NO180763B - Circuit for use in processing reflected signals - Google Patents

Description

The present invention relates to a circuit for use in detecting characteristics of the seabed on the basis of a return echo signal from a hydroacoustic transmitter-receiver device, the circuit comprising means for defining a time slot for extracting from the return echo signal a first component which forms part of the first bottom echo component , means for defining a time slot for extracting from the return echo signal a second component which constitutes the second, or a later, bottom echo component, as well as means for obtaining a first signal from the extracted first component.

The term transmitter-receiver equipment shall in the present context mean equipment for sending pulses and for receiving the resulting echo of the sent pulses, and is, unless otherwise mentioned, not limited to the type of pulse (which can, for example, be sound, ultrasound or electro-magnetic pulses) and also not limited to equipment that has common transmitter and receiver parts.

In the area of sound and ultrasound waves, there is equipment available, such as sonar and sonar, which give audible and visible indications that return echoes exist. Sonar, for example, gives results that are written on registration paper.

In order to achieve an improvement of such equipment, circuitry has been designed for processing the return echo signal in order to derive more information from it than has hitherto been readily available with this group of equipment. Such a circuit arrangement is shown in GB-2 102 573, where means are arranged to extract the second bottom echo component from the return echo signal in order to obtain information about the nature of the seabed. In the aforementioned patent document, it is shown that the unique nature of the seabed was revealed by recording whether the second bottom echo component had an amplitude above a determined

threshold value.

From GB-2 102 573, a circuit of the type indicated at the outset is thus previously known, and comprising devices for defining a time slot which serves to detect a first bottom echo component (primary echo), and devices for defining a time slot for detection of another bottom echo component (secondary echo).

It has now been shown that there is further useful information in the return echo of equipment, such as sonar, and that this information can also be found in return echoes from sonar, radar, medical ultrasound equipment and transmitter-receiver devices in general.

In particular, it has been shown that the information is not only in the timing and size of a return pulse, but also in its form. Due to saturation effects, such information is not normally available in the return echo of high-power equipment. When lower transmitted power is used, it has now been shown that return echoes have shapes that depend on the characteristics of the reflected surface or surfaces, for example seabed layers, which are of interest in mapping and underground profiling.

According to the present invention, a circuit of the kind indicated at the outset is thus provided, characterized by means for integrating the extracted second component to provide a second signal which gives an indication of the energy content of the second component, as well as means for providing a display which represent the individual values of the first and second signal.

The means for obtaining the first signal preferably comprise an integration device.

It has been shown that the area under a return pulse edge carries information that is characteristic of the nature of the body that reflects the transmitted signal, and that this area can be displayed on an analogue meter or digitally, so that the user can see a value that changes noticeably with the nature of the reflected body. By e.g. echo sounder, it has been shown that the leading edge of the bottom pulse (first, second or later, but especially the first) varies depending on the nature of the material near and on the seabed, and especially depending on the nature of the upper layers of the seabed. Thus, a very steep rising (front) edge gives a relatively low value of the representative signal and indicates a hard seabed without significant mud. On the other hand, a relatively high displayed value of the area under the leading edge will indicate that there is a layer of mud on the seabed (useful information for shrimp fishermen and when dredging and underwater mapping), or that there are fish near the bottom. Shown variations of the value are also of interest. It is believed that such an analysis can also be used in medical ultrasound use to determine differences in the characteristics of different tumors or other features found in human or animal bodies, especially in the eyes.

A reflected pulse would theoretically have a shape that depends almost exclusively on the nature of the receiving equipment, because one would think that the rise time of the received pulse would mainly depend on the response time (transient response) of the receiving equipment. Integration or other analysis of the leading edge would therefore only provide information about the receiving equipment. It has been shown that this is not the case. When using a standard sonar with a transmitted pulse width of the order of 300 microseconds, the received pulse from the bottom has a shape that depends on the characteristics of the upper layers of the seabed at a leading edge which can even be from one to two milliseconds. Possibly the return pulse, as it occurs in the sonar, is a summarized form of several reflections and scattering from different depths of the seabed layers. This would explain the fact that the leading edge provides information about these layers. A leading edge that is only a transient rise time would probably only last a hundred microseconds or so, not one or two milliseconds. As described below, the time slot is synchronized with the returning signals in such a way that it is not fixed, but is automatically varied in accordance with the length and shape of the leading edge. When the leading edge varies in length and shape with the characteristics of the seabed, the position and length of the time slot can thus vary. Using a fixed, short time slot would probably often only lead to the capture of a transient response.

If equipment is used that has a sufficiently short, transmitted pulse and a short transient response, it could probably happen that one could not see a smooth leading edge over the time slot, such as the one we are dealing with. One could probably see a series of spikes in that time slot, caused by wave fronts from successive areas under the seabed and from spreading. One could smooth or integrate that signal to produce the same kind of leading edge that would be formed by a conventional sonar. The smoothed signal could then be used to define the time slot used for capturing the "leading edge", but in this case one would in reality capture a series of leading edges.

The time slot can be defined in different ways. For examination of the leading edge of an echo pulse, the trailing edge of the time slot can thus be determined using level registration for determining the trailing edge in dependence on the level of the return echo signal, as it passes through a given level. If one takes a rising level for determining the time slot, in one embodiment a delay (small compared to a typical echo pulse) can be introduced to cause the end of the time slot to be more closely approximated to the tip of the relevant echo pulse. In this way, the time slot is automatically

mathematically adapted to the particular nature of the received pulse.

A smoothing device may be required to smooth tips at the leading edge to prevent response of the level detection device to such tips.

The leading edge of the time slot can be determined with the help of organs that react to a signal that is representative of the transmitted pulse, so that said end occurs at a given time after the occurrence of the transmitted pulse. The given time can be made adjustable, so that the user can extend or shorten the time slot as needed, e.g. to increase or decrease the effects of fish that had to swim close to the bottom of the sea on the result. But this is not always satisfactory. In a depth sounder, the echo of interest may for example be the bottom echo, but it moves in relation to the transmitted pulse when the seabed rises and falls. The time slot would thus change with rise and fall and thus distort the reading. In an example of use, the beginning of each time slot can therefore be set to occur at a given, adjustable time after the occurrence of the echo pulse in the previous cycle of the return echo signal, i.e. so that it is at an adjustable "height" above the seabed.

The principles outlined above can be applied generally in echo systems, whether the signals are acoustic or electro-magnetic (as in maritime radar) or whether the signal is projected vertically (as in sonar) or generally horizontally as in sonar and maritime radar. In all such cases, one can ensure that the time slot is synchronized with or connects to a particular echo pulse, whether it is representative of the seabed, a school of fish or another body in or floating on the water. A visual display can then be produced for the user on a meter of a distinctive feature at the leading or trailing edge of this echo pulse, and thus the user can obtain more information about the echo source than is normally available with conventional radar, sonar or sonar display.

In a typical application, the transmitter will send pulses periodically so that cycles of return echo signals are received. In this case, there are preferably means for keeping the value of said representative signal for display, when said time slot in a subsequent cycle cannot be determined as a result of poor signal reception. In this case, an indication given to the user of the value of the characteristic will not be seen to vary uncontrollably, should it result in bad echo signals, because in this case the user will be presented with the last, correctly determined value. This feature can remove problems that can arise as a result of bad weather conditions that interfere with the sending of a pulse or the reception of a return echo. As a continuation, the holding device itself can form an average, e.g. to emit a signal for display which is the average of several, e.g. three of four previous values. This is useful when the conditions are such that the display would otherwise oscillate in an unacceptable way.

In such processing of return echo signals which are formed as a result of sending a series of pulses, it may be appropriate to have timing means which respond to an echo pulse with a given property in the echo signals in each cycle of return echo signals, to determine a time period which is synchronized with respect to the said echo pulses, means for storing a special feature of the return echo signals, means for displaying the stored special feature at any time, sampling means for controlling the supply of data to the storage means, depending on said time period, so that the stored special feature is representative of the part of the return echo signals in a given time slot of the cycles, and means that react to the absence of said echo pulse in a cycle of return echo signals for controlling the sampling means to prevent sampling and thus prevent updating of the storage device.

In particular, a favorable result will be obtained by the use of a time slot to extract a desired part of a return echo together with means for processing the extracted part to obtain a signal representative of the mean value of the part in the time slot. Application of this idea may be appropriate in connection with pelagic fishing, where the time slot determines an area at a depth determined by the manually selected timing of the time slot. The resulting signal determines the level of the echo from this area, and increases when fish are present.

It has been shown that there is a further application of this idea for improving information extraction in forward-looking transmitter-receiver devices, i.e. in conventional sonar and maritime radar direction and/or range finder equipment. Such equipment conventionally provides an image of a cathode ray tube showing all return echoes of sufficient strength, including noise signals. In order to improve the effect of such equipment, attempts have been made to incorporate filters to filter out the noise content in such signals. It is nevertheless difficult to register small boats, for example, due to the level of noise and sea slips in the surroundings.

In this context, it may also be beneficial to provide equipment comprising a transmitter for sending a series of acoustic pulses at ultrasonic frequency in a direction with a horizontal component (and perhaps even with a small upward component), and a receiver for to process the return echo signals resulting from reflection of the transmitted pulses, the receiver comprising a circuit having: means for determining a time slot for extracting from the return echo signals the signals coming from a reflection within a selected distance; and means for averaging the extracted signal in each repetition. One can therefore visually or automatically compare the mean value with a preceding mean value or several preceding mean values, in order to get an indication of a change in the mean value which represents a change in data in the noise and thus represents possible useful information within the selected area. An alarm can be triggered by an appropriate change level.

This can be combined with features already discussed above, i.e. a time slot together with averaging or integration, but in this case the time slot is not necessarily connected to a specific echo pulse, but is set manually in terms of its area and its duration for to "look" at a specific zone that has a horizontal distance from the vessel carrying the equipment, i.e. the axis of the beam (or main loop) of transmitted pulses is not vertical. By looking at changes in the "noise", rather than trying to filter out the noise, data of interest can be recorded, although it is not necessarily visible on a conventional cathode ray tube or printer. Thus, a school of fish would cause an increase in the "noise" level. In another application, this equipment is used to detect objects on or near the surface that could damage a vessel. An audible alarm gives a signal every time the received echo within the relevant area rises above a certain value (e.g. in amplitude or in integral value) set by the user.

In this connection, a circuit can also be used for recording the characteristics of the seabed on the basis of return echoes from a hydro-acoustic transmitter-receiver device, where the circuit includes means for determining a time slot for extracting one of the bottom echo components from the return echo, and means for processing the extracted component to form a width signal when the width of the extracted component exceeds an adjustable value.

This aspect is particularly applicable to the second bottom echo component, where it has been shown that the width of the component is a property that is directly related to the hardness of the seabed. The resulting signal can thus be fed to an alarm device to alert when the signal represents the achievement of sufficient hardness of the seabed to damage a fishing net. Patent document GB-2 102 573 deals with the formation of a warning signal which depends on the size of the second bottom echo component and this idea is preferably incorporated in this aspect of the invention by reacting to width only when the second bottom echo has at least one given size . A further safety measure can be taken in that a warning signal is only sent out when the threshold and width criteria are met by two (or more) successive signals, so that a series of random responses is avoided. This aspect of the invention can also be applied to a third or later bottom echo.

In this connection and with regard to the leading edge analysis discussed earlier, it has been shown that the first bottom echo component can be exposed to noise and distortion especially in connection with bad weather conditions, while other subsequent bottom echo components are relatively undisturbed and thus can be more useful for leading edge analysis and also for seabed analysis as discussed under the further aspect of the invention.

An alternative bottom alarm can be based on an integration of the second or a later bottom echo, and then in particular in connection with means for determining a time slot for extracting a second or a later bottom echo component from the return echo signal and means for integrating the extracted component to obtain a indication of its energy content.

In order to make it easier to understand the invention and how it can be carried out, reference is made to the drawing, which illustrates the invention with some examples of execution, and where

Figure 1 is a general block diagram of a processing circuit unit for echo pulses,

figure 2 shows a series of waveforms that illustrate how the circuits according to figure 1 work,

figure 3 shows more detailed waveforms at a longer time scale,

figure 4a-c show return echo waveforms,

figure 5 is a diagram of a combined meter and alarm circuit,

figure 6 is a diagram of a bottom measurement circuit,

figure 7 shows a series of waveforms illustrating operation of the circuit according to figure 6,

figure 8 is a circuit diagram of a bottom alarm,

figure 9 shows waveforms in the bottom alarm circuit,

Figure 10 shows a circuit of a second bottom echo timing circuit device,

figure 11 shows a circuit which is a modification of a bottom-target circuit,

figure 12 is a circuit diagram of a fish registration circuit, and

Figures 13-16 show diagrams illustrating the operation of the circuit according to Figure 12.

Figure 1 shows a circuit device which is specially designed for processing the return echo signals A from a conventional sonar S with transmitter/receiver transducer T. The analogue return echo signal is, after a time-varying amplification (TVF), fed to input 1, and this signal form is generally shown at line A in figure 2. It is a periodic signal which at each repetition comprises a large transmission component TC, a first bottom echo component E1 and a second bottom echo component E2 which all have successively decreasing amplitude. Between the transmission component and the first echo component, a fish echo can occur as shown by EF. Various noise and other echo signals may also be present, while their magnitudes are of no importance in the present context and they are not shown in Figure 2. If the circuit were used in connection with a forward-looking sonar or maritime radar, a cyclical pulse pattern, but the shape of the signal would be different, but would nevertheless include pulses such as those shown at E1 when the surroundings include a reflective "body", such as a boat or school of fish.

Even when using a conventional sonar TVF, some distortion or saturation may appear in certain objects, and it may therefore be advantageous to provide an alternative TVF circuit to overcome saturation and make more information available. A microprocessor with a mathematical algorithm could, for example, take care of the TVF process.

The sonar S also forms a pulse TX at an input 2 in figure 1. This pulse indicates the time when the transmission pulse is sent out from the transducer T. From input 2 the pulse TX goes to a Schmitt trigger circuit 3 for transformation and then to a monostable circuit 4 which provides a stretched pulse B (line B in Figure 2), which is used in the circuitry as a mask to filter out the TX component of the echo signal and acoustic noise near the transmitter.

The analog signal that is fed to input 1 is applied to three analog amplifiers 5, 6 and 7, which have gains that can be varied by the user with variable resistors 8, 9 and 10.

The amplifiers have capacitors 11, 12 and 13, which are connected across the amplifier outputs and are connected to Schmitt trigger circuits 14a, 14b and 14c. The capacitors provide smoothing to smooth the leading edges of echoes in an attempt to avoid the leading edge detection process to be discussed being falsified by distorted leading edges. The circuit 14c feeds a fish alarm circuit 15, which will be described in more detail later, via a NOR gate 16, which is fed by signal B for blanking transmission pulses. Circuit 14b feeds a second bottom echo alarm 17 via NELLER port 18, which also receives signal B. The second bottom echo alarm 17 can be as described in GB-PS no. 8 221 670

(2,102,573). The amplifier 6 also feeds its analogue signal (corresponding to A in Figures 2 and 3) without smoothing to fish (19) and bottom (20) measuring circuits which will be discussed later.

The circuit 14c supplies a bottom pulse recording circuit and this will be discussed with reference to Figure 3, which shows more detailed waveforms over several cycles of the return echo signal A, including the signals B, C, D and E according to Figure 2. The Schmitt trigger circuit 14c emits at its output a digitized version of signal A, i.e. with its pulses reshaped into a series of square pulses of uniform amplitude. These pulses pass through a series of NOR gates 21, 22 and 23 and inverters 24 and 25 to a gate 36, whose Q output is signal F which is shown on line F in Figure 3. This signal F is a pulse starting from a given amplitude point on the rising edge of the first echo and ends at the beginning of signal B, fed to the reset input of the latch 26. The threshold of the rising edge of the echo signal E1 is selected by adjusting resistor 10. The periodic signal F is fed to a light-emitting diode (LED) at 27, which you will see light up during normal use of the equipment. The threshold will be increased during setup, until such information stops and then lowered again, until the diode just starts to light up again for setting a threshold for this channel which is suitable for recording a point high up on the rising edge of the current echo pulse. In order to prevent a response from the channel to other pulses, the transmission pulse itself is masked by feeding signal B to NELLER port 23 and other signals can be masked by applying additional signals via NELLER ports 21 and 22. Signal F is fed to a monostable circuit 28, which forms a 40 millisecond pulse G (line G in Figure 3). With a switch 29 in position B as indicated, pulse G is fed as an input to a monostable circuit 30 which at its Q output produces an indicator pulse having a duration determined by the user at the variable resistance 31.

The capacitors can also be connected in or out of the circuit with the circuit 30 by means of a part of the switch 29, for variation of the timing range of the adjustment created by the resistor 31. This pulse from the Q output of the monostable circuit 30 is signal H in figure 3 , and it will appear that it starts on the falling edge of pulse G and stops somewhat later, adjustable by the user, so that it comes the desired distance before the leading edge of the next first bottom echo pulse E1.

As described so far, it will be understood that several port switching signals are provided for determining areas where a sonar signal can be processed. These include signal G which determines the seabed, i.e. the bottom pulse, and signal H which determines a point at an adjustable height above the seabed.

Switch 29 provides other gate switching signals when in its second position P. In this case, a monostable circuit 30 is fed with signal TX after signal transformation from circuit 3, thus forming a version of signal H which determines an adjustable depth below the transducer . Similarly, a second monostable circuit 32 provides a second signal for determining a second adjustable depth below the transducer. Together, these signals form an adjustable range, e.g. for the fish alarm circuit 15 which will be discussed below.

Before discussing the details of the rest of the circuitry, the idea behind the fish and bottom measuring circuits must be mentioned. These circuits comprise integration or averaging means for averaging the areas under the return echo for specific time slots in each pulse repetition period, i.e. cycle. In the case of the fish measurement circuit, a time slot, as shown at C in Figure 2, can be used to determine a lower area of the sea, which ends at or near the top of the leading edge of bottom pulse El. The output goes to gauge 38, which will show a rising value when you move into an area with softer bottom and/or fish near the seabed.

In contrast, the bottom meter 20 will integrate or average over a time slot, to include essentially the entire bottom echo (first and/or second and/or third, etc.). When moving over harder bottom, the gauge 38' will indicate higher values.

If gauge 38 rises and gauge 38' falls, this will normally indicate a soft bottom. If gauge 38 rises and gauge 38' is constant, this will normally indicate fish. These gauges thus complement each other and are not only useful for anglers, but also for hydrographic mapping work.

A further mention of fish measurement district 19 can be found in the applicant's master application 85.1118 which is notified under 166 347.

From this it is clear that such a circuit does not just act like

a fish gauge, but also provides effective discrimination of bottom conditions, as "softer" bottom gives high readings on gauge 38 and "hard" bottom gives low readings. Pulse C can be delayed by adjusting a resistor to reduce the effect of fish. Figures 4a-4c show diagrams of seabed echoes at soft, medium, respectively. hard bottom. It will be obvious when integrated up to a given threshold, e.g. close to the maximum value on figure 4a-4c, that figure 4a will give a higher one

value than figure 4b or 4c. Why the waveforms have the shape they do is not entirely clear. A factor to be taken into account in any analysis is that the reflection from a point on the seabed is too simple a solution, because in practice the ultrasound beam will have an angle of up to approx. 30°, so that a detected signal includes reflections at different depths from normal incidence and also from scattering at abnormal incidence. The first part of the signal according to figure 4a will comprise low-level reflections from normal incidence when the wave front penetrates the "soft" bottom. Possibly increased reflections occur at a later time as a result of the material below the bottom being more condensed with depth. At the same time, these deeper reflections will have scattering from nearby areas added, so that the return echo is slowly built up as shown in Figure 4a. In the case of a "hard" bottom, reflection from the surface of the seabed, on the other hand, will essentially drown such secondary effects and give it a steep rise time according to figure 4c.

As dispersion is probably a significant factor and as it is frequency-dependent, it is noted that frequencies used in this connection are recommended to be in the range 30-100 kHz, e.g. 50-60 kHz. At a frequency as low as 20 kHz, it is conceivable that "soft" bottoms are not "seen", while "soft" bottoms at frequencies as high as 200 kHz can be perceived as hard.

It is also assumed that leading edge analysis will find other uses than those already mentioned in maritime sonar, especially in fishing and hydrography. A medical application for recording and distinguishing between tissues etc. in the human body is conceivable. A special application is the examination of the retina, when one thinks of the peculiarity of the eye which forms a boundary between the fluid in the eye and the retina. Such a boundary is analogous to the seabed.

The above description mentions the use of the aforementioned circuit for leading edge analysis of the first bottom echo component. It can also be used for leading edge analysis of another (or a later) bottom echo component.

A particular application is to prevent damage to yachts and other small boats at sea, which can often be caused by large, massive, floating objects, generally described as drifting goods. The list of objects is almost endless, but many of them can come from the ever-increasing pollution, e.g. containers, oil barrels, telegraph poles, etc.

Today's collision avoidance technology uses maritime radar warning devices. They are only effective when it comes to objects that provide a good reflection of the prevailing state of the sea. The circuit according to Figure 1 (and according to the parent application in Figure 4), as discussed so far and when it is switched to the P position, is intended to amplify signals received via a relatively inexpensive sonar system S, which has its transducer T arranged two or three feet below the surface of the water and has a fixed beam of about 30° inclined forward and towards the sea surface. An irradiated area is selected using the monostable circuits 30 and 22, corresponding to approx. 45.7-182.8 m from the vessel. A larger area would be preferable, but it is unlikely that the system would work over a greater distance without the use of far more sophisticated and thus more expensive equipment. The sample-hold technique would be used to integrate the total signals within the irradiated area, along with some noise filtering and an operator-controlled acoustic alarm set to a threshold that would replace the meter 38. If a large object such as is dangerous for the vessel's safety, enters the irradiated area, the integrated value will increase marginally but sufficiently to set off the alarm, and thus the skipper will be notified. Due to the relatively repetitive nature of swells and similar waves, it is possible to incorporate an appropriate filtering which can work effectively to isolate the necessary obstruction signal.

In the fish and bottom measurement circuits according to figure 1 (and figure 6 which will be discussed below), a gauge is used to generate a visual indication of the conditions. If desired, an acoustic alarm circuit can be added to both circuits, as shown in Figure 5. In each case, an alternative R2, C2 combination, denoted R2, is arranged at the output of the switch 35 (which is used in both the fish measuring device and the bottom measuring device/figure 6) '' and C2". Instead of feeding a meter, this additional circuit feeds a load resistor R3, across which is connected an acoustic alarm circuit (only shown at the fish meter 19 in Figure 5). An important condition is that the resistor R2 '' can be adjusted to give a different time constant than that set by resistors R2. Thus the meters may give readings averaged over, say, eight or more cycles, but this may give too slow a response for an alarm. This is resolved in that R2'' is adjusted to give an alarm in response to the mean value of, say, two cycles.

R2'<1> can be buffered from R2 if necessary. The following components are placed in parallel with the load resistor R3 to generate an alarm corresponding to a selected level: two operational amplifiers 49 and 50; five resistors (including two potentiometers 51 and 52); and a monostable circuit 53. The potentiometer 52 with fine adjustment knob sets the voltage threshold on the negative terminal of the operational amplifier 50. If the amplified voltage from R3 exceeds the voltage on the negative input by a small value, an acoustic warning is generated. The alarm will continue to sound until the amplified voltage is somewhat less than the threshold voltage on the negative input of the operational amplifier 50. This hysteresis is inherent in the circuit as shown. The operational amplifier 49 supplies a non-inverting direct current amplifier with a gain of approx. 42 which is selected by the potentiometer 51, so that the voltage at the output of the operational amplifier 49 is equal to 10v when the meter reads full scale, i.e. 1ma. Hysteresis is necessary in the comparator as the voltage through the meter varies somewhat before settling.

Figure 6 shows a form of the bottom measuring circuit 20. It should be noted that the same sample holding technique as mentioned in the parent application can be used here. In the upper part of Figure 6, corresponding components are thus given the same reference number, e.g. 33, and in the lower part, corresponding components are designated with the same reference number that is marked, e.g. 33'.

The analog signal from the amplifier 6 is fed to the base of the transistor 54, from which the signal is taken to the first and second bottom echo component processing circuits 55 and 56.

First bottom echo component processing circuit 55 comprises semiconductor switches 33, 35 and 39 (type 4016), an integrating capacitor C1, a leaky integrator and storage unit (composed of the variable resistor R2 and the capacitor C2), a resistor 57 and an amplifier 37. The switch 33 is controlled by signal G (figures 1 and 2), which is a signal of 20-40 milliseconds which begins when the first bottom echo component has reached a given amplitude. When signal G is not present, switch 33 is open and switch 39 is closed, so that the charge on capacitor C1 tends to zero. On the rising edge of signal G, switch 33 closes and switch 39 opens, so that capacitor C1 can be charged proportionally to the magnitude of the strength of the first echo component signal (an approximation to the integral of the signal).

On the negative trailing edge of signal G, switch 33 opens to stop the integration process and a monostable multivibrator 41 is activated to produce a short signal M (Figure 8), which closes switch 35, allowing the integrated led echo signal to charge capacitor C2 via resistor R2.

A delay circuit 40 composed of a diode, resistor, capacitor and inverter produces a signal N of signal G to open switch 39 and then closes it only after switch 35 is closed.

The elements R2 and C2 form an adjustable sample-hold circuit. If the resistor R2 is set to zero ohms, the capacitor C2 holds the charge from the preceding echo, so that the voltage across the capacitor C2 can then vary abruptly, e.g. fall almost to zero, if the echo component is not received during one cycle of the system. By increasing the value of the resistor R2, the capacitor stores the average value of more and more cycles (up to 85 when the resistor R2 is 100 kohm). The resistor is usually set to an intermediate value for storing the last approx. 10-15 bikes.

The second bottom echo circuit 56 is similar, in that it includes components that correspond to the components in circuit 55. The difference concerns further elements which are a NOR gate 58, a latch that includes NOR gates 59 and 60 and a monostable multivibrator 61 of 25 ms duration .

NELLER gate 58 is fed with signal G and also with the output from an inverter according to Figure 8. That output is designated as signal 0 in Figure 7, and is a pulse train that identifies echoes (especially bottom echoes) with sufficient amplitude. The NELLER gate 58 causes the part of signal 0 which appears during the first echo to be eliminated, while the subsequent part which appears at the second echo and is shown as signal P in Figure 9 remains. Signal P will thus appear at every other bottom echo with sufficient amplitude to trigger the circuit 35 in Figure 8.

Signal P sets the latch 59, 60 which is reset by signal B (the transmission pulse - figure 2). The output signal Q from the latch turns on the multivibrator 61 to produce signal R, which thus starts when the second bottom echo reaches a given amplitude and stops 25 milliseconds later. Signal R is used instead of signal G in the second circuit 56, thus extracting the second bottom echo.

The two circuits 55 and 56 are connected by means of switches 62 and 63 to an amplifier 64 and a meter 65, where the meter is as described above in connection with figures 1, 4 and 6.

When the switches 62 and 63 are in their upper positions, the meter 65 displays the first processing bottom echo signal. In the lower position, the second processed bottom echo signal is displayed. When one switch is up and one down, the sum of the signals is displayed.

It should be noted that Figure 7 shows four cycles, where the second bottom echo in the second cycle is too small to be captured, and in the third cycle both bottom echoes are missing. As can be seen from the other waveforms, the various switches are operated to prevent these low or missing signals from significantly reducing the meter reading for the sample-hold features.

Figure 8 shows a form of the bottom alarm circuit 17 which measures the width of the return echo signals. This circuit is particularly, but not exclusively, used in sonar, and the description below mentions the circuit in this connection, i.e. with first and second bottom echo components. In this case, the second bottom component is of particular interest, as its characteristic, i.e. width at a particular level, has been shown to be particularly sensitive to sea-bottom hardness, probably due to the double reflection involved.

The output of the amplifier 6 is shaped by the Schmitt trigger circuit 14B (Figure 1), with a threshold set by resistor 9, so that only echo components with a value above the threshold go to the circuit according to Figure 8. Inverters 65 and 66 and NELLER gates 67, 68 and 69 further process signal and gates it with the previously described signals B and F to eliminate or mask all components that appear before the first and second echo components in each cycle. To eliminate fish signals, signal F is delayed at 70. It is transformed at 71 and inverted at 72. The resulting signal 66 is signal 0 from Figure 8.

The NELLER gate 69 then gate-couples the output from the inverter 66 with signal G from the monostable circuit 30 or its inverse, depending on the position of the switch 73. Signal G has the effect that the first port echo signal is switched off, or everything except the first port echo signal is switched off, depending on the position of the switch 73.

The extracted echo component shown at I in Figure 9 sets latch 74, which is reset by signal B according to Figure 2 at the next transmission pulse. The output of the latch 74 triggers a monostable circuit 75 having a duration of 1 to 50 milliseconds, which is manually selected by means of the variable resistor 76. The output of the circuit 75 is shown on line J of Figure 5.

The output of the NOR gate 69 also goes to a resistor-capacitor combination 77 to delay the pulse I, which then continues to AND gate 78, inverter 79 and NOR gate 80. Line K shows pulse I in solid lines, as delayed and when it occurs at the input of NELLER gate 80. Its trailing edge occurs after the trailing edge of pulse J and this causes the output L of NELLER gate 80 to become high. This condition implies that the relevant echo had a size above the threshold set by the organs 9, 6 and 14b in Figure 1, and that at this threshold value it had a width that exceeded a value set by resistance 76. If the pulse had were as narrow as indicated by the dashed lines at K, the output from NELLER port 80 would remain low.

Signal L goes to a split-in-two circuit 81, which gives an alarm signal when the circuit 81 registers two pulses L in successive cycles, determined by signal B. The alarm signal goes to an acoustic alarm circuit 82 and can also sent to a visual display, e.g. of a sample-hold circuit, to maintain the alarm for a given time. By integrating the signal L, the visual display can give a value that determines the length of pulse L.

Figure 8 shows a further masking feature that can be used when the second or third or a later bottom echo is to be processed in this or another optional way, e.g. for cutting edge analysis. This masking feature is the feeding of pulse J via the network R, D and C, delaying the trailing edge of pulse J, to AND gate 78. This feature masks away all bottom echo components after the one being analyzed.

The analysis of the second bottom echo on the basis of width proves useful for recording the characteristics of the bottom conditions under water, as the echo increases in length with increasing hardness. A fisherman can, for example, look for conditions that are normally found at the boundary between the rock and sandy bottom. With the circuit according to Figure 8, it turns out that when passing from a sandy bottom to a rocky bottom, clear alarm patterns are formed that indicate the presence of the desired mixture of rocky and sandy bottom. During and after this process, an increasing reading on the bottom gauge is observed.

On the other hand, spikes in and at the end of the trailing edge of the second bottom echo may turn out to cause an alarm, because the time constants involved make the echo appear longer than it is. In this case, a bottom alarm as described in the above-mentioned UK-PS may be preferable. Alternatively, an alarm as shown in Figure 5 can be attached to the bottom gauge according to Figure 6, to issue a bottom alarm based on a selected gauge reading. Again, suitable bottom conditions for fishing can be identified with the help of a clear alarm series.

Figure 10 shows a modification, which is a digital method for identifying second and third bottom echoes, for forming a signal that corresponds to signal P in Figure 7. Here, the circuit arrangement according to

figure 8, which produces signal 0 together with gates 58-60 in figure 6. This modification makes use of the transmission pulse received at input 2 in figure 1 and first bottom echo pulse signal F produced by the monostable circuit 26 also shown in figure 1. The modification comprises an up/down counter 83, a clock 84, a comparator 85 and an AND gate 86 and an NOR gate 87.

The counter is designed to count up when signal F goes low (when TX occurs), and to count down when F goes high (when the bottom is reached). The counter is also reset by the monostable circuit 3, i.e. every time a transmission, in case the first echo is lost. The parallel output from the counter is compared to "zero" in the comparator which outputs a '1' when registering a 'zero<1>.

On transmission, the counter is reset, so that the comparator outputs a<1>1', but this is blocked by signal F at gate 86, so that Q is kept low at this stage.

The counter increments until F goes high, then the counter decrements to reach zero and causes the inputs to AND gate 86 to both go high, resulting in Q going high. Thus, port switching starts the second echo. If F does not go high (first echo lost), the counter will continue to increment until the next transmission. When Q goes high, gate 87 stops counting.

Since the second echo is somewhat more than twice the depth of the first echo (this is due to the depth of the transducer on the hull), gate coupling for the second echo will begin somewhat before this arrives, although the maximum error on a clock pulse occurs when the interval from transmission until the bottom pulse is measured. The maximum error in depth is designed not to amount to more than half a meter. Thus, a vessel with a draft of more than half a meter will begin to process the other echo somewhat before it arrives.

The meter's "LSB" should not represent more than half a meter of water; this determines the lowest clock frequency, which is preferably greater than or equal to 1.5 kHz. The maximum depth of operation can be calculated from the clock frequency and the number of bits in the counter. The maximum operating depth for the counter should preferably be equal to or greater than that corresponding to the period of the longest pulse repetition period, e.g. for a longest period of 3 sec-under, there will be a maximum depth of 2250 m. In this example, 12 bits will be required for the counter. In order to extract the third bottom echo component, the circuit according to Figure 10 can be modified to operate its mentioned counting sequence again to determine the expected arrival time of the third bottom echo. The maximum count reached will be stored, and either the counter will be reset to count back down from that value, or it will be designed to count up to that value.

Signal Q can be used in two ways. First, it can replace the circuitry already described for forming Q. Second, it can form the basis of an alternative bottom measurement circuit as shown in Figure 11. A monostable circuit 88 converts signal Q into a gate coupling signal of sufficient length to include the second echo, but excludes later echoes and ambient noise that would be included if the signal Q itself were used. A capacitor 89 above the meter can only charge while the monostable circuit 88 is active. The meter will thus show the integrated value of received echoes. Thus, the capacitor 89 together with the resistance in the circuit will provide a relatively large time constant, compared to the length of the time slot where the integration is to take place. A device which, for example, is appropriate for lobster fishermen in shallow water, can be based on the circuit according to Figure 11. In this case, it is sufficient to integrate all bottom codes, including the third and fourth, etc., as these echoes can be perceived under the conditions of shallow water. In this case, signal F can be fed to pass by circuit 88 (as indicated by the dashed lines). The signal Q itself may nevertheless be useful in such an application, as it will transmit the second, third, etc. echoes, all of which are particularly informative about the bottom conditions as a result of their reception after multiple reflections. An alternative port switching signal for third and later echoes alone could also be thought of as a possibility. A simple circuit that is specially designed for such an application can include the circuit according to figure 12 with the circuit devices for generating the necessary timing control signals, together with a depth measuring circuit, where the two feed a digital indicator that replaces the analog meter according to figure 11.

Attention is now drawn to the fish alarm 15 according to figure 1, which is shown in more detail in figure 12. This circuit also uses time slot analysis. The analog signal 14A passes through the variable amplifier 8.5 and the Schmitt trigger circuit 14A according to Figure 1, in order to digitize it, i.e. transform each echo component in signal A into a pulse when the relevant echo component exceeds a threshold set by the resistor 8. The transmitting component elimi- nered by signal B which is fed to NELLER port 16 in Figure 1. NOR gates 90 and 91 determine a time slot using the signals F and H. This time slot begins at a time determined by the pointer signal H (from the monostable circuit 30 in Figure 1), and ends at a time determined by the bottom pulse signal F This time slot is therefore synchronized by the first bottom echo pulse, and its width is determined by the variable resistor 31 (figure 1).

As this circuit's threshold is relatively low, the first bottom pulse that emerges from NELLER port 16 will tend to always be in the relevant time slot and thus have a serious impact on the registration of fish.

Accordingly, the circuit of Figure 12 comprises delay means formed by resistors R3 to R5, capacitors C3 to C5 and Schmitt trigger circuits 92 to 94. Resistors R4 and R5 are joined together to vary the delay. When R4, R5 are set for minimal delay, the first bottom pulse will produce an output at NELLER port 91. The delay can then be reduced by the operator, until the output is zero, whereupon it is known that the time slot ends at or just above the seabed. As the resistors R4, R5 delay the echoes, they will thus effectively move the time slot upwards in space, until its trailing edge is at the leading edge of the bottom echo. The delay devices also act as a filter for the digitized echo signal to reduce the effect of noise peaks in the signal.

This circuit supplements the fish measuring circuit 19 and the bottom measuring circuit 20. The fish measuring circuit 19 can also be controlled by a time gate that ends at the leading edge of the bottom pulse, and the meter reading (which represents the integral of the echoes in the gate) will thus be increased, i.a. of fish near the bottom. The alarm circuit according to Figure 12 is particularly useful as skippers cannot easily keep an eye on the gauges continuously, but can react when an acoustic alarm sounds. Tests at sea have shown that a sequence of five or six alarms or even more indicates a large concentration of fish very close to the bottom, and this concentration of fish will not be obvious on most sophisticated echo sounders with an expanded scale or color image.

The various theoretical and practical considerations regarding the registration of fish near the seabed will now be discussed. A feature to take into account is shown in Figure 13, which represents the seabed 97 and the partially spherical wavefront of the leading edge and trailing edge of the transmitted ultrasound pulse. As a result of the beam spread, signals return over a relatively wide area, so that fish A near the seabed, but not directly below the transducer, produces an echo that is received at such a time that it appears as B. If the fish is sufficiently high above the seabed, the echo of the EF appears as in figure 14, separated from the bottom echo E1. But in the situation shown in figure 13, fish A appears on the leading edge and thus EF has a higher amplitude. These signals are shown on an oscilloscope and depict half the undetected waveform as shown at the output (collector) of an input transistor for the amplifier 8. The total length of the echo E1 is usually 8-10 ms and it shows a characteristic hump at the tip.

Despite the fact that such fish echoes exist, recording fish near the bottom is one of the year-certain practical problems the fisherman faces. Conventional echo sounders show an image of depth in relation to the distance sailed, and the presence of fish in middle flow areas is revealed as echoes. However, there are problems when the fish are near the bottom. If the equipment is sensitive enough, such fish will cause echoes, but it is the identification of such echoes that causes problems.

R. B. Mitson (Fischeries Sonar, 1983, section 5.1.2) says that it is completely impossible to distinguish fish echoes from sea bottom echoes if they are less than half a pulse length above the bottom at normal angle of incidence. But as figures 14 and 15 show, a fish close to the seabed will influence the return echo and should in principle be detectable. What may be uncertain is whether this echo EF is a fish or simply a small lump in the bottom contour. In a case like 14, clear water will normally be seen between EF and E1 which can resolve the ambiguity. Mitson's analysis further assumes that a fish echo in shape and duration is an imitation of the "square wave" transmitter pulse, but we have reason to believe that it is Gaussian in shape and may be shorter in length. It would have the effect of reducing the impact of the half-length criterion, without necessarily drawing in the fish/bottom ambiguity.

One method for measuring the fish echoes is with the fish measuring circuit 19 for integrating the signal in a time gate blocked to the bottom, and which shows the result on a meter with a medium response time. Such a fish measuring circuit may in effect be similar to Figure 11, but with suitable gate switching signals fed to it. The gate is triggered past a bottom pulse (e.g. by signal H) to open it and is closed by the next bottom pulse, so that the gate thresholds are set above the fish echo levels. The gauge reading will rise in the presence of fish near the bottom. But this cannot in itself distinguish fish from bottom hardness variations that affect the leading edge of the bottom pulse. When using the bottom measuring circuit 20 also to monitor hardness, this distinction can however be made in most situations. The acoustic alarm system according to figure 12 has proven to be even more sensitive to fish registration and thus supplements the measurement. Any signal that exceeds the fish threshold and occurs within the gate will trigger the acoustic alarm. The gate is opened at a suitable and variable distance from the bottom by triggering past the preceding bottom pulse, and the gate is opened when the bottom echo reaches a suitable threshold. To avoid false alarm, the gate is moved by R3, R5 to close a few microseconds to a few milliseconds before the bottom echo reaches its threshold (determined by signalF), so that the gate actually closes just before the bottom echo reaches the fish threshold. Note that the leading edge used to form the gate has been smoothed by capacitor 13. The signal processed by the circuit of Figure 12 can also be smoothed somewhat to reduce noise, as shown by capacitor 11, but this effect can be made very small compared to the effect of the capacitor 13, the capacitor 11 can even be removed.

It will now be seen that fish near the bottom increase the amplitude of the leading edge (figure 15) and thus break through the threshold. In other words, a fish echo which in itself would be too small to trigger the alarm will nevertheless trigger the alarm if it is summed to the leading edge of the bottom echo (provided that the phase relationship is such that an additive effect occurs). The higher up the leading edge of the fish echo appears, the more sensitive the alarm is. In laboratory simulations, 'fish' only 75 mm from the seabed produced an alarm. Good operation may in practice depend on the weather conditions, but even with false alarms it may be possible for the human ear to distinguish the pattern in such alarms from a fish alarm.

Another theory regarding the favorable results of this method for registering fish is on figure 16. During observations of actual soundings made at sea with fish echoes near the bottom, it can be clearly seen that an alarm is generated when there are fish near the bottom, but in reality not on the bottom. However, some alarms are generated without the obvious presence of any echo at all. At such alarm moments, it was observed that the characteristic "hump" of the bottom echo is rounded off at Z. This causes the bottom pulse generated at the output of Schmitt 14c to spring back, ie be delayed, by perhaps 100-200 microseconds. Again, the effect is that a 'fishing' pulse is released through the system at the NELLER gate 90 according to figure 12. (It is of course actually a bottom signal in the fishing circuit that causes this.)

The theory behind this phenomenon is that a school of fish of sufficient density/volume lying on the seabed, deep within the "dead zone" due to the overlap of the echoes with the main bottom echo, causes a distortion of the leading edge of the main echo pulse from the seabed. It could be a Fresnel-type effect, where the fish echo EF is not in phase with the bottom echo. This distortion would be extremely difficult to see, even on an oscilloscope, but is recorded relatively easily using the techniques according to the present invention.

The embodiment examples discussed above use logic components that are fixed to achieve the desired functions with switches where it is necessary to change the mode of operation. Alternative circuitry can be used to achieve the same or similar effects. Microprocessor circuits plus software may also be used to implement at least some of the described functions. For example, waveforms can be digitized, stored in storage units and analyzed by software, for example to locate and identify the various echoes, especially the bottom echo components and leading edges. When time slots are referred to here, this can in some cases, especially when it comes to the leading edges of bottom echo components, be considered as including storage address areas. Such areas correspond to time slots and can be identified by analyzing the value and the change in value of storage locations, sooner than with real-time processing in relation to bottom pulses and sent pulses.

Claims (7)

1. Circuit for use in detecting characteristics of the seabed on the basis of a return echo signal from a hydroacoustic transmitter-receiver equipment (S), the circuit comprising means (28) for defining a time slot for extracting from the return echo signal (A) a first component constituting part of the first bottom echo component (E1), means (58, 59, 60) for defining a time slot for extracting from the return echo signal a second component (E2) which constitutes the second, or a later, bottom echo component, and means (55) to obtain a first signal from the extracted first component, characterized by means (C' ) for integrating the extracted second component to provide a second signal which gives an indication of the energy content of the second component, as well as means (62, 63, 64, 65) for providing a display representing the individual values of the first and second signal. 2. Circuit as stated in claim 1, characterized in that the means for producing the first signal comprise integration means (C1). 3. Circuit as stated in claim 1, characterized in that the means for producing a display are arranged to be able to function so that a value is displayed which is a further function of the value of the first and second signal. 4. Circuit as stated in claim 1, 2 or 3, characterized in that the second component consists of the second bottom echo component (E2). 5. Circuit as stated in one of the claims 1-4, especially for use with equipment where the pulses are sent out repetitively, so that cycles of return echo signals are received, characterized in that the circuit comprises organs (C2, R2, C2', R2 ') to hold the first and second signals obtained in one cycle for display in a subsequent cycle when, in said subsequent cycle, said time slot is not definable in relation to the return echo pulse due to poor signal reception. 6. Circuit as stated in claim 5, characterized in that it comprises sample holding means which are arranged to hold first and second signals and comprises breaking means (35, 35') which react to a return echo component with an amplitude which is below a given level to prevent updating of the output signal from the sample-holding means. 7. Circuit as specified in one of claims 1-6, characterized in that, for the purpose of defining a time slot for extracting the second bottom component, a digital counter (83) is provided which counts in one direction from an original transmission until -takes a first bottom echo component, and then counts in the other direction mainly to define the expected time of appearance of the second bottom echo component (E2).