Power electronic converters (part 2)

Commutation

The transitional period from blocking to conducting, and vice versa, is called commutation and the period during which a component turns on/off, is called the commutation period. During commutation, the component comes under electrical stress due to changes in the circuit conditions and the thermal stress due to losses. These losses produce heat in the component and also stress the insulation and current paths.

• In the blocking mode, losses are usually small and mainly due to the leakage current flowing through the device

• In the conducting mode, losses are relatively higher and mainly due to the current and forward volt drop across the component (I2 R losses)

• During commutation, losses are due to the transitional voltage and current activity within the component and in the control circuit to trigger the gate.

++++ 5 illustrates thyristor commutation for both the turn-on and the turn-off periods.

++++ 5: Simple commutation of an electronic switch

In modern PWM inverters, there is a tendency to use electronic switches operating at high switching frequencies to achieve faster responses or better output wave-shapes.

Unfortunately, the increased number of commutations results in higher losses both in the triggering circuits as well as the power circuits of the components.

Losses may be reduced by using devices that have the following characteristics:

• Low leakage current during blocking

• Low forward volt drop during conduction

• High switching speed, short commutation period

• Low triggering losses in the control circuit

Power electronic rectifiers (AC/DC converters)

The first stage of an AC frequency converter is the conversion of a 3-phase AC power supply to a smooth DC voltage and current. Simple bi-stable devices, such as the diode and thyristor, can effectively be used for this purpose.

Initially, when analyzing power electronic circuits, it will be assumed that the bi-stable semiconductor devices, such as the diodes and thyristors, are ideal switches with no losses and minimal forward voltage drop. It will also be assumed that the reactors, capacitors, resistors, and other components of the circuits have ideal linear characteristics with no losses. Once the operation of a circuit is understood, the imperfections associated with the practical components can be introduced to modify the performance of the power electronic circuit.

In power electronics, the operation of any converter is dependent on the switches being turned ON and OFF in a sequence. Current passes through a switch when it’s ON and is blocked when it’s OFF. As mentioned above, the word commutation is used to describe the transfer of current from one switch turning OFF to another turning ON. In a diode rectifier circuit, a diode turns ON and starts to conduct current when there is a forward voltage across it, i.e. the forward voltage across it becomes positive. This process usually results in the forward voltage across another diode becoming negative, which then turns off which stops conducting current. In a thyristor rectifier circuit, the switches additionally need a gate signal to turn them on and off.

++++ 6: Simple circuit to illustrate commutation from diode D1 to D2

The factors affecting commutation may be illustrated in the idealized diode circuit, which shows two circuit branches, each with its own variable DC voltage source and circuit inductance. Assume, initially, that a current I is flowing through the circuit and that the magnitude of the voltage V1 is larger than V2. Since V1 > V2, diode D1 has a positive forward voltage across it and it conducts a current I1 through its circuit inductance L1. Diode D2 has a negative forward voltage across it and is blocking and carries no current.

I1 = 1

I2 = 0

Suppose that voltage V2 is increased to a value larger than V1, the forward voltage across diode D2 becomes positive and it then starts to turn on. However, the circuit inductance L1 prevents the current I1 from changing instantaneously and diode D1 won’t immediately turn off. So, both diodes D1 and D2 remain ON for an overlap period called the commutation time tc.

With both diodes turned on, a closed circuit is established which involves both branches. The effective circuit voltage VC = (V2 - V1), called the commutation voltage, then drives a circulating current ic, called the commutation current, through the two branches which have a total circuit inductance of LC = (L1 + L2). In this idealized circuit, the volt drop across the diodes and the circuit resistance have been ignored. From basic electrical theory of inductive circuits, the current ic increases with time at a rate dependent on the circuit inductance. The magnitude of the commutation current may be calculated from the following equations.

If the commutation starts at a time t1 and finishes at a time t2, the magnitude of the commutation current I_C at any time t, during the commutation period, may be calculated by integrating the above equation from time t1 to t.

During the commutation period:

• It’s assumed that the overall current through the circuit remains constant.

I = (I1 +I2) constant

As the circulating commutation current increases:

• Current (I2) through the diode that is turning on increases in value

I2 = I_c increasing

• Current (I1) through the diode that is turning off decreases in value

I1 = I – I_c decreasing

++++ 7: The currents in each branch during commutation

For this special example, it can be assumed that the commutation voltage VC is constant during the short period of the commutation. At time t the integration yields the following value of IC, which increases linearly with time.

When IC has increased to a value equal to the load current I at time t2, then all the current has been transferred from branch 1 to branch 2 and the current through the switch that is turning off has decreased to zero. The commutation is then over.

Consequently, at time t2

I1 = 0

I2 = I_c = I

At the end of commutation when t = t2, putting I_C equal to I in the above equation, the time taken to transfer the current from one circuit branch to the other (commutation time), may be calculated.

It’s clear from this equation that the commutation time t_c depends on the overall circuit inductance (L1 + L2) and the commutation voltage.

From this we can conclude that:

• A large circuit inductance will result in a long commutation time.

• A large commutation voltage will result in a short commutation time.

In practice, a number of deviations from this idealized situation occur.

• The diodes are not ideal and don’t turn off immediately when the forward voltage becomes negative. When a diode has been conducting and is then presented with a reverse voltage, some reverse current can still flow for a few microseconds. The current I1 continues to decrease beyond zero to a negative value before returning to zero. This is due to the free charges that must be removed from the PN junction before blocking is achieved.

• Even if the commutation time is very short, the commutation voltage of an AC fed rectifier bridge does not remain constant but changes slightly during the commutation period. An increasing commutation voltage will tend to reduce the commutation time.

In practical power electronic converter circuits, commutation follows the same basic sequence outlined above. A typical 6-pulse rectifier bridge circuit to convert 3-phase AC currents IA, IB and IC, to a DC current ID.

++++ 3-Phase commutation with a 6-pulse diode bridge

This type of circuit is relatively simple to analyze because only 2 of the 6 diodes conduct current at any one time. The idealized commutation circuit can easily be identified. In this example, commutation is assumed to be taking place from diode D1 to D3 in the positive group, while D2 conducts in the negative group.

In power electronic bridge circuits, it’s conventional to number the diodes D1 to D6 in the sequence in which they turned ON and OFF. When VA is the highest voltage and VC the lowest, D1 and D2 are conducting.

In a similar way to the idealized circuit, when VB rises to exceed VA, D3 turns on and commutation transfers the current from diode D1 to D3. As before, the commutation time is dependent on the circuit inductance (L) and the commutation voltage (VB - VA). As can be seen from the 6-pulse diode rectifier bridge example, commutation is usually initiated by external changes. In this case, commutation is controlled by the 3-phase supply line voltages. In other applications, commutation can also be initiated or controlled by other factors, depending on the type of converter and the application. Therefore, converters are often classified in accordance with the source of the external changes that initiate commutation.

• In the above example, the converter is said to be line commutated because the source of the commutation voltage is on the mains supply line.

• A converter is said to be self-commutated if the source of the commutation voltage comes from within the converter itself. Gate-commutated converters are typical examples of this.

Line commutated diode rectifier bridge

One of the most common circuits used in power electronics is the 3-phase line commutated 6-pulse rectifier bridge, which comprises 6 diodes in a bridge connection.

Single-phase bridges won’t be covered here because their operation can be deduced as a simplification of the 3-phase bridge.

In the analysis of the various types of converter that follow, the procedure will be to assume initially that the conditions and components are ideal. Once the principles have been established, any deviations from the ideal will be discussed. The following ideal assumptions are made:

• The supply voltages are 'stiff' and completely sinusoidal
• Commutations are instantaneous and have no recovery problems
• Load currents are completely smooth
• Transformers and other line components are linear and ideal
• There is no volt drop in power electronic switches

These assumptions are made to gain an understanding of the circuits and to make estimates of currents, voltages, commutation times, etc. Thereafter, the limiting conditions that affect the performance of the practical converters and their deviation from the ideal conditions will be examined to bridge the gap from the ideal to the practical.

In the diode bridge, the diodes are not controlled from an external control circuit.

Instead, commutation is initiated externally by the changes that take place in the supply line voltages, hence the name line commutated rectifier.

According to convention, the diodes are labeled D1 to D6 in the sequence in which they are turned ON and OFF. This sequence follows the sequence of the supply line voltages.

++++ Line commutated diode rectifier bridge.

The 3-phase supply voltages comprise 3 sinusoidal voltage waveforms 120° apart which rise to their maximum value in the sequence A - B - C. According to convention, the phase-to-neutral voltages are labeled VA, VB and VC and the phase-to-phase voltages are VAB, VBC and VCA, etc.

These voltages are usually shown graphically as a vector diagram, which rotates counter-clockwise at a frequency of 50 times per second. A vector diagram of these voltages and their relative positions and magnitudes is shown below. The sinusoidal voltage waveforms of the supply voltage may be derived from the rotation of the vector diagram.

++++ Vector diagram of the 3-phase mains supply voltages.

The output of the converter is the rectified DC voltage VD, which drives a DC current ID through a load on the DC side of the rectifier. In the idealized circuit, it’s assumed that the DC current ID is constant and completely smooth and without ripple.

The bridge comprises two commutation groups, one connected to the positive leg, consisting of diodes D1-D3-D5, and one connected to the negative leg, consisting of diodes D4-D6-D2. The commutation transfers the current from one diode to another in sequence and each diode conducts current for 120° of each cycle.

In the upper group, the positive DC terminal follows the highest voltage in the sequence VA-VB-VC via diodes D1-D3-D5. When VA is near its positive peak, diode D1 conducts and the voltage of the +DC terminal follows VA. The DC current flows through the load and returns via one of the lower group diodes. With the passage of time, VA reaches its sinusoidal peak and starts to decline. At the same time, VB is rising and eventually reaches a point when it becomes equal to and starts to exceed VA. At this point, the forward voltage across diode D3 becomes positive and it starts to turn on. The commutating voltage in this circuit, VB-VA starts to drive an increasing commutation current though the circuit inductances and the current through D3 starts to increase as the current in D1 decreases. In a sequence of events similar to that described above, commutation takes place and the current is transferred from diode D1 to diode D3. At the end of the commutation period, diode D1 is blocking and the +DC terminal follows VB until the next commutation takes place to transfer the current to diode D5. After diode D5, the commutation transfers the current back to D1 and the cycle is repeated.

In the lower group, a very similar sequence of events takes place, but with negative voltages and the current flowing from the load back to the mains. Initially, D2 is assumed to be conducting when VC is more negative than VA. As time progresses, VA becomes equal to VC and then becomes more negative. Commutation takes place and the current is transferred from diode D2 to D4. Diode D2 turns off and D4 turns on. The current is later transferred to diode D6, then back to D2 and the cycle is repeated.

The conducting periods of the diodes in the upper and lower groups are shown over several cycles of the 3-Phase supply. This shows that only 2 diodes conduct current at any time (except during the commutation period, which is assumed to be infinitely short!!) and that each of the 6 diodes conducts for only one portion of the cycle in a regular sequence. The commutation takes place alternatively in the top group and the bottom group.

The DC output voltage VD is not a smooth voltage and consists of portions of the phase to-phase voltage waveforms. For every cycle of the 50 Hz AC waveform (20 msec), the DC voltage VD comprises portions of the 6 voltage pulses, VAB, Vac, VBC, VBA, VCA, VCB, etc, hence the name 6-pulse rectifier bridge.

The average magnitude of the DC voltage may be calculated from the voltage waveform shown above. The average value is obtained by integrating the voltage over one of the repeating 120 dgr. portions of the DC voltage curve. This integration yields an average magnitude of the voltage VD as follows.

VD = 1.35 × (RMS - Phase Voltage)

VD = 1.35 × VRMS

E.g., if VRMS = 415 volts, VD = 560 volts DC

If there is sufficient inductance in the DC circuit, then the DC current ID will be fairly steady and the AC supply current will comprise segments of DC current from each diode in sequence. As an example, the current in the A-phase is shown. The non sinusoidal current that flows in each phase of the supply mains can affect the performance of other AC equipment connected to the supply line that are designed to operate with sinusoidal waveforms. The effects of the non-sinusoidal currents: Electromagnetic compatibility (EMC).

++++ Voltage and current waveforms during commutation

In practice, to ensure that the diode reverse blocking voltage capability is properly specified, it’s necessary to know the magnitude of the reverse blocking voltage which appears across each of the diodes. Theoretically, the maximum reverse voltage across a diode is equal to the peak of the phase-phase voltage. E.g., the reverse voltage VCA and VCB appears across diode D5 during the blocking period. In practice, a factor of safety of 2.5 is commonly used for specifying the reverse blocking capability of diodes and other power electronic switches. On a rectifier bridge fed from a 415 V power supply, the reverse blocking voltage V_bb of the diode must be higher than 2.5 × 440 V = 1100 V. Therefore, it’s common practice to use diodes with a reverse blocking voltage of 1200 V.

The line commutated thyristor rectifier bridge

The output DC voltage and operating sequence of the diode rectifier above is dependent on the continuous changes in the supply line voltages and is not dependent on any control circuit. This type of converter is called an uncontrolled diode rectifier bridge because the DC voltage output is not controlled and is fixed at 1.35 × VRMS. If the diodes are replaced with thyristors, it then becomes possible to control the point at which the thyristors are triggered and therefore the magnitude of the DC output voltage can be controlled. This type of converter is called a controlled thyristor rectifier bridge and requires an additional control circuit to trigger the thyristor at the right instant. A typical 6-pulse thyristor converter.

From the previous section, the conditions required before a thyristor will conduct current in a power electronic circuit are:

• A Forward Voltage must exist across the thyristor and

• A Positive Pulse must be applied to the thyristor gate

++++ 6-pulse controlled thyristor rectifier bridge

If each thyristor were triggered at the instant when the forward voltage across it tends to become positive, then the thyristor rectifier operates in the same way as the diode rectifier described above. All the voltage and current waveforms of the diode bridge apply to the thyristor bridge. A thyristor bridge operating in this mode is said to be operating with a zero delay angle and gives a voltage output of:

VD = 1.35 × VRMS

The output of the rectifier bridge can be controlled by delaying the instant at which the thyristor receives a triggering pulse. This delay is usually measured in degrees from the point at which the switch CAN turn on, due to the forward voltage becoming positive.

The angle of delay is called the delay angle, or sometimes the firing angle, and is designated by the symbol a. The reference point, for the angle of delay, is the point where a phase voltage wave crosses the voltage of the previous phase and becomes positive relative to it. A diode rectifier can be thought of as a converter with a delay angle of a = 0°.

The main purpose of controlling a converter is to control the magnitude of the DC output voltage. In general, the bigger the delay angle, the lower the average magnitude of the DC voltage. Under steady state operation of a controlled thyristor converter, the delay angle for each switch is the same. ++++ the voltage waveforms where the triggering of the switches has been delayed by an angle of a degrees.

++++ Voltage waveforms of a controlled rectifier

In the positive switch group, the positive DC terminal follows the voltage associated with the switch that is in conduction in the sequence VA-VB-VC. Assume, initially, that thyristor S1 associated with voltage VA is conducting and S3 is not yet triggered. The voltage on the + bus on the DC side follows the declining voltage VA because, in the absence of S3 conduction, there is still a forward voltage across S1 and it will continue to conduct. When S3 is triggered after a delay angle = a, the voltage on + bus jumps to VB, whose value it then starts to follow. At this instant, with both S1 and S3 conducting, a negative commutation voltage equal to VB-VA appears across the switch S1 for the commutation period, which then starts to turn off. With the passage of time, VB reaches its sinusoidal peak and starts to decline, followed by + DC terminal. At the same time, VC is rising and when S5 is triggered, the same sequence of events is repeated and the current is commutated to S5.

As with the diode rectifier, the average magnitude of the DC voltage VD can be calculated by integrating the voltage waveform over a 120 dgr. period representing a repeating portion of the DC voltage. At a delay angle a, the DC voltage is given by:

VD = 1.35 × (RMS - Phase Voltage) × Cos a

VD = 1.35 × VRMS × Cos a

This formula shows that the theoretical DC voltage output of the thyristor rectifier with a firing angle a = 0 is the same as that for a diode rectifier. It also shows that the average value of the DC voltage will decrease as the delay angle is increased and is dependent on the cosine of the delay angle. When a = 90 dgr , then Cos a = 0 and VD = 0, which means that the average value of the DC voltage is zero. The instantaneous value of the DC voltage is a saw-tooth voltage as shown below.

++++ DC output voltage for delay angle a = 90 deg.

If the delay angle is increased further, the average value of the DC voltage becomes negative. In this mode of operation, the converter operates as an inverter. It’s interesting to note that the direction of the DC current remains unchanged because the current can only flow through the switches in the one direction. However, with a negative DC voltage, the direction of the power flow is reversed and the power flows from the DC side to the AC side. Steady state operation in this mode is only possible if there is a voltage source on the DC side. The instantaneous value of the DC voltage for a > 90 dgr. In practice, the commutation is not instantaneous and lasts for a period dependent on the circuit inductance and the magnitude of the commutation voltage. As in the idealized case, it’s possible to estimate the commutation time from the commutation circuit inductance and an estimate of the average commutation voltage.

++++ 15: DC voltage when the delay angle a > 90 dgr.

As in the diode rectifier, the steady DC current ID comprises segments of current from each of the 3 phases on the AC side. On the AC side, the current in each phase comprises non-sinusoidal blocks, similar to those associated with the diode rectifier and with similar harmonic consequences. In the case of the diode bridge, with a delay angle of a = 0, the angle between the phase current and the corresponding phase voltage on the AC side is roughly zero. Consequently, the power factor is roughly unity and converter behaves something like a resistive load.

For the controlled rectifier, with a delay angle of a, the angle between the phase current and the corresponding phase voltage is also roughly a, but normally called the power factor angle Ø. This angle should be called the displacement factor because it does not really represent power factor (see later). Consequently, when the delay angle of the thyristor rectifier is changed to reduce the DC voltage, the angle between the phase current and voltage also changes by the same amount. The converter then behaves like a resistive-inductive load with a displacement factor of Cos Ø. It’s well known that the power factor associated with a controlled rectifier falls when the DC output voltage is reduced.

===

Delay angle -- Converter behavior

===

A common example of this is a DC motor drive controlled from a thyristor converter. As the DC voltage is reduced to reduce the DC motor speed at constant torque, the power factor drops and more reactive power is required at the supply line to the converter.

++++ 16: Reactive power requirements of a DC motor drive with a constant torque load fed from a line-commutated converter.

++++ 17 summarizes the possible vector relationships between the phase voltage and the fundamental component of the phase current in the supply line for the various values of delay angle a.

++++ 17: Vector diagram of phase voltage and fundamental current for a controlled thyristor rectifier bridge.

The phase current on the AC side is, fundamentally, a non-sinusoidal square wave. By applying the principles of harmonic analysis, using the Fourier transform, this non sinusoidal wave can be resolved into a fundamental (50 Hz) sinusoidal wave plus a number of sinusoidal harmonics. The fundamental waveform has the highest amplitude and therefore the most influence on the power supply system. In a 6-pulse rectifier bridge, the 5th harmonic has the highest magnitude, theoretically 20% of the fundamental current.

++++ 18: The fundamental current and the 5th harmonic current.

The RMS value of the fundamental current can be calculated from the following formula, which is derived from fundamental principles:

The corresponding apparent power S1 kVA is given by:

S1 = 1.35 VRMS ID kVA

The active power component is given by:

P1 = S1 Cos f kW

P1 = 1.35 VRMS ID kW

This confirms that the active power calculated on the AC side is identical to the power calculated for the DC side (VD.ID), since from the previous formula VD = 1.35 VRMS Cos a.

The reactive power component is given by:

Q1 = S1 Sin f kVAr

Q1 = 1.35 VRMS ID Sin f kVAr

This formula illustrates that, if the load current is held constant (constant torque load on a DC motor), the reactive power will increase in proportion to Sin a as the triggering delay angle is increased.

Looking at the rectifier from the 3-phase supply, an effective phase-to-phase short circuit occurs across the associated supply lines during commutation, when the 2 sequential switches are conducting. E.g., when switch S3 is triggered and switch S1 continues to conduct, the voltage of VA and VB must be equal at switches themselves (except for the small volt drop across the switches). The commutation voltage VB-VA drives a circulating current through S1 and S3 and the circuit inductance 2L. Depending on the delay angle, the commutation voltage can be quite large. At the voltage source, the magnitude of the voltages VA and VB are depressed during this period by an amount dependent on the circulating current and circuit inductance. This additional non-desirable effect in the supply line is called voltage notching. The effect of notching is to slightly reduce the DC voltage VD, but this reduction is very small and may be ignored. However, notching is important when considering the losses in the converter.

++++ 19: Voltage notching in the supply line

Practical limitations of line commutated converters

The above analysis covers the theoretical aspects of uncontrolled and controlled converters. In practice, the components are not ideal and the commutations are not instantaneous. This results in certain deviations from the theoretical performance.

One of the most important deviations is that the DC load current is never completely smooth. The reason for this is fairly obvious. Accepting that the instantaneous DC voltage VD can never be completely smooth, if the load is purely resistive, the DC load current cannot be completely smooth because it will linearly follow the DC voltage. Also, at delay angles a > 60°, the DC output voltage becomes discontinuous and , consequently, so would the DC current. In an effort to maintain a smooth DC current, practical converters usually have some inductance LD in series with the load on the DC side. For complete smoothing, the value of LD should theoretically be infinite, which is not really practical.

The practical consequence of this is that the theoretical formula for the calculated value of DC voltage (VD = 1.35 VRMS Cos a) is not completely true for all values of delay angle

a. Practical measurements confirm that it only hold true for delay angles up to about 75°, but this depends on the type of load and , in particular, the DC load inductance.

Experience shows that for a particular delay angle a > 60°, the average DC voltage will be higher than the theoretical value.

++++ 20: Deviation of DC voltage from theoretical vs delay angle

Applications for line commutated rectifiers

An important application of the line-commutated converter is the DC motor drive. A single controlled line-commutated converter connected to the armature of a DC Motor. The converter provides a variable DC voltage VA to the armature of the motor, controlled from the control circuit of the converter.

++++ 21: Converter fed DC motor drive

When the delay angle is less than 90°, the DC voltage is positive and a positive current IA flows into the armature of the DC motor to deliver active power to the load. The drive system is said to be operating in the 1st quadrant where the motor is running in the forward direction with active power being transferred from the supply to the motor and its mechanical load.

The motor field winding is usually separately excited from a simple diode rectifier and carries a field magnetizing current IF. For a fixed field current, the speed of the motor is proportional to the DC voltage at the armature. The speed can be controlled by varying the delay angle of the converter and its output armature voltage VA. If the delay angle of the converter is increased to an angle greater than 90°, the voltage VD will become negative and the motor will slow to a standstill. The current ID also reduces to zero and the supply line can be disconnected from the motor without breaking any current.

Consequently, to stop a DC motor, the delay angle must be increased to value sufficiently larger than 90° to ensure that the voltage VD becomes negative. With VD negative and ID still positive, the converter transiently behaves like an inverter and transfers active power from the motor to the supply line. This also acts as a brake to slow the motor and its load quickly to standstill. In this situation, the drive system is said to be operating in the 2nd quadrant where the motor is running in the forward direction but the active power is being transferred back from the motor to the supply line.

The concept of the 4 operating quadrants has been covered, but is illustrated again below. It illustrates the 4 possible operating states of any drive system and also shows the directions of VD and ID for the DC motor drive application.

++++ 22: Operating quadrants for variable speed drives

The converters discussed so far have been single converters, which are only able to operate with positive DC current (ID = +ve), which means that the motor can only run in the forward direction but active power can be transferred in either direction. Single DC converters can only operate in quadrants 1 and 4 and are known as 2 Quadrant converters.

To operate in quadrants 3 and 2, it must be possible to reverse the direction of ID. This requires an additional converter bridge connected for current to flow in the opposite direction. This type of converter is known as a 4 quadrant DC converter, and sometimes also called a double or back-to-back 6-pulse rectifier.

++++ 23: 4-quadrant line-commutated rectifier

With a DC motor drive fed from a 4-quadrant DC converter, operation in all 4 quadrants is possible with speed control in either the forward or reverse direction. A change of direction of the motor can quickly be achieved. Converter-1 is used as a controlled rectifier for speed control in the forward direction of rotation, while converter 2 is blocked, and vice versa in the reverse direction.

Assume, initially, that the motor is running in the forward direction under the control of Converter-1 with a delay angle of < 90°. Converter-2 is blocked. The changeover sequence from running in the forward direction to the reverse direction is as follows:

• Converter-1 delay angle increased to a > 90°. This means that DC voltage VD < 0 and DC current ID is decreasing.

• When ID = 0, Converter-1 is blocked and thyristor firing is terminated.

• After small delay, converter-2 unblocked and starts in the inverter mode with a firing angle greater than 90°.

• If the motor is still turning in the forward direction, converter-2 DC current ID starts to increase in the negative direction and the DC machine acts as a generator and is braked to standstill, returning energy to the supply line.

• As the firing angle is reduced a < 90°, converter-2 changes from the inverter to rectifier mode and , as voltage VD increases, the motor starts to rotate in the opposite direction.

In a DC motor drive, reversal of the direction of rotation can also be achieved by using a single converter and changing the direction of the excitation current. This method can only be used where there are no special drive requirements for changing over from forward to reverse operation. In this case, the changeover is done mechanically using switches in the field circuit during a period at standstill. Considerable time delays are required during standstill to re-magnetize the field in the reverse direction.

There are many practical applications for both uncontrolled and controlled line commutated rectifiers. Some of the more common applications include the following:

• DC motor drives with variable speed control

• DC supply for variable voltage variable frequency inverters

• Slip-energy recovery converters for wound rotor induction motors

• DC excitation supply for machines

• High voltage DC converters

• Electrochemical processes