of the thin barrier is then zero.

When a forward bias is VF applied, the energy levels on N side are raised relative to those on P-side. Now, there exists a band of occupied energy states on the N-side (electrons in the conduction band of N-side) and corresponding unoccupied states on the P-side (empty states in the valence band of P-side). The electrons can tunnel from the N-side to the P- side.

Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 4

FIG: Tunneling at various Biasing levels

TTunneling stops

Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 5

This tunneling current will reach a maximum value IP (peak current) at a forward voltage bias voltage VF = VP (Peak Voltage) which is approximately equal to (Vp +Vn) / 3).

When the forward bias is further increased (VP < VF < VV valley voltage), the energy levels on N side are further raised and there are fewer available unoccupied states on the P-side and tunneling decreased.

Further increase in forward voltage (V = VV) makes the band uncrossed, and at this point tunneling current no longer flow or the current is minimal (IV valley current).

With still further increase in forward voltage (VF > VV) the normal thermal current will flow. From the operation of tunnel diode, in the forward bias condition the tunnel current increases from zero to a maximum value IP as the forward voltage increases. With the further increase in voltage the current decreases to IV. The decreasing portion after the peak current is negative differential resistance region. The tunnel diode at various bias conditions are shown in Figures 1 through 5 to the corresponding points in the characteristic curve.

The values of IP and IV determine the magnitude of the negative resistance and the ratio IP/IV is known the figure of merit of the tunnel diode.

Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 6

Tunnel diode Characteristics

Symbol and Equivalent circuit of Tunnel diode

Ls- Lead inductance (nH) , Rj - Negative Resistance (100 ohm) , Rs – Bulk Resistance (few ohm) , Cj – Junction capacitance (pf) The empirical form of Tunnel diode current is given by

The first term of the above equation is the Tunnel current and the second term is the normal thermal current. The negative differential Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 7 resistance can be obtained from the first term of the equation and is given by

The values of VP and VV depend upon the diode materials. The following figure shows typical V-I characteristics of some important materials at room temperature.

The current ratios of IP/IV are 8:1 for Ge and 12.1 for GaSb and GaAs. Because of its smaller effective mass (0.042M0), the smaller bandgap (0.72eV), the GaSb tunnel diode has the largest negative resistance of the three devices. Applications 1. The negative resistance gives negative power (= -I2R) and is the source of power. (ie) instead of absorbing the signal it boost the signal. This property found applications in Microwave amplifiers and oscillators. 2. Since the tunneling phenomenon is very fast, tunnel diodes are used in high speed switching circuits (The time of switching of a tunnel diode is only few nano seconds)

Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 8

Advantages  Insensitive to temperature - hence tunnel diodes are used in place of normal diodes in military applications .  Very low junction Capacitance  Extreme speed in switching and stable characteristics  Ability to operate under wide variety of critical environments  Low noise level and small size Drawback Tunnel diodes cannot replace rectifier diodes, because tunnel diode is very leaky in reverse bias. Tunnel diode based oscillator Tank circuits oscillate but “die out” due to the internal resistance (Positive Resistance). If a tunnel diode is properly biased to operate in the negative resistance region the “negative resistance” provided by the tunnel diode can overcomes the loses due to the positive resistance and maintains the oscillations

Tunnel diode oscillator Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 9

Transferred Electron Devices (TED) - Gunn Diode Ridely and Watkins (1961) theoretically discussed the possibility of negative resistance effect in . In 1962 Hilsum discussed the possibility of obtaining a voltage controlled bulk negative conductance in GaAs called as Transferred Electron Mechanism. In 1963 Gunn observed micro wave oscillations in GaAs sample. Gunn Effect Gunn observed that, above some critical voltage (corresponding to electric field of 2K-4K V/cm) the current passing through n-type GaAs material becomes a periodic fluctuating function of time (oscillation) . He also observed that, the frequency of oscillation is determined mainly by the material not by the external circuit. The period of oscillation is inversely proportional to the specimen length and is equal to the transit time of electrons between the electrodes. The Gunn Effect was successfully explained by the Two-Valley Model theory. Figure

Ososcillating current after V>Vth I

Vth V

Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 10

Gunn Diode Gunn Diode is also known as Transfer Electron Device. Even though it is called as a diode it does not contain PN junction. It is a Bulk device, and because of its two electrodes it is called as a diode. Thus the operation of a Gunn diode is based upon the bulk properties of the material and not on the properties of PN junction.

Construction and Working

Symbol and Equivalent Circuit

Gunn diodes are fabricated from a single piece of n-type semiconductor. The most common materials are Gallium Arsenide, (GaAs) and Indium Phosphide (InP). Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 11

The device is simply an n-type bar with n+ contacts. It is necessary to use n-type material because the transferred electron effect is only applicable to electrons and not to holes.

The most common method of manufacturing a Gunn diode is to grow and epitaxial layer on a degenerate n+ substrate. The active region is very thin and its thickness is between a few microns and a few hundred microns. This active layer has a low doping level between 1014 /cm3 and 1016 /cm3 - this is considerably less than that used for the top and bottom areas of the device. The thickness of the active region determines the frequency of oscillation. The base also acts as a heat sink which is critical for the removal of heat.

Operation The Gunn diode operation depends upon the very thin active region. When a voltage is placed across the device, most of the voltage appears across the inner active region. Since the active region is very thin the voltage gradient that exists in this region is very high. After a threshold voltage level, the device exhibits a negative resistance region on its V/I curve.

Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 12

This negative resistance region means that the current flow in diode increases in this region when the voltage falls. This phase reversal enables the Gunn diode to act as an and oscillator.

When the voltage across the active region reaches a certain point a current is initiated and travels across the active region. During the time when the current pulse is moving across the active region the potential gradient falls preventing the formation further pulses. Only when the pulse has reached the far side of the active region will the potential gradient rise, allowing the next pulse to be created. It can be seen that the time taken for the current pulse to traverse the active region largely determines the rate at which current pulses are generated, and hence it determines the frequency of operation. The pulse formation can be explained by the Two Valley Model Theory (Gunn Diode Operation in detail. Gunn diode operation at microwave frequencies At microwave frequencies, it is found that the dynamic action of the diode incorporates elements resulting from the thickness of the active region. When the voltage across the active region reaches a certain point a current is initiated and travels across the active region. During the time when the current pulse is moving across the active region the potential gradient falls preventing any further pulses from forming. Only when the pulse has reached the far side of the active region will the potential gradient rise, allowing the next pulse to be created.

It can be seen that the time taken for the current pulse to traverse the active region largely determines the rate at which current pulses are generated, and hence it determines the frequency of operation. Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 13

To see how this occurs, it is necessary to look at the electron concentration across the active region. Under normal conditions the concentration of free electrons would be the same regardless of the distance across the active diode region. However a small perturbation may occur resulting from noise from the current flow, or even external noise - this form of noise will always be present and acts as the seed for the oscillation. This grows as it passes across the active region of the Gunn diode.

Gunn diode operation The increase in free electrons in one area cause the free electrons in another area to decrease forming a form of wave. It also results in a higher field for the electrons in this region. This higher field slows down these electrons relative to the remainder. As a result the region of excess electrons will grow because the electrons in the trailing path arrive with a higher velocity. Similarly the area depleted of electrons will also grow because the electrons slightly ahead of the area with excess electrons can move faster. In this way, more electrons enter the region of excess making it larger, and more electrons leave the depleted region because they too can move faster. In this way the perturbation increases. Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 14

Gunn diode operation - electrons in the peak move more slowly The peak will traverse across the diode under the action of the potential across the diode, and growing as it traverses the diode as a result of the negative resistance. A clue to the reason for this unusual action can be seen if the voltage and current curves are plotted for a normal diode and a Gunn diode. For a normal diode the current increases with voltage, although the relationship is not linear. On the other hand the current for a Gunn diode starts to increase, and once a certain voltage has been reached, it starts to fall before rising again. The region where it falls is known as a negative resistance region, and this is the reason why it oscillates.) Two Valley Model Theory According to the energy band theory of n-type GaAs, other than valence band and conduction band there is a third band called sub-conduction band. The two levels of conduction band give two valleys called lower valley (LV) and upper valley (UV). The two valleys are separated by 0.36eV. Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 15

The effective mass of electron is given by the expression M = h2 / (d2E/dK2) ------1 The basic mechanism involved in the operation of n-type GaAs device is the transfer of electrons from lower valley conduction band to upper valley conduction band. Since the lower valley slope is sharper than the upper valley, the electron has different effective mass, and hence different mobility (µ) in lower and upper valley.

From 1, it is clear that , ML < MU

µL > µU (Since µ = e/M) GaAs Valley Effective Mass Mobility

2 Lower ML = 0.068M0 µL = 8000 cm /V-sec

2 Upper MU= 1.2M0 µU= 180 cm /V-sec

M0 is the rest mass Electron densities in the lower and upper valleys remain the same under an equilibrium condition.  When the applied electric field is lower than the electric field of the

lower valley (E < EL), no electrons will transfer to the upper valley. Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 16

 When the applied field is higher than that of the lower valley and

lower than that of the upper valley (EL < E < EU), electrons will begin to transfer to the upper valley.

 When the applied field is higher than that of upper valley (E > EU), all electrons will transfer to the upper valley.

If electron densities in the lower and upper valleys are nL and nU, the conductivity of the n-type GaAs is

σ = e (µLnL + µUnU) ------2 Where e – the electron charge Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 17

P n = nL + nU is the electron density and is proportional to E (where p is a constant) According to ohms law J = σE ------3 dJ/dE = σ + (dσ/dE) E = σ ( 1+ E/σ (dσ/dE) ) (1/σ) dJ/dE = 1 + [ (dσ/dE) / (σ/E ) ]

The condition for negative resistance is dJ/dE < 0 (i.e) 1 + [ (dσ/dE) / (σ/E ) ] < 0 Or (dσ/dE) / (σ/E ) < -1 Or -(dσ/dE) / (σ/E ) > 1 This condition can be achieved in semiconductors which have the following properties. 1. The presence of a sub-band in the conduction band. 2. The energy difference between two valleys must be several times larger than the thermal energy (KT~0.0259eV) 3. The energy difference between the valleys must be smaller than the bandgap energy (Eg) 4. Electron in lower valley must have a higher mobility and smaller effective mass than that of in upper valley Applications Gunn diodes are mainly used in the generation of Microwave

frequencies between 1 GHz and 100 GHz. As a discrete component, a Gunn diode can be used as an oscillator or amplifier in applications that require low-power radio frequency (RF) signals, such as proximity sensors and wireless local area networks (LAN). Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 18

Gunn diodes that are made from gallium arsenide can operate at frequencies up to 200 GHz. A Gunn diode made from gallium nitride can reach 3 THz.

Avalanche Transit Time Devices (ATTD) ATTD is proposed by Read in 1958. In an ATTD negative resistance is achieved by creating a delay (180o phase shift) between the voltage and current. The delay is achieved by, 1. Avalanche delay caused by the finite build up time of avalanche current - (produce 90o phase shift between V & I) 2. Transit time delay – the finite time taken by the carriers to reach the respective electrode (produce another 90o phase shift between V & I) When these two delays add up to 180o, the diode electronic resistance is negative.

Two modes of ATTD 1. IMPATT Diode – Impact Ionization Avalanche Transit Time Diode Also called as Read Diode ( 5 -10% efficiency) 2. TRAPATT Diode – Trapped Plasma Avalanche Transit Time Diode (20-60% efficiency)

IMPATT Diode Impact Avalanche current Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 19

The avalanche is generated by carrier impact ionization. If a free electron with sufficient energy strikes an atom in the semiconductor like Si, it can break the covalent bond and liberate an electron from the covalent bond. The liberated electron gains energy by being in an electric field and liberates other electrons from other covalent bonds. This process can cascade very quickly into a chain reaction producing a large number of electrons and a large current flow. This phenomenon is called impact avalanche.

Structure The IMPATT-Diode is a two terminal which operates by a combination of avalanche multiplication and transit time effects. Generally, it has a PN structure which is reverse biased to operate in the avalanche breakdown region. Two important regions of IMPATT diode are

1) Thin p region (High field/Avalanche region) – avalanche multiplication occurs

2) Intrinsic n region (Drift region) – generated electrons must drift towards the n+ contact.

A reverse biased IMPATT diode with various regions is shown in figure Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 20

Operation The doping profile and the corresponding electric fields in a reverse biased IMPATT diode is shown in the following figure. Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 21

 When the diode is operated in reverse bias near breakdown both the N and N- (intrinsic) regions are completely depleted

 When the reverse bias voltage is above the breakdown voltage, the space charge region always extends from ‘p+ n’ junction to the ‘i n+’ junction through the n and the i regions.  Because of the difference in doping between the "drift region" and "avalanche region", the electric field is highly peaked in the avalanche region and nearly flat in drift region.  Carriers (electrons) in the high field region near the p+ -n junction acquire energy to knock down the valence electrons in the covalent bond and hence electron hole pairs are generated. This is avalanche multiplication  In operation, avalanche breakdown occurs at the point of highest electric field, and this generates a large number of hole-electron pairs by impact ionization. Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 22

 The holes are swept into the cathode (- terminal), but the electrons travel across the drift region toward anode (+ terminal). The transit time of an electron across the drift intrinsic region L is given by

 = L / Vd And the avalanche multiplication factor is given by

n M = 1 / ( 1- V/Vb) V – Applied reverse voltage,

Vb – Avalanche Breakdown voltage, n – numerical factor depending on the doping concentration. For silicon normally 3 to 6.

Carrier Current and External Current  An ac voltage is maintained at a given frequency between the diode terminal in such a way that the total field across the diode is the sum of ac and dc fields. The sum of both fields causes avalanche breakdown at the P+ -n junction during the positive half cycle of the ac voltage cycle if the field is above the breakdown voltage.

 The carrier current Io(t) generated at the p+ -n junction by the avalanche multiplication grows exponentially with time while the field is above critical voltage.  During the negative half cycle, when the field is below breakdown voltage, the carrier current decays exponentially. Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 23

 I0(t) is in the form a pulse of very short duration and it reaches its maximum in the middle of the ac voltage cycle or one quarter of the cycle later than the voltage.  Under the influence of electric field the generated electrons are injected into the space region towards the negative terminal.

As the injected electron traverse the drift space,

1) They induce a current Ie(t) in the external circuit. 2) Cause a reduction of the field Since the velocity of the electron in the space charge is constant

Ie (t) = Q/ = VdQ/L Q – total charge of the moving electrons

The external current Ie(t) because of the moving electrons is delayed by

o 90 relative to the pulsed I0(t). Since the carrier current I0(t) is delayed by one quarter cycle or 90o relative to the ac voltage, Ie(t) is then delayed by 180 relative to the voltage. Hence negative conductance occurs and the diode can be used for microwave oscillation and amplification. Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 24

The following figure shows a close-up of the current and voltage waveforms after oscillations have stabilized. It is clear from the figure that the current is 180° out of phase with the voltage.

The following figure shows the buildup of microwave oscillations in the diode current and voltage when the diode is embedded in a resonant cavity and biased at breakdown

Fig 4. Buildup of microwave osc

Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 25

Advantages and drawbacks

 A main advantage is their high power capability. These diodes are used in a variety of applications from low power radar systems to alarms.  The diodes make excellent microwave generators for many applications. The output is reliable and relatively high when compared to other forms of diode The main drawback of using an IMPATT-diode is the high level of phase noise that the device generates. In view of its high levels of phase noise it is used in transmitters more frequently than as a local oscillator in receivers (since in receivers the phase noise performance is generally more important).

Applications IMPATT diodes are semiconductor devices that generate relatively high-power microwave signals at frequencies between about 3 GHz and 100 GHz or more. IMPATT diodes are used in low-power radar systems and alarms. The main drawback of using an IMPATT diode is the high level of phase noise that the device generates.

PARAMETRIC AMPLIFIERS The parametric amplifier is named for the time-varying parameter, or value of capacitance, associated with the operation. Since the underlying principle of operation is based on reactance, the parametric amplifier is sometimes called a REACTANCE AMPLIFIER. Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 26

The conventional amplifier is essentially a variable resistance that uses energy from a dc source to increase ac energy. The parametric amplifier uses a nonlinear variable reactance to supply energy from an ac source to a load. Since reactance does not add thermal noise to a circuit, parametric amplifiers produce much less noise than most conventional amplifiers. The most important feature of the parametric amplifier is the low- noise characteristics. Electronic noise is the primary limitation on receiver sensitivity and is the name given to very small randomly fluctuating voltages that are always present in electronic circuits. The sensitivity limit of the receiver is reached when the incoming signal falls below the level of the noise generated by the receiver circuits. At this point the incoming signal is hidden by the noise, and further amplification has no effect because the noise is amplified at the same rate as the signal. The effects of noise can be reduced by careful circuit design and control of operating conditions, but it cannot be entirely eliminated. Therefore, circuits such as the parametric amplifier are important developments in the fields of communication and radar. The basic theory of parametric amplification centers around a capacitance that varies with time. Consider the simple series circuit shown in figure. When the switch is closed the capacitor charges to value Q. If the switch is opened, the isolated capacitor has a voltage across the plates determined by the charge Q divided by the capacitance C. Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 27

Voltage amplification from a varying capacitor. An increase in the charge Q or a decrease in the capacitance C causes an increase in the voltage across the plates. Thus, a voltage increase, or amplification, can be obtained by mechanically or electronically varying the amount of capacitance in the circuit. In practice a voltage-variable capacitance, such as a varactor, is used. The energy required to vary the capacitance is obtained from an electrical source called a PUMP.

Varactor diode, more commonly referred to as a varicap, is an electronic semiconductor device very closely related to a standard diode but with certain capabilities similar to a capacitor. The difference between a varactor and a standard diode is that a standard diode is designed to minimize the device’s capacitance while a varactor diode is designed to use and exploit capacitance. Varactor diodes find common use in parametric electronics, such as parametric amplifiers and other tuning circuits that can be varied by a change in voltage. Although ordinary PN junction diodes exhibit the variable capacitance effect, varactor diodes are optimised to give the required changes in capacitance.

Varactor diode basics Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 28

The varactor diode or varicap diode consists of a standard PN junction, and it is optimized to function as a variable capacitor. The diode is operated under reverse bias conditions and this gives rise to three regions. At either end of the diode are the P and N regions where current can be conducted. However around the junction is the depletion region where no current carriers are available. As a result, current can be carried in the P and N regions, but the depletion region is an insulator. This is exactly the same construction as a capacitor. It has conductive plates separated by an insulating dielectric. The capacitance of a capacitor is dependent on a number of factors including the plate area, the dielectric constant of the insulator between the plates and the distance between the two plates. In the case of the varactor diode, it is possible to increase and decrease the width of the depletion region by changing the level of the reverse bias. This has the effect of changing the distance between the plates of the capacitor. Varactor diodes are always operated under reverse bias conditions, and in this way there is no conduction. They are effectively voltage controlled capacitors, and indeed they are sometimes called varicap diodes, although the term varactor is more widely used these days. Varactor diode applications Varactor diodes are widely used within the RF design arena. They provide a method of varying the capacitance within a circuit by the application of a control voltage. Although varactor diodes can be used within many types of circuit, they find applications within two main areas: Voltage controlled oscillators, VCOs: Voltage controlled oscillators are used for a variety of applications. One major area is for the oscillator Class Notes by. K.Elampari, Associate Professor of Physics, S.T.Hindu college, Nagercoil 29

within a phase locked loop - this are used in almost all radio, cellular and wireless receivers. A varactor diode is a key component within a VCO. RF filters: Using varactor diodes it is possible to tune filters. Tracking filters may be needed in receiver front end circuits where they enable the filters to track the incoming received signal frequency. Again this can be controlled using a control voltage. Typically this might be provided under microprocessor control via a digital to analogue converter.