JPH03166042A - Device for detecting scale - Google Patents

Abstract

PURPOSE:To use effective data to detect the position of a scale with a high accuracy by providing a data diagnosing circuit which diagnoses data latched by a data latch circuit as effective only when the present speed is roughly a predetermined speed. CONSTITUTION:The timing at which the input and output ends of a mark pass is detected, and the data of moving positions of one of the ends according to the timing of which the closed loop of a moving body is controlled is latched by a data latch circuit 24. The present speed is compared with a predetermined speed by means of measurement of the time at which the input and output ends pass, and thereby the data latched by the data latch circuit is diagnosed as effective by a data diagnosing circuit 27 only when the present speed is roughly the predetermined speed.

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial Application Field) The present invention relates to a scale detection device useful for correcting errors in a drive system, etc. in an NC grinding machine.

(Prior Art) As drive systems in conventional NC processing machines, there are examples of drive systems using a semi-closed loop control system and those using a full closed loop control system.

As is well known, in the former, a servomotor is equipped with a position detector such as a rotary encoder or a resolver, and a moving body is driven via a power transmission mechanism using a hole screw or a rack/binion. In the latter case, the actual moving position of the moving body is detected using a highly accurate position detector such as a so-called optical scale.

However, in a semi-closed loop control system, there is a problem in that the reduction mechanism is affected by thermal distortion of the power transmission mechanism including the ball screw or rack and pinion, and thermal distortion of the frame, resulting in machining errors.

Comparing a ball screw and a rack and pinion, the ball screw has a greater thermal effect, but the rack and pinion has poorer positioning accuracy, so it is hard to say that the rack and pinion is superior.

Furthermore, although thermal distortion of the power transmission mechanism is removed by using the full-closed-lobe control, errors due to thermal distortion of the workpiece or frame cannot be detected.

Further, optical scales are very expensive, and are susceptible to physical damage due to vibration during processing, resulting in problems such as loss of synchronization of scale signals. Furthermore, since a feedback signal is always given to the servo system, there is a limit to the response speed, which limits the motor speed. Furthermore, there are other problems such as the servo system becoming unstable due to the rigidity of the drive system and hunting occurring, which has advantages and disadvantages compared to semi-closed loops.

Therefore, by detecting the movement status of the moving body relative to a fixed position in a manner that can be converted into a constant temperature, and comparing this detected value with the detected value in the closed loop control system,
It is conceivable to control the closed loop control system so that the actual machining position matches the command value at the constant temperature.

In this case, the movement status of the moving body is detected using a different scale separately from the position detection of the closed-loop control system.
This detection is compared with the position detected by the closed loop control system.

(Problem to be Solved by the Invention) However, in the scale detection device considered as above, the specially provided scale is detected by, for example, an optical sensor as the moving body moves. , it is difficult to read the marks on the scale accurately.

For example, a commercially available optical sensor has a detection timing delay of 30 μs or more, and a proximity sensor has a detection timing delay of 500 μ6 or more, so if the moving speed is 50 m/min and the response delay is lms, a deviation of 0.833 mm will occur.

Furthermore, even if an overall offset is applied to the response delay, it is difficult to apply this type of offset if speed variations occur when real-time detection is desired in line with actual machining.

Therefore, the present invention provides a scale detection device that detects the movement status of a moving object using a scale provided separately from a position detector of a closed-loop control system, in which the moving speed of the moving object is also actually measured. , the purpose is to make only data regarding a predetermined speed valid and to perform highly accurate position detection using valid data.

[Structure of the Invention] (Means for Solving the Problems) The present invention achieves the above object by using a scale having a plurality of marks arranged along a moving body of a processing machine controlled in a closed loop. A device for detecting movement of the mark detects the timing of passage of an input end and an output end of the mark, and latches movement position data managed by the closed loop of the moving body at one end thereof. By comparing the current speed with the scheduled speed by measuring the passing time between the data latch circuit and the input end and output end, the data latched by the data latch circuit is valid only when the current speed is approximately the expected speed. The device is characterized by being equipped with a data quality determination circuit.

(Function) The scale detection device of the present invention measures the passage time of the input end and output end of the mark on the scale, and only when this time is constant, that is, at a predetermined speed, the data detected this time is considered valid. , a fixed offset corresponding to the speed is given to this data to obtain detected position data.

(Example) Examples of the present invention will be described below.

FIG. 2 is an explanatory diagram showing an example of a positioning device in which the present invention is applied to a punch press machine. The drive system is shown as an example of a semi-closed loop.

In the figure, the te is movable in the left and right direction (X direction). - the pull 1 is fixed to the table bracket 2;

This table bracket 2 is made movable in the X direction according to the rotation of the ball screw 4 by screwing the ball screw 4 extending in the X direction into a nut 3 embedded above the table bracket 2.

Both ends of the ball screw 4 are rotatably supported by bearings 5. Further, one end of the ball screw 4 is connected to a servo motor 7 via a reduction gear 6.

The servo motor 7 is provided with a tacho generator 8 and an incremental rotary encoder 9.

A work clamping device 10 is provided on the table 1 and is movable in the Y direction perpendicular to the X direction (perpendicular to the paper surface in the figure), and moves the gripped work W in the Y direction on the table 1. It is considered possible. Therefore, by driving the servo motor 7 in the X direction and driving the work clamp device 10 in the Y direction, the work W can be moved freely within the X and Y planes.

An X-axis mask scale 12 is attached to a fixed frame portion below the table 1, which is supported at one point by a pin 11 at the origin position and is extendable and retractable in the X direction.

The mask scale 12 is formed into a band shape using a homogeneous material with a known coefficient of thermal expansion, and has a plurality of optical sensor dog holes Pn (PL, P2, P3,...,PN) are provided. Since the pitch is measured and used as described later, it is not necessarily necessary to manufacture it with high precision. However, it is necessary to manufacture each hole Pn with constant hole dimensions.

Further, an optical proximity sensor 1 is provided on the lower surface of the table 1 to detect the dog hole Pn as the table 1 moves.
3 is provided.

In this example, a crankshaft 13 is provided above the table 1, and a ram (not shown) attached to the crankshaft 13 is driven up and down to press molds provided above and below the workpiece W. It is designed to be punched. A top dead center dog 14 is provided at one position of the crankshaft 13,
By detecting this with the proximity sensor 15, it is possible to identify the top dead center, that is, the non-punch state.

The frame is provided with a temperature sensor TS for detecting the temperature as a representative value of the environmental temperature.

On the other hand, the control device that controls the punch press with the above structure is N.
It is mainly composed of a C device and a programmable controller connected thereto, and a counter circuit 16 is provided in the programmable controller of this control device, and this counter circuit 16 is connected to a transmitting/receiving arithmetic processing device 17.
and a servo system 19 having a servo parameter storage section 18. General servo system 19
has a position loop and a speed loop, and inputs a positioning target value output from the NC device, and controls the movable body, that is, the table 1, at a commanded speed according to this target value.

The counter circuit 16 manually inputs the detection signal of the dog hole Pn of the mask scale 12 at a predetermined timing, latches the position signal detected by the encoder 9 at that time, and transmits this value to the transmission/reception calculation device 17. .

The transmission/reception arithmetic processing device 17 has a buffer therein, inputs detection data of each pitch distance and the dog hole Pn, sets servo parameters in the servo parameter storage section 18, and rewrites the set parameters. It is.

In the device with the above structure, the initial settings, correction principle,
We will explain the process of changing servo parameters during machining, the machine material compatible method, and the dog hole detection method in this order.

At the time of assembly, a laser mirror is attached to a bracket or the like of the table 1, and the difference δ between the actual operation and the NC command value is stored using the laser distance 51.

For example, if the actual movement amount due to the laser projection is 100.05 mm with respect to the command value of 100 mm, the difference 100, 05-100-0.05 is recorded and the value is given to the arithmetic processing unit 17.

Therefore, after making the measured value of the encoder accurate using the difference δ accurately measured by the laser distance meter, the pitch interval is measured, and this is converted to 20°C to obtain the true pitch interval T (n, f4) This is stored in the servo parameter storage section 18.

Specifically, the assumptions here are that the machining accuracy of the gauge is not known on the micron level, that a latch delay occurs due to the response speed of the optical sensor, and that the measurement latch may be delayed due to changes in the axis speed setting. Recognizing that the data are different, the initial value of the scale 12 must be determined.

First, the scale itself can be measured with a measuring device to find it, but since it depends on the mounting position (the condition with the sliding surface and the mounting reference hole position), it is not recommended to use this method to find the reference value of scale 12. do not have.

Therefore, the scale 12 is installed as shown in the drawing.

Also, set the mold and prepare for processing.

Set the shaft speed to the override value F4, and manually input the command value Clmm (so that it is as large as possible within the measurable stroke of the measuring instrument) from the NC console.

Next, C2mrn (around 10 mm near the origin of the workpiece) is manually processed.

At this time, the processed plate was left to stand for a day and night under the control of 20°C and measured with a measuring device, and the measured values at this time were Mlmm and M2, respectively.
Suppose it was mm.

Also, following the past measurement (under the same environment as machining), measurement was performed at shaft speed F4 and the counter clutch coordinates Lnmm (xl, x2
, ..., xN) are stored at all pitch points.

Similarly, the above measurements are repeated for each of the shaft speeds F3, F2, and F1, and the latch coordinates are stored.

From these, the latch coordinate table shown in FIG. 3 is created.

Here, the average delay D (4-3
).

D(4-3) So-called Σ(L (n, f 3) −L (n, f
4) ) /N Similarly, calculate the average delay D (4-2) and D (4-1) at F2 and F1 compared to F4.

D (4-2) - Σ (L (n, f 2) - L (n, f 4)) / ND
(4-1) 1Σ(L (n, f 1) −L (n, f4))/N
The reason for calculating the average value is to keep the amount of data to be managed as small as possible, and the data that is actually managed on the software is T (
n, f4) and D (4-3), D(4-2), D
(4-1). The relationship between the error and the speed f4 is expressed as the fourth
Shown in the figure.

Next, find the dimension T(n, f4) to be found.

T (n, f 3) - T (n, f4) + D (4-
3)T (n, f2)-T (n, f4)+D (4-
2) T (n, f 1) - T (n, f4) + D (4
-1) The values obtained above are stored in the buffer in the transmitting/receiving arithmetic processing unit 17 in advance as 20°C conversion values, and the values of each pitch are determined according to the environmental temperature, and these are used as servo parameters to servo system 19 , and the well-known pitch error correction is applied.

Correction Principle> As described above, each dome hole Pn position of the master scale 12 is converted to 20° C. and the position is managed. In other words, no matter how the environmental temperature changes, the actual machining deviation can be detected by detecting the dog hole Pn position of the mask scale 12, and an appropriate correction value can be given according to the difference. .

To show the basic operation, if a processed plate made of iron is placed in an environment of 25°C, 11 points per meter at 20°C.
．． 7μm/”C・mX (25-20)-58.5
It is elongated by μm. If this plate was made to have high rigidity and processed according to the theoretical value, and then cooled to 20°C, it would have been processed to be 58.5 μm smaller.

Therefore, in order to avoid this, mask scale 12
It is sufficient to correct the positioning of the processing machine at the latch point of the dowel hole Pn.

It should be noted that this correction value is the actual machining position detected on an accurate scale with a temperature correction function, so it takes into account not only the mechanical distortion and temperature distortion of the power transmission mechanism, but also the temperature of the workpiece and frame. This also cancels out the effects of mask scale 12 and its measurement system, and allows processing with the measurement accuracy of the mask scale 12 and its measurement system.

To give a specific example, suppose that the ball screw 4, which has a power transmission function, is slightly higher than the workpiece temperature Tw, and the strain of the ball screw is △1. Assuming that the error due to the expansion of the workpiece W is Δ2 and other distortions of the frame are Δ3, these values Δ1,
This means that it does not matter what Δ2 and Δ3 are.

Therefore, if the servo parameters are appropriately corrected in response to changes in conditions, especially changes in temperature, machining with almost no errors can be maintained permanently.

In addition, when comparing the effects of this semi-closed loop with that of a fully closed loop, it is said that it is rather superior from a practical point of view, as it allows high-speed machining, there is no risk of step-out, and it can be designed at a lower cost. It can be said that semi-closed loops are superior. Errors due to temperature are basically the same.

Furthermore, although the ball screw 4 is shown as an example in this example, it may also be a rack/binion. However, in terms of the inherent accuracy due to pack lash, etc., the ball screw has better processing accuracy.

Figure 5 shows the servo parameter setting method.

At the start of machining, when the origin is returned in step 501, the start to NC is first prohibited in step 502.

Next, in step 503, a reset signal is sent to the current value counter, and in step 504, a signal to turn on the correction function is sent.

Next, in step 505, data from the temperature sensor TS is transmitted, in step 506 the frame extension at that temperature is calculated, and in step 507, the NC is permitted to start.

In step 506, servo parameter values corresponding to the temperature θ are set based on the 20° C. equivalent value T(n−f4).

In this example, it is assumed that the correction work is performed by one servo system, but it is also possible to perform correction by changing the original target value.

Correction Parameter Changing Process> The servo parameter is changed by the process shown in FIGS. 6 and 7. FIG. 6 shows the change request procedure, and FIG. 7 shows the change procedure when a change request is made.

In FIG. 6, in step 601 during processing, the proximity sensor 15 shown in FIG. Determine whether or not.

If it is the predetermined direction, proceed to step 603,
The latch coordinates of the dog hole Pn are sent, and in step 604,
Depending on whether the latch coordinates are for the dog hole Pn, the latch value for that dog hole Pn is loaded into the memory.

Therefore, in step 605, the value loaded in step 604 is compared with the reference value, that is, the value that should be detected at the previously detected temperature based on the value converted to 20°C, and if it is not within the allowable value, For example, the process moves to step 606 and the parameter change request nobit is turned on. Note that in step 607, the frame temperature is received.

Such a parameter change request is issued mainly when the temperature of the ball screw 4 changes due to a change in temperature, that is, a change in the ring filling temperature, or a change in load.

Next, in FIG. 7, when a parameter change request is turned on in step 701, the controller waits for the return to the origin position in step 702, and outputs a start prohibition to the NC in step 703.

Next, in step 704, if the material t4 of the workpiece W is iron as before, the change value of the servo parameter is calculated in step 706 and set in step 707, and the change value in step 706 is calculated.
8 to allow NC to start. The calculation in step 706 is to calculate servo parameters according to the current situation using the actually measured latch data. Step 705 will be described later. The measured latch data is corrected according to the speed at that time.

Material compatible method> There are two types of material handling methods:

■ One is the expansion rate of each material, for example iron...11.
7 μm/m·°C Copper: 16.7 μm/m·°C Aluminum: 23 μm/m·°C This is a method of appropriately correcting the material of the mask scale 12 currently used.

In this case, for example, when the NC is returned to the origin, the material change is identified in step 704 of FIG.
05, calculate the temperature at 20°C for each material, and store the calculated value.

■ Also, the other one is Figure 8(a) and Figure 8(b)
In this method, a cartridge-type mask scale 20 is created for each material as shown in FIG.

A bolt 22 attached to the guide 21 is for preventing displacement in a direction perpendicular to the direction of movement. The tip of this bolt 22 is formed into a spherical shape, and the cartridge scale 20 is slidably moved in the direction of movement by the guide 21.
Press lightly against the The support point of the pin 11 is also spherical in order to align the fixed point with the origin.

If the fixed point is shifted from the origin, it is necessary to perform a predetermined shift correction. Note that in the case of a punch press, processing is performed at the punch center, so all standards may be set at the punch center.

Dog Hole Detection Method (Part 1)> This item makes it possible to identify whether the detection data is good or bad when detecting the dog hole Pn.

FIG. 9 shows a specific example of the detection circuit.

In the figure, the detection circuit of the dog hole Pn is a buffer (differential TTL
) 23 and a counter clutch control unit 24 that manually outputs the detection signal of the optical sensor 13, and both circuits 23 and 24 are
The counter circuit 16 is manually operated. The counter circuit 16 is supplied with an internal clock signal CLK.

A sensor-on buffer 25 and a clock counter 26 are connected to the counter circuit 16, and the outputs of both channels 25 and 26 are connected to a data quality control section 27.
7 is connected to a data path 28. The clock counter 26 is also supplied with the internal clock signal CLK.

In the above configuration, as shown in FIG. 10, assume that the sensor 13 moves in one direction and obtains a signal 29 for one dog hole Pn. This signal 29 is a pulsed signal that turns on at a certain threshold and then turns off. Assume that the commanded shaft speed is bomm/S. Dog hole P
The width of n is known and is assumed to be a mm.

The signal processing method is shown in Figure 1.

In step 101, latch data when the sensor is on is temporarily stored in the buffer 25, and in step 102, the clock counter 26 calculates the time t from when the sensor is on to when the sensor is off.
get.

Therefore, in step 103, the data quality control unit 2
7, b. ta is determined, that is, the current speed value h is the scheduled speed b. If it is within the allowable value, step 104 validates the latch data when the sensor is on, and sends it to the data bus 28. On the other hand, if the speed b is outside the allowable value, the latch data is invalidated in step 105.

Therefore, in the wood example, the reliability of the latch data detected by the sensor 13 is improved, and the processing accuracy itself can be improved.

In the tree example, the detection of the dog hole Pn is shown, but the mark on the mask scale 12 is not limited to this, and may be printed with a different color, for example. In this case, the quality of the latch data can also be checked in the same way. It can be determined.

Dog Hole Detection Method (Part 2)> This section shows an example of correcting latch data according to speed.

In FIG. 11, in the detection circuit of this example, a velocity calculation section 30 based on waveform detection and a latch coordinate extraction section 31 are connected to the previous generation counter circuit 16. Of which, speed calculation section 3
0 is connected to the average delay parameter value calculation unit 32,
The calculation section 32 and the latch coordinate extraction section 31 are connected to a measurement result output section 33.

The average delay parameter value calculation unit 32 has a correspondence table of errors ε caused by speed and signal delay as shown in FIG. Calculate. Since the errors vary, the average value is used.

The measurement result output unit 33 applies the error ε calculated by the calculation unit 32 to the latch coordinates extracted by the latch coordinate extraction unit 31, and outputs a measurement value close to the true value. The measured value may be used by averaging a plurality of data for the same dog hole.

As shown in FIG. 13, in step 1301 the coordinate value when the sensor is on is taken out, in step 1302 the velocity value is calculated from the pulse width when the sensor is on, in step 1303 the error ε is calculated, and in step 1304 the corrected measurement Output the value.

Therefore, by calculating a velocity equivalent value from a waveform such as a scale, and subtracting the corresponding delay parameter from the measurement data, measurement accuracy and eventually processing accuracy can be improved.

In this example, measurements can be made at any shaft speed. Furthermore, if the speed is constant, accuracy can be further improved. Note that in this example, the speed value was obtained by waveform detection, but
An NC command value can also be used on the condition that the speed is stable.

As described above in detail, according to this embodiment, by properly detecting the dog holes of the mask scale 12,
The table 1 controlled by the semi-closed loop can be moved with precision determined by the mask scale 12.

Furthermore, at this time, the mask scale 12 is configured to be able to compensate for temperature, and is also configured to detect the actual machining position, so that not only distortion of the power transmission mechanism but also thermal effects of the workpiece and frame can be removed. 0.01, which was previously difficult. High accuracy of mm or more can be easily achieved under all environmental conditions, making ultra-precision processing possible.

Unlike a fully closed loop, there is no need to worry about loss of synchronization or damage.

In the above embodiment, a punch press was used as an example, but other NC machines such as a laser processing machine, punch/laser combined processing machine, lathe, etc.
The same applies to machine tools.

Further, in the above embodiment, an example in which a workpiece is moved is shown, but the same applies to a machine tool in which a tool is moved.

The present invention is not limited to the above-mentioned embodiments, but can be implemented in other suitable embodiments by making appropriate design changes.

[Effects of the Invention] As described above, the present invention provides a scale detection device that detects the movement status of a moving body using a scale provided separately from a position detector of a closed-loop control system. The moving speed of the moving object is also measured, and only the data regarding the predetermined speed is made valid, so that highly accurate position detection can be performed using the valid data.

[Brief explanation of the drawing]

FIG. 1 is a flowchart showing a method for determining the quality of data for a scale detection device according to an embodiment of the present invention, FIG. 2 is an explanatory diagram showing the configuration of a positioning device implementing the present invention, and FIG. An explanatory diagram of latch data for each speed,
Figure 4 is an explanatory diagram showing the error situation due to speed, Figure 5 is a flowchart showing the servo parameter setting method, and Figure 6 is a flowchart showing the servo parameter setting method.
The figure is a flowchart showing the output method of a parameter change request, FIG. 7 is a flowchart of the parameter change method, FIG. 8(a) is a front view showing a cartridge-type mask scale created for each material, and FIG. 8(b) is a right side view thereof, FIG. 9 is a block diagram showing an example of a latch data detection circuit, FIG. 10 is an explanatory diagram showing its detection action, and FIG. 11 is a block diagram showing another example of a latch data detection circuit. figure,
FIG. 12 is an explanatory diagram of data used by the circuit, and FIG. 13 is a flowchart showing a method for correcting measured values of latch data. 1...Table 3...Nut 4...Ball screw 7...Servo motor 9...Encoder 11...Pin 12...Mask scale 13...Optical sensor 16...Counter Circuit 17...Transmission/reception processing unit 18...One servo parameter storage section...Servo system 26...Clock counter 27...Data quality control section

Claims (1)

[Claims] In an apparatus for detecting a scale having a plurality of marks arranged along a moving body of a processing machine controlled in a closed loop as the moving body moves, detecting the passage timing of an input end and an output end of the mark, a data latch circuit that latches movement position data managed in the closed loop of the moving object at one end thereof; and a data latch circuit that latches movement position data managed in the closed loop of the moving object at that time, and a data latch circuit that changes the current speed to the planned speed by measuring the passing time of the input end and the output end. A scale detection device characterized by comprising a data quality determination circuit that makes valid the data latched by the data latch circuit only when the current speed is approximately the expected speed by comparison.