CN1173151C - Grid Displacement Sensor - Google Patents

Abstract

A grating type displacement sensor, including a moving ruler and a fixed ruler, the fixed and moving rulers are designed to have unequal number of grid lines in the range of 0-360 degrees or within the same length range, and the difference between the number of grid lines is used Realize a substantial increase in the number of equivalent grid lines, and the number of equivalent grid lines λ satisfies the relation: λ=NN'/n, where N is the number of moving-scale grid lines, N' is the number of fixed-scale grid lines, and N'>N , n is the greatest common divisor of N and N ', and the present invention also utilizes the relationship between the difference between the moving and fixed-scale grating lines to realize each periodic error other than ni (i=1, 2, ...) times The elimination of , so that the accuracy is greatly improved. The present invention changes the current situation that one pulse is generated by one grid line in the prior art, and can obtain higher resolution with fewer grid lines for precision measurement of displacement. The designed differential grid displacement sensor is easy to process, simple in structure and low in cost. Inexpensive, strong anti-interference.

Description

Grid Displacement Sensor

technical field

The invention belongs to the field of precision measurement of geometric quantities, and relates to a displacement precision measurement device.

Background technique

Commonly used grating displacement sensors often use the regular periodic changes of a certain physical quantity during its movement to form "grid lines" that are evenly distributed along the space, so that the displacement can be obtained by counting the grid lines. For example, the grating uses the change of light and shade of the Moiré fringe, the magnetic grating uses the recorded NS pole magnetic field change, the toothed grating uses the top tooth groove to cause the magnetic flux change of the probe, and the capacitive grid uses the change of the pole plate area to cause the periodic change of the capacitance, etc. The existing grating type displacement sensor has the same density of moving scale and fixed scale, or the same number of grating lines within the same distance, limited by the limited density of grating lines in space, and limited resolution, which cannot meet the requirements of engineering measurement. A common problem of grid sensors is that the number of spatial grid lines is difficult to further increase, so the original signal must be subdivided electronically to meet the resolution requirements of engineering applications, which will cause complex systems, high costs, and the possibility of failure Increase, a series of disadvantages such as more stringent requirements for uniformity of motion.

Contents of the invention

The purpose of the present invention is to change the current situation that one grid line generates one pulse in the prior art, and can obtain higher resolution with fewer grid lines. Amplification of the number of equivalent grid lines, that is, a new sensor that uses fewer lines to achieve more "equivalent grid lines" - a differential wire grid displacement sensor.

To achieve the above object, the technical solution of the present invention is as follows,

A grating type displacement sensor, including a moving ruler and a fixed ruler, the fixed and moving rulers are designed to have unequal number of grid lines in the range of 0-360 degrees or within the same length range, and the difference between the number of grid lines is used To achieve a substantial increase in the number of equivalent grid lines, the number of equivalent grid lines λ satisfies the relational expression: $\mathrm{\λ}={\mathrm{NN}}^{\′}/\mathrm{no},$ Among them, N is the number of moving-scale grid lines, N' is the number of fixed-scale grid lines, N'>N, and n is the greatest common divisor of N and N'.

In the grid type displacement sensor, the uniformly distributed signals of n channels are superimposed, which can eliminate all periodic errors except ni (i=1, 2, . . . ) times.

The grid type displacement sensor is characterized in that the selection of the grid line width duty ratio should be based on the following conclusions:

(1) When S _N =0.5, the amplitude A of Y _(x) is equal to the difference between W _N' and its closest special grid width kG (k=0,1,...,q) and the greatest common divisor n product, ie

A=|W _N' -Gk|·n(k=0, 1,...,q) (1)

(2) When the fixed-length grid line width is the special grid line width $W_{{\mathrm{KN}}^{\′}}=\frac{G}{2}(2k-1)$ (k=1, 2, ..., q) and S _N =0.5, the shape of the two-foot coincidence angle change curve Y _(x) is a triangular wave, and the amplitude A reaches the maximum, and

$\mathrm{<span\; class="notranslate">\; A</span>}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; G</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 2</span>}<span\; class="notranslate">\; )</span>$

Among them, N-the number of moving scale grid lines N'-the number of fixed-scale grid lines, and set N'>N n-the greatest common divisor of N and N'

W _N - Duty Angle of Line Width of Moving Scale W _N′ - Duty Angle of Line Width of Fixed Scale

S-duty cycle, S ∈ [0, 1]

λ-the number of equivalent grid lines after amplification G-resolution,

    Y- total coincidence angle of two feet

Y _(x) - coincidence angle change curve (wave change curve)

A-wave curve amplitude (peak-to-peak value), A=Y _max -Y _min

The number of special values of qW _N′ is related to the value, $q=\frac{W_{N^{\&prime;}\max }}{G}=\frac{N}{\mathrm{no}}$

Take the rack as an example to further illustrate the technical solution of the present invention:

1 Increase the number of equivalent grid lines by using the grid line difference

The gear grid uses the total overlapping area of the tooth tops of the inner and outer coaxial gears as a variable, which can cause regular changes in capacitance, inductance, etc. to form a grid sensor.

As shown in FIG. 1 , suppose there is a pair of internal teeth 1-2 and external teeth 1-1 opposite to each other, the number of internal and external teeth is 100, and the width of the tooth tip is the same as the width of the tooth space, each of which is 1/2 pitch. When there is relative motion (unified as fixed external teeth and rotating internal teeth), it is obvious that the total overlapping area of the two gears is the largest when the internal and external teeth are aligned or completely overlapped. As the internal and external teeth are gradually staggered, this area gradually decreases. When the top and the alveolar face each other, the overlapping area is zero, and then gradually increases to form a cycle. The movement changes 100 times in one week, so 100 grid lines are formed. If the internal and external teeth are regarded as two plates, this is the principle of the capacitive grid sensor. The curve of the capacitance value C changing with the angle  is shown in Figure 2, and W is the grid pitch angle, where W=360°/100= 36°.

And as shown in Fig. 3, we design the number of teeth of the outer teeth 3-1 to be 101, and the number of teeth of the inner teeth 3-2 is still 100. Assuming that at a certain moment, only tooth 1 and tooth 1' are coincident and aligned (meaning that the positions of the duty angles are as close as possible), all other teeth cannot be completely coincident, but are staggered by 1/100, 2/100, ..., 99/100 outer tooth pitch. When the inner tooth rotates 1/101 inner tooth pitch, only tooth 2 is aligned with tooth 2'; and then rotated 1/101 inner tooth pitch, only tooth 3 is aligned with tooth 3'... 101 rotations in this way are 1 inner tooth After the pitch, tooth 1 and tooth 2 are aligned to complete a small cycle, and after 100 small cycles, that is, the internal teeth rotate 100 internal pitches, that is, one week, and a large cycle is completed. Therefore, in this week, 100×10 1 =10100 aligned states occur in total. If the inner and outer gears are regarded as two pole plates or wound into two coils, the capacitance or inductance between the teeth may undergo 10,100 periodic changes similar to those in Figure 2 within one round of motion, thus forming an enlarged The number of grid lines (the reason why it is said "possible" is that although the overlapping state of a single tooth has changed, in some cases the total overlapping area of the two gears will remain unchanged, which is the proportion of the teeth pre-tooth will be discussed in detail later).

Comparing Fig. 1 and Fig. 3, that is, comparing ordinary reticles and difference-type reticles, it can be seen that: under the same mechanical reticulation density, the latter obtains a much larger number of equivalent grid lines than the former.

2 Use different grid lines to improve accuracy

It is not difficult to see that the periodic change of the coincidence area in Figure 1 occurs evenly in 100 teeth within one week, which has an average effect and is of obvious benefit to improving the accuracy. However, the periodic change of the overlapping area in Fig. 3 occurs on one tooth each time, and the specific position of this tooth changes tooth by tooth against the rotation direction of the internal gear. Because there is no average effect, the mechanical manufacturing error will be fully reflected, and the accuracy is poor. The remedy is to adjust the number of teeth.

For example, if the number of internal teeth is 4 and the number of external teeth is 6, then two teeth will be aligned at the same time, as shown in Figure 4, and the difference between the two teeth is 180°, and the specific positions of the two teeth are opposite to each other. The rotation direction of the internal gear changes tooth by tooth, but always maintains a mutual difference of 180°. In this case, the periodic change of the overlapping area will occur 4×6/2 times in one week (once every 30°), that is, the number of equivalent grid lines is 12 lines/cycle.

For another example, if the number of internal teeth is 8 and the number of external teeth is 12, then 4 teeth will be aligned at the same time at the same time, and the 4 teeth are 90° apart from each other. The specific positions of these 4 teeth are against the rotation direction of the internal gear. The teeth are gradually changed, but always maintain a mutual difference of 90°. In this case, the periodic change of the overlapping area will occur 8×12/4 times in one week, that is, the number of equivalent grid lines is 24 lines/cycle.

From the above two examples, it can be seen that the proper selection of the number of teeth can obtain the effect of multi-reading head diameter readings, thus playing a very good role in suppressing and offsetting geometric errors such as manufacturing errors and installation eccentricity.

3 Calculation method of grid line width duty cycle

In the actual design, it can be found that the differential gate has many unique rules. For example, the choice of the number of teeth and the difference between the number of teeth can determine whether or not the teeth are aligned at the same time or how many teeth are aligned at the same time, which has an important impact on the accuracy and the number of equivalent grid lines; The size and change law, which have an important impact on the waveform of the induced electrical signal and so on. Its detailed derivation is as follows:

i) Definition

N-number of internal teeth N'-number of external teeth, and set N'>N n-the greatest common divisor of N and N'

W _N - Duty angle of internal tooth width W _N′ - Duty angle of external tooth width

S-duty cycle, S ∈ [0, 1]

λ-the number of equivalent grid lines after amplification G-resolution,

Y-total coincidence angle of two gears, use coincidence angle instead of the above coincidence area to represent the periodic change of physical quantity.

Y _(x) - coincidence angle change curve (wave change curve)

y _i - coincidence angle of the ith pair of gears x - relative angular displacement of the two gears

A-wave curve amplitude (peak-to-peak value), A=Y _max -Y _min

The number of special values of qW _N′ is related to the value, $q=\frac{W_{N^{\&prime;}\max }}{G}=\frac{N}{\mathrm{no}}$

W _K'N' - special tooth width value

ii) Conclusion

次。(1) When the greatest common divisor of N and N' is n, the number of equivalent grid lines $\mathrm{\&lambda;}={\mathrm{NN}}^{\&prime;}/\mathrm{no},$ That is, the number of periodic changes of Y _(x) in one week is

Second-rate.

(2) The greatest common divisor of N and N' is n, then there are n pairs of teeth in the same overlapping state during the rotation process, and the positions of the n pairs of teeth change gradually with the rotation (forward rotation direction or reverse rotation direction), but Always maintain a geometrically symmetrical position with a mutual difference of 360°/n.

This conclusion is very important. If n=2, it is equivalent to the radial readings of two reading heads, which can eliminate all periodic errors except for 2i (i=1, 2,...) times; if n=4, it is equivalent to For the radial readings of 4 reading heads, it can eliminate all periodic errors except 4i times; the uniformly distributed n-channel signal superposition can eliminate all periodic errors except ni times ^[2] , so it can be used for manufacturing And installation errors cause a good offset. The disadvantageous effect is that the larger n is, the smaller the number of equivalent grid lines λ is.

(3) When the external tooth width is a special tooth width $W_{{\mathrm{KN}}^{\&prime;}}=\frac{G}{2}(2k-1)$ (k=1, 2, ..., q) and S _N =0.5, the shape of the curve Y _(x) of the coincidence angle of the two gears is a triangular wave, and the amplitude A reaches the maximum, and

$\mathrm{<span\; class="notranslate">\; A</span>}<span\; class="notranslate">\; =</span>{\mathrm{<span\; class="notranslate">\; Y</span>}}_{\mathrm{<span\; class="notranslate">\; max</span>}}<span\; class="notranslate">\; -</span>{\mathrm{<span\; class="notranslate">\; Y</span>}}_{\mathrm{<span\; class="notranslate">\; min</span>}}<span\; class="notranslate">\; =</span>\frac{\mathrm{<span\; class="notranslate">\; G</span>}}{\mathrm{<span\; class="notranslate">\; 2</span>}}\mathrm{<span\; class="notranslate">\; no</span>}<span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; -</span><span\; class="notranslate">\; (</span>\mathrm{<span\; class="notranslate">\; 1</span>}<span\; class="notranslate">\; )</span>$

From the perspective of signal pickup, we hope that the larger the A, the better, so that a larger capacitance or inductance change can be obtained, and thus a stronger electrical signal can be obtained.

(4) When the external tooth width $W_{N^{\&prime;}}\&NotEqual;\frac{G}{2}(2k-1)$ (k=1,2,...,q) and S _N =0.5, the amplitude A of Y _(x) is equal to the difference between W _N' and its closest special tooth width kG (k=0,1,...,q) The product of the value and the greatest common divisor n, that is

A=|W _N' -Gk|·n(k=0, 1, ..., q) (3) and as the difference between W _N' and its closest special tooth width value W _kN' becomes smaller and smaller, coincidence The angle change curve will gradually change from a pulse wave to a triangular wave, and the amplitude A becomes larger and larger.

The above conclusions provide a theoretical basis for the selection of the number of teeth and tooth width in the design of differential wire grid sensors. In the actual design, the number of teeth should be reasonably selected taking into account the number of equivalent grating lines and the radial reading of multiple reading heads. After the number of teeth is determined, the tooth width that meets the conclusion (3) should be selected as much as possible.

The above description takes angle measurement as an example. If it is a straight line measurement, the marking method of the differential grid is similar to that of a vernier caliper, but the vernier caliper needs to be aligned with the naked eye, and the differential grid can be counted and displayed by a counter by sending out more pulses. Therefore, the above differential wire grid principle is fully applicable to the measurement of linear displacement.

The invention uses fewer grid lines to obtain higher resolution to carry out displacement precision measurement, and the differential wire grid displacement sensor designed by the invention is easy to process, simple in structure, low in cost and strong in interference resistance.

Description of drawings

Figure 1: Gear grid sensor, 1-1-outer teeth 1-2-inner teeth;

Figure 2: Capacitive sensor capacitance variation curve, W-grid pitch angle;

Figure 3: Differential wire grid displacement sensor, 3-1-external teeth 3-2-internal teeth;

Figure 4: Example of simultaneous alignment of 2 teeth;

Figure 5: Differential wire grid angular displacement sensor, 5-1-fixed gear 5-2-moving gear.

Figure 6: Differential wire grid linear displacement sensor, 6-1-fixed ruler 6-2-moving ruler.

Detailed ways

As shown in Figure 5, the number of teeth of fixed gear 5-1 is designed to be 100, and the number of teeth of movable gear 5-2 is 96, then n=4, the number of equivalent grid lines λ=100×96/4=2400 lines/circle, and the diameter 4 reading head readings, eliminating all periodic errors except for 4i (i=1, 2, ...) times. According to the grid line width duty cycle calculation method, select the appropriate tooth tip width of the moving and fixed gears, and take the moving and fixed gears as two plates or wind them into two coils, then the capacitance or inductance between the teeth is Periodic changes similar to those in Figure 2 occur 2,400 times within one cycle of movement to form an enlarged number of grid lines. The differential wire grid displacement sensor designed according to this method eliminates all periodic errors except for 4i times, and the resolution G=0.15°, that is, the ordinary grid displacement sensor needs to use 2400 with no more than 100 grid lines The resolution that can only be achieved by raster lines.

As shown in Figure 6, the number of teeth of the fixed scale 6-1 is designed to be 104, and the number of teeth of the moving scale 6-2 is 96, then n=8, the number of equivalent grating lines λ=104×96/8=1248 lines/L, and the diameter 8 reading head readings, eliminating all periodic errors except 8i (i=1, 2, ...) times. Select the appropriate inner and outer tooth tip widths according to the calculation method of grid line width duty ratio, and take the inner and outer gears as two plates or wind them into two coils, then the capacitance or inductance between the teeth will move in one circle There are 1248 periodic changes similar to those in Figure 2 to form an enlarged number of grid lines. The differential wire grid displacement sensor designed according to this method eliminates all periodic errors except for 8i times, and uses no more than 104 grid lines to achieve the resolution that ordinary grid displacement sensors need to use 1248 grid lines to achieve. Rate.

Claims (3)

1. grid formula displacement transducer, comprise scale and moving chi, it is characterized in that: fixed, moving chi in 0-360 degree one circle scope or the grid line number in the equal length scope unequal, the difference of their grid line number has realized increasing considerably of equivalent grating line number, equivalent grating line number λ satisfies relational expression: $\mathrm{\&lambda;}={\mathrm{NN}}^{\&prime;}/n,$ N is moving chi grid line number, and N ' is a scale grid line number, and N '＞N, n are the highest common factor of N and N '.

2. the described grid formula of claim 1 displacement transducer is characterized in that with uniform n road signal stack, eliminate remove ni (i=1,2 ...) inferior each second periodicity error in addition.

3. according to the described grid formula of claim 1 displacement transducer, it is characterized in that choosing of the wide dutycycle of grid line should be according to following results:

(1) works as S _N=0.5 o'clock, Y _(x)Amplitude A equal W _{N '}Wide kG of immediate special grid line with it (k=0,1 ..., the product of difference q) and highest common factor n, promptly

A＝|W _N′-Gk|·n(k＝0，1，…，q) (1)

(2) be that special grid line is wide when the scale grid line is wide $W_{{\mathrm{KN}}^{\&prime;}}=\frac{G}{2}(2k-1)$ (k=1,2 ..., q) and S _N=0.5 o'clock, two chi overlap angle change curve Y _(x)Be shaped as triangular wave, it is maximum that amplitude A reaches, and

$A=\frac{G}{2}n---\left(2\right)$

Wherein the moving chi grid line of N-is counted N '-scale grid line number, and establishes the highest common factor of N '＞Nn-N and N '

W _N-moving wide the duty of chi grid line angle W _{N '}The wide duty of-scale grid line angle

The S-dutycycle, S ∈ [0,1]

Equivalent grating line number G-resolving power after λ-amplification,

The total overlap angle of Y-two chis

Y _(x)-overlap angle change curve (ripple varied curve)

A-ripple varied curve amplitude (peak-to-peak value), A=Y _Max-Y _Min

Q-W _NParticular value number correlation, $q=\frac{W_{N^{\&prime;}\max }}{G}=\frac{N}{n}$

Images