AMAZON multi-meters discounts AMAZON oscilloscope discounts
<< cont. from part 1
VVI (Variable Voltage Inverter)--Input
This design takes the supply voltage (e.g., 230 or 460 V), rectifies it, and sends the variable voltage to the DC bus and then to the inverter section.
The inverter section then inverts (changes DC to AC) the variable voltage DC to a variable-voltage and variable-frequency AC. The inverter section contains power semiconductors such as transistors or thyristors (SCRs).
To deliver variable voltage to the inverter, the input rectifier section or front-end consists of a controllable rectifier-SCRs. The control logic fires the SCRs at the appropriate time during the sine wave, thereby providing the variable voltage to the DC bus. FIG. 30 shows a block diagram of a VVI drive.
FIG. 30. VVI drive block diagram
Some of the advantages of a VVI drive include good speed range, ability to connect multiple motors to the drive (within drive current limitations), and a fairly simple control regulator. However, there are some limitations.
One of the major disadvantages, in terms of AC drives, is the input power factor. It decreases as the speed of the drive/motor decreases. This is due to the "controllable" rectifier front end being constructed of SCRs. This issue is identical to that of DC drives.
Another disadvantage is the inability of the drive to "ride through" a low input voltage situation. The term power loss ride-through is defined as the ability of a drive to ride through a low- or zero-voltage input and still remain in operation. This may be 2 to 3 cycles of the AC Sine wave, or more. (Note: Each AC cycle lasts about 16 ms). At low operating speeds (low hertz output), the SCR rectifier is not constantly keeping the DC bus charged at full potential. This means that the motor has voltage to draw from, in the event of a low line input.
Other disadvantages include the requirement of an input isolation trans former or line reactor. This is needed because of the SCR control technology (line spike generation). In addition, the generation of additional output harmonics is a possible result of the technology being applied to the AC motor.
The VVI drive has one additional disadvantage-the characteristic of low speed motor cogging (shaft pulsations/jerky motion). Though not an issue at mid to high speeds, cogging at low speeds can cause equipment problems. This limitation is best illustrated in FIG. 31.
FIG. 31. VVI voltage and current waveforms
As seen in FIG. 31, the voltage waveform approximates a series of steps. Because of this characteristic, this drive is sometimes called a six-step drive. During low-speed operation (>15-20 Hz), the rotor actually searches for the next available magnetic field in the stator. The result is a jerky rotation of the motor shaft. Because of this, gears or gear reducers connected to the motor shaft will suffer additional friction and wear. At high speeds, the inertia of the motor will provide continuous movement of the motor shaft. Therefore, cogging is not a problem.
As shown in the current waveform, there are several spikes that occur at regular intervals. These spikes, or transients, are caused by the SCRs gating on, or triggering. The DC bus filter circuit (shown by an L and C) does reduce the effects of these spikes, but they are not eliminated. These spikes translate into additional motor heating and inefficiency.
The VVI drive was one of the first AC drives to gain acceptance into the industrial drives market. It may be considered one of the most economical drives in the 25- to 150-HP range if a 6:1 speed range is acceptable (operation from 10-60 Hz). This type of drive is also widely used in high-speed drive applications-400 to 3000 Hz.
CSI (Current Source Inverter)
This type of AC drive (sometimes referred to as current source input) has basically the same components as a VVI drive. The major difference is that it is more of a current-sensitive drive as opposed to a VVI, which is more of a voltage-sensitive drive.
This design also takes the supply voltage (e.g., 230 or 460 V), rectifies it, and sends the variable voltage to the DC bus and then to the inverter section. As with the VVI, the CSI drive inverter section inverts (changes DC to AC) the variable-voltage DC to a variable-voltage and variable-frequency AC. The inverter section is made up of power semiconductors such as transistors or thyristors (SCRs).
To deliver variable voltage to the inverter, the input-rectifier section also consists of a controllable rectifier-SCRs. The control logic fires the SCRs at the appropriate time during the sine wave, thereby providing the vari able voltage to the DC bus. FIG. 32 shows a block diagram of a CSI drive.
FIG. 32. CSI block diagram
Some of the advantages of a CSI drive include high efficiency, inherent short-circuit protection (due to the current regulator within the drive), inherent regenerative capability back to the AC line during overhauling load situations, and the capability of synchronous transfer (bringing other motors on-line during full-voltage output).
However, there are also some limitations. As with a VVI drive, the input power factor decreases as the speed of the drive/motor decreases. Also, this drive has a limited speed range due to low-speed motor cogging (shaft pulsations/jerky motion). This drive is also unable to "ride through" a low input voltage situation. This drive also has the requirement of an input isolation transformer, due to the SCR control technology (line spike generation).
Unlike the VVI drive, the CSI drive cannot operate more than one motor at a time. The motor is an integral part of the drive system and its characteristics must be matched to the drive. (Usually the motor and drive are sold as a complete package.) Multiple motors would cause malfunctions in the drive-control system. In addition, the motor normally requires a feed back device (e.g., tachometer) to provide information to the drive current regulator.
Also related to motors is the requirement for the motor to always be connected to the drive. This means that the drive cannot be tested without the motor connected. In some cases, one additional disadvantage is the drive size. Typically, it is physically larger than other drive types because of internal power components.
As mentioned earlier, the VVI and CSI drives produce low-speed cogging.
This is illustrated in FIG. 33.
FIG. 33. CSI voltage and current waveforms
As seen in FIG. 33, line notching or spikes developed from the gating of SCRs in the drive front end. Compared with a VVI drive, the voltage waveform is somewhat closer to the sine wave voltage required by the motor. The current waveform appears to simulate a trapezoid. In addition, there are times when no current flows. These gaps in current cause the rotor to search for the next available magnetic field in the stator. This characteristic, like that of the VVI, results in jerky rotation of the motor shaft at low speeds (<15-20 Hz).
As with the VVI drive, the DC bus filter circuit (shown by an L) does reduce the effects of these spikes, but they are not eliminated. Here again, these spikes translate into additional motor heating and inefficiency.
CSI drives are the latest addition to the line-up of AC variable-frequency drives. They are usually used in applications requiring 50 HP or larger.
These VFDs are well suited for powering pumps and fans because of the inherent synchronous transfer capability. The cost of a CSI drive may be less than either a VVI or PWM in powering pumps, fans, or similar applications. However, the efficiency of the CSI drive matches that of the DC drive and may not provide a total energy-saving package compared with the PWM drive.
PWM (Pulse-Width Modulated)
As seen earlier, the power conversion principle of this drive is different from that of VVI and CSI. One of the major differences is that of a fixed diode front end, not a controllable SCR front end. This fixed diode bridge provides a constant DC bus voltage. The DC bus voltage is then filtered and sent to the inverter section. Another difference between PWM and the other types is the operation of the inverter section. The inverter in the PWM drive has a dual purpose-it changes fixed-voltage DC to variable voltage AC and changes fixed frequency to variable frequency. In the other types, the inverter's primary purpose is to change the fixed-frequency to a variable-frequency output.
PWM drives use several types of power transistors; IGBTs, and GTOs (gate turn-off-SCRs) are examples. These semiconductors offer the advantages of PWM technology without the expense of commutation circuits. (Com mutation circuits are required to turn off the SCRs once they start con ducting. They are found in early VVI or CSI units.) Another major difference is the actual voltage output of the inverter itself.
The DC bus voltage is fixed and approximately equal to the RMS value of the drive input voltage (e.g., 460 V × 1.414 = 650 V). By chopping or modulating the DC bus voltage, the average voltage (output voltage) is increased or decreased. The output voltage value is controlled by the length of time the power semiconductors actually conduct. As seen earlier, the longer the on time for the semiconductors, the higher the output volt age. The longer the off times occur in the process, the lower the frequency output. Thus the inverter accomplishes both variable voltage and frequency. FIG. 34 shows a block diagram of a PWM drive.
FIG. 34. PWM block diagram
Some of the advantages of a PWM drive include high efficiency, the capability of optional common bus regeneration (operating several inverter sections off of one DC bus), and a wide controllable speed range (in some cases up to 200:1, with no low speed cogging under 20 Hz operation).
The PWM drive offers other advantages, such as power loss ride-through capability, open circuit protection, and constant input power factor. This is due to the fixed diode front end and DC bus inductor. Constant power factor is not seen by CSI, VVI, or DC drives.
Like the VVI drive, the PWM also allows multi-motor operation (within the current capability of the drive). However, there are a few limitations.
Extra hardware is required for line regenerative capability (discussed later in this section). Also, the regulator is more complex than a VVI. However, microprocessor control has nearly eliminated significant economic differences between the two drives.
As mentioned, low-speed cogging is not an issue with PWM drives. This is illustrated in FIG. 35.
FIG. 35. PWM voltage and current waveforms
FIG. 35 has been seen before. Of particular interest is the fact that there is no line notching or spikes developed, thanks to the diode front end. The voltage waveform, which could be superimposed on the modulations, very closely approximates the sine wave voltage required by the motor. If the carrier frequency is high (8-16 kHz), the quality of low-speed operation is improved. The higher the carrier frequency, the smoother the motor operation. (Remember--carrier frequency is the speed at which the power semiconductors are switched on and off.) Another benefit of high carrier frequencies is that of reduced audible noise. The higher the frequency, the less motor noise is generated. Audible motor noise can be an issue with low switching rates (e.g., 1-3 kHz). The current waveform, though it contains some ripple, is the smoothest of the three types of drives. It closely approximates the AC sine wave. The efficiency is therefore very high with little motor heating.
Continued improvements in drive technology have enabled PWM drives to deliver a response almost equal to that of DC servos. High response applications such as machine tools and robots require very precise control of motor speed and torque. PWM flux vector drives provide this type of capability and are covered later in this section.
AC Drives--Braking Methods
Braking methods of DC motors has already been reviewed earlier in this section. In this section, attention will be given to AC drive braking methods, which, for the most part, are similar to DC drive braking methods, with a few exceptions. FIG. 36 is a review of the stopping methods of an AC motor, with a minor variation.
FIG. 36. Braking methods for AC drives
As indicated, the easiest way of bringing an AC motor to a stop is the simple method of coast-to-stop. This is followed by the next fastest means, called ramp-to-stop.
During this method, the drive actually forces the motor down to a stop by systematically reducing the frequency and voltage. This is done in a deceleration ramp format, which is set through a parameter in the drive. It should be noted that the motor will contain energy or inertia that must be dissipated-in this case voltage. The DC bus circuit will have to absorb the back fed voltage. When this happens, the DC bus voltage rises-possibly to a point of a voltage trip (called over-voltage fault or DC bus fault). A typical drive will automatically protect itself by shutting down at ~135% of nominal DC bus value. (For example, a 460-VAC drive will carry ~650 VDC on the bus. The trip point would be ~878 VDC.)
The DC bus of a typical AC drive will take on as much voltage as possible without tripping. If an over-voltage trip occurs, the operator has three choices-increase the deceleration time, add DC injection braking, or add an external dynamic braking package. If the deceleration time is extended, the DC bus has more time to dissipate the energy and stay below the trip point. This may be a trial-and-error approach (keep setting the deceleration time until the drive does not trip). A few of the recent drives offered on the market automatically extend the deceleration time, without an operator having to do so. If 30s is a deceleration time and the drive stops the motor in 45s, the motor cannot be stopped in 30s without DC injection braking or external hardware (e.g., dynamic braking). If a 30-second stop is required by the application, DC injection braking is a possibility.
DC Injection Braking
As the name implies, during this braking process, DC voltage is "injected" into the stator windings for a preset period of time. In doing so, a definite north and south pole is set up in the stator, causing the same type of magnetic field in the rotor. Braking torque (counter torque) is the action that results, bringing the motor to a quicker stop, compared with ramp. The rotor and stator dissipate the energy within itself through heat. This method of braking is usually used in lightly loaded applications, where braking is not often used. Repetitive operation of injection braking can cause excessive heat buildup, especially in high-inertia applications, such as flywheels or centrifuges. Excessive heat can cause permanent damage to the stator windings and rotor core.
Dynamic Braking
If DC injection braking cannot bring the motor to a stop in the required time, then dynamic braking will need to be added. FIG. 37 indicates a typical dynamic braking system for an AC drive.
FIG. 37. AC drive dynamic braking
This form of stopping uses a fixed, high-wattage resistor (or bank of resistors) to transform the rotating energy into heat. When the motor is going faster than commanded speed, the energy is fed back to the DC bus. Once the bus level increases to a predetermined point, the chopper module activates and the excess voltage is transferred to the DB resistor. The chopper is basically a sensor and is constructed of a transistor or IGBT switch device. The DB resistor is not mounted within the drive box or inside a drive cabinet. It is always mounted in an area where the heat developed cannot interfere with the heat dissipation requirements of the drive.
As previous indicated, the main stopping power of a DB system occurs when the resistor is cold, during the first few seconds of the process. Once the resistor heats up, the amount of braking torque diminishes. The number of times per minute DB is engaged will also determine the effectiveness of braking torque. Duty cycle, as it is called, is the number of times per minute the DB resistor is used. Many DB circuits consider a maximum of 10% duty cycle (6s on, 54s off-time to cool).
Regenerative Braking (Four Quadrant)
The process of regenerative braking has already been discussed. However, this type of braking uses different components in the AC drive, compared with DC. The end result is still the same-generation of voltage back to the AC line in synchronization with utility power.
To accomplish this, a second set of reverse-connected power semiconductors is required. Some AC drives use two sets of fully controlled SCRs in the input converter section. The latest AC regenerative drives use two sets of IGBTs in the converter section (some manufacturers term this an active front end).
The reverse set of power components allows the drive to conduct current in the opposite direction (taking the motor's energy, and generating it back to the line). FIG. 38 indicates a block diagram of a regenerative braking (four quadrant) system.
FIG. 38. Regenerative AC drive (two IGBT bridges)
As expected with a four-quadrant system, this unit allows driving the motor in the forward and reverse directions, as well as regeneration in both the forward and reverse directions. The control board contains the microprocessor that controls the status of the forward and reverse IGBT bridges. When the speed of the motor is faster than commanded, the motor's energy is fed back into the DC bus. The regeneration circuit senses the increase in reverse voltage and turns on the reverse IGBT bridge circuit.
In this method, the reverse IGBTs need to be able to conduct in the reverse direction. Therefore, if power is removed from the drive, the microprocessor and the reverse IGBTs would not operate. Therefore, this method would not be used for emergency stop situations. However, one method of working around this issue is to include brake resistor and chopper across the DC bus. This provides the best of both worlds, regeneration and e-stop capability.
Drives (AC)--Torque Control
Up until now, standard PWM voltage-controlled drives have been discussed. In this type of drive, the voltage and frequency applied are the controlling variable, when talking about motor torque produced. Torque produced is actually a product of the amount of slip in the motor. The motor has to have a certain amount of slip present to produce torque. As the motor load increases, slip increases and so does torque. This type of AC drive is termed a volts per hertz drive, primarily because of the two control ling elements-volts and hertz. It is also given the label of a scalar drive.
The drive technology of today has moved beyond a "motor turner" philosophy. Drive systems of today need to accurately control the torque of an AC induction motor. Controlled torque is required by automation systems such as wind-unwind stands, process-control equipment, coating lines, printing, packaging lines, hoists and elevators, extruders, and any place where standard motor slip cannot be tolerated (typically 3-5%). Enter the realm of controlled-slip drives-called flux vector or simply, vector drives.
Flux Vector Drives
One of the basic principles of a flux vector drive is to simulate the torque produced by a DC motor. As indicated in the DC drive section, one of the major advantages of DC Drives, and now Flux Vector Drives -- is full torque at zero speed. Up until the advent of flux vector drives, slip had to occur for motor torque to be developed. Depending on motor design, 30-50 rpm of slip might be needed for torque to be developed. An output frequency of 3-7 Hz may be needed from the drive before the motor actually starts turning. With flux vector control, the drive forces the motor to generate torque at zero speed.
A flux vector drive features field-oriented control-similar to that of a DC drive where the shunt field windings continuously have flux, even at zero speed. The motor's electrical characteristics are simulated in the drive controller circuitry called a motor model. The motor model takes a mental impression of the motor's flux, voltage, and current requirements for every degree of shaft rotation. Due to the way the drive gathers information for the motor model, it would be termed a closed loop drive. Torque is indirectly controlled by the creation of frequency and voltage on the basis of values determined by a feedback device. FIG. 39 indicates a block diagram of a closed loop, flux vector-controlled AC drive.
To emulate the magnetic operating conditions of a DC motor, that is, to perform the field orientation process, the flux vector drive needs to know the spatial angular position of the rotor flux inside the AC induction motor. With flux vector PWM drives, field orientation is achieved by electronic means rather than the mechanical commutator and brush assembly of the DC motor.
FIG. 39. Closed loop flux vector AC drive (block diagram)
During field orientation, information about the rotor status is obtained by feeding back rotor speed and angular position. This feedback is relative to the stator field and accomplished by means of a pulse encoder. A drive that uses speed encoders is referred to as a closed-loop drive. In addition, the motor's electrical characteristics are mathematically modeled with micro processors, processing the data. The electronic controller of a flux vector drive creates electrical quantities such as voltage, current, and frequency.
These quantities are the controlling variables, which are fed through the modulator and then to the AC induction motor. Torque, therefore, is con trolled indirectly.
The advantages of this type of drive include good torque response (<10 ms). Some manufacturers consider this response as the limiting response of standard AC induction motors because of the inherent inertia of the machine. Other advantages include full torque at zero speed (at ~0.5 Hz output).
Note: Special caution must be taken when an "off-the-shelf" motor is used to pro vide full torque at zero speed. A specialized cooling system may be needed, in addition to the internally mounted fan. This is due to drastically reduced airflow. The motor is developing full torque and increased heat buildup.
Accurate speed control is possible because of the pulse tachometer feed back. This speed control approaches the performance of a DC drive. Accurate speed control would be stated as ±5% of rated torque.
Depending on point of view, there may be several drawbacks to this type of control. They may be considered drawbacks when compared with the next version of vector control discussed-sensorless flux vector control).
Sensorless Vector Drives
To achieve a high level of torque response and speed accuracy, a feedback device is normally required. This can be costly and adds complexity to the traditionally simple AC induction motor. Also, a modulator is used, which is a device that simulates the AC sine wave for output to the motor. A modulator, slows down communication between the incoming voltage and frequency signals and the ability of the drive to quickly respond to signal changes.. Although the motor is mechanically simple, the drive is electrically complex. FIG. 40 indicates a simple sensorless flux vector control scheme, which is achieving increased recognition in recent years.
FIG. 40. Sensorless flux vector control block diagram (ABB Inc.)
Sensorless flux vector control is similar to a DC drive's EMF control. In a DC drive, the armature voltage is sensed, and the field voltage is kept at constant strength. With sensorless flux vector control, a modulator is used to vary the strength of the field, which is in reality, the stator.
The role of sensorless flux vector fits generally in between the standard PWM open loop control method and a full flux vector, closed loop control method. As previously indicated, the role of sensorless flux vector control is to achieve "DC-like" performance, without the use of a shaft position feedback device. This method provides higher starting and running torque, as well as smoother shaft rotation at low speed, compared with standard V/Hz PWM drives. Additional DC-like performance comes from benefits such as a wide operating speed range and better motor-speed control during load variations.
As indicated, there are many advantages of sensorless flux vector over standard PWM control, a main advantage being higher starting torque on demand. However, standard sensorless flux vector drives may not accomplish torque regulation or full continuous torque at zero speed, without more complex circuitry. Several drive manufacturers use a software design that estimates rotor and stator flux. The result of the flux calculations (estimations) is current that produces a type of regulated motor torque.
If more accurate torque control is required by the application, even more sophisticated control technology is needed. Though more complex in design, high-speed digital signal processors and advanced micro circuits make the electronics design easier to manage. These newer designs also do not require a feedback device and provide the smooth control of torque, as well as full torque at zero speed.
The Direct Torque Control Method
The idea of vector control without feedback (i.e., open loop control) has been researched for many years. A German doctor, Blaschke, and his colleague Depenbrock published documents in 1971 and 1985 on the theory of field-oriented control in induction machines. The publications also dealt with the theory of direct self control. One manufacturer in particular, ABB Inc., has taken the theory and converted it into a refined hardware and software platform for drive control. The result is similar to an AC sensor less vector drive, which uses a direct torque control scheme. The theory was documented and tested in lab experiments for more than 30 years. How-ever, a practical drive ready for manufacture was not possible until the development of application-specific circuitry.
Specific circuits such as the DSP (digital signal processor) and ASIC (application specific integrated circuit) are imbedded in IC (integrated circuit) chips. These chips perform a certain function in the overall production of direct torque control. The controlling variables are motor magnetizing flux and motor torque.
With this type of technology, field orientation is achieved without feed back using advanced motor theory to calculate the motor torque directly and without using modulation.
There is no modulator used in direct torque control and no need for a tachometer or position encoder for speed or position feedback of the motor shaft. Direct torque control uses the fastest digital signal processing hardware available and a more advanced mathematical understanding of how a motor works.
The result is a drive with a torque response that is as much as 10 times faster than any AC or DC drive. The dynamic speed accuracy of these drives are many times better than any open-loop AC drive. It is also com parable with a DC drive that uses feedback. One drives manufacturer indicates this drive is the first universal drive with the capability of performance like either an AC or DC drive. It is basically the first technology to control the induction motor variables of torque and flux.
FIG. 41 shows a block diagram of direct torque control. It includes the basic building blocks upon which the drive does its calculations, based on a motor model.
FIG. 41. The direct torque control (DTC) method (ABB Inc.)
The two fundamental sections of direct torque control are the torque control loop and the speed control loop. During drive operation, two output-phase current values and the DC bus voltage value are monitored, along with the IGBT switch positions. This information is fed to the adaptive motor model.
The motor model calculates the motor data on the basis of information it receives during a self-tuning process (motor identification).
During this automatic tuning process, the drive's motor model gathers information such as stator resistance, mutual inductance, and saturation coefficients, as well as the motor inertia. In many cases, the motor is operated by the drive automatically for a short period of time, to gather the information required.
The output of this motor model is the representation of actual motor torque and stator flux for every calculation of shaft speed. The values of actual torque and actual flux are fed to their respective comparators, where comparisons are performed every 25 µs.
The optimum pulse selector is a fast digital signal processor (DSP) that operates at a 40-MHz speed. Every 25 µs, the inverter IGBTs are sent information for an optimum pulse for obtaining accurate motor torque. The correct IGBT switch combination is determined during every control cycle.
Unlike standard PWM control, in this control scheme there is no "predetermined" IGBT switching pattern. The main motor control parameters are updated as much as 40,000 times per second. This high-speed processing brings with it static speed control accuracy of ±0.5% without an encoder.
It also means that the drive will respond to changes in motor torque requirements every 2 ms.
The speed controller block consists of a PID controller and a circuit that deals with dynamics of acceleration. The external speed reference signal is compared with the actual speed signal given by the motor model. The resulting error signal is fed to the PID section of the speed controller. The flux reference controller contains circuitry that allows the drive to produce several dynamic motor features. Flux optimization is performing just-in-time IGBT switching. This IGBT switching method is controlled by a hysteresis block, which controls the switching action-when to switch, for how long to switch, and which IGBT switches are to be used. This reduces the resulting audible noise emitted from the motor and reduces energy consumption. In addition, flux braking is also possible, which is a more efficient form of injection braking.
Field-oriented control is a term commonly used by one manufacturer when describing continuous torque control. Similar to the direct torque control method, an advanced motor reference model acquires motor parameters during actual operation. An auto-tuning procedure determines the motor values to be used in the motor reference model. These voltage values are fed back to an adaptive software control block, which controls output cur rent, thereby controlling torque. The proper amount of slip is provided, thereby maintaining field orientation (precise stator flux control).
Drives (AC)--Technical Concerns
SCR and GTO control of AC drive power structures have been around since the 1960s. Forced commutated SCR PWM drives gained increased acceptance in the mid-1970s. This was followed by GTO and bipolar transistor-based PWM drives in the mid-1980s. In the late-1980s, IGBT PWM drives were emerging as the drive to take the variable-speed industry into the 21st century. By the early 1990s, several manufacturers were promoting a full-line of IGBT based AC drives for the industrial, as well as HVAC, marketplace. With these AC drive offerings came several advantages and some challenges. FIG. 42 illustrates to one of the advantages and challenges.
FIG. 42. Voltage reflection/standing waves (ABB Inc.)
As seen in FIG. 42, as the technology era of power semiconductor devices changed, so did the number of circuit boards needed to support that technology. In the 1960s and 1970s, SCRs and GTOs needed more than a dozen circuit boards to support the gating of the power device.
Given the fact that each board had a retail value of $500-$1000, it is easy to understand the high cost of AC drives in that era. Separate gate driver boards were needed for each SCR or GTO device to turn off the device and control timing circuitry. In relative terms, the device turn-on time was rather slow, compared with the other emerging technologies. With slower turn-on or switch times, the drives caused audible noise in the motor of 500-1000 Hz, quite a noticeable level. The laminations in the stator winding vibrated at the switch frequency, producing the noise much like an audio speaker.
With the advent of bipolar transistors came the requirement for fewer boards. Less sophisticated control circuitry was required since there was no need for separate gate driver boards. Fewer circuit boards meant less overall cost for the drive. In addition, the relative size of the drive was reduced, compared with SCR-based products. On the positive side, the bipolar transistor switched 3-6 times faster than SCRs or GTOs (in the 1- to 3-kHz range). This meant that the audible noise was also reduced to a more tolerable level.
When IGBT technology emerged in the early 1990s, it was considered the power technology of the future. The device switched over 10 times faster than bipolar transistors (3-12 kHz), which meant a drastic reduction in audible noise. The circuit board count was reduced to two. The control board contained all the circuits for timing and signal processing. The motor control board contained all the circuits to turn on and off the device. With only two circuit boards needed, the drives industry realized the lowest cost drive possible. There was also a reduction in size to about 1/3 that of bipolar transistor drives. With this technology advancement came an acute challenge for the device connected to the drive-the motor.
With the extremely fast switching times, came the rise of a phenomenon called voltage reflection. Voltage reflection is caused by the fast-rising voltage waveform versus unit of time. In essence, the IGBT turns on immediately compared with 30 times longer with other devices. When a drive switches at this high rate, a reflected wave back from the motor adds to the voltage leaving the output of the drive. The result is a voltage at the motor terminals greater than the original voltage output from the drive. This is illustrated in FIG. 43.
FIG. 43. Voltage reflection characteristics
As seen in FIG. 43, this situation is more of an issue when an impedance mismatch exists between the drive output/motor cables and the motor terminals. The phenomena is similar to the standing wave ratio (SWR) that exists in citizens band (CB) radio antenna setups. The coil of the CB antenna must be installed and tuned correctly, so that there are no waves reflected back to the transmitter, which could cause damage. If tuned properly, the antenna absorbs all the energy the transmitter can deliver.
The amount of increased voltage at the motor terminals is a function of the drive output voltage, length of motor cable, and the amount of mis match. This situation is a possibility more often in smaller motors, which have a higher impedance compared with motors in the 150-HP range or more. In some cases, it is not uncommon to see more than twice the drive output voltage at the motor terminals. Many 460-V motor insulation systems are not designed to handle that type of spike voltage. The motor volt age spike issue can be seen in FIG. 44.
FIG. 44. Motor terminal voltage (Trans-Coil, Inc.)
The spike voltages created in this particular case are close to 1500 V (460 V drive). Because AC, IGBT drives are installed at significant distances away from the motor, the impedance mismatch can be present for various types and brand names of motors. For example, it has been determined that a typical IGBT drive would cause twice the output voltage at the motor terminals. This would be true if the motor is installed greater than 75 feet away from the drive.
There are several possibilities in protecting motors against damage or coping with the issue. On new drive installations, verify that motors installed a significant distance away from the drive meet NEMA MG1, part 31.4.4.2 standards. These motors are designed with insulation systems that are able to handle the over-voltage stress. FIG. 45 indicates the construction of a random wound versus a form wound motor. The concentric wound or form wound motor is designed to handle spike voltages generated.
FIG. 45. Random wound vs. concentric wound motors
Motors that meet the MG1 standard are concentric wound and are termed inverter duty motors. These motors contain stator windings that are care fully formed around the stator slots, so that the first winding turn is not next to the last winding turn. Voltage spikes poke minute holes in the insulation. When that occurs in a random wound motor, the likelihood is that the first and last turn are next to each other. A voltage spike hole would therefore short out the winding and make the motor useless until rewound. Inverter duty motors also have extra slot paper insulation separating the windings of different phases. In addition, these motors are typically dipped in lacquer insulation after the windings are complete to add to the insulation strength and cover insulation holes that may have occurred.
Some inverter duty motors are actually dipped a second time to improve the dielectric (insulation) strength.
Another means of protecting the motor against possible damage is to install output reactors, (similar to line reactors) at the output of the drive.
The drive manufacturer can make recommendations. Usually 1.5-3% impedance will protect existing motors to about 500 feet. If distances greater than 500 feet are encountered, dv/dt filters can be installed at the output of the drive. These filters are usually effective up to distances of about 2000 feet. (Note: dv/dt means change of voltage vs. change in time.) This is a special resistor-inductor-capacitor filter designed to drastically reduce the over-voltage spikes at the drive output.
Additional precautions include installing a sine filter at the output of the drive, which is not limited to motor distance. In addition, a snubber circuit installed at the motor will have over-voltage reduction similar to dv/dt filters. Snubber circuits do not usually have any distance limitation. FIG. 46 shows a reduction in spike voltage generation with the installation of an output dv/dt filter.
FIG. 46. Effects of a dv/dt filter on voltage reflection (Trans-Coil,
Inc)
At long motor cable lengths (e.g., 250 to 300 ft or more), another phenomenon can occur - that of capacitive coupling. Conductors separated by an insulator make up a capacitor. With additional capacitance at the VFD output, higher current is calculated by the VFD. To the VFD, the motor will appear to consume increased amounts of current, which can cause nuisance "overcurrent" trips. This condition can be exaggerated by higher IGBT switch frequencies and the lack of output snubber circuits. In most cases, the motor is consuming appropriate current, but the motor cabling causes inappropriate readings by the VFD. Typically, an output reactor, as previously indicated, can serve a dual purpose-protect motor insulation from damage, and improve the VFD current calculations. This ultimately results in more stable VFD performance with less nuisance faults. Some manufacturers include hidden parameters to assist in tuning up the VFD current sensing circuit.
Harmonics Generation Harmonics are basically a distortion of the original waveform. In the case of AC drives, harmonics are a distortion of the three-phase waveform, with the harmonic components fed back onto the AC power line. Harmonics are caused by the fact that AC drives draw current from the supply line in "bursts" rather than in a "linear" fashion. Because of this characteristic, AC drives are considered nonlinear loads. As a matter of fact, any electronic device that draws nonlinear currents causes a certain amount of harmonics.
VCRs, big screen TVs, stereos, and laptop and desktop computers all fall into the nonlinear load category. They include a switch mode power supply that changes AC to DC-a rectifier. Because rectifiers draw current in bursts, they create harmonic currents, which are fed back to the power source.
A six diode bridge AC drive (called a 6 pulse drive) produces the 5th, 7th, 11th, and 13th harmonic. The values of these harmonics is enough to distort the AC supply waveform. An example of distortion created by the 5th and 7th harmonic is illustrated in FIG. 47.
Note: The fundamental frequency is 60 Hz. Harmonic frequencies are multiples of the 60-Hz waveform (e.g., 5th = 300 Hz, 7th = 420 Hz, and so on). Users of AC drives need to be concerned about harmonics, which are generated back to the line supply. Harmonic currents do not provide any useful work. Harmonic current distortion generates additional heating in transformers and cables, reducing the available capacity of the equipment.
Current distortion can also create resonance conditions between the line supply reactance and power factor correction capacitors (if used). In addition, the high frequencies of harmonics can cause electronic interference with telephone and telecommunications equipment.
FIG. 47. Distortion caused by the 5th and 7th harmonic (Trans-Coil,
Inc.)
Harmonic voltage distortion causes increased heating in motors. Voltage distortion can also cause malfunctions in sensitive communications and computer equipment. In many areas, local electrical codes or drive-installation specifications require compliance with IEEE 519-1992. The locations where voltage and current harmonics can be an issue are termed the point of common coupling or simply, PCC.
The local power utility deals with the PCC, where the customer's building connects directly with utility power. This is the current harmonic distortion concern and is termed the total demand distortion or simply, TDD. The utility customer is faced with the voltage harmonic distortion concern, where non linear loads meet other loads, such as linear loads (inductors, line operated motors). FIG. 48 indicates the locations of concern for both harmonic current and voltage distortion.
Overall, harmonics are a system issue. Harmonics that are produced by an individual drive are only important when they represent a significant portion of the total system. For example, if the drive load on a transformer represents over 1/4th of the kVA load, then harmonics could be an issue and requires further investigation. If the total drive load is 5 HP on a 1000 kVA transformer, harmonics would not be an issue. It is worth noting that the addition of linear loads, such as line-operated motors, tend to reduce the overall system harmonic levels.
FIG. 48. PCC and harmonic distortion (ABB Inc.)
IEEE 519-1992 indicates limits of THD (voltage distortion) as 5% for general systems (e.g., factories and general office buildings, not including hospitals, airports, and power systems dedicated to drive loads). Current distortion limits (TDD) are based on a ratio. The ratio is the short-circuit current available at the PCC divided by the maximum fundamental load current. Therefore the limits will vary on the basis of the amount of the electrical current tank available. If the current capacity ratio is high, the allowable TDD will be high, compared with a low ratio. (Example: If a 10 lb rock is dropped in a bathtub full of water, the waves created represent the current harmonics generated-quite a significant amount. If that same rock is dropped off the Golden Gate bridge in San Francisco, the amount of waves hitting the shoreline would be almost non-existent-no significant current harmonics generated.) It is important to note that an 80% THD nonlinear load will result in only a 8% TDD if the non-linear load is 10% of the total system load. With that in mind, there are several ways of reducing (mitigating) harmonics.
Using the above 80% THD example, the following comparisons could be drawn. Line reactors could be added to the input of the drive. A 5% line reactor (or equivalent DC bus inductor) could drop the THD from 80% down to 28%. Adding a 5th harmonic trap filter to the line reactor could drop the THD down to 13%. The cost of a line reactor may be 15-25% the cost of the drive (depending on drive horsepower). A harmonic trap filter could add 25-50% the cost of the drive (depending on drive horsepower).
Beyond these techniques, more serious mitigation could be realized, including higher associated costs. A 12-pulse drive input rectifier could be installed, along with a 5% impedance transformer. (Note: A 12-pulse drive is two six-diode bridge rectifiers, with a special delta-delta-wye input transformer, which could be 1/2 the cost of the drive unit itself. A 12-pulse drive effectively reduces the 5th and 7th harmonic. The total cost of the drive and transformer could be about double that of a six-pulse drive, depending on horsepower.) With this configuration, the THD could be reduced to 8%. With a 12-pulse drive, a 5% impedance transformer and an 11th harmonic trap filter, a reduction of THD down to 4% could be realized. Installing an active harmonic filter would reduce the THD down to 3%.
An active harmonic filter is essentially a regenerative drive. As stated above, this type of drive would yield the highest amount of harmonic mitigation of all the techniques. It could also be over twice the cost of a standard 6-pulse drive.
When dealing with harmonics, it is helpful to work with the drive's manufacturer or a company specializing in harmonic mitigation techniques.
Some drive manufacturers offer a "no-cost" analysis service-submitting a harmonics report on the basis of the installation of their drive in a specific system. Harmonics will be even more of an issue in the future, with VFD's being applied in applications traditionally deemed fixed speed. The positive side to harmonics is that cost-effective techniques are available for a wide variety of installations.
Power Factor
When discussing electronic power conversion equipment, there are two ways to identify power factor: displacement and total or true power factor.
Displacement PF is the power factor of the fundamental components of the input line voltage and current. Total PF indicates the effects of harmonic distortion in the current waveform. No matter how PF is viewed, the power utility imposes penalties for customers that use equipment with a poor PF.
Displacement PF for an AC drive is relatively constant. It is approximately 0.96-0.97. This value is primarily independent of the speed of the motor and its output power. The harmonic current distortion is determined by the total values of inductance, capacitance, and resistance, from power source to the load. The power distribution system has impedance, which also enters into the calculations for total PF. FIG. 49 gives a general indication of displacement and total PF for AC as well as DC drives.
The curves shown in FIG. 49 are for drives at full load and constant torque operation. The shaded area indicates a variation of total PF for typical drive installations. For an AC drive, the total (true) PF varies from roughly 0.94 at rated load, down to below 0.75 when under a light load.
As you may recall, a lagging PF is seen for an AC induction motor used in a power system. The AC drive does an effective job in isolating the input power source from the lagging PF at which the motor operates. In a certain sense, AC drives could be considered PF correctors by means of its isolation from the AC line. Because of this fact, PF correction is not normally applied to AC, PWM drives. If existing PF capacitors are connected of the motor, they must be removed when an AC drive is installed in place of a full-voltage starter. PF capacitors between the drive output and the motor can cause physical damage to the drive output IGBTs.
FIG. 49. Total and displacement PF (DC and AC drives)