In the field of industrial drives, the vector control of the induction motor has been widely used to achieve the high performance torque control. In this paper, a new high precision torque control of the induction motor has been proposed. The control strategy proposed here is developed according to the conventional vector control law, but it has a feature that the secondary is controlled directly by selecting the appropriate voltage vector of the voltage-fed inverter.
The prototype was constructed using a 2.2kW induction motor where the torque control was implemented by software of DSP-TMS 32010. All control processing including the identification of the rotor resistance could be executed within 100μs. From the experimental results, the linear relationship between torque and slip frequency was established for the wide frequency range from rotor rocked condition to 50Hz. Due to identification of rotor resistance, the accuracy of the torque control within ±2% was obtained regardless of the temperature change of the motor.

View full abstract

A great deal of efforts have been made in order to overcome the dynamical nonlinearity of servo system. One of the proposed methods is the acceleration control method based on the disturbance observer. However, as the disturbance observer requests the information of motor speed, this control method is often under the influence of the noise of speed sensor in the low motor speed range. In order to overcome these problem, this paper proposes a new DSP-based acceleration controller. This acceleration controller is designed on the basis of both the dual passive adaptive control loop and the direct compensation element G(s). It is called the model following acceleration control system in this paper. The proposed acceleration control system is fixed to the nominal mechanical dynamics in the presence of parameter variations and force disturbance. Hence, both the fast motion control and the accurate force control can be carried out easily. For the purpose of realizing the fast and accurate control, the proposed system is constructed by the software argorithm of DSP (digital signal processor). The experimental examples show that the proposed method is valid for both the fast motion control and the accurate force control.

View full abstract

Direct drive motor drives the mechanical load without a reduction gear and, therefore, the required precision of torque control is higher than that of the ordinary AC servo motors. The current control is the key technology to realize the high precision torque control.
A new software-implemented current control is proposed. Since the proposed method is the extension of the conventional current hysteresis controlled PWM, the algorithm is quite simple and can be implemented within 20 μs using TMS 32010. The processing time is so short that 3-axis control by 1 DSP can be realized.
Further, the analysis of the torque ripple due to the non-sinusoidal flux distribution is given and it is shown that the ripple can be compensated by the proposed current control. The additional processing time is 12 μs.
Torque ripple is also caused by the offset involved in the current pick up loop. Based on the analysis of torque ripple due to the offset, the compensation algorithm with processing time of 26 μs is proposed and proved to be excellent by experiments. This means the realization of tuning-less drive system.

View full abstract

Use of G3 facsimiles has been spreading in response to new communication demands. Their cost has been reduced along with their size. The modem used in G3 facsimiles has also been integrated in accordance with these trends.
In order to fabricate a high cost performance modem, we have produced a 1-chip VLSI Modem for G3 facsimiles by using a DSP (DSP-E). To realize this VLSI Modem, we took into consideration the following items. The first was to raise the device production efficiency by lowering the amount of analog circuits which are difficult to include in the VLSI. The second was to incorperate most of the functions by the software processing in order to make a version changes simple.
We had to design the construction architecture of the VLSI and to develop the signal processing method best matched to this modem. This paper deals with the architecture, the processing performance of the DSP, and the specialized monitor software used to arrange software in program ROM efficiently.

View full abstract

As two-phase motors used for small fan drive and so on has zero torque point round one rotation, self-starting is serious problem. Therefore it is necessary to reduce torque ripple and to improve ability of self-starting by choosing the most suitable combination of cogging torque and electrical torque. But some reported analysis using Finite-Element-Method has a limit on separation between permeance distribution and flux distribution. Considering these problems, we developed a new method to analyze cogging torque using Schwarz-Christoffel transformation. We calculate effect of inter-pole shape, magnetization distribution of parmanent magnet and position of Hall-sensor, and the following conclusion is obtained:
(1) The inter-pole reduces ripple of permeance and cogging-torque.
(2) It is necessary to make the minimal shift of inter-pole from center position.
(3) By setting inter-pole gap equal or a little smaller than main teeth gap, total torque ripple can be reduced.
(4) Magnetizing permanent magnet into sinusoidal distribution reduces cogging torque effectively.
(5) Shifting the position of Hall-sensor against rotation direction improves ability of selfstarting.

View full abstract

A new variable speed control method for the Commutatorless Motor is proposed on the basis of Modern Control Theory, to provide not only a quick response for speed reference but a high efficiency operation. The summarized contents of the proposed method are the following.
(1) The optimum ratio of armature current to field current making the efficiency maximum is derived.
(2) Ordinarily, as a commutatorles motor has an unstable tendency when it is driven in the manner of a constant commutation margin angle, the air gap flux is necessary to keep constant by sensing the air gap flux. However, the proposed method does not need the flux sensing because of estimating all the state variables with accessible state variables.
(3) For reducing the production cost, the technique that constructs the feedback control system using the accessible state variables is employed.
(4) As a result from considering thyristor commutating mechanism, a new current controller which protects the motor system from the commutation failure is introduced.
Finally, the proposed control method is applied for the numerical commutatorless motor model, and its usefulness is confirmed through the simulation.

View full abstract

In many industrial applications, high performance and light position controllers for robot drives are required. To satisfy these needs, a new backlashless servomechanism using twin motors is proposed. Both twin motors are coupled with reduction gears and are controlled to generate opposite torques each other in any load for compensating the backlash error of the gears.
The motor controller consists of positioning control for one motor coupling with a position sensor and torque control for the other. For positioning control, I-PD control implemented with a position feedback and a speed feedback are employed. In the accelerating or decelerating action, both motors are controlled to generate the same direction torque to achieve the highest speed response. In the steady state, one motor is controlled to generate the opposite constant torque for compensation of the backlash errors.
In the experimental system, twin 100 WDC motors and reduction gears with the double-reduction by spur gears of the ratio of 1: 50 are employed. The two-phase rotary encoder of 500 ppr directly coupled with the motor shaft is used for sensing the position. The DC tachogenerator is also coupled with the same shaft.
The experimental results prove that the proposed system provides high precision about 13 (1/100, 000 rev.) and the weight is about 1/3 of the conventional direct drive motor.

View full abstract

For detection of AC voltage with non-contacting way, a potential transformer is widely utilized, However, when the voltage to be detected is low in the frequency or includes any DC component, the well-known problem of magnetically biased field or magnetic saturation is often experienced and does not put the high accurate detection into practice.
This paper discusses the voltage sensor which solves these problems and intends to achieve high accurate detection. The constitution is composed of the following procedures: (i) Generate a pair of positive and negative voltages in proportion to the voltage to be detected: (ii) Take out the positive and negative voltages from the voltage of (i) through an alternative switching of which frequency is sufficiently high compared with that of the voltage to be detected; (iii) Set the resultant voltage as a reference value for the voltage induced in the search coil arranged on the isolating transformer; (iv) Take the inverse operation of (i) to the secondary coil which is so wound as to make a close coupling with the search coil and reconstruct the original voltage to be detected. This sensor can also detect DC voltage.
The factors causing the detection error are, in general, due to both the existence of primary winding resistance, leakage inductances and magnetizing impedance as well as the dependence of these values on the magnitude and the frequency of the voltage to be detected. The proposed sensor reduces these effects. In practice, for ± 10 V full scale, the minimum detectable voltage is 10 mV when the % error of 0.08% is allowed.

View full abstract

The vector control of an induction motor has made an epoch-making progress in the torque control. If the system is ideal, the torque transfer function should be constant under the vector control. Actually, however, the change of rotor resistance, the control capability of stator currents and the sampling of digital controller cause the transient state. In order to study these unfavorable effects, this paper proposes a linear model for the current source inverter-fed induction motor drive system.
Firstly, a linear continuous system model is derived by using a small displacement method, and the ideal characteristics of the vector control are theoretically clarified in the proposed analytical model.
Secondly, the system in which the vector control is carried out by a microprocessor is analyzed. A linear sampled-data system model is derived considering the effects of the sampling period and the dead time.
The last part of this paper discusses the computed results concerning the trajectories of poles and zeros, and the transient responses to the interesting system parameters such as PI current control gain K_i, rotor resistance and sampling period. What is acquired in the analysis gives vital directions for the design of the system, for example, the selection of the gain K_i, should be a large one to stabilize the system, but it should not exceed an upper bound value caused by the sampling.

View full abstract

Recently, large rolling mills have been used in steel industry, so that cycloconverter-fed AC motor drives have been applied to the systems. For the systems, there are induction motor drives with advantage of easiness of maintenance, and synchronous motor drives with advantage in view of power factor and converter capacity. Both kinds of motor drives are controlled by vector control method, because the rolling mill drives need high performance variable speed control. The control method and its characteristics for the vector-controlled AC motor drives have been discussed. However, the characteristics of the method for synchronous motor drives have not sufficiently been described.
In this paper, decoupling control method for the vector-controlled synchronous motor with field and damper windings is proposed. The necessary transfer functions for the designing the vector control system are derived and it is shown that the method of high performance variable speed and power factor control are obtained. Then, the characteristics of the method are confirmed by the simulation results. In the decoupled system, the transfer functions between the voltages (armature and field) and currents are the first order lag and the transfer functions between the currents and armature flux are the first order lead-lag. Consequently, stability and response performance of the decoupled synchronous motor drives can be designed with suitable characteristics by this control strategy.

View full abstract

The half-bridge high-frequency inverter using Static Induction Transistor (SIT) described in this paper can be formed as a current-fed type inverter or a voltage-fed type inverter. The SIT inverter is a low-cost one which has a quick response for output power control by Pulse Width Control (PWC). The circuit operation including the practical PWC mode was elucidated by introducing some normalized circuit constants and operating frequency. The characteristics of a half-bridge high-frequency inverter using SIT were discussed in this paper in comparison with current-fed type and voltage-fed type. The merits and demerits in output power control by PWC were studied for both type inverters. The characteristics of the output power control by PWC were substantiated on the basis of the numerical analysis and experimental results. Results of the characteristics analysis of the SIT half-bridge high-frequency inverter indicate that the inverter is effectively applicable to the static power controller.

View full abstract