JPH0368869A - Speed detector - Google Patents

Abstract

PURPOSE:To make speed detection in a low speed region accurate by generating basic pulses in a phase A and a phase B whose phases are different by 90 degrees to each other every time when the rotating position of a rotary body is changed by a specified unit amount, generating the total of four multiplying pulses in the one period, and detecting the speed of the rotary body by using the pulses. CONSTITUTION:An optical encoder is used as a pulse generating means. Basic pulses 100A and 100B in a phase A (b) and a phase B (c), whose phases are different by 90 degrees every time when the rotating position of a rotary disk 10 is changed by a specified unit amount, are generated. A rotating-direction detecting circuit 32 detects the rotating direction of the disk 10 based on the basic pulses 100A and 100B. In a multiplying-pulse generating circuit 34, multiplying pulses 200 are generated in synchronization with the rise and fall of the pulses 100A and 100B. Thereafter, the pulses 100A and 100B are inputted into a speed operating part 38a and a rotating-data adding part 38b constituting a low-speed-region speed-operating circuit 38. The output of the operating part 38a is added to the adding part 38b, and the rotating speed is found based on the data between the pulses.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a speed detection device, and particularly to an improvement in a speed detection device that digitally detects the speed of a rotating body in a low speed range.

[Prior Art] Background Art In the field of mechatronics such as robots and NC controllers, various servo motors such as DC motors and C motors are widely used. In such fields, in order to improve the control characteristics, it is necessary to accurately detect the speed of a rotating body such as a servo motor and perform control based on the detected value.

Conventionally, various types of rotating body speed detection devices have been proposed, and in particular, rotary encoders (particularly incremental type) that are small and easy to interface with a microprocessor are used as the rotation detection section.
(hereinafter, an incremental rotary encoder will be simply referred to as an encoder).

The encoder is basically coupled to a rotating body to be measured. Then, every time the rotating body rotates by a predetermined angle, the encoder outputs A-phase and B-phase basic pulses having a phase difference of 90°.

In speed detection devices, in order to increase the number of pulses output from the encoder, multiplied pulses are usually generated in synchronization with the rising and falling edges of the A-phase and B-phase basic pulses. A total of four multiplied pulses are output per basic pulse period, and as compared to the case where only basic pulses are used, the resolution in the low speed range can be improved and the speed detection accuracy can be improved.

Conventional speed detection is performed by counting the number of multiplied pulses generated as the rotating body rotates at regular intervals, and determining a speed proportional to the counted value.

However, while such a speed detection device can accurately detect the speed in a high speed rotation range where the number of generated pulses is large, the speed detection accuracy is extremely degraded in a low speed rotation range where the number of generated pulses decreases.

For this reason, speed detection in a low-speed rotation range where the number of generated pulses decreases has been carried out by measuring the generation time interval of the multiplied pulses.

Prior Art However, such speed detection has low detection accuracy in a low speed rotation range where the frequency of multiplied pulse generation is low.

For this reason, proposals have been made for a speed detection device that uses some kind of estimation or 4V to estimate the speed at the time when there is no multiplied pulse and improves the resolution in the low speed range.

This type of speed detection device can be broadly divided into those using an analog method and those using a digital method.

As a speed detection device using an analog method, a proposal disclosed in, for example, Japanese Unexamined Patent Publication No. 182988/1982 is known. This conventional technology uses the phase angle of a high-precision sine wave encoder to improve low-speed resolution, but since it includes an analog section, it suffers from changes over time and the equipment is expensive and large. be.

In addition, speed detection devices using digital methods do not have the same problems as those using analog methods, but the duty of the multiplied pulses itself is affected by temperature changes in the encoder.
Since this includes large error factors caused by changes over time, there is a problem in that good resolution cannot be obtained unless these are corrected.

That is, the A-phase and B-phase basic pulses output from the encoder are set so that their duty ratios are 50=50. However, due to various causes such as changes in temperature of the encoder and changes over time, the duty ratio is often not exactly 50:50.

Therefore, in a speed detection device using a digital method, correcting the error included in the pulse duty itself by some means is an important issue for accurate speed detection. ■■
The proposals shown in , ■ have been made.

■: First, JP-A-83-49807. E13-498
In the proposal according to No. 13-49809, a constant speed rotation measurement value is obtained in advance before regular operation, and a speed in a low speed range is obtained by a correction pulse duty based on this.

■: Also, in the proposal in JP-A No. 59-225355, the rising fundamental period and falling fundamental period are measured from the A-phase and B-phase fundamental pulses, and the last counted fundamental period before the multiplication pulse is generated. It is configured to calculate the speed in the low speed range based on the period (the basic period here is
(This is the period of the basic pulse output from the encoder, not the period of the multiplied pulse).

[Problems to be Solved by the Invention] However, in the prior art described in item (2) above, it is necessary to perform low-speed rotation measurements on all encoders in advance and obtain correction data as initial data, and it is necessary to obtain correction data as initial data. There was a problem in that the initial data could not sufficiently correct for changes over time.

In addition, in the conventional technology related to (2) above, although it is possible to correct the pulse duty using periodicity, in a low-speed region where there is no multiplied pulse in the sampling period,
Since the last counted fundamental period is held until the next pulse is generated, there is a problem in that the responsiveness of speed detection deteriorates.

[Purpose of the Invention] The present invention has been made in view of such conventional problems, and its purpose is to: ■ eliminate the use of expensive and complicated analog methods that change over time; To provide a speed detecting device which is not easily affected by the noise and can detect the speed of a rotating body with good responsiveness and high precision even in a low speed range where not even a single multiplied pulse is generated in a sampling period.

[Means for Solving the Problems] In order to achieve the above object, the present invention provides the following features: Each time the rotational position of the rotating body changes by a predetermined single position,
a pulse generating means for generating A-phase and B-phase basic pulses having different phases by 0 degrees, and generating a total of four multiplied pulses in synchronization with the rising and falling edges of each basic pulse during one period of the basic pulse; a rotational direction detection means for detecting the rotational direction of the rotary body based on the A-phase and B-phase basic pulses; and an elapsed time elapsed since the multiplication pulse generation means outputs the final multiplied pulse. Multiplied pulse information observation means for measuring as time information and sequentially updating and storing pulse interval information of the last five multiplied pulses, rotational direction information of the rotating body output from the rotational direction detection means, and multiplied pulse information The elapsed time information from the time when the final multiplied pulse was generated and the last five pulse interval information outputted from the observation means are based on the assumption that the acceleration in the low speed range is constant, the periodic conditions of the waveform in the pulse generation means, low-speed range speed calculation means for calculating and outputting a speed in a low-speed range based on a predetermined speed calculation formula derived from a pulse phase angle preservation condition; Moreover, the speed detection is performed while compensating for the duty error of the multiplied pulse signal.

[Principle of the Invention] Next, the present invention will be explained in detail. For convenience of explanation, the rotational angular velocity of the rotating body will be referred to as speed, the rotational angular acceleration will be referred to as acceleration, and the period of the basic pulse before multiplication will be referred to as basic period.

Basic pulse In the present invention, each time the rotational position of the rotating body changes by a predetermined single position, the pulse generating means generates A with a phase difference of 90° from each other.
Basic pulses of iII and B phases are output. Various types of pulse generating means can be used as required, but the most common type of pulse generating means is a magnetic or optical rotary encoder.

FIG. 7 shows an example of an optical encoder, in which a rotating disk 10 is fixedly attached to a rotating body to be measured, and a plurality of pores 12 are formed around the rotating disk 10 at regular intervals. is drilled. A light source 14 and light receiving elements 18a and 18b are provided on both sides of the rotating disk 10.
are arranged facing each other, and the light emitted from the light source 14 toward the pore 12 of the rotating disk 10 is received by the light receiving elements 1.8a and 1.8b through the fixed slit 16 with light having a 90° phase difference [7]. It is configured like this.

Therefore, as the rotating disk 10 rotates, the light receiving element 18
A and 18b output sinusoidal electrical signals with a phase difference of 90 degrees, and this signal is converted into the A-phase and BMJ fundamental pulses 1.
Output as 00A and 100B.

Multiplied Pulse FIG. 8 shows the A-phase and B-phase basic pulses 100A and 100B output in this manner. The multiplied pulse generation means outputs the A phase and B phase output in this way.
Phase basic pulse 100A.

A multiplied pulse 200 is output in synchronization with the rise and fall of 100B. Therefore, during one cycle of the basic pulse 100, a total of four multiplied pulses 200 are output as shown in the figure.

In FIG. 9, the multiplied pulse 200 is shown. As shown in the figure, the time intervals of the multiplied pulses output from the multiplied pulse generation means are expressed as t [01, t [1], t [2], etc. in order from the newest pulse to the past. Make it.

Moreover, each multiplied pulse t [01, t [11, t [2
]... The basic periods of the basic pulses 100 corresponding to t) [0] in order from the newest to the past based on the following formula:
, p[1].

Let it be expressed as p[2]...

p[i] = t [1] + t [1+1] + t
[1+2] + t [l+3] However, i -0,1,2
,... (1) In other words, the multiplied pulse is the basic pulse - a total of 4 per period
will be output. Therefore, it is understood that the period of the t-base water pulse 100 corresponding to the multiplied pulse 200-i is obtained as the total value of four multiplied pulses, as shown in equation (1) above.

Further, as shown in FIG. 9, the elapsed time t from the time when the last multiplied pulse 200-0 is outputted from the multiplied pulse generating means to the present time is determined by the elapsed time information (t-f'reerun).
).

As described above, in the conventional speed detection technique, the number of multiplied pulses 200 thus obtained is counted, and the speed is detected based on the pulse interval. However, the multiplied pulse train obtained in this way has
As mentioned above, since the duty error is included, even if the rotation speed is constant, each multiplied pulse 200-0,
Pulse interval t[0], t[1] of 200-1...
1, t [2], t [3].

t[4] will not have equal values. The main causes of such a duty error are, for example, the output levels of the sine waves and cosine waves output from the light receiving elements 18a and 18b of the encoder shown in FIG. 7, and the waveform shaping circuits 20a and 20.
This is because the threshold level used to output the basic pulses 100A and 100B in b varies depending on temperature and changes over time.

Countermeasures for Duty Error The present inventor studied such duty errors included in a multiplied pulse train, and found that even if the multiplied pulse train includes a duty error, the properties described in (a) and (b) below do not change. I discovered that.

Note that the time axis and rotation angle axis mentioned below do not indicate mechanical axes, but correspond to the meaning of the axes in the Cartesian coordinate system, and the time axis represents the observation of a phenomenon corresponding to the independent variable time. It is used in the sense of looking from. The same applies to the rotation angle axis.

Here, we purposely distinguish between the rotation angle axis and the time axis because the speed estimation method proposed below also takes into account temporal changes in speed (that is, Δθ≠K・Δt; is a constant). . Finding speed means rotation angle (axis) and time (axis)
The question is how to deal with the relationship between If the speed is constant, the rotation angle axis and the time axis will have a proportional relationship.

Characteristics specific to the encoder structure are expressed relative to the angular axis of rotation. On the other hand, pulse information is observed along the time axis. In this regard, the two conditions described below focus on the structural properties of the encoder, and are therefore expressed regarding the rotation angle axis.

(a): Encoder (pulse generator) white waveform period condition At constant speed rotation, basic period p of basic pulse 1-00
is kept almost constant when viewed from above the rotation angle axis θ.

In other words, even when the rotating body is rotating at a constant speed,
Although the multiplied pulse includes a duty error, the fundamental period p of the fundamental pulse 100 itself does not include such an error and remains almost constant when viewed from the rotation angle axis θ (see Fig. 11 (A)). ).

For example, a light receiving element 18a as shown in FIG.

Let us consider a case where the magnitudes of the sinusoidal waveforms of the A-phase and B-phase signals output from the 18b are equal. The quadrupled pulse is generated at the point where the sinusoidal waveform and the pulsing threshold level intersect (point in the figure).

The original threshold levels of A phase and B phase are exactly Le.
If vel O, Δθ' [1] corresponds to the interval from the generation of the quadrupled pulse until the next generation.

Δθ' [l+1], Δθ' [142], Δθ' [
143] are both equal and Δ θ ′
[l] −Δθ' [l+1]
-Δθ'[l+2]-Δθ' [l
+3]. In other words, there is no duty error.

Next, consider the case where an error occurs. If the threshold levels of A phase and B phase change to Level A and Level B, the corresponding angular displacement becomes Δθ[l as shown in the figure.
], Δθ[1+l], Δθ[1+21, Δθ[1+
3], which causes a duty error to occur. However, in this case, if we pay attention to the fundamental period of the sinusoidal waveform, it can be seen that it is not affected by fluctuations in the threshold level.

If this is shown for phase A, the fundamental period (angle) ΔPθ[
I] − Δθ[1] + Δθ[141] + Δθ[142] + Δθ[
1+3] corresponds to being kept almost constant.

(b): Pulse phase angle preservation condition The ratio of four multiplied pulse intervals (duty ratio) occurring within the basic period p of the basic pulse 100 is itself:
If observed along the rotation angle axis, it remains approximately constant in each neighborhood. In other words, the pulse phase angle observed with a shift of one fundamental period on the angular axis is preserved.

This means that even if the threshold level changes, the pulse interval (angle) displacement by one fundamental period (angle) is almost preserved, and as shown in Figure 13, Δθ[1+3] −Δ θ[1-1]Δ θ[1+
2] −Δ θ [i−2] Δ θ [1+1]
This corresponds to the relationship -Δ θ [i-3].

The above discussion of (a) and (b) has been explained for the case where the magnitude of the sinusoidal waveform is equal in the A and B phases, but the same applies even when the magnitude of the waveform changes as shown in Fig. 14. ,
This also applies when both the threshold level and the waveform size change.

Speed Calculation Formula The present inventors focused on the properties shown in (a) and (b) and established the method for speed calculation described below.

(2): First, it is assumed that the time change in the target system is a constant acceleration α in the interval until the next multiplied pulse is generated.

At low speeds, for example, the position of the servo system is close to the target, so this assumption is not unreasonable. In other words, the time point when the last multiplied pulse occurs is the time origin 1-0
If we assume a constant acceleration α, the angle θ at time t is expressed by the following equation (2), and the velocity V# is expressed by the following equation (3).
).

θ−(α/2)・t2+β・t+θ.

(2) ■, −α ◆ t +β
... (3) ■: Here, what we can observe is the number of pulses and the pulse interval along the time axis, as shown in Figure 11 (B), and as shown in Figure 11 (A). Direct observation is not possible along the rotation angle axis shown.

In other words, if the rotational motion is not constant, the waveform on the rotation angle axis shown in FIG. 11(A) and the waveform on the time axis shown in FIG. 11(B) will be different.

Therefore, the pulse intervals of the multiplied pulses obtained from such a waveform naturally become different, and the correspondence between the actual rotation angle of the rotating body and time becomes the relationship shown in FIG. 12.

In this case, the constant α in equations (2) and (3) above.

In order to find β, the above-mentioned duty error property (a) is used.
, (b) may be used to perform the following calculations.

(a'): Cycle condition of encoder internal waveform (a) above
When the condition is applied to the case of constant acceleration motion of the above equation (2), the following equation (4) is obtained.

(2r/N) ・revol[O]-1(α/2
)・ (+)[0]) ' + β・ (p
[mouth])... (4) Here p[0ko - t [01+ t [13+
t[2] +t[3]p[l] =t[1]
] +t[2] +t[31+t[41revol
[0] : Rotation direction corresponding to t [01 (positive direction +
1. Negative direction -1) N: Represents the basic number of pulses generated per encoder rotation (before multiplication).

That is, it is assumed that, for example, N basic pulses]-00A are generated from the waveform shaping circuit 20a per revolution of the rotating disk 10. At this time, the amount of rotational position displacement of the rotary disk 10 corresponding to one period p[0] of the basic pulse]00A is expressed by the left side of equation (4).

Furthermore, assuming that the rotating body is moving at a constant acceleration as described above, the fundamental period p[0
] The amount of rotational position displacement of the rotating body within
It will be expressed on the right side of the equation.

Therefore, it will be understood that if the condition (a) above is applied to the case of uniformly accelerated motion, the above equation (4) can be obtained.

(b'): Pulse phase angle preservation condition Further, when the condition (b) is applied to the case of uniform acceleration motion of the above equation (2), the following equation (5) is obtained.

−(α/2)・t[0]” +β・t[0]
-(α/2) (p[012-(p[1
]+t[01)")+β・t[4]... (5) In other words, the left side of the above equation (5) is the output of the next multiplied pulse after the multiplied pulse 200-1- shown in FIG. 9 is output. 20
It represents the amount of change in angle of the rotating body until 0-0 is output.

In addition, the right side of the above equation (5) is the multiplied pulse 200-4
The amount of change in angle between when is output and when 200-0 is output is subtracted from the amount of change in angle between when multiplied pulse 200-5 is output and when multiplied pulse 200-0 is output. This value is the amount of angular change between the multiplied pulses 200-5 and 200-4.

Therefore, from condition (b) pulse position angle preservation condition, the above-mentioned (
5) It is understood that the left and right sides of the equation have equal values, and the open equation holds true.

From equations (4) and (5) obtained in this way, the constants a and β
By finding and substituting this into the above equation (2),
Time t (t
−freerun) has elapsed, V,
is given by the following equation (6).

V e = (2r /4N) ・revof[0]
・l(t[4]=t[0])・``reerun+p[0]
]2+2t[4]・p[0]=t[4]21/[p
[0]・p[1]・(t[ol+t[4])1...
(6) Characteristics of the velocity calculation formula ■: As mentioned above, this formula (6) compensates for the duty error using a method that takes into account the periodic conditions and pulse phase angle preservation conditions that exist within the encoder. The speed obtained will be of high precision.

■: Furthermore, since Equation (6) above estimates the speed in the interval until the next multiplied pulse is generated based on the assumption that the acceleration is constant, it is necessary to avoid generating a single multiplied pulse in the sampling period. Even at low speeds, it is possible to perform speed detection with good responsiveness (accurate speed estimation).

Therefore, by digitally calculating the speed of the rotating body in the low speed range using Equation (6) above, the above-mentioned conventional problems can be solved, the response in the low speed range is good, and the duty error of the multiplied pulse is This makes it possible to perform highly accurate speed detection that compensates for the

[Work] Next, the present invention will be explained in detail.

In the present invention, for example, an encoder as shown in FIG. 7 is used as a pulse generating means, and each time the rotational position of the rotating body changes by a predetermined single position, the Basic pulses 100A and 100B of the phase are generated.

The basic pulses 100A and 100B are as described in (a) above.

As described above, the two conditions of (b), the cycle condition of the waveform in the encoder and the pulse phase angle preservation condition, are satisfied.

The rotational direction detection means detects the rotational direction of the rotating body based on the A-phase and B-phase basic pulses 100A and 100B.

Further, the multiplied pulse generating means generates a multiplied pulse 200 as shown in FIG. 8 in synchronization with the rise and fall of the basic pulses 100A and 100B output in this manner.

FIG. 9 shows a multiplied pulse 20 output in this way.
The relationship between the pulse interval t of 0 and the fundamental period p of the fundamental pulse 100 is shown. In the figure, 200-0 represents the last multiplied pulse output.

The multiplied pulse information observation means measures the elapsed time t from the time when the last multiplied pulse 200-0 was output to the current time as elapsed time information (t"rreerun), and also measures the last five multiplied pulses. 200-0.20
Pulse interval information of 0-1...200-4 (t[0]
, t[11,...t[4]) are sequentially updated and stored.

It is preferable that such a pulse information observation means is constructed using a timer means and a storage means.

Then, the timer means measures the elapsed time t after the final multiplied pulse is output from the multiplied pulse generation means, and outputs this as elapsed time information (t - rreorun ). At the same time, this elapsed time information t when the next multiplied pulse is output is output as the latest pulse interval information.

The storage means stores the latest five pulse interval information t [0], t 01, . . . outputted from the timer means in this way.
- Sequentially update and store t[4].

As a result, the timer means always outputs the final multiplied pulse 2.
Information on the elapsed time since 00-0 was output (t −f'
reerun) can be obtained, and the latest five pulse interval information 1[0] can be obtained from the storage means.

t[1]...t[4] can be obtained.

The feature of the present invention is that each data obtained in this way is
The low-speed The goal is to calculate and output the speed in the area.

In this way, speed calculation is performed using an arithmetic expression that takes into account the cycle condition of the waveform in the encoder and the pulse phase angle preservation condition, so the duty error included in the multiplied pulse signal is compensated for.
Speeds in low speed ranges can be detected accurately.

Furthermore, according to the present invention, the speed in the section until the next multiplied pulse is generated is estimated based on the assumption that the acceleration is constant, so that it is possible to estimate the speed in the section until the next multiplied pulse is generated. However, it is possible to perform speed detection with good responsiveness.

As such a speed calculation equation, it is preferable to use the above-mentioned equation (6).

[Effects of the Invention] As explained above, according to the present invention, the constituent elements of each means can be configured with digital circuit elements (including a computer), so compared to the conventional speed detection device that uses an analog method, There is no performance deterioration due to changes, and the configuration of the device can be made simple and inexpensive.

Furthermore, according to the present invention, by using a speed calculation formula that satisfies the waveform period conditions and pulse phase angle preservation conditions within the pulse generation means, the duty error of the multiplied pulse is compensated for, and the effects of temperature and aging changes are compensated for. It becomes possible to perform speed detection in a low speed range with a high degree of rust without being affected.

Furthermore, according to the present invention, by using a speed calculation formula based on the condition that acceleration in the low speed range is constant, speed detection with good responsiveness is possible even in the low speed range where multiplied pulses do not occur for a long time. It can be carried out.

[Embodiments] Next, preferred embodiments of the present invention will be described in detail with reference to the drawings.

First Embodiment FIG. 1 shows a preferred first embodiment of the speed detection device according to the present invention.
An example is shown.

In the speed detection device of the present invention, a pulse generator 30 is attached to a rotating body to be measured, and each time the rotational position of the rotating body changes by a predetermined single position, the
Phase basic pulse 100A.

Generates 100B.

The pulse generator 30 of the embodiment is constructed using the optical encoder shown in FIG. 7 above.

Then, the basic pulses 100A of A phase and B phase shown in FIG. 8 are generated from the waveform shaping circuits 20a and 20b.

100B is the rotation direction detection circuit 32. The signal is output to the multiplied pulse generation circuit 34.

The rotational direction detection circuit 32 detects the rotational direction revol of the rotating body in real time from the phase relationship of the A-phase and B-phase 100A, 100B input in this way, and sends this detection signal to the low-speed region speed calculation circuit 38. Output direction.

Further, the multiplied pulse generation circuit 34 generates a multiplied pulse 2 as shown in FIG.
00 is output and inputted to the multiplied pulse information observation circuit 36.

FIG. 9 shows the basic pulses 2 that are sequentially output in this way.
00-0.200-1. . . . and the relationship between each pulse interval is shown.

Then, the multiplied pulse information observation circuit 36 detects the elapsed time t (rreo
run) is defined as API. Furthermore, the pulse interval information t[0], t[lgo...t[4] of the latest five multiplied pulses 200-0.200-1...200-4 is updated and stored in sequence. Then, this information is outputted to the low speed range speed calculation circuit 38.

FIG. 2 shows a specific configuration of the multiplied pulse information observation circuit 36 of the embodiment. The observation circuit 36 of the embodiment is
Free run timer 40, pulse generation detection circuit 42ε,
Five time interval data registers 50, 52, . . . , 58 connected in series to the free-run timer 40, and time interval data transfer switches provided in front of each register 50, 52, . . . , 58. 60.62...68.

Then, the free run timer 40 generates a multiplied pulse 20
When 0 is entered manually, the elapsed time from that point is counted, and the value is output as elapsed time information (1-rreerun) to the low speed range speed calculation circuit 38 via line 216.

Further, the pulse generation detection circuit 42 operates based on the multiplied pulse 200 and the measurement cycle pulse signal 210, which are manually inputted.
In synchronization with the multiplied pulse signal 200, a data shift pulse signal 214 is output to each data transfer switch 60, 62, . . . , 68, and a reset signal 212 is output to the free run timer 40.

Therefore, when the data shift pulse signal 214 is output due to the input of the multiplied pulse 200, the free run timer 40
The output is connected to a data transfer switch 60. The last data of the elapsed time information (Rreerun) is then transferred to the time interval data replenishment register 50 as the latest pulse interval information.

Similarly, the pulse interval information in the registers 50, 52, 54.56 is transferred to the next register 52, 54° 56.58 via the switch 62.64° 66.68. Immediately after the transfer operation ends, the pulse generation detection circuit 42 resets the free run timer 40, and the free run timer 40 starts measuring elapsed time information (freevn) for the next multiplied pulse 200.

In this way, as shown in FIG. 9, the free run timer 40 outputs the elapsed time information (rrcerυn) since the last multiplied pulse 200-0 was output in real time, and the five registers 50° From 52, . . . 58, the latest five multiplied pulses 200-0.200-1 .・・・
・Pulse interval t [0], t [1] of 200-4
...t[4] will be sequentially updated and output.

Then, each data output in this way is transmitted to line 21.
6, 218...226 to the low speed range speed calculation circuit 38
The direction is output.

The speed calculation circuit 38 of the present invention substitutes each data thus input into the speed calculation formula shown in equation (6) above, and calculates and outputs the speed V in the low speed range.

The speed calculation circuit 38 of the embodiment includes a speed calculation section 38a and a rotation information addition section 38b.

The speed calculation section 38a includes a multiplied pulse information observation circuit 36
Based on the data input from
The calculation of the above equation (6) excluding l[0] is performed, and the calculation result is manually input to the rotation information adding section 38b.

The rotation information addition unit 38b calculates the rotation direction rev[ from the rotation direction detection circuit 32 when the final multiplied pulse 200-0 is generated, based on the data manually input in this way.
01 is added and the final calculation of equation (6) is performed. Then, the speed (2) obtained by this calculation is output as the detected speed in the low speed range.

As described above, according to the present invention, a digital method is used to improve the resolution in the low speed range, and
Because it does not use an analog method like the proposal in issue B,
Compared to this conventional device, there is less change in measurement data over time, and the overall structure of the device can be made simple and inexpensive.

Further, according to the present invention, the encoder (pulse generator 3
The duty error of the multiplied pulse is compensated for using the speed calculation formula (6) above, which utilizes the waveform period condition within 0) and the pulse phase angle preservation condition. Therefore, the conventional technique (Japanese Unexamined Patent Application Publication No. 63-1989) involves determining the constant speed rotation measurement value in advance before regular operation and correcting the pulse duty error based on this value.
49807, etc.), it is not necessary to obtain correction data in advance for all encoders (pulse generator 30), and moreover, it is possible to perform highly accurate speed detection that is less affected by temperature and changes over time than conventional devices. I can do it.

Further, according to the present invention, the speed calculation formula shown in equation (6) above is created based on the assumption that the acceleration in the low speed range is constant, and the speed calculation in the low speed range is performed based on this speed calculation formula. ing. Therefore, the basic pulse of A phase and B phase is 100A.
, 100B, the rising basic period and the falling basic period are totaled 711+1, and the speed in the lower speed range is calculated from the last counted basic period before the sampling pulse is generated. Even in a low speed range where no pulse is generated in a sampling period, the responsiveness of speed detection is not reduced, and moreover, it is possible to perform highly accurate speed detection.

3 and 4 show data measured using the speed detection device of the present invention and data measured using conventional speed detection technology.

FIG. 3 shows a case where the speed of the rotating body changes stepwise in a low speed range, and FIG. 4 shows a case where the speed of the rotating body changes sinusoidally.

In each measurement date, the solid line represents the actual speed, the thin line represents the data measured using the device of the present invention, and the broken line represents the data measured by the prior art, which simply determines the speed using the previous time interval. represents.

As is clear from this experimental data, according to the present invention, the responsiveness in the low speed range is good in any case, and speed detection can be performed while compensating for pulse duty errors.

In addition, in the present invention, the speed calculation circuit section 38a includes:
It is also possible to configure software using a microcomputer.

FIG. 5 shows an example of the operation of the microcomputer in this case. In this case, the initialization routine 150 is executed first, and then the operation wait routine 15
1 and the counting period pulse generation detection routine 152 waits for the next counting period pulse to occur.

Then, when a counting period pulse is generated, the pulse interval information manual routine 153 calculates each pulse interval information t[0], t[11] from the multiplied pulse information observation means 36.
,...t[4] and elapsed time information (freerun
) into the memory area and execute the following basic period calculation routine 1.
Execute 54.

This basic period calculation routine 151 calculates the basic period of the encoder (the basic period of the basic pulse 100) p[0] and p[l] from the pulse interval information.

Then, using the data obtained in this way, a part of the numerator and a part of the denominator of the formula (6) are respectively calculated in the molecular product calculation routine 1552 and the denominator term calculation routine 156, and using this calculation data, Next speed calculation routine 1
57, find the velocity V.

Then, the obtained speed V is applied to the next speed output routine 15.
Output at 8.

In this way, the speed calculation section 38a can also be configured in software using a microcomputer or the like. Note that the low speed range speed calculation circuit 38 shown in FIG. When the speed is determined by software, the microcomputer itself can act as the rotation information adding section 38b.

Note that in this embodiment, the measurement period pulse signal 200 is used as one of the timing signals.

This is due to the following reasons.

Generally, in control using a microcomputer, the first
As shown in Fig. 4, control is performed in a discrete time system with a certain fixed time interval (Tc in the figure) as the control period. Correspondingly, external observation inputs required by the control side are also taken in at regular time intervals.

This take-in signal corresponds to the measurement cycle signal 210 that is input to the pulse generation detection circuit 42. This is a timing signal issued by the control side that requires detection speed (
In other words, it is captured based on this periodic signal).

Second Embodiment FIG. 6 shows a preferred second embodiment of the speed detection device according to the present invention.
Examples are shown.

The features of this embodiment include, in addition to the configuration of the first embodiment,
Pulse counting circuit 80. This is because a high speed range speed calculation circuit 821 and a selection circuit 84 are provided.

The pulse counting circuit 80 counts the multiplied pulses 200 output by the multiplied pulse generation circuit 34, and outputs the count value to the high speed range speed calculation circuit 82 and the selection circuit 84.

The high speed range speed calculation circuit 82 of the embodiment includes a speed calculation section 82a.
and a rotation information adding section 82b.

Then, the speed calculation unit 82 calculates the rotational speed of the rotating body based on the manually inputted multiplied pulse count number, and inputs the calculated rotational speed to the rotation information addition unit 82 . Then, the rotation information addition unit 82b adds the rotation direction revo I input manually from the rotation direction detection circuit 32 to the speed input manually in this way, and obtains the speed in the high speed range.

The speed thus determined is then output to the selection circuit 84.

The selection circuit 84 receives the detected speed in the low speed range output from the low speed range speed calculation circuit 38 and the speed in the high speed range output from the high speed range speed calculation circuit 82. Based on the count value output from the pulse counting circuit 80, this selection circuit 84 determines whether the rotation speed of the rotating body is in a high speed range or a low speed range, and in the low speed range, the low speed range speed calculation circuit 38 In the high speed range, the speed output from the high speed range speed calculation circuit 82 is selectively output.

As a result, according to this embodiment, it is possible to perform accurate speed detection in a wide range from low speed ranges to high speed ranges.

Note that the present invention is not limited to the above embodiments,
Various modifications are possible within the scope of the invention.

For example, in the 14th embodiment, an optical incremental encoder is used as the pulse generating means. However, the present invention is not limited to this. The encoder used herein and various other pulse generating means can be used.

[Brief explanation of drawings]

1 to 5 show a first preferred embodiment of the speed detection device according to the present invention, FIG. 1 is a block circuit diagram of the speed detection device of the embodiment, and FIG. 2 is a block circuit diagram of the speed detection device of the embodiment. , a block circuit diagram showing a specific configuration of the multiplied pulse information observation circuit of the circuit shown in FIG. FIG. 4 is an explanatory diagram showing a comparative example of accuracy. FIG. 4 is an explanatory diagram of a comparative example of speed detection accuracy between the present invention and the conventional device when the speed changes in a sinusoidal manner in a low speed range. FIG. FIG. 6 is a block circuit diagram of a second preferred embodiment of the speed detection device according to the present invention; FIG. FIG. 1 is an explanatory diagram showing an example of an optical encoder used as the pulse generator, FIG. 8 is a timing chart of the basic pulse and multiplied pulse outputted in each part of the circuit in FIG. 1, and FIG. , an explanatory diagram showing the relationship between the multiplied pulse and the observed pulse interval, FIG. 10 shows the fluctuation of the analog waveform output from the light receiving element when the encoder shown in FIG. 7 is used to output the basic pulses 100A and 100B. , and an explanatory diagram of the cause of duty error caused by variations in the threshold value for obtaining the fundamental pulse from this analog waveform. Figure 12 is an explanatory diagram showing the relationship between the encoder rotation angle and time. Figures 13 and 14 are the period conditions of the encoder waveform and the pulse phase angle preservation conditions. The explanatory diagram, FIG. 15, is a timing chart of a measurement cycle pulse signal in control using a microcomputer. 30... Pulse generator, 32... Rotation direction detection circuit, 34... Multiplication pulse generation circuit, 36... Multiplication pulse information observation circuit, 38... Low speed region speed calculation circuit, 40... Free-run timer, 50-58...Register, 60-68...Switch.

Claims (3)

[Claims] (1) Pulse generating means that generates A-phase and B-phase basic pulses having phases different by 90 degrees each time the rotational position of the rotating body changes by a predetermined single position; A multiplied pulse generating means for generating a total of four multiplied pulses synchronized with the rising and falling edges of the basic pulse; and a rotational direction detecting means for detecting the rotational direction of the rotating body based on the A-phase and B-phase basic pulses. , a multiplied pulse information observation means for measuring the elapsed time since the final multiplied pulse output of the multiplied pulse generating means as elapsed time information, and sequentially updating and storing pulse interval information of the last five multiplied pulses; and the rotation direction. The rotational direction information of the rotating body outputted from the detection means, the elapsed time information since the last multiplied pulse generation and the last five pulse interval information outputted from the multiplied pulse information observation means are used to calculate the acceleration in the low speed range. low speed region speed calculating means for calculating and outputting a speed in a low speed region based on the assumption that the speed is constant, a periodic condition of the waveform in the pulse generating means, and a predetermined speed calculating formula derived from the pulse phase angle preservation condition; What is claimed is: 1. A speed detection device comprising: and, characterized in that the speed detection device has good responsiveness in a low speed rotation range of a rotating body and performs speed detection that compensates for a duty error of a multiplied pulse signal. (2) In claim (1), the multiplied pulse information observation means measures the elapsed time since the final multiplied pulse is output from the multiplied pulse generation means, outputs it as elapsed time information, and also measures the elapsed time after the final multiplied pulse is outputted from the multiplied pulse generation means, and outputs the elapsed time information as elapsed time information. and a storage means for sequentially updating and storing the latest five pulse interval information sequentially output from the timer means. A speed detection device featuring: (3) In either of claims (1) and (2), the low speed range speed calculation means calculates the latest five pulse interval information from the newest to the past, t[0], t[1], t[2], t[3], and t[4], and the elapsed time from the generation of the final multiplied pulse is expressed as t=freerun, and the most recent multiplied pulse t
The rotation direction corresponding to is expressed as revol[0], the basic number of pulses per rotation of the rotating body is expressed as N, and based on the speed calculation formula shown in the following equation, V_■=(2π/4N)・revol[0]・{(t[
4]-t[0]・freerun+p[0]^2+2t
[4]・p[0]−t[4]^2}/{p[0]・p[
1]・(t[0]+t[4])} However, p[0]=t[
0]+t[1]+t[2]+t[3]p[1]=t[1]
]+t[2]+t[3]+t[4] A speed detection device that calculates a speed V_■ of a rotating body in a low speed range.