SU1311024A1 - Angular displacement-to-digital converter - Google Patents

Abstract

The invention relates to automation and measurement technology and can be used in digital automatic control systems, in particular in computer numerical control systems for the movement of machine mechanisms. In order to increase the accuracy of converting and expanding the field of application due to the possibility of changing the output code size into a tracking converter of the following type, an adder 3, a switch 5, a high-pass filter 6, a controlled inverter 7, a switch 10, reversible counters 13, 4, block 15 control and synchronization unit I6. Since the control of the mismatch sign occurs after it is formed, the instrumental depth excludes the analog keys, which are inevitably present in the block diagram with the quad selector. The presence of separate reversible counters 13 and 14 for the formation of the functions of quasi-cone and quasi-cosine allows you to eliminate the bogs associated with the inequality of full and random complements of codes when implementing triangular functions. The ability to preset the N / Z code as well as digital correction of the systematic error of the sensor provides high accuracy of conversion and allows you to select the required output code width. 2 hp f-ly, 4 ill. c (L SL9

Description

eleven

The invention relates to automation and measurement technology and can be used in digital automatic control systems, in particular in systems of numerical program control. By the movement of machine mechanisms.

The purpose of the invention is. improving the accuracy of conversion and ensuring that the output code can be changed.

Improvement of accuracy is achieved by compensating the first harmonic component of the intra-step error of the sensor, reducing the error in generating a mismatch signal and eliminating the systematic error of the conversion. The change in the output code size is provided by building a converter with adjustable discreteness, which allows the use of the converter in systems with different information code lengths. ,

FIG. 1 shows a block diagram of an angular displacement transducer into a code; in fig. 2 - diagram of the converter; in fig. 3 is a block diagram of the correction and control; in fig. 4 is a diagram of a synchronization unit;

The angular displacement transducer to the code contains multipliers 1, 2, adder 3, subtractor 4, switch 5, high-frequency filter 6, controlled inverter 7-, integrator 8, voltage-frequency converter 9, switch 10, elements OR 11, 12, reversible counters 13, 14 correction and control unit 15, synchronization unit 16 5 element 17 EXCLUSIVE OR, power supply unit 18, reversible counter 19, quadrature signal sensor.

The correction and control unit 15 comprises a switch 21, a reversible-counter 22, an element 23 EXCLUSIVE. OR, element 24 OR, read only memory 25.

The synchronization unit 16 contains elements 26-29 AND-NOT,. Inverters 30, 31, triggers 32, 33, distributor

34 ticks, element 35ШШ.

The Converter operates as follows.

Movement sensor 20 (FIG. 1) of induction type, for example, resolver or inductosin included in am 10242

bulk mode (transformer) mode, powered by single phase voltage

. ,

5 where and 1x) o is the amplitude and frequency of the supply voltage. From the two windings of sensor 20 displaced in space by 90, the output voltages Ug, are removed, which to are supplied to the analog inputs of multipliers 1 and 2, respectively. The stresses Ug, Uc are determined from the following expressions:

15

U5 K UoSinp0, U K-rUoCOSp0,

where K is the transformation ratio

sensor 20;

p is the coefficient of electrical reduction of the sensor 20; the number of steps (periods) of the sine or cosine function contained in one revolution of the sensor

Q is the angle of rotation of the sensor 20.

The converter is constructed according to the compensation principle. The signal of the ostib is formed in the form of the sine of the difference of two angles:

s1p (p0-f),

where P is the output angle of the transducer. Angle 9 of the rotation of the sensor 20

It is used as a master transducer. Adjustable (output) values of the converter are unipolar binary n-bit codes N, N and code number

quadrant, formed in reversible counters 13, 14 and in block 15 correction and control, respectively.

The digit of counters 13, 14 determines the decisiveness (discreteness) of the conversion, which is equal to

.

where Ksch is full digital converter scale.

The value of LF determines the price of the unit of the least significant digit of the counters 13, 14. When using

multi-pole sensor (p 1) the resolution of the converter (with constant conversion hardware) increases by a factor of p and the discrete price; is the value

.. pNu

Codes N, N are equivalent to each other, carry the same information about the magnitude of the output angle of the transducer within the quadrant of the sensor. They perform in the converter the role of digital arguments of the functions of the quasi-sinus and quasicosine-sine of the output angle Φ, which serve to form the error signal. The values of the digital arguments N and N vary within the same quadrant of the sensor. 3 g-olT75 also get I, oiIcho

to. correspond in each square to 5 values of the function f (sin) and f (cos), that

,: f -. f (sin)

- function f (tp)

Run two output angles that complement each other up to / 2. For example, in the first quadrant, code N corresponds to the angle φ, and code N corresponds to the angle (F / 2-Φ).

Binary codes N and N complement each other 20 to the value Nj ..:

, (2)

where Z is the quadrant number of the sensor;

N2 is the discrete number df dividing quad-25 gauge of the sensor with the number z.

Ideally, K2 No const. Then

their attitude

will approximate the tangent function with a high degree of accuracy equal to several minutes.

Expressions (3) for the approximating functions f (sin) and f (cos) are exactly the same and differ only in the argument - the N and N codes, i.e. as well as the arguments of the approximable functions sin Ч and 11n (I / 2-F) differ.

From functions zhffi sin (-) it is visible,

s m i

The codes N and M vary according to the laws of symmetric linear triangular functions (Fig. 2) of the output angle F. With their help, you can (roughly) approximate the trigonometric functions / sin f / x / cos Φ /, respectively.

As is known, the accuracy of the servo converter with a sensor included in the amplitude mode does not depend on the accuracy of the approximation of the sine and cosine functions, but is determined by the accuracy of the approximation of the tangent function. When approximating the function of sine and cosine by linear functions N and N, the accuracy of the realization of the tangent function determined by the ratio of the functions N and N is small. The accuracy of the approximation of the tangent function increases if the sine and cosine functions are approximated by nonlinear rational positive functions of the form,

f (sin)

BUT -,

 five

f (soi)

l N 2

1 + B 7

2

, (3

where A and B are constant coefficients ,.

The optimal choice of the coefficient B provides approximating functions f (sin), f (cos), which is close to sinusoidal. The coefficient A determines the amplitude of the approximating functions.

The functions f (sin) and f (cos) are the functions of quasi-sinus and quasicosine-sine, taken modulo, and they more accurately approximate the functions (hfn) and / cos f /, respectively. When choosing the optimal value of the coefficient 1

3 g-olT75 also get I, oiIcho

 the value of the function f (sin) and f (cos), that

their attitude

0

25

thirty

will approximate the tangent function with a high degree of accuracy equal to several minutes.

Expressions (3) for the approximating functions f (sin) and f (cos) are exactly the same and differ only in the argument - the N and N codes, i.e. as well as the arguments of the approximable functions sin Ч and 11n (I / 2-F) differ.

From functions zhffi sin (-) it is visible,

that their arguments complement each other

35

40

ha up to the value. A similar requirement must be presented to the numeric arguments for codes N and N, i.e. equality (2) is necessary to ensure high instrumental accuracy of the conversion.

In the prototype, equality (2) is satisfied with an error equal to 1 MLR, since the forward N and the inverse N codes are used. Therefore, using expressions similar to (3):

50

)

55

f (sin)

. N 2

A n

2

N

1- -in- |

-; f (cos) 1 + B

the functions sin Φ and sin - - (Φ + DF) are approximated | cos (f + df), respectively. Errors in the approximation of the sine and cosine functions in comparison with the approximation in expressions (3) are not present here, but there is an error due to the fact that the approximating functions are functions of different arguments. As a result, ap5

the proxy function f (tg) is realized — with an error.

Codes N and N are fed to the digital inputs of multipliers 1, 2,

Two bipolar signals U and and are taken from the two outputs of the multipliers 1, 2

- 2.

and K ySinp 0-f (cos) ,, ,, cosp 9 f (sin),

where Ku is the transfer coefficient of the multipliers 1, 2.

The error signal is obtained from the inverters and and and with the help of successively connected blocks 3, 5, 7 or blocks 4-7. The choice of the required chain of passage of signals U and U2 is carried out according to the number of the quadrant. At the output of block 7, the error signal is generated as a function

sin (p9 - () sinp9-f (cos) K (j + cosp6 xf (sin) Ky,

where K. D .. are the signs of the transfer coefficients of the converting circuits, respectively, from multipliers 1, 2 to the output of the phase inverter 7. Depending on the number of quadrants K and K, they take the values +1 or -1, forming the necessary signs. By approximating functions f ( cos) and f (sin), respectively.

In tab. 1 shows the required signs of Ku, Ku, providing the work of the converter in four quadrants

four quadrant Tbl and c

The adder 3 generates a signal equal to the sum of the signals 11 and U, + Uj,

and subtractor 4 is a signal UjZ equal to the difference of these signals:

-, -U ..

five

0

Each of the two signals carries information about the magnitude of the mismatch in only two quadrants of the sensor. The values of the signals 221 of these quadrants are proportional to the mismatch sip () function. For control, signals 11 are used alternately, depending on the quadrant number: signal j ° and 4th quadrants, and signal in the 1st and 3rd quadrants. In other quadrants, these signals do not carry information about the magnitude of the error and are not used to form the error signal.

Since each of the signals 11 and ILy is proportional to the mismatch function only in a certain quadrant, the choice of the desired signal, from the two available ones, can be done by the current quadrant number, which is implemented using switch 5, which is switched when the quadrant is changed. filter 6 high frequencies from the output of the adder 3 to the output of the subtractor 4 or vice versa - from the output of the subtractor 4 to the output of the adder 3.

Thus, the switch 5 is located in the converter directly behind the adder 3 and the subtractor 4, therefore the parameters in its composition of keys, which are characterized by temperature instability and inconsistency of the passage resistances in the open state, are excluded. from the summation scheme, i.e. not available; Their effect on the transfer coefficients of the adder 3 and subtractor 4 is important. This feature provides increased accuracy and temperature stability in the formation of the error signal.

The construction of a precision converter is impossible without filtering the constants of the useful signal, which are the bias voltages and the zero drift of the operational amplifiers of the multiplier 1, 2, su -1Mator 3 and the subtractor. 4. This function in the converter performs a high-frequency filter 6. Block 6 also serves to suppress low-frequency imag, for example, a network of 5 pickups with a frequency of 50 Hz,

From the output of block 6, the filtered variable sig0 is removed and fed to the input of the controlled inverter 7

five

0

tal with frequency, Hecynteft u) d- Its value is proportional to the mismatch.

Block 7 simultaneously performs two functions: the function of a phase-sensitive rectifier and the function of an inverter. It completes the formation of the error signal by providing the required sign Kc and K with the help of the transmission sign Kf of block 7 (j in Table 1. This operation is accomplished by inverting the input signal of block 7 in accordance with the sign Кf у, which changes in phase with the change the sign of the cos function (p. Simultaneously, the block 7 demodulates the carrier of the error signal, inverting the input signal depending on the polarity Up of the sensor supply voltage. 20.

Thus, the signs; Q and K are formed using adder 3, subtractor 4, switch 5 and controlled inverter 7. If we designate the transfer coefficients of the adder 3 for the signals U and U2 as responsible, and the transfer coefficients of the subtractor 4 for the same signals as Kg and Kg., then the principle of formation of the sign coefficients Ku and Ku can be illustrated using the table. 2

table 2

On

According to the data given in table 2, it is possible to obtain the values of K and Ku, which are indicated in table. 1. The calculation is performed on the quadrant number using the following expressions:

In the 1st and 3rd quadrants

e b -

JO

f5

thirty

55

In the 2nd and 4th quadrants,

The corrective link, which contains the integrator 8, forms the required law of regulation, for example, the integral proportional-integral integra- tion.

The output of block 8 is connected to the input of a voltage-frequency converter 9, at one of the outputs of which, depending on the size of the input signal, a sequence of frequency pulses is formed

The signal from the output of block 9 is fed to one of the outputs of switch 10. Switch 10 serves to generate N and N codes, which are symmetrical triangular

20 functions of the output angle.

The reversing counter J9 accumulates the pulses of the voltage-frequency converter 9, as a result of which in block 19 the output code is generated

25 converter.

In order to obtain the triangular law of changing the output code, it is necessary, when a certain value of the output code is reached, to switch the counting direction from summation to I subtraction, i.e. switch input pulses from one counting input to another.

Similarly, by switching the counting direction in the counters 13, 14 at the moments of changing the quadrant number, Tpeyronbnyk can provide the output codes N and N. The amplitude of the triangular functions N, N can be set using the N code of the initial conditions of the quadrant. The N code is recorded when the quadrant is changed into a reversible counter that will work for subtraction. Its value determines the instant of the next quadrant shift.

The function of switching the counting direction in blocks 13, 14 is performed by the switch 10, which is controlled by the signal a. The logic function is Q is the low-order bit of the current quadrant code and therefore changes its value when the quadrant is changed, thereby switching the pulse sequence of the frequency fjj from one output of the switch 10 to another output.

Two outputs of the switch 10 are connected to the counting inputs of blocks 13 and 14 through the elements OR 11 and 12, respectively. An item change OR occurs

u131

when changing the direction of rotation of the rotor of the sensor 20 (changing the sign of the speed (jL) of the driving force) or changing the number of the quadrant.

Tab. 3 illustrates the dependence of the counting direction in blocks 13, 1A on the sign of the velocity w and on the quadrant number.

Table 3

The signs + and - in the table, 3 indicate the counting direction of the pulses in blocks 13 and 14: the summation mode and the pulse subtraction mode, respectively.

The change in the counting direction of the pulses in blocks 13 and 14 is carried out as follows.

When the sign of the velocity changes and) the sign of the input signal of block 9 changes and the frequency pulses switch from one of its output to another, i.e. the switching of the inputs of block 10 occurs, which changes the counting direction in blocks 13 and 14, thus, the counting direction inside the quadrant changes for any value of the output angle F.

When the quadrant number is shifted, the logical variable a is changed, which controls the switch 10, and the frequency pulses are switched from one output of the switch 10 to another output and, therefore, 5 from one OR element to another.

The quadrant change signal in the transducer is negative transfer pulses P and P, generated in blocks 13 and 14, respectively. Negative transfer impulse is generated by a reversible subtracting counter in the engine.

The transition point from code 00 ... 00 to code 11.

0

Ranks a,

2n

The quadrant number code controls the operation of the converter: the bit controls the switch 5 and the switch BO, and the sum of the bits a and a, modulo two and equal to the logic function, is used to form the signal d controlling the unit 7. Logic Function d is formed by the element 17 EXCLUSIVE OR, and is determined from the reflection

J5

P

0

five

0

0

five

0

where arp the logical signal obtained from the sensor supply voltage, takes the value I or O depending on the sign of the voltage U. The correction and control unit 15 also generates codes N of initial conditions of quadrants, which, in general, may be different for different sensor quadrants. To store the codes Nj, the block 15 contains a permanent storage device 25. With the help of the N codes, the resolution of the df transform can be changed. In this case, not only the amplitude of the triangular functions N and N is changed, but also the law of their behavior, which repeats the asymmetry of the sensor. Thus, with the help of codes N. one can compensate the first one with the harmonic component of the intra-step error of the play sensor, which is harmonically changed as a function of displacement. In the simplest case, when the N code is the same for all quadrants of the sensor (const), in block 25, only one word is stored.

In a real induction sensor, the values of the output voltages are different from the ideal ones. The main reason for this is the intraperiod (intrastage) error, the components of which can be approximated using the function of the form sinKp0 and cosRvr, where Klj254, The most powerful weight among them is the first harmonic which is systematic because it is due to the manufacturing technology, namely the inaccuracy of manufacturing the sensor windings. The first harmonic channel, one norpetuHOCTH, is the result of the geometric orthogonality of sine and cosine5

sensor windings, it manifests itself in the electrical asymmetry of the sine od and cosine signals U.

Therefore, when the rotor is rotated by angle 0 of the real sensor, the signals Gc are taken from its outputs; and Uc, which correspond to the angle RE + c (instead of the angle PB in the case of an ideal sensor). The values of the signals Ug and U are determined from the expressions:

.io51p (re + L,

Uj.KT.UoCos (p0 + d).

Since the error is the angular value (the error in the argument for the function Us and Uj,), it can be compensated for in the measuring transducer, artificially entering the error in the value of the digital arguments N, N when forming the function f (sin) and f (cos). mismatch in the converter will be formed in this case in the following form:

11п (р0-Ф) 81п (re-с /) - (ф + .сУ).

Without compensation of the error, the law of mismatch will have a different form: sin (re + c /) -, which inevitably leads to the oyibke when forming the output angle f.

Accounting for errors with in values

,

13

digital arguments n. and n are implemented using the initial condition N2 codes, which define the values of real quadrants, obra-. Baths of the sensor's electrical axes. They are turned relative to the Cartesian axes and form sensor quadrants that do not coincide with the Cartesian quadrants. This is the sensor asymmetry, which in the converter needs to be repeated in the values of N. When the cG error (in the values of N2) is taken into account, the required number NUJ is maintained and the following equality is ensured for each sensor step:

Niy Ni + N2 + Nj + N4.,

where NI-N4 is the initial conditions 1-4

sensor quadrants. The introduction of asymmetry into the law of formation of triangular N and N is carried out by redistributing a small amount of disk space between the values of the N codes inside — each step of the sensor is based on

1024

12

received as a result of certification of the induction sensor.

FIG. 3 shows a functional diagram of the correction and control unit 15.

0

,

Negative carry pulses

Pr and

P come through element 24

CM on the sample input of the ROM 25 and on the switch 21, which is controlled by a logical signal e ,. Depending

l p

on the value of the signal ej, the negative transfer pulses commute

from to the summing or quantifying inputs — a reversible counter 22, which forms the t-address address z for sampling word 11- from ROM 25. The lower bits ap and the z numbers of the quadrant are input to the element 23 EXCLUSIVE OR, which forms a logical function a , From the outputs of block 15, the code Nj is removed, and the two logic functions a and a. Counters 13,

5 1 work only in the zone of uni-positive positive codes. However, when changing the quadrant number of the converter, the reversible counter that generated the impulse goes into the forbidden one, the area of negative codes (at its output is set to code 11 ...). To ensure the triangular law of variation of the N and M codes, it is necessary to return this counter to the zone of positive codes and set

 in it the required code is 00 ... 01. In another reversible counter, a setting is required

,

0

update code (). This initial setup of the counters 13, 14 should be performed when the quadrant changes between the time the negative transfer pulse appears and the next time it arrives (its counting frequency pulse.

The function of the initial installation of the counters 13, 14 is performed by the synchronization unit 16. At its two inputs, negative transfer pulses P2 and P are received, each of which initiates the formation of three pulses spaced apart at synchronizer 16 outputs, resulting in a reversible counter that worked on subtraction and formed a negative transfer impulse C code 00 .. .01, and on another reverse counter, the code jCNj-I).

Formula

13

and 3 o r e t

e ni

Claims (3)

1. The angular displacement transducer in the code containing the power supply unit, the first output of which is connected to the sensor input, the first and second outputs of which are connected to the analog inputs of the first and second multipliers, a switch, an integrator, the output of which is connected to the input of the voltage-frequency converter, The first and second outputs of which are connected respectively to the summing and subtracting inputs of the first reversible counter, the subtractor, characterized in that, in order to increase the accuracy and expand the area of the changing the converter by compensating for errors and the possibility of changing the output code, it has an adder, a high-pass filter: frequencies, a controlled inverter, a switch, a second and a third reversing meter, two elements OR, an EXCLUSIVE OR element, a correction unit, and control unit and the synchronization unit, the output of the first multiplier is connected to the first input of the adder and the inverse input of the subtractor, the output of the second multiplier is connected to the second input of the adder and the summing input of the subtractor, the outputs of the adder and subtract l are respectively connected k.pervomu the second information inputs switch, reversible counter, element 40 45 a key switch whose output is connected via a high-pass filter to the information input of a controlled inverter whose output is connected to the integrator input, the first and second outputs of the voltage-frequency converter are connected respectively to the first and second information inputs of the switch, the first and second outputs of which are connected to the first the inputs of the first and second elements, respectively, OR, the output of the first element of the RSHI is connected to the summation of the information input of the second reversible counter and the subtractive information Discount input of the third down counter, the output of the second OR gate is connected to the data input of the second subtractor summiruyup and he data input of the third down counter 5, second and third group of reversible counters outputs connected respectively to the digital inputs of the first 50 OR, the EXPLOSIVE OR element and the persistent storage, whose outputs are the information outputs of the correction and control unit, the inputs of the OR element are the first and second inputs of the correction and control block, the control input of the switch is the third input of the unit correction and control element output OR is connected to the control input of the persistent storage device and informational input of the switch, the outputs of which are connected to the inputs of the reversible counter. the output group of which is connected to the information inputs of the permanent reference of 1; its devices, the outputs of two low-end bits of the output counter group of the counter are connected to the inputs of the EXCLUSIVE OR element, whose output is the first control output of the correction and control unit, the output of the first time low yes groups of outputs reverse counter311024 14 and the second multipliers, the transfer outputs of the second and third reversible counters are connected respectively to the first and second inputs of the correction and control unit and respectively to the first and second inputs of the synchronization unit, the first and second outputs of which are connected to the second inputs of the first and second OR elements, respectively, the third and the fourth outputs of the synchronization unit are connected to the inputs of the preset control, respectively, of the second and third reversible counters, the third output of the voltage converter is the frequency connected to the third input of the correction and control unit, the information outputs of which are connected to the preset inputs of the second and third reversible meters, the first control output of the correction and control unit is connected to the first input of the EXCLUSIVE RSHI element, whose output is connected to the control input of the controlled inverter, the second the control output of the correction unit, and the control is connected to the control to the common inputs of the switch and the switch, the second output of the power supply unit is connected to the second input of the element EX1CH1SP1EE OR. 2. The converter according to claim 1, i.e., in that the correction and control unit contains a commutator, a reversible counter, the OR element, the EXCLUSIVE OR element, and the persistent storage, whose outputs are the information outputs of the correction and control unit, the OR element inputs are the first and second inputs of the correction and control unit, the control input of the switch is the third input of the correction and control unit , the output of the OR element is connected to the control input of the permanent storage device and the informational input of the switch, the outputs of which are connected to the inputs of the reversible counter, the output group of which is connected to the information inputs of the permanent memory 1; its devices, the outputs of the two younger and 1 bits of the output counter group of the reversible counter are connected to the inputs of the EXCLUSIVE OR element, the output of which is the first control output of the correction and control unit, the output of the first low digit of the reverse output group counts. 1513 This is the second control output of the correction and control unit. 3. The converter according to claim 1, which is characterized in that the synchronization unit contains four AND-NOT elements, two inverters, two triggers, a clock distributor and an OR element, the first and second inputs of which are the first and second inputs synchronization unit and connected to the S-inputs of the first and second triggers, respectively, whose outputs are connected to the first inputs of the first, second and third, fourth AND-NOT elements, respectively, and the outputs of the first and fourth AND-NOT elements are respectively the first and second exits bl Single synchronization output of the third elemenQn QW OIL %%%%but BUT/ 2416 This AND-NO is the third output of the synchronization unit and is connected to the output of the first inverter, whose input is the fourth output of the synchronization unit and connected to the output of the second inverter, whose input is connected to the output of the first inverter, the output of the second AND-NOT element is connected to the output of the second the inverter, the first output of the clock distributor is connected to the second inputs of the second and third elements of the NAND, the second output of the distributor of clock pulses is connected to the second inputs of the first and fourth elements of the NAND, the third Exit distributor ipylcov clock and is connected to R-inputs of the first and second flip-flops, vpsod OR gate connected to the input clock pulse distributor. F .f F J6ff