4.3 Servo Mechanisms

  

A servomechanism may be broadly defined as a closed-loop control system in which a small power input controls a much larger power output in a strictly proportionate manner. In applying such a mechanism to the automatic control of an aircraft, the system must be capable of continuous operation and have the ability to (i) detect the difference between an input and an output (error detection); (ii) amplify the error signals; and (iii) control the closing of the servo loop by providing the feedback.

There are two main classes of servomechanism: (i) position control and (ii) speed control; both classes may be independently applied to automatic flight control systems depending respectively on whether they are of the displacement type or the rate sensing type. In some control systems they may also be used in conjunction with each other.

4.3.1 Open-Loop Control System

An open-loop control system is controlled directly, and only, by an input signal. The basic units of this type consist only of an amplifier and a motor. The amplifier receives a low- level input signal and amplifies it enough to drive the motor to perform the desired job.

The open-loop control system is shown in basic block diagram form in figure 4.3.1. With this system, the input is a signal that is fed to the amplifier. The output of the amplifier is proportional to the amplitude of the input signal. The phase (ac system) and polarity (dc system) of the input signal determines the direction that the motor shaft will turn. After amplification, the input signal is fed to the motor, which moves the output shaft (load) in the direction that corresponds with the input signal. The motor will not stop driving the output shaft until the input signal is reduced to zero or removed. This system usually requires an operator who controls speed and direction of movement of the output by varying the input. The operator could be controlling the input by either a mechanical or an electrical linkage.

There is no means of precisely controlling these factors and, therefore, the open loop system is not good enough for close tolerance control.

In practice, the accuracy of the open loop system is unsatisfactory. Many factors contribute to this e.g.

a) Variations of load conditions.

b) Frictional forces within the motor and its load, and the mechanical interconnection (gear

trains, clutches, linkages, drives etc).

c) Variation of power supply.

d) Value of the demand voltage.

e) Variations of amplifier gain.

4.3.2 Closed Loop Control System

A closed-loop control system is one in which control action is dependent on the output. If a radar operator, for example, could see the scanner, he could cancel the movement when it has moved the correct amount. This is now a closed loop, the operator completing the loop between output and input. He compares the desired effect and adjusts the control panel to reduce the error between them. He is what we call an error detector. The amount of error which he observes determine how he adjusts the input to produce the desired results. Modern systems incorporate automatic error detector to replace human interaction.

The essential features of a closed loop system are as follows:

a) The feedback of information concerning the behaviour of the load.

b) The comparison of this information with the behaviour demanded by the input

c) The production of an error signal proportional to the difference between the desired

behaviour and the actual behaviour.

d) The amplification of the error signal to control the power into a servomotor.

e) The movement of the load by the servomotor in such a direction as to reduce the error

signal to zero.

f) The error detector fulfills the requirements of sub para (b) and (c).

g) The position feedback must be in anti-phase with the demand signal.

4.3.3 Response of servomechanism

The response to servomechanism is the pattern of behaviour of the load when a change is made to the input condition, the most important factors being the form which the input change takes and the various restraints, friction, etc. which act on the output. There are two types of input change to be considered and these are referred to as step input and ramp input, the names being derived from the shape of the curves of input against time as shown in figure 4.3.3.

Figure 4.3.3: Response of servomechanisms (a) Step input (b) Ramp input

A step input is one whereby the input is suddenly changed to a new angular position θi from a null position. Because of the inertia of the load an angular change at the servomechanism output will not be able to follow exactly that at the input, with the result that a large error signal is produced initially. This causes the load to be accelerated to its required position, and thereby reduces the error to zero.

A ramp input is one whereby the input is suddenly moved at a constant speed. In the early stages of the input, and while the error signal is small, the load accelerates slowly and lags behind the input.

4.3.4 Damping

Oscillatory or transient responses of a servomechanism from whatever cause are obviously undesirable, and so it is necessary to provide some form of damping by which a load can be brought to rest in its required position with the minimum of overshoot.

Varying degrees of damping can be applied (fig. 4.3.4). Using only inherent friction light damping is achieved. If there is too much extra viscous friction, the system is heavily damped and a very sluggish response is produced. The degree of damping which just prevents any overshoot is known as critical damping. Damping which allows one small overshoot and gives the smallest settling time is known as optimum damping.

Figure 4.3.4: Degrees of damping: (a) step input (b) ramp input

Servomechanisms possess various inherent factors which, together, have the general effect of reducing the amplitude of each successive oscillation; such factors include static friction, kinetic friction, eddy current, lubricant viscosity, etc., and while contributing to damping requirements they do have certain detrimental effects, e.g. power is wasted, and errors can be introduced with the servomechanism operating in the steady state. The effects are partly due to a small force of constant magnitude known as coulomb friction, and to viscous friction which increases with speed.

4.3.5 Synchro System

With the introduction of large multi-engine aircraft the problem arose of how to measure various quantities such as pressure, temperature, engine speed and fuel tank contents at points located at greater distances from the cockpit. Many of the instruments then available operated on mechanical principles which could be adapted to suitably transmit the required information. For example, on one very early twin-engine aircraft, mechanically-operated engine speed indicators were designed with large-diameter dials so that by mounting the indicators in the engine nacelles they could be read from the cockpit.

Synchronous systems fall into two classes: direct-current and alternating-current. The principles of some of those commonly used form the subject of this chapter. Although varying in the method of data transmission, all the systems have one common feature: they consist of a transmitter located at the source of measurement and a receiver which is used to position the indicating element.

4.3.5.1 Basic Desynn System

The electrical element of the transmitter consists of a resistor wound on a circular former (called the 'toroidal resistor') and tapped at three points 120° apart. Two diametrically-opposed wiper contact arms, one positive and the other negative, are insulated from each other by a slotted arm which engages with a pin actuated by the appropriate mechanical element of the transmitter.

The wiper contact arms are assembled in the form of a bar having rotational freedom about a pivot which carries current to the positive arm. Current to the negative arm is carried via a wiper boss the underside of which is in contact with a ring fitted on the inner side of the terminal moulding. A circlip, fitted at the end of the pivot, holds the complete assembly in place against a spring which gives the required contact pressure on the toroidal resistor.

The receiver element consists of a cylindrical two-pole permanent magnet rotor pivoted to rotate within the field of a laminated soft iron stator, carrying a star-connected three-phase distributed winding supplied from the toroidal resistor tappings. A tubular brass housing is fitted inside the stator, and together with its end cover, provides a jewelled bearing support for the rotor spindle. The front end of the spindle projects through the end cover and a dial mounting plate, to carry the pointer. Electrical connection between the transmitting and receiving elements may be either by terminal screws or plug-type connectors.

The electrical elements of the receivers are common to all three circuit arrangements of the Desynn system.

Operation

When direct current is applied to the transmitter contact arms (which are in contact with the toroidal resistor) currents flow in the resistor causing the three tapping points to be at different potentials. For example, with the contact arms in the position shown in Figure 4.3.9 the potential at tapping No 2 is greater than that at No 1 because there is less resistance in the circuit between the positive arm and the No 2 tapping. Thus, currents are caused to flow in the lines between transmitter and receiver, the magnitude and direction of which depend upon the position of the contact arms on the toroidal resistor.

In turn, these currents flow through the coils of the receiver stator and produce a magnetic field about each coil similar to that of a bar magnet; thus either end of a coil may be designated as a Npole or a S-pole, depending on the direction of the current through a coil. The combined fields

extend across the stator gap and cause the permanent-magnet rotor to align itself with their resultant.

A pull-off magnet is fitted to the end plate, its purpose being to act as a power-failure device by exerting an attractive force on the main magnet rotor so as to pull it and the pointer to an off-scale position when current to the stator is interrupted. The strength of this magnet is such that, when the system is in operation, it does not distort the main controlling field.

Figure 4.3.9: Circuit diagram of basic Desynn system

4.3.5.2 The Micro-Desynn System

In applications where the movement of a prime mover is small and linear, the use of a basicsystem transmitting element is strictly limited. The micro-Desynn transmitter was therefore developed to permit the magnification of such small movements and to produce, by linear movement of contacts, the same electrical results as the complete rotation of the contact arms of the basic transmitter.

In order to understand the development of this transmitting element, let us imagine that a toroidal resistor has been cut in two, laid out flat with its ends joined together, and three tappings made as before together with positive and negative arms in contact with it. Movement of the contact arms will produce varying potentials at the tapping, but as will be clear from Figure 4.3.10(a), the full range will not be covered because one or other of the arms would run off the resistor. We thus need a second resistor with tappings so arranged that the contact arms can move through equal distances.

If we now take two toroidal resistors and join them in parallel then, by cutting them both in two and laying them out flat, we obtain the circuit arrangement shown at (b). By linking the contact arms together and insulating them from each other, they can now be moved over the whole length of each resistor to produce voltage and current combinations which will rotate the receiver through 3600. Since the contact arms have to traverse a much shorter path, their angular movement can be kept small (usually 450), a feature which helps to reduce the energy required to operate the transmitter.

The resistors are wound on bobbins which may be of round or square section, the latter type being designed to help reduce cyclic and friction errors.

Each resistance bobbin is secured in place against a set of miniature spring contact fingers accurately positioned so as to provide the necessary tapping points.

The contact arms are mounted on a rocker shaft supported between the vertical parts of a Ushaped bracket, and movement of the transmitter's mechanical element is transmitted to the arms via a spring loaded operating pin and crank arm connected to the rocker shaft. Two berylliumcopper hair-springs conduct current to the contact arms and also act together to return the rocker shaft and contact arms to their starting position.

4.3.5.3 The Slab-Desynn System

In addition to the cyclic error present in the basic and micro-type systems, small errors also arise due to friction set up by the contact arms having to move over a considerable surface of resistance wire. Although such errors can be reduced by providing a good contact material and by burnishing the resistance wire surface, the cyclic error is still undesirable in certain measurements.

Figure 4.3.11: Slab-Desynn transmitter

A solution to this problem was brought about by modifying the basic system so as to change its three sawtooth waveforms into sinusoidal waves, the instantaneous sum of which is always zero. The transmitter so developed is shown schematically in Figure 4.3.11, from which it will be noted that the resistor and contact arms have, as far as electrical connections are concerned, virtually changed places with each other. The resistor is now wound on a slab former, hence the term `slab-Desynn', and is connected to the direct-current supply, while the contact arms themselves provide the three potential tapping points for the indicator stator.

The three contact arms are insulated from each other and pivoted over the centre of the slab, and are each connected to a slip ring. Spring-finger brushes bear against these slip rings and convey the output currents to the stator coils. Movement of the mechanical element is transmitted to the brushes via a gearing system.

4.3.5.4 Torque Synchros

These are the simplest form of synchro and are used for the transmission of angular position information by means of induced signals, and for the reproduction of this information by the position of a shaft at an output or receiver element. A typical application of torque synchros is in flight instrument systems.

A torque synchro system comprises two electrically similar units interconnected as shown in Figure 4.3.12, and by convention one is designated the transmitter (TX) and the other the receiver (TR).

Each unit consists of a rotor carrying a winding, and concentrically mounted in a stator carrying three windings the axes of which are 120° apart. The principal physical differences between the TX and the TR are that the rotor of the TX is mechanically coupled to an input shaft, while the TR rotor is free to rotate. The rotor windings are connected to a source of single-phase a.c. supply, and the corresponding stator connections are joined together by transmission lines. The similarity between these connection arrangements and a conventional transformer may also be noted; the rotors correspond to primary windings and the stators to secondary windings.

When the rotors are aligned with their respective stators in the position indicated they are said to be at `electrical zero'; this refers to the reference angle standardized for synchros at which a given set of stator voltages will be produced, and by this convention enables replacement synchros to be matched to each other.

With power applied to the rotors, due to transformer action a certain voltage will be induced in the stator coils the value of which will be governed, as in any transformer, by the ratio of the number of turns of the rotor (primary) and stator (secondary) coils.

When the rotors of TX and TR occupy the same angular positions and power is applied, equal and opposite voltages will be produced and hence no current can flow in the stator coils. The system (and any other type of synchro) is then said to be at `null'.

When the rotors occupy different angular positions, for example when the TX rotor is at the 30° position and the TR rotor is at electrical zero, an unbalance occurs between stator coil voltages causing current to flow in the lines and stator coils. The currents are greatest in the circuits where voltage unbalance is greatest and their effect is to produce magnetic fields which exert torques to turn the TR rotor to the same position as that of the TX.

As the TR rotor continues to turn, the misalignment, voltage unbalance and currents decrease until the 30° position is reached and no further torque is exerted on the rotor.

In considering this synchronizing action one might assume that, since currents are also flowing in the stator coils of TX, its rotor would be returned to `null'. This is a reasonable assumption, because in fact a torque is set up tending to turn the rotor in a clockwise direction. However, it must be remembered that the rotor is being actuated by some prime mover which exerts loads too great to be overcome by the rotor torques.

4.3.5.5 Control Synchros

Control synchros differ from torque synchros, in that their function is to produce an error voltage signal in the receiving element, as opposed to the production of a rotor torque. Typical uses of control synchros are in servoed altimeters and airspeed indicators which operate in conjunction with central air data computers.

The interconnection of the two elements of a control synchro system is shown in Figure 4.3.13. By convention, the transmitter is designated as CX, and the receiver designated as a control transformer (CT). The CX is similar to a torque transmitter, and from the diagram it will be noted that the a.c. supply is connected to the CX rotor only. The CT rotor is not energized since it acts merely as an inductive winding for detecting the phase and magnitude of error signal voltages which are supplied to an amplifier. The amplified signals are then fed to a two-phase motor which

is mechanically coupled to the CT rotor. Another difference to be noted is that a control synchro system is at electrical zero when the rotor of CT is at 90° with respect to the CX rotor.

Figure 4.3.13: Control synchrc system

If the rotor of CX is rotated through a certain angle, the resultant flux in the CT stator will be displaced from its datum point by the same angle, and relative to the CT rotor position at that instant. An error voltage is therefore induced in the rotor, the phase and magnitude of the voltage depending on the direction of CX rotor rotation, and on the degree of misalignment between it and the CT rotor. The error voltage is then amplified and fed to the control phase of the motor, the other phase (reference phase) being continuously supplied with alternating current. Since the control phase voltage of a two-phase motor can either lead or lag the reference phase voltage, then the phase of the error voltage will determine the direction in which the motor will rotate, and its magnitude will determine its speed of rotation. As the motor rotates, it turns the rotor of the CT in the appropriate direction, thereby reducing its displacement relative to the CX rotor. Rotation continues until both rotors are in alignment (bearing in mind, of course, that the electrical zero points are at 90° from each other) at which position no further error voltage is induced.

4.3.5.6 Differential Synchros

In some cases, it is necessary to detect and transmit error signals representative of two angular positions, and in such a manner that the receiver element of a synchro system will indicate the difference or the sum of the two angles. This is achieved by introducing a third synchro into either a torque or control system, and using it as a differential transmitter. Unlike TX or CX synchros, a differential transmitter (designated TDX or CDX) has an identically wound stator and rotor which in the application to a torque synchro system are interconnected as shown in Figure 4.3.14.

At (a) the TX rotor is shown rotated clockwise through 600 while the rotor of TDX remains at electrical zero; all the magnetic fields rotate, and the rotor of TR takes up the same angular position as the rotor of TX. If now the TX rotor remains at electrical zero, and the TDX rotor is rotated clockwise through 150 say, the fields of both synchros remain in the electrical zero position because their position is determined by the orientation of the TX rotor (diagram (b)). However, a 150 clockwise rotation of the TDX rotor without a change in the position of its field is equivalent to moving the rotor field 150 anticlockwise whilst leaving the rotor at electrical zero. This relative angular change is duplicated in the stator of TR and so its rotor will align itself with the field i.e. for a 150 clockwise rotation of the TDX rotor, the TR rotor will rotate 150 anticlockwise.

Assume now that the TX rotor is rotated through 600 clockwise, and the TDX rotor through 150 clockwise, then because the TR rotor will rotate 150 anticlockwise, its final angular movement will be equal to the difference between the two input angles i.e. it will turn through 450 (diagram (c)). The differential effect is of course reversed when the TDX rotor is rotated in the opposite direction to the TX rotor, so that the TR rotor rotates through an angle equal to the sum of the two input angles. By reversing pairs of leads either between TX and TDX, or between TDX and TR, any one of the rotors can be made to assume a position equal to the sum or difference of the angular positions of the other rotors.

In the same way that differential transmitter synchros can be used in torque synchro systems, so they can be used in systems utilizing control synchros to transmit control signal information on the sum or difference of two angles. The basic arrangement is shown in Figure 4.3.13

4.3.5.7 Resolver Synchros

The function of resolver synchros (RS) is to convert alternating voltages, which represent the cartesian coordinates of a point, into a shaft position and a voltage, which together represent the polar coordinates of that point. They may also be used in the reverse manner for voltage conversion from polar to cartesian coordinates. Typical applications of resolver synchros are to be found in flight director and integrated instrument systems.

A typical arrangement of an RS for conversion from polar to cartesian coordinates is shown in Figure 4.3.15 and from this it will be noted that the stator and rotor each have two windings arranged in phase quadrature, thus providing an eight-terminal synchro. An alternating voltage is applied to the rotor winding RI-R2, and the magnitude of this voltage, together with the angle through which the rotor is turned, represent the polar coordinates. In this application, the second winding is unused, and as is usual in such cases, it is short circuited to improve the accuracy of the RS and to limit spurious response.

In the position shown, the alternating flux produced by the current through rotor winding R1-R2 links with both stator windings, but since the rotor winding is aligned only with S1-S2 then maximum voltage will be induced in this winding. Winding S3-S4 is in phase quadrature so no voltage is induced in it. When the rotor is at a constant speed it will induce voltages in both stator windings, the voltages varying sinusoidally. The voltage across that stator winding which is aligned with the rotor at electrical zero will be a maximum at that position and will fall to zero after rotor displacement of 900; this voltage is therefore a measure of the cosine of the displacement. The voltage is in phase with the voltage applied to R1-R2 during the first 900 of displacement, and in anti-phase from 900 to 2700, finally rising from zero at 2700 to maximum in

phase at 3600. Any angular displacement can therefore be identified by the amplitude and phase of the induced stator voltages. At electrical zero, stator winding S3-S4 will have zero voltage induced in it, but at 900 displacement of rotor winding R1-R2, maximum in-phase voltage will be induced and will vary sinusoidally throughout 3600; thus, the S3-S4 voltage is directly proportional to the sine of the rotor displacement. The phase depends on the angle of displacement, any angle being identified by the amplitude and phase of the voltages induced in stator winding S3-S4. The sum of the outputs from both stators, i.e. rcosθ plus rsinθ, therefore defines in cartesian coordinates the input voltage and rotor rotation.

Figure 3.16 illustrates an arrangement whereby Cartesian coordinates may be converted to polar coordinates. An alternating voltage Vx = rcosθ is applied to the cosine stator winding S1-S2, while a voltage Vy = rsinθ is applied to the sine stator winding S3-S4. An alternating flux representing cartesian coordinates is therefore produced inside the complete stator. One of the rotor windings, in this case R1-R2, is connected to an amplifier, and in the position shown it will have maximum voltage induced in it, which will be applied to the amplifier. The output from the amplifier is applied to a servomotor which is mechanically coupled to a load and to the rotor. When the rotor is turned through 900 the induced voltage in winding R1-R2 reduces to zero and the servomotor will stop. The rotor winding R3-R4 will now be aligned with the stator flux, and a voltage will be

induced in it which is proportional to the amplitude of the alternating flux as represented by the vector r i.e. a voltage proportional to √( Vx 2 +Vy 2). This voltage together with the angular position of the rotor therefore represents an output in terms of the polar coordinates.

Figure 4.3.16: Conversion of cartesian coordinates to polar coordinates

4.3.5.8 E and I Transformer

When the armature is located in the center of the E-shaped core, as shown in the figure 4.3.17, equal and opposite voltages are induced in the secondary coils. The difference between them is zero. Thus, the voltage at the output terminals is also zero. But, if the armature is moved, say to the tight, the voltage induced in coil 1 increase, while the voltage induced in coil 3 decreases. The two voltages are then unequal, so that the difference is no longer zero. A net voltage results at the output terminals. The amplitude of this voltage is directly proportional to the distance the armature has been moved from its center position. The phase of this output voltage, relative to the ac on the primary, controls the direction the load moves in correcting the error. The basic Etransformer will detect movement of the armature in one axis only (either the horizontal or vertical depending upon the way the unit is mounted). To detect movement in both the horizontal and vertical axes, a CROSSED-E-TRANSFORMER is used. If you place two E-transformers at right angles to each other and replace the bar armature with a dome-shaped one (figure 4-208), you have the basic configuration of what is known as the crossed-E transformer, or pickoff. In most applications the dome-shaped armature is attached to a gyro, and the core assembly is fixed to a gimbal, which is the servo load.

4.3.5.9 Inductance Transmission

Inductance transmissions consist of a metal rod inside an electrical coil (Figure 4.3.19). The position of the rod in the coil depends on the degree of movement of measuring element. As the metal rod moves, it changes the current passing through the coil. The transmitting coil and rod can be duplicated in the receiving instrument and combined with the receiving coil into an inductance bridge circuit (Figure 4.3.20)

When the measuring element moves the metal rod in the transmitting coil the bridge balance is upset. The unbalance is then corrected by a movement of the metal rod in the receiver coil. Usually there is an amplifier and a motor in such a system. The unbalance is fed into the amplifier which provides a strong enough electrical impulse to cause the motor to run. The motor then moves the metal rod in the receiving coil to a position which restores bridge balance.

The inductance bridge (Figure 4.3.21) is generally supplied for use with alternating current power supply

4.3.5.10 Capacitance Transmitter

Capacitance Transmitter utilizes the capacitance formed between two probes to calculate the level of the liquid/medium inside the vessel. The capacitance between the two probes increases when a dielectric is introduced between them.

Thus, if the liquid is a dielectric then corresponding capacitance (between the probes) will be proportional with its height.

In aircraft this types of sensors are mostly used in the fuel tanks to measure the fuel heights. When the tank is empty the capacitance will be the lowest.