Control systems for AC Motor VSDs--part 3

AC drive vector control

The term vector control is probably one of the more abused terms in industrial control and consequently has caused considerable confusion amongst users of VS drives. Vector control for AC variable speed drives has been available from some drive manufacturers since the mid-1980s. The technique of vector control has only become possible as a result of the large strides made in solid-state electronics, both with microprocessors and power electronics.

It has been promoted as an AC drive equivalent to DC drives and claimed to be suitable for even the most demanding drive applications and this is where the confusion arises.

The statement is true, but only to the extent that the principles of vector control are implemented. There are degrees to which this enhanced type of control can be applied to AC variable speed drives. Some manufacturers have encouraged this confusion in an effort to attribute higher performance characteristics to products that only partially apply the technology of vector control. The meaning of the various terms is covered later in this section after the fundamental principles of vector control are explained. Today, the term 'vector control' has become a generic name applied to all drives which provide a higher level of performance (compared to the fixed V/f drives). Referring back, electric motors produce torque as the result of the interaction of two magnetic fields, one in the fixed part (stator) and the other in the rotating part (rotor/armature) and their interaction across the air-gap. The magnetic fields are produced by the current flowing in the windings of the stator and rotor. The motor torque depends on the strength of both of these magnetic fields. In fact, torque is proportional to the product of the currents producing these two magnetic fields.

In a DC drive, it’s fairly well understood that the output torque is proportional to the product of two current vectors, the armature current Ia (torque producing current) and the field current If (flux producing current), at 90 deg. to one another. In practice, the field current (flux producing current) is normally held constant. Consequently, the armature current Ia is directly proportional to output torque of the motor. Armature current (Ia) can be used as the torque feedback in the cascaded closed-loop controller. Both these currents can readily be measured and accounts for the simple control of the DC drive.

In an AC induction motor (refer to equivalent circuit in the schematic), the flux producing current (Im) and torque producing current (Ir) are 'inside' the motor and cannot be measured externally or controlled separately. As in the DC drive, these two currents are also roughly at 90 deg to one another and their vector sum makes up the stator current, which can be measured. This is what makes the vector control of an AC motor more difficult than its DC counterpart. The challenge for the AC flux-vector drive is to distinguish and control these two current vectors without the benefit of two separate circuits and only being able to measure and control the stator current.

The strategy of an AC vector control drive is to calculate the individual current vectors to eventually enable separate control of the flux current and/or the torque current under all speed and load conditions. As in the DC drive, the aim is to maintain a constant flux current in the motor.

The calculation of the current vectors involves the measurement of the available variables (such as the stator current (Is), stator voltage (Vs), phase relationship, frequency, shaft speed, etc) and applying them to a 'motor model', which includes the motor constants (such as the stator resistance and inductance, the rotor resistance and inductance, the magnetizing inductance, number of poles, etc). Because of the many variables, there are many possible applications of a motor model, from simple estimation of motor conditions to those that are very comprehensive and very accurate. The more detailed the motor model, the more processing power is required.

++++ Simplified equivalent circuit of an AC induction motor

Under motor no-load conditions, almost all the no-load stator current IS comprises the magnetizing current. Any torque-producing current is only required to overcome the windage and friction losses in the motor. Slip is almost zero, stator current lags the voltage by 90deg , so power factor is close to zero (Cos_f = 0). At low motor loads, the stator current IS is the vector sum of the magnetizing current IM (unchanged), with a slightly increased active torque-producing current. Stator current lags the voltage by a large angle f, so power factor is poor (Cos f << 1). Slip is still small.

At high motor loads, the stator current IS is the vector sum of the magnetizing current IM (unchanged), with a greatly increased active torque-producing current, which increases in proportion to the increase in load torque. Stator current lags the voltage by the angle f, so power factor has improved to be close to the full load power factor (Cos f = 0.85).

++++ Current vectors in an AC induction motor

Therefore, the central part of the vector control system is the active motor model, which continuously models the conditions inside the motor:

• Continuously calculates in real time the torque-producing current by implementing the following activities:

- Stores the motor constants in memory to be used as part of the calculation

- Measures stator current and voltage in each phase

- Measures speed (with encoder) or calculates speed (no encoder)

• Continuously calculates in real time the flux-producing current

• Implements the speed control loop by comparing the speed feedback with the speed setpoint to provide an error output to the torque control loop

• Implements the torque control loop by comparing the active torque, calculated from the current and speed feedback, to provide an error output to the PWM switching logic controller

• Constantly updates this information and maintains tight control over the process.

For adequate dynamic response of the drive, the model calculations need to be done at least more than 2000 times per second, which gives an update time of less than 0.5 ms.

Although this is easily achieved with modern high speed processors, the ability to continuously model the induction motor at this speed only became viable within the last 10 years or so with the development of 16-bit microprocessors. Initially, sufficient processing power for vector control was quite expensive, but over a period of time, the cost of the processors has come down and processing speed has increased significantly.

The main difference between a traditional fixed V/f ratio VVVF converter and a modern vector control drive is almost entirely in the control system and the extent to which the active motor model for vector control is implemented to control the switching pattern of the IGBTs of the inverter.

The power circuit for a vector converter is almost identical to that used by a VVVF drive:

• Rectifier to convert 3-phase AC to a DC voltage

• Inductive choke to reduce harmonics on the supply side

• Dc link with capacitor filter to provide a smooth and steady DC voltage

• An IGBT semiconductor inverter bridge to convert the DC to a PWM variable voltage variable frequency output suitable for an AC induction motor

• A microprocessor based digital control circuit to control the switching, provide protection and provide a user interface

Today, 'standard' AC variable speed drives from most reputable manufacturers implement vector control to some degree. E.g., sensorless vector control is advertised as a performance feature with almost all modern AC drives.

There are essentially 3 basic types of control for AC variable speed drives today:

• Basic fixed V/f drive, provides fair speed control at a reasonable price and is suitable for the control of centrifugal pumps and fans

• V/f sensorless vector drive, provides better speed regulation, better starting torque and acceleration by implementing more/better control of the flux producing current vector (flux-vector)

• Closed loop field oriented vector control drive, provides excellent speed and torque control with DC like performance using cascaded PI control over speed, torque as well as flux regulation. Dynamic performance is excellent.

Fixed V/f drive

The control strategy of a fixed V/f drive is essentially open-loop control as shown below.

• The speed reference is taken from an external source and controls the voltage and frequency applied to the motor.

• The speed reference is first fed into a ramp circuit to convert a step change in the speed request to a slowly changing signal. This prevents electrical and mechanical shock to the speed control system. The acceleration and deceleration ramp times can be set by the user.

• The signal is then passed to a section that sets the magnitude of both the voltage and frequency fed to the motor. The V/f ratio between the voltage and frequency is kept constant at all times. It also sets the rate of change of these two values, which determines the motor acceleration.

• The base voltage and base frequency used for this ratio are taken from the motor nameplate.

• Finally, the signal passes to the PWM switching logic module, that controls the switching pattern of the IGBT switches to provide the voltage pattern at the output terminals according to the PWM algorithm (sine-coded, etc).

• There is usually no speed feedback from the motor. It’s assumed that the motor is responding to and following the output frequency (open-loop control).

• The current feedback from the current transducer is there mainly for protection, indication and to set a current limit, it’s NOT used as part of the control strategy.

++++ Block control diagram of fixed V/f drive It’s necessary to monitor the stator current flowing to the motor. The drive usually monitors total current and cannot distinguish between Im and Ir. This current is not used to control torque, but is aimed at the following functions:

• Measures actual current for the I^2 t overload protection of the motor

• Provides protection of the power electronic components

• Provides a current limit, the control system reduces the frequency command signal when the current exceeds a predetermined value. Usually, current limit is set to 150% of the rated motor current.

• Some newer V/f drives provide slip compensation as a strategy for improving the speed holding capability in an attempt to maintain relatively constant motor speed even with changes in the motor load torque. As the output torque increases, the motor current increases, which can be used to adjust the output frequency of the converter. E.g., at full rated load, the full slip value can be added to the output frequency. With slip compensation, improved speed regulation can be obtained from an induction motor without a speed feedback device.

This method of open loop fixed V/f control is adequate for controlling steady-state conditions and simple applications, such as pumps, fans and conveyors, which allow a lot of time for speed changes from one level to another and where the consequences of the changes in the process are not severe.

This type of drive is not well suited to the following:

• Applications where motors run at low speeds (below 5 Hz). The torque at low speed is generally poor because the stator volt drop significantly affects the magnitude of the flux-producing current. Many V/f drives include a 'start boost' when allows the V/f ratio to be boosted at starting in an attempt to improve the flux and consequently the starting torque.

• Applications which require higher dynamic performance.

• Applications that require direct control of motor torque rather than motor frequency.

• The dynamic performance of this type of drive with shock loads is poor.

Sensorless flux-vector drives (open-loop vector)

The development of sensorless flux-vector drives was aimed at overcoming the main shortcomings of the fixed V/f drives, mainly the loss of torque at low speeds.

This type of drive is often also called an open loop vector drive because its basic core is still the fixed V/f ratio controller. But wrapped around this core are several additional control components:

• A current resolver (mathematical model) that uses the measured stator current to calculate (in real time) the two separate current vectors which represent the flux-producing current (Im) and the torque-producing current (Ir)

• A high performance current limiter which uses the torque-producing current (Ir) to rapidly adjust the frequency command to limit current

• A flux regulator which continuously adjusts the V/f ratio to maintain an optimum control of the flux-producing current (Im)

• A slip estimator that provides accurate estimation of the rotor speed based on the known motor parameters, without the use of an encoder. This provides improved slip compensation under all conditions of speed and load.

The result is greatly improved torque, particularly at low speeds, to provide high breakaway and acceleration torque and an improved dynamic response to shock loads.

However, this type of drive does not provide torque control, it’s still a speed control device. In addition, speed holding capability is substantially improved.

This type of drive can also be operated with an encoder, providing closed-loop control of the speed. This substantially improves the speed holding capability of the VS drive with speed regulation of 0.1%.

Closed-loop field oriented vector drive

Up to the end of the 1980s, high performance drive applications inevitably required the use of a DC drive. However, the high maintenance requirements of DC drives have encouraged the development of alternative solutions. Vector controlled AC drives have evolved to provide a level of dynamic performance that has now exceeded that of DC drives.

Closed-loop vector control is not required for every AC VSD application, in fact only on a minority of applications. But there are a number of applications that inherently require tight closed-loop control, with a speed regulation better than 0.01% and a dynamic response better than 50 radians/sec. This dynamic response is about 10 times better than that provided by standard V/f drives.

The control block diagram for a high performance vector control AC drive system is essentially a cascaded closed-loop type with speed and torque control loops:

• There are two separate control loops, one for speed and the second for current. This control strategy is similar to that used for the control of a DC drive.

- Speed loop controls the output frequency, proportional to speed

- Torque loop controls the motor in-phase current, proportional to torque

• The speed reference command from the user is first fed into a comparator, from where the error controls the speed regulator

• The speed error signal becomes the setpoint for the torque (current) regulator.

This signal is compared to the calculated current feedback from the motor circuit and the error signal determines whether the motor is to be accelerated or decelerated

• There is a separate control loop for the flux current (V/f regulator)

• Finally, the signal passes to the PWM and switching logic section, that controls the IGBTs in such a way that the desired voltage and frequency are generated at the output according to the PWM algorithm (sine-coded, star modulation, VVC, etc).

++++ Block diagram of the flux-vector converter control circuit

Although a shaft mounted incremental encoder can be used to measure speed in an AC drive, it’s often considered to be an additional expense. In some cases it’s difficult to mount on the motor, E.g. when motors have integral brakes. Even when an encoder is not used, the cascaded closed loop control can still be implemented because speed can be calculated by the active motor model, but with a lower level of accuracy due to the difficulty of calculating slip, particularly at very low speeds. Vector controlled drives which don’t use encoders are usually referred to as sensorless vector drives. The dynamic response of vector control drives, which don’t use an encoder, is usually inferior to those that do.

The following are some interesting figures that have been presented by one of the leading manufacturers of variable speed drives:

Typical applications for this type of high performance VS drive are:

• Crane and hoist drives

• Rewinders on paper and steel-strip lines

• Paper machines

• Printing machines

• Positioning systems for automated manufacturing lines

• etc.

When setting up high performance VSD controllers, a modest proportional gain gives a good transient response, while the integral gain gives high steady state accuracy. PI controllers have the advantage that they can maintain a non-zero output to drive the converter although their input is zero. This is an advantage in closed-loop control because high accuracy should lead to zero error at the controller input.

Suitable values of P and I determine the step and ramp parts of the response respectively and have to be calculated for each inverter-motor-load combination.

• The values of P and I for the speed loop are dependent on the motor flux, load friction and inertia as they influence the response of speed to current.

• The values of P and I for the current loop depend on the inverter gain, motor resistance and leakage inductance, since they influence the response of current to the motor frequency.

In modern digital drives, the P and I values for both current and speed loops can be set by keypad or, alternatively, most modern digital drives usually include an algorithm for self-tuning. This removes the difficulties of 'tuning the loops', which was traditionally necessary with older analog DC drives. The P and I gains of the speed loop can be setup during commissioning to meet application requirements and seldom need to be changed.

There are a number of disadvantages of the vector controlled AC drive, when compared to a DC drive:

• The vector controller is far more complex and expensive when compared to the simple cascade controller of a DC drive.

• Encoder speed feedback is usually necessary to obtain accurate feedback of the motor shaft speed. Fitting these encoders to a standard squirrel cage AC induction motor is often difficult and makes the motor more expensive. In recent years, 'Sensorless' vector control has been developed where an encoder is not required. The approximate speed is calculated by the processor from the other available information, such as voltage and current. However, the speed accuracy and dynamic response of these drives is inferior to those using encoders.

• The nature of the drive itself often requires the AC motor to operate at high torque loadings at low speeds. The standard squirrel cage AC induction motor then requires a separately powered cooling fan, installed at the ND end of the motor.

• Regenerative braking is more difficult with a vector drive than with a DC drive. Resistive type dynamic braking systems are most often used with AC vector control drives.

Current-feedback AC VSDs

Methods of measuring current in VSDs

Current feedback is required in AC variable speed drives for a number of purposes:

• Protection, short circuit, ground fault and thermal overload in motor circuits

• Metering, for metering and indication for the process control system

• Control, current limit control and current loop control. Several methods have been developed over the years to measure the current and convert it into an electronic form suitable for the drive controller. The method chosen depends on the required accuracy of measurement and the cost of implementation. The main methods of measurement are as follows:

• Current shunt, where the current is passed through a link of pre-calibrated resistance. The voltage measured across the link is directly proportional to the current passing through it. This method was often used in drives with analog control circuits.

• Hall effect sensor, where the output is a DC voltage, which is directly proportional to the current flowing through the sensor. High accuracy and stability over a wide current and frequency range are amongst the main advantages of this device. This device is commonly used with modern digital control circuits.

The performance of a normal core type current transformer is usually not adequate for power electronic applications because its performance at low frequencies is poor and accuracy of measurement of non-sinusoidal waveforms is inadequate. The main methods of current measurement are described in detail.

Current feedback in general purpose VVVF drives

The primary need for current feedback in general purpose VSDs is inverter switching device protection. During short circuit or ground fault conditions, the device current will rise rapidly. If the power electronic switching device, such as an IGBT, BJT, GTO or MOSFET is not switched off quickly, it will be damaged and will fail. VSD reliability depends on the fast and accurate sensing of over-current conditions.

The secondary need for current feedback is to perform current limiting. Early versions of AC VVVF converters did not have a current limiting feature and would simply shut down if the load became too high, requiring manual reset by an operator. This increased downtime and gave VVVF converters a poor reputation in many industries, where overload trips were common. Modern VSDs use current feedback to limit the output current when high loads are encountered.

Current limiting is not the same as current control. Current control means that the current is being controlled at all times, whether it’s high or low. Current limiting means that some action is taken to stop the current exceeding the desired limit point. This action may be only indirectly related to current, such as a change in frequency or voltage.

A third need for current feedback is to provide a current signal roughly proportional to load. This signal may be used internally by the drive to optimize motor volts/hertz or provide slip compensation, where the frequency is increased slightly as load increases to improve speed regulation. The signal may also be made available for external use, by the user, as a load indication signal. As outlined earlier in this section, the stator current of the motor is only roughly proportional to the mechanical load, since the stator current is the vector sum of the magnetizing current IM and the torque-producing current IR. Motor current feedback can also be used to provide thermal protection of the motor.

This requires a thermal model of the motor to be implemented in the drive control system, using frequency and current feedback and motor parameters to estimate the internal temperature of the motor, using an I^2 t replica in the converter. If current level exceeds a set point for a period of time, the motor protection will trip the drive and give an indication of a motor thermal overload.

Current feedback in high-performance vector drives

High performance drives, such as vector controlled drives, employ field oriented control and require current feedback as an integral part of their control loops. In these cases motor current is not simply limited at a pre-defined level. It’s controlled to match a continuously changing torque demand. The vector components of the stator current in each phase are calculated, which requires current from all three phases. This can be achieved preferably with one hall effect CT in each output phase or alternatively two in the output phases and one on the DC bus. If only two-phase sensors are used, the third phase can be calculated from them, however the bus current sensor is still required for device protection.

High accuracy motor current feedback is also necessary to provide control of motor torque. Torque control is necessary in applications such as rewind/unwind systems, hoists, winches, elevators, positioning systems, etc.

DC bus current feedback

DC bus current feedback is suitable for switching device protection and current limiting in most AC VSDs. To a lesser extent, it can provide some load indication if suitably scaled. However, this is usually only accurate over a narrow range of speeds and loads, as the signal must be synthesized from the bus current waveform. It’s the preferred method in general purpose drives, as it only requires a single current feedback device, reducing complexity and cost.

Robust performance for a large variety of load types can be achieved through careful implementation of DC bus current limiting. This is achieved by controlling the motor frequency to maintain the bus current at or below the preset limit point. E.g., excessive loads may be encountered if a high inertia load is accelerated too quickly. This may occur if the acceleration time on the VSD is set without regard to the load dynamics.

E.g., consider an application where a 22 kW motor would take 10 secs to accelerate a high inertia load at 150% rated torque and current. If the operator sets the acceleration time to 5 seconds, this would require 300% rated torque and around 500% current to accelerate the load. Clearly, a drive rated at 150% current overload won’t be able to achieve the desired acceleration time. In this situation, a modern well designed VSD won’t trip, but will modify its acceleration time to maintain the DC bus current at the current limit point. While the operator may not have been able to achieve the desired acceleration time, this is clearly preferable to the drive tripping on over current every time it starts.

Speed feedback

In closed-loop speed control of electric motors and positioning systems, the speed and position feedback from the rotating system is provided by transducers, which convert mechanical speed or position into an electrical quantity, compatible with the control system.

The following techniques are commonly used today:

• Analog speed transducer, such as a tachometer generator (tacho-generator), which converts rotational speed to an electrical voltage, which is proportional to the speed, and transferred to the control system over a pair of screened wires.

• Digital speed transducer, such as a rotary incremental encoder, which converts speed into a series of pulses, whose frequency is proportional to speed. The pulses are transferred to the control system over one or more pairs of screened wires.

• Digital position transducer, such as a rotary absolute encoder, which converts position into a bit code, whose value represents angular position. The code is transferred digitally to the control system over a screened parallel or serial communications link.

Analog speed transducers are increasingly being replaced by digital devices, which are more compatible with modern digital control systems.