Antenna

This invention relates to an antenna; particularly to a receiver antenna of the type which includes means for producing a rotation of the antenna pattern and uses the received signal, modulated by the rotation, to derive a control signal to track the received signal source. Antennas of this kind are known as "conical scanning" antennas, because the beam pattern rotates around the surface of a cone (the apex being at the antenna).

Conical scanning antennas were first applied in radar tracking of targets (the antennas acting both to transmit and to receive but generally scanning the transmitted beam). More recently, conical scanning antennas have been employed as ground station antennas for satellite telecommunication links tracking non geo stationary satellites.

A particular problem occurs when an antenna is mounted on a vessel at sea, since a vessel is subject to endless rolling, pitching and yawing motion due to the normal swells and tides and to the wakes of other passing vessels. It is not unusual for a small boat to roll through 50 to 60 degrees; the period of the roll is variable, but is on the order of ten seconds or so. The problem is of course exacerbated for smaller pleasure craft (which generally try to avoid extreme conditions).

For a water vessel (or other vehicle) to receive satellite communications it is therefore necessary that the receiver antenna be controlled to point at the satellite. Most seaborne satellite antennas are either gimballed or are mounted on drive motors which are responsive to sensors sensing the motion of the ship. An example of such an antenna is shown in EP0154240. Such arrangements are however mechanically complex and expensive. It is also known to mount an antenna on a gyro stabilized platform, but this limits the antenna size and weight since the capacity of such platforms are restricted.

Another problem with such arrangements is that the antenna is maintaining its orientation relative to the vessel or vehicle. However, when the vessel moves to a different geographical location, the relative inclination required to point at the satellite changes and consequently the antenna is mis-aligned.

These problems make such antennas unattractive for application as vehicle-borne receiver antennas for satellite television, where a simple, robust and inexpensive antenna is essential.

In one aspect, the invention therefore provides a water vessel comprising a satellite television receiver antenna which employs conical scanning. Such an antenna may receive Direct Broadcast by Satellite (DBS) signals, or other television formats (for example the transmission format used by the Astra Satellite).

Another problem encountered in providing a tracking antenna for satellite television reception is that the received signal may be strongly periodically amplitude modulated; for example, a triangular envelope, typically harmonically related to the line or frame period, may be imposed on the FM carrier. This modulation interferes with the signal derived from conical scanning. In a further aspect of the invention, there is therefore provided a conical scanning antenna which employs a scan frequency harmonically related to the above signal modulation frequency; this enables this signal modulation to be taken account of.

Preferably the scan frequency is a sub-harmonic of the modulation frequency, and the arrangement is such that a signal is derived as the difference of received signal samples separated in time by an integer number of modulation periods so that the effect of the modulation signal on the satellite tracking is cancelled.

Similarly, where power for the antenna is derived from an AC power supply such as the 50 hertz or 60 hertz mains, a mains ripple may be super imposed at various points in the scanning system. In another aspect of the invention, therefore, the scanning frequency is arranged to be harmonically related to the power supply frequency, and preferably to be a sub-harmonic of it, as above. Where the movement of the ship can be expected to be lively, and the antenna response must therefore be particularly rapid, the scan frequency must also be increased. If a scan frequency that is a harmonic of any signal modulation is used, the latter may be easily reduced or eliminated by filtering at a later stage. Any mains ripple that may be present will be seen as a slight offset of the antenna if the scan frequency and the mains frequency are the same. In any event, where, ( as in the above aspects of the invention), the scanning frequency is to be harmonically related to an external modulation frequency, close control over the scanning frequency is also essential. The triangular wave form envelope observed in satellite television signals is usually at the frame rate of the television signal (25 or 30 hertz) and the AC mains power supply is generally 50 or 60 hertz; the scan frequency of the antenna will therefore be a sub multiple, or a multiple, of the external modulation frequency.

Where the conical scanning is effected by mechanical rotation of an element of the antenna it is difficult to maintain close position and rotational speed accuracy, especially at low speeds, especially where the element is small and consequently has a low angular momentum and mechanical inertia. Known techniques, for example employing a phase locked loop, are often unstable under these conditions.

According to a further aspect of the invention, there is therefore provided a rotary speed control system comprising a rotary position signal generator generating a periodically varying signal including a constant average level related to the rotary speed, and a reference signal generator generating a corresponding signal, further comprising comparator means arranged to generate an output corresponding to the difference between said signals, whereby when the rotary speed approximates a desired speed said comparator means generates a pulse width modulated output arranged to stabilize said rotary speed, and when said rotary speed diverges substantially from desired speed said comparator output varies to adjust said rotary speed.

Preferably such a system is employed to control the scan speed of a rotated sub-reflector.

In general, a conical scan can be produced either electrically (by varying electrical parameters of the antenna) or mechanically (by rotating a component of the antenna). It is known to provide an antenna with an off-set feed and rotate the feed around the central axis of the antenna. One type of antenna employed for satellite television reception is the Cassegrain reflector antenna, which comprises a parabolic dish focusing the received signal onto a secondary reflector, or sub-reflector, having a hyperbolic profile, which refocuses the signal onto a feed, located in the centre of the parabolic reflector. GB1136174 shows a Cassegrain antenna for producing a transmitted scanned elliptical beam in which the sub-reflector is mounted axially aligned with the parabolic antenna axis, but is eccentrically mounted with respect to that axis. However, when the sub-reflector has an appreciable weight and is rotated at an appreciable speed, this arrangement can lead to undesirable mechanical vibration of the whole antenna since the center of gravity of the sub-reflector is oscillating, reducing the accuracy of the positioning and the lifetime of the antenna.

In a further aspect of the invention, therefore, we provide a Cassegrain antenna for conical scanning in which the sub-reflector is mounted to rotate about its center of gravity substantially on the main axis of the main reflector, but the main axis of the sub-reflector is angularly misaligned with that of the main reflector. This arrangement produces a conical scan but with no substantial mechanical vibration as a result.

The above aspects of the invention make it possible to provide an antenna which directly tracks the satellite and consequently is correctly aligned irrespective of movement of the vehicle or vessel upon which the antenna is mounted, providing good accuracy whilst employing a relatively simply and inexpensive construction.

Other aspects and preferred features of the invention will become apparent from the following description and appended claims.

The invention will now be illustrated, by way of example only, with reference to the accompanying drawings in which:

FIG. 1 shows schematically a water vessel including a satellite receiver antenna;

FIG. 2 shows schematically the general structure of a conical scanning antenna;

FIGS. 3a to 3f illustrate the principal of conical scanning;

FIG. 4 shows schematically a cassegrain receiver antenna;

FIG. 5 shows schematically the sub-reflector of an antenna of one embodiment of the invention;

FIG. 6 shows schematically the electrical components of a sensor in FIG. 5;

FIG. 7 shows schematically the electrical circuit used to rotate the sub-reflector of FIG. 5;

FIG. 8 shows schematically a wave form occuring at points in the circuit of FIG. 7;

FIG. 9a shows schematically an embodiment of a component of FIG. 7;

FIG. 9b shows the arrangement of two such components in the circuit;

FIG. 10a shows schematically the control circuit used to generate an error signal to position the antenna in one embodiment of the invention;

FIG. 10b shows an alternative arrangement of part of the circuit of FIG. 10a;

FIG. 11 shows wave forms produced at points of the circuit of FIG. 10a;

FIG. 12 shows schematically the position control system of the antenna according to one embodiment of the invention;

FIG. 13 shows schematically the external appearance of an antenna according to one embodiment of the invention;

FIG. 14 shows schematically a semi-sectional side elevation through the upper part of FIG. 13; and

FIG. 15 shows a corresponding front elevation.

Referring to FIG. 1, there is a water vessel comprising a hull A upon which is mounted a satellite reception antenna B directed towards a geo stationary satellite C. The antenna B effectively receives the signal from the satellite C provided the satellite C lies within the effective angular beam width of the antenna, which is defined by the antenna geometry and its operating frequency. It is usually quoted as the half power beam width, given as 57.3 L/D.degree., where L is signal wave length and D is antenna diameter. The hull A and antenna B are subject to rolling motions in three dimensions, namely pitching (fore and aft rotation), rolling (side to side rotation about a horizontal axis) and yawing (side to side rotation about a vertical axis). The magnitude of any one of these motions is sufficient normally to cause the antenna B to lose the signal from the satellite C over portions of each movement, thus perodically disrupting the received signal.

Referring to FIG. 2, in general a conical scan antenna comprises an antenna body 1 (shown as a cassegrain antenna comprising a main reflector 1a and a sub-reflector 1b), a detector 2 receiving the signal acquired from the antenna 1, a scan control generator 3 modifying the antenna properties to produce a conical scan antenna beam pattern, an error signal generator 4 receiving a signal from the detector 2 and generating, in response to this signal and to the angular position of the antenna 1, an error signal or signals which indicate the mis-alignment of the antenna with its target satellite, and a position control drive 5 receiving the error signal from the error signal generator 4 and modifying the position of the antenna 1 so as to reduce the magnitude of the error signal and hence improve the alignment of the antenna.

Refering to FIG. 3a, the conical scan produces a small angular mis-alignment between the center of the antenna beam pattern and the central axis of the antenna body 1, and rotates the antenna beam pattern so that the direction of mis-alignment rotates. Referring to FIG. 3b, when the main reflector 1a is directly aligned with the satellite, the view from the main reflector 1a would notionally show the center of the antenna beam pattern rotating symmetrically about the satellite position so that the degree of mis-alignment with the satellite is equal through each rotation, and consequently the strength of the signal received from the satellite is constant as shown in FIG. 3c.

Referring to FIG. 3d if the satellite is not aligned with the central axis of the main reflector 1a (due, for example, to rolling of the vessel), then as shown in FIG. 3e the degree of mis-alignment or eccentricity between the antenna beam pattern and the direction of the satellite has a minimum (when the satellite is most closely approached) and a maximum, and consequently, as shown in FIG. 3f the strength of the signal received from the satellite is modulated by a periodic variation, the amplitude of which corresponds to the degree of mis-alignment between the antenna and the satellite, and the phase of which indicates the direction of mis-alignment of the antenna.

The antenna could thus be exactly aligned by extracting the amplitude and phase of this signal strength variation and employing these as position setting signals to exactly align the antenna; or alternatively, amplitude and phase or related (eg quadrature) signals can be extracted and employed as feed back control signals for a position control system seeking to continually reduce or minimize the mis-alignment (rather than to produce completely correct alignment).

Refering to FIG. 4, in this embodiment the antenna 1 comprises the main reflector 1a consisting of a dish having a paraboloid profile; such dishes are commonly produced from aluminium or other metals by spinning and are available in a range of sizes. A diameter of 0.5-2 meters (eg. 90 cm) is generally adequate for reception.

The secondary reflector, or sub-reflector, 1b has an essentially hyperboloid reflector surface, and is mounted on a support structure ie positioned so that the focal point of its hyperboloid surface lies at the focus of the parabolic reflector 1a. An incoming signal is thus reflected from the surface of the main reflector 1a, off the surface of the secondary reflector 1b, which is focused upon a feed horn 2a at the center of the main reflector 1a acting as the signal receiver. The feed horn 2a is coupled to a commercially available radio frequency down convertor 2b. Preferably two convertors are provided, for respectively horizontal and vertical polarizations, to allow two different signals to be watched on two separate television sets.

Following the down convertor 2b, a signal strength measuring circuit is provided which produces an output corresponding to the amplitude of the envelope of the signal; where a television or radio tuner is provided it may be convenient to utilize the automatic gain control (AGC) signal output, but any other convenient circuits such as a diode mixer circuit or other type of envelope detector could be employed.

The dimensions and shape of the secondary reflector 1b and of the feed horn 2a, are determined within the constraint that the received signal must focus into the feed horn 2a. The secondary reflector 1b needs to be wide enough to receive substantially all the signal from the primary reflector 1a so as to maximize the strength of the signal focused onto the feed horn 2a, but on the other hand the wider the secondary reflector 1b becomes, the more it blocks the aperture of the main reflector 1a. Such blockage is inevitable, however, to some extent because of the supporting structure 1c behind the sub-reflector 1b (which include a scan motor as discussed below). Similarly the feed horn 2a should be small enough to remain in the shadow cast by the sub-reflector 1b so as not to interfere with the reflecting system, but ideally wide enough to receive the entire beam width from the sub-reflector 1b. These parameters are easily determined from the dimensions of the primary reflector 1a and the supporting structure 1c, and a sub-reflector 1b of an appropriate size and profile is easily produced, for example, by turning a metal blank on a lathe.

Referring to FIG. 5, in this embodiment, the sub-reflector 1b comprises a mushroom shaped metal component the upper surface of which is machined to a hyperbolic profile. An axial blind bore runs into the stalk or shaft of the sub-reflector 1b. The bore does not follow the axis of the subreflector 1b exactly; instead, it is arranged so that when mounted upon a spindle 10 supported on the support structure 1c and co-axial with the axis of the main reflector 1a, the focii for the two reflectors 1A, 1b co-incide (shown with the "o" symbol in FIG. 5) and the axis of the sub-reflector 1b diverges from that of the main reflector 1a by a small angle which determines the angle of mis-alignment of the conical scan. The angle of mis-alignment is to some extent a compromise between the effectiveness of the scanning (which favours a large mis-alignment) and the effectiveness of the antenna as a receiver (which is inevitably degraded since the antenna is never ideally aligned). It is found that a scan angle of around the theoretical antenna half power beam width (that is, the angular width around the main antenna axis at which the received signal strength falls to half the value received on the main antenna axis) is suitable. For example, with an antenna half power beam width of 1.7.degree. the angle of mis-alignment could be 0.5.degree. to 0.75.degree..

The elements comprising the scan control 3 of FIG. 2 in this embodiment will now be explained referring to FIGS. 5 and 6.

The sub-reflector 1b is secured to the spindle 10 by a grub screw 11 screwed through a bore in the reflector 1b to contact the spindle 10. A mounting plate 12 is connected by three legs 13a (only one leg 13a is shown) to the main reflector 1a. The spindle 10 running through the mounting plate 12 is an extension of the shaft of a DC motor 30 which consequently rotates the sub-reflector 1b and thereby causes its axis, and the beam pattern of the antenna as a whole, to revolve around the axis of the reflector 1a.

Also mounted upon the mounting plate 12 is a sensor 31 aligned with the shaft of the sub-reflector 1b. In the present embodiment, the sensor 31 is responsive to the angular position of the sub-reflector 1b and is conveniently provided by an optical encoder; for example, the Radio Spares reflective opto switch 2601 which comprises an infra-red light emitting diode (LED) 31a and photo transistor 31b arranged so that a reflective surface at a distance from the device reflects radiation emitted from the LED 31a to the photo transistor 31b which provides an output photo current, as shown in FIG. 6.

A reflective position defining mark 31c is provided on the sub-reflector 1b; conveniently, this is a strip of adhesive reflective tape of a length sufficient to cover half the circumference of the shaft of the sub-reflector 1b such that the output of the sensor 31 is high for half each rotational cycle and low for the other half.

The DC power supply to the sensor 31 and DC motor together with the output line from the sensor 31 are routed via a cable along one of the support legs 13a.

Referring to FIG. 7, the electrical circuit for driving the scan motor 30 to rotate the sub-reflector 1b and produce a concical scan comprises a reference frequency generator circuit 32 generating a stable signal at the frequency at which the motor 30 is to rotate.

Where the scanning frequency is to be harmonically related to the mains power supply frequency, it would be possible to derive the reference signal from the mains supply, but AC power supply generators for use on small boats often do not generate a stable supply frequency, however, and so in this embodiment it is preferred to employ a crystal oscillator 32a running at some convenient frequency (for example 3.2768 MegaHertz) and a digital pulse divider circuit 32b producing an output pulse every N input pulses, where N is the dividing ratio (for example 2.sup.18, or 262144 in this case). The divider circuit 32b may for example comprise commercially available counter-timer integrated circuits. One suitable arrangement comprises an M706BI (divide-by-2.sup.16) circuit followed by a 4013 dual flipflop device. The output of the reference signal generator circuit is therefore a square wave signal at a frequency of 12.5 hertz, as shown in FIG. 8d.

The output of the optical sensor 31b when the sub-reflector 1b is rotating alternates between a high and low level depending on whether the dark or reflective areas, respectively, of the shaft of the sub-reflector 1b are facing the sensor 31; corresponding optical inputs to the sensor 31 and electrical outputs of the sensor 31b are indicated respectively in FIGS. 8a and b. The transition between high and low levels in FIG. 8b is of finite width due to the finite aperture of the sensor 31, and the output of the photo sensor 31b is therefore supplied as an input to a comparator 33 the other input of which is supplied with a reference threshold line between the high and low output levels of the the photo sensor 31b. The comparator 33 may be an operational amplifier acting as an inverting comparator. The output of the comparator 33 is therefore a train of square pulses at the frequency at which the sub-reflector 1b is actually rotating.

A control circuit 34 receives the reference signal at the desired frequency and the sensor signal (output by the comparator 33) indicating the actual frequency and phase of the rotation of the sub-reflector 1b, and generates a control signal to control a power supply 35 feeding the motor 30 so as to bring the actual rotational speed towards the desired rotational speed.

The power supply 35 is conveniently a switched mode power supply acting as a voltage follower arranged to deliver a power output, for example from a 12 volt DC power source to the motor 30 on receipt of a switching signal. The control circuit 34 operates as follows. The reference pulse train and the sensor pulse train are each supplied to a signal conversion circuit which produces, in response, an output signal having two components; a DC component related to the frequency of the input pulse train and a small AC component superimposed thereon. FIG. 8e shows two signals of this type, the magnitude of the DC component (not to scale) being indicated as X and that of the AC component being indicated as Y.

One suitable convertor device is provided by the Radio Spares IC 2917 tachometer integrated circuit, which comprises essentially a frequency to voltage converter providing an output DC level X proportional to the input frequency. A small, AC ripple (approximately saw tooth in shape, as shown in FIG. 8e) of magnitude Y and frequency double the input frequency also occurs as a result of a charge pump within the device responding to each zero crossing in the input signal. To generate zero crossings, a capacitor (not shown) is positioned in the signal path prior to the input to each tachometer. In the prior art, this ripple is viewed as undesirable. However, in this embodiment, the ripple is utilised as follows.

As shown in FIG. 8e while the two input signals are at approximately the same frequency, the DC levels of the corresponding outputs of the two signal convertors 36a, 36b will be approximately the same and consequently the two output signal levels will cross at four points within each 12.5 hertz cycle. When the two frequencies differ, however, to an extent causing a difference in DC components X greater than the magnitude Y of the AC ripple the two, signal levels will not cross at all.

A comparator 37 (typically comprising an operational amplifier followed by a transistor acting as an inverting comparator), as shown in FIG. 9a receives the two outputs of the two signal convertors 36a, 36b and generates a high output while the magnitude of the signal from the reference frequency signal convertor 36b is greater than that from the sensor signal convertor 36a. Accordingly, if the motor revolution frequency is much slower than the reference frequency, the output from the reference signal convertor 36b is always higher than that of the sensor signal 36a and consequently the output of the comparator 37 is permanently high, causing the power supply 35 to permanently supply power to the motor 30 which consequently accelerates rapidly.

On the other hand, when the rotational frequency of the sub-reflector 1b, and consequently the pulse frequency of the signal from the sensor 31b, is considerably higher than that of the reference signal frequency the DC level of the output of the sensor signal convertor 36a is sufficiently high that it remains permanently above the level of the output of the reference signal convertor 36b and consequently the output of the comparator 37 remains low, so that the power supply unit 35 supplies no power to the motor 30 which consequently rapidly decelerates.

As a result either of such an acceleration or such a deceleration, inevitably the levels of the outputs of the two signal convertors will approximately co-incide, and, as shown in FIG. 8e, resulting in the comparator producing a series of output pulses having a width corresponding to the degree of overlap between the two signals (or, more precisely, to the time for which the reference frequency signal level is above the sensor signal level). Should the rotational frequency of the sub reflector 1b momentarily drop, the arrival of the zero crossings of the output of the comparator 33 is delayed and consequently the corresponding output of the signal convertor 36a will be delayed, resulting in an increase of the width of the output pulses from the comparator 37 and consequently an immediate increase of power supply to the motor 30 to restore the rotational speed. Likewise, a rise in rotational speed causes a decrease in the width of the output pulses of the comparator 37 and consequently a reduction of the power supplied to the motor 34.

This type of speed control operates almost instantaneously, twice within each rotional cycle. Should the rotational speed deviate from the reference frequency by more than a few cycles, the DC levels of the two signals differ to the extent that the signal levels do not overlap and the output of the comparator 37 stays high to accelerate the motor 30 to bring the rotational speed back to the reference frequency.

Referring to FIGS. 9a and 9b, the above referenced Radio Spares IC 2917 tachometer device includes a comparator. The circuit 34 thus comprises two such devices, the output of one 36b being supplied to the input of the comparator 37 of the other. The other input of the comparator 37 is connected to the output of its own tachometer 36a.

The two signal levels are set such that they overlap at the desired motor speed by a potentiometer circuit 38.

One embodiment of the invention using such devices therefore provides power to the motor 30 as a pulse width modulated signal when the motor frequency lies within a predetermined band (eg .+-.2%) around the reference frequency, and when the frequency lies outside this band, supplies power at either a 100% duty cycle to accelerate the motor or zero percent to decelerate the motor. Use of this type of device therefore provides fine control of the motor when it is close to the desired frequency and rapid acceleration or deceleration of the motor when it is far from the reference frequency.

Other advantages accrue from this embodiment of the invention; firstly, because the signal conversion device is responsive to the input signal frequency and zero crossings it is relatively insensitive to the shape or absolute level of the input signals, and two devices 36a, 36b of the same type will produce a similar output signals even in response to differing input signals. Secondly, by employing a pair of devices of the same type supplied from a common power supply, the effects of temperature variations (which can be quite marked when the antenna is mounted outdoors on a water vessel) are substantially the same on the output of each device and are thus eliminated at the comparator 37; much the same is true of other extraneous or intrinsic factors causing drift or variation in the counter devices.

The operation of the control signal generator 4 will now be discussed in greater detail. Briefly, the control signal generator 4 operates to sense the magnitude of the received signal at predetermined antenna orientations, and uses these to derive error signals indicating the mis-alignment of the antenna.

The first requirement is therefore to accurately determine the angular position of the antenna beam. This could be determined in a number of ways; for example, a further optical encoder could be provided associated with the sub-reflector 1b. It is however economical and convenient to employ the existing optical encoder 31 to provide a positional signal as well as a rotational speed signal. However, since the optical encoder 31 produces only one pulse per rotation of the sub-reflector 1b, it is necessary to further process the output to derive position signals for a plurality of rotational positions.

It would be possible to provide, instead of a single reflective area 31c, a plurality of radially distributed reflected bands. However, this is in practice not as convenient as providing a single detachable reflective strip 31c since it is harder to align a plurality of reflective areas accurately.

Accordingly, in this embodiment, a plurality of position signals are generated by interpolation from the optical encoder 31.

A phase locked loop is well known to comprise a controllable oscillator, the control signal for which is supplied from the output of a phase detector or (for example multiplier) circuit comparator. The phase detector receives an input signal and a reference signal and generates the control signal as a function of the phase of the input signal relative to the reference signal. The reference signal is supplied from the output of the controlled oscillator. If the phase of the input signal changes, a change will occur in the control signal, altering the frequency of the controlled oscillator. If the frequency of the input signal changes, a phase shift occurs and the control signal changes in such a manner as to vary the frequency of the oscillator to cause the reference signal to follow the input signal.

Accordingly, referring to FIG. 10a, the output of the comparator 33 is processed by a rotational positional signal generator 39 comprising a phase locked loop 39a (for example a 4046 device) the oscillator of which is set to run at a frequency a multiple of the desired rotational frequency of the sub-reflector 1b or, in other words, the frequency of pulses received from the comparator 33. Preferably the multiple is a power of 2; for example, 16. Thus, for a scan frequency of 12.5 hertz, the phase lock loop oscillator is set to run between around 150 and 250 hertz depending upon the control voltage applied.

The output of the voltage controlled oscillator of the phase lock loop 39a is supplied to a counter circuit 39b (responsive, for example, to positive or rising edges in the oscillator output). The counter circuit 39b is advantageously one of the many commercially available flip flop devices; for example a 4 bit continuously circulating counter which generates in response to successive inputs each successive binary digit between zero and fifteen.

The number to which the counter 33b counts before recirculating is related to the ratio of the phase lock loop frequency to the frequency input from the comparator 33; in a simple case the two are equal so that the counter or divider 39b counts through its range once each rotation of the sub-reflector 1b. The state of the highest order bit output line 40a from the counter 39b therefore changes at the same frequency as the signal input to the phase locked loop 39a from the comparator 33, and is fed back to the phase detector or multiplier of the phase locked loop 39a to provide the reference signal for the phase locked loop. The state of the lowest order bit in this embodiment changes at half the phase locked loop frequency.

The output of the phase locked loop 39a therefore tracks variations in the rotational speed off the sub-reflector 1b as they occur whilst maintaining a fixed phase relationship with the rotational position of the sub-reflector 1b at the correct rotational speed.

The digital output of the counter 39b therefore directly represents the rotational position of the sub-reflector 1b. This digital output, comprising 4 bit output lines 40a-40d in order of significance, is connected as the control input 41c of an analogue multiplexer device 41 (such as the 4067B CMOS 16-channel analogue multiplexer/demultiplexer device). Such a device comprises a single input line 41a receiving an analogue input signal and a plurality (eg 16) of output line 41b each selectively connectable to the input line 41a on the application of a corresponding multi-bit digital word to the control input lines 41c of the multiplexer 41. A further output line from the phase locked loop 39a at the phase locked loop frequency is connected to enable and disable the analogue multiplexer 41 at a rate of 200 HZ, so as to reduce (by half) the time during which the multiplexer passes the signal to 2.5 milliseconds and consequently enable a higher sampling accuracy.

The input line 41a of the analogue multiplexer 41 is connected to the signal detector 2 of the antenna to receive a signal indicative of the signal strength received by the antenna. During each rotation of the sub-reflector 1b, therefore, this signal is selectively switched successively to each of the outputs 41b of the analogue multiplexer 41.

A plurality (in this case, 4) of sample and hold circuits 42a-42d are connected to spaced output lines of the analogue multiplexer 41. In a preferred arrangement, pairs of sample and hold circuits 42a, 42c; 42b, 42d are connected to multiplexer output line separated by half a revolution one from the other. In a particularly preferred arrangement, the rotational spacing between the sample and hold circuits is equal.

Each sample and hold circuit may comprise a simple feedback amplifier storing charge upon an associated input storage capacitor during a period of 2.5 milliseconds in which the signal from the detector 2 is routed to that sample and hold circuit, and retaining the stored charge for the remaining 77.5 milliseconds of the rotational cycle thereafter. The circuit comprising the analog multiplexer 41a, associated input resistor 41d, and sampling capacitors provides a Commutating Analogue bandpass filter which, in known fashion, sharply attenuates frequencies not near the scan frequency or harmonics thereof.

The output of each sample and hold circuit therefore represents the signal strength sensed by the detector 2 at a respective antenna inclination angle relative to the satellite. The misalignment between the antenna and the satellite is thus determined by combining the sample and hold circuit outputs.

Referring briefly to FIG. 10b, in one simple arrangement, the sample and hold circuits corresponding to points separated by 180 degrees are subtracted by a pair of differential amplifiers 43a, 43b to provide respective error output signals. If the antenna is optimally aligned with the satellite, the signal strength received will be equal throughout the rotational cycle and the outputs of the differential amplifiers 43a, 43b will correspondingly be zero; in any other orientation of the antenna, the error signals will represent in two orthogonal axes the magnitude of mis-alignment of the antenna.

Where, as preferred, the rotational speed of the sub-reflector is an even multiple of the frequency of any modulation of the signal received and/or any electrical interference present, identical modulation and/or interference levels will appear at alignment positions separated by 180 degrees, and consequently at both inputs to each differential amplifier 43a, 43a so as to be cancelled by the differential amplifiers from the error signal outputs.

Since each sample and hold circuit 32a-32d is refreshed with a new signal once per revolution (for 12.5 hertz, once every 80 milliseconds), the output of the corresponding differential amplifiers 43a, 43b of FIG. 10b changes twice per revolution (ie every 40 milliseconds). In some applications, it may be desirable to update the error signal more frequently than this.

referring once more to FIG. 10a, in a preferred embodiment of the invention, the outputs of the 4 sample and hold circuits 42a-42d are connected each to one of its immediate neighbours. The sample and hold circuits 42a-42d are connected to output lines of the analogue multiplexer 41 selected such that they correspond to antenna inclinations at 45 degrees to the inclinations (eg horizontal and vertical) in which the antenna is steerable or to which the error signals generated correspond.

Each differential amplifier 43a, 43b therefore generates a signal responsive to the difference between the sums of corresponding opposed sample and hold circuits, so as to generate, as before, a pair of orthogonal error signals but since each error signal is now responsive to the outputs of all four sample and hold circuits 42a-42d its value changes four times each rotation of the sub-reflector 1b (or 20 milliseconds) so that the antenna responds quicker to mis-alignment.

It will of course be apparent that other arrangements of sample and hold circuits could equally be used to generate a pair of error signals, which need not themselves correspond to an orthogonal axis.

FIG. 11 illustrates the waveform outputs of the components of FIG. 10a.

It is important that the rotational positions which the sample and hold circuits operate (or, to be more precise, positions at which the sample and hold circuits stop sampling and start holding), relative to the vertical axis of the antenna, should be aligned to allow for any phase lag or other delays introduced within the position determining system. If the antenna is not properly aligned, mis-alignment in one axis will lead to correction in a different axis so that the antenna does not properly track the satellite.

In the above embodiment, alignment may be performed by manually aligning the antenna directly on a satellite or other signal source, and then elevating the antenna to introduce a vertical error but no horizontal error. The adhesive reflective strip 31c the sub-reflector 1b is then moved, whilst observing the vertical and horizontal error signal outputs on an oscilliscope, until the horizontal error signal output is eactly zero volts. Once a first antenna has been aligned, a second antenna of identical construction should not require separate alignment or calibration.

Referring to FIG. 12, the antenna position control drive 5 shown in FIG. 2 will now be discussed in greater detail. The error signals from the differential amplifiers 43a, 43b are connected to respective summing nodes 44a, 44b (comprising, for example, operational amplifiers). The respective outputs of the summing nodes 44a, 44b are connected as inputs to a pair of drive control units 45a, 45b supply respective output power levels to a pair of drive units 46a, 46b connected to physically move the antenna 1 in different directions.

Preferably, the drive units are arranged to move the antenna 1 in orthogonal directions; conveniently, they comprise a horizontal or azimuth drive 46a arranged to rotate the antenna in a horizontal plane and a vertical or elevation drive 46b arranged to rotate the antenna in a vertical plane. Conveniently, both drive units 46a, 46b are electrically powered motors; conveniently DC motors. The motors are arranged to run at a relatively high rotational speed (up to 1000-2000 rpm) for accuracy, and the drive units 46a, 46b in this case further comprise reduction gears (for example reducing the rotational speed by a ratio of 240), connected to respective vertical and horizontal rotation axes on the antenna 1.

The horizontal and vertical control units 45a, 45b each comprise a switch mode DC power supply, delivering a motor drive current proportional to the control signal from the respective summing nodes 44a, 44b to the corresponding drive motors 46a, 46b. The drive current comprises a DC supply pulsed at approximately 20 kilohertz, the pulse width being controlled to determine the motor current. One suitable control unit 45 comprises the L292 motor driver integrated circuit device supplied by SGS, connected as shown in FIG. 15 (page 36) of "A designers guide to the L290/2L291/L292 DC motor speed/position control system", Power Linear Actuators Databook - 2nd Edition, Jan 84.

Associated with each drive unit 46a, 46b is a velocity sensor 47a, 47b mounted to sense the rotational speed of the motor. The motor speed signal generated by the sensor 47 is fed back and subtracted at the respective summing node 44a, 44b. Conveniently the velocity sensor comprises an optical encoder comprising a light source, a light sensor and a rotating disc including a plurality of radially distributed reflective or transmissive elements arranged to modulate the light path between the sensor and the source, together with a frequency to voltage converter which converts the output of the sensor into a voltage level supplied to the respective summing node 44a or 44b. One suitable arrangement is the L290 integrated circuit described in the above referenced publication connected to the output of the Radio Spares Shaft Encoder Kit No. 631-532, described in Radio Spares Data Sheets 9394 (March 1989). This arrangement uses two phase-related outputs of the encoder to provide a bipolar, dependent, voltage level.

It will thus be seen that when a significant mis-alignment voltage appears at a summing node 44, the respective control unit 45 generates a significant motor drive current supplied to the drive unit 46 which correspondingly rotates the antenna to reduce the mis-alignment. As the rotational speed of the motor rises, the output of the sensor 47 also rises, causing the output of the summing node 44 to decrease and the motor drive current to decrease to a level sufficient to maintain a speed corresponding to the misalignment voltage. When alignment is reached, the error voltage applied to the node 44 from the preceding differential amplifier 43 becomes insignificant but the output of the sensor 47 remains high and therefore the control voltage supplied to the control unit 45 becomes negative, decelerating the motor 46 rapidly, so the output of the sensor 47 falls towards zero and the motor 46 stops.

This arrangement allows a very widely variable motor speed control permitting high accuracy and rapid response of antenna positioning.

Referring to FIG. 12, also provided at the summing nodes 44a, 44b are a pair of lines from a coarse alignment control unit 48 provided to allow the antenna to be initially positioned to point towards the satellite; for example this may comprise manual elevation and azimuth controls each comprising a manually variable potentiometer connected to a voltage source, to allow a user to manually align the antenna by variation of the potentiometers.

The elevation of a satellite will vary between nought and 90.degree. from the horizontal, and accordingly, the course control elevation potentiometer should be variable over, say, five volts corresponding to a range of 0.degree.-90.degree. from the horizontal. The azimuth of a satellite can vary 360.degree. degrees depending upon the alignment of the vessel or other item upon which the antenna is mounted, and accordingly, the azimuth course control potentiometer should have a maximum range corresponding to at least 360.degree. and preferably 720.degree..

Instead of manually aligning the antenna, it would instead be possible to supply store signals corresponding to predetermined inclinations to the coarse control unit 48.

The signal supplied by the potentiometers acts to control the azimuth and elevation drives in an equivalent manner to that in which the error signals do so, as described above.

Advantageously, the azimuth and elevation drives 46a, 46b are also arranged to generate azimuth and elevation position outputs (these may be generated by potentiometers or the velocity sensors 47a, 47b) which may be used, for example, to stabilize other ship borne machinery.

Referring to FIGS. 13 and 14, the antenna 1 is mounted on a horizontal pivot axle 49 by a pair of brackets 51a, 51b each carrying a weight 51, 51a which in combination counter balance the weight of the antenna 1. The horizontal axle 49 passes through a vertical axle 52 carried within a vertical outer sleeve 53 supported by a tripod comprising legs 54a, 54b, and a third leg (not shown). The azimuth drive 46a is mounted to the outer sleeve 53, and comprises a DC electric motor connected through two consecutive worm and screw gear boxes having reduction ratios of 20 and 12 to the inner vertical axle 52 so as to rotate the antenna 1 about the vertical axle 52. The elevation drive unit 46b may be essentially identical to the azimuth drive unit 46a, and is mounted through similar reduction gears to the horizontal axle 49 so as to pivot the antenna 1 about that axle. It is mounted so as to rotate with the antenna 1 (50a, 50b) around the vertical axle 52 when driven by the azimuth drive unit 46a.

Referring to FIGS. 14 and 15, the horizontal pivot axle 49 is mounted in a yoke 55 which comprises a pair of parallel plates bolted together, each plate including a semi-circular recess carrying the bearing for the axle 49. In order to prevent damage to the reduction gears in the event that movement of the antenna is obstructed or jammed, both the elevation and azimuth drive units 46a, 46b are mounted by their shafts to the antenna to allow slippage. As shown in FIG. 14, the axle 52 pivotting the antenna in azimuth is connected to the yoke 55 by a clamp comprising the two halves of the yoke 55 bolted by bolts 56a, 56b tightly around a recessed portion 57 at the top of the vertical axial 52. A similar clamp is provided on the other axle 49.

Due to the unavoidable manufacturing tolerances, backlash can occur in the reduction gear system making it difficult to rapidly halt the antenna and leading to unwanted vibration and instability on the drive system. Accordingly, a brake is provided, acting against rotation in azimuth. Referring to FIG. 15, the brake may comprise a split cylinder 58 consisting of a pair thin metal half shells 58a, 58b rigidly connected to the yoke 55 to rotate therewith, and clamped around the tube 53 (which is of reduced thickness at this point 53a) by a clamp exerting circumferential compression (eg, a hose or Jubilee clamp 59). Between the tube 53a and the shells 58a, 58b is a brake lining ring 60, comprising a material having an essentially constant co-efficient of static and dynamic friction (for example a strip of tape coated with Teflon (TM)). The brake assembly 53a, 60, 58, 59 thus acts to damp backlash.

The bearing within which the shaft 52 rotates is provided by a bush 61 of low friction material (preferably a ring of Teflon (TM)). It is important that the tube 53, vertical axle 52, legs 54 and mounting bush 61 should all be structurally rigid and without play or looseness, and that the antenna shall be rigidly mounted, or the tracking system can cause the antenna to vibrate.

To avoid backlash in the elevation drive system, the counter balance weights 51a, 51b do not quite balance the dish 1a so that the antenna is slightly "nose heavy".

In use, the legs 54a, 54b and third leg are bolted to the deck of a water vessel. Preferably, the antenna 1 is protected by a radome or hood transparent to radio frequencies.

In one embodiment, the sub-reflector is rotated at 50 revolutions per second. Any 50 Hz mains supply ripple manifests as a minor, constant offset alignment error. The 25 Hz triangular signal appears as an opposite error on alternate cycles, filtered out by averaging the output signal over two or more rotations.