The present invention relates to measuring apparatus and methods and is applicable to distance measuring apparatus and methods, position location apparatus and methods and to apparatus and methods for adjusting the subjective balance, for a listener, between the channels of a multi-phonic (e.g. stereophonic) sound reproduction facility to maintain the desired audio effect as the listener moves about the room or other space served by the facility.

It is also envisaged that position location apparatus and methods according to further aspects of the invention may be applied to robotics, e.g. to solve the problem of locating the position of a robot arm in relation to a workpiece or other structural member.

At various points in this dicclosure, reference is made to ultrasonic waves, to sound, and to pressure waves. For the avoidance of doubt, it is hereby declared that pressure waves of any frequency may be employed within the scope of the present invention, the exact choice depending on the particular application and the circumstances of each individual case, as will be appreciated by those skilled in the art.

One object of the present invention is to provide simple and effective apparatus for measuring distance between two or more fixed or moving points.

Another object of the invention is to provide a simple and accurate means for locating the position of an object with respect to reference coordinates, for example the position of a robot arm relative to other mechanical components.

According to one aspect of the invention, there is provided measuring apparatus characterized by: at least two transmission devices for transmitting respective first signals; transponder means for receiving said first signals from each said transmission means and for transmitting second signals in response thereto; receiving means for receiving said second signals; and timing means coupled to said transmission means and to said receiving means and arranged to time the time interval between transmission of each first signal by said transmission means and receipt of the corresponding second signal by said receiving means, at least one of said transmission devices and said transponder means being arranged to transmit signals in the form of pressure waves.

In one embodiment, said first or said second signals are electromagnetic signals, preferably infrared signals.

Where electromagnetic signals are used, preferably the transponder means is arranged to transmit electromagnetic signals as said second signals. If, on the other hand, ultrasonic signals are employed as said first and second signals, the second signals are preferably of a different frequency from the ultrasonic first signals.

Preferably, the receiving means is positioned at the same location as said transmission means. It is also conceivable however that the receiving means may be positioned elsewhere, for example at a central control unit.

The transmission means may include in one embodiment, first and second transmitters at spaced reference locations to transmit respective first signals, and said transponder means may be at an unknown location and responsive to both of the first signals, said timing means being arranged to measure respective timings for said first signals.

Preferably, the transmitters of said transmission means are arranged to transmit at mutually differing frequencies. This permits the individual transmitters to transmit simultaneously without mutual interference.

In an alternative embodiment, the transmitters of the transmission means are arranged to transmit at mutually differing times. This embodiment enables the first and second transmitters to utilize the same frequency without mutual interference.

Thus, time division multiplex or frequency division multiplex systems may be employed.

in an alternative embodiment, said transponder means includes first and second transponders at spaced reference locations which are arranged to transmit respective further signals in response to respective transmissions from said transmission means at an unknown location, said timing means being arranged to measure respective time intervals for said further signals.

The transmission means may be arranged to transmit at mutually different frequencies for respective transponders, or may be arranged to transmit at mutually differing times for respective transponders. In either

case, interference between the transmissions for respective transponders is avoided in a simple manner.

The first and second transponders may be arranged to transmit respective further signals at mutually different frequencies.

In order to obtain positional data in respect of the unknown location from said time intervals, computing means may be coupled to said timing means.

The spaced reference locations my be fixed relative to a pair of loudspeakers, and said computing means may be arranged to further compute a control signal based on said positional data, control means being provided for controlling the power supply to said speakers in response to said control signal in a manner such as to equalise the audio intensity from respective loudspeakers at said unknown location.

Expediently, the reference locations are within respective housings of said loudspeakers.

According to a further aspect of the invention, there is provided position location apparatus for determining the position of an object characterised by: first and second transmission means at respective spaced reference locations; transponder means on said object for receiving first and second interrogation signals from respective transmission means and for transmitting respective first and second response signals in response thereto; receiving means for receiving said response signals; timing means coupled to said first and second transmission means and to said receiving means and arranged to time a first interval between transmission of said first interrogation signal

and receipt of said first response signal and a second time interval between transmission of said second interrogation signal and receipt of said second response signal; and computing means coupled to said timing means and arranged to compute positional data representing the position of said object from said time intervals.

It is well known that the spatial illusion created by stereophonic sound reproduction is, for a given balance setting between the channels, effective only over a relatively small zone within the space served by the system. For example, where the volume settings of two channels of a stereophonic system are equal, the stereophonic effect is obtained only at points equi-distant from the speakers.

Consequently, if the listener is moving about the room attending to other matters whilst listening to sound reproduced by the system, the stereo effect will be intermittently lost. One object of the present invention is to provide means whereby the above mentioned defects may be avoided.

According to another aspect of the invention there is provided apparatus for maintaining the subjective balance, for a listener, between the channels of a stereophonic sound reproduction facility, or among the channels of a multi-phonic, e.g. quadraphonic, sound reproduction facility, as the listener moves around a space in which the loudspeakers of the facility are disposed, said apparatus being characterised by: a device adapted to be carried by the listener and capable of transmitting signals to receiving means fixed with respect to said speakers, and said receiving means being arranged to determine, from the signals received, the relative

distances of the device, and thus of the listener carrying the device, from said loudspeakers and to vary the respective volumes of sound reproduction from said channels accordingly.

Alternatively, the receiving means may be arranged to compute the absolute distances of the device from respective loudspeakers.

In a preferred embodiment of the invention, the device adapted to be carried by the listener takes the form of a small unit which may be carried in the listener's pocket or strapped to his wrist after the fashion of a wrist watch, or clipped to his clothing. The device may be arranged, for example, to emit, at short intervals, low-intensity, radio signals which are picked up by receivers mounted close to the respective loudspeakers for the respective channels. A central control facility may then process the signals received by the respective receivers to determine the relative distance from the transmitting device to the two receivers, the control facility being arranged, on this basis, to adjust the balance between the channels to maintain the subjective balance as judged by the listener.

In another arrangement, the speakers of the two channels may have respective transmitters associated therewith, arranged to transmit signals to the device carried by the listener, and the latter device may be arranged to compare the signals received from the two transmitters and to transmit corresponding data to the central control facility which on this basis determines the position or relative distances of the listener and adjusts the balance accordingly.

The ultrasonic frequency may be in the range of 50-100 kHz, although the application is not limited to this particular range.

While certain aspects of the invention have been described by reference to a stereophonic system, it will be appreciated that the balance adjusting system described may be used in conjunction with a system utilizing more than two channels, for example in relation to a quadraphonic system which utilizes four separate channels for sound reproduction.

Further aspects of the invention are defined in the claims.For a better understanding of the invention, and to show how the same may be carried into effect, reference will now be made by way of example to the accompanying drawings, in which:

Figure 1 is an explanatory schematic block diagram of ultrasonic measuring apparatus to illustrate the principles of embodiments of the present invention;

Figure 2a is a block circuit diagram of control and timing apparatus for use in the embodiment of Figure 1;

Figure 2b is a functional representation of the circuit of Figure 2a;

Figure 2c is a waveform diagram illustrating the principles of operation of the circuit of Figures 2a and 2b;

Figure 3a is a block circuit diagram of a transponder for use in the embodiment of Figure 1;

Figure 3b is a functional diagram to illustrate the circuit of Figure 3a;

Figure 3c is a waveform diagram illustrating the principles of operation of the circuit of Figures 3a and 3b;

Figure 4 illustrates a typical radiation pattern of a typical loudspeaker;

Figure 5a is a block circuit diagram of distance measuring apparatus according to one embodiment of the present invention;

Figure 5b shows a reset circuit for use with the embodiment of Figure 5a;

Figure 6 is a block circuit diagram of a transponder for use with the embodiment of Figure 5a;

Figure 7 is a waveform diagram illustrating operation of the circuit of Figure 5a;

Figures 8a and 8b illustrate a block circuit diagram of distance measuring apparatus according to a second embodiment of the present invention;

Figure 9 is a block circuit diagram of a transponder for use with the embodiment of Figures 8a and 8b;

Figure 10 illustrates apparatus for stereo balancing according to a further embodiment of the invention;

Figure 11 illustrates a block circuit diagram of a circuit for use with the apparatus of Figure 10;

Figure 12 is a block circuit diagram of a further part of the circuit of Figure 11;

Figure 13a illustrates characteristic curves of a typical JFET;

Figure 13b illustrates waveform diagrams illustrating operation of the circuit of Figures 11 and 12;

Figure 14a is a resistor arrangement for use with the circuit of Figures 11 and 12;

Figure 14b is a circuit arrangement for use with the circuit of Figures 11 and 12 in place of the circuit of Figure 14a;

Figure 15 is a schematic diagram of position sensing apparatus according to another embodiment of the present invention;

Figure 16 is a block circuit diagram for use with the apparatus of Figure 15;

Figure 17 is a block circuit diagram of timing circuitry for use with the circuit of Figure 16;

Figure 18 illustrates system scaling apparatus for use with the embodiment of Figure 16;

Figure 19 is a computing circuit for use with the embodiment of Figure 16;

Figure 20 is a further computing circuit for use with the embodiment of Figure 16;

Figures 21 and 22 are flow charts illustrating the operation of the circuit of Figure 20;

Figure 23 is a waveform diagram relating to the circuit of Figure 16;

Figure 24 illustrates application of apparatus according to the invention for ploughing a field; and

Figure 25 schematically illustrates an interrogator provided with a sound-absorbing collar.

GENERAL PRINCIPLES

Figure 1 illustrates schematically ultrasonic measurement apparatus comprising first and second interrogator devices 4 and 5 coupled to a central control unit 3. A transponder unit 6 is positioned at a location spaced from the interrogators 4 and 5. Within the control unit 3, is provided processing circuitry for deriving distance and/or position data from the outputs of respective interrogators 4 and 5. It will be appreciated that the transponder 6 comprises a receiver and a transmitter and is arranged such as to transmit a signal in response to reception of a transmission from either of interrogators 4 and 5.

In outline, the measuring process takes place by the transmission of signals from respective interrogators 4 and 5 to the transponder 6 which then operates to send response signals for reception by respective interrogators. By timing the transit times for the various signals, the distances from the interrogators may be computed and the position of the transponder relative to the interrogators may if desired be computed from this distance information.

The interrogators 4 and 5 preferably each contain a respective ultrasonic transmitter and an infrared receiver, in which case the transponder comprises an infrared transmitter. Alternatively, the interrogator receiver may be an ultrasonic receiver and the transponder transmitter may be an ultrasonic transmitter. Ultrasonic frequencies in the range of 50-100 kHz may be employed although this will depend upon the particular application. To avoid environmental problems, it is preferable that the frequency is above the audible range of any animals likely to be in the vicinity, such as dogs.

It is preferred that the transmitters associated with units 4 and 5 have a directional characteristic such that they transmit over a range of angle less than 180 degrees. On the other hand the transmitter and receiver contained in transponder should be omni-directional.

Reference will now be made to Figure 2a which shows a block schematic diagram of one channel of the processing and control block 3 of Figure 1. A receiver interface 10 receives signals on line 13 from the receiver of interrogator 4 or 5 of Figure 1. Similarly, a transmitter control block 11 supplies control signals on line 14 to the transmitter of block 4 or 5 of Figure 1. In addition, when block 11 causes the transmitter to transmit (by producing the control signal) it simultaneously provides a timer start signal on line 16 for starting a timing operation by a timer 12. When interface 10 receives a signal on line 13 from the receiver, it emits a timer stopping signal on line 15 for stopping the timing operation of timer 12. The result of the timing operation is then supplied on an output line 17.

In functional terms, the circuit of Figure 2a may be considered as shown in Figure 2b which could be implemented as an integrated or hybrid circuit. inputs and outputs RCV2 and TRNSl are provided for the received signal and signal for transmission respectively. Also shown are a reset input, a data output bus and a cyclic signal CYCLE on terminal START which will be explained in more detail hereinafter.

Figure 2c illustrates the signal CYCLE together with an indication of the data out information and the timing of the transmitted and received signals.

Figure 3 shows a block schematic diagram of the transponder 6 of Figure 1. A receiver 20 detects receipt of a signal from the transmitter of interrogator 4 or 5 and in response provides a signal to a receiver interface 21. Interface 21 provides a control signal to a transmitter interface 22 which thus causes a signal to be transmitted by a transmitter 23.

In a similar manner to Figure 2b, Figure 3b illustrates a transponder circuit as a functional block. In this case on an input signal RECEIVE and an output signal TRANSMIT are shown, the timings of these signals appearing in the waveform diagram shown in Figure 3c. Again, the transponder circuit may be implemented as an integrated or hybrid circuit if desired.

DISTANCE MEASURING (FIRST EMBODIMENT)

Reference will now be made to Figure 5a which shows a more detailed circuit diagram of interrogator 4 or 5 of Figure 1 and the associated control and processing circuitry.

Upon receipt of a cyclic signal START of pulse length DT, a switch 31 closes and allows a sinusoidal signal F]_ to pass from a signal generator 33 via an amplifier 32 to an ultrasonic transmitter 30 as signal TRNS1. The resulting ultrasonic wave is received by the transponder, not illustrated in Figure 5a, which operates to send an infrared signal in response. This signal is detected by a photo-transistor 34 supplied with operating voltage via a load resistor R _c .

In response, the transistor provides a signal RCV2 to an inverting amplifier 35 whose output STOP is connected to both a delay circuit 36 and to the RESET input of a bistable flip-flop 37 whose SET input is connected to receive signal START. The output of the delay circuit 36 is connected to the clock input of a data latch 38 and to the input of a non-retriggerable monostable multivibrator 39.

The output of the signal generator 33 is connected to the non-inverting input of a comparator 40 whose inverting input is grounded. The output of the comparator 40 and of the multivibrator 39 are connected to respective inputs of an AND gate 41 whose output AND is connected to the input of a further non-retriggerable monostable multivibrator 42, the output of which provides the signal START. The Q output of the bistable flip flop 37 is connected to the count input of a counter 43, which is clocked at a frequency F2 and may be cleared by a signal CLR. Data from the counter 43 is read by the data latch 38 in response to the output signal from the delay circuit 36 and is subsequently supplied as output data representing a timed measurement result. The CLR signal is derived from the output of monostable circuit 39 via a differentiator 44, across which a diode is connected.

When the infra-red wave is received by the phototransistor 34, the flip flop 37 is reset so that its output Q stops the counting process of the counter 43. After the delay introduced by circuit 36, the data latch 38 reads the counter output. The delay is necessary to ensure safe latching of valid data at the counter output. The delay is slightly longer than the time necessary from generation of signal STOP until the counter output is valid. However, the delay is a very short inverval and will be negligible in comparison with the time necessary for sound waves to cover a.practical measuring distance. The output of the AND gate 41 will rise only when the following two conditions are fulfilled: (1) The counter output has already been latched and (2) the output of comparator 40 is high. The comparator output will rise only when the sinusoidal output of the clock generator 33 is rising through zero. This ensures that the output of gate 41 will be true only after the data has been latched and when the sinusoidal output of the generator is greater than zero and rising. The interval timed by monostable circuit 39 should be about 2.5 times the period of the generator 33. The output START of monostable circuit 42 will be high for the transmitting period DT and will be synchronised with the transmitted ultrasound wave. Before the signal START goes high, the counter 43 will be cleared by signal CLR and only after it is cleared will the signal START set the R.S. flip-flop 37 to initiate counting. At the same time, the transmitting section is enabled. The generator 33 supplies its output to the transmitter 30 when the switch 31 is closed by the signal- START. It will be understood that the signal START is also available for peripheral circuits wishing to read the data latched in the data latch 38. That is to say, a peripheral circuit may read out the data in response to the START signal.

Naturally, in any particular measuring cycle data from the preceding cycle will be read out.

There are a few considerations that should be borne in mind in respect of the frequency control of the ultrasound transmitter. As has been explained, switch 31 is closed by signal START for a controlled interval DT to cause a "pulse ¹¹ of ultrasound to be transmitted. The length of the transmission interval DT is set by monostable circuit 42. It can be shown by Fourier transform techniques that where a sinewave of frequency Fi is transmitted for an interval DT the resulting transmitted pulse contains a band of frequencies centered on frequency F]_. The majority of the frequencies are however concentrated within a bandwidth DF, where DF=1/DT. It is clear from this relationship that the bandwidth of the pulse is larger as the pulse interval is reduced. The consequence is that the interval DT should be as large as possible. On the other hand, DT cannot be allowed to exceed the the time taken for the ultrasound wave to traverse the distance required to be measured.

In this connection, it is also necessary to consider the effect of DT on the actual frequency transmitted. If it is desired to maintain the transmitted frequency within X% of the center frequency, the minimum transmitted frequency will be 100X/DT. This must be considered when there is a limit on the frequency which may be transmitted. Generally, the frequency should be maintained as low as possible, because the transmission loss increases as a function of frequency, as will be discussed in more detail hereinafter.

The frequency F2 of the counter clock depends on the required accuracy. If it is desired to measure distance

to within an error of E, then F2=V/E, where V is the velocity of the ultrasound wave.

Figure 5b shows a circuit for use in conjunction with the circuit of Figure 5a to effect system reset. The output of an OR gate 56 provides a signal for clearing and resetting the counter 43 of Figure 5a. The OR gate 56 has two inputs taken respectively from the output of the debouncing circuit 44 and the output of a similar circuit 55 comprising a capacitor 53, a resistor 54 and a diode 55. The signal input to the capacitor 53 is derived from a monostable circuit 52 which is triggered by operation of a RESET switch 50. The input of the monostable circuit 52 is connected to ground via a resistor 51.

Reference will now be made to Figure 6 which is a more detailed circuit diagram of the transponder 6 of Figure 1.

An ultrasound receiver device 60 is connected to an amplifier 61 which in turn is connected to the input of a band-pass filter 62. The output of the filter 62 is connected to the non-inverting input of a comparator 63 whose inverting input is connected to a reference voltage ^V REF• ^A non-retriggerable monostable circuit 64 having a period 2 receives its input from the comparator 63 and supplies its output via a buffer amplifier 65 and a differentiator circuit (comprising capacitor 66, resistor 67 and diode 68) to a Schmitt trigger circuit 69. The output of the Schmitt trigger is connected to the base of a first PNP transistor Qτ_ whose collector is connected to base of a second NPN transistor Q2. In the collector emitter path of the transistor Q2 is provided an infrared transmitter in the form of an LED 74. Resistors 71 and 72 provide correct biassing conditions for the transistor Q _] _.

The circuit operates as follows. After filtering by the band-pass filter 62, the received signal is compared with ^V REF *°y comparator 63 which thus sets the receiver sensitivity. The first rising edge of the received signal at the output of the comparator 63 causes the monostable circuit 64 to be triggered. Further triggering of the circuit 64 is not possible until its time period has expired. The output pulse of the circuit 64 is differentiated by the capacitor-resistor pair 66,67 and any negative pulse resulting is limited to 0.2 volts by the Germanium diode 68. The resulting signal operates the Schmitt trigger 69 which thus operates the infrared transmitter circuit composed of transistors l a d Q2, the LED and the resistors Rj and R2.

It will be appreciated that the pass band of the filter 62 should correspond to the bandwidth DF of the transmitted ultrasound having center frequency Fτ_ as discussed above.

The detection time of the transponder depends upon V-^gp at the comparator and may be taken into account when computing the transit time of the ultrasound wave.

The operation of the transponder may be more clearly understood by reference to Figure 7 which shows six waveforms occurring in operation of the circuit of Figure 6. Waveform I represents the signal at the output of the amplifier 61, waveform II represents the output of the comparator 63, waveform III represents the output of monostable circuit 64, waveform IV represents the output of the capacitor- 66, waveform V represents the output of the Schmitt trigger 69 and waveform VI represents the voltage across the LED 74.

DISTANCE MEASURING (SECOND EMBODIMENT)

Reference will now be made to Figures 8a and 8b which show apparatus generally similar to that illustrated in Figures 5a and 6 but modified to employ ultrasonic transmissions not only from the interrogators but also from the transponder. Thus bidirectional ultrasonic transmissions are employed and infrared transmitters and receivers are not used. It will be appreciated that this "double ultrasound" implementation is fundamentally less accurate than the apparatus discussed with reference to Figures 5a and 6, but may be of advantage in situations where infrared radiation cannot effectively be employed, such as under water.

In Figures 8a and 8b those components which are equivalent to those already explained and described with reference to Figures 5a and 6 have been provided with the same reference numerals and further detailed description thereof will be omitted. It will merely be pointed out that the receiving section of the circuit of Figure 8a corresponds to that of the transponder of Figure 6 and the storage and timing section of Figure 8a corresponds to that of Figure 5a. The only qualification to this statement is that only a single ultrasonic transducer 30(60) functions both for receiving and for transmitting, so that a change-over switch 31a is required in place of the single pole switch 31 of Figure 5a.

Figure 9 illustrates a transponder circuit for cooperation with the interrogator circuit illustrated in Figures 8a and 8b. Again, components having similar construction and function to those of Figures 5a and 6 are provided with the same reference numerals. Monostable circuit 70 which

receives the output of monostable circuit 64 and provides an output to control a change-over switch 31b has no direct equivalent in Figure 5a (although it is analagous to monostable circuit 42) and has therefore been provided with its own reference character.

The apparatus illustrated in Figures 8a, 8b and 9 operates with the same ultrasound frequency transmitted from both the transducers and the transponder. If it were desired to employ two different frequencies, two separate ultrasonic transducers could be employed both in the transponder and in the transmitting and receiving circuit.

In the illustrated example, the minimum error will be of the order of one wavelength. To improve the accuracy to half a wavelength, a rectifier circuit could be inserted between the amplifier 61 and the filter 62 of Figure 8a. This will permit the comparator 63 to respond at each zero transition of the signal and not only on a positively travelling transition.

Comparing this implementation with that of Figures 5a to 6, the main difference is that here the accuracy depends directly on the frequency value selected, so that accuracy can be improved by increasing the frequency. In contrast, accuracy in the method described earlier depends mainly on the counter clock frequency employed to measure the transit times.

In the present embodiment, the frequency of the counter clock should be selected according to the possible accuracy. If measurements are being made to the nearest wavelength and the center frequency of the return wave is F2, the counter clock frequency should be at least F2. On the other hand, if measurements are made to the nearest

half wavelength, the clock frequency should be at least 2F ₂ .

in the present embodiment, measures must be taken to ensure that reflections of the first transmission are not interpreted by the associated receiver as transmissions from the transponder. This problem can be avoided by using different frequency bands for the two transmissions. Of course, this problem does not exist in the case of the first embodiment, since the interrogator is expecting electromagnetic waves and not ultrasonic waves to be returned.

As regards the system timing, the waveforms I, II and III of Figure 7 are also applicable here, where waveform III is the output of monostable circuit 64. This circuit triggers monostable circuit 70 with its positively travelling edge to initiate the transmission interval DT for the transponder. Although the transmission intervals for the transponder and the interrogation circuit are here considered as the same for simplicity, it will be understood that if desired they may differ.

MULTI-PHONIC BALANCER

Referring now to Figure 10, there is illustrated schematically apparatus for adjusting automatically the stereo balance perceived by a listener as he moves around a space served by a stereo music reproduction system. The apparatus includes first and second loudspeakers 100 and 200 coupled to a central control unit 300 of a stereophonic reproduction system having an amplifier 350. Associated with the respective loudspeakers are interrogator devices 400 and 500. The devices 400 and 500 may be located within the respective speaker housings, or

may be mounted in separate units attached to respective speakers. A transponder unit 600 is adapted to be carried by a listener and takes the form of a small unit which may be carried in the listener's pocket or strapped to his wrist after the fashion of a wrist watch, or clipped to his clothing.

In addition to the control circuitry 300, there is provided processing circuitry 250 for deriving distance information from the outputs of respective interrogators 400 and 500 to enable the control unit 300 to control the power supplied by the stereo amplifier 350 to respective speakers 100 and 200.

In outline, the measuring and adjusting process takes place in three stages. First of all, the relative or absolute distances of the unit 600 from the stationary interrogators 400 and 500 is measured, secondly the effect of these distances on the stereo balance is computed, and thirdly the balance is adjusted in accordance with the computed result.

The devices 400 and 500 each contain a respective ultrasonic transmitter and preferably also an infrared receiver. The ultrasonic frequencies used are preferably in the range of 50-100 kHz, although the frequency used will depend on the circumstances. To avoid environmental problems, it is preferable that the frequency is above the audible range of any pets likely to be in the vicinity, such as dogs. The system works most effectively when the walls of the room in which it is contained are reasonably non-reflecting. Most domestic living rooms containing curtains, carpets etc. will satisfactorily meet this requirement. This aspect will be discussed in more detail hereinafter.

It is preferred that the transmitters associated with units 400 and 500 have a directional characteristic such that they transmit over a range of angle less than 180 degrees. On the other hand, the transmitter and receiver contained in unit 600 should be omni-directional.

It will be understood that the circuit for the transponder 600 will be constructed in accordance with that illustrated in Figure 6.

Referring now to Figure 11, the construction of the control and processing circuitry of Figure 10 will be described in more detail. Blocks 701 and 702 represent infrared receivers, block 810 a timing control unit, and blocks 901 and 902 ultrasonic transmitters. The system operates to send ultrasonic interrogation signals to the transponder by means of the transmitters 901 and 902. When an interrogation signal is transmitted, a timer of block 810 is started. When the transponder has responded, such that an infrared signal is received by receiver 701 or 702, the timer is stopped, and the timed interval is employed as a measure of the distance to the transponder from the respective interrogator and the associated loudspeaker.

The processing and control circuit 810 is in fact constructed in an analagous manner to that illustrated in Figures 2b and 5a. It will therefore suffice if it is explained that circuit 810 comprises blocks 32, 33, 35, 36, 37, 39, 40, 41, 42, 43, and 44 of Figure 5a so that further description thereof is omitted. Similarly, a reset arrangement similar to that shown in Figure 5 is provided and includes a monostable circuit 820, a first resistor 821 grounding the input of circuit 820, a

capacitor 822, a second resistor 823, and a diode 824. No data latch is provided in the block 810 of Figure 11; this is instead arranged as shown in Figure 12.

In addition to the components already described, the apparatus of Figure 11 includes timing logic to ensure correct system sequencing and timing. The object of such timing is to cause interrogation signals to be transmitted alternately from the two transmitters 901 and 902 and to ensure that a response signal is received and processed for each transmission before a further transmission is sent. Such timing is provided by a D-type flip-flop 850 in conjunction with two AND gates 851 and 852. As illustrated, the flip-flop 850 is clocked by the signal START from the processing circuit 810 and provides complementary signals ENl and EN2 on its Q and ^" Q outputs. The AND gates receive respective signals ENl and EN2 and are also connected to receive the signal START. Output signals from the AND gates are designated LTCHl and LTCH2 and are employed for latching data from the counter of circuit 810 as will be explained with reference to Figure 12. Signals ENl and EN2 are used to control switches SI and S2 for enabling the receiver/transmitter pair 701,901 or 702,902. The system timing appears more clearly from Figure 13b which is a waveform diagram illustrating the signals START (CYCLE), ENl, EN2, LTCHl, and LTCH2.

Referring now to Figure 12, two data latches 812 and 813 receive data from the DATA output of the processing circuit 810 on a data bus 811 in accordance with latching signals LTCHl and LTCH2. This data is then passed through respective digital-to-analogue converters 814 and 815 to provide signals Vg _S τ_ and V _gs 2 respectively.

Of course, to maintain the stereo balance it is only necessary to ensure that substantially equal sound intensities are received at the unit 600 from respective speakers 100 and 200. This will ensure that the stereo effect is maintained irrespective of the position of the unit 600 and irrespective of the absolute values of the sound intensities present. For adjusting the stereo balance, it is therefore only necessary to determine the ratio of the respective distances from the transponder to the speakers, and for this purpose the ratio of the two time intervals determined by the timer will suffice.

It remains to be described how the adjustment of the stereo balance is effected in response to the output signals Vg _S χ and Vg _S 2« This may be achieved either by direct electronic control using, for example a variable resistor such as a JFET connected in the loudspeaker volume control circuitry, or alternatively by using a mechanical system which mechanically rotates a balance control spindle whose rotation effects a necessary change in the stereo balance. Electronic control will be discussed in more detail with reference to Figures 13a, 14a and 14b.

In order that the balance between the channels will be continually maintained, the power (in Watts) transmitted from each speaker will have to be changed when the distances d]_ and d2 are changed. Since the power transmitted from the respective loudspeakers is proportional to the square of the voltages (Vτ_, V2) applied across respective loudspeaker impedances, and since the power received by a listener falls off as the square of the distance, it may readily be shown that:

Vχ/V ₂ = d _x /d ₂ ,

where d_ and d2 are distances from the transponder to respective speakers.

Since the final power stage of an audio amplifier (at the input of which stage the balance is generally controlled) has linear behaviour (in the active region), which means that the ratio of the power on each speaker to the power on the final stage input is constant, changes of voltage ratio at the final stage input would have the same effect as changes of voltage ratio at the speakers.

Thus, where Uι_ is the input voltage to the first input power amplifier and U is the input voltage to the second input power amplifier, the result is obtained that:

Therefore, the balance can be controlled by automatically adjusting the ratio U ^* 2/U]_ according to changes in the ratio d2/d]_.

In a stereophonic system, for each location of the listener (transponder) in the room, there will be a different value for the ratio U2/U1 so that the listener will not need to adjust the balance at all, not even initially. All adjustments may be perfomed automatically as a function of the transponder (listener) location at any given moment.

One way to implement the voltages ratio changes would be with JFETS: (active resistances) . A set of characteristics for a typical JFET is shown in Figure 13a.

The beginning of each curve of I _D (drain current) against V _D g (drain-source voltage) for a constant value of V _Q S (gate voltage relative to the source) is linear. In other words, for each VQ- a different value of RDS* ^the drain-source resistance, will be obtained for small signals.

Thus, operation in the small signal regions of the characteristics can provide controllable resistance.

If the audio amplifiers are provided with variable passive resistances at the input of each final stage, the arrangement could be as shown in Figure 14a. As shown, each channel has a series connection of two resistors, ^R 2' ^R 1 ^{and R} 1' ^R 2 respectively. The output voltage is taken from the interconnection node of each resistor pair in each case.

Increasing the balance in one side decreases the balance for the other side.

This sort of arrangement is often used for conventional manual stereophonic balance adjusting, and uses a logarithmic potentiometer, so that linear changes in the balance potentiometer will compensate for the logarithmic sensitivity of our ears. However, for automatic balance adjusting the changes will be linear.

In order to implement the variable resistances with JFETs, the JFETs all have to be compatible with the same characteristics. A possible arrangement is shown in Figure 14b. In this case, the passive resistors of Figure 14a are replaced by respective JFETs Qτ_, Q2, Q3 and Q4 which are connected in pairs for respective channels and are supplied with current from respective preamplifiers of the amplifier block 350.

When Vr- is the threshold voltage of each JFET, and Vg _S ι and Vg _S 2 are the signals applied to the gates of respective JFETS, it may be shown that:.

'out = ( gsl - V _T )/(V _gsl + V _gs2 - 2V _T )

Since in our case V2/V1 = d2/d , gate control signals Vg _S τ_ and Vg _S 2 for the JFETs may be derived as shown in Figure 12. The outputs are taken from the central node of the respective JFET pairs and are connected to respective power amplifiers of block 350.

It would also be possible, on the basis of the timed intervals and the known velocity of sound in air (331.7 metres per second) to derive a measure of the distance travelled by the two ultrasonic signals. This distance is of course the distance between the loudspeaker unit 400 or 500 and the transponder unit 600. From this distance value, a power compensation circuit could compute a measure of the sound intensity at the position of the transponder 600 caused by the loudspeaker associated with the relevant unit 400 or 500.

A similar measurement and computation process could be effected for each loudspeaker 100 and 200 and associated unit 400 and 500. From the resulting information, bearing in mind that received sound intensity falls off as the square of the distance, an appropriate adjustment could be made by a control device contained within the unit 300 to maintain the intensity from each speaker at a constant value as the distance changes. It will be noted that the position of the unit 600 could also be determined from the distance information, since it is located at the intersection of two circular arcs centred at respective

units 400 and 500 of radius equal to the computed distances. It is true that this information gives two possible positions for the unit 600, but one of these positions is excluded in the normal case since the loudspeakers are usually mounted adjacent a wall and one of the positions would be outside the room. However, it is not actually necessary to compute the coordinates of the unit 600 in order to correctly adjust the stereo balance or the absolute volume.

Since there are two separate loudspeakers, it is clearly necessary to provide some means for distinguishing the signals from unit 500 from those from unit 400. This may be achieved in one of at least two ways. In one possible method, unit 400 and unit 500 operate in differing frequency bands. In this case, the transponder in unit 600 must be capable of receiving and transmitting in two separate frequency bands. In a second method, as described in the illustrated embodiment., signals from units 400 and 500 are time multiplexed under the control of unit 300 and in this case the transponder need only receive and transmit on a single frequency or in a single frequency band. This reduces the complexity of the transponder, but of course correspondingly increases the complexity of the control unit 3.

As an alternative embodiment, it is possible to place the transponder in each of the units 400 and 500 and to replace the transponder in unit 600 by an interrogator. This system however would require some means of transmitting data indicating instants of transmission and reception from the unit 600 to the timing unit 250. This could be achieved by means of a wire or by means of a wireless link. However, the added complexity makes this alternative less attractive.

Another alternative would be to provide a simple transmitter in unit 600 and receivers in units 4 and 5. Whilst the time of receipt could readily be transmitted to unit 3 in such an embodiment, the difficulty is again presented that a wire or wireless link is required to transmit data indicating the instant of transmission from unit 600 to timing unit 250.

It will be appreciated that when units 400 and 500 are built into the housings of respective speakers 100 and 400 all necessary control and data lines may be conducted within a common speaker cable, preferably the same cable utilized for supplying the audio signals to the speaker.

Control of the system may be effected either by dedicated electronic control logic circuits as illustrated in conjunction with Figure 5a, or by means of a microprocessor in conjunction with an appropriate program stored in a read only memory.

Considering the implementation of the multi-phonic balancing system in more detail, there are three important factors to consider:

1. The source (the speaker).

2. The environment (closed space).

3. Symmetrical organization.

For simplicity, the description and analysis for the first two factors will refer to one speaker.

1 . The Source

For an isotropic source (ideal source) transmitting in an isotropic medium (open space), the sound power (P) at each point at distance D from the source will be proportional to 1/D ² .

Figure 4 shows a typical radiation pattern of a non-omnidirectional (approx. 180°) speaker.

Although at any instant there will be a difference between the frequencies transmitted by respective speakers, the transmission loss at distance D from each speaker may be regarded as about the same for each speaker because the difference in frequency is usually slight and because there is a logarithmic relationship between the frequency and the transmission loss.

Therefore, the loss characteristic as a function of frequency need not be taken into account.

The speaker is a non-isotropic source because it has a non-zero size. This also need not be taken into account because of the following: a) its volume is negligible compared to the volume of the space occupied; and b) in stereophonic systems the supposition is made that all speakers have the same volume, size and structure.

2. The Environment (Transmission Medium)

When the speaker is transmitting inside a closed space, there are reflections from the floor, the walls, the ceiling and other objects, and the assumption that P is

proportional to 1/D ² may not be exactly true. In order that this relation would be applicable, reflections should be avoided as much as possible. For example, when the speaker is inside a room that is well carpeted, has curtains covering all the walls and has an acoustic ceiling, the reflection effect would be negligible. In this case we can suppose that we are dealing with an isotropic medium, and so we will assume this for the user.

3. Symmetrical Organization

In a multiphonic (e.g. stereophonic or quadraphonic) system, it would be helpful to keep more or less symmetrical organization of objects in the room with reference to each speaker.

POSITION LOCATION SYSTEM

The techniques disclosed in this specification may also be applied according to another aspect of the invention to position location for accurately locating the position of an object in two or three dimensions.

For example, the position of a robot arm may be determined by using similar techniques as schematically illustrated in Figure 15. A transponder 45 can be carried by a workpiece or other object and can be positionally located relative to respective transmitter/receiver units 41, 42 and 43 fixed relative to the robot. Alternatively, the transponder could be carried by the robot arm itself so that the exact position of the arm relative to transmitter/receiver units fixed relative to a stationary member of the robot may be determined.

In order to have a conyinuous indication of the object position with reference to fixed coordinates, three pairs of transmitter and receiver (e.g. interrogator) will normally be provided. Each pair will be located at a different point and will interact with the transponder i.e. will transmit signals thereto and receive response signals therefrom. By timing the signal transit times, the distances from respective interrogators may be computed.

Geometrically, the distance from each interrogator to the transponder represents the radius of a sphere centered at the interrogator position. Thus, three spherical surfaces are defined by the data obtained by the three interrogators.

These three surfaces will have two common points: one at the transponder location and the other outside the area being considered. Thus the transponder position may be unambiguously determined.

Where the transponder is not on the arm, its position with reference to the robot arm will be continually known to the robot processing units because of the following:

A) The common point outside the space being considered will be outside of the calculation range, as mentioned above.

B) Each point of the robot arm position with reference to each point of the robot body will be continually known to the robot processing units.

C) The distance from each point of the robot body to each of the three interrogator locations will be known to the robot processing units. These three interrogators can in this embodiment be considered as an integral part of the robot body.

The transponder will preferably not be located on the object, but rather it will be located at a fixed location with reference to the object.

The object position with reference to the robot arm will thus be continually known to the robot processing units.

Controlling the arm position with reference to the object position may be effected as a function of:

A) The task which was specified.

B) The transponder position.

C) The interrogator location.

An example of the application of this technique will now be discussed with reference to Figure 15.

Figure 15 schematically represents a production room for assembly of an electrical circuit board, or for producing some mechanical structure, such as an automobile.

The three interrogators 41, 42 and 43 are here located for example in three different wall/ceiling corners, where they would have "line-of-sight vision" of the transponder. They are electrically connected to a control and processing circuit which will be described with reference to Figure 16.

The transponder 45 will be fixed with reference to the object to be located (which may be a workpiece or the robot arm) .

While the production line is running, when the task for one object is completed, the associated transponder 45 (which is located with reference to this object) will be

disabled and a further transponder (not illustrated) which is located with reference to the next object will be enabled, and so on until the production line is completed.

When there is more than one production line in the production room, one timing processing unit will alternate between the lines. Alternatively, two separate systems could transmit in two different frequency bands, and two sets of interrogators with associated timing circuitry could be employed.

Figures 16, 17, 18, 19 and 20 illustrate details of a hardware implementation for the system illustrated schematically in Figure 15.

Figure 16 shows a time measuring system 51 for three distance measuring channels and one scaling channel. The construction of system will be substantially identical to the construction of block 810 of Figure 11, except of course that two additional channels are provided. It is therefore believed that detailed description of block 51 is unnecessary. The output from the timer 51 is supplied to a distance computing circuit 52 which is arranged to compute the distance of the relevant interrogator from the transponder as will be explained in more detail hereinafter. This process is effected for each channel and results in three signals X, Y and Z representing the coordinates in three dimensions of the object. These coordinate signals may be supplied to the processing circuitry of the robot. The computing circuit is controlled by a control circuit 63 which also controls the timing of the time measuring system 51 and the transmit and receive timing.

Details of the interrogators 41 to 43 and transponder 45 are similar to the details of the interrogator and transponder already discussed with reference to Figures 5a and 6, so that further discussion is omitted in the interests of brevity.

Figure 17 illustrates details of the control unit 63 of Figure 16. As shown, it comprises three D-type flip-flops 631, 632 and 633 in addition to three AND gates 634, 635 and 636. The flip-flops each receive the START (CYCLE) signal on respective clock inputs CLK and receive a reset signal REST on their respective clear inputs. The respective Q outputs of the flip-flops are connected to respective AND gates 634, 635 and 636 and carry signals ENl, EN2 and EN3 respectively. The signal START (CYCLE) is applied to a further input of each AND gate and the outputs of the AND gates provide respective signals LTCHl, LTCH2 and LTCH3. The timings of these signals are illustrated clearly in Figure 23. It will be observed that in each cycle defined by the signal CYCLE, the data read out from the latch is relevant for the preceding measuring cycle. The timing circuitry of Figure 17 ensures that the correct latch pulse is produced to read the correct data in each cycle. The signals ENl to EN3 are arranged to close the correct one of the switches Si to S3 to enable "one cycle measurement" from the relevant interrogator.

The time measuring block 51 is connected by lines TRNSl and RCV2 to the respective transmitters and receivers of the three interrogators via the switches Si to S3 so that operation of the switches by means of the signals ENl to EN3 connects the timer to the correct interrogator in each cycle.

The reset circuit ensures that the timing and measurement operations are properly coordinated so that the data will be related to the signals LTCHl to LTCH3 in such manner that the data latched with signal LTCHl will be the correct data for distance Ll, and similarly the remaining signals and data are correctly associated with the distances L2 and L3.

The reset circuit also is preferably arranged to cause the scaling function to be operated once upon initialisation. During scaling, the three main interrogators are disconnected from the timing circuit, whilst the fourth interrogator is connected. This condition remains until completion of scaling, i,e. until the scaling data has been latched. The system then measures times for the three distances in the following sequence: Ll, L2, L3, Ll, etc.

Reference will now be made to Figure 18 which illustrates a further part of the control unit 63 for controlling the timing of the scaling interrogator 44 which cooperates with a scaling transponder 45a at a known distance from interrogator 44. A D-type flip-flop 640 receives an initial reset signal RES at its clock input and provides the signal REST at its Q output (cf. Figure 17) which is supplied to one input of an AND gate 641. The output of the AND gate 641 supplies a signal LTCH4 to the computing unit 52 as shown in Figure 16 and also to the clock input CLK of a counter 642 which is arranged to produce a carry output at a count of two. The carry output of the counter 642 is effective to clear the flip-flop 640 at its CLR input and is supplied via an amplifier 643 and a capacitor 644 to the clear input CLR of the counter 642. As usual, the output electrode of the capacitor is connected to ground through a resistor 645 and a diode 646. Thus the counter 642 clears itself after each counting operation.

The computing unit 52 may be implemented in various ways, and two methods will be described here. A hardware implementation will be described with reference to Figure 19 and a software implementation will be explained with reference to Figures 20, 21 and 22.

Turning first to Figure 19, unit 52 comprises three functional blocks: a scale factor computing circuit 521, a distance computing circuit 522 and a position computing circuit 523. The circuit 521 is connected to receive timing data on line 524 and the signal LTCH4 from AND gate 641 as explained with reference to Figure 18. From the known distance between the scaling interrogator 44 and the scaling transponder 45a, the circuit 521 computes from the timing data a scaling factor V which is in effect a measure of the local velocity of the ultrasound being employed. Data representing this factor is supplied to the distance computing circuit 522, which also receives data representing the timed intervals for each of the interrogators 41, 42 and 43 from the timer 51 and receives the system timing signal CYCLE for synchronisation purposes. Circuit 522 supplies distance information for each of the interrogators in turn to the positioning circuit 523 on line 525 in synchronism with respective latching signals LTCHl, LTCH2 and LTCH3. Having received data for three distances, the circuit 523 computes the position of the transponder from this data and supplies signals X, Y and Z representing the coordinates of the transponder 45. The computation methods for the distance computing circuit 522, the scaling circuit 521 and the position computing circuit 523 present no particular difficulty and will be clear to those skilled in the art, since the computation of various mathematical functions using logic circuits is a well-known technique. Further

detail of the computation process will therefore be omitted to avoid obscuring the main principles here disclosed. The mathematics of the position computation will however be discussed hereinafter.

Reference will now be made to Figure 20 which shows a microcomputer 530 for carrying out the functions required for the computing unit 52 of Figure 16. As is conventional, the computer will comprise a microprocessor, a ROM for storing an operating program, a RAM for working storage and an input/output circuit for interfacing the microprocessor with the timer 51, the control unit 63 and any peripheral apparatus such as a robot controller. Any suitable microprocessor may of course be used, but it is suggested that the INTEL 80286 sixteen bit processor would be appropriate. The computer 530 is supplied with an 8 MHz clock signal on line 531 and with data defining the maximum possible values X and Y of the x and y coordinates of the transponder (a maximum value Z for the z coordinate may also be supplied if desired but is not necessary in this implementation) . _ In addition of course the data line 524, the CYCLE signal and timing signals LTCHl, LTCH2, LTCH3 and LTCH4 are supplied to the computer.

Figure 21 shows the main routine stored in the ROM of the computer 530. Following system reset (step 540), the processor initialises itself (step 541) and then performs the scaling function on the basis of the scaling data derived from the scaling interrogator 44 (step 542). On the basis of the computed scale factor and the time data generated by the timer circuit 51, the processor then computes the coordinates (x,y,z) of the transponder 45 in step 543. The routine then advances to a decision step 544 and awaits an interrupt. Upon the receipt of an interrupt, the routine returns to step 543 and computes

the coordinates again on the basis of the next set of timer data.

Figure 22 illustrates the step 543 of Figure 21 in more detail. Upon receipt of an interrupt signal (CYCLE) at step 550, the program advances to an initialisation step 551 and then initiates a data reading step 552 in which timer data is read from the timer 51 into predetermined storage locations in RAM. Having collected the data for each interrogator, the processor next performs a distance computation step 553 to determine the distances Lτ_, L2 and L3 from the respective interrogation devices to the transponder. When the distances have been computed, steps 554, 555 and 556 are performed to compute the respective position coordinates of the transponder. The processor then returns to the main routine.

Further details of the mathematics of the position computation process will now be described with reference to Figure 15. The transponder 45 is located at position (x,y,z) in a system of Cartesian coordinates having its origin at one corner of a parallelepipedal space in which the transponder is located. The three interrogators 41, 42 and 43 are positioned at corners of the space at distances Ll, L2 and L3 from respective interrogators 41,42 and 43. If the coordinates of the corner of the space diametrically opposite the origin are (X,Y,Z), interrogator 41 is at position (0,Y,0), interrogator 42 is at the origin (0,0,0), and interrogator 43 is at position (X,0,0). It will be understood that the following relationships apply:

Li ² = _x 2 _{+ (} y_ _yι) 2 + _z 2 ₍ T

L2 = x ² + y ² + z ² (2)

L3 ² = (x-Xχ) ² + y ² + z ² (3)

From these relations, it can easily be shown that:

x = (X ² + L ₂ ² - L ₃ ² )/2X (4) y = (Y ² + L ₂ ² - Lτ_ ² )/2Y (5) z ² = L2 ² - χ ² -y ² (6)

It should of course be borne in mind that the desired values of x, y and z are constrained to lie within the defined space, which means that negative values are excluded, as are any values greater than X, Y and Z respectively. This enables the values of x, y and z to be unambiguously computed.

The preferred implementation for the position computation will be by means of the arrangement described with reference to Figure 20 using an INTEL 16 bit processor 80286. A 32 bit processor such as the INTEL 80386 would operate four times faster if required for any particular application.

For illustrative purposes, it will be assumed that the maximum distance to be measured is 20 meters and that an accuracy of 0.1mm is required. This requires a calculation ratio of 0.1/2.10 ⁴ , i.e. 1:200,000. To perform the necessary calculations in one operation, at least an 18 bit processor would be required, since 2------ =

264,000. Therefore, the calculations would have to be done in two stages with a 16 bit processor.

Using an 8 MHz clock, the approximate computing times for addition and subtraction would be 0.5 microseconds, the approximate times for multiplication and division would be about 7 microseconds, and the operation of taking a square root would occupy about 50 microseconds. The square root

operation can be performed either by means of a polynomial approximation or, if this is inappropriate for a particular application, by use of a co-processor such as the INTEL 80287.

In order to compute the x coordinate, four multiplication and division actions and three additions and subtractions are necessary, making a total of 36.5 microseconds. To compute the y coordinate, a further 36.5 microseconds are similarly required. For the z coordinate, three multiplications, two subtractions and one square-rooting operation are required, making a total of 71 microseconds.

The total computing time required is therefore (roughly) 144 microseconds.

In order to determine the time available for the computation, it is necessary to consider both the speed of the transponder (if it is moving) and the minimum transit time for the pressure waves. If the assumption is made that the transponder speed will be small compared with the speed of the waves, the possible motion of the transponder may be disregarded. If the minimum measurement distance is taken as one meter, and the speed of the waves is taken as 350 meters per second, the minimum transit time is 2800 microseconds, which is roughly twenty times the required computation time. It will therefore be apparent that a 16 bit processor is quite adequate for the purpose, and that much shorter distances may be measured and/or the accuracy of measurement may be considerably better than that assumed in the above.

Another possible application for the techniques described in this specification according to another aspect of the invention is the control of farm machinery apparatus, such

as a plough during ploughing operations in an open field. Figure 24 illustrates schematically how this could be achieved. By placing a transponder on the plough and positioning three interrogators TR]_, R2 and TR3 at corners of the field, the position of the plough may be accurately determined and therefore controlled. (Only two interrogators will be required to determine the position; the third may be used for scaling or omitted since extreme accuracy is not essential in this application) .

A further possible application would be for locating the submerged position of an underwater object, such as a remotely controlled submarine craft. Such a craft may be equipped with surveying equipment for surveying the sea bed, and the present invention could be employed to track the position of the craft and to ensure that it followed a predetermined course or search pattern. in this case, the transponder would be carried by the submerged craft and the three interrogators would be placed at various known locations which could be on the surface of the sea or submerged. Where the invention is applied to use under water, naturally account must be taken of the fact that the velocity of pressure waves is different in water from that in air. Also, it will be apparent that infrared signals would not be practical in this application, so that bidirectional ultrasonics would be employed.

It is also possible to control an object which is moving in three dimensions according to signals from different directions, such as from other moving objects.

Three interrogators may be provided for generating distance information in order to find th.e object position, and according to the distances from the other two moving objects, the object position can be changed. It would of

course be possible to measure also the angle to the other moving objects by different methods

When working with a moving transponder, there are three points which should be considered:

A) The object velocity with reference to the velocity of the pressure waves will have to be taken into account to improve the accuracy and to achieve proper timing.

B) When the transponder is moving, the transmitted frequency will change as a result of the Doppler effect.

C) It is also possible to allow the interrogators to move with respect to the transponder. Both the above points A and B will then need to be considered. .

ACCURACY CONSIDERATIONS

The general question of the accuracy of the methods discussed above will now be considered. Naturally, accuracy is more important when considering position location systems, but in principle the following discussion applies also to the multi-phonic balancing techniques discussed above.

Generally speaking, the accuracy that can be obtained depends upon the wavelength of the ultrasonic waves used, especially when embodiments employing bidirectional ultrasonic transmission are employed. The wavelength is of course equal to the velocity of the waves divided by their frequency. The velocity changes according to the environment and depends upon the mass density of the medium carrying the waves.

When ultrasonic transmissions are exclusively employed, i.e. in both directions, the maximum accuracy which can be achieved is of the order of half the wavelength. A frequency of 40 kHz will give an accuracy of about 4.15 mm, a frequency of 60 kHz will give an accuracy of about 2.77 mm, and a frequency of 100 kHz will give an accuracy of about 1.69 mm, etc.

For the robotic application, in order to improve the accuracy it would be necessary to increase the frequency. Using for example a frequency of 1 MHz, an accuracy of about 0.17 mm can be obtained, whilst at 500 kHz an accuracy of about 0.33 mm is possible.

In order to achieve optimum results, it is necessary to consider at least four problems:

A) The point of the cycle of the transmission wave at which the wave starts out from the interrogator to the transponder must be exactly predetermined so that the transponder knows in advance what part of the cycle to expect first upon detection.

B) The receiving circuit in the interrogator takes a finite time DTi to respond following reception.

C) The transponder takes a finite time DT2 to respond following reception of a signal.

D) It is not possible to synchronise the transmission clock generator of the interrogator with the clock generator of the transponder.

Problem A can be solved by detecting a predetermined point, such as a rising zero transition, of the sinusoidal signal for driving the transmitter and ensuring that the transmission starts at this fixed point of the sinusoidal signal.

As regards problem B and C, the error can be reduced to negligible proportions by use of suitable circuitry, such as that disclosed in this specification. However, the errors -rDT _] _ and +DT2 can always be taken into account because they will be exactly known in advance.

Problem D can be reduced by half by sensing the rising and falling edges of the transmitted wave so that the total loss of accuracy is no more than about half a wave period.

As a result of these considerations, it is clear that the total error E = (+L/2V) + (-.DT].) + (+DT2), where L and V are the wavelength and velocity of the response wave. It will be appreciated that all the errors will always be positive, since they are all such as to increase the measured time.

Two possible measurement methods are suggested:

1. one transmission ultrasonic, one transmission electromagnetic, e.g. infrared.

2. Both transmissions ultrasonic.

In method 1, the factor +L/2V is negligible compared with +DT]_ and +DT2, so that the total error will about (+DTτ_) + ⁽ +DT2) to a first approximation. For many applications, this error will be quite acceptable and in fact the accuracy in this method is virtually dependent solely on the accuracy of the timing technique used to time the transit time. In practice, this means that the accuracy is dependent upon the frequency of a clock generator employed to drive a counter for timing the transit time, the possible error being positive or negative. it is unnecessary to take account of the transit time for the electromagnetic wave because its velocity is greater than that of the ultrasonic wave by a factor of about 10^. Thus, accuracy may be improved, if necessary, simply by increasing the clock frequency.

ln method 2, the factors *DTτ_ and +D 2 are negligible compared with +L/2V, so that the total error is approximately +L/2V and may be reduced by increasing the frequency of the response wave. Unfortunately, this will also increase the transmission loss since this is proportional to the logarithm of the frequency. Therefore, in any particular case, the frequency of the response wave will have to be selected in accordance with these conflicting requirements.

SYSTEM SCALING AND CALIBRATION

When it is desired to provide a measure of actual distance, (this not being necessary for the stereo balancing application), it may be necessary to calibrate the apparatus if optimum accuracy is required.

Since the velocity of pressure waves varies as a function of barometric pressure, humidity, etc., system scaling or calibration may become necessary. Such scaling will be more important where the robotic application is implemented outdoors. In a closed space such as a production room the velocity changes will be insignificant. If necessary, scaling may be carried out by sending a transmission over an accurately known distance and measuring the time taken. Since the distance is known, the velocity may be calculated. Such scaling techniques have been described hereinbefore with reference to Figures 16 to 22.

REFLECTIONS AS A RESULT OF THE ULTRASONIC WAVE

Reflections of the ultra-sound waves from surrounding objects have to be considered.

It has already been mentioned in connection with the multi-phonic balancer that the reflections effect must be taken into account for the audio signal, that we can control the volume according to the distance changes bearing in mind that the intensity P varies in proportion to 1/D ² , that for this purpose we will have to create a medium whose behaviour would be uniform, such as isotropic medium, and that the way to achieve this would be to carpet the room very well, to use an acoustic ceiling and to put curtains on the walls all around, so that in this way the reflections would be negligible.

Especially in the position sensing and robotics applications however, the reflections as a result of the ultrasonic waves will have to be reduced to a minimum in order to avoid incorrect measurement results.

However, for the measurement technique we are using, the crucial measurement time is the interval from the actual arrival time of the wave at the transponder until a few (about 4) wave periods have elapsed. This is because by that time no direct receiving for this measurement by the transponder would be possible, so that any signals that are received may be attributed to reflections. in fact, the transponder will recognize the received frequency from the first wave to arrive. This will be discussed in more detail hereinafter.

When the transponder transmits ultrasonic waves as a reaction to the reception, any reflections as a result of the first transmission will normally be negligible, and this is why the same frequency may be used for both transmissions. However, in order that the reflections of the first transmission would die out fast, we still have

to try to create an isotropic medium. (In this way we will have less noise and more reliability and the system can be faster). As mentioned above, the most important time to avoid reflections would be the time of the first few waves because this is the critical interval during which detection occurs, and this will function more reliably without interference from reflections.

In the case of robotics applications, when the, frequency is 1 MHz, the wave length is about 0.35 mm. Therefore, a distance of 1.75 mm, or about 2 mm, corresponds to five wavelengths.

An non-reflecting area of at least 2 mm width should therefore be provided around each transmitter. In practice, a surrounding area of about 1 cm width may as well be provided since it does not require much extra effort or expense.

In the case of audio applications, when the frequency is 100 (kHz), the wavelength will be about 3.5 mm and five wavelengths represents a distance of about 1.75 cm.

In this case, it could be stated that the listener will have to make sure that reflections from around the transponder within a range of 2-5 cm will be avoided. In practice, this means not locating the transponder in a position surrounded by a good reflector, or possibly supplying a piece of unreflecting material hooked around the transponder. Figure 25 indicates a possible arrangement satisfying these requirements. An interrogator 70 is surrounded by a collar 71 of non-reflecting material. It will be appreciated that reflections could occur both in the vicinity of the source of waves and in the vicinity of the receiver. Both types

of reflection are to be avoided and therefore the collar will be applied to both the transponder and the interrogators.

It should perhaps be mentioned again at this point that most of the reflection problems can be avoided if the transponder is arranged to respond with an electromagnetic wave, e.g. an infrared wave, instead of an ultrasonic wave.

DISTANCE AND SENSITIVITY FOR THE ULTRA-SOUND WAVE

The transmission loss for ultra-sound wave is given by the following relationship:

^TL (dB) ⁼ 01ogF + 201ogD ^• *• ,

where TL is the transmission loss, F is the frequency, D is the distance travelled and K is a constant depending on the specific surface density of the environment.

From this relation, the required transmission power can be derived as follows:

TL ^ (Transmission power)^ - (Receiving sensitivity)a _B .

Various types of ultrasonic transducers are commercially available, such as those available from International Specialists Inc and referred to as the Pulse Transit (PT) type which includes a transmitter and a receiver and can operate over a frequency range of from 20kHz to 60kHz. Transducers with various case diameters, e.g. from 12 mm to 24 mm, may be obtained.

Many modifications will occur to those skilled in the art and it is intended that all such modifications are included within the scope of the present invention as

defined by the appended claims. It will only be mentioned by way of illustration that besides arrangements with one transponder and 2 or 3 interrogators, it is also possible to envisage arrangements with two transponders and two interrogators, or three transponders and one interrogator.

It would need to be borne in mind in this connection that when employing more than one transponder the number of frequency bands will have to at least equal the number of transponders. It is also possible to operate in parallel during the measuring intervals. For example, three different signals in differing bands could be transmitted simultaneously to a single transponder, if the transponder were constructed to receive and transmit in the three bands.