CH683641A5 - Converting measurement signal of absolute position measuring pick=up - Comparing measurement signal of each cycle at interface with incremental sum produced in process - Google Patents

Abstract

The method is esp. applicable to a rotary or linear measuring pick-up with a parallel digital signal as measurement signal. The latter (M) is compared with an incremental sum (sigma) produced in the interface. An incremental signal (A/B) for the output is produced from the comparison corresp. to the signal of an incremental pick-up for negative, positive or no change. The incremental sum for the next process cycle is actualised. The interface can be hardware, software or a selected combination. USE/ADVANTAGE - Absolute rotary pick-up with gray code disc and associated photodiode array. Extendable to processing system, e.g. MPS, NC or CNC, replacing physical zero crossing by virtual one.

Description

5

10th

15

20th

25th

30th

35

40

45

50

55

CH 683 641 A5

description

The invention relates to a method for signal conversion, in particular for converting the measurement signal of a position encoder, and to an interface to an absolute encoder, in particular to a position encoder, according to the corresponding, independent patent claims. The output signal of the absolute encoder is converted in the interface according to the procedure.

Encoders, in particular angle encoders or linear encoders, are known as two types: as incremental encoders and as absolute encoders. While the output signal of the incremental encoder only contains the change in the measured value and the direction of this change, the output signal of the absolute encoder also contains the absolute value related to a zero point at all times. The signal from the incremental encoder, which essentially consists of pulses per change unit, is switched to a counter during further processing, for example for control or regulation, in cases where an absolute value is also to be registered, which adds up the individual increments and thus generating information that roughly corresponds to the information of an absolute encoder. This with one important difference: When switching on or after a power failure, such systems must be brought into a defined state through an initiation step, i.e. the counter is set to zero. However, this requires that the random position of the encoder, which it takes up after switching on or after a power failure, is interpreted as zero position without special measures. For this reason, incremental encoders are usually controlled in such a way that they are automatically initiated by zero travel after switching on or after a power interruption. In other words, the corresponding control moves, for example, the machine part monitored by the corresponding incremental encoder into a zero position defined by other means and then sets the counter to zero. Such an initiation represents a physical zero travel. This requires time and additional sensor technology and, above all, it must be possible to carry out the machine in any state, since a power failure is possible in every state of the machine. This condition limits the use of incremental encoders, especially in the area of robotics.

For applications in which the absolute value of the measured variable must be known, but in which a physical zero travel is not desirable or not possible, absolute encoders are used whose output signal contains the absolute value at any time, i.e. even after switching on or after a power failure. The output signals of such absolute encoders differ from the output signals of incremental encoders. They have to be processed in a more complex process, which is also difficult to achieve the very high measurement frequencies required today.

For most PLC (programmable logic controllers), NC (numeric control) and CNC systems, various interface cards and modules are available on the market with which incremental rotary and linear encoders can be connected directly. Such interface modules usually already contain some “intelligence” for positioning tasks. No such interface modules are available for absolute encoders. This means that absolute encoders cannot simply be connected to tried and tested control and regulation systems, so that the use of absolute encoders is always relatively complex. It would obviously be desirable to create absolute encoders that could be wired like incremental encoders, because this would combine the advantages of both types of encoders.

The object of the invention is to demonstrate a method with which the output signal from absolute encoders, in particular from position encoders, is converted such that it can be processed by a further processing circuit designed for incremental encoders without information being lost in the process. It is also the object of the invention to provide an interface for absolute encoders, in particular for position encoders, by means of which it is possible to cover the inputs of further processing systems, for example PLC, NC or CNC systems, which have been designed for incremental encoders, with absolute encoders.

This object is achieved by the method for signal conversion and by the interface to an absolute encoder in accordance with the characterizing parts of the corresponding independent claims.

The central idea of the method according to the invention is based on evaluating the information about the absolute measured value, which is contained in the output signal of the absolute encoder, upon initiation, that is to say after switching on or after a power interruption, for a virtual zero travel. After this virtual zero travel, which is treated by the further processing exactly the same as the physical zero travel of an incremental encoder, an output signal reduced to change pulses is sufficient, which is summed up by a counter of the further processing to determine the current absolute measured value, just like for the further processing of the signals one Incremental encoder. The method according to the invention therefore fulfills two tasks: on the one hand, it generates a virtual zero run before operation, that is, it generates as many incremental pulses from the signal of the absolute encoder as would be generated by a corresponding physical zero run, and thus brings the counter to further processing to an effective, instantaneous measured value of the sensor,

2nd

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

CH 683 641 A5

current status and on the other hand it converts the signals of the absolute encoder into pulses during operation which correspond to the signals of an incremental encoder.

The method according to the invention and the interface according to the invention are described in detail with reference to the following figures. Show:

1 shows a general process diagram;

2 shows a block diagram for an exemplary variant of the method according to the invention for converting a parallel digital signal (Gray code) into an incremental signal, which at the same time corresponds to a diagram of a hardware embodiment;

3 shows a diagram for the conversion of a Gray code into a natural binary code;

4 shows a diagram and a truth table for the second code converter CW.2;

Fig. 5 shows a block diagram for the same method variant as shown in Fig. 2, but for a soft-warm execution.

1 shows a very general method scheme for converting the measurement signal of an absolute encoder to an incremental signal according to the inventive method. The method according to the invention basically consists of three sub-steps, which are repeated in the form of a method cycle. The sub-steps are: a comparison (M = X?) Between a measurement signal M and an increment sum 2 generated by the method, a generation SE of incremental signals (A / B) for the output corresponding to the comparison result and an update of the increment sum likewise corresponding to the comparison result X.

In a process cycle, the measurement signal M generated by the absolute encoder is compared with an increment sum x generated by the process. If the measurement signal M is greater than the increment sum s, a signal is generated for the output that corresponds to a positive increment of an incremental encoder (measurement value change by one unit in the positive direction: incrementation). At the same time, the increment sum E is increased by one increment. If the measurement signal M is smaller than the increment sum x, a signal corresponding to a negative increment is generated and the increment sum E is reduced by one increment (decrementing). If the measurement signal M is equal to the increment sum X, no increment signals are generated and the increment sum remains unchanged. In each next process cycle, the new increment sum E generated in the previous process cycle is compared with the current measurement signal M and the process cycle is carried out in the same way as described.

The measurement signal M can be an analog or a digital signal.

The increment sum E is initialized (init) in such a way that the increment sum E is set to zero whenever the further processing unit is initialized. If the measurement signal M is the same size as the increment sum X, a signal can be generated (M = x), which can be used for further control purposes or displays. If the increment sum becomes zero, a signal (x = 0) can also be generated, which corresponds to the zero or index pulse (N signal) of an incremental encoder.

The function of an absolute encoder with the interface according to the invention is now as follows: After switching on, the increment sum E is set to zero by an initiation signal (init), just like the corresponding counters for further processing. The encoder, which is in any position, delivers an absolute measurement value, which in most cases will not be zero. This leads through the corresponding comparison result to the generation of a signal for an increment, to the corresponding updating of the increment sum E, to a further comparison and so on until the increment sum X corresponds to the measurement signal M. The incremental signals A / B / N are passed to a further processing unit, in which they are interpreted and processed like the incremental signals of an incremental encoder during its physical zero travel. In this way, a virtual zero run is carried out during which nothing needs to move mechanically. The signal M = E can be used, for example, to report the virtual zero travel as complete and to enable operation.

During operation, any change in the measurement signal M in the same process cycles as they are carried out during virtual zero travel leads on the one hand to the generation of corresponding incremental signals A / B / N, to the corresponding update of the increment sum X and thus to a change in the starting position for the next comparison .

So that there are no disadvantages with regard to resolution and speed compared to incremental encoders and directly connected absolute encoders, it is necessary to design a circuit that carries out the method according to the invention in such a way that the measurement frequency made possible by the encoder is not reduced. It is also advantageous to carry out the method so quickly that, if operation is started immediately after switching on, the error resulting from a virtual zero travel that may not yet have ended remains negligible. In other words, a machine does not necessarily have to come to a standstill during the virtual zero run if the zero run is carried out sufficiently quickly, so that no deviations between the effective measured value and the level of the increment which are intolerable for the control of the machine can occur.

There are three options for circuits for carrying out the method:

3rd

5

10th

15

20th

25th

30th

35

40

45

50

55

CH 683 641 A5

- Firstly, for a so-called hardware implementation, commercially available elements can be used for the individual steps, which may make further intermediate steps necessary for converting the signals. Such an implementation results in a rapid process sequence (order of magnitude 1 to 5 ns per cycle), in a cost-effective version, but which will always be associated with a relatively large space requirement, which is not possible for every sensor. Such an embodiment is described in connection with FIG. 2.

- Secondly, the method for a so-called software execution can be carried out in a microprocessor, which leads to inexpensive, space-saving designs, but which will not work fast enough for many applications (order of magnitude 10 ms per cycle), but this is due to a distinction between different large deviations in the comparison of measured value and increment sum and correspondingly generated pulse series could be weighed out. Such an embodiment variant is described in connection with FIG. 5.

- Thirdly, the process can be implemented with an integrated circuit specially created for this, which leads to a very fast process sequence (of the order of 0.1 jis per cycle) and to extremely small embodiments, which are, however, very expensive.

The method according to the invention proves to be particularly advantageous for absolute rotary and linear encoders whose measurement signal is a parallel, digital signal in the form of a binary or Gray code. Such encoders are widely used in connection with positioning tasks in mechanical engineering and in particular in robotics, applications in which the control and regulation systems mentioned at the beginning are also used with advantages.

2 now shows the method according to the invention applied to a sensor with a gray code signal. The diagram shown is both a process diagram and a schematic diagram of a hardware circuit with commercially available components that are used for one process step (one block in the diagram).

The measuring transducer MG, which corresponds to a known rotary or linear encoder, supplies a measured value M.1 in the form of a number of parallel signals as a Gray code. The number of signals in this gray code depends on the measurement resolution of the encoder and is the same for a gray code encoder as for a binary code encoder, for example twelve for a resolution of the measurement range in 4096 steps. In a first code converter CW.1, the Gray code is converted into a binary code, which is fed as a parallel signal M.2 to a comparator (M = £?). The comparator has two outputs, at which signals are present corresponding to M> 2 or M <£. These signals are fed into a controller S. On the one hand, the control S generates an up or. down signal for a following binary counter £ and a synchronization signal T for the output. The binary counter £ is counted up or down by the signals of the control and its current status as a parallel binary code is on the one hand a further code converter CW.2 and on the other hand back in the comparator M = £? guided. In the code converter CW.2, the binary code is converted into an incremental signal A / B / N. The incremental signal is fed to the output of the interface according to the invention via a blocking circuit (L, latch). The blocking circuit is controlled by the synchronization signal T from the control S. This synchronization ensures that no illegal (incorrect) intermediate values are present at the output. Such intermediate values can occur briefly during the counting process in the internal counter £ or through the code conversion in the code converter CW.2

The output of the interface according to the invention corresponds to the output of an incremental encoder. In addition, a measuring sensor with an interface according to the invention requires an input for an initial signal (init) with which the counter £ can be set to zero and the controller S can be brought into a defined state. Such a signal is also necessary for a corresponding further processing unit, for example a control unit, and can be routed from there to the transmitter.

The individual blocks in the diagram in FIG. 2 can all be implemented using commercially available building blocks:

- The encoder MG is, for example, an absolute encoder with a gray code disk and a corresponding photodiode array.

- The code converter CW.1 is, for example, an EPROM that is programmed as a gray-binary converter. The addressing is used as input and the data lines as output. Fig. 3 shows the functional diagram of another code converter for a Gray code (left input) with four bits into a natural binary code (right output) with four bits, which has the form of a simple EXOR gate and can be expanded for any number of bits can. The variant with the EXOR gate has the disadvantage of longer conversion times compared to the variant with the EPROM, especially with higher bit numbers, for example for more than eight bits.

- The comparator M = £? has, for example, three outputs that are activated depending on the two mutually independent code inputs M and £. The comparator has, for example, the following function table. The comparator therefore has three different output states for practically any number of possible input states. The number of possibilities of the input states is only limited by the code width used (number of bits).

4th

5

10th

15

20th

25th

30th

35

40

45

50

55

60

65

CH 683 641 A5

Entrances

Exits

M> s

M = I

M <T

M = s

0

1

0

M> s

1

0

0

Mer

0

0

1

The control S essentially generates only the synchronization signal T for the output latch L.

- The internal counter X is, for example, a binary up / down counter, which acts as a control element in the control loop and, according to the definition, contains the same value as the external counter in the corresponding processing system.

- The second code converter CW.2 converts the binary absolute value of the internal counter z into a (step-by-step) corresponding incremental signal with a zero pulse. As with CW.1, this code conversion can be implemented with one or more EPROMs or with other logic components. A possible circuit variant and a truth table for the code converter CW.2 are shown in FIG. 4.

- The Latch L must have the ability to "freeze" the three output signals A, B, N from CW.2 (latch) or to transmit them unchanged (transparent). The respective state is specified by the controller S.

The electronics required for the interface according to the invention are advantageously accommodated in the encoder itself, so that it forms a unit that can be installed and connected exactly like an incremental encoder. The interface depends on the MG encoder used in relation to the number of outputs of the encoder for its measurement signals and in relation to the code of this measurement signal. If the MG encoder supplies a natural binary code, the first code converter CW.1 is not necessary; it is necessary for Gray code or BCD code. If the measuring transmitter MG supplies an analog signal, for example from a potentiometric measurement, an analog / digital converter with a binary output is used instead of the first code converter CW.1.

5 shows the block diagram for an exemplary software implementation of the method according to the invention using the example of its use with a rotary or linear encoder which delivers a measurement signal in the form of a Gray code.

First you are asked whether an initialization is necessary. If this is the case, an internal value W is set to zero; if no initialization is necessary, the internal value W remains unchanged. The absolute measured value A of the sensor is read in and compared with the internal value W. If the measured value A is larger or smaller than the internal value W, this is incremented or decremented accordingly. If there is no difference between the measured value A and the internal value W, the internal value W remains unchanged. Incremental signals A, B, N are generated and output from the current internal value W. Then you are asked for an initialization again and the cycle is worked through again and again in the same way.

Since software implementation of the method is relatively slow, it is advantageous to differentiate the comparison step, in which the measured value A is compared with the internal value W, in such a way that it not only represents a comparison of larger and smaller, but also that The size of the difference was determined and classified. Series of incremental pulses can then be generated for larger differences without further comparisons, which saves many passes through the process cycle and thus processing time. A code conversion in the sense of the first code converter CW.1 (FIG. 2) is unnecessary since it is readily possible to compare gray codes or other codes by software and to calculate with them.

Furthermore, the software implementation of the method can be accelerated by using an external, hardware counter.

In general, any combination of software and hardware components is conceivable, whereby processing speed is probably always the primary selection criterion.

Claims (1)

Claims 1. A method for converting a measurement signal from an absolute encoder, in particular a position encoder, characterized in that in each cycle of the method the measurement signal (M) of the encoder is compared with an increment sum (x) generated in the method, depending on the sign of the difference between the measurement signal (M) and the increment sum (z) at least one signal is generated which corresponds to the signal of an incremental encoder for an increment in the same direction, and the increment sum (I) is incremented or decremented by a corresponding number of increments, in each further Cycle of the method, the current measurement signal of the transmitter is compared with the increment sum updated in the previous cycle. 5 5 10th 15 20th 25th 30th 35 40 45 50 55 CH 683 641 A5 2. The method according to claim 1, characterized in that the increment sum (X) is set to zero in an initiation step. 3. The method according to any one of claims 1 or 2, characterized in that the measuring signal (M) of the transmitter is a parallel digital signal or an analog signal and that the converted output signal is an incremental signal (A / B). 4. The method according to claim 3, characterized in that a zero pulse (N) is generated as an output signal when the increment sum (x) is equal to zero. 5. The method according to any one of claims 1 to 4, characterized in that a signal (M = X) is generated when the measurement signal (M) is equal to the increment sum (x), and that this signal for the release of the operation after a virtual zero travel is used. 6. The method according to any one of claims 1 to 5, characterized in that the measuring signal of the transmitter is a Gray code. 7. Interface for an absolute encoder, in particular for an absolute position encoder, for performing the method according to one of claims 1 to 6, characterized in that it has inputs for a measurement signal (M) of the absolute encoder, an input for an initiation signal (init), means for Comparing, means for adding up or counting, means for generating incremental signals and two outputs for incremental signals (A / B). 8. Interface according to claim 7, characterized in that it additionally has an output for a zero signal (N). 9. Interface according to one of claims 7 or 8, characterized in that it has a comparator (M = x?), A control unit (S) and a counter (x) and that the measurement signal (M) of the absolute encoder (MG) a first input of the comparator (M = x?), the output of the comparator (M = x?) to the control unit S, the output of the control unit S to the counter (x), the output of the counter (x) to a second input of the comparator (M = 2?) and is directed to an output code converter (CW.2). 10. Interface according to claim 9, characterized in that it has a converter between the absolute encoder (MG) and the comparator (M = x?). 11. Interface according to claim 10, characterized in that the converter is a code converter or an analog / digital converter. 12. Interface according to claim 7, characterized in that it has a microprocessor. 13. Absolute encoder, in particular rotary or linear encoder, characterized in that it has an interface according to one of claims 7 to 12. 14. Absolute encoder according to claim 13, characterized in that it has two outputs for incremental signals (A / B) and one input for an initiation signal (init). 15. Absolute encoder according to claim 14, characterized in that it has a further output for a zero signal (N). 6