Device for controlling a member in a system

TECHNICAL FIELD

The present invention relates to a device for preventing the function of a system being affected on account of a first member in the system being impaired by, for example, manufacturing irregularities, supplier varia¬ tions, being exposed to external and internal environmental influences, or operating with a function which is dependent on the material, units or parts which the first member will act upon in the system. This can in this case be a product-generating system. The device is furthermore of the type in which the first member is controlled from a second member, for example as a function of the movements, state etc. of the second member. The invention also relates to a method in conjunction with the said device.

PRIOR ART

It is previously known to correct or compensate for deficiencies and manufacturing variations in a given member which forms part of a system. The member can, for example, be manufactured with greater tolerance accuracy or be made to cooperate with incorporated or connected compensating devices which compensate for changes in, for example, temperature. In cases where the system requires rapid functioning of the first member and the latter operates with a comparatively great inertia, it is pre¬ viously known to arrange for the first member to be controlled such that it is acted upon in advance in order to compensate for the inertia. It is also previously known, with communication protocols which operate with well-defined time transmission of signals and in similar arrangements, to use a message system in which the messages are numbered and transmitted in a predetermined

order, for example in an order which corresponds to the message numbering.

DESCRIPTION OF THE INVENTION

TECHNICAL PROBLEM

In systems of the relevant type, there is often a sig¬ nificant requirement to be able to make the system as such operate with great accuracy. This is made difficult by the fact that components incorporated in the system are sensitive to external and internal factors which result in their having a function which is dependent on tolerances upon manufacture, exchange of brand, temperatures or other factors. It is also a great advantage for the system to have a structure which is such that it can comprise technically simple and economically advantageous components. The last-mentioned requirements conflict with the requirement for great system accuracy, since such accuracy presupposes accurately manufactured components which are incorporated in a non-critical cooperating function in the system.

The present invention proposes a device and a method which solve, inter alia, the abovementioned problems. By means of the invention, the mechanical inadequacies of the components involved can be compensated by a system function which is built into a software used. This can be designed so that the mechanical inadequacies are easily compensated.

In systems of the relevant type, there are also difficulties in generating and transmitting the signalling from the second member to the first member so that clear function executions are obtained from the first member. It is also desirable to be able to use and deal with information from a non-intelligent pulse source (sensor) for producing exact function executions in the first member. It is also desirable to be able to make the components easily exchangeable in the system without the

latter having to be subjected to modification on account of the replacement. It is also desirable to be able to make the system useable for system changes. In some connections it will perhaps be desirable to use cheaper/simpler components than those already in place, or vice versa. The signal transmission will be non- critical and receiving and separating true signals from interference signals will be salient. It is also in some cases advantageous to be able to use known, well-proven communication protocols, even where these are in principle intended for other signal handling, for example prioritising messages. This concerns in particular those circuits in which the protocol handling is built into hardware, i.e. existing known systems (for example CAN, SDLC, GPIB, Multibus, etc.) are used in an entirely different way.

In a system where certain events will give functions in other parts of the system, the signal is normally trans¬ mitted directly from the sensor to the unit which will execute the function, the magnet. Since many events will take place in a machine, this requires a great quantity of conductors, and the unit which delivers the signal on these conductors must take account of the delays which can occur on the way from when the output signal arises to when the desired function is effected.

SOLUTION

The new device and the new method are intended to solve the said problems in an advantageous manner.

The features which can principally be regarded as being characteristic of the new device are, inter alia, that the system comprises transmitter and receiver members and, arranged between these, a connection in the form of a bus connection. The transmitter member receives signals which emanate from the particular parameter or parameters of the second member. The receiver member can form or be

connected to a control module for the first member. The transmission between the transmitter and receiver members is effected by means of a message system in which the messages are preferably numbered and can be transmitted from the transmitter member to the receiver member in a certain order via the said bus connection. The first detection member's signals/said emanating signals are assigned message numbers as a function of each parameter value. The lowest parameter value is preferably assigned the lowest message number, the next-lowest parameter value is assigned the next-lowest message number etc. A communication protocol which operates with great time accuracy is used in this connection, and examples which can be used are communication protocols CAN, TRAINET, etc. which are known on the market. The transmission times in the signalling between the first detection member and the transmitter member and between the transmitter and receiver members are in this way well defined. An adjustment of the first member to the desired parameter value by means of a control member incorporated in the system, for example a master unit, causes the receiver member/control module to receive a message which contains the parameter value in question. The control module receives, preferably from the control member, time information for the desired function execution of the second member. The control module generates, as a function of the parameter value and the time information, an activation signal for the first member, which activation signal is produced in such a way that it takes into account the first member's present inertia, which can vary in accordance with the above. Upon receiving the message number in question, the control module preferably generates, with the aid of this, a trigger signal for the first member so that the latter effects its function execution. A second detection member is moreover arranged to detect the actual function execution of the first member on account of the trigger signal and, in the event of a difference from the desired function execution, a signal is generated for the control unit which as a

function of the last-mentioned signal adjusts the receiving member for receiving a message number which is different from the preceding message number and which represents another parameter value which comprises compensation for the existing inaccuracy between the desired and actual function execution and generates a trigger signal which is delayed or accelerated in relation to the first trigger signal. A variant of the above is that the receiving unit can transmit a message to the transmitting unit in order to cause the latter to transmit other parameters which it is expedient to obtain for achieving a good trigger point.

During system operation, the first detection member preferably emits in an essentially continuous manner signals associated with each parameter, and in a cor¬ responding manner there is essentially continuous transmission of the message numbers which represent the parameters in question on the bus connection. The second detection member can comprise a sensor member which detects the first member's actual execution of its function. The second detection member can also comprise a time-measuring member which generates a signal which is dependent on the difference between a time instant for delivery of activation energy to the first member and a time instant when the first member has executed/carried out its function. The control member can in this case be arranged to calculate a lead time, dependent on the time difference, in the form of one or more messages for activation of the first member. One transmitter member at a time preferably operates on the bus connection in order to avoid risks of collision between messages from dif¬ ferent transmitter members and, consequently, time delays which can jeopardise the trigger signal generation. The transmitter member continuously transmits the detected parameter/parameters or all the detected parameters, for which the receiver member can be adjusted from the control unit. Each message thus represents with its message number a value for the detected parameter(s) and

at the same time the message can contain data which represents another quantity in respect of the action, state, etc. of the second member, for example angular velocity, direction, acceleration, etc., an action, state, etc. of an article controlled by the first member, or a number of sensor elements arranged on the first member. With the aid of the information from the said action, state, etc. and of data associated with a message transmitted and received in the receiver member regarding another quantity/other quantities concerning the second member, for example angular velocity, acceleration, direction of rotation etc., an actual event, for example final function establishment, in the first member can be calculated (in a manner known per se) afterwards.

The system for the communication protocol operates with different phases. A primary adjustment phase for start¬ up of the system is used, as well as a secondary adjust¬ ment phase in which a main processor/control member incorporated in the system gives instructions regarding trigger signals which are to be selected. In addition, an operating phase is used in which each subunit/slave incorporated in the system receives an assigned message/assigned messages in order to effect one or more trigger signals as a function of this/these. The secon- dary adjustment phase and the operating phase can be repeated cyclically on instructions regarding new trigger points/trigger signals. In a case where the trigger points are not altered, the secondary adjustment phase is omitted. Further features of the new device will emerge from the description which follows.

The method is characterised, inter alia, in that signals emanating from the particular parameter/parameters of the second member and dependent on each parameter value are transmitted on a transmission system operating with a well-defined time function and of the type which com¬ prises, on the one hand, a transmitter member to which the said emanating signals are conveyed, and a receiver

member which is connected to or forms a control module for the first member. A bus connection is arranged between the transmitter and receiver members. The transmission system operates with a message system in which the messages are preferably assigned message numbers and are transmitted in a certain queue order, preferably successively as a function of the number value. The said signals which represent the different parameter values are preferably assigned different message numbers, for example the lowest parameter value is assigned the lowest message number, the next-lowest parameter value is assigned the next-lowest message number, etc. The control module is controlled from a control unit incorporated in the system, for example a master unit, or is pre-programmed in order to be responsible for one or more message numbers. Time information, position, temperature, etc. are or will be assigned to the control module regarding the time, the position, the temperature etc. when the first member will have carried out its function.

With the aid of the said parameter value and time information etc., an activation signal is generated which results in the function being executed at the desired time regardless of the inertia with which the first member operates in the case in question. In one embodi¬ ment, a trigger signal is generated in the control module for the first member as a function of each received message number for which it has been made responsible by the control unit. The function execution of the first member is detected, which can take place in a manner known per se by means of sensor members or with the aid of the action, state, etc. of the material, element, wire etc. which is controlled or acted upon with the aid of the first member, which in this case can serve as an actuation, holding or stopping member, etc. In the case of a function detection where the function execution of the first member differs from that which is desired, a correction signal is generated and is used for producing,

for example calculating, new control information for the control module which, when the new control information is received, is made responsible (instead) for a message number which differs from the first-mentioned message number or the first applicable message number. With the aid of the new message number, there is generated in the control module a trigger signal which is delayed or accelerated in relation to the first-mentioned trigger signals.

ADVANTAGES

With the aid of the directions in accordance with the above, an exact and accurate system can be built up in conjunction with a product-manufacturing machine, unit, installation etc. (for example a textile machine) . Even if the transmission paths between the signal-transmitting (signal-generating) point at the second member and the signal-receiving point have a considerable length, great accuracy can be obtained in the trigger signal generation for the first member, despite the fact that this is not constructed or chosen with advanced tolerance charac¬ teristics and little spread between different examples. Short distances between sensors and the communication modules together with data communication on which interferences can easily be detected and separated out (bit test and/or check total) gives an extremely reliable transmission system. The system permits transmission of other information when the transmission system is not used before the trigger signal messages.

LIST OF FIGURES

A proposed embodiment of the invention is described hereinbelow with reference to the attached drawing, in whichJ

Figure 1 is a block diagram of a transmission system used in the invention, with a master 1014 and four slave modules. The buses 1006 and 1013 can in some cases be one and the same.

Figure 2 shows in block diagram form parts incor¬ porated in the system in the module which transmits from the sensor the messages on the bus 1005' which gives trigger response in other receiving modules.

Figure 3 shows in block diagram form the parts which can be incorporated in the module which receives the trigger messages and as a function of the received trigger message performs a desired action on the received message.

Figure 4 shows how back coupling can be performed on the receiver module in order to achieve the correct delay between the trigger message and execution of the function.

Figure 5 shows a system with control member, control module and transmitter and sensor members.

DETAILED EMBODIMENTS

As an example a reference may be made to the case in which, in an additional installation for a machine, an object, unit etc. is to be released at a certain angle on a drive shaft used in the machine. The first member can consist of a magnet member which serves as a holding member and which as a function of a trigger signal will release the object, part etc. in question. The magnet member is subject to a certain delay between the supply of activation energy and the time when its member has

carried out or executed its function. Such a delay differs from magnet to magnet and can vary with temperature, material stress, material quality etc. for one and the same electromagnet. Even greater variations can occur when magnets or the like from different sup¬ pliers and/or of different marks or models are used. The efficiency in the machine is increasing with today's requirements, which thus imposes increased requirements for a small spread between magnets and for each magnet used to be able to operate with unaltered function under different conditions. According to the invention, the machine is used for ordering material release at a certain angle, at the same time as information is con¬ tinuously picked up regarding angles so that a necessary lead time can be calculated. By virtue of the fact that the receiver module itself can determine which trigger message will be used, this can take into account all the variations which can occur at the function-increasing point without adjustments having to be made in the other parts of the system. The precision can thus increase over the system as a whole, without it being necessary to increase the requirement for each individual sub¬ component. According to the invention, use is made of a known communication protocol/transmission system, see above. In this, a known circuit (Philips) can be used for transmitting messages, for example from 0 to 2032. In addition, an Intel circuit can be used for receiving one or more of the said messages.

According to the concept of the invention, a sensor is used or coupled to the main shaft/drive shaft and is read off by a microprocessor (CPU). For each degree, this transmits a message with a number, which corresponds to the degree number in question. It is also possible to add to the message the actual angular velocity and, if appro- priate, also the direction of rotation and acceleration in order to give the receiver module further information in order to make a better calculation of the time at which the function execution is to be started. There is

thus transmitted on the bus connection a continuous stream of messages, one for each degree. Since only one module transmits trigger messages or other information on the connection, there is no risk of collision and, there- fore, the transmission speed can be precisely known. The time from which the pulse sensor indicates until the message is transmitted can also be precisely known, which can take place in a manner known per se. With a pulse sensor with considerably more pulses than one per degree, the said transmitter/transmitter member can also extra¬ polate values so that it transmits the correct value at the correct time. As speed, acceleration and the like are transmitted, the same calculation can also be carried out in the receiving module.

The receiver, the magnet control module, comprises a sensor which indicates the passage of the material, object etc. beyond the armature of the magnet. A main processor/master unit incorporated in a CAN system orders the module to release the object or similar at 90°. With adjusted default value, for example 88°, the module provides current to the magnet, and the time for the object, material etc. to pass from a holding position to a sensor position is measured. In order to obtain the trigger point 88°, the module ordered a CAN circuit incorporated in the CAN system to receive message number 88. When the circuit receives the message, it transmits an interrupt/activation signal to the microprocessor (CPU) which sets off the magnet coil current and starts a time counter. When the passage of the object or material is indicated, the counter is stopped and read off. If data is transmitted with the message, the data can also be read off from message 88 (for example angular velocity and acceleration) . The module can then calculate when the object or the material was in actual fact released. Assume that this occurred at 91.45°. It then asks the CAN circuit or similar function module to pick up message 86 instead next time. When this message arrives, the module waits a time which corresponds to,

for example, 0.55° and then sets off the magnet. If the machine/drive shaft does not change speed, the function execution for the first member will this time be correct. The circuits in the system CAN, TRAINET, etc. are rapid and can operate with a resolution of half a degree without problems, and this too in the case of rapid machines, i.e. rapid drive shaft rotations (for example 1200 rpm) . Of course, not all the degrees of a turn need to be transmitted via the communication, but only those which are important for achieving a sufficiently good functioning of the function which is to be triggered. An example which can be mentioned is that one function is to be obtained at 10 degrees and another at 180 degrees. In this case, all angles between 0 and 15 and between 160 and 190 degrees are expediently transmitted. The other degrees do not need to be transmitted, since it is not necessary to obtain any function at or around these degree numbers. It is also conceivable that in the one case between 0 and 15 degrees each degree is transmitted, whereas, between 160 and 190 degrees, only every second degree need be transmitted. In a more complex form, it is conceivable for the receiving unit to transmit such messages so that the transmitter only transmits those messages which are necessary for fulfilling the desired function. The transmitter must then take into account all the receiving units so that for these a sufficient number of trigger messages is transmitted.

A detailed exemplary embodiment is described below, with the following reference numbers being used.

Figure 1

1001 Rotating shaft, direction of rotation arbitrary.

1002 Sensor, for example position in degrees.

1003 CAN adaptation, circuit which sends different messages for serving the said sensor, sends the trigger messages and receives adjustments. This CAN circuit should be of the Philips type so that all

message numbers in the CAN system can be used. The circuit can also be connected to a communication bus 1013. 1004 Conductor for trigger signal from sensor. 1005 First connection to CAN bus for trigger messages.

1006 CAN bus for trigger messages.

1007 Second connection to CAN bus for trigger messages.

1008 First handling processor which receives specific trigger messages from the CAN bus for trigger messages. The processor executes operation on trigger message, checks that the operation is executed at the correct time and adjusts the delay. The system is back-coupled. The processor receives and transmits information on the CAN bus for control.

1009 Connection to CAN bus for trigger messages.

1010 Second handling processor which receives specific trigger messages from the CAN bus for trigger messages. The processor executes operation at the correct time and adjusts the delay. It is incor¬ porated in the back-coupled system and receives and transmits information on the CAN bus for control.

1011 Connection to CAN bus for trigger messages.

1012 Third handling processor which receives specific trigger messages from the CAN bus for trigger messages. It is incorporated in the back-coupled system and receives and transmits information on the CAN bus for control.

1013 Second CAN bus for control. 1014 Main processor (master) which ensures that the system fulfils the desired total function.

1015 First connection to the said second CAN bus for control.

1016 Second connection to the said second CAN bus for control.

1017 Third connection to the said second CAN bus for control.

1018 Fourth connection to the said second CAN bus for control.

Fiσure 2

1101 Processor.

1102 Internal memory RAM and ROM.

1103 CAN circuit for connection to CAN bus for control. 1104 Adaptation circuit for sensor (input member).

1105 Internal bus in slave unit.

Fiσure 3

1101' Processor.

1102' Internal memory RAM and ROM. 1103' CAN circuit for connection to CAN bus for trigger messages. 1104' Adaptation for sensor (input member) gives the delay from when the control pulse has been fired from the output member (1106). 1105' Internal bus in the slave.

1106 Adaptaiton circuit for control member (output member) . The circuit gives a control signal to the physical unit which carries out the operation (in this case a magnetic hammer) . 1107 CAN circuit for connection to CAN bus for trigger messages.

1108 Connection to control member (magnetic hammer) .

1109 Connection to accelerometer or other suitable sensor. 1110 Control member in the form of magnetic hammer. 1111 Stop for magnetic hammer with sensor.

The system according to Figures 1-3 operates with three phases, and two of these can be repeated cyclically.

Phase 1: Primary adjustment phase takes place upon start-up (current supply) of the system. Basic adjustments are carried out.

Phase 2: Secondary adjustment phase takes place when the main processor gives instructions regarding new trigger points which are to be chosen.

Phase 3: Operating phase is when each slave (1008, 1010, 1012) waits, receives trigger message and/or executes handling.

Phases 2 and 3 can be repeated cyclically. Phase 1 can give instructions on trigger points and, if these are not altered, phase 2 disappears and in this case only phase 3 is repeated cyclically.

Phase 1: The main processor (1014) sends adjustment information to each slave (1008, 1010, 1012) regarding which messages are to be used. One of the slaves (1008, 1010, 1012) adjusts the para¬ meters of the sensor slave (1003) and the messages which are to be used as trigger messages. Alternatively, the main processor (1014) with an extra CAN port which is con¬ nected to CAN bus (1006) for trigger messages can be used instead of one of the slaves (1008, 1010, 1012) being used for adaptation between the main processor (1014) and the sensor slave (1003). (Not shown in Figure 1).

Phase 2: The main processor (1014) indicates new trigger points if so required. Each slave (1008, 1010, 1012) calculates its trigger point parameters based, on the one hand, on data from the main processor and, on the other hand, on informa¬ tion from external/internal sensors (not included in the figure) . The slaves then wait for each trigger event to occur.

Phase 3: Trigger event occurs when sensor (1002) trans- mits trigger signal (1004) to sensor slave

(1003) which then transmits the correct trigger message on CAN bus for trigger messages. Slave which accepts this trigger message performs the necessary operation and reconciles the calculated trigger point with the chosen

trigger point which has been obtained from the main processor.

The function can alternatively be described as follows:

Phase 1: The main processor (1014) sends out a command to each handling slave (1008, 1010, 1012) which is then assigned its basic parameters. The main processor (1014) then sends a command to the sensor slave (1003) either through the handling slaves (for example 1010) or via a separate CAN bus connection to CAN bus for trigger message

(1006, which is not shown in the figure concerned) . These commands include information on the messages which are to be used, for example if the sensor (1002) gives the position in degrees between 0 and 359 but instead trans¬ mits messages with numbers in the range 1 to 360 where 360 corresponds to 0.

Phase 2: The main processor sends the different trigger points to each handling slave (1008, 1010, 1012) and appropriate commands to the slaves

(1003, 1008, 1010, 1012) for altering resolu¬ tion and interpretation of message numbers.

Phase 3: The handling slaves calculate which message number will start each operation. Here there is a back-coupled loop (see Figure 12, 1104 and

1106). The shaft (1001) rotates in an arbitrary direction. The sensor (1002) indicates the direction and step and sends information on via connection (1004) to the sensor slave (1003) which immediately sends the corresponding trigger message on CAN bus for trigger message (1005, 1006) (implemented with interruption or polling). The handling slave (1008, 1010, 1012) which has this message as trigger point then executes phase 2 or phase 3 again.

In the above example it is assumed that the effecting module/control module receives only a message that activation will take place. However, the control module knows that the time for the execution of the function of the first member will take place later, i.e. knows that there is a certain inertia in the first member. In one exemplary embodiment, the control module receives the message regarding time information, for example from the control member. A function sequence can then be as follows:

1. The control module receives or possesses informa¬ tion on how long a period of time is available for execution of the function of the first member in relation to the message stream with the parameter values/degree numbers. The control module deter¬ mines on the basis thereof which parameter value A will be received in order to initiate the control module's own operation. In addition, the CAN module is adjusted to receive a desired parameter value B regarding time information when the control module is expected to precisely fulfil the desired effect. The sensor (cf. 1111) gives a signal when the execution of the function has taken place exactly.

2. When the parameter value A is received in the control module, this starts the actual sequence.

3. Thereafter, the CAN module either generates a signal when B is received or when the signal from the sensor (cf. 1111) is received internally. The said signals start a time counter which also indicates which signal was obtained first, for example + for B and - for the sensor signal.

4. When the sequence has been carried out, the micro¬ processor (CPU) of the module reads off the time in the counter, and in this way the module establishes how well the task has been achieved. If B and the

sensor signal come simultaneously, a random selec¬ tion of the signals can be made for example.

The sensor signal need not necessarily be generated by the control module, but can be obtained from any other suitable source in the system. The only requirement is that the time delay in the first member will be known or can be determined.

In the exemplary embodiment according to Figure 4, the preconditions are as follows:

1. The software time and/or hardware time of the handling slave for starting time-critical opera¬ tion on the output member is known, let this be t _b .

2. The time delay for "PRE TRIGG" from the sensor through the sensor slave to interruption in the handling slave is known, let this be t _Et .

3. The time delay for "REAL TRIGG" from the sensor through the sensor slave to interruption in the handling slave is known, let this be t _rt .

4. The time delay for "CONTROL TRIGG" from sensor used to interruption in the handling slave is known, let this be t _ct .

5. The system function of the output member (delay) is approximately known, let this be t _ct .

6. Shaft with sensor slave's system function (trans- mission function) is approximately known.

The following procedure is assumed:

Assume that an angular velocity w _r of the shaft 1001 in Figure 1 is approximately constant and function execution will take place at an angle w _rt =150°. The hammer 1110 in

Figure 3 with srtike control sensor 1111 at the saWβ't±me as the sensor slave indicates the angle w _rt .

Event:

Turn 1 The handling slave calculates the trigger point shift with given data.

(t _pt +t _b +t ₀ )*w _r *=Dw

This means that the operation takes place DW degrees too late.

w _pt =w _rt -Dw

conclusion, choose w _pt as "PRE TRIGG", but this can be improved by t _rt and t _ct being known, as well as the times when these two events have taken place (see block diagram) . A hardware calculator can be used if maximum time resolution is to be achieved and started and stopped, respectively, by the said two signals. The time difference t _j between these two signals will be used for adjusting w _pt at the next turn.

Turn 2 Now t _j is known and compensated by t _rt and t _ct . Now calculate new "PRE TRIGG"

(t _pt +t _b +t ₀ +t _j )*w =Dw

w _pt =w _rfc -Dw

At the end of this cycle, a new t _j is obtained.

A corresponding model is used if w _r is not constant.

In Figure 4, calculating circuits of known type have been shown by 1201, 1202 and 1203. A control circuit is shown by 1204 and a processor, textile machine etc. is shown by

1205. The function of the different circuits will emerge

from the context.

Figure 1 describes more exactly a system where every¬ thing originates from the sensor 1002, which in this case is considered to be a sensor, for example, which indi- cates the angle of a shaft. There can be a number of such sensors with similar tasks in a system, but in Figure 1 only one is shown. The sensor is acted upon by any physical property, in this case a rotating shaft 1001, which sits in a machine which the control system will control. The invention is particularly applicable where different parts in the machine are to be synchronised to some part in the machine. In this case, the shaft 1001 constitutes a mechanical master, on the basis of whose position other events in the machine will be carried out. The angle information from the sensor is transmitted via a conductor 1004 or other information carrier to the transmitter member 1003. The transmitter member 1003 can have functions in the system other than transmitting angle information, but in this example this is not shown. The fact that in a machine or a system there is a main shaft to which all other functioning in a machine will be synchronised is well known. This is usually carried out mechanically by other parts in the machine being connected to the said shaft mechanically with gear wheels, belts, drive shafts etc. It is also possible for the synchronisation with the main shaft to be carried out by angle or position information being transmitted to other units by virtue of the fact that the conductor 1004 is connected to all other units so that all the units concerned can read off the position directly. In this case the information goes via a transmitter member 1003 which transmits the information as a message on a communication conductor 1005 which, for example, can follow the specification for the CAN system. The transmitted information can be picked up by the modules 1008, 1010 and 1012 and, in cases where the communication channels 1006 and 1013 are one and the same, also by the module 1014. In Figure 1, the system is divided into two

communication channels 1006 and 1013. The channel 1006 is used for trigger messages and the channel 1013 is used for other information. The advantage of this is that the trigger information can be transmitted without being disrupted by other transmission in the case where there is a restriction that it is only 1003 which may send on 1006. All other control and information transmission is then conducted via the channel 1013. In the case where the transmitter member 1003 must also obtain information, this too must have two communication ports like the modules 1008, 1010 and 1012, so that connection to the communication channel 1013 can take place. The connections 1005, 1007, 1009 and 1011 are physical connections of each module 1003, 1008, 1010 and 1012 to the communication bus 1006. These connections consist first and foremost of a physical connection to the information carrier 1006, which can be a conventional electrical conductor, but optical and radio transmission are also possible. In the connection there must also be an adaptation between the way in which information is present on 1006 and how the information will appear in the modules where it is generally treated with the aid of a computer which in this connection generally operates with digital levels of 0 and 5 volts in accordance with the CMOS or TTL specification. In a similar way the modules 1008, 1010, 1012 and 1014 are connected to the channel 1013. In cases where trigger information can be mixed with the remaining information in the system, the connections 1006 and 1013 can be one and the same conductor. In such a case, only one connection of the modules 1008, 1010 and 1012 to the communication conductor (1013/1006) is needed, with the result that, for example, the connections 1007, 1016 and 1011 can be removed. There can be, however, other reasons, for example limited function in the communication circuits 1107 or 1103', which mean that the same module can have a number of connections to one and the same bus.

Figure 2 shows a more detailed embodiment of the module

1003 according to Figure 1. There is normally a processor in the module, which manages the function of the module. This processor 1101 follows a program which is stored in a memory 1102. This memory is normally divided into two parts, one for the program, which normally it is only possible to read after it has once been programmed, and a part which can both be read and written for storing occasional information, data and programs. In order for the processor to be able to cooperate with the surroundings, different types of connection possibilities are needed. Two such pos-sibilities are shown in the figure. One connection 1104 is included in order to convert the information from the angle sensor 1002 via the conductor 1004'/1004 to a digital word which can be read and thereafter interpreted in the processor's logic unit or directly written in the memory or transmitted directly to 1103 via the adaptation 1005'/1005 and the communication line 1006. In order for the different parts 1104, 1102, 1101 and 1103 to be able to cooperate, they are coupled together with the databus 1105 in the module 1003. This databus consists of address and data lines, and control lines to the latter. In addition to this there are voltage feed, clock signals and other signals which can go to parts or all of the units which are connected to the bus. The figure does not show the voltage feed, clocks, capacitors and other parts which are included in all modules with microprocessors. The communication unit 1103 is an important part if the invention is to achieve the desired function. This unit is normally manufactured using the same basic technology as a microprocessor, i.e. it operates with CMOS or TTL levels and operates digitally with two binary levels. Since the communication can take place with other signal levels than CMOS/TTL, in optical or modulated manner, in most cases an adaptation 1005'/1005 is needed before information can be transmitted on the communication conductor 1006. The task of the communication unit 1103 is to ensure that informa-tion which is to be transmitted arrives at the bus in such a way that the other units

which read in the information from the bus receive it in an unambiguous way. The task can be divided into the following parts.

1. Ensure that the information arrives at the bus 1006 in such a way that it does not collide with other information transmission.

2. Ensure that information arrives at the correct speed, so that the other units interpret each part of the information in a correct and unambiguous manner.

3. Ensure that the different parts in the transmission are in the order which the receivers require.

4. In addition, in some cases the circuit can check that information appears correctly on the co - munication line by simultaneously reading in information from it and comparing with transmitted information. It is also possible for the transmit¬ ter to transmit a check total which the received units can then compare with the check total which they calculate on received information.

Figure 3 shows in detail the module 1008 in Figure 1. The module receives the information from the transmitter module 1003. The structure of this module 1008 is in many ways similar to that of 1003. It comprises a processor 1101' which is connected via an internal databus 1005' to a memory 1102' and a communication unit 1103', and an adaptation to the surroundings 1104' which can be the same as or of a similar type to 1104. There can be times when 1003 and 1008 can be identical in purely physical terms. After start-up, different tasks are received. By choosing from different programs in the program memory 1102 or by transmitting the function program after start¬ up, they can thereafter perform different tasks. In the same way the hardware can be programmed or used in different ways after start-up. In comparison with Figure

2, the unit in Figure 3 has a further two units which are not present or shown in Figure 1. This is a communication unit 1107 and control output/adaptation for an external unit 1110, the hammer, which is controlled via an infor- mation carrier 1108. In this case, no angle sensor is connected to the module via 1104', but another type of sensor 1111 which has been connected via the conductor 1109.

Figure 4 describes with a block diagram how the unit 1008 (Figure 1) or 503 (Figure 5) can process the information in order to obtain the function desired in the invention. Figure 3 shows a typical set up where this algorithm can be used. Also 503 together with 506 and 507 is a good example, where 506 and 507 can be compared with 1111 and 1110. The important point for solving the existing problem is to keep track of the delays in the system. This is quite common. The new aspect of the invention is, inter alia, how some of these times can be calculated and how others can be adjusted in order to obtain a better function for the overall system. The time tpt from when a certain angle at 1003 is known until this information has reached 1008 will be known with sufficient accuracy, as will the time tb which it takes before the unit 1008 is able to transmit the signal to 1110 via 1108. In addition, the reaction time to can be calculated roughly, in addition to the variations which occur on account of environmental factors, manufacturing method and brand of manufacture. When the angular velocity wr is known, it is possible to calculate the size of the angle Dw which the shaft 1001 moves through during the total period of this time. By placing the trigger point wpt at Dw degrees before wrt, which is the angle in which the function is desired, the correct function can be obtained despite time delays in the system. With the aid of the sensor 1111 it is possible to record at what time tc the function was obtained. This signal from 1111 too can receive a delay. If the time is long, compensation can be carried out by taking from the recorded time this known

delay tct, which is carried out in 1202. The variations in the delay on the signal from 1111 may not be too great, and precision will then be maintained. By virtue of the fact that, for example, 1003 transmits a message when tripping is to take place, 1008 can check that triggering has taken place at the correct time. As the delay from 1003 to 1008 is known, 1008 can compensate for this delay, which is carried out in 1203. The time when it actually took place is obtained from 1202, and the time when it should have occurred is obtained from 1203. By taking the difference between these, the difference tj between these times is obtained. By introducing this time difference into the calculation of Dw, a better wtp can be calculated. By continuously adjusting in this way for unknown time delays, the time difference tj can be kept close to zero. The fact that uncertainty is great upon start-up is often no problem, since a xrachine in most cases goes relatively slowly, for which reason the time is not so critical. In the case where it is critical right from the start, 1008 may carry out a blind firing of 1110 and measure the time delay before the machine starts.

Figure 5 is a system similar to that in Figure 1, but constructed in a somewhat different way. The overriding control member 501 will control the system. This unit is master in the system or at least master over communcia- tion. In some cases this master is not needed, but in these cases the communication method is unlimited since uncontrolled alteration on the communication would create the risk of chaos. The communication channel 509 is necessary for the module 501, 502 and 503 to be able to exchange information. In the case where the trigger information is extremely critical and requires great reliability in the time it takes to transmit the informa- tion, there can be a further communication channel 510. The communication 510 can go to all or parts of the system. Connected to the transmitter member 502 is a sensor, compare with 1002 in Figure 1. In a start-up, the

communication must take place in accordance with estab¬ lished rules which are programmed in each module 501, 502 and 503. If appropriate, further information on the com¬ munication procedure can be indicated by 501, for example which trigger messages are to be transmitted from 502 on 509 and/or on which trigger messages 503 will activate its first member 507. It can happen that all sensor values from 504 give one message or that on 502 there is a selection by means of a combination of hardware and software 505 or as is indicated in the figure directly with hardware by virtue of the fact that there is a direct coupling between the connection for the sensor member 504 and the communication unit which transmits on 509. The receiver member 508 can be constructed in such a way that it receives all the messages which are transmitted on 509 or that it has selection possibilities so that only those messages which have significance for the module 503 activate a function of the latter. When a trigger message arrives at 503, there is after a certain time an activation of the first member 507, which has intrinsically in itself delays dependent on physical properties. These delays in 507 can vary depending on manufacturing variations, ageing and environmental factors, for example pressure and temperature. With the aid of a detector 506, it is possible to measure when 507 has attained the desired function, for example when a drawing magnet has reached the desired position. The invention is based on the fact that it is possible to guarantee the time, with sufficiently great accuracy, from when a certain angle is obtained at 504 to when a signal for activation is transmitted from 503 to 507. With the aid of 506, it is in addition possible to measure the time from when a signal is given to 507 to when the desired function at 507 has been achieved. It has also been assumed that the time from when the signal is transmitted from 503 to when 507 has become activated does not vary more than the interval within the fault tolerance between two consecutive activations. By means of a calculation in 503 it is then possible to calculate

the time when the signal to 507 will be transmitted so that activation of 507 takes place at the correct point in time relative to a certain angle. The activation of 507 will normally take place at a certain position at 504. This means that the time when the signal is transmitted from 503 to 507 also depends on which angular velocity applies at the time. Naturally, a trigger message must be transmitted in such good time before activation of 507 that 503 has time to execute the activation, alternatively that 503 chooses a trigger message which is so far in advance that the task can be executed. This choice of which trigger message is to be chosen can be made in a number of ways. The simplest way is that 502 transmits all angles and 503 chooses the trigger point which it considers to be best for the case in question. Alternatively, 503 can transmit information to 502 and indicate which angle is to be transmitted as trigger angle. Another alternative is that 502 transmits one or more trigger points and that 503 can compensate for the delay between receiving and trigger point until the signal to 507 is transmitted in order to achieve activation at the correct time.

The invention is not limited to the embodiment shown above as an example, but can be subjected to modifications in accordance with the following patent claims and the inventive concept.