JPH0844486A - Coordinate input device and method for correcting coordinate error - Google Patents

Abstract

PURPOSE:To provide a coordinate input device capable of executing accurate coordinate input. CONSTITUTION:Vibration inputted from a vibration pen 3 to a vibration transmitting plate 8 is detected by a vibration sensor 6 and a transmission distance from a vibration source and coordinates based upon the distance are calculated by a signal waveform detecting circuit 9 and an arithmetic and control circuit 1. The detected coordinates are checked by the detected transmission distance to judge whether an error is included or not. When an error is included, the length of one wavelength is added to the transmission distance of a sensor having the highest probability of including an error and most close to the vibration source out of plural calculated transmission distances, its coordinates are calculated again and the judgement of an error is repeated. A sensor most far from the vibration source can also be selected as a sensor corresponding to a transmission distance including an error. Thus error detection controlling preparatory correction can quickly be executed by simple constitution and accurate coordinate input can be attained.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a coordinate input device, and more particularly, to detecting an input elastic wave vibration at a plurality of points on a vibration transmitting plate, and then based on the transmission time from the input position to the detection The present invention relates to a coordinate input device that detects the coordinates of an input point.

[0002]

2. Description of the Related Art A coordinate input device using ultrasonic waves is a method of calculating a position coordinate by detecting a delay time of a wave propagating on a tablet which is an input surface. Since no work is done, it is possible to provide an inexpensive device. Moreover, if a transparent plate glass is used for the tablet, it is possible to construct a coordinate input device having higher transparency than other methods.

In such a coordinate input device, a vibration pulse having a predetermined frequency is input to the tablet, the vibration is detected at a plurality of points on the tablet, and the delay time from the input of the vibration to the detection is measured, According to the delay time and the propagation velocity of vibration, the distance from the position where the vibration is input to each vibration detection point is obtained, and the coordinates are calculated from the distance.

Here, the vibration propagating through the tablet has a waveform as shown in FIG. 9, and the propagation velocity of the vibration also has a group velocity vg in which the envelope propagates and a phase velocity vp in which the phase advances. ing. Therefore, the delay time of the detected vibration includes the group delay time tg detected for the propagation delay of the envelope signal and the phase delay time tp for the propagation delay of the phase signal. When vibration is input to the tablet from the vibrating pen, the vibration sensor at a predetermined position on the tablet detects the vibration propagating through the tablet. The distance d between the vibrating pen and the vibration sensor is the velocity vg of the detected envelope signal, It is expressed by the following formula by the delay time tg.

D = Vg · tg (P1) Furthermore, in order to determine the coordinate position with higher resolution, processing based on the detection of the phase signal is performed. If the delay time tp is the time from the specific detection point of the phase signal, for example, the application of vibration by the vibrating pen to the zero cross point after the signal level detected by the vibration sensor exceeds a certain level, the vibration sensor And the distance d of the vibrating pen is: d = n · λp + Vp · tp (P2) Here, λp is the wavelength of the elastic wave, and n is an integer.

From the equations (P1) and (P2), the integer n is expressed as n = int [(Vg · tg-Vp · tp) / λp + 1 / N] (P3).

Here, N is an appropriate real value other than "0". For example, if N = 2.0, even if the detected values of tg and tp having an error within ± 1/2 wavelength are obtained, the value of the equation (P2) n can be correctly determined. By substituting the value n obtained as described above into the expression (P2), the distance d between the vibrating pen and the vibration sensor can be accurately measured.

If at least three vibration sensors are arranged on the tablet so that they do not overlap on the same straight line,
Three positions of the vibration pen and the vibration sensor are obtained, and the coordinates can be calculated from the Pythagorean theorem.

[0009]

However, in the above-mentioned conventional coordinate input device, the pointing point is often erroneously detected due to noise or a physical cause, and such erroneous detection point is the input character or figure. The technique for removing the false detection points is very important because it deforms the writing shape of.

As a conventional technique for determining erroneous detection, erroneous detection is set by deleting the vibration transmission distance from each calculated coordinate value to each vibration sensor and inspecting the obtained information to delete erroneous input point coordinates. There is a configuration.

Since the ultrasonic type coordinate input device is a system in which vibration (ultrasonic wave) is interposed as a detection principle, electromagnetic waves are not interposed unlike other highly accurate coordinate input systems. Therefore, the electromagnetic wave noise from the outside has a characteristic that it is not easily affected by the principle, and at present, due to technological progress, only the false detection due to the physical cause remains as a problem.

The causes of physical erroneous detection in the conventional ultrasonic system are as follows. 1) On the entire input surface of the vibration transmission plate 8, the vibration phase velocity (V
Coordinate calculation error (propagation speed varies depending on frequency) that occurs because p) is non-uniform. The non-uniformity of the thickness of the vibration transmission plate 8 and the dispersion of the velocity of vibration (plate wave) (the propagation velocity differs depending on the frequency) are the causes of the phase velocity change. 2) An error of several millimeters or more due to a distance error that occurs when n is erroneously calculated from the expression (P3). Mainly when the detected group delay time collapses from the linear relationship between distance and transmission time, the causes are superposition of reflected waves from the end face of the vibration transmission plate, deformation according to the transmission distance of the detected waveform, and pen There is waveform deformation due to angle (details will be described later).

In addition to these, the vibration to be detected (Lamb
Wave vibration (Lamb wave S)
o mode) is erroneously detected, or when the detection level exceeds a predetermined range, an error may occur due to the deviation of the detection point by the wavelength from the specific point of the predetermined phase waveform signal. This can be easily avoided by adopting a circuit configuration that is not easily affected by the characteristics of the vibrating pen and the detection level.

The erroneous detection of 1) has a maximum error amount of about sub-millimeter. Therefore, it does not pose a problem if it is used in consideration of such an error as the performance specification of the coordinate input device. However, the error detection of 2) occurs as a discontinuous point when the coordinate input device is actually used because the error amount occurs in units of wavelength order, and the coordinate value is deleted (the coordinate value is not output). Correction of coordinate values is necessary. Therefore, in order to correct the n calculated in error, the vibration transmission distance to each vibration sensor is calculated backward from the calculated coordinate value, and the obtained information is inspected. A method of changing the value one by one (+1 or-1) has been proposed. The coordinate position obtained from the calculated distance obtained by the correction is re-examined each time, and the correction processing of the value of n is performed until it is determined that the coordinate is correctly detected. However, after the correction processing is performed a predetermined number of times, the processing is interrupted even if the coordinate is erroneously detected.

However, in the conventional n-value correction method, a huge amount of processing procedure is required for the correction, the coordinate input sampling rate is lowered, or an expensive arithmetic control circuit capable of high-speed processing is required. is there. This is because the vibration transmission delay time (tg and tp) obtained from multiple vibration sensors
For example, it is not possible to determine which of the
When the vibration sensor 6d erroneously detects one of the two vibration sensors 6a to 6d, the n values are sequentially corrected one by one from the vibration sensor 6a, and the correction is performed four times before the vibration sensor is corrected and the coordinate is correctly detected. N-value correction / detection procedure is required. Further, the n value does not always shift to the plus side (erroneously detected), and if the ± n value correction is taken into consideration, the n value correction / inspection procedure which is doubled 8 times is required. Further, since it is not uncommon for a plurality of vibration sensors to be erroneously detected, it is clear that a huge amount of procedure is required if the n value correction / inspection procedure for a combination of a plurality of vibration sensors is assumed.

The cause of erroneous detection of n value (hereinafter referred to as n jump) will be described below with reference to the drawings. First, it is the cause of waveform deformation associated with the transmission distance of the detection signal of the vibration sensor. FIG.
Frequency characteristics obtained by analyzing the detection signal of the vibration sensor with a spectrum analyzer are shown in (a). This is when the repetitive pulse frequency of the drive signal of the vibrating pen is set to 500 kHz, and since one of the natural vibration frequencies of the vibrating pen is 500 kHz, the frequency peak is 500 kHz as shown in the figure. However, it is difficult to say that vibrations of a single frequency are propagated because a plurality of peaks also exist near 500 kHz.

As described above, the vibration used in the ultrasonic coordinate input device is a plate wave, and the one detected by the vibration sensor in the configuration of FIG. 2 is the Lamb wave in the plate wave. A
This is o-mode vibration. The propagation speed of the plate wave differs depending on the thickness of the propagating vibration transmitting plate and the frequency of the vibration itself. FIG. 8B shows frequency characteristics of the propagation velocity of the plate wave.
Assuming that the glass plate thickness is constant at 1 mm in the figure, when the vibration frequency changes from 600 to 700 kHz, the phase propagation velocity Vp is about 120 m / s and the group propagation velocity Vp is 90 m / s.
It can be seen that it changes about s. When vibrations of a plurality of frequency components as shown in FIG. 8A propagate at the speed as shown in FIG. 8B, waveform deformation occurs.

FIG. 9 shows a signal waveform detected by the vibration sensor when the distance L between the vibration pen and the vibration sensor is changed. The signal waveforms are clearly different between near and far. As described with reference to FIG. 8B, since the propagation velocity differs depending on the frequency, as the vibration propagates, the high frequency component having a high velocity spreads to the front of the waveform and the slow component spreads to the rear. That is, the detected waveform spreads on the time axis at a long distance. If the peak position is actually detected as the group delay time tg with the specific point, the peak position moves in proportion to the distance according to the group velocity of the vibration of the center frequency component of 500 kHz, so that a linear tg is obtained.

However, when the peak position is to be detected, the vibration sensor is provided in the peripheral portion so that the vibration component traveling faster than the peak is reflected and interferes at the tablet end. When trying to miniaturize the device,
It is desirable to detect tg at a point in front of the waveform that is less susceptible to reflection. Therefore, the conventional method is to detect the inflection point of the envelope. If the time difference between the peak position and the inflection point position increases from Δt1 to Δt2 in proportion to the transmission distance in FIG. 9, the peak position should move in proportion to the distance and the inflection point should also move in proportion. Is. But,
This is not the case with the actual detection signal waveform of the vibration sensor.

FIG. 10 shows a plot of vibration transmission distance (vibration pen = distance between vibration sensors) and detected values of tg and tp. The relationship between the phase delay time tp and the vibration transmission distance L is a stepwise discontinuous parallel group of straight lines A1 to A.
4, the relationship between the group delay time tg and the vibration transmission distance L is a curved curve B as shown in the figure. The asymptote C near the group delay time tg is a straight line (ideal tg) indicating the peak position. As described above, tg is detected from the inflection point, and the method of waveform deformation is not linear (Δt in FIG. 9 is not proportional to the transmission distance), so the group delay time tg does not become a straight line. It is considered that this is because the frequency distribution of the propagating vibration is not symmetrical as shown in FIG. 8 and the vibrations of the plurality of frequency components are not simultaneously transmitted from the vibrating pen to the vibration transmitting plate. The reason for the latter is that when a vibration sensor is placed directly below the vibrating pen and the obtained waveform on the time axis is observed, a low frequency component (around 400 kHz) in front of the waveform,
It can be easily confirmed because a high component (500 kHz or more) is distributed behind the waveform. It is considered that this is because vibration is emitted from the pen tip as a result of multiple reflection inside the vibrating pen 3.

Using tg and tp detected as shown in FIG. 9, the n value is obtained from the above equation (P3) and converted into an integer (in
The difference ΔN from that before processing (t processing) is calculated by the following equation and is calculated as shown in FIG.
It is shown in FIG.

ΔN = (Vg · tg-Vp · tp) / λp-int [Vg · tg-Vp · tp) / λp + 1/2] (P4) In FIG. 11, the distance between the vibrating pen and the vibration sensor is 125 mm
By actually measuring and correcting the point O in
Is zero.

As is apparent from FIG. 11, the group delay time tg
The influence of the non-linearity appears as a result of ΔN swinging to the negative side at a short distance and a long distance. By the way, the above-mentioned erroneous detection of n jumps is a symptom that ΔN exists as a discontinuous point (or straight line) in FIG.
No jumps have occurred. However, for example, if the waveform deformation becomes extremely large and the curvature of FIG. 11 is extremely bent, it exceeds −0.5 at a short distance and a long distance, and the discontinuous difference is returned from the plus side (+0.5). Will be plotted as a curve.

However, it is unlikely that n jumps will occur due to the non-linearity of the group delay time tg as a sole factor. In addition to this, factors such as the influence of waveform deformation due to the superposition of reflected waves from the end surface of the vibration transmission plate and the waveform deformation due to tilting the pen (changing the pen angle) are added, resulting in n jumps. .

The mechanism of occurrence of n jumps is shown in FIG.
Will be schematically described again. It is assumed that the relationship shown by the ideal asymptote C in FIG. 10 is obtained. When the pen indicates L1, the tp on the straight line A3 is also obtained.
The value of n is 2 because it is the second one counted from the straight line A1.
If the writing pressure of the pen becomes extremely strong and the detection signal level becomes extremely large, the value of tp will be the value obtained by extending the straight line of the straight line A4, and will be a smaller value. However, since the value of n still becomes 3, L calculated from tg and tp
Can be correctly obtained from the equation (P2). That is, the value of tg is uniquely determined according to the distance, but the value of tp is an extension of a step-like parallel line determined by the phase velocity Vp and the wavelength λp (wavelength = interval between zero cross points in the phase detection waveform). It may be on a line and may have multiple values, but if the value of n is calculated correctly, the calculated distance can be uniquely determined.
In other words, even if one phase delay time tp is obtained, L1 obtained as an intersection with the straight line A3 and the straight line A1.
It is possible that there are two or more transmission distances such as L0 obtained as an intersection with the extension line of However, the distance L1 can be calculated as a true value based on the value of the group delay time tg.

However, in reality, if the relationship between the group delay time and the transmission distance is non-linear as shown by the solid line, the group delay time tg.
Therefore, tp corresponding to any of the straight lines A1 to A4
When it is finally decided, it will be wrongly calculated as a different staircase. A large deviation from an ideal state (a state in which the group velocity Vg is constant and a delay time tg linear to the distance is obtained) is n jumps. Even if the value of ΔN is inspected, the value of ΔN is always +0.5 to − when inputting coordinates.
Since it is in the range of 0.5 (from the formula (P4)), it is difficult to determine the occurrence of n jumps.

Since the possibility of n jumps increases as the value thereof deviates from zero, it can be said that ΔN is exponential indicating the possibility of n jumps. However, the amount of ΔN generated by tg nonlinearity is as shown in FIG. It occurs on the minus side at short and long distances. This is because the measurement is performed at the origin O in FIG. 7 and the correction is performed, so that Δ near the transmission distance intermediate distance (hundreds of tens of mm)
N is zero.

By the way, the other factors causing ΔN are as follows:
First, due to the interference of the reflected wave from the end surface of the vibration transmitting plate, ΔN swings in plus and minus. This is because a plurality of waves having different phase and peak positions (substantially the same frequency) are superimposed, and the interference distance changes depending on the designated coordinate position. Therefore, tg is the time axis according to the relative phase difference (interference distance). This is because the results will be mixed up and down.

Secondly, ΔN generated by the tilt of the pen also has a value that swings plus or minus. That is, when the vibrating pen tilts toward the vibration sensor, ΔN swings negatively, and when it tilts backward, ΔN swings positively. That is, when the pen is tilted, the vibration detection waveform (envelope) differs depending on the vibration transmission direction. In one direction, it becomes a short waveform on the time axis, and in another direction it becomes a waveform extended on the time axis. There is.

Third, even when the pen is vertically supported, the second symptom remains, though slightly. Although it is called pen directivity, it is impossible to construct a vibrating pen with a material and a shape that are perfectly symmetrical with respect to an axis. Therefore, the waveform varies slightly depending on the direction of vibration transmission. Depending on which direction is used as a reference, ΔN swings either positively or negatively.

Fourth, there is a problem in the performance of the circuit that detects the signal waveform. An ideal circuit stably detects a specific point = inflection point on the envelope of the detection signal at any detection signal level, but in an actual circuit, the group delay time tg slightly varies depending on the level. ΔN due to the fourth factor is 0.05 or less at the maximum, which is not a big problem.

The factors that cause ΔN are summarized as follows.
It is known that it is uncertain whether it takes a positive or negative value, and when these other factors are combined and the worst value is added with the same sign, it is known that ΔN becomes a value close to 0.5.

As described above, in the conventional ultrasonic type coordinate input device, there are cases where the coordinate misdetection of several millimeters of n jumps occurs, and it is easy to detect the coordinate value including an error, but There is a problem that a complicated and time-consuming process is required to correct the value correctly.

[0034]

[Means for Solving the Problems] and

In order to solve the above-mentioned problems, a coordinate input device of the present invention defines a plurality of vibration detection positions of a vibration transmitting body for propagating vibration, and a coordinate indicator for generating vibration and each of the vibration detection positions. Means for calculating the coordinate value based on the distance while measuring the distance from the arrival time of the vibration, calculating and outputting the point designated by the coordinate indicator on the coordinate input surface on the vibration transmitting body as the coordinate value. And, the presence or absence of a wavelength error is determined from the calculated coordinate value and the measured distances, and the vibration detection position closest to the coordinate indicator is selected from the arrival time of the vibration. By correcting the distance from the generated vibration detection position to the coordinate indicating tool in wavelength units, it is possible to determine erroneous detection of coordinates with a simple configuration and correct the error.

Also, a first detecting means for detecting a predetermined point of the envelope signal of the detection signal detected at the vibration detecting position and detecting a group delay time based on the group velocity of vibration with reference to the predetermined point. And a second detecting means for detecting a phase delay time based on the phase velocity of vibration with reference to the zero-cross point of the detection signal, and transmission of vibration based on the group velocity detected by the first detecting means. A coordinate input device for deriving a position coordinate of the vibration input means based on time and a transmission time of vibration based on a phase velocity detected by the second detection means, wherein the wavelength error is a wave number of vibration. Is a quantization error in quantizing, and the distance is corrected according to the value of the quantization error.

Further, the correction of the wavelength error is performed by adding the distance for one wavelength.

The position coordinates are calculated by using the group velocity obtained from the movement corresponding to the distance of the envelope predetermined point when the vibration transmission distance is a long distance. The first detection means sets the detection reference position to a distant position so that the quantization error becomes small when the vibration transmission distance is long.

[0038]

FIG. 2 is a block diagram showing the schematic arrangement of an ultrasonic coordinate input device according to this embodiment. In the figure, reference numeral 1 denotes an arithmetic control circuit for controlling the entire apparatus and calculating a coordinate position. Reference numeral 2 denotes a vibrator drive circuit for vibrating the pen tip inside the vibrating pen 3. Reference numeral 8 is a vibration transmission plate made of a transparent member such as acrylic or glass plate. Coordinates are input by the vibration pen 3 by touching the vibration transmission plate 8. That is, the position coordinates of the vibrating pen 3 can be calculated by designating with the vibrating pen 3 an area (hereinafter referred to as an effective area) indicated by a solid line A in the figure.

In order to prevent (reduce) the propagating wave from being reflected at the end surface of the vibration transmitting plate 8 and returning to the central portion, the vibration isolator 7 is provided on the outer periphery of the vibration transmitting plate 8. As shown in FIG. 2, a vibration sensor 6a for converting mechanical vibration into an electric signal, such as a piezoelectric element, is provided near the inside of the vibration isolator as shown in FIG.
6d is fixed. Reference numeral 9 denotes a signal waveform detection circuit that outputs a signal in which vibration is detected by each of the vibration sensors 6a to 6d to the arithmetic control circuit 1. Reference numeral 11 is a display such as a liquid crystal display capable of displaying in dot units, and is arranged behind the vibration transmission plate. Then, by driving the display drive circuit 10, it is possible to display a dot at a position traced by the vibrating pen 3 and see it through the vibration transmission plate 8 (made of a transparent member).

The vibrator 4 built in the vibrating pen 3 is driven by the vibrator driving circuit 2. The drive signal for the vibrator 4 is supplied as a low-level pulse signal from the arithmetic control circuit 1, amplified by a predetermined gain by the vibrator drive circuit 2, and then applied to the vibrator 4. The electric drive signal is converted into mechanical vibration by the vibrator 4, and is transmitted to the vibration transmission plate 8 via the pen tip 5.

Here, the vibration frequency of the vibrator 4 is selected to a value capable of generating a plate wave on the vibration transmission plate 8 such as glass. When the vibrator is driven, a mode in which the vibration transmitting plate 8 vibrates in the vertical direction of FIG. 3 is selected. Also,
By setting the vibration frequency of the vibrator 4 to the resonance frequency including the pen tip 5, efficient vibration conversion can be performed. The elastic wave transmitted to the vibration transmission plate 8 as described above is a plate wave,
It has the advantage that it is less susceptible to scratches, obstacles, etc. on the surface of the vibration transmission plate as compared to surface waves.

<Description of Arithmetic and Control Circuit> In the above-mentioned configuration, the arithmetic and control circuit 1 has a predetermined cycle (for example, 5 msec).
Every time), a signal for driving the vibrator driving circuit 2 and the vibrator 4 in the vibrating pen 3 is output, and the internal timer (which is composed of a counter) starts timing. Then, the vibration generated by the vibrating pen 3 propagates on the vibration transmission plate 8 and arrives with a delay according to the distance to the vibration sensors 6a to 6d.

The signal waveform detecting circuit 9 includes the vibration sensors 6a to 6a.
The signal from 6d is detected, and a signal indicating the vibration arrival timing to each vibration sensor is generated by the waveform detection processing described later. This signal for each sensor is input to the arithmetic control circuit 1, and each vibration sensor is input. Detection of the vibration transmission time from 6a to 6d, and driving the display drive circuit 10 based on the position information of the vibration pen 3 to control the display on the display 11, or output the coordinates to an external device by serial or parallel communication. (Not shown).

FIG. 4 is a block diagram showing a schematic configuration of the arithmetic control circuit 1. Each component and its operation outline will be described below.

In the figure, 31 is a microcomputer which controls the arithmetic control circuit 1 and the coordinate input device as a whole, and is a nonvolatile memory which stores an internal counter, a ROM storing a processing procedure, a RAM used for calculation and the like, a constant and the like. It is composed of memory. Reference numeral 33 is a timer (not shown) that counts a reference clock (for example, is composed of a counter), and inputs a start signal to the vibrator driving circuit 2 to start driving the vibrator 4 in the vibration pen 3. Then, the timing starts. As a result, the start of timing and the vibration detection by the sensor are synchronized, and the sensor (6a
It is possible to measure the delay time until the vibration is detected by ~ 6d).

The circuits that are the other constituent elements will be described in order.

The vibration arrival timing signals from the vibration sensors 6a to 6d output from the signal waveform detection circuit 9 are latched via the detection signal input port 35 to the latch circuits 34a to 34.
It is input to d. Each of the latch circuits 34a to 34d corresponds to each of the vibration sensors 6a to 6d, and when receiving a timing signal from the corresponding sensor, it latches the measured value of the timer 33 at that time. When the determination circuit 36 determines that all the detection signals have been received,
A signal to that effect is output to the microcomputer 31.
When the microcomputer 31 receives the signal from the determination circuit 36, the vibration transmission time from the latch circuits 34a to 34d to each vibration sensor is read from the latch circuit, a predetermined calculation is performed, and the vibration on the vibration transmission plate 8 is vibrated. Pen 3
Calculate the coordinate position of. Then, by outputting the calculated coordinate position information to the display drive circuit 10 via the I / O port 37, it is possible to display a dot or the like at a corresponding position on the display 11, for example. Alternatively, the coordinate value can be output to an external device by outputting the coordinate position information to the interface circuit via the I / O port 37.

<Description of Vibration Transmission Time Detection (FIGS. 5 and 6)
> Next, the principle of measuring the vibration transmission time from the vibration pen 3 to the vibration sensors 6a to 6d will be described.

FIG. 5 is a diagram for explaining a detection waveform input to the signal waveform detection circuit 9 and a vibration transmission time measuring process based on the detection waveform. In addition, although the case of the vibration sensor 6a will be described below, other vibration sensors 6b, 6c,
The same is true for 6d.

It has already been described that the measurement of the vibration transmission time to the vibration sensor 6a starts simultaneously with the output of the start signal to the vibrator drive circuit 2. At this time, the drive signal 41 is applied from the vibrator drive circuit 2 to the vibrator 4. The ultrasonic vibration transmitted from the vibration pen 3 to the vibration transmission plate 8 by the signal 41 progresses for a time tg corresponding to the distance to the vibration sensor 6a, and then is detected by the vibration sensor 6a. The signal indicated by 42 in the figure shows the signal waveform detected by the vibration sensor 6a. Since the vibration used here is a plate wave, the relationship between the envelope 43 and the phase 42 of the detected waveform with respect to the transmission distance in the vibration transmission plate 8 changes according to the transmission distance during the vibration transmission. To do. Here, the traveling speed of the envelope 43, that is, the group speed is Vg,
The speed at which the phase 42 advances and the phase speed are set to Vp. If the group velocity Vg and the phase velocity Vp are known, the distance between the vibration pen 3 and the vibration sensor 6a can be calculated from the vibration transmission time.

First, focusing only on the envelope 43, its speed is Vg, and a point on a certain specific waveform, for example, the first zero-cross point of the signal 44, which is the second differential waveform of the envelope 43, is the inflection point of the envelope 43. , The distance d between the vibration pen 3 and the vibration sensor 6a is detected.
Is given by d = Vg · tg (1) with its vibration transmission time being tg. Although this formula relates to the vibration sensor 6a, the other three vibration sensors 6b to 6d are expressed by the same formula.
The distance between the vibrating pen 3 and the vibrating pen 3 can be similarly expressed.

Further, in order to determine a finer coordinate position, processing based on the detection of the phase signal is performed. A time from a specific detection point of the phase waveform signal 42, for example, vibration application to a zero cross point after a certain predetermined signal level 431 is t.
p'47 (obtained by generating a window signal 46 of a predetermined width from the time when it exceeds the level 431 and comparing it with the phase signal 42)
Then, the distance between the vibration sensor and the vibration pen is: d = nλp + Vptp (2) (in the formula, tp is a value obtained by subtracting the delay time of a circuit or the like from tp '). Here, λp is the wavelength of the elastic wave, and n is an integer.

From the above equations (1) and (2), the above integer n
Is expressed as n = int [(Vg · tg−Vp · tp) / λp + 1 / N] (3).

Here, N is an appropriate real value other than "0". For example, if N = 2.0, even if the detected values of tg and tp having an error within ± 1/2 wavelength are obtained, the value of the equation (2) n can be correctly determined. By substituting N obtained as described above into the equation (2), the vibration pen 3
Also, the distance between the vibration sensors 6a can be accurately measured.

By the way, what is actually measured by the signal waveform detection circuit 9 is tg 'and tp' which include the offset for the delay time in the vibrating pen 3 and in the circuit.
When substituting into the equation (2) or the equation (3), it is necessary to subtract the offset amount and restore it to tg and tp. Due to the above-mentioned measurement of the two vibration transmission times tg 'and tp', the generation of the signals 45 and 47 is performed by the signal waveform detection circuit 9 as shown in FIG.

FIG. 6 shows a signal waveform detection circuit 9 of the coordinate input device.
FIG. 3 is a block diagram showing the configuration of FIG. In FIG. 6, the output signal of the vibration sensor 6a is filtered by the band pass filter 511 to remove excess frequency components of the detection signal, and is input to the envelope detection circuit 52 including, for example, an absolute value circuit and a low pass filter. , Only the envelope of the detection signal is taken out. The timing of the envelope inflection point is detected by the envelope inflection point detection circuit 53. In the peak detection circuit, a signal tg '(signal 45 in FIG. 5), which is an envelope delay time detection signal having a predetermined waveform, is formed by the tg signal detection circuit 54 composed of a mono-multivibrator or the like, and is input to the arithmetic control circuit 1.

On the other hand, 55 is a signal detection circuit, which is the envelope signal 43 detected by the envelope detection circuit 52.
The pulse signal of the portion exceeding the threshold signal 431 of the predetermined level is formed. 56 is a monostable multivibrator, which opens the gate signal 46 of a predetermined time width triggered by the first rising edge of the pulse signal. 57 is a tp comparator, which is a phase signal 4 while the gate signal 46 is open.
The first rising zero-cross point of 2 is detected, and the phase delay time signal tp′47 is supplied to the arithmetic control circuit 1. The circuit described above is for the vibration sensor 6a, and the same circuit is provided for other vibration sensors.

<Description of Circuit Delay Time Correction> The vibration transmission time latched by the latch circuit includes a circuit delay time et and a phase offset time toff.
The error caused by these is always included in the same amount when the vibration is transmitted from the vibration pen 3 to the vibration transmission plate 8 and the vibration sensors 6a to 6d.

Therefore, for example, from the position of the origin O in FIG.
For example, assuming that the distance to the vibration sensor 6a is R1 (= X / 2), the measured vibration transmission time from the origin O to the sensor 6a measured by inputting with the vibration pen 3 at the origin O is tg.
z ′, tpz ′, and tgz, tpz, which are the true transmission times from the origin O to the sensor, tgz ′ = tgz + et (4) tpz ′ = with respect to the circuit delay time et and the phase offset toff. tpz + et + toff ... (5).

On the other hand, the measured value tg 'at an arbitrary input point P
Similarly, tp 'is tg' = tg + et (6) tp '= tp + et + toff (7). When the difference between (4), (6), (5) and (7) is calculated, tg'-tgz '= (tg + et)-(tgz + et) = tg-tgz (8) tp'-tpz Since '= (tp' + et + toff)-(tpz + et + toff) = tp-tpz (9), the circuit delay time et and the phase offset toff included in each transmission time are removed, and from the position of the origin O. It is possible to obtain the difference in the true transmission delay time according to the distance from the position of the sensor 6a between the input points P as the starting point.
The distance difference can be obtained by using the equation (3).

Since the distance from the vibration sensor 6a to the origin O is stored in advance in a non-volatile memory or the like and is known, the distance between the vibration pen 3 and the vibration sensor 6a can be determined. The other sensors 6b to 6d can be similarly obtained.

The measured values tgz 'and tpz "at the origin O are stored in the nonvolatile memory at the time of shipment,
The equations (8) and (9) are executed before the calculation of the equations (2) and (3), and highly accurate measurement can be performed.

<Description of Coordinate Position Calculation (FIG. 7)> Next, the principle of actual detection of the coordinate position on the vibration transmission plate 8 by the vibration pen 3 will be described. Now, if four vibration sensors 6a to 6d are provided at positions S1 to S4 near the midpoints of four sides on the vibration transmission plate 8, each of the vibration sensors 3a to 6d from the position P of the vibrating pen 3 based on the principle described above. The linear distances da to dd to the positions of the vibration sensors 6a to 6d can be obtained. Further, the arithmetic control circuit 1 can obtain the value from the linear distances da to dd by the following formula from the Pythagorean theorem of the coordinates (x, y) of the position P of the vibrating pen 3.

X = (da + db) · (da−db) / 2X (10) y = (dc + dd) · (dc−dd) / 2Y (11) where X and Y are vibration sensors 6a and 6b, respectively. And the distance between the vibration sensors 6c and 6d.

As described above, the position coordinates of the vibrating pen 3 can be detected in real time, but the coordinates obtained here may include an error due to n jumps as described above. Therefore, the error is corrected according to the following procedure.

<Correction of Coordinate Position (FIG. 12)> FIG. 1 shows a flowchart of a processing procedure by the arithmetic control circuit 1 when detecting and correcting an error due to n jumps when calculating coordinates in the above procedure. .

First, when the operator designates the coordinates with the vibrating pen 3, the vibration is detected by one vibration sensor to obtain the transmission delay times tg 'and tp' as detection data.
According to the transmission delay time at the origin O stored in the memory, t
Off is deleted and converted into original data of tg and tp (step S11). This is processing according to equations (8) and (9).

Next, based on the original data tg and tp, the equation (3) is used.
Then, the distance n between the vibration sensor and the vibration pen is calculated by the equation (2) (step S12).

This processing is performed for a plurality of vibration sensors, and the designated coordinates of the vibration sensor 3 are obtained from the respective vibration sensors = distance between the vibrating pens by the equations (10) and (11) (step S13). This first coordinate calculation is performed in step S1.
The calculation is the same as the data acquired in 1.

When the first calculation is completed, the equation (1
For the coordinate values calculated by 0) and (11), error determination is performed using the following formula (step S14).

Da ² − (X / 2 + x) ² −y ² ≦ Eth (13) Dc ² − (Y / 2 + y) ² −x ² ≦ Eth (14) Da: Pen = Sensor 6a Distance Dc: Pen = Distance between Sensors 6c Eth: Judgment Threshold Here, as can be seen from FIG. 7, the right side of both equations should be 0 if the correct coordinates are calculated, but the judgment threshold Eth is n. A coordinate error other than the jump is also taken into consideration, and the value is set to a value equal to or smaller than the minimum value when the wavelength error (n jumps) occurs.
That is, when these expressions are not satisfied, it can be determined that the calculated coordinates include a wavelength error due to n jumps.

Regarding the calculated coordinate values, the above (1
Error determination is performed using the equations (3) and (14) (step S
14), the judgment result is passed (formulas 13 and 14 are both satisfied).
If so, the coordinates are output (step S18), and the process proceeds to the next coordinate point acquisition.

If the result of the determination is "NO", that is, if either of the equations (13) and (14) is not satisfied, it is determined in step S15 whether or not the correction has been completed, and the coordinate value is the first time, so the process proceeds to the correction process. (Step S15-NO).

In the correction process, first, the delay time data obtained from all the sensors are compared, and the sensor closest to the vibrating pen 3 (the latest sensor) is selected (step S16). According to FIG. 11, the closest sensor is most likely to have a negative ΔN fluctuation due to the influence of tg nonlinearity. Since the other factor, ΔN, should have occurred, the sensor may not always fly n times recently, but ΔN due to the non-linearity of tg is the largest (compared with the worst value). Is the highest. Therefore, as the next correction process, 1 is added to the n value in the equation (2) for the latest sensor (step S
17). If n jumps in the negative direction, step S
The correct n value is obtained by the correction process in 17.

When the value of n is corrected, the corrected n value is used to calculate the coordinates again (step S13) and the judgment is made again (step S14). Here, if the result is passed, that is, if the formulas (13) and (14) are satisfied, the coordinate is output (step S18), but if the result is not passed, the coordinate is not output and the coordinate is not output. (Step S15 → Step S11).

By performing the processing in accordance with the flowchart described above, distance correction can be performed based on the error determination result, and accurate coordinate output can be quickly performed by simple processing. In addition, since these processes can be performed by the arithmetic and control unit, they can be realized by executing the program by the microcomputer 31, and the hardware can be left as it is.

In the present embodiment, the high precision method using tg and tp has been described, but in the method using only tg, the time for one wavelength is added to tg for correction instead of the correction of the n value. Of course, it goes without saying that a coordinate input device that does not output coordinates with an error of one wavelength or more can be obtained.

Further, in this embodiment, the plate wave asymmetric wave (L
Although the amb wave Ao mode) has been described, the present invention is not limited to this, and this embodiment can be applied to any device that has a plurality of distance references and performs coordinate calculation. [Modification of Embodiment] In the above-described embodiment, the correction processing is completed once, but as can be easily inferred from FIG. 11, even if the vibration transmission distance is long, the same tg nonlinearity causes a minus. ΔN occurs. Therefore, it can be said that the sensor that has the highest possibility of n jumps next to the sensor recently is the vibration sensor (farthest sensor) located farthest from the input point by the vibration pen. In fact, it is not uncommon for recent sensors to fly n to the farthest sensor at the same time. Therefore, the flowchart of FIG. 12 is shown as a processing means for improving the pass rate after correction by correcting the n value of the farthest sensor when the determination does not pass even in the previous correction processing. In the figure, steps having the same reference numerals as in FIG. 1 perform the same processing.

As shown in the figure, the coordinates are calculated again using the n value that has been subjected to the first correction process of the latest sensor (step S13),
The determination is made again (step S14). Up to this point, the procedure is the same as the processing in FIG. If it fails, step S
20, the farthest sensor is selected (step S20) and correction processing (step S21) is performed. As can be understood from FIG. 11, the correction here is performed by adding 1 to n and increasing the wavelength by one wavelength. After this, the coordinate calculation is performed again. If the result is acceptable, the coordinate output is performed. If not, the acquired data is abandoned and the process proceeds to the next input data acquisition.

In this example, the number of correction processes is increased to two, but a significant reduction in processing procedure can be realized as compared with the conventional correction method. Conventionally, since the sensors are corrected in order, it is necessary to correct the n jump of two sensors in the worst 15th pass (the shortest second time). Further, compared with the above-mentioned embodiment, it takes extra time, but since the sensors to be corrected are the latest one and the farthest one, the number of cases that can be correctly obtained by the correction increases, and new sampling is performed again. Then, compared with the time and effort of recalculating, it cannot be said that the processing is always slow, and conversely, the processing can be speeded up. [Another Modification] Further, an embodiment will be described in which the possibility of passing judgment (correction output) by correction is increased and the correction processing procedure only needs to be performed once.

As described above, FIG. 11 shows ΔN obtained from the transmission delay time detected by the conventional circuit delay time correction method. By the way, as is clear from FIG.
The vibration transmission distance to a sensor other than the recent sensor does not become extremely small no matter where the pointing coordinate of the vibrating pen is located in the effective area A in FIG. Specifically, in the sensor arrangement of FIG. 7, if X = 250 mm, there is no more than two vibration sensors having a pen = sensor distance of 50 mm or less and having a large ΔN as shown in FIG. However, conversely, there are sensors other than the farthest sensor that have a long transmission distance. That is, the problem of ΔN generation at a long distance may exist in a plurality of sensors, but it can be said that the generation of ΔN at a short distance is a problem limited to one sensor.

Therefore, it is desirable to use means for calculating n as shown in FIG. The group velocity Vg used in FIG. 11 (conventional) is an average value over the entire distance, and the relationship with the actual tg is as shown in FIG. FIG. 13 shows that the slope of the asymptote C of the curve of the group delay time tg obtained at a long distance from the middle is used as the group velocity Vg in the expression (P4) (also used in the coordinate calculation processing expression (3)) and the long distance. (20 in the figure
It is a plot of ΔN obtained when the circuit delay time is corrected with reference to the vibration transmission time at the 0 mm position.

As can be seen from FIG. 13, the large negative deviation of ΔN due to tg nonlinearity is due to the vibration transmission distance of 5
It is 0 mm or less. Since the vibration sensor existing below 50 mm has only one sensor in the sensor arrangement configuration of FIG. 7,
The effect of “the possibility of n jumps is recently concentrated on the sensor” appears. Therefore, as the correction processing procedure, the processing procedure shown in FIG. 1 (correction processing only once for the latest sensor)
The coordinate correction is realized with a sufficient probability.

The circuit delay time is corrected with reference to the vibration transmission time at a long distance (200 mm in the figure) position, and
The above effect can be obtained by performing the coordinate calculation process with the gradient of the asymptote of the curve of tg obtained from the middle to the long distance as Vg.

When correcting the circuit delay time, a point equidistant from each vibration sensor (for example, origin O in the figure) is not used, and one correction coordinate point is prepared for each vibration sensor. Is complicated because coordinate conversion processing when calculating coordinates from the obtained pen-sensor distance becomes N.
By changing the value, a calculation formula may be prepared such that ΔN has a predetermined negative value at the origin O (125 mm position) in FIG. This makes the coordinate calculation process easier.

Further, even if the gradient of the asymptotic line C of the curve obtained from tg obtained by measuring at a distance farther than the middle between the vibration sensor and the vibration pen is not Vg, a predetermined value is added to the conventional Vg, A large estimated value may be set as Vg.

The present invention may be applied to a system composed of a plurality of devices or an apparatus composed of a single device. Further, it goes without saying that the present invention can be applied to the case where it is achieved by supplying a program to a system or an apparatus.

[0088]

As described above, in the coordinate input device according to the present invention, the wavelength error at the time of distance detection is not random, but is affected by the non-linearity of the group delay time tg, and the nearest sensor is negative. Focusing on the fact that errors occur most often on the side, the error in detecting coordinates is determined, and the distance from the detection location is corrected by one wavelength to the positive side, so that high-precision coordinate output is always stable. The effect is that it is obtained by a simple structure.

[0089]

[Brief description of drawings]

FIG. 1 is a processing flowchart of an arithmetic control circuit according to an embodiment.

FIG. 2 is a block configuration diagram of a coordinate input device according to an embodiment.

FIG. 3 is a diagram showing a configuration of a vibrating pen.

FIG. 4 is an internal configuration diagram of an arithmetic control circuit.

FIG. 5 is a time chart of signal processing according to the embodiment.

FIG. 6 is a block diagram of a signal waveform detection circuit according to an embodiment.

FIG. 7 is a diagram showing a coordinate system of a coordinate system input device.

FIG. 8 is a frequency characteristic diagram of a detection signal and a frequency characteristic diagram of a propagation velocity of a plate wave.

FIG. 9 is an explanatory diagram of waveform deformation according to a vibration transmission distance.

FIG. 10 is an explanatory diagram of tg nonlinearity.

FIG. 11 is a diagram showing an integerization error (ΔN) obtained by a conventional delay time correction and coordinate calculation method.

FIG. 12 is a processing flowchart of an arithmetic control circuit according to another procedure.

FIG. 13 is a diagram showing another example of an integer error (ΔN).

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Arithmetic control circuit 2 Vibrator drive circuit 3 Vibration input pen 4 Vibrator 5 Pen tip 6a-6d Vibration sensor 7 Vibration isolator 8 Vibration transmission plate 9 Signal waveform detection circuit

 ─────────────────────────────────────────────────── ─── Continuation of front page (72) Inventor Katsuyuki Kobayashi 3-30-2 Shimomaruko, Ota-ku, Tokyo Canon Inc. (72) Inventor Hajime Sato 3-30-2 Shimomaruko, Ota-ku, Tokyo Canon Incorporated (72) Inventor Yuichiro Yoshimura 3-30-2 Shimomaruko, Ota-ku, Tokyo Canon Inc.

Claims (6)

[Claims] 1. Input means for inputting vibration of a predetermined wavelength to a desired input position on a vibration transmission body, and detection means for detecting vibration input by said input means at a plurality of detection positions of said vibration transmission body. A distance calculation for calculating the transmission distance from the input position to each of the detection positions for each of the detection positions, based on a delay time from when the vibration is input by the input unit to when the vibration is detected by the detection unit. Means, coordinate calculation means for calculating the coordinates of the input position based on the transmission distance calculated by the distance calculation means, coordinates calculated by the coordinate calculation means, and transmission distance calculated by the distance calculation means. Determination means for determining whether or not the error included in the coordinate is an allowable amount based on the above, and a predetermined amount for the transmission distance from the input position based on the determination by the determination means. Selecting a detection position candidate including an error depending on the transmission distance, a correction unit correcting the transmission distance at the detection position of the candidate selected by the selection unit in the predetermined wavelength unit, and the correction unit. Control means for controlling to correct the coordinates by using the transmission distance corrected by the means, the coordinate input device. 2. The distance calculating means calculates a transmission distance in wavelength units from the phase delay time and the phase speed according to the distance calculated from the group delay time and the group speed of the vibration detected by the detecting means, The coordinate input device according to claim 1, wherein the correction unit corrects an error in a wavelength unit included in the transmission distance calculated in the wavelength unit. 3. The coordinate input device according to claim 1, wherein the selection unit selects a detection position having the shortest transmission distance from the input position as a candidate for a detection position including an error in the transmission distance. . 4. The selecting means selects the detection position having the shortest transmission distance from the input position as a first candidate for the detection position having an error in the transmission distance, and detects the longest transmission distance from the input position. The coordinate input device according to claim 1, wherein the position is selected as the second candidate. 5. The coordinate input device according to claim 1, wherein the distance calculating means preliminarily corrects a transmission distance calculated at a predetermined transmission distance so as not to include an error. . 6. A vibration having a predetermined wavelength is input to a desired input position on the vibration transmitting body, and a vibration is input by the input means to a transmission distance from the input position to a plurality of detection positions of the vibration transmitting body. A coordinate error correction method in a coordinate input device for calculating the coordinates of the input position by calculating based on the delay time from the detection to the detection means, the method being based on the calculated coordinates and the transmission distance. , A determination step of determining whether or not the error included in the coordinates is an allowable amount, and based on the determination in the determination step, a detection position candidate including an error that exceeds a predetermined amount in the transmission distance from the input position is determined as the transmission distance. A selection step of selecting according to the correction step, a correction step of correcting the transmission distance at the detection position of the candidate selected by the selection step in units of the predetermined wavelength, and a transmission distance corrected by the correction step. Coordinate error correction method characterized by and a control step of controlling to correct the coordinates used.