JP2625705B2 - Digital current control servo driver - Google Patents

Description

Description: A. TECHNICAL FIELD The present invention relates to a digital current control servo driver for driving a DC motor, and particularly to a digital hardware without using an analog operation by an operational amplifier or a software processing by a microcomputer. The present invention relates to a digital current control servo driver realized by the above.

B. Summary of the Invention The present invention relates to a servo driver that uses a PWM pulse as a servo driver and performs current control when a DC motor is driven, and includes an error caused by each PWM pulse between a target value and a control output feedback value. By performing high-speed PI calculation by hardware and operating the drive unit, a servo driver that does not perform analog calculation, does not use frequency conversion and uses many signal lines as in the case of software processing is obtained. .

C. Conventional technology and its problems Conventionally, the current control servo driver that drives a DC motor is an analog type composed of an operational amplifier and the like, and the adjustment of offset and temperature drift is complicated.
With the advent of high-speed, large-capacity microcomputers and digital signal processors (DSPs) as technology advances, digital methods have been adopted and adjustments have become easier.

As the digital type servo driver, a frequency-input type servo driver shown in FIG. 5 and a parallel signal input type shown in FIG. 6 are used. In FIG. 5, the current detector 1
The V / F converter 2 converts the current detected in to the frequency f ₀ , and the servo controller 3 converts the serial signal into a parallel signal together with the command frequency f _S , and then performs a servo operation by software processing. P at converter 4
A WM pulse is output, and a voltage is output as each gate signal of the power bridge 5 shown in FIG. 5 (B).

However, in this method, the frequency becomes higher or lower based on the reference frequency, so that the control cycle for controlling the current as the servo driver changes depending on the frequency. That is, f _{S is} higher than the reference frequency.
If the speed is increased, high-speed operation is performed. If the speed is lowered, low-speed operation is performed. However, the low-speed operation in this case is delayed due to low frequency. This causes a dead time in control and causes instability.

In the case of the parallel signal input type shown in FIG. 6, the A / D converter 6 converts the current detected by the current detector 1 into, for example, a 12-bit digital quantity (D _{00 to} D ₀₁₁ ). Te, the command digital quantity (D _S0 ~D _S11), and performs servo operation by software processing by the servo controller 3. In this method, it is necessary to receive the command digital amount as a parallel signal from the host controller. For this purpose, the number of signal lines is increased and, for example, 12 lines are required.

Therefore, the present invention does not execute analog software using an operational amplifier or the like, and does not perform digital software processing based on a frequency input type or a parallel signal input type having a large number of signal lines causing a problem at a low frequency. PW
The purpose is to provide a digital current control servo driver realized by digital hardware as an M signal input.

D. Means for Solving the Problems In order to achieve the above object, the present invention provides a digital current control servo driver for driving a DC motor, which is provided with a PWM converter for converting a detected current into a PWM pulse. P from host controller
Converting the WM pulse into parallel data, respectively, then the in consideration of the PI controller output m _n-1 of this etc. Each of the parallel data on the basis of the deviation e _n of press and previous deviation e _n-1 and the previous The present invention is characterized in that a PI controller that sequentially performs a PI operation by a timing pulse based on a PWM pulse is configured by hardware, and includes a PWM conversion unit that uses the PI controller output _mn as a gate signal.

E. Operation Each of the DC motor control command and the motor current control feedback is a PWM pulse, and this PWM pulse is converted into parallel digital data by digital hardware in the servo controller, and PI calculation is performed by digital PI algorithm. The calculation result is processed into the gate signal of the FET bridge, and the PWM signal is processed without processing the analog signal as before, and without performing V / F conversion or digital software processing based on the parallel signal.
Reduction of the number of signal lines based on signals, signal processing of "0" and "1" in principle, which is resistant to noise due to PWM signals, PW
The calculation timing by the M signal can be speeded up.

F. Embodiment An embodiment of the present invention will now be described with reference to FIGS. In FIG. 1, a current detected by a current detector 1 is converted by a PWM converter 10 into a PWM pulse corresponding to the magnitude and a direction pulse (sign pulse). on the other hand,
The servo controller 11 that inputs a PWM command pulse and a sign pulse from the host controller and a feedback PWM pulse and a sign pulse from the PWM converter 10 converts the PWM pulse into a parallel digital amount, and performs a servo operation (digital PI operation). The result is processed into a gate signal of the FET bridge 12 and amplified. Here, in FIG. 1, both the PWM pulse and the sign pulse are separately produced and sent, but FIG. 2 shows an example in which the sign (polarity) pulse is included in the PWM pulse itself. That is, the code can be determined from the position information of the PWM pulse, and it is determined whether the information at a certain point is 0 or not.

Here, the FET bridge 12 drives the motor by the FET assembled in the bridge as shown in FIG. 1 (B).

The servo controller 11 converts a PWM pulse into parallel data, performs a PI operation by a PI algorithm, and performs a PWM conversion as a signal for output to the FET bridge 12. In this case, a PI operation unit (a PI controller) is included. All processing is performed by hardware. That is, PWM
In FIG. 3, which shows the inside of the servo controller, the PWM servo controller includes a clock, a timing generator a, a PWM command pulse parallel converter b, a PWM feedback pulse parallel converter c, a command pulse I ^*, and a feedback pulse I. control of the _ANS deviation e _n = I ^* -I a computing section d of _ANS, the deviation e _n the integral time T _I, operation when the sampling period _{ΔT (ΔT / T I) ·} e n becomes calculation section e ,
From further _{e n + (ΔT / T I} ) calculation of -e _n f, n-1-th sampling time of the PI controller output m _n-1 and n-1 th control deviation e _n-1 Tokyo the time of sampling _{(m n-1 / P)} -e n-1 to obtain a calculation unit _{i, e n + (ΔT /} T I) is obtained from the output of the output operation portion f of the operational unit i e _n + (m _{n- 1} / P) -e _n-1 of the arithmetic unit g, by a proportional gain _{P m n = P (e n} + (ΔT / T I) e n -
_{e n-1) + m n} -1 of the arithmetic unit h, is composed of a PWM converter section j, n-th sampling by the n-th control deviation of the sampling is e _n from step d to h to PI operation When the PI controller output is getting _mn . That is, the control deviation e _n, a proportional gain P, expressed an output m _n of the PI controller as integral time T ₁ at controlled following equation _{m n = P (e n +} (1 /
T _I) ∫e _n dt) becomes consistent with operational expression in the arithmetic unit h.
The operation units in steps d to h perform digital PI operations, and are sequentially performed by hardware.

In this case, the operation timing is PW as shown in FIG.
This is performed in synchronization with six pulses based on the pulse _PGf by the M feedback pulse. That is, the computation of the operational pulse P ₁ in e _n = ^I * -I _ANS, the calculating pulses P ₂ (delta
T / T _I) calculation of e _n, the calculation pulse _{P 3 e n + (ΔT /} T I) · e n
In the calculation, calculation pulse _{P 4 e n + (ΔT /} T I) e n + (m n-1 /
P) calculation of -e _n-1, operational pulse P ₅ in _{m n = P (e n +} (ΔT
_{/ T I) e n -e n} -1) + m n-1 calculations in the calculating pulse P ₆ (m
_n-1 / P) -e _n-1 is calculated. These calculations can be processed at a high speed and end in about 1 μsec after the fall of the PWM feedback pulse. The PWM converter j processes the gate signal of the FET bridge from the rising point of the PWM feedback pulse and outputs it. Is done.

In FIG. 3 and FIG. 4, each terminal input is as follows.

PWM _C: PWM command pulse I ^* _S: command code pulse PWM _f: PWM feedback Pulse I _ANS.S: Feedback code pulse _{A D ~D D: (ΔT /} T I) · e n arithmetic multiplication constants A _M to D _M: m _n arithmetic multiplication constant I ^*: command digital quantity I _ANS: feedback digital quantity e _ns: e _n of code pulse _{S 7, S 8, S 6} : computing section h, i, code pulse of the operation result of the j Here, the PI calculation shown in FIG. 3 will be described more specifically.

(1) First, in the PWM command pulse parallel conversion unit b, P
The rise time of the WM command pulse is counted and the command value I ^* of the parallel data (same as the data used for ordinary digital numerical calculation) is calculated.

(2) Next, the PWM feedback pulse parallel converter 5
In, the rise time of the PWM feedback pulse is counted by a counter to obtain the parallel data _IANS .

(3) e _n = ^I * In -I _ANS computing unit d, by using the ^{I *} and I _ANS in synchronization with the timing pulses P ₁ e _n = ^{I *} -
Calculate I _ANS .

_{(4) (ΔT / T I} ) · e n in the arithmetic unit e, in synchronization with the timing pulses P ₂ (3) multiplied by the e _n obtained in claim constants A _D
From to D _D to calculate the _{(ΔT / T I) · e} n.

Here, the constants _{AD to DD} are 4-bit numerical values that can be set by a dip SW of 0 or 1, and specifically, the circuit is configured to have the following relationship.

In ^{_{ΔT / T I = 2 -1 ×}} A D +2 -2 × B D +2 -3 × C D +2 -4 × D D (5) e n + (ΔT / T I) · e n calculation section f, the timing in synchronization with the pulse P _{3 (3)} (4) using the results of term
to calculate the _{e n + (ΔT / T I} ) · e n.

(6) In the calculation unit _{g, e n + (ΔT /} T I) from the previous sampling operation results of the operation unit i which will be described below with the results of the synchronously (5) to claim the timing pulse P ₄ · e _n + (m _n-1 / P) -e _n-1 is calculated.

 Here, the initial value of the calculation unit i when n = 1 is 0.

(7) In the calculation unit h, results and multiplication constants in synchronization with the timing pulses _{P 5 (6) A M ~D} m n = P (e n + (ΔT / T I from the results of _M) · e _n Calculate + ( _mn-1 / P) -en _-1 .

At this time, the constants A _{M to} D _M are 4-bit numerical values that can be set by a dip SW of 0 or 1, and specifically, the circuit is configured to have the following relationship.

_{^{P = 2 0 × A M +2}} 1 × B M +2 2 × C M +2 3 × D M The resulting PI controller output m _n a PWM conversion unit j
(Description omitted) to create a gate signal for the FET bridge.

(8) In the operation unit i, calculates the synchronously _{(m n-1 / P)} -e n-1 to the timing pulse P _6.

Here, (m _n-1 / P) uses the data before multiplying by P in the calculation part h. E _n−1 uses the result of the operation unit d.

In the arithmetic unit i, unlike the other arithmetic units, n-1 is added as a suffix despite the n-th sampling. However, as apparent from the item (6), the arithmetic result at the time of the sampling is obtained by the n + 1 sampling. This is because it is used in the PI operation (in the operation part g of the item (6)).

G. Effects of the Invention As described above, according to the present invention, since the servo controller is configured by digital hardware using the PWM signal as input, the adjustment is simple in the digital system, and the signal line is reduced because of the PWM signal. It is strong against noise and can be increased in distance from the host controller, enabling remote control. In addition, the adoption of a PWM converter can increase the speed of A / D conversion and increase the speed by processing using arithmetic pulses. The problems of conventional analog arithmetic and digital software processing can be entirely eliminated.

[Brief description of the drawings]

1 (A) and 2 are block diagrams of two examples of a PWM servo driver according to an embodiment of the present invention, FIG. 1 (B) is an FET bridge, FIG. 3 is an internal configuration diagram of a servo controller,
FIG. 4 is a time flowchart, and FIGS. 5 and 6 are circuit diagrams of a conventional example. In the figure, 10 is a PWM converter, 11 is a servo controller, and 12 is a FET bridge.

Claims (1)

(57) [Claims] In a digital current control servo driver for driving a DC motor, a PWM converter for converting a detected current into a PWM pulse is provided, and this PWM pulse and a PWM pulse from a host controller are respectively converted into parallel data. then PI sequentially by a timing pulse based on the respective parallel data etc. the PWM pulses in consideration of the PI controller output m _n-1 of and the previous deviation e _n-1 and the previous based on the deviation e _n dumb A digital current control servo driver, comprising: a PI controller that performs an operation; and a PWM converter that uses the PI controller output _mn as a gate signal.