CH617017A5 - Device for processing data, in particular seismic data, derived from acoustic underwater signals. - Google Patents

Description

It is an object of the invention to provide a data processing device which enables an increase in the resolution of large-scale devices in comparison to devices known today, without increasing the weight or the mass of the underwater cable arrangement.

The device according to the invention is characterized by the characterizing part of claim 1.

A two-stage multiplexed seismic data processing device is thus proposed. As will be explained further below, this principle enables the formation of shorter arrangements or rows, so that a relatively high resolving power of the device combined with a sensitivity to relatively high frequencies of the acoustic signals are achieved.

The seismic data processing device according to the present invention is designed in such a way that the electronics that process the signals are decentralized, normally housed in a central registration vessel, in that the electronics are installed in the individual cable sections that form the cable arrangement. Seismic analog signals from individual seismic sensors are digitally displayed and transmitted to a common, central station via a data transmission connection by using the aforementioned two-stage multiplexing method which delays the running time.

In one embodiment of the invention, the seismic cable arrangement has at least 10 identical cable sections. A number of seismic sensor groups are included in each cable section. A transceiver is assigned to each cable section; the term "transceiver" is intended to refer to the electronics assembled in connector modules which form part of the cable assembly and an output circuit for the transmission of digitally represented seismic signals to the cable in addition to the common electronic circuits for amplification and digital representation of analog seismic data input information and the switching arrangement for successively connecting the elementary seismic sensor units to the common electronic circuit. The cable sections and connector modules with the transceiver are mechanically and electrically connected to one another. The signals from each of the transceivers are successively applied to the seismic cable and thus to the central station, while the switching arrangement is connected in such a way that it receives signals from the assigned seismic sensors. In this way, digital signals from all sensor units are successively fed to the central station. The length of each seismic sensor group is a certain fraction of the length of a seismic wave, the frequency of which corresponds to the successive delivery rate of the signals applied to the cable, as will be explained below.

According to a further embodiment of the invention, the seismic cable arrangement consists of a number of active cable sections, each of which contains an interrogation link, a data transmission connection and a number of seismic sensor groups when the sensors are accommodated in the cable. The connection modules with the transceivers connect adjacent cable sections to one another. Each transceiver has a plurality of input channels, a data repetition network, and an interrogation network. The polling network of each transceiver is connected in series with the polling link. The seismic sensor groups in intermediate cable sections are coupled to corresponding input channels of the assigned, specific transceiver, which is arranged at one end of an assigned cable section. The data repetition network of each transceiver is coupled to the data transmission link.

Another embodiment of the present invention has a multiplex switch in each transceiver. The multiplex switch is controlled by a control network

617 017

sequentially operated, which is connected to the query network. Depending on polling pulses from the central station, which are transmitted via the polling element, the polling network causes the control network to actuate the multiplexer frequently and digitally display the analog data from the seismic sensor groups and to transmit a self-clocking digital data word to the transmission link via the data repetition network.

A plurality of preamplifiers / filters, each with an input and an output, are preferably provided in each transceiver. The inputs are coupled to an equal number of seismic sensor groups which are fastened within the assigned cable section, in particular in the case of seismic marine cables. The preamplifier / filter outputs are connected to corresponding inputs of the multiplexer, the single output of which is coupled to the common amplifier, digital display device, buffer and output register, code converter and repetition network. The output of the repetitive network is in turn connected to the data transmission connection. The interrogator is coupled to the multiplexer and to the output register via a control network. Depending on an interrogation signal, which is a pulse which is transmitted to the transceivers via the interrogator, the multiplexer in each transceiver is sequenced by the control network at a selected, e.g. switched to the next channel so that an analog data test is achieved '- *. The data check is prepared for amplification, digitized so that a digital data word is formed, and clocked into the repetition network by the output register and code converter. When the interrogation pulse arrives at each of the corresponding transceivers, a corresponding, self-clocking, phase-coded data word is transmitted to the central station via the data transmission connection. A second, next interrogation pulse brings the multiplexers in each of the transceivers to a second, i.e. next channel for testing and digitizing the next signal, while a phase encoded word from the first, i.e. previously checked channels of the transceivers are transmitted. Accordingly, all analog input channels in all transceivers are checked over a period of a sampling cycle.

According to further features of an embodiment of the invention, the data and query connections have three parallel redundant lines. A majority selector circuit in each transceiver receives a signal from two of the three lines. A fault indicator circuit, coupled to the data and polling connections on each transceiver, indicates an interrupted line within the associated cable section if the majority selector circuit does not indicate three identical inputs. The transmission link can be a broadband transmission channel with a bandwidth of at least 100 MHz.

According to a further feature of an embodiment of the invention, each transceiver has a data repeater, which in turn has a signal receiver, regenerator and transmitter for receiving, regenerating and retransmitting signals from downlink transceivers to uplink transceivers . There may be an artificial delay line in the interrogation network connected in series with the interrogator in each transceiver, thereby delaying the arrival of the interrogation pulse at the downlink transceivers and thereby separating the data words from each other when they are separated from one another be transmitted to another, adjacent cable section.

5

S

10th

15

2 \)

25th

311

35

40

45

50

55

60

fi5

617 017 6

Furthermore, according to the present invention, a calibration signal is connected that it provides a multiplicity of seismic analog signals which is transmitted by the central station in order to drive signals via the input channels into one of the transmit / receive groups of each seismic sensor. The output feeds devices, with each transceiver device of each sensor group being compared with the input signal using a data processing device for conversion so that a sensor calibration factor is obtained. s has an analog signal given to it in digital form. According to a further embodiment, and each transceiver has a device, it should be noted that the seismic data processing involves self-clocking digital words, they are a seismic signal display system according to the present invention Number of digits from each of the input channels to all of the transmit / receive transmitters that are remote from a broadband transmission link on central station. The transceiver, "admit.

are connected to the central station by a broadband data transmission and a reflection-based seismographic transmission processing and an interrogator. Exploration method is thus a seismic fault - Each unit has several input channels, each of which is routed, at least two direction-sensitive cables are connected to a short seismic sensor subassembly, which is also housed in a cable as a sensor group, and is referred to as a mixed sensor group is, which -> a geophysical exploration by individual behavior consists of three sensors, which are about 2 m apart from each other in the directional sensitivity of each group. A channel selector, which is the above course of recording reflections of the seismic interference-mentioned multiplexer, connects sequentially the input detection performed, whereby group signals are obtained, channels with a common, the signals propagating and a special feature of an embodiment of the Erfin-digital performing network including an output extension sees the use of ten or more groups, ie channel. The output channel specifies digital data signals from the sets of sensors which are arranged along a cable offset from seismic sensor groups (seismic subassemblies), each such set having a multiplicity of elements to the data transmission connection to the multiplex-transmissive sensor groups, and the signals from Sensing to the central station. Digit groups are combined in the central station, so that each group with total data words, which represent seismic signals from the subordinate delays between the combined signals, is evaluated and provided by a format control that is just as large that the group is composed in the direction of new data words, which represent ice-adjacent underground locations in the mixed signal to be explored, which is targeted from a desired, essential area, whereby the delay between the larger seismic sensor series would have been obtained. Units for different signal groups is different

According to one embodiment of the invention, the central and also changes with time, so that the groups so tralstation a core memory, and a device is directed that they are provided signals from progressively lower to store the digital data words when recording shifts. The resulting tracking signals can be obtained from the elementary seismic sensor groups, i.e. which are then combined, whereby the usual movement corrections

Subgroups can be used in a memory matrix in channel sequencing rectification and display techniques around which were taken. The central station also has a complete ^ seismic section to gain.

mat control, which takes data words from the memory. According to further developments, a seismic data processing device that arises from selected seismic sensor groups is thus proposed, which has a combination

and represents these signals as new composite data words nation from the following individual constituents: one combined, which represent desired larger groups. central unit, a variety of remote transmit-receive

The central station may also have a fixed coefficient -jo devices, a broadband transmission link that the

Have memories (RCM) in order to send the data words desired transceivers to the central station

Unmatched evaluation coefficients in front of individual subgroups, a number of spaced apart before they are combined as a composite data word elementary seismic sensors that are switched so that. they have a variety of analog seismic signals in each

A control unit in the central station transmits the query .is transceiver, a data processing

pulses to the transceivers via the interrogation device in each transceiver. When responding to the query signals, i.e. Pulses convert each analog signal into a digital signal, and tests the electronics in the corresponding transceiver - a device for selectively combining digital signals

Devices the analog signal that is present on an input signal from sets of elementary seismic sensor groups, the sample digitally represents and transmits the digital data word as a self-clocking, phase-co-determined sets of composite group signals and transmits the digital, and for selective combining digital signals

Data word in the central station via the broadband transmission different sets of seismic sensor groups around the environment connection. When the controller forms the arrival of different sets of group signals, being

If data words are reached from a first channel in all transmit / receive desired sensor groups configurations, the device sends a second interrogation measure without the seismic cable or another pulse being sent out to a second channel in all of the transmit / receive pre-detection device needs to be changed in the work area.

directions to check, etc. as mentioned above until all. According to another embodiment of the present

Input channels tested in all transceivers The invention is an analog-digital signal acquisition arrangement

have been, the interval between successive, preferably a seismic data processing device

the query impulses to the same channel in a typical manner, which receive a large number of signals.

up to 1 mill sec. the input channels, one of which receives no signals

With the present invention, a seismic and is therefore grounded, a multiplexer with a multiplicity of inputs

Data processing system proposed that a plurality of gears and a common output line, where

Transceiver devices and a broadband data transfer - a capacitor between each of the input channels and transmission link, the transceiver - "5 is connected to one of the multiplexer inputs and connects to each line with a central station, a large number of capacitors, an interference voltage is generated, a switching arrangement

sets of spaced elementary seismination with an amplifier to connect the multiplexer

rule provided sensor groups, each set so specialized line with an evaluation device, and a tracer

7 617 017

cherschaltung, between the switching arrangement and the amplification of the amplifier stages connected in cascade

Value device is switched on, the key memory switch being a bypass switch and a series capacitor equal to the successive powers of n or 0 to 2m.k tion.

In a further embodiment of the invention, polarity to the interference voltage is generated each time a desired switching sequence in the components becomes a desired one if the grounded input channel is switched by the multiplexer switching sequence in the components of a desired subgroup is scanned and the bypass switch is closed, introduced by successive data acquisition units

has, the test voltage during operation night; the subgroup is made up of the multitude of data

the interference voltages selected by the signal acquisition units. The sub-group has a first one which cancels the channels when the multiplexer has the unit selected by the signals m and a last selected unit.

scans the receiving channels. An interrogation pulse in the first state is from the central

A variable gain amplifier processing arrangement is transmitted through the interrogation channel. According to the degree of efficiency, at least two, preferably four, amplifiers of a selected time delay have a long control

with two amplification levels, which are transmitted in a cascade between pulse via the control channel. The leading edge of the one input connection and one output connection of the long control pulse overtakes the interrogation pulse in the first tet. Each amplifier normally has a state on the first selected unit and intercepts it.

low gain factor and can be controlled by a control unit. The trailing edge of the long control pulse overtakes the switch to a state with a high gain factor and passes through all units before the request pulse.

the. A voltage reference device selectively yields a distance beyond the last selected unit. The length of the crete reference voltage that corresponds to each amplifier. A: n long control pulse is equal to a first integral multi-

The comparison device carries out a comparison between the compensation of the signal transit time difference by the two channels between

The system's output voltage and the selected reference voltage see any two data acquisition units. By. If the comparison shows that the reference voltage desired switching sequence only occurs in the units in which the output voltage is greater, the control unit sets the first amplifier in its state with high amplification, which are present approximately simultaneously with the interrogation and control pulses. The first integral multiple is equal to the number of factor. Subsequently, the comparison device leads components reduced by one that exist in the sub-group

are a second comparison between the output voltage of the. The selected time delay is a second integral

System and the reference voltage, which corresponds to the second magnitude multiple of the signal delay difference between two more closely. If the comparison again indicates that any data acquisition units, the second variety

the reference voltage is greater than the output voltage, represents vi che is equal to the number of units between the first, the control unit, the second amplifier in the state with the selected unit and the central processing device.

high gain factor. This process will be repeated.

catches up until the comparison indicates that the output voltage of the

Systems greater than the reference voltage according to the invention, three parallel control channels are provided. At least one particular amplifier, or until all amplifier majority selection circuitry on each data acquisition unit is set to the high gain state, is coupled to the three control channels. Have been delayed a long time. The total gain factor of the system is control impulse is transmitted in parallel over the three control channels.

thus in increments of a value with low comparisons. The simultaneous recording of an interrogation pulse in the amplification factor, preferably 1 for incoming signals with the first state on a data acquisition unit over the large amplitude to a value with a high amplification factor interrogation channel and a long control pulse over at least two gate adjustable for incoming signals with small amplitude, the three control channels conduct a first desired switching sequence

This means that no amplifier runs the risk of being overdriven.

will. According to a further embodiment of the invention

In one embodiment of the invention, arrays of a delayed short control pulse are transmitted over the first of the three settings to automatically clear DC or 15 control lines. The simultaneous arrival of a very low frequency noise signal is provided which occurs at the interrogation pulses in the first state at a selected input connection to each amplifier due to the individual data acquisition unit and a delayed short control duel amplifier characteristics. The noise impulses via the first control channel conduct a second quenching circuit, a capacitor and a switching-desired switching sequence.

ter between the input-output circuit of a 50 Furthermore, it is proposed that a third desired each amplifier for periodically isolating the amplifier switching sequence in a selected data acquisition unit and for charging the capacitor with the self-generated the simultaneous arrival of an interrogation pulse in the first

Amplifier noise are switched. The charge at the condensed state via the interrogation channel and a delayed short gate has an amplitude and a polarity equal to and opposite control impulse via the second of the three control lines.

against the self-generated noise, such that the 55 is conducted.

DC noise signals appearing at the input terminals can additionally substantially cancel a fourth desired switching sequence in each amplifier. selected data acquisition unit by the simultaneous According to a further example of the invention, an arrival of an interrogation pulse in the first state via the amplifier arrangement with variable amplification factor, interrogation channel and a delayed short control pulse will be initiated in particular for use in connection with the here m via the third of the three control lines , described seismic data processing system according to further embodiments, interrogation and striking, which has a number of cascaded amplifier control pulses repeatedly on the data acquisition units in core stages, each of which transmits an input and a preselected sampling interval after a first course and Has gain states where a switching sy recording cycle has been initiated. The width and the stem for controlling the amplification state of the amplifier - <.1 duration of transmission of the control pulse are graded as a function of the output from the amplifier - such that a desired sequence of events is provided in a first subsystem; the amplification states are so grappled by data acquisition units that a preselected one is selected and the switching system is designed so that the total number of component units is made effective.

617 017

After the first recording cycle has been completed, a second recording cycle is initiated and the width and duration of the transmission of the control pulse are adjusted so that a desired switching sequence is activated in a second subset of data acquisition units. The above-mentioned steps can be repeated several times, so that a device is achieved which makes a desired switching sequence effective in successive subgroups of subsequent data acquisition units.

It is also proposed to keep the width of the control pulse constant for each recording cycle. For each recording cycle, the duration of the transmission of the control pulse with respect to the transmission duration of the interrogation pulse is different by an integer multiple of the signal delay difference, i.e. Delay, delayed between any two units. For example, by increasing the delay by a multiple of one unit after each recording cycle, successive subgroups of data acquisition units are activated one after the other, as a result of which the desired rollability described above is achieved.

In one embodiment of the present invention, a seismic data processing device and a method for transmitting messages to components of a subset of seismic data acquisition devices, which are selected from a multiplicity of remote seismic data acquisition devices, are proposed, the subset including at least one such device; this method is to transmit a first signal to the seismic data acquisition device, to transmit a faster second signal to the seismic data acquisition device, and to delay the transmission of the second signal with respect to the first signal so that the first and second ones Signals arrive approximately simultaneously at a selected subset of the seismic device.

The invention is explained below in connection with the drawing using exemplary embodiments. Show it:

1 is a schematic overall view of an embodiment of a seismic data processing device which is operated in the water and which is towed behind a vehicle,

2a-2d cross-sectional views of a section through the seismic cable of FIG. 1,

3a-3d cross-sectional views of a connector module with transceiver as part of the seismic cable of FIG. 1,

4 shows a circuit diagram of the energy feed connections for the transceivers,

5 is a schematic circuit diagram of a transceiver;

6 shows a schematic illustration of the priority selection and error display circuits,

7a shows a schematic illustration of a pressure transducer, FIG. 7b shows the electrical connections of the auxiliary channels, FIG. 8a shows a sectional view of an end section,

8b is a schematic representation of the electrical connections in the end section,

9 is a cross-sectional view of the lead-in cable section, FIG. 10 is a timing diagram illustrating a kind of self-clocking code for the transmission of data words,

12 is a timing diagram showing the sequence of interrogation signal and data signal transmission with respect to two cable sections,

13 is a timing diagram illustrating a method for activating three successive transceivers, but not other devices.

14 is a schematic block diagram of a typical filter circuit for a seismic analog-digital signal processing system,

15 shows a schematic illustration of an embodiment of a commutated high-pass filter according to the invention,

16 shows a schematic illustration of an analog-digital converter in which the commutated filter according to FIG. 15 is used,

17 is a schematic block diagram of the seismic data processing system according to the present invention.

18 is a schematic illustration of the electronics contained in the transceiver;

19 shows a detailed schematic illustration of the repetition network in the transceiver according to FIG. 18,

20 shows time diagrams for explaining the mode of operation of the system according to FIG. 17,

21 is a schematic illustration of a group forming device as contained in the central station of the device according to the present invention;

22 modifications of the form of different pulse codes as used in the device according to the invention,

23 shows a diagram of a marine seismic exploration system, on the basis of which certain features according to the present invention are explained,

24 and 25 are diagrams showing the response of an unguided seismic cable group at different frequencies as a function of the incidence or inclination of the different layers from which signals are reflected; Fig. 26 is a diagram showing the configuration of a typical known undirected seismic Grouping indicates

Figure 27 is a diagram illustrating a tapered grouping of the type used in practicing certain features of the present invention.

28 is a graphical representation of the root mean square of the velocity of the seismic waves as a function of the duration of reflection from the initial shot to the time when the seismic reflection reaches a sensor.

29 is a simplified diagram used in calculating the delay on neighboring sensor groups that form a multi-group seismic grouping.

30 shows the response of the steerable seismic grouping as a function of an angle of inclination,

31 is a block diagram of an embodiment of the large-scale seismic data processing device according to the present invention.

32 is a block diagram of the cable electronics and seismic sensor groups used in one embodiment of the invention.

33 is a detailed block diagram of a grouping molding apparatus which is part of the circuit of FIG. 31;

34 is an overall block diagram of the specific data processing device used in executing the beam steering grouping molding device according to an embodiment of the invention;

35 is a diagram showing the formation of groupings in the beam steering apparatus of FIGS. 31 and 34; FIG. 36 is a simplified block diagram illustrating a typical digital-to-analog conversion system;

37 is a simplified schematic block diagram of such a device, which is used for the purpose of the present invention

5

10th

15

20th

25th

30th

35

40

45

50

55

60

for \ 5

9

617 017

modified to achieve DC deviation elimination,

38 is a diagram of a modified circuit corresponding to that of FIG. 37, but in which the filter of FIG. 16 is inserted into the input channels,

39 is a block diagram of a variable gain binary gain system used in the illustrated embodiment of the invention;

40 shows a schematic diagram of an amplifier stage of the system according to FIG. 39 with a noise canceling device,

41 and 42 are schematic representations of the principle of canceling the noise as used in the circuit of FIG. 40,

Fig. 43 is a highly simplified, schematic illustration of the multi-channel seismic data processing system showing the data processing devices which are the transceivers in connection via a multi-channel remote measurement connection to the central signal processing device which is the common central station, this illustration being used for explanation a special modification of the illustrated embodiment is used,

44 is a block diagram of the signal conditioning logic included in each transceiver;

45 is a circuit diagram of the signal property identification device and the first and second signal coincidence detectors in a transceiver;

46 shows a circuit arrangement for delaying the transmission of a control signal following the transmission of an interrogation signal and for giving the control signal a time shift,

47 shows an expansion of the representation according to FIG. 43,

showing seven data acquisition devices, of which three successive devices are activated, and a timing diagram showing the temporal relationships of the control and interrogation pulses with respect to the transceivers and

48 shows a modified embodiment of the control and interrogation circuits according to FIG. 45.

1, a ship 10 tows a seismic sensor cable assembly 12 through water 14. The seismic sensor cable assembly 12 is connected to an elastic shock absorber section 16 and an insertion section 17. The trailing end of the cable assembly 12 is provided with a short end section 18. The cable arrangement 12 is divided into individual active cable sections 20, each of which is 60 m long. Each of a number of connector modules 13, including an electronics assembly, referred to as transceiver 111, connects active cable sections 20 to each other both electrically and mechanically. A typical seismic cable arrangement 12 consists of 50 or more active sections 20 and can have a total length of 3000 m or more. Each cable section can contain ten sensor groups 21, each of which represents a single channel. The entire cable arrangement 12 therefore generates output signals from 500 individual channels. The sensors can be hydrophones if the embodiment shown is a marine seismic cable.

The signal outputs from the individual sensor groups 21 are coupled to one of the transceivers, which transmits the signals to a central station 2 on the ship 10. The central station contains a control circuit 4 for the transmission of query, command, power and test signals and a device 6 for recording and recording digital data words from a data transmission connection in the cable.

When the ship is dragging the cable assembly 12 through the water, a seismic switching source 19, e.g. an air cannon or a gas explosion device, at intervals acoustic waves in the water. The acoustic waves travel down through the water 14, e.g. along the jet path 15 and meet the water bottom surface 22, where they are broken in the direction of arrow 27 due to the different speed between water 14 and earth layer 24. When penetrating the ground, the acoustic waves continue their path along the broken beam path 28 and are reflected by an underground layer of earth 26. The reflected acoustic waves return along the Ul beam path 28 to the bottom surface 22 of the water and from there continue their way up along the beam path 30. The reflected waves are displayed by sensor groups 21, which convert the reflected acoustic waves into electrical signals. The acoustic waves i 5 also take other beam paths, e.g. 31-32-34-36, where they are separated by sensor groups, e.g. 21 ', which are further away from the ship 10 than the sensor group 21. Although there are similar beam paths between the switching source 19 and each of the 500 seismic sensor groups in 2o of the seismic cable assembly 12, only two such paths are out Shown for clarity.

FIG. 2a shows a schematic longitudinal sectional view of the leading end of an active seismic cable section 20.: s For the sake of the drawing, the longitudinal dimensions have been significantly shortened. The section consists of an outer plastic skin 40, three steel load members 42, 43 (the third member is not shown), a plurality of end wall spacers 44 and an end spacer 46 at each end of the section. The plastic skin 40 has an inner diameter of 7 cm and a wall thickness of 0.475 cm. The end wall spacers 44 are provided at 60 cm intervals within the plastic skin 40 for internal support. Each end wall spacer 35 44 has three holes 48, 48 ', 48 "(FIG. 2b) for the load components to be passed through and a central hole 50 for the connection cable bundle 52 to be passed through. The skin is fixed to the end end wall 46 via steel bands 54, 56 entire skin is filled with light kerosene, 40 which gives a neutral buoyancy in the water.

A variety of sensors 23, e.g. Hydrophones are arranged at intervals of 2 m within the cable section 20. Each sensor is supported between two closely spaced end walls 44 by means of connecting elements of the conventional type. Each cable section preferably has at least 30 seismic sensors 23. In a preferred arrangement, 3 sensors 23 are connected in parallel by local data lines 58, 60, so that they form a single instrument and thus an elementary group 21 of seismic sensors. Since the sensors are offset 2 m from each other, the length of the elementary sensor group is 4 m and the separation between group centers is 6 m. The local data lines 58, 60 are connected to cable bundles 52 and carry the signals from the sensor units to corresponding 55 pins of a multi-conductor connector 62. In this arrangement, each elementary seismic sensor group 21 feeds signals from three sensors 23 into a single, common data channel. The parallel connection has the effect that the electrical outputs of the individual sensors 23 are summed alge-60.

The summation of the signals tends to amplify desired, systematic, reflected signals and to suppress unwanted, arbitrary noise signals, provided that the seismic wavefront is approximately parallel to the level of the grouping. In this ideal case, “see”

all sensors 23 in the elementary sensor groups 21 and 21 '(FIG. 1) the wave front and record the seismic waves in phase. The angle increases with increasing distance from

617 017

the source greater. In addition, the beam path angle depends on the slope of the bottom surface 22 of the water, the reflective interface 26, and many other factors.

A wavefront, e.g. 35, which can be a wave guide, proceeds so that its extension remains perpendicular to the beam paths 30, 36. When Wellenbert 35 strokes cable assembly 12, it is first detected by sensor group 21, and somewhat later by sensor group 21 '. At the moment when the wave crest 35 hits the sensor group 21, it is possible that the wave trough of a preceding wave is still scanned by a sensor group 21 '. If all sensors between 21 and 21 'were connected together in a long group, signals from the sensor outputs would weaken rather than amplify each other. It is therefore desirable that the length of an individual elementary sensor group be short compared to the wavelength of the seismic signal of interest with the highest frequency.

The wavelength of a seismic wave, as received by a group of seismic sensors that are electrically connected, depends in a complicated manner on the angle of inclination and the depth of the underground layers of earth, the speed of the seismic wave, the distance between the acoustic source and the Sensors and many other factors. Below is considered a group of electrically connected seismic sensors located at or near the surface of the water. The group has a length X. If a horizontally traveling wave (angle of incidence = 90 °) is incident at one end of the group, the time T is required for the wave to cross the group

T = X / V (B)

where V is the acoustic velocity in the spreading medium. At a speed in water of 1,500 m / sec and a group length of 69 m of the known type, the transit time for the wave is 0.046 sec. For an additive response in the longitudinal direction of the sensor group, the group length should be less than the XU wavelength. The running time of a wave is thus 0.046 X 4 or about 0.184 sec. This period corresponds to a cutoff frequency of about 6 Hz. Waves that come in at the end of the group and have frequencies that are much larger than the 6 Hz limit tend to that they are strongly dampened.

In the system shown, the length of an elementary sensor group is 4 m. The running time for a wave is 0.0026 seconds. The frequency corresponding to a quarter wavelength is f = 1 / (0.0026 X 4) = 96.1 Hz.

By using a short elementary seismic sensor unit or group, the upper limit frequency has been significantly expanded. Assuming an angle of incidence of 30 ° for a wide-angled, shallow reflection signal that is picked up towards the end of the cable arrangement, the upper limit frequency is raised to f = 96.1 / sin 30 ° = 192.2 Hz.

In addition to the seismic sensors 23, auxiliary sensors, e.g. Pressure transducers 64, loss indicator conductors 66, 68 and a water break indicator 72 are provided in the cable section 20 near the leading end. Electrical connections from the additional sensors are connected to the cable bundle 52 and transmit sensor output signals to corresponding pins of a connecting plug 62.

10th

For a typical cable section 20, for example: 10 data channels and 3 auxiliary channels available.

FIG. 2c is a cross-sectional view of the cable section along line 2c-c and shows the configuration of the seismic s and additional sensors within the outer layer 40. The reference symbols in FIG. 2c correspond to the same reference symbols in FIG. 2a. FIG. 2d is a cross section through an end wall 44 along the line 2d-d according to FIG. 2 and shows holes 48 for the load component 42 and an opening 50 for the cable bundle m 52.

The cable cuts are interconnected as shown in Fig. 3a. In this representation, the ends of adjacent cable cuts are symmetrical, so that only one end is described in detail. The load components 42 and 43 extend i 5 through an end end wall 46 and are terminated by pulling hooks 45, 47 or the like which are known per se. The cable bundle 52, which extends through a central opening in the end end wall 46, ends at a connecting plug 62.

A connector block 13 is provided between adjacent sections 20. Each connector module 13 contains a transceiver, the purpose of which is to receive analog signals from seismic sensor groups and auxiliary sensors, to digitally represent the signals and to transmit the digital data to the ship 10 via a data transmission connection in the cable bundle 52. The connection module 13 has an end wall connector 76 at each end, which mates with a connector 62. The matching connector plugs allow the sensor groups to be coupled to the inner transmitter! » Receiving device and represent a device for coupling the transceiver in series with the transmission link, the interrogator, the energy and test signal channels in the connection cable bundle 52. Short stress components 78, 80 (a third corresponding component is not shown i5), which are completed by pulling hooks 82, 84, which fit together with the pulling hooks 45, 47, are fixed to the connector module 13 by steel clips 86, 83. The housing 75 of the connector module 13 and the end wall connector 76 are designed such that they can withstand an ambient pressure of up to 140 kg / cm 2. The outer dimensions are 6.25 cm × 35 cm.

If two cable sections 20, 20 'are to be connected to one another, the pull hooks 45 and 47 of the load components 42, 43 are connected to the associated pull hooks 82, 84 45 of the short load components 78, 80 via bolts 90, 92. A connector plug 62 is associated with each end wall connector 76 at each end of the connector module 13. A plastic cuff 94 with an inner diameter that is somewhat larger than the outer diameter of the outer layer 40 is slipped over the end walls 46. The cuff 94 is fixed to the end walls 46 by steel straps 96,98. The inside of cuff 94 can be filled with light kerosene (for buoyancy). A better buoyancy in water can be achieved by syntactic foam (55, for example, manufactured by Dow Chemical Corp.).

The connector module 13 is shown in partial cross section in Fig. 3b. Each end of the cylindrical housing 75 of the connector assembly 13 (Fig. 3b) is terminated by an end wall connector (, a 76 which slides into a recess 100 machined into the end of the housing 75. O-rings 102, 104 form a fluid seal around the The end wall connector 76 is secured in place by a snap ring 106.

65 The connector module 13 is shown in cross section along the line 3c-c in Fig. 3c and along the line 3d-d in Fig. 3d. The electronics of the transceiver contained in the receiver module 13 (hereinafter in detail

11

617 017

)) is attached to three printed circuit boards 108, 110, 112 (FIGS. 3c and 3d). The three plates, which represent the transceiver 111, are designed in the form of a triangular prism. They are designed so that they can be inserted into the housing 75. Before insertion, the interior of the housing 75 is lined with a thin glass fiber layer (not shown) so that the electronics are isolated from the metal wall. After assembly, the connector module 13 is filled with mineral oil of a known type which does not damage the electronic components. This results in good heat conduction and prevents water from entering.

In rough weather at sea, the towing vehicle 10 (FIG. 1) is subject to unpredictable accelerations about the pitch, roll and yaw axis. So that such accelerations are not transmitted to the seismic sensor cable arrangement 12, one or more elastic cable sections 16 are connected between the insertion section 17 and the cable 12. The elastic cable sections are similar in construction to an active cable section, except that they do not contain seismic or auxiliary sensors. Ropes made of nylon or other elastic material are used instead of the steel load components. A cable bundle, which corresponds to the cable bundle 52 according to FIG. 2a, is guided through the central openings 50 of the end walls 44. A sufficient sag is provided in the cable bundle 52 so that the section can stretch up to 50% of its relaxed length. In a preferred embodiment, two such stretching sections are used. A connector member is inserted between the leading end of the first active section and the trailing end of the second draining elastic section. A second connector block 13 'is inserted between the trailing end of the lead-in cable section 17 and the leading end of the leading stretching section.

The cable bundle 52 in each cable section 20 contains two sets of current conductors. A set of local data conductors, e.g. 58, 60, transmits analog signals at the elementary seismic sensor units 21 and the auxiliary sensors within each cable section to the transceiver within an adjacent connector module 13. The local conductors are preferably coaxial cables, e.g. RG-174. The other set of cables are feed-through trunks for transmitting query, command and control signals from the ship 10 to each transceiver 111 in a connector block 13 and for transmitting data signals from each transceiver 111 back to the ship 10. Die Trunks have a data transmission link, an interrogation connector, two command connectors, two test lines and a power transmission line. With the help of the plug connections on the plug module 13, the long-distance lines extend over the entire length of the cable arrangement 12.

In a preferred embodiment of the invention, the broadband data transmission connection consists of three coaxial cables RG-58 / CU. Coaxial cables are required to accommodate the 20 megabits per second transmission speed (40 MHz for a word that consists only of ONE). Three cables are used instead of one for redundancy reasons. If one cable fails, two more are available for operation.

The interrogation connector consists of three redundant transmission lines with twisted pairs. Twisted pairs of conductors are permissible for this and all other signal transmission lines because the transmission speed of the interrogation pulses is comparatively slow, namely in the kHz range. The two command connectors with twisted pairs transmit the control signals. The test

and test control lines are twisted pairs of conductors through which a test signal and a test control pulse are transmitted.

The power transmission line consists of two twisted pairs of 14 AWG wire connected in parallel. AC power is transmitted via this line in order to excite the energy feeds for the transceiver which are present in each of the connector modules 13.

A block diagram of a transceiver 111 mounted on printed circuit boards 108, 110, 112 (FIG. 3b) of a connector module 13 is shown in FIG. 5. The essential components are a repeater network 114, a query network 116, a command network 118, preamplifier 120, a multiplexer 122, the gain-conditioning amplifier 124, an analog-to-digital converter (digital display device) 126, a code register and a code converter 128, an error detector 130, a control network 132, an energy supply source 134, a test drive device 136 and a test control relay 138. The transceiver circuit elements are explained in detail below, the functions of the transceiver 111 are shown in block form in FIG. 5 schematically shown for a better understanding of the mode of operation.

Repeater network 114 transmits a local, self-clocking, phase-coded data word to central station 2 (FIG. 1) via data transmission connection D 1, D 2, D 3, and then receives, regenerates and transmits self-clocking, phase-coded data words from the downlinked ones Transceivers. Depending on a first interrogation pulse, these functions are initiated and completed before a second interrogation pulse arrives. On command or in the event of a power failure in a transceiver, the receiver can be shunted, as will be explained below.

The interrogation network 116 receives, buffers and transmits interrogation pulses via the triple redundant interrogation connector IP 1, IP 2, IP 3. This unit identifies the interrogation pulse by means of a pulse width identification circuit, since it is either a wide pulse S1 or a narrow pulse S2. The S1 pulse is 1500 nanoseconds wide; the S2 pulse is 600 nanoseconds wide. The query network 116 has an artificial delay line in series with the query connector. The preferred delay is 600 nanoseconds. The artificial delay line is tapped so that small adjustments can be made to compensate for small differences in the signal propagation times through the interrogation connector.

The command network 118 picks up the two command signals DATA (activate data) and DATA BYPASS, buffers and transmits them to downlink transceivers. The simultaneous arrival of an S1 interrogation pulse and a DATA BYPASS pulse at a selected transceiver 111 causes phase encoded words to be bypassed around the corresponding repeater network 114, as will be described below. DATA is a pulse, the width of which is set by multiples of the artificial delay time input to query network 116. The transceivers in one or more adjoining cable sections are activated only by the simultaneous presence of a DATA pulse and an SI query pulse, both of which are signal pulses in their nature.

The interrogation network 116, the command network 118 and the repetition network 114 are provided with power failure shunt lines that are operated by relays, as will be described below. In the event of a power failure

S

10th

l s

20th

2s ok

35

40

45

50

55

60

(»5

617 017

12

the relays are deactivated so that incoming phase-coded words and query and command pulses are diverted around the defective transceiver.

Analog data are transmitted from the elementary seismic sensor groups 21 via local coaxial cables 58, 60 through amplifiers 120 and filters to the inputs of the multiplexer 122. Upon responding to an S1 pulse that is picked up and displayed by query network 116, control network 132 sets multiplexer 122 to channel # 0. Depending on the leading edges of a series of S2 pulses, the multiplexer becomes a normal scan cycle ordered sequentially so that it checks the input channels one by one. In the preferred embodiment there are 14 input channels. Channel # 0 is a blind or pseudo channel. Analog seismic data signals are handled through channels # 1-10. Analog signals from auxiliary sensors are transmitted through channels no. 11-13.

When the multiplexer 122 is reset to channel # 0, certain so-called organizational operations and test functions are performed: the gain conditioning amplifiers 124 are set to gain level 1, while the DC deviation is automatically removed from the multiplexer and the amplifier inputs. At this time, a fault indicator circuit, described below, also gives a warning if one or more of the three redundant data transmission lines are defective.

When each of channels # 1-13 is tested, the analog data signal under test is treated in the gain-conditioning amplifier 124. As is well known, seismic signals have a wide dynamic range of up to 120 db (1 million: 1). Signal gain conditioning involves compressing the dynamic range of the seismic signals to keep the range within the limits of the analog-to-digital converter. The signal conditioned in the gain factor is converted by the analog-digital converter 126 into a binary number which forms the sign and mantissa part of a number with a floating point. The gain states of the amplifier 124 conditioning the gain factor are encoded as a four-bit code. The four-bit code is combined with the mantissa in the output register 128, so that a number with a floating point from 10 to 16 bits resolution is created. The floating-point number represents the amplitude level of the seismic data signal at the time of sampling or testing. Four additional bits, including a parity bit, can be added to the data word as a preamble to ensure proper identification of the beginning of the phase-coded words.

The twenty bits that represent the data word are encoded in any convenient self-clocking code and are transmitted in the return-to-zero mode (RZ) over a broadband remote measurement connection in the direct digital data transmission mode. In a preferred embodiment, a self-clocking code, e.g. Two-phase M used. An example of an encoded data word is shown in FIG. 10. Self-clocking codes like the ones shown here

are described in "The Interface Handbook" by Kenneth M. True, published by Fairchild Instrumment Co., 464 Ellis Street Mountain View, California, on pages 4 to 18. Circuits using such codes are discussed in this publication. The lack of data is represented by a logic level of zero. Fifty nanoseconds before the first data bit, the logic level drops to - 5 V, so the first data bit must be a positive going pulse. Each data bit takes up a cell time of 50 nanoseconds. A binary "1" is represented by a polarity reversal in the middle of a cell time of 50 nanoseconds,

while a binary "O" is represented by no polarity reversal. Successive binary zeros are represented by successive 50 nanosecond polarity changes at the cell time limits. Since there are 20 data bits? a phase encoded word takes a time slot of 1000 nanoseconds, i.e. a millionth of a second. At the end of a phase-coded word, the logic level drops to —5 V for 75 nanoseconds and then goes to zero. The logic circuit in the repeater network 114 always looks for positive pulses within an interval of any two cell times. If no such pulse is detected, the logic determines the end of a data word.

The maximum phase change frequency of the phase encoded words is 40 MHz (for all ONE). Due to the fast rise time at the leading edge of the pulses, the bandwidth of the data transmission connection must be at least 100 MHz.

The inputs to multiplexer 122 are AC coupled through capacitors 123. The multiplexer output is coupled to amplifier 124 through series resistor 140 and buffer amplifier 142 to gain 1. The capacitors 123 and the series resistor 140 together with the multiplexer 122 form a high-pass commutation RC filter; is the cutoff frequency of the filter

25th

f = 1 / (2 RC) (D / T)

where D is the channel switch-on time and T is the channel switch-off time. The filter is described in more detail below in connection with FIGS. 15 and 16.

The operation of the preferred gain conditioning amplifier 124 in conjunction with analog-to-digital converter 126 and output register 128 to form a floating point data word is further described below in connection with FIGS. 39-42.

In a typical cycle of operation, fourteen analog channels of each transceiver 111 are tested within one scan cycle. A new scan cycle is run at a desired test speed, e.g. once every 4 »initiated half a millisecond. At a test speed of one millisecond, the fourteen channels of each transceiver are tested at 71.4 microsecond intervals. The completion of a scan cycle requires that the interrogator in the control device 4 of the common central station 2 of the ship 10 transmit one Si pulse and thirteen S2 pulses per millisecond.

FIG. 11 shows the time sequence of the interrogation signals, namely the S1 and S2 pulses within a sampling cycle of such a millisecond duration. When a Si polling pulse progresses along polling link IP1, IP2, IP3 in cable assembly 12 to transceivers 111 (Fig. 5), the corresponding multiplexers are reset to channel 0. On the other hand, a data word is clocked by the control network 132 from the output register and code converter 128 via the repeater network 114 into the data transmission connection D1, D2, D3. The phase-coded words are processed in the delay multiplex mode in the data connection due to the impressed delay of the 60 interrogation pulse between adjacent transceivers and the artificial delay time that is built into the interrogation network. The first S2 pulse is transmitted after 71.4 microseconds. The multiplexers in the transceivers 111 are sequentially 65 shifted to channel # 1 when the S2 pulse reaches each transceiver, and phase encoded words are again clocked into the data link from each transceiver in time order.

Additional S2 pulses are transmitted until all channels in all transceivers have been checked.

The sequence described above is shown in the time diagram according to FIG. 12. The interrogation pulses (IP) run outwards, i.e. downward in relation to the connection, right to left (the time base increases to the right) from the central station 2 to the transceivers 111 in the fifty connector modules 13A, 13B, 13C, etc., in that order. In the top three representations according to FIG. 12, the time therefore increases from left to right. An S2 pulse arrives at connector component 13A, for example, and shifts the multiplexer 122 of the transceiver to channel No. 1. After passing through the artificial delay line, the S2 pulse leaves connector module 13A 600 nanoseconds later. The distance between the transceiver in the connector module 13A and the transceiver 13B is 60 m. Assuming a propagation speed of 1.305 nanoseconds per foot in the interrogation connector IP1, IP2, IP3 with twisted pairs, the S2 pulse arrives at connector module 13B 256.8 nanoseconds later. The total delay between devices 13A and 13B is 856.8 nanoseconds.

As soon as the leading edge of the S2 pulse for channel no. 1 has been detected by the transceiver in the connector module 13 A, a phase-coded word from the register 128 is clocked into the data transmission connection D1, D2, D3. The data flow is from left to right (the time base increases to the left), upwards in relation to the connection to the central station. Similarly, when the S2 pulse arrives at connector chip 13B 856.8 nanoseconds later, the data word for connector chip 13B is clocked out. The signal travel speed in the coaxial cable, which represents the data transmission connection, is 1.542 nanoseconds per foot. Therefore, the leading edge of the phase-coded Wc ^ es arrives from connector block 13B on repeat network 114 in connector block 113 A 1160.3 nanoseconds later than the leading edge of the phase-coded value from connector block 13A has left repeater network 114. A phase-coded word separation of 160.3 nanoseconds is therefore provided.

A two-stage runtime sequence / channel sequence multiplex system is explained below. The phase-coded words transmitted from successive ones of the fifty transceivers in the connector modules are ordered depending on the propagation time of the interrogation pulse between the central station and the corresponding transceivers. Phase encoded words from the fourteen channels within each corresponding transceiver are ordered depending on the channel selection sequence during a scan cycle.

The DATA input from the central station to the command network 118 in Fig. 5 controls the operation of the system in the manner indicated above. The timing for the delivery of a DATA pulse allows either all sections of the seismic cable assembly to be used or only a portion thereof, for example the front half of the cable sections. As mentioned above, it may be desirable to move the elementary seismic sensor groups in the near half of the cable at a scanning speed immediately following the generation of a seismic pulse, i.e. Shot, to be checked randomly and then to check the signals from the entire cable at a different speed. The use of DATA pulses according to the length and timing can be done to perform these functions. In the description below, the required timing and length of the DATA pulses are general

13 617 017

explained so that any number of transceivers can be selectively activated.

Before examining the timing diagrams in detail, it appears appropriate to consider the entire data acquisition scheme 5 and the time frame in which the data generated in each of the five hundred seismic sensor groups is transmitted from the cable. First of all, it should be pointed out that each of the fifty cable sections has ten seismic sensor groups along its length and an associated io transceiver which processes the data from these ten sensor groups. At the command of the control unit 4 in the central station 2 on board the ship (FIG. 1), phase-coded words are transmitted from the last channel of each of the fifty 15 cable sections from the cable in sequence over the individual data link Dl by transmitting a wide S1 pulse , D2, D3 (which consists of three redundant coaxial cables). Thereafter, upon receipt of a narrow S2 pulse, the transceiver 111 associated with each of the fifty cable sections 20 transmits channel 1 information from each cable section in turn.

Then, after receiving another Sl pulse, channel 2 information is given from each of the fifty cable sections, etc.

; s In terms of timing, each cycle for sampling the signal present on all five hundred channels occurs for one millisecond. This cycle is defined as a sampling cycle and corresponds to the length of time between successive ^ jn Sl pulses, with each sl pulse .m being followed by thirteen S2 pulses before the next Sl pulse is generated. The transmission of an individual binary bit of a phase encoded word takes only fifty nanoseconds. For example, each phase-coded word is represented by twenty bits, and each phase-coded word is transmitted in approximately 35 thousand nanoseconds. Each one millisecond sampling interval is a thousand microseconds, so that there is sufficient time to transmit data signals from the five hundred seismic channels through the cable during each sampling interval, i. Sampling cycle, in a systematic manner, as will be described below.

The activation of one or more transceiver devices requires the simultaneous presence of an S1 pulse and a DATA (data control) pulse, as will be explained in connection with FIG. 13. A variety of cable sections are located remotely from the central station. At the leading ends of each section, connector modules 13 A-G are arranged,

each of which includes a separate transceiver, for example, assume that it is desired to have only the three successive transceivers in connector modules 13C, 13D, and 13E active, but no others. The circuit for performing this function is explained in more detail below in conjunction with FIGS. 43 to 48, but a brief description will be given here for a better understanding of the illustrated embodiment of the invention.

A Sl pulse is transmitted from the central station 2 via the interrogation connector to each connector module 13 in sequence. The moment of arrival of the signal pulse S1 at block 13A is t = 0, the arrival time at block 13B is t = 856.8 nanoseconds, the arrival time of the signal pulse S1 at block 13C is t = 1713.6 nanoseconds, etc. The six timing lines in Fig 13, designated by (, 5 IPA-IPF, represent the locations of the same S1 pulse with respect to connector modules 13A-F, each of which contains separate transceivers, at the end of every 856.8 nanosecond polling pulse transit time -

617 017

14

valles. Some time after an SI pulse has been transmitted, a DATA pulse is transmitted via the command connector (Fig. 5). The signal propagation speeds in twisted wire pairs that contain the interrogation and command connectors are the same. However, because of the 600 nanosecond delay line in each transceiver turned on in the interrogation network 116, the effective S1 pulse rate is lower than the command pulse rate because there are no delay lines in the command connector. Accordingly, a DATA pulse that has been delayed by 1200 nanoseconds with respect to a corresponding IP pulse receives the S1 pulse on the third transceiver in connector component 13C. The six timing lines, labeled DATA-AF, show the position of a data pulse with respect to S1 pulse at the end of every 856.8 nanosecond polling pulse transit time.

When an S1 pulse arrives at the transmitter-receiver device in the connector module 13A, this has no effect on the Buastein 13A because the DATA pulse is delayed by 1200 nanoseconds. The DATA pulse lags 600 nanoseconds at the transceiver in connector module 13B, so that there is no effect at module 13B. The DATA pulse intercepts the S1 pulse at the transceiver in connector module 13C, so that the transceiver in the connector module is deactivated.

At block 13D, the leading edge of the DATA pulse leads the S1 pulse by 600 nanoseconds, but due to the width of the DATA pulse, it is still available for activating the transceiver in connector block 13D. At the block 13E, the trailing edge has not yet passed the IP pulse S1, although the leading edge of the DATA pulse leads the pulse S1 by 1200 nanoseconds. Thus, the transceiver in the connector component 13E is activated. Finally, at the point in time when the SI pulse arrives at the unit in connector block 13F, the trailing edge of the DATA pulse leads the SI pulse. Therefore, the transceiver unit in the connector module 13F and all subsequent transceivers are not activated. All transceivers that are activated by the coincidence of SI and DATA pulses remain active for an entire scan cycle, so that they respond to all subsequent incoming S2 pulses. The width W of a DATA pulse is

W = [(L — 1) x DLY] + dt where L = number of transmit / receive signals to be activated

devices DLY = artificial runtime dt = a small amount of time increment at random length to enable smaller runtime differences.

In the example according to FIG. 13, the width of the DATA pulse is [(3-1) x 600] + 300 = 1500 nanoseconds.

The initial runtime ED given to the DATA pulse is

ID = M x DLY,

where M is the number of transceivers skipped between the central control station and the first active transceiver.

As discussed above, a DATA BYPASS pulse that coincides with an SI pulse is used to route data around a defective transceiver. The delay BD given to the DATA BYPASS-s pulse relative to an associated S1 pulse is

BD = Kx DLY

m where K is the number of transceivers located between the central station and the defective transceiver.

5, the data and query connections, which form part of the cable bundle 52, each consist of three parallel lines. If one of the lines is damaged, the other two are available. Majority elects two good lines. A majority selector circuit 131 is coupled to the input conductors of repeater network 114, and another such circuit (not shown) is associated with query network 116. The circuit 131 of FIG. 5 is shown in detail in FIG. 6 and consists of AND gates 136, 138, 140 and an OR gate 142. A logic 1, which is present on two of the three data conductors simultaneously, generates a logic 1 am Output of OR gate 142.

A fault indicator circuit 130 is coupled to the majority selection circuit in repeater network 114 and provides a signal in the event that at least one of the data lines D1, D2 or D3 is defective. The circuit (Fig. 6), n consists of a NAND gate 144, a diode 146, a storage capacitor 148 and a bias resistor 150. A negative voltage is applied to the fault indicator line 152 to keep it slightly negative when no output is off the gate 144 is present. When one of the lines D1, "D2, D3 becomes inactive, the output of the NAND gate 144 is opened. During a normal sampling cycle, phase-coded signals with a frequency of 40 MHz pass through the majority selection circuit and into the NAND gate 144, which is connected to the Inputs of repeater network 114. As long as all three lines are good, no output from NAND gate 144 occurs, but if one line is defective a 40 MHz signal is present at the output of NAND gate 144. The signal is through rectifies diode 146. The resulting rectified 4S voltage is stored in capacitor 148, generating a positive fault voltage on fault indication line 152. During the period in which multiplexer 122 is reset to channel # 0, switch arm 154, which is arranged between the amplifier 124 and the analogue, n digital converter (FIG. 5), moves from the contact 156 to the contact 160 and he allows analog-to-digital converter 126 to check the voltage level present on error display line 152. A mechanical switch 154 is shown in FIG. 5 for simplicity, but in practice, however, a bit rate Schottky field effect transistor switch can be used.

Marine streamer cables of the type described above tend to stretch up to 1% when towing. With a cable with a length of 3,000 m, the total extension is of the order of 30 m. The seismic sensor groups are offset to 6 m group centers within each cable section. With a extension of 30 m, the sensor groups are in the leading end of the cable arrangement 12 os shifted by almost five group intervals in relation to the sensor groups in the running end of the cable arrangement When assembling a larger group from a number of sensor groups, it is necessary to determine the exact one

15

617 017

Knowing the sensor group distance. If the distance is not known exactly, the effectiveness of the composite grouping will deteriorate significantly. The relationship between cable extension and drag voltage is known. Accordingly, a stretching strip 11 (FIG. 1) of a known type is expediently connected to the load components between the running elastic cable section 16 and the first active cable section 20. The output from the strain gauge 11 is fed to the input of an auxiliary channel, which is arranged in the transceiver in the connector component 13 'at the leading end of the first elastic cable section. From the knowledge of the trailing voltage, errors in the sensor group spacing that result from the cable extension can be corrected.

As is known in seismics, the individual hydrophones rarely have identical sensitivity. A deviation of ± 25% is not unusual. An arrangement is therefore provided to calibrate the hydrophones, an analog test signal of known amplitude is transmitted to the test control circuit 136 via the test signal line 162 (FIG. 5). A preferred test signal frequency is 15.625 Hz. A test control signal is transmitted via test control line 163 to test control relay 138, which moves switch contact arm 164 from contact 165 to contact 166. A test signal is now applied to drive the seismic sensor groups 21, each via a resistor 168. A normal multiplexer scan cycle is initiated to transmit the output from each sensor group 21 to the central station 2 in the ship 10. The amplitude of the output signal for each sensor group is compared with the test signal amplitude so that the calibration factor is obtained for each sensor unit. The sensor groups are calibrated at any time when seismic data are not recorded.

The test signal is used to accurately measure the sensitivity of all sensor groups throughout the seismic sensor cable assembly 12, which can be up to two miles long. Over such a distance, the test signal at the outgoing end of the cable is greatly weakened due to the ohmic voltage drop if the test signal control circuits 136 are connected in parallel to the test signal line. A resistor 167 is therefore placed in series with the test signal line 162 in each transceiver. The inputs of the test signal control circuit are connected in parallel to the series resistor 167; since all resistors have the same resistance value, all test signal control circuits receive identical input voltages. In this way, a test signal of constant amplitude is guaranteed for each transceiver.

An energy supply source 134 is provided in each transceiver 111. Energy is transferred from the ship to connector module 13 via a pair of twisted conductors 170, 172. Each energy supply source has a current transformer and a shunt regulator. Transformer primary windings in the corresponding transceiver devices in the connector modules 13 are connected in series. By crossing lines 170, 172 in each cable section 20, each transformer is alternately placed on an opposite side of the power line, as shown in Fig. 4, thereby maintaining a balance in the line load. Since the energy supply sources are connected in series, the voltage drop along the cable arrangement 12 between the ship 10 and the termination section 18 depends on the number of connector modules 13 which are connected to one another. In the case of a cable arrangement with fifty sections, the voltage drop is in the order of magnitude of 400 to 500 V. Energy is transmitted at 2000 Mz and 4A. This frequency is significantly higher than the normal seismic frequencies and therefore does not affect them. In the power source 134, the AC power from the power line is rectified and converted to ± 15 V and + 5 V for use in the logic circuitry in the transceiver. In the event of damage, e.g. of an open circuit in a transceiver, the voltage on the primary winding of the power transformer rises to a high value. A protective triac-i, i rectifier bridge circuit of known type short-circuits the primary winding when the voltage rises above a predetermined value. When the energy supply source is shut down, the fail-safe bypass relays (not shown) in the repetition network 114, query network 116 i s and command network 118 drop out due to non-appearance, as a result of which command impulses and phase-coded data words can bypass the harmful component.

As described above, the multiplexer 122 has fourteen inputs, of which channels Nos. 11 to 13 are used to: o transmit data from the auxiliary sensors, as will be explained below.

The pressure transducer 64, which has been mentioned in connection with FIG. 2a and is described in detail below, is a transducer known per se in seismics. Its function will be briefly explained below. As shown in FIG. 7 a, the transducer has a tubular spring 174. A soft iron pole piece 176 is connected to the movable end of the spring 174. The pole piece 176 moves in the longitudinal direction in a coil 178, which is attached to the fixed end of the spring 174 via a support arm 180. An oscillator, which contains an LC resonant circuit, is accommodated in a housing 182. The coil 178 is the inductive part of the resonant circuit. In a fluid, a change against the tube spring 174 causes the pole piece 176 to move in the coil 178 i-i, thereby changing the inductance and thus the frequency of the oscillator. The output signal of the pressure transducer 64 is therefore a frequency-modulated signal, the frequency of which is related to the surrounding fluid pressure. The signal is transmitted via a coaxial cable on channel No. 11 of the 40 multiplexer 122, as shown in Fig. 7b.

7b, a leak detection device 186 is provided, which indicates the presence of salt water within the outer skin 40 of a cable section 20. The leak detection device 186 consists of two wires 66, 68, which are poorly insulated by porous plastic material. The porous coating prevents physical contact between the wires, but allows the water to make fluid contact. The two wires 66, 68 extend over the length of the cable section 20. A wire 66 is connected to the oscillator output 184 of the pressure transducer 64. The other wire 68 is connected to the auxiliary input channel No. 12 of the multiplexer 122. As long as there is no water inside the cable section 20, no signal is given to the leak detection channel. When water penetrates the cable section 55, a conductive path is established between the two wires 66 and 68. An amplitude modulated signal then appears on the leak detection auxiliary channel. The amplitude of the signal is proportional to the resistance of the leak path.

A water interrupt indicator 72 is connected to the auxiliary input channel No. 13. This display device 72 is a special hydrophone that is used to detect an acoustic wave that arrives directly from the sound source along a path near the water surface (.s.

Although all active cable sections are identical and interchangeable, it is necessary to provide an impedance matching termination to ensure that the data

617 017

16

Command and interrogation signal lines, i.e. complete the transmission links on the last section to prevent unwanted reflections. Furthermore, the twisted, series-connected energy, test and test control conductor pairs must be provided with a feedback circuit. Correspondingly, a terminating section 18 is placed on the running end of the last cable section 20. The structure of the end section is shown in Fig. 8 a.

8a, the end wall 46 of the last cable section 20 is shown. The outer layer 40 is fixed to the end wall via steel strips 54, 56. Cable bundle 52 and plug 186 extend beyond the end wall 46 together with load components 42, 43 and pull hooks 45, 47. A corner swivel connector 188 is provided at the end of the terminating section 18. One end of three short load members 190, 192 (the third is not shown) is embedded in the connector body. Towing hooks 194, 196, which interact with towing hooks 45, 47, are fixed via bolts 198, 200.

The termination module 202 is fastened with load components 190, 192 via a steel band 204. The conductors, which are contained in a cable bundle 52, are electrically connected to the termination module 202 via plug 186 and associated plug 187. A plastic cuff 206 is slipped over the plug 188 and the end wall 46 of the last active section 20. The stuple is fixed in place by steel strips 208, 210, 212, 214. The space inside the cuff 206 is filled with light kerosene for buoyancy.

The electrical connections within termination module 202 are shown in FIG. 8b. Coaxial data transmission lines D1, D2 and D3 are terminated via resistors 216 with 50 ohm, 'h watts. The twisted wire pairs for the lines IP1, IP2, IP3 of the interrogation connection and the command wires for the signals DATA and DATA BYPASS are terminated with resistors 218 of 130 ohms, 'A watts. Power, test signal and test control lines are shorted by jumper wires 220. The feed cable section 17 is coupled to the leading end of the elastic section 16 via a transmitter / receiver unit in the connector module 13 '. The other end of the feed cable section 17 is fixed to the ship 10, which creates the possibility of towing the cable section 12 and also establishing a connection to the central station 2. The feed cable section 17 is shown in cross section in FIG. 9. It consists of a central load component 230, preferably a non-rotating steel cable with a diameter of 0.9375 cm. The load component 230 is encased with neoprene or other plastic 232. The current conductors, which form the cable bundle 52 in the active cable section, are spirally wound around the encased load component 230. The current conductors themselves are enclosed in a plastic sheath 234. The lines are shown in FIG. 9 as coaxial data lines 236, power lines 238 with a double twisted wire pair, command lines 240 with twisted wire pair and coaxial local data lines 242. Since the feed cable section 17 can be up to 180 m long, the data lines 236 RG / 59U are coaxial cables which prevent signal deterioration. Local data lines 242 extend from ship 10 to connector block 13 'where they are connected to the inputs. Local data lines 242 can be used to introduce into the system the signals formed by special sensors 222 (FIG. 1) near the ship 10.

To explain the filter used in the illustrated embodiment of the invention, reference is made below to Fig. 14, which shows a simplified representation of the digital-to-analog conversion system in which the multiplexer 122 has input channels Q, C2, Cn with a variety of signal sources, e.g. seismic sensors 21 is connected. The input channels are all connected to the multiple-xer output multiple line 312 via coupling capacitors 123, resistors 315 and switches 316. Each resistor 315, in conjunction with its capacitor 123, represents a high pass RC filter for its channel. Switches 316 are known high speed field effect transistor switches.

m The output of the multiplex line 312 is connected to the non-inverted input of a buffer amplifier 320 with gain 1, which is an operational amplifier, e.g. 5, with a high input impedance. The output of the buffer amplifier 320 is connected to an i5 test and hold circuit 322, which has a shunt capacitor 324 and a series switch 326. The output of circuit 322 is coupled to a variable gain binary amplifier 124 as shown in FIG. 5 and described in connection with FIGS. 39 and 42 further below.

The control network 132 is connected to the switches 316, 326 via a control multiple line 341, which has a plurality of control lines. The control network 132 sequentially switches the multiplexer 122 to connect the signal input channels C, -Cn to the output multiple line 312 sequentially during a multiplexer scan cycle in the manner described above. The control network 132 also controls the operation of the binary amplifier system 124 via the control lines 350.

m The amplifier system 124 is coupled to the analog-to-digital converter 126, which converts the tested analog signals into corresponding digital signals. Other networks which do not contribute to the essence of the invention have been omitted from FIGS. 14, 15 and 16 for the sake of a better overview.

In the circuit of FIG. 14, each RC high-pass filter has a series capacitor 123 and a shunt resistor 315 connected to ground. For use in seismics, each high pass filter has a low cutoff frequency f0 in the order of 1 Hz, although other cutoff frequencies can also be used.

The capacitance C is related to the resistance R and the cutoff frequency f0 by the following equation:

45 C = l / (2Rjtf0) (C)

For practical reasons of design, R must be relatively small, namely in the order of 10,000 ohms. Thus for f0 = 1 Hz the value of C = 16 is 50 microfarads. Such large capacitors are relatively expensive and bulky. The high-pass RC filters according to FIG. 14 are therefore not only expensive, but also prevent miniaturization.

The description of the filter arrangements according to FIGS. 15 and 16 is simplified by the fact that the same parts are provided either with identical reference numerals as in FIG. 14 or with identical reference numerals and a ('), so that the correspondence existing in certain points (, o becomes visible.

15 illustrates the possibility of using a single, common shunt resistor 315 'on the output line of multiplexer 122 instead of an individual shunt resistor 315 for each channel, os when multiplexer 122 is over its channel selection cycle, i.e. Sampling cycle is advanced, capacitor 123 ', which is in series with each channel selected by the multiplexer, e.g. Q, C2 etc. is connected in connection with the common

17th

617 017

shunt resistor 315 'represents the desired RC filter for this channel.

When the circuit according to FIG. 15 was tested, however, it was found experimentally (which was later proved theoretically) that the capacitance C is determined by:

C = [1 / (2 jtRfJ] [D / T]

(D)

where D is the dwell time, i.e. the length of time that each switch 316 is closed and T is the channel shutdown time, i.e. the length of time that each switch 316 is open. Dwell time D is a function of the number of analog channels and the total time required to sample all of these channels. For a multiplexer with sixteen channels and a sampling time of 1,000 microseconds, the maximum dwell time D is 62.5 microseconds per channel and the channel switch-off time T is 937.5 microseconds.

The dwell time D can be changed with the aid of the dwell setting circuit 360 according to FIG. 16. Dwell adjustment circuit 360 may be a monostable circuit, e.g. National Semiconductor DM 74121. A monostable circuit is a circuit or device that can be used to modify the duration of a control pulse by stretching or shortening the pulse width. The pulse width adjustment is achieved with the aid of a known resistance-capacitance feedback circuit. The desired dwell time remains set to the same value over a multiplexer sampling cycle.

Equation (D) can be written as follows:

f0 = [1 / (2 jcRQ] [D / T]

(E)

With constant values for R, C, T, the cutoff frequency of the filter can be changed by changing the dwell time. In the numerical example given above, the maximum dwell time for a sampling period of one millisecond is 62.5 microseconds. The shortest dwell time is determined by the acquisition time of the test and hold circuit 322 '. For example, the acquisition time is eight microseconds. The dwell time setting circuit 160 must be programmed to obtain a selectable dwell time range, for example ranging from a minimum dwell time of eight microseconds to a maximum dwell time determined by the sample duration, which is 62.5 microseconds in the example above. The cut-off frequency of the filter can thus be set to a desired value over a range of almost 8: 1.

The manner in which the improved multiplexer commutated RC filters are used in the seismic data processing device of the present invention is shown in FIG. Since FIGS. 14 and 16 are very similar, the description of FIG. 16 is limited only to the differences between the two representations.

The most important feature of taking advantage of the multiplexer channel selection cycle used in the multi-channel seismic data recorders is the combination with a common shunt resistor connected to the output of the multiplexer. Not only has the number of shunt resistors been reduced, but the size of the required capacitors 123 'has also been reduced by a factor of 15: 1 in the case of sixteen input channels. In addition, the cut-off frequency can be set by changing the dwell time over a range of almost 8: 1.

16 shows that it is expedient to use a low-impedance operational amplifier 320 '. The series resistor 315 ", which corresponds to the resistor 140 in FIG. 5, is the input resistance in the inverting input of the amplifier 320 '. The non-inverting input in the amplifier 320' is connected to earth. Since both inputs must have the same potential, the connection point 333 is practically at ground potential, which is why the filter according to FIG. 16 is the equivalent s to the filter according to FIG.

Accordingly, the low input impedance operational amplifier 320 'using the filter common resistor 315 "as the input resistance is the preferred filter circuit. The use of the io 320' operational amplifier eliminates the essential design limitations imposed by the amplifier 320 relatively high input impedance (Fig 14) are given.

It has been found that the use of a common resistor 315 "(or 140 in Fig. 5), which is successively connected 15 by switches 316 to capacitors 123, results in a considerable saving in the number of resistors required and that of the resistors and the capacitors take up space. These advantages are particularly important in miniaturized systems 20 for purposes of the present invention.

At this point in the description of the seismic data processing device according to the present invention, some general explanations appear necessary, and for this purpose reference is made to FIG. 17, which is a block diagram of the general structure of the multi-channel seismic device, which comprises the central station 2 and a multiplicity of Transceiver 11 11 - 11 In, which are arranged remote from the central station 2. The transceiver llla-llln are connected in series with the central station io 2 via broadband data transmission connections 1014a-1014n and interrogation connections 1016a-1016n, which correspond to the channels or connections Dl, D2, D3 and IP1, IP2, IP3 and they represent, these channels being arranged in the manner described above within a section of the seismic streamer cable.

Each transceiver 111 is shown as having the polling network 116 and a repeater network 114. Each repeater network 114 includes a number of input channels 1022, 1022 ', 1022 ". Three such channels 40 are shown in FIG. 17 for simplicity, but fourteen or more such channels can be used in practice, as previously discussed Seismic sensor groups 21, 21 ', 21 "are coupled to input channels 1022, 1022', 1022" and up to ten or 45 more such sensor groups are provided in each cable section, so that fifty or more such sections are five hundred or more individual elemental seismic Contain sensor units, each of which represents a separate information channel.

18 shows one of the transceivers in a detailed illustration; each analog channel 1022 is connected to a separate preamplifier and an "ALIAS" filter 1036, which is coupled to an input port of multiplexer 122 via a DC coupling capacitor 55123. Multiplexer 122 has input ports C1, C2, C3, C4, C5 (five channels are shown, but fourteen or more channels can be used), one for each channel 1022. The CO channel is the test channel. The channel CO contains a capacitor 1043, which is 60 connected to ground. The output of the multiplexer is connected to the operational amplifier 142 via the common series resistor 140. The capacitors 123, the multiplexer 122 and the common series resistor 140 form the above described, multiplexed commutated high-pass RC-65 filter for each channel.

The output of amplifier 142 is applied to test and hold circuit 1044, which has a series switch 1045 and a shunt capacitor 1046, which of the above

617 017 18

described arrangement corresponds. The output of the test arrives is guaranteed. A gap between data words and hold circuit 1044 is desirable with another test and hold adjacent transceiver to

Circuit 1050 connected via a buffer amplifier 1048. to have a proper distinction.

The buffer amplifier 1048 provides electrical isolation between the two circuits 1044 and 1050 as explained above. The test is 5 1000 nanoseconds long. The lines 1016a and 1014a and holding circuit 1050 have a series capacitor 1051 (FIG. 17) are each 60 m long and the rate of progression

and a shunt switch 1052 connected to earth through these lines is 1.6 nanoseconds / foot (5.33

is on. The output of circuit 1050 is at binary nanoseconds / m). Lines 1014a and 1016a require

Gain system 124 coupled. The test signal 1041, i.e. thus a pulse transit time of (120 X 5.33 nanoseconds / m)

Channel CO, in connection with the test and hold circuit m = 640 nanoseconds. The desired gap or dead center

1050 represents a DC offset elimination network, thus between successive data words is' A

the word length explained in detail in connection with Fig. 37, e.g. 250 nanoseconds. Thus the delay

is tert. The binary amplifier 124 with variable amplifier line 1029 (FIG. 18) is switched on for a delay time.

Degree of efficiency has four amplifier stages with double gain, which is given to degree of efficiency, namely amplifiers 124a, 124b, 124c and 15

124d, which are connected in cascade. Each amplifier has D = L + S - T (F)

usually a low gain state,

e.g. gain 1. The feedback can be entered so that L is the length of the phase-coded data word, S is that a discrete state of high gain desired word separation and T is the sum of the degrees of run for each amplifier, and this can be done by 20 a pulse through lines 1014a and 1016a. In

Switches 1055a, 1055b, 1055c and 1055d happen. The binary one numerical example is the artificial delay D amplifier arrangement 124 is described in more detail below in

Connection described with FIGS. 39 to 42. D = (1000 + 250 - 640) = 610 nanoseconds

The output of the binary amplifier arrangement 124 is the

Analog-to-digital converter 126 abandoned. This analogue 25 The mode of operation of the interrogation of each send / receive

Digital converter 126 receives an amplification device 111 through the interrogation connection 1016, dated analog signal sample and converts it to a binary number in connection with FIGS. 43 to 48 below in a directional pulse code. The exit described.

of converter 126 is sent to the intermediate output memory register. A data word consists of twenty bits, of which the bits ster and code converter 128 are supplied. The code converter m 1-3 the preamble, bit 4 the parity bit, bits 5-8 the exponent

128 converts the binary number of the directional pulse, bit 9 the sign bit and bits 10-20 which are Matisse code into a self-clocking, phase-coded pulse code. From the analog-to-digital converter 126, the data of the transmission connection 1014 via the line 1057, data in binary alternating code (NRZ code) are formatted

the regenerator 1060 and the transmitter are abandoned controls. This code is shown in Fig. 22 by diagram 6a. The data transmission connection 1014 is shown schematically. A bit interval is fifty nanoseconds. Since the data connection D1, D2, D3, which are present above with twenty bits, is table, the word length is 1000 nanometers.

Fig. 5 is described. Seconds. In the example shown, bit 9 is considered a

The timing functions of the various networks are assumed to be bits, while the rest are zero bits. Each transceiver 111 is controlled by the data. In code converter 128, the data is controlled in control unit 132, already mentioned in the phase control, which format-controls an interrogated NRZ pulse code. as shown in diagram 6b pulse from the query network 116 in series with the query of FIG. For a zero bit, there is a logic level transition on each bit interval connection 1016, which is schematically limited by connection IP1, IP2. For a one-IP3 shown above. A main control bit (e.g. bit 9) is a logic level transition in the middle of the interval device 1019 in the central station 2 transmits via the connection. In the absence of data, e.g. between data words, dung 1016 the wide Sl pulse, to which a number of 45 remains the signal level zero. The leading edge of the first bit of narrow S2 pulses within each scan cycle of a data word must always be followed by a positive pulse. With the shown thirteen input channels and the.

a test channel requires a sampling cycle to transmit a phase-coded NRZ data (diagram 6c in FIG. 22)

Sl pulse, followed by thirteen S2 pulses. are converted into a phase-coded data center (return to zero)

The interrogation network 116 consists of a pulse width-so pulse code through the transmitter 1066 for transmission into the condenser 1031 and a delay line 1029. When the current flow mode is converted into the data link 1014. Pulse width detector 1031 indicates a wide S1 pulse, “No data” is represented by a logic zero level, it provides a control SYNC pulse via line 1035. Fifty nanoseconds before the start of a data word controller 132 that drops the multiplexer 122 on the channel CO tes, the logic level drops to minus 5 V for a bit interval, causing a DC offset correction 55, which results in a positive leading edge for the first and the Transmission of a phase-coded data word is guaranteed by data bits. A polarity transition must be initiated from the last channel of the previous scan - V to + V (or + V to - V) at every bit interval. When the pulse detector 1031 is made a narrow S2 pulse limit for a zero bit. In addition, an On indicates it sends a pulse over line 1033 so that bit requires a polarity change in the middle of the interval, controller 132 will multiplexer 122 to the next one (e.g., bit 9). At the end of a data word, the logic level channel switches forward and the data from the preceding 75 nanoseconds goes on - V and then goes to zero, which the scan transmits. The S1 and S2 pulses step out. From diagrams 6b and 6c of FIG. 22 it follows that the delay line 1029 continues and results in an at least one positive pulse within two sufficiently high time delays so that there is a separation between bit intervals have to be. For a complete end of a local data word, as recorded by the transmitter description of various pulse codes, refer to The Interface Receiver purple and the handbook by Kenneth M. True.

Start of the data word that is linked from the next, downward. It is desirable to have a self-clocking pulse code, e.g.

those transceiver, e.g. unit 111b to use the one described above because the clocks 1108

19th

617 017

in each transceiver (Fig. 19) are asynchronous to each other, although they operate at identical frequencies. In the absence of a self-clocking pulse code, separate timing pulses would have to be assigned to the data word, which would complicate the system.

Data arriving at a transceiver 1 IIb (FIG. 17) from a transceiver 11 le below the connection is received by the data receiver 1068 (FIG. 18) in the repeater network 114. The regenerator 1060 either receives a local data word from the memory register and the code converter 128 or a data word below the connection from the receiver 1068. The regenerated word is fed via line 1063 into data transmitter 1066 for transmission over a data connection 1014 to the next transceiver 11a above the connection.

19 shows one of the repetition circuits 114 in detail. The receiver 1068 is a linear amplifier 1100 with positive feedback to achieve hysteresis. The amplifier 1100 amplifies incoming phase-coded RZ data that is received via the data link 1014 and, based on the hysteresis, converts it into corresponding phase-coded NRZ logic levels. The regenerator 1060 has an OR gate 1102 that is on the line

1103 provides either data from receiver 1068 or local data from line 1057 which is connected to the output of memory register 128 (Fig. 18).

A data detector circuit 1104 indicates a zero crossing, in particular the transition from negative to positive at the first bit interval limits, so that a data synchronization device 1106 is activated. The data detector

1104 also indicates the absence of data, such as the gap between data words, if no polarity transitions occur at the output of OR gate 1102 within a certain time interval. Thus, line 1101 only accepts data words from below the connection, while line 1103 accepts either local data words or data words below the connection.

Each transceiver 111 has a crystal oscillator or clock with exactly 80 MHz. The clocks in the corresponding transceivers are freely oscillating clocks and are therefore asynchronous to one another.

The phase-encoded data words are distorted due to noise interference and radio frequency degradation through the data transmission link 1014. Furthermore, the signal levels are weakened due to the line losses. It is therefore expedient to amplify and regenerate the data pulses at each transceiver, but in order to regenerate the data pulses at each transceiver, the data pulses must first be re-synchronized with the local clock.

Local clock 1108 (FIG. 19) synchronizes phase encoded NRZ data words in data synchronizer 1106. Line 1107 receives resynchronized data that is sent to transmitter 1066 under control of a control pulse from line 1112, which is connected to data detector 1104 ,. be sent. Line 1113 from data detector 1104 also sends control pulses to data synchronizer 1106.

The regenerator and synchronization network 1060 has flip-flops 1070, 1071, a counter 1072, a crystal oscillator, which is the clock 1108, and a component 1074 that is divided by two. First, flip-flop 1071 is reset, thereby setting its Q output to ZERO, which resets the Q output of flip-flop 1070. The divider 1074 is reset so that no clock pulses are transmitted over line 1109 to the counter 1072.

When the first bit of a data word is received on line 1103, the first data pulse on line 1103 becomes a positive going pulse, causing a change from ZERO to ONE. This positive going pulse pulses the Q output of flip-flop 1071 to a ONE, thereby releasing flip-flop 1070 and divider 1074 via line 1113. Divider 1074 transmits a 40 MHz clock pulse over line 1109. The clock pulse that is closest in phase with the positive m data pulse on line 1103 triggers flip-flop 1070 so that the logical ONE at the D input is transmitted to the Q output as a regenerated, newly synchronized data pulse.

The data detector 1104 has the counter 1072 so that! 5 the presence of a data word on line 1103 is tested. The first positive pulse of a data word on line 1103 results in a reset of counter 1072. Clock line 1109 sends a gated 40 MHz clock pulse to counter 1072. The counter counts three in clock pulses. Since the pulses occur every 25 nanoseconds, the three counts take 75 nanoseconds, 25 nanoseconds being more than one bit interval. If the counter is not reset by a positive data pulse within the 75 nanosecond interval, the 25 counter 1072 will expire after three pulses. When counter 1072 expires, it sends a reset pulse over line 1114 to reset flip-flop 1071 and causes line 1113 to go zero, resetting flip-flop 1070 and divider 1074 and transmitter 1066 over m line 1112 is disabled.

20 shows a set of timing diagrams that explain the operation of regenerator 1060. The first diagram 1200 shows binary data that is format controlled as a phase encoded NRZ data word. The bit interval is 50 i5 nanoseconds. There is no polarity reversal between the bit interval boundaries for a ZERO bit, while for a ONE bit the polarity changes in the middle of the interval. A polarity reversal must occur at every bit interval limit. The bit interval limits are indicated by arrows above the 4d diagram. The bit values are written between the parts. A complete data bit consists of 20 bit intervals and is therefore 1000 nanoseconds long. The diagram shows that there must be at least one positive going pulse within two bit intervals. 45 The function of the data detector 1104 (FIG. 19) is based on this feature.

The second diagram 1202 of FIG. 20 represents a sequence of 80 MHz clock pulses 1224. Diagram 1204 shows the waveform of an incoming data word that has been translated by the 50 receiver 1068 from the phase-coded RZ pulse code into the phase-coded NRZ pulse code. Due to noise and interference, the original impulses shown in diagram 1200 have deteriorated. The recorded impulses are out of 55 phase and asymmetrical.

19 and 20, the first positive transition 1220 diagram 1204 of the data word clocks the Q output of flip-flop 1071 to a ONE. The logical 60 ONE on line 1113 generates a control signal 1222 to divider 1074 and flip-flop 1070 timing diagram 1206. The next 80 MHz clock pulse 1224, which is in phase closest to the rising bit edge 1220, becomes the first 40 MHz clock pulse 1226 - Timing diagram 1208. The clock pulse 1226 (, 5 clocks the logical ONE at the D input of flip-flop 1070 into the Q output via line 1107 as the leading edge 1228 of the first newly synchronized, regenerated data bit. So long a logical ONE on lines 1112 and 1113

617 017

20th

is present, a transmit control signal 1230 - timing diagram 1212 - enables the output transmitter 1066. The data bits on line 1107 can then be converted from phase encoded NRZ data to phase encoded RZ data for transmission to the next transceiver above the link. When the first rising bit edge has been re-synchronized, the 40 MHz clock pulses are re-synchronized and regenerate the remaining data bits. The timing diagram 1214 shows the output from line 1114 of counter 1072. If a data word remains zero over three 40 MHz clock pulses, counter 1072 sends out pulse 1232 to reset flip-flop 1071 and counter 1074 and transmitter 1066 to render ineffective.

Transmitter 1066 (FIG. 19) has a voltage bipolar to current converter 1076 and an NRZ to RZ converter 1077. In order to reliably transmit data via a transmission connection, e.g. To send a coaxial cable, the signals should be AC signals or low-frequency components. Preferably, constant current transmission is used to avoid the need to use broadband dynamic receivers 1068. Voltage losses per unit length of data link 1014 result in a significant signal weakening. A transmission in constant current mode results in essentially constant voltage levels: at the terminating resistor 1150 at the input of the receiver 1068, regardless of losses in the data connection 1014 between adjacent transceivers 111.

In order to achieve transmission in constant current mode, the phase-coded data words are converted from a voltage into a constant current by the current converter 1076. The logic voltage gate 1075 has both inverting and non-inverting outputs 1075a and 1075b. Resistors 1073a, 1073b (which have the same value R) and transistors 1078a, 1078b convert the voltage levels of the logic gate 1075 into constant currents.

The bases of transistors 1078a and 1078b are set at a voltage Vt, where Vt is equal to the ZERO state of outputs 1075a and 1075b of the logic gate. The exact voltage of the ZERO state depends on the type of logic selected. The voltages VR that appear across resistors 1073a and 1073b are output voltage states (Va-V,) and (Vb-Vj). That is why the current is through the two resistors

A

Ihoch = (VR, ONE state - V,) / R,

or

Low = (VR, ZERO state - VJ / R = O

The output current in data link 1014 will therefore

I ^ high 1fr low *

that in

50

55

Jfrhoch ^^ nii low for a bit interval polarity transition changes (Fig. 22). Since I low and I low = zero, the data transmission link 1014 current becomes high, -Ib high on a bit polarity transition. These data are in bipolar NRZ current operation. At the end of a data word, a low-frequency ft5 component results due to the time constant of the transformer 1079 and the coaxial data connection 1014. In order to prevent the low-frequency component, the NRZ signal is in

RZ converted by circuit 1077. When the transmitter is activated by a pulse on line 1112, switch 1077a is applied to voltage V1. For simplicity, switch 1077a is shown as a mechanical switch; in practice, however, it is a solid-state switch, e.g. a transistor (field effect transistor). When the end of a data word has been detected by data detector 1104 (FIG. 19), a pulse through control line 1112 connected to data detector 1104 causes switch 1077a to switch to V2. V2 is equal to the ON state of outputs 1075a and 1075b, so that transistors 1078a and 1078b are turned off. The current in transmission link 1014 goes to zero, so that the bipolar NRZ output of transformer 1079 is converted to the bipolar, i-phase encoded RZ pulse code. The windings 1079a, 1079b, 1079c of the transformer 1069 have a turn ratio of 1: 1: 1.

17 and 21, the central station 2 has a data receiver 1028 which receives the phase-coded RZ data) words which are transmitted by the transceiver devices 111 via the data transmission connection 1014 which the triple redundant coaxial connection D1, D2, D3 in 5 is can be transferred. Data receiver 1028 first translates the phase-coded RZ data words into a binary NRZ pulse code as floating-point numbers, and then converts them from floating-point numbers to fixed-point numbers that represent the data words. The fixed comma data words are arranged in a channel sequence matrix in an allocation memory in the group shaping device i 1030, as will be explained below. Each elementary seismic sensor unit represents a subgroup, since it has three individual detectors in the case of the described embodiment. The output signals from a number of such sub-groups are combined in the group shaping device 1030 in such a way that they produce a new composite signal characteristic of a much larger, i.e. form an extended group. The composite signal formed is transmitted to a format control device 1032 and from there to a magnetic tape device 1034, in which the signals are recorded so that they can later be used for a seismic cross section of the layers of the earth.

At the beginning of a seismic operation, a calibration address operation is initiated at the central station 2. An interrogation pulse is sent from the main controller 1019 via the interrogation link 1016 to the corresponding transceiver 111. Since each transceiver indicates the polling pulse, a data word is sent back to the data receiver 1028 over a data link 1014. A time converter 1015 measures the length of time between the transmission of the calibration interrogation pulse or the arrival of the resulting data words from the transceiver 111. Of course, a time delay is associated with each of the data words from each transceiver. The time delays are encoded as binary numbers to form an address code and are stored in the address memory 1017 over line 1027. During a normal operation, the time converter 1015 measures the interval between corresponding scan pulses and the returning, phase-coded data words. The time intervals are compared with the stored address codes. The address memory 1017 then identifies each recorded data word with respect to the originating transceiver and causes the group shaper 1030 to insert the recorded data word into the correct location in the matrix in a core memory.

21 shows a schematic illustration of the group shaping device 1030. This group shaping device 1030

21 617 017

has a control unit 1081, an allocation memory 1083a, the seismic data channel signal in the scan data channel storage

1083b, a memory write controller 1037, a data channel, rather 1047 is read out in the data channel sequence, the scanning memory 1047, a read controller 1049, a multipli data channel 1 being the first and the data channel 10n being the last.

Ornamental circuit 1053a and 1053b a coefficient fixed value- The data channel signals are on the data line

Memory (ROM) 1055, adders 1064a and 1064b, 5 1082 to multipliers 1053a and 1053b.

Accumulator register 1065a and 1065b, an output memory gen. Under control of the control unit 1080 and the program

1080 and a microprogram memory (ROM) 1067, group pattern that is stored in the control memory 1067

which is read-only storage. chert is, each data channel signal with the corresponding

The receiver 128, as stated above, multiplies the weighting coefficients stored in the coefficient-phase encoded data words from the transceiver memory 1055. When the signal from a line 11 la-11 In below the link and converts If the data channel is not to be used in a group, the phase-coded RZ data words in NRZ binary numbers with its coefficient will be zero and thus is the result of the multi-floating point order, which again results in numbers with a fixed zero. The evaluated data channel signals are then format-controlled by commas At the same time, in adding devices 1064a and 1064b to the outputs of the reception of the data words identify the time counters or converters, 5 add the registers 1065a and 1065b, the accumulate -1015 and the address memory 1017 accumulate each number in relation to the results in the registers 1065a and 1065b correspond to their origin in the form of the transceiver -Number of the sum of the evaluated data channel signals. The control unit and the data channel number. The fixed comma numbers from 1081 are transferred depending on the memory 1067 and the receiver 1023 are transferred to the group mover, contents of registers 1065a and 1065b are output to an output register 1030 and stored in the map memory 1083 of 2u1080 as a data sample for each composite group, group mover where they are written to the data channel - if the pre-programmed number of individual sequences under control of the control circuit 1081 and the data channel signals forming each composite group - write controller 1037 after one has been summed. The process continues until all data channel identification via the address memory 1017 data channels in the scan data channel memory 1047 has taken place together. 25 set group signal samples processed and on the

The data channels are numbered from 1-500, with 1080 being transmitted with the output memory. The content of the first seismic sensor group is started, which the output memory 1080 then contains a sample together.

next signal signals for the composite groups are placed below the connection.

catching device purple is assigned, and that with the last pen. Auxiliary channel duration are similarly under control

Seismic sensor group ends that are the most distant of the control unit 1081 format-controlled.

distant, below the connection arranged send-Emp. For example, assume that a seismic trap Hin is arranged. The phase-coded cable arrangement five hundred individual short sub-groups

However, data words are not owned in numerical order from the. The output signals from successive sentences

Data transmission connection 1014 established, they can be combined from twenty subgroups in such a way that, as explained above, as transmit / receive channel 1, all 35 composite signals are generated, the twenty-five

Transmit-receive devices 11 la, 11 in, then represent much longer groups as transmitters. By using

Receive channel 2 from all transceivers double multipliers 1053a, 1053b, double addie

llla-llln etc. In the data channel numbering, channel 1 will be 1064a, 1064b and double registers 1065a, 1065b may assign purple channel 1 to the transceiver form successive groups,

Channel 1 becomes data channel 11, transceiver with adjacent groupings, i.e. Sentences, for generation

111c, channel 1, becomes data channel 21, and transceiver pre-composed signals that are fifty overlapping, long

direction lln, channel 1, data channel 1 + 10 n. The transmitters represent groups, will be connected.

Receiving channels 0, 11, 12 and 13 are described above. The content of the output memory 1080 is sequentially opened

Embodiment auxiliary channels. Therefore, the function of the 45 is the format controller 1032 and then on the record

Allocation memory 1083a (Fig. 21), the data are transferred with fixed device 1034. If this processing

Comma arriving from the receiver 1028 in which the cycle for the first scan is completed will put together the correct data channel sequence in a memory matrix - next scan of new data in the memory 1083a and 1083b and rearrange the auxiliary channels in their correct order in one. When the processing of the first ends

Separate auxiliary channel location 1083b. If a scan, as determined by controller 1081,

of the seismic data and the data in the auxiliary channel, the new data for scanning the data control

Mapping memory 1083a, 1083b has been transferred to chers 1047, where processing to form a data channel signals from the mapper memory will sample the second sample of signal patterns for the next one.

1083a into data channel scan memory 1047 and the auxiliary channel set begins.

data 1083b is written into the output memory 1080. 55 The process described above continues until the

The digital numbers, as stored in the map memory 1083a recording and processing cycle, are the signals from the five hundred seismic sensor groups defined by controls in the main controller 1019, each of which is a short sub.

group forms. The control memory 1067 and the coefficient- The operation of the system described so far is like

Memories 1055 are preprogrammed to follow the digital M): A number of transceivers 111

Combine signals from selected sub-groups to form a spaced at regular intervals with respect to the central composite digital signal that characterizes station 2. Each transceiver has a table for a much larger desired group. The fourteen input channels with which seismic sensor sub-

Evaluation or the weight that the digital signal of an individual group, as well as a channel selection device, namely the duel subgroup to the entire, composite t, s multiplexer 122, and a common signal conditioning elec-

Contributes digital signal is connected through the coefficient memory tronics of the transceiver. From

1055 controlled. The methods and techniques for using the fourteen channels are ten data channels. The rest of the evaluation coefficients are explained below. Test and auxiliary channels. The output of the common signal

617 017

22

Conditioning electronics is connected to repeater network 114. The polling network 116 is arranged within each of the transceivers.

The repeater networks 114 of the transceivers are all connected in series and connected to the data receiver 1028 in the central station via the broadband transmission link 1014. The polling networks 116 are all connected in series and connected to the main control device 1019 in the central station 2 via the polling connection 1016, which is the triple redundant line IP1, IP2, IP3 of FIG. 3.

Periodically at the beginning of each sampling interval, e.g. every millisecond, the main controller 1019 sends out a wide SI pulse, which is a polling pulse, over a polling link 1016. Since the polling network 116 identifies the S1 pulse in each transceiver 111, the channel selector is reset to channel # 0. The digital data word from the last channel of the preceding scan is converted into a self-clocking, phase-coded RZ data word and is sent via the transmitter 1066 of the repeater network 114 of the transmission link 1014 for transmission to the data receiver 1028 in the central station 2. An S2 pulse is then transmitted from the main controller 1019 exactly 71.4 microseconds later. When each transceiver picks up and identifies the S2 pulse, the channel selector, i.e. the multiplexer 122, switched to channel No. 1. The seismic signal that is present in channel no. 1 is checked, conditioned and digitally represented as a phase-coded NRZ digital word as a local data word. The local data word is converted into a phase-coded data center data word and is given to the data transmission connection 1014.

Each transceiver first transmits a local data word. Then it awaits the arrival of a distant data word from the next transceiver downlink, and then the arrival of a data word from the downlink transceiver adjacent to the next, etc Each remote data word that is picked up in this way is regenerated and retransmitted with respect to the connection to the data receiver 1028 in the central station 2. Thus, the transceiver 11a transmits the local data word and then sequentially picks up, regenerates and transmits removed data words from the other forty-nine transceivers (if there are fifty transceivers in total). The last, the fiftieth transceiver, of course, only transmits its own local data word.

In order to test all channels in all transceivers, one S1 pulse and thirteen S2 pulses are thus transmitted during one scan cycle. Thus, data words are transmitted into the data receiver 1028 in a time-sequence / channel-follow two-stage multiplex cycle. This means that data words from the corresponding transceiver 11 a-111 n are separated depending on the query pulse transit time between successive transceivers. An artificial delay is introduced into each query network 116 to ensure separation. Data words from the corresponding channels within each transceiver are separated from one another depending on the channel selection sequence.

When the data words arrive at the data receiver 1028, they are directed to the allocation memory 1083a, where they are ordered in a channel sequence matrix. The first data channel of the first transceiver purple is channel 1. The last data channel of the last transceiver

device 11 In is channel 500. Thus, data words from data channels 1, 11, 21, etc. are recorded as a function of the first query pulse. Depending on the second interrogation pulse, data words from data channels 2, 12, 22, etc. are recorded, and so on.

In the central station 2, the control memory 1067 is preprogrammed in such a way that it extracts signal samples which come from selected sets of seismic subgroups from the allocation memory 1083a. The selected signal samples io are transferred to the data channel sample memory 1047. From the memory 1047, the signal samples are transferred to the multipliers 1053a 1053b where they are multiplied by selected weighting coefficients under the control of the coefficient memory 1055. The 15 samples evaluated are then combined in adders 1064a, 1064b and form a single, composite signal sample which represents a much larger group with preselected properties. The composite data is stored in output memory 1080 for subsequent transfer to format controller 1032 and subsequent recording in recorder 1034, e.g. stored on a magnetic tape.

As a result, the previously described seismic data processing system has a single seismic cable that enables the formation of a desired seismic sensor group configuration from a plurality of sub-groups. It is not necessary to physically change the seismic cable or other components of the data processing system on site in order to obtain different groups that are able to cope with changing geological conditions.

Another specific feature of the described embodiment of the invention will be explained below in connection with FIGS. 23 to 35.

FIG. 23 shows, similar to FIG. 1, a schematic illustration of seismic exploration arrangements, which show an embodiment of a seismic exploration or exploration system. A ship 10 with a large cable reel 2054, which is attached to the stern, tows a long sea seismic cable 2056, which is unwound from the reel 2054. The seabed is indicated by the horizontal line 2058 and various geological strata interfaces are indicated by horizontal lines 2060, 2062 and 2064.

Cable 2056 has a first section 2056 'which is .15 closer to ship 10 and a distant section 2056 "which is further away from the ship. As is common with marine seismic cables, a large number of seismic sensors are included in this As mentioned above, the cable may be 3,000 meters long, for example, 50 and 500 sets of seismic sensor groups, each set containing three interconnected seismic sensors, in which arrangement successive sensors are between 1.8 and 2.1 m, preferably 1.875 m apart, and each sensor group consisting of 55 three sensors has a length of approximately 3.75 m, the distance from center to center for adjacent sensor groups being 5.625 m.

23 shows various aspects of a special feature of the illustrated embodiment, which will now be described in detail below. First, the group of sensors located at location 2066 will be explained. The sensor group at location 2066 picks up reflections that occur in succession, after which an explosion charge is detonated at location 2068 near the (.5 stern of the ship and near the adjacent end of cable 2056. Following the initial pulse Known as the "first break" and usually directly through the top layer of the

23

617 017

Water travels to the sensor group 2066, the first reflected seismic wave is that from the sea floor 2058. This first reflected signal, which is recorded at the sensor group 2066, travels over the relatively short path 2070. The beam or path 2070 forms when striking the Sensor group at point 2066 with the horizontal a relatively flat angle 0 ,. Subsequent reflections from geological interfaces 2060 and 2062 follow paths that include lines 2072 and 2074. These rays successively form larger angles © 2 and © 3 with the horizontal. Thus, the direction of the signals striking the sensor group 2066 changes during the recording of the seismic reflections and in particular the angle of the recorded signals increases with time with respect to the horizontal. Furthermore, the signals reflected along beams 2070, 2072 and 2074 and sequentially picked up by sensor groups 2066 may be viewed from a simplified standpoint as the representation of a trace of reflected signals from points along vertical line 2075 (including points 2070 ', 2072' and 2074 ') can be viewed.

23 shows an additional sensor group 2076 at the distal end of the seismic cable. For reasons of convenience, the inner half of the sea seismic cable is designated by the reference number 2056 'and the outer half by 2056 ", the sensor group 2076 being arranged at the outer end of the removed, outer half of the cable 2056". At a later time, the sensor group 2076 picks up signals that are reflected by the geological interface 2062 along the path 2078. At an even later time, the sensor group 2076 picks up reflections along the path 2080 from the deep geological interface 2064. In particular, it should be noted

that the sensor groups 2066 and 2076 record signals from a specific layer with different angles and at different times.

As with other types of wave progression analysis, the waves can be represented by moving spherical wave fronts or by rays that form part of such a moving wave front. Furthermore, the beam shapes 2072 and 2074, which are shown in a straight line in FIG. 1, are refracted at the interfaces 2058 and 2060 with angles that are related to the physical properties of the layer, as is known in the art and shown in FIG. 1 .

Of course, exactly horizontal geological interfaces, such as the horizontal interfaces 2058, 2060, 2062 and 2064 shown in FIG. 23, are unusual and not particularly interesting for geologists. Of greater interest are geological anomalies such as Fault, clumps and other tilting, sloping or "collapsing" geological interfaces. FIG. 23 shows the inclined plane 2082, which gives a positive angle of incidence relative to the horizontal plane 2064. The dashed line, which shows a plane of incidence 2082, represents the object which is explained in connection with FIGS. 24 and 25 below.

Seismic path 2083 (FIG. 23), which is shown in dashed lines, meets incident level 2082 at location 2083 '. In the case of a plane that is incident downward from the direction of movement of the ship, using the geometry of FIG. 2, the resulting seismic signal, which is displayed on the sensor group 2076, is compared in intensity with the signal which is caused by reflection a plane that is tilted or collapses in the opposite direction.

This is explained in detail in connection with FIGS. 24, 25 and 30. However, on a qualitative basis,

that conventional seismic groupings are more sensitive to signals that arrive vertically at the grouping and are less sensitive to waves that arrive at relatively small angles of incidence. Furthermore, this increased sensitivity is more pronounced for conventional seismic groupings at higher acoustic signal frequencies. As shown in FIG. 23, seismic path 2083 is directed against sensor unit 2076 at a smaller angle of incidence to the horizontal than path 2080. If the plane 2082 is tilted in the opposite direction, the angle of incidence at the sensor unit 2076 is even closer to the vertical and the response intensity would be increased accordingly. This phenomenon is considered on a quantitative basis further below in connection with the description of FIGS. 24, 25 and 30. 15 In the present description, the difference in seismic monitoring results obtained in one direction of movement is compared to the opposite direction of movement when performing a sea investigation. As stated above, this difference is due to the difference in the direction of transmission of the seismic energy that arises from the ship for marine exploration with a seismic pulse. In systematic land surveys, of course, the seismic impulse can be initiated from various locations, including 25 locations on the back or on the front of a linear cable arrangement along the traverse. Applying the present analysis to land surveys, the location of the seismic pulse source relative to the seismic cable is a determining factor.

24 is a graphical representation of the relative response to reflected signals from an unguided sensor group with the known configuration shown in FIG. 26, located 300 m from the launch site along cable 2056 in FIG. 3, at various indicated-.15 showed frequencies and following a time interval of one second from the firing along the reflection path to the sensor group. The response at 200 Hz is identified by consecutive x, at 100 Hz by consecutive +, at 50 Hz by small triangles and at 20 Hz by small circles. 4 (i The one-second time interval, together with the speed, determines the penetration depth of the reflected signals. In FIG. 24, the speed is taken from the high-speed plot according to FIG. 28 and is therefore 1800 m / sec; the penetration depth is accordingly approximately 45 900 m In Fig. 24, the layer tilt angle corresponding to the tilt angle between the dashed line 2082 and the horizontal line 2084 in Fig. 23 is plotted in the horizontal axis. At an offset of approximately 300 m from the launching site, it follows that the maximum Response occurs at all frequencies at a layer tilt angle of approximately -10 °, which would reflect the seismic waves from the launch site approximately vertically against the sensor group.This maximum response condition is represented by vertical line 2084 in FIG.

55 Most of the seismic work so far has been at relatively low frequencies for the reasons indicated in Fig. 24. It should be noted here that on the second vertical line 2086 in FIG. 24, which corresponds to a positive tilt angle of 10 °, in the direction shown in FIG. 23 with the tendency to transmit signals to the seismic sensors at a more horizontal angle of incidence to reflect, practically no energy is absorbed by the sensor group at 200 Hz.

It should also be noted that along the vertical line 2086, the energy absorbed at 100 Hz has a level of (.5 about -18 db and relative to the normal amplitude, which is indicated by the 0 db level on the top of the graph in FIG. 24 25 is a more extreme condition than that in FIG

617 017

24th

shown for an unguided arrangement. It applies to a distance of approximately 1200 m from the terminating point to the sensor receiving point, a relatively low speed of approximately 1500 m / sc, a characteristic of water or near-surface materials in some parts of the world, and a period of corresponding to reflections from a relatively shallow geological layer in just one second. As can be seen by referring to vertical line 2088 in Fig. 25, even at a tilt angle of + 10 °, even the 50 Hz seismic signals are cut off and only the very low frequencies, e.g. the 20 Hz frequencies shown by plot 2090 are indicated by the sensors. Certain 200 Hz side loops are shown at 2092, 2094 and 2096. However, these do not provide important information because they are faulty or out of phase or have other anomalies.

24 and 25 illustrate in detail the problems encountered with fixed unguided groupings of the type discussed in some of the publications mentioned above.

In particular, FIG. 26 shows a group of twenty-six elements that use a uniform evaluation of the sensor inputs and a variable distance. The total length of the group is 63 m, and the distance is given by the following numerical values: ± 3 ', ± 8', ± 14 ', ± 19', ± 25 ', ± 30', ± 38 ', ± 44', ± 52 ', ± 61', ± 71 ', ± 80', ± 105 ', the twenty-six elements being offset from the center of the group by the number of meters specified.

In Fig. 26, the uniform height of lines 2028 indicates the uniform evaluation of the sensors, the horizontal locations indicate the relative distance of the sensors along the seismic cable. The resulting sensitivity is symmetrical about the vertical center line, is relatively wide and only differentiates against horizontally running waves. The sensor group of FIG. 26 was used in the creation of the diagrams according to FIGS. 24 and 25.

The group of FIG. 27 consists of ten sensor groups, each of which contains three detectors. 27, the sensor groups 2102 and 2104 have a rating of "1" at the respective end in comparison with an increasing rating of 2,3,4,5 for the sensor units towards the center of the group, the two Sensor groups 2106 and 2108 have ratings of «5» in the middle. The elements of the group are arranged at a uniform distance from each other and extend over a total distance of 69 m from the first sensor to the last sensor.

This slanted group according to FIG. 27 has a response characteristic with a fairly sharply defined main loop. As will be explained further below, the beveled configuration of FIG. 27 can be used in conjunction with the aspect of the invention discussed herein.

Fig. 28 is a graph plotting speed in meters per second versus duration of reflection in seconds. In Fig. 28, the low speed line 2110, designated VL, appears as a horizontal line indicating a constant speed of 1500 m / sec. This is the speed of the seismic waves in water or near the surface of the earth; it is particularly important for a wave that runs almost horizontally in water. However, the high speed line 2112, designated VM, increases in speed with increasing depth through the earth (as opposed to water conditions). With longer reflection periods, the quadratic mean of the speed in meters per second increases sharply up to a maximum value of four seconds

Reflection time of 3300 m / sec too. The high speed line represents the actual speed conditions in many parts of the world. The analysis of FIG. 24 is based on the high speed characteristic VH of FIG. 28, while FIGS. 25 and 30, which represent more extreme conditions, on the characteristic VL 28 are based on the low speed.

According to another important feature of the embodiment of the present invention, a large number of groups are formed in the longitudinal direction of a seismic cable, and each of these groups can be directed individually so that it is sensitive to seismic reflections that arise from different, predetermined depths of a geophysical terrain, that is being examined, being reflected. This can be achieved by initially transmitting a large number of seismic signals from groups of seismic sensors that are staggered along the cable. Then a large number, perhaps 30.50 or more groups at 30.50 or more spaced locations along the cable are determined from the signals picked up by the sensor groups. These groups, which are staggered along the cable, are directed to sense seismic signals from selected depths along adjacent vertical lines of the geophysical area under investigation. This directional sensitivity can be achieved by appropriate delay signals that come from neighboring sensor groups. The seismic signals are then combined along vertical lines so that they produce a cross-section or a composite geophysical measurement of the terrain to be investigated.

In order to achieve the correct delay value between signals from neighboring sensor groups which form a group, it is important to determine the differences in the arrival times of seismic signals at the neighboring sensor groups. Figure 29 and the mathematical study below show how this delay can be calculated. In FIG. 29, a seismic pulse from the launch site 2116 is reflected from the geological boundary layer 2118 into the sensor group 2120, the seismic signals traveling along the beam paths 2122 and 2124.

In the following mathematical analysis, the letters “s”, “x” and “d” refer to the distances and points shown in FIG. 29. For convenience, point 2116 is reflected in point 2126, which is the virtual image of point 2116 relative to plane 2118.

The following definitions apply to the mathematical examination:

x - distance of the sensor group from the termination point v - traveling speed of the seismic wave d - depth of the reflecting boundary layer s - path length of the reflected signal t - transmission duration of the reflected seismic wave along the path t0 - transmission duration of the reflected seismic wave over the distance 2d.

The following applies s = vt (G) in the geometric representation according to FIG. 29, s = Vx2 + 4d2 (H)

s = vt = Vx2 + v2t02 (I)

Squaring this equation reveals that t and t0 are given by the following expression:

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

25th

617 017

v2t2 = x2 + v2t02 (J) on the low speed line 2110 (VL) of Fig. 28,

and is therefore fully comparable to FIG. 25. In Fig. 30, similar to 0 (as in Figs. 24 and 25, the response at 200 Hz by successive x, at 100 Hz by successive v2t02 = v2 {2 _ x2 +, at 50 Hz by successive small triangles and at 20 Hz is characterized by successive small circles - From equation (I) the following is drawn. In each instant of time the group is directed so

energy from a horizontal geological boundary

t = V () 2 + t02 (L) layer, which is arranged at the correct depth,

v - i h so that reflected signals are generated at the sensor groups.

In contrast to the arrangements according to FIGS. 24 and 25 and the derivation, the diagram according to FIG. 30 shows a maximum response at x all frequencies at an inclination angle of 0 °.

= v yx2 _ | _ yit i (M) speaking the center line 2128 in Fig. 30. The response to if you insert the equation (K) into equation (M), the result is 15 of each frequency including the curve 2130 of the highest

Frequency 200 Hz is essentially symmetrical around the line dt x _ _x_ 2128. For negative sloping angles, the 200 Hz

(yyy / 2 <22 i 2 response curve the value of - 40 db at approximately 22 °, as shown by vvx + vt x vt, position 2132, and for positive layer inclination

Assume the following values: 20 angle, the intersection is 28 °, as by digit 2134

x _ j gQQ m shown. The loss of the 100 Hz signals at the layer

v = 2 100 m / sec angles of ± 15 ° is less than 3 db, and the t - 1 000 sec and loss of the 50 Hz signals is of the order of 1-2

a distance between the elementary sensor groups dx = 6 m, db. This is basically in contrast to de can be resolved to dt as follows: 25 o C ï St '^ tu T * nformation be, the 50

Hz, 100 Hz or 200 Hz frequency levels at the shift

dt = dx4 = 20 ^ = 2.45 millisecond (O) angles of + 15 ° can be obtained-

v2t (2 100) 2 (1,000) arrangements that use individually and continuously changeable, directional groups in the longitudinal direction of the seismic, this represents the desired delay between elementary "> cables, have the additional advantage that they have sensor units that are 6 m apart are, which is largely insensitive to the direction of the traversal required so that the seismic signals are at the time t tion. If, in particular, a seismic traversal at neighboring units in a group runs simultaneously, for example from east to west, where a system is coming. 24 or 25 is used with the response,

The angle © in FIG. 23 at the point in time t = 1,000 seconds would achieve a significantly different result than with the same system when crossing from west to

0 = arc cos * = arc cos = arc cos% = 31 ° (P) 0st If, on the other hand, an arrangement with the s vt 7 response properties according to FIG. 30 is used, no significant

As the time period t increases, reflections from deeper deviations between the seismic traverses occur

Layers, © increases and determine the required delay in two opposite directions between the elements of the group, which at the point 2120 Now the basic principle of this feature according to the

(Fig. 29) are arranged, explained for the maximum response illustrated embodiment of the invention and and the maximum signal-to-noise ratio decreases. some advantages of the new process and the new arrangement

have been explained, an embodiment is given below

30 is a diagram for comparison with FIGS. 24 and 45 of the new seismic arrangement and the new one

25, in which the response over the sloping angle method is discussed in detail.

is applied. In FIG. 30, the individual groups that are shown in FIG. 31 are a block diagram of such an embodiment.

27 shows a strong directional configuration of the type of the invention shown in FIG. 27, the cable 2152 having the cable 12 according to FIG. 1 and the cable 2056 according to FIG. 23 arranged in the longitudinal direction of the seismic cable

and directed so that they absorb energy that can correspond to a 50, a large number of cable sections are reflected in the horizontal plane at depth that the distorted 2156a, 2156b, 2156c .... 2156n. These sections of the cable, the length of time from the time of the launch to the time, correspond to one another by means of electronic connector modules for recording seismic reflections. This connected 2164a, 2164b, 2164c .... 2164n, which

corresponds, for example in connection with FIG. 23, to the cattle blocks 13 according to FIG. 1 and thus the transmitter

ten of the group located at location 2066 contained in such 55 receivers. The cable arrangement 2152 can be

Way that maximum energy from the sea floor 2058 ser is arranged in the course of a marine seismic exploration along the path 2070. To a small extent, as shown in FIG. 23 or in the event of a much later time, the group 2066 is spread further by land cable over the terrain to be explored, being directed downwards so that it draws energy from the interface 2060 to the seismic sensors which are then geophones that record along the 2072 path. This change in direction M, surface of the earth.

can, for example, by changing the delay between the remaining part of the arrangement according to FIG.

the various seismic sensor groups can be reached at the central station 2 in the ship 10 or a train

the one that forms the group at 2066 along the cable. stuff or the like arranged; for certain applications

FIG. 30 shows the addressing of a steerable group. In certain cases, certain processes that are shown in FIG.

which is centered at a location 2400 m from the launch fi5, is located at a central processing location distant from the site and which "looks" in one direction of seismic exploration. As in Fig.

Given an elapsed time for the arrival of reflected 31, the data word output from the cable into the seismic signals will equal 1000 seconds. It is based on data receiver and system control unit 2172, for example

617 017 26

17, and the capture device converted and processed by a second multi-seismic data plexed by the cable on unit 2172 of broadband transmission link 1014 can be processed to give as above explained. The timing of the transmission a) a seismic monitor section 2174 for testing by the multiplexed signals is controlled by the exploration team and b) a final cross-interrogation signal which cuts the transmission link 2176 producing high resolution for the geologist. Be given to the query network 116 1016, as in the mixing data from the control unit 2172 are described in a first one above.

Transferring group forming device 2178, which corresponds to the group- Fig. 33 shows part of the block diagram of Fig. 31 in the forming device 2130 in Fig. 17, and further to a detail and corresponds in some respects to Fig. 21. In Fig. 33

conventional monitor recorder 2080 for the seismic, (] is the group forming device 2178 in the large block, the

Transfer sections. The output from the group form is shown in dashed lines and has the reference number 2178,

Direction 2178 can also be recorded on a digital tape recorder, and corresponds to the group forming device 2182. That of the curve 1030 of FIG. 17. The system control unit 2172 of FIG. 31 is

Recorder 2180 recorded seismic on the monitor in Figure 33 represented by blocks 2212 and 2214. Data

Sections match those that have been in use so far u from the transmission link 1014 of the cable 2056 using various different seismic mixers in the data acquisition and processing device 2214

Cables have been achieved. Various can now be recorded that the data collection device 1028 of

Types of explorations, which up to now correspond by exchanging a Fig. 17, which the data words of the group form

Cable against another have been achieved by using 2178 under control of the main control device 2212

extension of the special cable according to the present invention and the 2o, which its counterpart in the main controller 1019 of Fig.

special control circuit 2172 and the group form 21 finds. The capture and processing device 2214

lines 2178 can be realized without the seismic cable converting the transmitted seismic digital data words with a length of two miles are spatially exchanged from the data channel 1014 into binary numbers and causes one of the must. Format control of this data in binary numbers with fixed

The output from the control unit 2172 is also a 25 point, which is given for processing in the digital group form second group form device 2184, which are also suitable for device 2178. Simultaneously with the recording is referred to as the beam steering device. If desired, the data information on the recording device 2214, the beam steering device 1284 can be operated directly from the control unit, a memory address and time control circuit 2216 (corresponding to 2172. Often, however, it is desirable to only operate the 1017 in FIG. 21), in order to decorate the original seismic information from the control unit 2172 on the position and number of the seismic sensor units at high speed and high speed and to record them to the recorded seismic information density and to assign them to the seismic information at a later point in time. The seismic data from the processing pre-beam steering device 2184 using the additional direction 2214 is fed to the allocation memory 2218 (corresponding tape device 2188. The recording devices 2186 corresponding to 1083 in Fig. 21) where they are reordered and and 2188 can e.g. RCA-Versabit 35 type video recorder according to instructions from the write control circuit. The output from the beam steering device 2184 can be written 2230 (corresponding to 1037 in FIG. 21) into the data channel sequence of the conventional digital tape recorder 2190. From the allocation memory 2218 and a normal seismic data processing device, the seismic information will be fed onto the data channel scan memory 2192 to transmit the final cross section 2176 rather 2222 (corresponding to memory 1047 in Fig. 21), high resolution using a conventional seis 40 The data channels are numbered from 1-500, using the seismic chart recorder 2194. A normal blending of sensor group 21a is started which is closest to the system shift correction for the group signals by the control unit and which is achieved with the most processing device 2192. On the other hand, remote seismic sensor group 21n can be ended (cf. a normal displacement correction also within FIG. 32). This numbering scheme for the sensor groups beam steering device 2184 (see below) can be achieved. 45 is also shown in FIG. 35. The assignment step is

As explained in detail above, this can be used to accommodate the different sequence of abandonment cable 2152 of FIG. 31, shown in FIG. 32, a series of data signals mixed into data link 1014 among data transmitter / receiver units within the Bring connectors as explained above. The detailed modules 2164a, 2164b, 2164c .... 2164n contain, so that internal control of the group shaping device 2178, FIG. 32 is essentially achieved in the upper part of FIG. 17 by the control unit 2224 (cf. control unit 1081 in FIG. at this point, however, it is repeated to see FIG. 21). The control unit 2224 is assigned the group control read understanding for the explanation of the beam steering to improved memory 2226, which memory 1067 in FIG. 21 also learns. The two broadband transmission connections correspond. Information relating to the desired group ends, the control unit 2172 (FIG. 31) with the transmission-reception combinations of the seismic signals from the five hundred devices; the first broadband transmission connection is 55 seismic sensor groups, the data connection 1014 is introduced into the control read memory, the second the control and interrogation 2226. This group information may, for example, link 1016, as shown in FIG. 32, which also shows that wise form a group, such as that shown in FIG. 27. This would be each transceiver having a polling network 116 and a group of ten elements with a changed rating a repeater network 114. coefficients as described above. The desired group-As mentioned above, the fifty cable sections 60 pen rating coefficients are in the group form-2156a, 2156b, 2156c .... 2156n of Fig. 31 uniformly in their direction 2178, especially in the read memory 2228 in a longitudinal direction Sensors, with three leading each (corresponding to memory 1055 of FIG. 21).

Sensors are connected to a sensor group. Three of the When Forming Combinations of Seismic Values

ten seismic sensor groups 21, each cabalab, which are required to form the evaluated groups,

are assigned are shown in Fig. 32. Seismic 65 will store the data stored in 2222 depending

Analog signals from each of the seismic sensor groups 21 from coefficients stored in the read memory 2228 are multiplexed and evaluated in digital form in multipliers 2230 and 2232, and those of the repeater converter circuit 114 in the transmit-emp elements of each group subsequently about it in

27th

617 017

Adders 2234 and 2236 added, these components being identified as 1047,1055,1053a, 1053b, 1064a and 1064b. A read controller 2238 (corresponding to 1049 in FIG. 21) is of course also provided for the time-controlled output of the seismic data from the memory 2222. The sums of seismic data that make up each group are temporarily stored in registers 2240 and 2242 (corresponding to 1065,1065b in Fig. 21). The dual channels, which include multiplier 2230, adder 2234, and register 2240, and multiplier 2232, adder 2236, and register 2242, are used to overlap groups that use seismic data from a single elementary sensor group in two different groups Include groups, with optionally different assessment of seismic information from a single channel as used in different groups. From the output registers 2240 and 2242, the group signals are stored in the output memory 2224 (corresponding to 1080 in FIG. 21), from which they are input to the format control device 2246 and the recording device 2248 (see the circuits 1032 and 1034 in FIG. 21). As indicated in FIG. 31, a real time writer 2180 may be connected to the output of the format controller 2246. On the other hand, the curve recorder 2180 can be stored on tapes whose data are recorded in the recording device 2248.

31, the beam steering group formation control device 2184 is similar in some respects to the group format control device 2178 of FIG. 33, but also has the additional essential property of selecting group signal elements from different arrival times at the different sensor units.

The group format controller 2184 (No. 2) of Fig. 31 is shown in block diagram form in Fig. 34, and its operation will be explained below in connection with the diagram of Fig. 35. In Fig. 34, the high speed, high density recorder 2128 is shown to the left and the main control circuit, shown as block 2252, controls all of the functions in the beam steering apparatus. Within beam steering device 2184, the high capacity of matrix input memory 2254 is one of the main differences from group format controller 2178. Instead of memory that stores a single value of seismic data from each of the five hundred channels, matrix memory 2254 stores data words that are 128 values of seismic Display information from each of the five hundred channels. The memory 2254 can be designed as a core memory, for example. 34 has the group read memory 2256, the channel coefficient memory 2258, the delay selection and processing device 2260 and the special data processing circuit 2262. In addition, the group shaping device includes input and output buffer circuits 2264 and 2266.

In operation, the group shaping device of FIG. 34 selectively combines a large number of seismic data input signals to form groups which continuously change in their direction of maximum signal reception. This change in direction is intended to accommodate the changes in angle der discussed in connection with FIG. 23 when it is expected that signals from successive deeper boundary layers will be reflected along the seismic section corresponding to line 2075 in FIG. 23. In order to direct the groups, the delays along the individual channels of seismic information that are combined with one another must be changed when successive complete cycles of group output formation are completed during successive periods of one millisecond. This is accomplished by the delay selector 2260 (Fig. 34) which provides address information in the large memory 2254 to deliver seismic data from each channel s from the memory 2254 which is properly delayed with respect to seismic data from neighboring channels.

The operation of the beam steering device 2184 is explained below from certain other points of view in order to express its operation more clearly.

FIG. 35 shows an embodiment of the invention in which a recording vehicle 10 is shown on the right side and the cable 2056 extends to the left. 15 The five hundred seismic sensor groups (each of which has three spaced detectors) arranged in the longitudinal direction of the cable are represented by numbered dots drawn along the cable 2056. The first eight sensor groups, which are arranged in the first cable section, are shown in FIG. 35 between the point 2302 and the point 2304. Groups # 251 through 270 extend from location 2306 to location 2308, and end sensor groups # 491 through 500 extend from location 2310 to location 2312. Each of the points shown in Fig. 35 25 in the matrix 2254 'represents a multi-digit binary number stored in the large memory 2254 of FIG. 34. In matrix 2254 of FIG. 35, data taken from certain channels is located below the associated numbered sensor group, and the data taken from a particular channel at 3 'consecutive time intervals of one millisecond is along a vertical one Line arranged, the time intervals at the edge of the right side are shown in Fig. 35. During a one millisecond interval, the total of five hundred channels corresponding to the five hundred seismic sensor groups along the cable are tested and the resulting seismic data is stored in memory 2254. As the time cycle progresses, old data is deleted from the memory 2254 and new, updated information is introduced. However, a "time window" or cycle duration interval of 128 milliseconds corresponding to 128 tests from each of the five hundred seismic sensor groups is stored in the memory 2254. This enables the combination in the group form device 2184 of seismic data from the sensor groups to be included in a group, the permissible time shift for the recording of the tests being up to 128 milliseconds.

So to give a specific example of how the system of Fig. 34 works in practice, the diagram of Fig. 35 identifies the location for three overlapping signal groups by lines 2314, 2316 and 2318 which are diagonally across the memory map 2254 'in Fig. 35 55. Each of these signal groups has signals from eight sensor groups (each of which has three sensors), as shown in FIG. 27. The example is based on a distance x (cf. FIG. 29) corresponding to the 251st sensor group, which is arranged at a distance of 1800 m from the launch site, usually-60, near an end of the cable. Furthermore, it is assumed that the time t = one second and that the seismic speed is 2100 m per second. The distance dx between the sensor groups is 6 m in the case of the present example. If you insert these values into 65 the formula (0), dt = 2.45 milliseconds. This is the desired delay between seismic signals that come from neighboring seismic sensor groups that are switched into the group at a distance of 6 m.

617 017

28

In the embodiment, as can be seen in connection with line 2314 in FIG. 35, the first sample from sensor group No. Taken 251 at the time equal to 1,000 seconds. The second group signal is derived from sensor group No. 252 taken at time t equal to 1.002 seconds. Similarly, the third test is taken from sensor group 253 at time t = 1.005 seconds. The remaining five groups of the eight groups tested in this way are those that appear above the dots along line 2314, i.e. up to the 258th group, which is checked at time t equal to 1.017 seconds. These selected delays correspond to the desired 2.45 milliseconds per channel and result in the in-phase summation of the reflected seismic signals.

35, lines 2316 and 2318 represent overlapping groups. In particular, the group indicated by line 2316 has seismic sensor groups No. 256 to 262, and the group represented by line 2318, sensor groups No. 261 to 268. The group represented by line 2316 begins 1830 meters from the launch site, and the arrangement represented by line 2318 begins 1860 meters from the launch site. Using formula (0), dt becomes 2.49 and 2.52 milliseconds for these two groups, the time being 11,000 seconds. The desired 2.49 milli: second delay difference per channel for the group corresponding to line 2316 gives the same scheme of relative delays for line 2316 as for the group of line 2314. The group represented by line 2318 has however, a much larger delay difference between channels, so that the fourth element of the group is selected from the memory slot according to a time period t equal to 1.008 seconds instead of 1.007 seconds as in the case of the groups corresponding to lines 2314 and 2316. Similarly, the sixth and eighth tests are taken at times 1.013 and 1.018 seconds instead of 1.012 and 1.017 seconds for the groups corresponding to lines 2314 and 2316. This increase in the desired delay between tests combined to give in-phase summation would, of course, be expected in the case of groups further away from the launch site and with a larger angle of the seismic waves impinging . In general, the stored seismic tests can be selected for each group in the longitudinal direction of the cable depending on the formula ■ (0).

In the case of FIG. 35, groups which comprise the first sensor groups which are closest to the launch site receive signals along paths which are oriented largely perpendicular to the cable and therefore do not require a great delay between channels to be combined. On the other hand, for the end group at the end of the cable, the incoming reflections run at a smaller angle than that in the middle of the cable, and therefore require a considerably greater delay between adjacent channels since the signal information is combined. In addition, the desired delays between the channels change with time and are reduced with increasing time because reflections from progressively deeper geological interfaces arrive at the groups with an almost vertical angle of incidence. From a mathematical point of view, the groups can be expressed by equations of the following form:

YJt = e »• YK (

(t + mz)

(Q)

YJi, ooo - Q, j 'Y2S11ioo + C2> J • Y25ij002 ... + C8 j • Y25f 0i7 (R)

where a YJt is a group test of the delivery of the Jth group at time t, YK (t + mz) is a test from the Kten input channel at time t, where t increases by the inclination z multiplied by the test number m, and Cu the ite coefficient ; is given to an input channel for the output of the Jth group. Equation (Q) is for short groups, e.g. limits the groups of eight elements indicated by lines 2314, 2316 and 2318 in Fig. 35, since equation (Q) assumes that the slope z is constant i, and this assumption applies only to short groups .

When evaluating the equations (Q), (R) by computer, it should be noted that the successive group signal outputs are a combination of signals from eight adjacent sensor groups, which are taken at different, separate test intervals, which are selected such that they approach the slope of the delay versus the distance of the incoming seismic waves. In the present case, this inclination is 2.54 milliseconds for the sensor group spacing of 6 m. Correspondingly, the selected samples are offset from one another by two or three milliseconds.

For the group with ten sensor groups according to FIG. 27, the ten coefficients are 1,2,3,4,5,5,4,3,2,1. Thus, for example, for the eighth of the ten coefficients in one; Group of samples to be included according to FIG. 27 C8J-3.

On the other hand, and for other explorations, an assessment can be applied to all samples. In addition, of course, a larger or smaller number of channels can be used in the formation of group signals. i Usually, however, between eight and thirty-two seismic sensor groups are provided in each group. The present invention opens up the additional possibility of processing data using groups that are in the expected direction of arrival of seismic signals; and the subsequent processing of the data to steer the groups in modified directions that are tailored to the specific geological conditions, so that a better "picture" of the terrain to be explored is obtained without additional work on the spot i is required.

In connection with marine seismic systems, the seismic impulse is normally initiated on the ship when the cable is towed behind the ship. The beam steering device accordingly shapes the groups so that their: Direction of maximum absorption is directed towards the ship and down and with increasing time points further downwards when seismic reflections return from deeper geological boundary layers. Similarly, in conjunction with land surveys where the 1 seismic impulse occurs either at the end of the seismic cable or near the center of the cable, the arrangements are initially targeted at the expected reflection points of the seismic impulse from horizontal geological boundary layers. It should be noted that in relation to the; normal correction of this, if necessary, can be performed within beam steering device 2184 rather than in processing device 2192. This is achieved in connection with FIGS. 34 and 35 by selecting seismic group samples from the memory 2254, which are time shifted as a grouping from the samples included in neighboring groups. In Fig. 35 this would quantitatively shift the line 2316 downwards, i.e. later, in field 2254 ', so that the corresponding group is shifted i; this would cause the group signals from the groups represented by lines 2314 and 2316 to represent both reflections from the same depth at any time instant. Of course, memory 2254 would have to be in its

29

617 017

Capacity will be increased so that it can accommodate the large delays required. In this way, normal correction can be achieved in addition to the group formation in the beam steering device 2184.

To continue the description of the seismic data processing device according to the present invention, a feature of the illustrated embodiment is explained below, which is carried out in the area of analog signal recording by the seismic sensor groups.

For a better understanding of the feature to be explained, reference is made to FIG. 36, which is a simplified schematic illustration of the arrangement according to FIG. 14. 14 shows the multiplexer 122 with a multiplicity of recording channels CUC2. ■ .Cn connected, all of which are connected to the multiplexer multiple line 312 via DC coupling capacitor 123, resistors 315 and switch 316. The signal recording channels are coupled to seismic sensors 21. Each resistor in conjunction with its capacitor 123 forms a high pass RC filter for its channel. The filter removes the DC components of the incoming analog signals. The switches 316 are high speed FET switches of known construction.

The output from the multiplexer 122 on the multiple line 312 is connected to a signal conditioning and amplifier network (SCAN) 3011, which schematically represents the switching arrangement shown in FIG. 16 and thus has, for example, the amplifier 320 'and the test and hold circuit 322'. This network conditions and amplifies the tested analog signals before they are sent to an evaluation device 3033. Network 3011 may include a fixed gain buffer amplifier, test and hold circuitry, and other circuitry for preconditioning a signal sample prior to transmission to evaluator 3033, which evaluator in practice includes amplification system 124 and analog-to-digital converter 126 of the foregoing Description is. Internally generated erroneous noises in multiplexer 122 appear on each capacitor 123 as interference voltage Vj, which may have one of the two polarities with respect to earth, and the internally generated interference noise voltage V2 on network 3011 occurs at the output and is generally added to voltage Vt. If the voltage amplitudes of the incoming desired signals become comparable with the interference voltages Vj + V2, the noise-signal ratio becomes too high. In seismics, a track that is too noisy often has to be erased, as a result of which valuable seismic information is lost.

The interference or stray voltages Vt + V2 have various causes, e.g. : thermoelectric influences, Peltier effects, offset drifts of the amplifier stages in the network 3011 etc. Each of the FET switches 316 used in the multiplexer 122 induces the noise voltage Vj on a capacitor 123. It is known that FET switches, which are semiconductor switches, Have feed-through capacitors and bleeder resistors between their control and switching elements (not shown). In addition, each control element in each FET switch is activated by a relatively large control pulse, which is provided by the network 3011, to a control line 3021. This control pulse reaches each DC coupling capacitor 123 via the feedthrough capacitor and the leakage resistance of the associated FET switch 316.

The capacitors 123 are identical, and the feedthrough capacitors and bleeder resistances of the FET switches are also identical to one another. Therefore, the stray voltages across the capacitors 123 are uniform.

37, according to an improved modification, one of the input channels to the multiplexer 122 is grounded.

This channel, referred to as the CO or "test" channel, is in all respects similar to the signal receiving multiplexer input channels Ci-Cn_, where n is the number of channels, except that the input port to the s channel CO is connected to earth so that no external signals can be given to the test channel CO. The test channel has a capacitor 123 ', which has the same capacitance value as the capacitors 123. A multiplexer switch 3023 is assigned to the capacitor 123 '. i H Between the network 3011 and evaluation device 3033, the test and hold (SH) network 1050 (FIG. 18) is connected, which comprises the series capacitor 1051 and the normally open shunt FET switch 1052, which is connected to ground, as in 18 explained. 15 The output from multiplexer 122 on multiple line 312 is connected to network 3011, which typically has amplifier 142 with gain ONE (FIG. 18), the output of which is connected to signal test and hold circuit 1044 (FIG. 18). 37, which is not shown in FIG. 37, is connected.

The state control network 132 controls all operational networks of the switching arrangement according to FIG. 37 via the control lines 3021. For example, the control network 132 sequentially switches the multiplexer channels C0-Cn_! over a multiple scan cycle and controls the signal evaluation device 3033, which has the amplification system 124, which is coupled to the analog-digital converter 126 in the manner described above. The converter 126 converts the multiplexed analog signals from the channels m Cj-Cn_! into corresponding digital numbers. Other networks that are not important in connection with the feature discussed here are omitted from FIG. 37 for the sake of clarity.

37, at the beginning of a multiplexer scan cycle, multiplexer 122 is reset to test channel CO, and at this point switch 3023 and another switch 3026 are closed. The stray voltage Vl which is generated at the capacitor 123 'is generally added to the stray voltage V2, which arises at the network 4011.

The voltage V! + V2 is transferred to a capacitor 3025 so that a test voltage V 'is generated there which, when switch 3026 is open, has the opposite polarity to voltage Vj + V2. Every-45 times when the multiplexer is reset to the test channel CO, the test and hold network 3024 receives the stray voltages from the test channel CO and from the network 3011. The switches 3023 and 3026 then remain open, while the multiplexer 122 sequentially the active channels Cj, C2 ... .Cn_ (thus receiving signals).

If the capacitors 123 and 123 'have the same capacitance values and if the overall gain of the network 3011 remains constant, then V' = Vx + V2. When the multiplexer 122 scans the channels Ci ~ Cn_i, each 55 channel at the output of the network 3011 has a signal voltage • VCo as well as the scattering noise voltage V! + V2, which is added successively and generally to the test voltage V ', which is held by the capacitor 3025 according to the following equation:

M>

Vc "+ V, + V2 + (-V ') = VCn so that each stray voltage Vj + V2 from each signal receiving channel is largely deleted by the test voltage V'.

FIG. 38 shows the principle of deleting stray voltages as given to the circuit according to FIG. 16. The single resistor 315 "is used instead of the individual resistor

617 017 30

stands 315 in Fig. 37 used. From the above, however, there are as many control lines as there are amplifiers, i.e. For this reason, the capacitors 123 and 123 'have a substantial number of stages.

is a smaller capacitance value than the corresponding condensers. In a preferred embodiment of the invention, Fig. 37. Accordingly, a given scattering control network 132 generates a synchronous, programmable current from a switch 316, for example a larger 5 counter, e.g. an integrated circuit 74S161 made voltage across capacitors 123 or 123 'of Fig. 38 as from Texas Instruments. Reference decoder 3144 is one on the corresponding larger capacitors of Fig. 37. Digital to analog converter, e.g. MC 1408L, manufactured by In Fig. 38 have other circuit elements that are not Motorola. The comparator 3138 can be an LM 311

37 but identical to the circuit voltage comparator, and each amplifier corresponds to that of FIG. 18, a multiplexer buffer, 0 Aj-A ,,, is an operational amplifier with high input impedance, stronger 142, the signal test - and holding capacitor 1046 and e.g. LF156, both manufactured by National Semiconductors, the 1048 buffer amplifier. Aside from these minor- Inc.

37, each of the amplifiers A] -Am has a control circuit for FIG. 37 fully applicable to FIG. 38, and the compensation, gain, and noise cancellation circuitry is generated on capacitor 1051 (also 40 only (both shown in detail in Fig. 40 (that of the more detailed circuit shown in Fig. 18), around the control circuit has a voltage divider network that eliminates stray voltages across capacitors 123 through resistors 3170 , 3172, FET switch 3174.3176 and compensate. the special gain control line 3150 is formed. At the

To continue the description of the seismic data process of adjusting the gain when the processor of the present invention closes 20 switch 3176 and switch 3174 is open, below a more detailed explanation of the gain is provided by the output signal of the amplifier from output 3135 to systems with variable gain in connection the inverting input 3134 fed back, whereby given with FIGS. 39 to 42. Amplifier in its state of low gain

In Fig. 39, the left side shows the analog-digital acquisition that is equal to ONE. When the switch 3174 system including the multiplexer 122 is closed with signal 25 and the switch 3176 is open, the switch becomes channel channels Ct, C2 ... .Cn, the output of the multiplexer being in its high gain state (eg G! For the on of the multiple line 312 with the non-inverting single amplifier Aj), since the resistor 3170 is then connected to a buffer amplifier 320 with the amplification factor feedback loop of the amplifier. The desired gate ONE is connected, which is an operational amplifier, e.g. an LF gain condition may depend on a control signal from National Seniconductor Inc., the signal of which is connected from the control network 132 to the test and hold circuit 322 via the control line 3150 output; set to the specific amplifier stage, e.g. A, these circuit components are referenced above, particularly in FIG. 40.

14 has been explained in detail. The noise canceling circuit has FET switch 3182,

The variable gain amplification system 124, 3184.3186, a capacitor 3188 and a noise cancellation factor has an input terminal 3130 and a control line 3160. If the switch 3182 is closed, the output port 3132 is open, and basically consists of the switches 3184 and 3186 opened. In the noise or at least two, preferably four, amplifiers with double displacement erase operation, with the aid of which the characteristic gain factor, i.e. Amplifier stages Aj, A2 ... on, which are corrected in the properties of each amplifier, are cascade connected. Any amplifier, i.e. each stage, the control network 132 sends a signal over the control line a pair of input ports 3133,3134 and an output port 40 3150 to close switch 3176 and switch port port 3135 (FIG. 40). The non-inverting 3174 open, causing each amplifier to be on state

Input 3133 to the first amplifier stage A1 is brought with the one gain factor ONE. At the same time, the connection 3130 separates, which receives the output signal from the control network 132 and the input connection 3130 to the amplifier test and hold circuit 322. The output voltage system 124 by opening the switch 3182 and setting the last amplifier stage to the output terminal 45 by closing the switch 3184 (FIGS. 40 and 41) the Ver-3132 is given to a signal evaluation device which is responsible for the connection between the connection point 3190 and the From analog-to-digital converter 126 in the output connection 3135 of the amplifier described here. The input system is like the comparison device 3138. The closure 3133 to the amplifier is grounded by closing the sound amplification factor of each amplifier stage to one of ters 3186. Switches 3182, 3184 and 3186 are set to two states, i.e. can be set, namely a 50 actuated by a signal that arrives on the control line 3160 state of low amplification factor, preferably immediately. Any DC noise ("displacement") appears as ONE, or a high gain state Gj, then at the output terminal 3135 of the amplifier. A Kon-whereby G; a discrete high gain value for the capacitor 3188 connected in the manner shown is i-th (i = 1.2 ... .m) amplifier. is built with the polarities on its plates 3188a, 3188b

55 load, which in Fig. 42 represents an offset voltage + V-The control network 132, which is in connection with above, which occurs at the input terminal 3133. If Fig. 14 was mentioned, between the switches 316 and 326 the offset cancellation operation is completed, this closes via the control multiple line 341, which has a multi-control network 132, the switch 3182 and opens the switch number of control lines. The control network 132 3184 and 3186 using a second signal via the control sequentially switches the signal input channels Ci-Cn via an ft0 line 3160, so that the input connection 3130 again with a multiplexer sampling cycle as described above. amplifier system 124 and the amplifiers

A variable reference decoding is returned to the control network 132 in normal operation.

rer 3144 and the window comparison device 3138 assigned. The charge voltage + V at the capacitor 3188 is now added to the control signals 3150 and the noise-canceling control all-in-all with the incoming signal Vin and lines 3160 connect the state control network 132 to (, 5 completely by the offset impressed on the amplifier in each case one of the amplifier stages Ax -Am. The simplicity signal + V cleared as shown in Fig. 42. The output signal is only one gain control line 3150 and one of the amplifier is thus free of DC noise, unde noise canceling control line 3160 shown in Fig. 39, it are dependent on the gain factor to which the amplifier is based

31

617 017

has been discontinued. The factors that determine how often it is necessary to go into offset cancellation operation depends on the speed at which the offset noise drifts and the value of the amplifier's input impedance.

When operating with a variable gain factor, the gain factor determination cycle for each channel (e.g. channel Ct) begins by bringing all amplifiers to gain factor 1. State control network 132 provides a sequence of digital codes that represent a corresponding sequence of variable reference voltage steps. The digital codes are converted to corresponding discrete reference voltages VR by the reference decoder 3144. The reference voltages can result from this

Vr, = Vb / Gj, VR2 = Vb / G2 .., VRm = VB / Gm, where G1; G2 ... Gm are the higher values for the amplification factor of the amplifiers A, A2 ... .Am and VB is a base voltage equal to a preselected fraction of the full voltage of the analog-to-digital converter 126. The gain factors G1; G2 etc. are proportional to a selected power of the same number base, e.g. two. The preselected energy is unique for each amplifier. The variable reference voltage steps VRi are inversely proportional to the unique gain values of their corresponding amplifiers. The control network 132, as part of the offset removal operation and the changing gain operation, has the amplifier set to the gain 1 state. This common step simplifies the implementation of these functions.

In a preferred embodiment, Gx = 2 exp 2 (mi) for a four amplifier amplifier system 124 (m = 4), the corresponding high gain values for the amplifiers are: G] = 256, G2 = 16, G3 = 4, G4 2. With only four high gain settings, the gain of binary gain system 124 can be designed to go from a minimum of two to zero power (ie one) to a maximum of 2 to fifteen power (ie 32,786) in steps of powers ranges from 2. To achieve this, only four gain decisions are necessary. The gain comparisons must be made so that the amplifier that has the highest gain setting is made first, and then the comparisons are made in order of decreasing amplifier gain regardless of the electrical position of each amplifier in the cascade.

The operation with changing gain factor is explained in detail below. When all four amplifiers are set to unity gain, the absolute value of the system output voltage IV0I that appears at the output terminal 3132 is compared to the first reference voltage VR1 = Vb / 256 by the comparator 3138. The comparator 3138 rectifies the voltage V0 and compares its absolute value with the reference voltage VR1. The decision based on this comparison is given to control unit 132. When IV0I ^ VR1, controller 132 causes the gain of amplifier At to remain at one. If IV0I ^ VR1, controller 132 sets the gain of amplifier Ax to Gb i.e. 256. The variable reference voltages VR2, VR3, VR4 are then sequentially compared to the successive values of the output voltage VQ that occurs after the sequential switching decisions have been made. When each comparison has been made, a gain decision is made by the control network 132 for the amplifier

50

55

leads, which has the next lower gain factor corresponding to the reference voltage used for the comparison.

Comparator 3138 thus compares output voltage IV0I at terminal 3132 with reference voltage VR1 corresponding to amplifier Av. If the comparison is negative, that is, if the absolute value of V0 is less than VR1, comparator 3138 issues a command to control network 132 so that normally closed switch 3176 is opened and normally open switch 3174 is closed (Fig. 40). The amplifier Ax now has the setting with a high amplification factor. The signal under test, which is held by the test and hold circuit 322, is again amplified by the binary system 124 with a changing gain factor, with the amplifier Aj set to its high gain setting and the amplifiers A2-A4 to their gain settings Stay one. If the new value IV0I is less than VR2, i.e. if the result of the second comparison is negative again, the comparator 3138 issues a command to the control network 132 to bring the amplifier A2 into its high gain setting. The signal under test is again amplified by the changing gain binary system 124 having amplifiers Aj and A2 at their high gain settings and amplifiers A3, A4 at their gain one setting. If the next value is IV0I <VR3, i.e. if the result of the third comparison is negative again, controller 132 causes amplifier A3 (not shown) to assume its high gain setting. If, after the fourth comparison, IV0I <VR4, controller 132 causes amplifier A4 (which may be amplifier Am) to assume its high gain setting, but if IV0I ^ VR4, controller 132 holds amplifier on Gain level one. After the fourth comparison is complete, the comparator 3138 issues a command to the control network 132 to cause the analog-to-digital converter 126 to take the voltage V0 that then appears at the output terminal 3132 and to convert it into a digital number.

Upon completion of each gain selection sequence, the total gain of the cascade of four amplifiers is encoded by the control network 132 as a digital gain codeword containing as many bits as there are stages, i.e., amplifiers, in the gain system 124. Each bit of the gain code word represents the state of switch 3176. The gain code is ONE when switch 3176 is open, it is ZERO when the switch is closed. Thus, in the case of four amplifiers, when all amplifiers are set to gain one, the gain code is 0000. When the total gain is 64, the gain code becomes 0110.

Thus, with four amplifiers that require only four decisions, the binary gain system 124 with changing gain, which is discussed here, gives 215 gain levels. To achieve this result, known systems have required fifteen amplifiers and sixteen separate decisions and procedures. Thus, the binary gain system 124 with the control circuit described can select an appropriate gain in a fraction of the time previously required in known systems and takes up only a fraction of the volume previously required to change a conventional binary gain system Gain gain factor.

In the course of the explanation of the diagram of FIG. 13, the principle of using a signal pickup is between

617 017

32

a relatively slowly progressing query signal and a relatively rapidly progressing command signal, namely the DATA signal. This feature is further discussed below in connection with FIGS. 43-48 because it is used generally for the purpose of initiating a shift.

43 is a schematic, simplified overall illustration of the seismic data processing device explained so far, which shows the common, central data processing device, ie station 2, and the multiplicity of identical, spaced-apart multichannel data acquisition devices, ie transceiver devices purple, 111b , 111c, 11 id, which are connected in series and connected to the central station via transmission lines which have been explained above and which is shown in simplified form in FIG. 43 as a signal transmission connection 4016 with three channels. The separation between the acquisition, i.e. Transceivers are preferably constant, and are approximately 60-90 m.

The central station 2 has a control circuit 4018 and a data processing and recording unit 4020, which are schematic representations of the circuit explained. The simplified representation of circuits 4018 and 4020 is to be understood to represent the system control unit and the data receiver (block 2172 in FIG. 31), a more detailed block diagram of the central station 2 being explained and shown above in connection with FIG. 21. Unit 4020 can be a known type of magnetic tape recorder. The control circuit 4018 has a signal transmission device, e.g. a clock circuit of known type, for the transmission of the multistate interrogation signals IP, e.g. Pulses S1 and S2, on preselected test signals and / or a control signal through the channels 4090 and 4091 of the three-channel signal transmission link 4016, as explained above.

After each transceiver lilac, 111b has finished transmitting its local data, it picks up data from more distant data acquisition or transceiver devices, regenerates it and transmits it to the central station 2. The data acquisition device lilac,. that closest to station 2 thus transmits its local data first and then picks up and transmits data from the remaining 99 downlink related devices (assuming there are 100 such devices in the system). The last data acquisition device, of course, only transmits its local data.

As also explained above, an interrogation signal can have one of a number of states or properties. The preferred interrogation and control signals are square wave pulses, although other types of signals can be used. The rate of propagation of a pulse through polling channel 4090 is different from the rate of propagation of a pulse through control channel 4091, polling channel 4090 schematically representing transmission lines IP1, IP2, IP3, while control channel 4091 represents the DATA and data shunt lines of FIG. 5 and Figure 8b. In the embodiment discussed here, the propagation speed through control channel 4091 is greater than through query channel 4090.

If a data acquisition unit, e.g. a transmit / receive device 111b becomes defective, it must be shunted so that data coming from a more distant device, e.g. of the device 111c are transmitted, are not influenced thereby. A control pulse is transmitted from the control unit 4018 via the control channel 4091. With a selected device, e.g. the send-receive

device 111b the receiver pulse overtakes and coincides with the interrogation pulse due to the different propagation speeds in channels 4090 and 4091. The coincidence of the two pulses on the device 111b; causes the device to be shunted.

43 and 44, the data acquisition, i.e. Transceiver devices purple, 111b, 111c, llld are provided with a multiplicity of input channels, each input channel being connected to a seismic sensor group 21 in the manner described above. Each of these devices includes the signal conditioning logic that includes the multiplexer 122, the test and hold circuits 3024, the gain conditioning amplifier 124, the analog-to-digital converter 126, and the output signal register 128. These circuit components connect the seismic sensors 21 to the data channel 4092 'at the signal input channels. They can be of a conventional type, as is known per se in seismics. The gain conditioning amplifier 124 is conveniently the variable gain, floating point binary amplifier as described in connection with Figs. 39-42, which has a four bit gain code to indicate the gain setting for each data test . Analog-to-digital converter 126 may be, for example, a 12-bit; Micronetics MN 5212 converter, although a converter with higher or lower resolution can also be used. Output signal storage register 128 may be a conventional 16-20 bit serial input to serial output shift register. In a preferred embodiment, register 128 has a capacity for at least 12 data bits from the analog-to-digital converter and four gain code bits for the gain conditioning amplifier.

As described above and shown in Figure 44, this is; Control network 132 is provided, which is activated by signals S1 or S2 on lines S1 or S2. The signals S1 (sampling interval interrogation pulse) or S2 (divider interrogation pulse) are generated as a function of interrogation pulses which assume either a first state or a second state. The corresponding interrogation pulses, which are generally designated by the letters «IP» and which are transmitted via channel 4090, e.g. the connections IP1, IP2, IP3 of FIG. 5 are also designated S1 and S2, where S1 is one state (namely a certain width) and S2 is another; State (namely a different width). Depending on a signal S1, the control network 132 resets the multiplexer 122 to the channel CO, the test or empty channel. Depending on a signal S2 following a signal S1, the control network 132 shifts the multiple ixer 122 to the first input channel in the sequence, so that the test and hold circuit 3024 checks an input from the first channel. The S1 pulse sets the control network 132 in such a way that the data output as a function of S2 pulses is made possible over the duration of the sampling cycle, the sampling; cycle is the multiplexer mode for checking all fourteen input channels.

If the signal sample is amplified by the amplification factor conditioner 124 and presented to the analog-digital converter 126, the amplification factor is correspondingly expressed as an amplification factor code with four bits. When the next S2 signal is received, the control network 132 sets the multiplexer 122 to the next channel and, at the same time, causes the analog-to-digital converter 126 to amplify the gain-conditioned sample from the first channel into a digital number converts. At the beginning of the conversion cycle, the four-bit gain code is transferred in series from the gain conditioner 124 to the output register 128 over line 3036.

33 617 017

As analog-to-digital conversion proceeds, such logic circuits will normally be present

the twelve bits that represent the digital number in series into these two states can be represented as binary number signals

Output registers 128 from analog-to-digital converter 126 are considered to be lend, and they are often given as logic one. In register 126 the 12 data bits are designated with the four and logical ZERO. In addition, the low gain code bits to form a digital word with 5 gen and high voltage states are sometimes called "binary zero"

16 bits corresponding to the sample from the first channel combined and «binary one» or as «false» and «real» signals or kidneys. Four preamble bits can be added to indicate states. If in the case of an AND gate e.g.

a word of 20 bits is created. the two inputs to a predetermined voltage level

If the conversion cycle for a channel, e.g. Channel raised (which is referred to as "real") changes

K begins, the control network 132 transmits the digital 10 the output also to this voltage level (as "real"

Data word from the channel Kl, which previously referred to in the output register), while when one of the inputs on a

128 was stored on the data channel 4092 '. A counter-different, lower voltage level lies (with

Decoder 3037 counts the bits that are denoted in series from the register "wrong"), the output of the AND gate on the

128 have been blanked, and gives the control network 132 a low level (in the "wrong" state). Similarly, the command to end the data bit transmission when 15 ways are used in the discussion below, the two the counting is complete. The data channel 4092 'of a logic circuit state as "real" and "false"

each transceiver 111 is usually referred to as states.

connected to the data transmission channel 4092, as can be seen from the fact that if the control network 132 (FIG. 44) receives a query

45 described below. The connection pulse transmits in the first state, the pulse proceeds as in

4092 of FIGS. 43 and 45 corresponds to the connection D1, D2, D3 20 FIG. 45 shown by the query channel 4090, by the switching

(see above). ter 4044 to the line receiver 4048, through the switch

Fig. 45 shows further details of one of the data acquisition 4050 to the line driver 4052, the switch 4046 and to the solution units, e.g. Transceivers purple, 111b next transceiver in sequence. The etc. including a signal property identifier pulse also passes through delay line 4072. At device 4038 and first and second signal coincidence detectors 25 at the end of 1000 nanoseconds the leading edge of this occurs

4040 and 4042, which are shown in dashed lines and represent the pulse from the output of the delay line. In series with the query channel 4090, in addition to this point, however, the trailing edge of the pulse at the closing switch 4044.4046 with loss of power, a line input of the delay line is "visible", i.e. available,

receiver 4048, a switch 4050 to disable the Thus both inputs go into the AND gate 4074 "real",

Interrogation signals and a line driver 4052 connected. The 30 i.e. to active, high levels to activate the AND gate

Control channel 4091 is provided with a line receiver 4054 and 4074, whereby a 200 ns signal on line S1 generates a line driver 4056. Data channel 4092 will have a positive leading edge. As instructed by a line receiver 4058 and an OR gate / described above, controller 132 (FIG. 44)

Power driver 4062 on. The two inputs in the line- if it indicates a signal on line S1, the multiplexing device driver 4062 are the inputs 4092 from downstream 35 122. The trailing edge of the second interrogation pulse lies data acquisition units, i.e. Transceiver generates a positive logic level on the direction and the input 4092 'from the local data output line S2, which is the output of the inverter 4076, 200 ns register 128 (see FIG. 44). The Schlater 4064 and 4066 after S1 became "real".

cause data to be delivered over the bypass line 4068. It is now assumed that a query pulse will later be shunted when deactivated. The 40 second> narrow state through the channel 4090 to delay

Direction of data flow in FIGS. 44 and 45 is reversed about line 4072 and progressing to inventory 4076. Since the pulse width like that in Figs. 43 and 46 is too narrow to be at the input at the same time

The signal property identifier 4038 seen from uncj at the output of delay line 4072

a tapped delay line 4072, the AND gate which can> kdn signal is generated on line S1. The

4074 and the inverter 4076, identifies the state, 45 running edge of the narrow pulse occurs at the exit

i.e. the property of an interrogation signal in the following course of the inverter 4076 as a positive going signal in the manner described. The interrogation signal is essentially on line S2. If controller 132 is positive a square wave with a certain width. The state of going S2 signal, as mentioned above,

or the property of a pulse is here defined by its width, it switches the multiplexer 122 to the next input channel, although with a corresponding circuit arrangement, it initiates a conversion cycle and it gives another property, e.g. the pulse height as a data signal via line 4092 'to channel 4092 and can be used as a divorce feature. A wide pulse thus into the 4020 recorder.

is an interrogation pulse in the first state. The width of a wide pulse must be greater than the delay time of the. As described above, any transceiver

Delay line 4072, however less than half of the 55 device have fourteen analog input channels, at pre-selected sample or test intervals. To be able to check a query pulse for each input channel one after the other, the second state must be clearly different from a query pulse.

be distinguishable in the first state; preferably its device 4018 is transmitted. If the wide interrogation pulse Sl is longitudinal

Width less than half a wide pulse. In the embodiment described of polling channel 4090 for each transceiver, the delay 60 purple, 111b, etc. progresses in order that it sets the duration of delay line 4072 1000 nanoseconds (ns), multiplexer 122, which is included in each unit. In the arrival

a wide pulse is 1200 ns long, and a narrow pulse is followed by a series of thirteen interrogation pulses

400 ns long. Additional pulse widths can be used S2 transmitted in the second state. Each pulse in the second state, in order to achieve a pulse with multiple properties, switches the multiplexer 122 on so that, if appropriate, changes are made to the signal property identifier in turn after each of the input channels from the seismic ornamental logic. Sensor groups S1 checks and the corresponding data

In the following description of logic circuit signals from the transceiver 111 to the programs, reference is made to the two states which the drawing unit 4020 transmits through the data channel 4092.

617 017

34

In the described embodiment, the fourteen channels are checked within one millisecond. Accordingly, the interval between S2 pulses is 71.4 microseconds (10-6 sec). The two-way pulse propagation delay through the transmission link 4016 between two transceivers in each case results in a time gap during which the data signals can be transmitted by the transceiver 111 without interfering with one another.

The shunt switches 4044, 4046 and 4064, 4066 are actuated in a known manner by relays and are shown in FIG. 45 in their position with the power switched on. In the event of a power failure in a particular transceiver, both sets of switches switch such that shunt lines 4068 and 4070 are connected in channels 4092 and 4090, respectively. Then the interrogation pulses and data words to remote and remote transceivers pass freely through the defective device via shunt lines 4070 and 4068.

A transceiver, e.g. device 111b may become defective requiring it to be shunted, or may stop further transmission of an interrogation pulse on a particular device if desired. These special functions are activated by a control signal in the manner described below.

The total transit time of the polling pulse to a remote transceiver depends on the delay time through the polling channel to the device. The runtime to device n is the sum of the delay times between all preceding data acquisition devices. Similarly, the delay time of a control pulse through the control channel to device n is the sum of the delays in the control channel between all of the previous devices that are closer to control unit 4018 than device n Transceiver a pulse that progresses through the faster channel by a time interval (n-1) R earlier than the pulse through the slower channel, where n-1 is the number of intervals between the first n transceivers and R is the signal transit time difference through the two channels between successive devices. Preferably, the feed cable section 17 between the central station and the first transceiver is designed so that the delay times for both signals, namely command signal and interrogation signal, through channels 4090 and 4091 are the same. Thus, all different delays are generated in a corresponding manner in the lines between successive transceivers.

It is assumed that the pulse propagation speed in the control channel 4091 is greater than in the query channel 4090. If a query pulse from the control circuit 4018 (FIG. 43) and (n-1) R later a control pulse is transmitted via the control circuit 4018, the control pulse overtakes the Interrogation pulse at the device and fields it. Both the polling and control transmission links can be characterized by identical propagation speeds. Delay lines can be inserted into one of the two channels on each transceiver to create an effective propagation rate difference. For example, a delay line 4078, shown by the dashed line box in Fig. 45, can be inserted into the polling channel 4090 between the line receiver 4048 and the shutdown switch 4050. In addition can

30th

the delay line 4078 serve as a replacement for the delay line 4072.

The first signal or pulse coincidence detector 4040 has a D-type flip-flop 4080 and a relay 4082, which is assigned to switches 4064 and 4066. The switches are shown in the position where the relay is energized. The D-type flip-flop 4080 can be one half of a 74S74 positive-edge triggered double flip-flop, as manufactured by Texas Instruments Co. A D-type flip-1 flop is a bistable memory circuit with a single input D and outputs Q and Q. The logic level present at the D input is transferred to the Q output when the correct edge (i.e. the transition from one logic level to another) occurs at the CK (clock) input. The flip-flop remains in its state until it is reset. The flip-flop 4080 responds to the rising edge_ (transition from negative to positive) of a pulse. The Q output always assumes a logic level that is opposite to that of the logic level of the Q-1 output. The flip-flop can be reset by applying a pulse to the CL (clear) input. After resetting, the logic level of the Q output is a logic ZERO and the Q output is a logic ONE.

Depending on the simultaneous presence of both a control pulse and an interrogation pulse in any state, the coincidence detector 4040 is effective for the first signal. The leading edge of an interrogation pulse sets (i.e., activates) the D input of the D-type flip-flop 4080 to a logical ONE. The Q output of flip-flop 4080 is normally "false" (logic ZERO), which energizes relay 4082 to close switches 4064 and 4066, as shown in FIG. 45. If a control pulse from connection 4091 arrives at the CK input while the D input is a logical ONE, a flip-flop 4080 is switched to turn the Q output to a logical ONE, i.e. relies on "real". When "Q" becomes "real", relay 4082 drops out, causing switches 4064 and 4066 to make contact with shunt line 4068 because the voltage level of the logical ONE is the same as 4 (1 kw + V).

In Fig. 46, control network 132 (Fig. 44) transmits an interrogation pulse over interrogation channel 4090, which is believed to have the slower propagation rate. The control channel 4091 is connected to the control network 132 via 45 a tapped delay line 4132, which has taps to be integral multiples of the delay time, e.g. 0, R, 2R, 3R, (n-1) R via the tap selector switch 4100.

so In order to bypass the transceivers, an interrogation pulse is first transmitted through the control circuit 4018 (Figs. 43 and 46) and then (n-1) R later a control pulse. The control pulse picks up and coincides with the interrogation pulse at the transceiver n, the 55 relay 4082 (FIG. 45) being de-energized; this switches switches 4064 and 4066 to bypass line 4068. In simple terms, the control pulse is delayed with respect to the interrogation pulse by an integer multiple of the delay time R, where the integer multiple 60 is equal to the number of transceivers connected between the transceiver n and the central station 2.

It may be desirable to prevent a further migration of an interrogation pulse to transceivers beyond ftS of unit n. To perform this function, the control pulse is time shifted so that it follows an interrogation pulse after a delay of (n-1) R-1-d, where d is the time shift. This feature will

35 617 017

through the second pulse coincidence detector 4042 through part of the subset of only the transceivers. Form 11 le-11 le, whereby the delivery of data from the input pulse in the second pulse coincidence detector 4042 (FIG. 45) is made effective only to this subset of this subset of the D input of the flip-flop 4084 with a tap 4085. Thus, the front of the tapped delay line 4072 remains connected. The 5 directions purple, 111b, Ulf and 111g inactive. The circuit delay time between the entry of the leading edge for activation, i.e. for effecting the desired pulse on tap 4085 is the same or some subset of devices is shown in FIG. 48.

larger than the width of the pulse, whereby the delay in FIG. 48, the data shunt circuit in duration d for tapping 4085 in the described embodiment is essentially the same as that in FIG. 45. However, the approximation form is 600 ns. When the interrogation pulse arrives, m signal property identifier 4038 in FIG. 48 by which it first activates the flip-flop 4080 in the pulse coincidence de- selectively delay line 4078 of FIG. 45. The tector 4040. At a time d later (e.g. 600 ns later) the pulse coincidence detectors 4040, 4042 according to FIG. 45 are designed differently in the leading edge of the pulse at the tap 4085 of FIG. 48, so that a higher flexibility

Delay line 4072 on. The time-shifted tax activity is possible.

pulse is too late to trigger the CK input of flip-flop 4080 15. Query channel 4090 is the only physical channel to trigger; thus, the pulse coincidence detector 4040 does not speak in FIG. 47. Control channel 4091 is also on. Now, however, the D input of flip-flop 4084 is represented by a single line in FIG. 47. However, it passes the delayed polling pulse activated. That is why a switch actually consists of three redundant lines. As described above, the time-shifted control pulse, which is explained in more detail at the CK input of the flip in the illustrated embodiment, flop 4084 arrives, flip-flop 4084 and causes the 20 to be twisted wire pairs of the interrogation and control channels. The normally "real" Q output goes after "false", a pulse propagation speed over the conductors that activates the polling relay 4086, which opens the breaker 4050, causing control channels to represent the same value, then a further migration of the polling pulse to have devices. However, the delay line 4078 is terminated in each transmission that is in line with the polling channel beyond the transceiver n receiving device. 25 tet. Thus, the effective one in the illustrated device

In FIG. 46, the time shift d is given to the control pulse speed in the query channel 4090 smaller than in the control channel via a fixed delay line 4102 if the nal 4091. Thus, the delay line 4078 delays the

Switch 4104 occupies the position shown. The time delay propagation of an interrogation pulse by a fixed time interval r d over the delay line 4102 is the same as that at each transceiver. The delay line

Time delay at the tap 4085 of the delay line 30 4078 has a maximum delay of 1000 ns with tap

tung 4072. i.e. 600 ns in the illustrated embodiment. to achieve shorter delays and smaller ones

If either flip-flop 4080 or 4084 can be set due to the simultaneous difference in the lengths of the presence of an interrogation pulse and a control im conductor forming the interrogation channel. The preferred delay puises is switched, it remains in the switched state up to 600 ns.

for release. The flip-flops 4080 and 4084 are only in 15 In Figure 48, an interrogation pulse that comes from the control

Dependence on an interrogation pulse in the first state, i.e. device 132 is transmitted along line 4090 via the

Sl released, but in the absence of a coincident line receiver or buffer amplifier 4048 above the energy

Control pulse when the output of the AND gate 4074 in giausfall-shunt switch 4044 in the delay line

a logical ONE goes. tung 4078 continued. Six hundred nanoseconds later it arrives

In a particular embodiment of the invention, 4 »impulse through the tapping 4101 to the line driver 4052 in addition to the features described above is also desirable, and on the transceiver which will activate the line on certain switching sequences, i.e. to activate next lies, i.e. the next downward with respect to the link and data from a subset of consecutive manure transceivers. Submit transceivers that are selected from the total- AND gate 4103 showing the presence of a white all devices. These functions are 45 th Sl pulse, as described in connection with the AND gate also by using two transmission links 4074 in FIG. 45. Since the Sl pulse is 1200 ns wide with different delays. The selected and the maximum delay of the delay line 4078 subset can, for example, also be only one transceiver 1000 ns, the output line 4105 of the device changes, or else all transceiver devices AND gate 4103 change to a logical ONE, as explained above, include. If the selected subset comprises more than 30 trig-direction, the CK input of the D-type flip-flop 4106 is a first selected device and device. If there is a logical ONE at the D input of the flip, a last selected device, with the first flop 4106 present (from a control pulse as closest to the next selected device of central station 2 will be explained), the Q output becomes a logical one ONE

In Fig. 47, which is similar in some respects to Fig. 13, it remains a logical ONE causing the output of the AND-

Shown are seven transceivers llla-llg, 55 gates 4108 merging into a logical ONE. The running edge of the SI pulse with the central station 2 via the transmission connection therefore also generates an S2

4016, which consists of three channels, namely the pulse that causes the control device 132 (FIG. 44) to be one

Polling channel 4090, control channel 4091 and the data conversion cycle. As long as the D input of flip channel 4092 exists. 47 is in the direction of flop 4106 on the logical ONE, the AND remains

conversely compared to the group of FIG. 43, and the 60 gate 4108 turned on. For the rest of a sampling cycle

For the sake of simplicity, the input channels 21 are not shown. subsequently incoming S2 pulses occur at the S2 output

47 also show the gear, i.e. on line 4110. Conversely, if the D-

Double arrows on control channel 4091 that the signal propagation input of flip-flop 4106 is a logic ZERO, the speed in this channel is greater than that blocked by the output of S2 pulses via AND gate 4108

Query channel 4090. 65 and the circuit of Fig. 44 is not activated.

In the exemplary embodiment shown in FIG. 47, it is control pulses which correspond to the DATA pulse described above.

Desired to correspond to a sampling cycle or another switching sequence in (DATA = data enable) are to be activated and initiated in parallel via the multiplexers 122, which transmit the triple redundant control channel 4091. The

617 017

36

DATA pulses are received via line receivers 4112, 4112 ', 4112 "and transmitted to another majority selection circuit 4114. Circuit 4114 consists of AND gates 4116, 4116', 4116" and OR gate 4118. A DATA pulse, which is present on any two of the three lines CON 1, CON 2 and CON 3, represented schematically by the DATA connection in Fig. 5, causes the output of OR gate 4118 to become a logical ONE, with the D Input of flip-flop 4106 is also set to logic ONE so that AND gate 4108 is enabled. Thus, the simultaneous presence of an S1 pulse from AND gate 4103 and a DATA pulse from any two of the three lines CON 1, CON 2, and CON 3 produces a single signal that outputs the data from control network 132 (FIG. 44) in FIG Dependence on subsequent S2 pulses, which are recorded during the rest of the sampling cycle, is made effective. The system remains in effect as long as a DATA pulse is present each time an SI pulse is received. DATA pulses advance outward from any particular transmitter / receiver unit to more distant units via power drivers 4126.4126 '4126 ".

A desired switching effect in a selected transceiver can be set by a DATA control pulse over a single control line, e.g. only the CON 1 line is sent. If a DATA pulse on only one line, e.g. occurs on line CON 1, the output of majority selection circuit 4114 is a logic ZERO. The output from the inverter therefore becomes a logical ONE, making the AND gate 4122 effective. If an S1 pulse is received at the same time as the DATA pulse over the single line CON 1, the output of AND gate 4122 becomes "real", which sets the CK input of flip-flop 4124. Since the The D input of the flip-flop 4124 is also "real", because the DATA pulse is present on the CON 1 line, the Q output becomes "real", which generates a C pulse. A Cl control pulse can be used, for example, to deactivate the data bypass relay 4082 '. The bypass relay 4082' is normally held in the position shown, unless a Cl pulse or a power failure (PF) deactivates the relay, via NOR gate 4142, thereby bypassing data from a more distant data acquisition unit around the particular transceiver considered in this discussion. Similarly, relay 4140 is used to switch to shunt conditions in the event of a power failure, as can be seen from the circuit of FIG. 48. DATA pulses sent through individual lines CON 2 or CON 3 similarly generate control signals C2, C3 so that other selected control functions can be performed. The width of a DATA pulse used to activate a desired switching effect in a selected device is half the width of an SI query pulse, i.e. about 600 ns.

Fig. 47, which corresponds in some respects to Fig. 13, shows a special effect based on the fact that the activation of one or more transceivers means the simultaneous presence of an S1 pulse and a DATA pulse on each which requires units. In Fig. 47, a variety of data acquisition i.e. Transceivers purple up to 111g away from the central unit, i.e. Central station 2 arranged. It is believed that it is desirable to activate only the three successive devices IIIc-IIle, but no others, i.e. to make it effective.

A S1 pulse is transmitted from the central station 2 to each device 111 one after the other via the interrogation channel. The moment of arrival of S1 at device 1 IIa is denoted by ta-0. The arrival time of Sl at the device 111b is then tb = 856.8 nanoseconds. The pulse delay between the devices purple and 111b is composed of the cable delay and the 600 ns delay in the delay line 4078 (Fig. 48). The length of the cable between the two units is 60 m; the pulse propagation speed is 391.5 m / ns (1305 feet / ns). Since the cable delay is 256.8 ns and the line delay time is 600 ns, the total delay is 856.8 ns. Thus, the arrival time of Sl at the device 111c tc = 1713.6 nanoseconds, etc., - as shown in Fig. 47. The six timing lines in FIG. 47, labeled IPA-IPF, represent the location of the same SI polling pulse with respect to each of the transceivers llla-lllf at the end of each 856.8 nanosecond polling pulse transit time.

- <»Some time after an SI pulse has been transmitted by query channel 4090, a DATA pulse has been transmitted by control channel 4091. The signal propagation speeds in the twisted wire pairs of the query and control channels 4090 and 4091 are the same. Because of the 600 nanosecond 25 delay line 4078 in each transceiver, the effective S1 pulse rate is less than the control pulse rate because there are no corresponding delay lines in the control channel. In another embodiment, the cable speeds for the two channels can of course be selected so that the delays are impressed in the channels themselves. The taps in delay line 4078 are then used only to compensate for small differences in cable lengths.

35

In Fig. 47, a DATA pulse transmitted 1200 nanoseconds after transmission of a corresponding SI pulse is the SI pulse on the third device, i.e. at the transceiver 111c. The six 40 timing lines labeled DATA A-DATA F show the position of a DATA pulse with respect to the SI pulse at the end of every 856.8 nanosecond interrogation pulse transit time. When a Sl pulse arrives at the purple device, no action occurs on the purple device because the DATA pulse lags 1200 nanoseconds after the Sl pulse. On the device 111b, the DATA pulse is 600 nanoseconds behind the S1 pulse, so that again no action occurs on the device 111b. The DATA pulse receives the S1 pulse on device 111c so that the data processing circuitry in device 111c is activated. At the device llld, the leading end of the DATA pulse leads the S1 pulse by 600 nanoseconds, but due to the width of the DATA pulse, a control signal is available to make the device llld effective. Although the leading edge of the DATA pulse is 1200 nanoseconds before the Sl pulse, the trailing edge has not yet passed the Sl pulse; thus the device 11 le is made effective. At the point in time at which the SI pulse arrives at 60 of the device Ulf, the trailing edge of the DATA pulse finally leads the SI pulse. Therefore, the transceiver Ulf and all subsequent downlink devices are not activated. All devices that are activated by coincident SI and DATA pulses remain active over an entire scan cycle. This means that they respond to all subsequent incoming S2 pulses over the remainder of the scan cycle. The

37

617 017

Desired delays are applied via the tapped delay line 4132 (FIGS. 46 and 47).

The width W of a DATA pulse is

W = (L-1) X DLY + dt where

L = number of transceivers that are activated

DLY = artificial delay line duration (delay line 4078), and dt = small increase in time of arbitrary length in order to enable small propagation time differences.

In the example of Fig. 47, the width of the DATA pulse is W = (3 - 1) X 600 + 300 = 1,500 nanoseconds.

The width of the control pulse can be changed with the aid of the pulse width setting circuit 4130 (FIG. 47) which is connected to the control device 4018 in the central station 2. The pulse width adjustment circuit can be a monostable multivibrator, for example of the National Semiconductor DM 74121 type. A monostable multivibrator is a circuit arrangement that can be used to modify the duration of a control pulse by stretching or shortening the pulse width. The pulse width adjustment is achieved by changing the time constant of a conventional RC feedback network, which is connected to the control inputs of the monostable switching arrangement.

The initial delay time ID given to the DATA pulse by delay line 4132 is

ID = M X DLY,

where M is the number of intermediate transceivers to be skipped between the central station 2 and the first active transceiver (Fig. 47).

As stated above, a CON 1 pulse, when coincident with an S1 pulse, is used to route data past a selected transceiver. The delay ED, which is given to the CON 1 pulse via the delay line 4132 relative to an associated SI pulse, is

BD = K X DLY,

where K is the number of transceivers located between the central station and the selected transceiver.

The triple redundant control lines 4091 of FIG. 48 are shown as single lines in FIGS. 43 to 47 and in FIGS. 5 and 8b, where they are labeled DATA to simplify the drawing.

It follows from the above description and the formulas that any suitable quantity of successive transceiver devices can be made effective by a suitable selection of the control pulse width and control pulse delay with respect to an S1 pulse. For example, the three devices purple, 111b and 11 lc are activated for a first scan. Devices 111b, 111c and llld are enabled for the second scan, third scan may include devices 111c, llld and ll le, etc.

Using this so-called rolling technique, the control pulse width for the first scan is (3-1) X 600 + 300 = 1500 ns. The initial delay is zero because there are no intermediate transceivers between the purple device and the central station 2. On the next scan, the control pulse width remains the same, but the initial delay is 600 ns because a device, namely device 11 la, is between the first selected device 111b and the central station 2 5, etc.

When performing seismic exploration, of course, an acoustic wave is generated during a recording cycle of many seconds and seismic reflection data is recorded. Accordingly, many H) consecutive scans are performed using the same subset of data acquisition units. 6000 samples are taken for a six second record using a one millisecond check interval. After the first recording cycle, the i5 system is "rolled" onto the next subset of data acquisition devices by shifting the delay and a new 6000 cycle recording cycle begins.

As mentioned above, an S1 pulse is transmitted from the central station to the data acquisition units once every millisecond, so that a test interval of one millisecond is defined. If there are fourteen input channels CO-C13 with seismic sensor groups 21, a series of thirteen S2 pulses are transmitted, the pulses 25 being generated at 71.4 microsecond intervals. As explained above, the frequency of the transmitted interrogation pulses is related to the frequency of the reflected seismic signals. For high frequency signals on the order of 200.. the test interval should be half to one millisecond (2000 to 1000 samples per second). For seismic signals at the lower end of the spectrum, e.g. 20 to 30 Hz, the test interval can be two or even four milliseconds (500 Hz or 250 Hz).

As is known in seismics, for the first part 35 of a seismic recording cycle, e.g. the first half to one second receive high frequency signals from shallow underground geological strata. Furthermore, these signals are received at sensor units that are closer to the launch site, since the reflected signals did not have time to reach the 40 sensor units at more distant parts of the cable. Later in the recording cycle, the seismic signals that are reflected from deeper geological strata are characterized by significantly lower signal frequencies.

45 At the start of a recording cycle, for example over the first second, it may be desirable to test the seismic data with a test interval of half a millisecond, using only transceivers and associated seismic sensors that are close to the central station 2 lie, e.g. the devices purple, llld. Accordingly, an S1 pulse and a DATA control pulse are transmitted from the central station 2. The width of the DATA pulse is purple, llld for the four devices

55 (4-1) X 600 + 300 = 2,100 ns.

The initial delay of the DATA pulse is zero because there are no devices that are purple between the central station 2 and the first transceiver 60.

At the end of the 2,000th scan (one second) and over the remainder of the recording cycle, the seismic data from the seismic sensor groups 21 can be checked on the f, 5 input channels in Fig. 43 at lower frequency intervals, e.g. Check intervals of two milliseconds and all transceivers are activated. Thus, a new one is created for the 2001th scan

617 017 38

Control pulse with a larger width correspond simultaneously with the S1 control transmission channels. In addition, impulse can be transmitted. The width of this new control pulse is the gain states of the amplifier stages

Ai-Am can be turned off on a ternary, ocalt or decimal number basis instead of (7-1) X 600 + 300 = 3,900 ns binary basis. The cable arrangement is in

s connection with the marine application are explained, the tow vehicle acting around the seven transceivers 11 la, 11 lg is a ship.

to make sam. The initial delay is the same as above. It should also be noted that the seismic data

ZERO. On the other hand, if all of the transceiver processing equipment of the present invention is to be disabled, the DATA pulse linkage with any seismic pulse source may be "ON" during the entire recording cycle, i.e. I () can be detected, e.g. with explosives, air cannons, Freeine have "infinite" length. frequency sweep vibrators. When signals from a frequency

While the present invention is recorded using a special wobble vibrator source, the exemplary embodiment has been explained with modifications, recorded signals before processing in the beam - the invention is not restricted to this exemplary embodiment. Control device, which is the group shaping device 2184. For example, query and control channels can be correlated to a 15 and transmitted on impulse response form, namely physical transmission channel according to one of the known ones for each of the five hundred sensor group locations or channels. Multiplexing techniques, e.g. the code modulation can be combined. The above description gives the preferred embodiment.

the. For example, query signals and control signals can again form the invention; however, different circuits, switching arrangements and devices modified on each transceiver can be decoded differently, and then different genes can be used instead of those given here without delays being given up to the two signals before the nature of the Invention is deviated. For example, they can be transmitted to the next transceiver, other techniques for multiplexing signals. Although the physical transmission lines are the same as the ones from the individual transceivers, they are used according to the terminology. Special circuits for straightening technology can also have two separate channels. While the invention is being carried out, the beam control and group formation are used in conjunction with at the same distance.

arranged transceiver devices. Furthermore, parts of the invention, e.g. that are amplifying, it is also possible to apply the same principles to a changing amplification system, the multi-spaced devices, plexer-commutated high-pass filters and various others by the sequential delay taps according to FIG. 46 30 switching arrangements and devices entirely general and are designed to be used in a manner that is subsequently the actual use case other than that described here

Differences between the delays in the signal and are.

C.

19 sheets of drawings

Claims (23)

617 017 2 PATENT CLAIMS 9. Device according to claim 7, characterized in 1. A device for processing from data that the transmission time for the control signals is equal to data derived from water signals, in particular seismic multiples of the data mentioned by the delay line, with a large number of individual sections (20, Fig. (4132, Fig. 46) conditional delay time. 1) arranged acoustic sensor groups (21, Fig. 1), the s 10. Device according to claims 4 and 7, characterized in sections electronic transceiver devices that the pulse width setting circuit (4130, (111, Fig. 47) are assigned, which during operation Fig. 47) the duration of the control signals sent to the first control channel (4091, Fig. data signals from the sensors via a transmission link 47) changes, which control channel is an extension (Dj, D2, D3, Fig. 5) in one Cable (12, Fig. 1) has a higher signal propagation speed than a whole common central station (2, Fig. 17) using a multiple of the control signal delay time (Fig. 47). Time-division multiplexing device (114, 116, Fig. 17) transmitted, 11. Device according to claim 6, characterized, characterized by a channel multiplexing device (122) that the first control channel (4091, Fig. 48) is an interrogation channel, Fig. 5) in each transmission -Receiving device for scanning the output signal of each of the sensor selection circuits (4114, FIG. 48) in succession to the majority in each transceiver device in order to respond to like groups of one of the mentioned sections and to transmit an electronic î * in time via the two control channels incoming control signals circuitry (114, 124, 128, Fig. 5) in each transceiver, and that the majority selection circuitry (111, Fig. 47) to output the signals of each transceiver from the data channel receiving device to the transmission link switched off when the mentioned simultaneous arrival of the controls, so that during operation the Output signals of each first signal occurs (Fig. 48). Sensor group per section through the channel multiplexing: »12. Device according to claim 1, for processing seismic direction (122, Fig. 5) and the output signals, characterized in that each transmit / receive signals from at least some transceiver devices (111) a plurality of inputs (21, Fig. successively to the central station via the transmission link 44) for seismic signals and an output channel (4092 ', by a time-division multiplexing stage (131, FIG. 5) of the time-division multiplexing (44) that said multiplexing device (122, plex device are transmitted. : 5 Fig. 44) between the inputs mentioned and the exit 2. Device according to claim 1, characterized in that the channel is arranged, and that the multiplexing device that the transmission link responds to a data channel (4090, FIG. To control signals of the central station (2), in order to 43), a first control channel (4091, FIG. 43), and to connect a second named input to the output channel, control channel (4092, FIG. 43) comprises that the signal propagation so that the seismic signals reach the central station. Switching speed in the two control channels is different. 13. Device according to claims 5 and 12, characterized in that it is large, and that in each transceiver device the control circuit (4018) is a means for (111, Fig. 47) a control network (132, Fig. 5) is present, sending a first control signal to the first control channel which responds to the simultaneous arrival of control signals received via the two control (4091, Fig. 46) and a delay line for sending a signal channel and in the second control signal a first characteristic on the Depending on the response triggers a control command (Fig. Includes second control channel (4090) and that a signal characteristic 5 and 43). shaft identification device (4038, Fig. 45) for resetting 3. Device according to claim 2, characterized by one of the multiplexing device (122) on a selected one of the delay line (4132, Fig. 46) in the central station (2) said inputs are present when the first and second for delayed transmission of control signals via the Control control signal arrive simultaneously in the relevant transceiver channels (4091, Fig. 46) to the transceiver devices -tu device (Fig. 43, 44, 45, 46). at different times, so that the control signals at the same time. tig arrive in the selected transceiver that the control circuit (4018) control signals with a (Fig. 46). second characteristic on the other control channel (4090) 4. Device according to claim 3, characterized by a, and that the multiplexing device (122) on the control pulse width setting circuit (4130, Fig. 47) in the central 45 signal with the second characteristic responds to the switching station for changing the pulse work of one of the two another of the said inputs (21) on said control signals, so that the control signals, if desired, in output channel (4092 '), (Fig. 43,44). more than one of the selected transceivers 15. Device according to claim 12, characterized by coinciding (Fig. 47). a control circuit (4018, Fig. 46) in the central station for 5. Device according to claim 4, characterized by a sc transmission of a series of control signals via the first control control circuit (4018, Fig. 46) in the central unit to the channel (4091, Fig. 46), so that the multiplexing devices (122 ) change the other of the two control signals so that additional control functions can be carried out in a subgroup by the transceivers (FIGS. 46, 47). (111 c to 111 e, Fig. 47) each same at the same time 6. Device according to claim 3, characterized by said inputs of each subgroup to the associated several s.5 output channels arranged in one of the control channels (4091, Fig. 46) in a predetermined sequence connecting delay lines (4132, Fig. 46) for delaying the outside (Fig. 43,44,47). Transmission of the control signal via said control channel (Fig. 16. Device according to claim 15, characterized in 46). net that first all inputs of all send / receive 7. Device according to claim 6, characterized in that directions (111) of the sub-group one after the other and then that said delay lines (4132, Fig. 46) every m all second inputs of the transceiver have the same delay time. Subgroups etc. successively to the output channel (4092 ', 8. Device according to claims 5 and 6, thereby Fig. 44) are switched on. characterized in that the delay lines (4132, Fig. 46) 17. Device according to claim 12, each characterized by an input and an output, and that each of a high-pass filter in each input for the seismic signals, Transceiver devices a detector circuit, which filters each have a capacitor connected in series (123 ', (4103, Fig. 48), which refers to the simultaneous occurrence of each Fig. 16) and a common resistor (315 ", Fig. 16) of a control signal at the input and output of the said include that the resistance in the output of the Multiplexvor- Delay line responds (Fig. 46 and 48). direction (122) is arranged (Fig. 16). 18. Device according to claim 17, characterized by a cut-off frequency setting circuit (360, Fig. 16) for changing the characteristic of said high-pass filter by changing the time during which one of said inputs through the multiplexing device (122, Fig. 16) with said resistor ( 315 ") is connected (Fig. 16). 19. The device according to claim 17, characterized in that the high-pass filter has the cut-off frequency fQ = [1 / (2 3t RC)] [D / T], where R is the ohmic resistance of the common resistor (315 "), C is the capacitance of the said capacitor (123 '), D is the time during which one of said inputs is connected to the common resistor, and T is the time it takes for the multiplexing device (122) to sample all said inputs (Fig. 16). 20. Device according to claim 17, characterized in that an input of the multiplexing device (122, Fig. 38) is connected to a grounded test channel (Co, Fig. 38), that a series capacitor (1051, Fig. 38) between the common resistor (140) and the output channel (4092 'Fig. 44) is arranged, and that there is a switching circuit (3026, Fig. 38) for applying a bias signal to the series capacitor, the bias signal of the same size and of opposite polarity to the unwanted The signal at the connection point between the multiplexing device and the series capacitor is when the multiplexing device applies the test channel (Co) to the output channel (FIGS. 38, 44). 21. Device according to one of claims 12 to 20, characterized by an amplifier cascade circuit (124, Fig. 39 and 44) from a number of amplifiers (Aj - Am, Fig. 39), which between the multiplexing device (122) and the output channel (4092 44), the gain of each amplifier being switchable from one to a higher value and the higher gains of the individual amplifiers being different, an adjustable reference voltage source (3144) for generating a reference voltage for each of the amplifiers, each of which Reference voltages is inversely proportional to the amplification factor of the amplifiers, a control network (132, FIG. 39) responsive to the second control signal with a second characteristic for switching the amplification factors of the amplifiers, a comparison device (3138, FIG. 39) for comparing the absolute value of the output voltage the amplifier cascade circuit with a signal that corresponds to the first ref voltage of the amplifier with the greatest gain factor, the control network switching said amplifier to the high gain state when said output voltage is less than said reference voltage, and an additional circuit, which cooperates with the comparison device and the control network, for switching the Gain factor of the corresponding amplifiers depending on the comparison of the output voltages of the amplifier cascade circuit with the individual reference voltages of the amplifiers in question (Fig. 39 u. 44). 22. Device according to claim 21, characterized in that the larger amplification factor of a selected amplifier is Gj = 22 (m ~ '', where G; the larger amplification level of the i-th amplifier if the amplifiers are ordered according to the decreasing amplification factor, and m is the number of amplifiers. 23. Device according to claim 21, characterized in that a capacitor (3188, Fig. 40) is connected in series with the input of each of the amplifiers, that in the control network (132) a reset circuit (3174, 316, Fig. 40) for connection the amplifier cascade (124) to a test input (Co) of the multiplexing device (122), to which input no seismic signals are supplied, and for setting all of them 3,617,017 Amplifier of the amplifier cascade to gain level one is present, and that feedback means (3184, Fig. 40) are provided for switching on the output signal of the amplifier to the capacitor mentioned, in order to store the output signal of the same size and opposite polarity (Fig. 38 , 39.40). 24. Device according to claim 23, characterized in that an analog / digital converter (126, Fig. 44) is arranged between the amplifier cascade (124) and an output signal register m (128, Fig. 44) for digitizing the amplified output signal and sending the digitized signal on the data channel (4092 ', Fig. 44) that a sensor group (21) is connected to each input (Q - Cn, Fig. 44) of the multiplexing device, and that in the central station (2) one that is seismic Signal processing signal conditioning device (4020, Fig. 43) on the data channel for receiving the digital data from the sensor groups and recording the seismic signals (Fig. 43 and 44). 25. The device according to claim 24, characterized by 2 (i a calibration circuit (136, FIG. 5) arranged in each transceiver for determining normalization coefficients to compensate for relative differences in the signal output voltage of each sensor group, one in the central station ( 2) arranged weighting coefficient memory (1055, Fig. 21) for the programmed weighting of selected digital signals from the sensor groups and multipliers (1053 a, 1053 b, Fig. 21) for multiplying a weighting coefficient by a corresponding normalization coefficient and adding the w product to selected digital signals before they are combined one after the other (Fig. 5.21). 26. Device according to claim 25, characterized in that said cable (12, Fig. 1) by connector modules (13, Fig. 1) into a plurality of seismic cable Î5 cut (20, Fig. 1) is divided into which cable sections contain the data channel and the control channels that the outputs of the sensor groups (21, Fig. 14) with the inputs (Q - Cn, Fig. 14) of the associated transceivers (III) are connected and that in the processing device for 4i) introducing selected phase delays into the digital seismic data before the combination thereof, a group shaping device (2184) is present (FIG. 1.34) 27. Device according to claim 2 for preparing seismic signals, for executing one of the central station (2) 45 triggered desired control sequence in a subset of the substantially identical and equally spaced transceiver devices (llla-lllg, Fig. 47), said devices via the first and second control channels (4090, 4091, Fig. 47) 50 and the data channel is connected to the central station, characterized by a control circuit (4018, Fig. 47) arranged in the central station (2) for transmitting a first control signal to the first control channel at specific intervals, one delay line each (4102, Fig. 46) to delay the first control signal so that the propagation speed of the first control signal in the first control channel is reduced by a predetermined amount in each transceiver, delay circuits (4132, Fig. 47) in the central station for transmitting a second control signal signals on the second control channel after a delay after sending the first step signal, wherein said delay time is a first integral multiple of said predetermined amount, and a signal coincidence detector (4114, Fig. 48) in each transceiver to determine the simultaneous occurrence of the two control signals (Fig. 46,47,48). 28. The device according to claim 27, characterized in that the delay circuits in the central station 617 017 a pulse width setting circuit (4130, Fig. 47) comprises for extending the duration of the second control signal by a second integer multiple of the aforementioned predetermined amount, so that one of the transceivers of the subgroups can be controlled by changing the delay of the second control signal (Fig. 47 ). 29. Device according to claim 28, characterized in that each transceiver has a number of input channels (Cj-Cn, Fig. 37) for analog signals, that each transceiver has the same number of seismic sensor groups (21, Fig. 37) and that each transceiver includes electronic circuitry (3011,3Ö33, Fig. 37) responsive to the output of the signal coincidence detector for filtering, sampling, amplifying and converting the analog seismic signals into digital data while digital data is from another channel are transmitted to the central station via the data channel (FIG. 37). 30. The device according to claim 29, characterized in that the control circuit (4018) is constructed such that the digital data is received and in the central station (2) in columns which correspond to the sequential arrangement of the transceiver in question and in lines which the Order of scanning of the analog signals in the transceivers are stored (Fig. 47). 31. Device according to claim 29, characterized in that the data channel (4092, Fig. 43) comprises a triple broadband conductor (Dj, D2, D3, Fig. 5), and in that the central station (2) an error detection and majority selection device (1028, Fig. 21) for receiving digital data from two of the three broadband conductors and for detecting an interruption on one of the three broadband conductors (Fig. 5, 21,43). 32. Device according to claim 29, characterized in that the control channel (4091, Fig. 48) has delay circuits (4078) and triple conductor pairs (D1; D2, D3, Fig. 6), that each transceiver has a majority selection circuit (131, 6) for selecting control signals supplied via two of the three pairs of conductors and an error display circuit (130, FIG. 6) for indicating an interruption of the connection on one of the three pairs of conductors (FIG. 6). Seismic exploration generates an acoustic wave signal at or near the surface of the earth. The acoustic wave travels downwards and is reflected by underground layers, from where the wave returns to the surface of the earth. The reflected seismic waves are displayed by sensitive seismic sensors, which are arranged in a linear grouping or row distributed on or near the surface of the earth. In current practice, 25 to 30 seismic detectors, often referred to as sensors, are equally spaced along a single section of cable that can be 30 to 100 meters in length. The sensors are all electrically connected to each other so that they form a single elongated data channel. 50 or more such cable sections, each of which represents a channel for signal transmission, are coupled to one another to form a cable arrangement of 3,000 m or more. The output of each or more channels is connected to a central signal processing device which is provided at one end of the cable arrangement. A separate pair of cables is required for each channel. There must therefore be 50 or more such pairs of lines. Because of the cost and weight of such a large number of Lines have been proposed in which the output signals from each channel are time-division multiplexed through a single data transmission channel. The foregoing invention is explained below in connection with geophysical exploration 5, but it can equally be used to indicate objects that are under water, e.g. big fish, schools of fish, ships or the like. Known similar devices belonging to the prior art result, for example, from US Pat. Nos. 3,652,979,3,748,638,3,851,302 and 3,873,961. All of these aforementioned patents relate to telemetry devices in which a signal processing device uses the seismic data digitally from a seismic sensor or a seismic i 5 sensor group. Using various multiplexing methods, the data are transmitted to a recording device via a telemetry channel. As mentioned above, a seismic sensor group is an arrangement with a length of 30 to 100 m. A zo acoustic wave, the wave front of which runs approximately parallel to the plane in which the seismic sensors are arranged, arrives at all sensors approximately simultaneously, as a result of which the output signal is amplified by summation. This ideal case does not occur under actual seismic prospecting conditions, except in the case where the wavefront is vertically reflected by a reflecting, underground interface that lies directly below the sensor arrangement. In the far more common case, the acoustic wavefront is at an angle transverse to the arrangement, so that the wavefront arrives earlier at one end of the arrangement than at the other end. The traveling time of the wavefront over the arrangement depends on the phase velocity of the seismic wave, the wavelength of the acoustic wave, the length of the arrangement and the depth of the reflecting interface, but also also on other factors. If the length of the arrangement is long compared to the wavelength of the acoustic wave, the arrangement dampens the acoustic wave. Therefore, the seismic sensor arrangement (or series) should be shorter than the wavelength of the incident 4 "seismic waves in order to avoid damping, and preferably the arrangement should be shorter than a quarter of this wavelength in order to obtain a good response. So far, the energy at the lower end of the seismic spectrum with wavelengths of many hundreds of meters has been of interest. The use of fairly long arrangements was considered satisfactory. More recently, however, a greater resolution of geological stratifications, e.g. required when researching underground structures. A higher resolution requires the use and thus 5o display of seismic waves of higher frequencies. However, the wavelengths of energy at the upper end of the seismic spectrum are comparable to the length (and often much shorter than the length) of conventional seismic sensor arrangements that are in operation today. It is therefore desirable to replace a long seismic sensor array in each cable section with many shorter seismic sensor arrays. However, such a solution to this problem has so far been considered impossible in view of the difficulties in data processing and the large number of conductors that would require 60, because it would require extremely expensive seismic cable arrangements, the size and mass of which from a practical point of view would no longer be portable.