The DFE method consists of a rectifier that converts DC to AC using a diode that cannot control the amount of current. Figure 1 shows a typical converter that uses a diode bridge.

Figure 1: 
Circuit diagram of a conventional rectifier

This rectifier consists of six diodes, and the output current flows when one of the upper diodes (D₁, D₃, or D₅) and one of the lower diodes (D₂, D₄, or D₆) are turned on at the same time. The diode with the largest amount of voltage in the positive direction among the upper part’s 3-phase power supply is turned on, and the remaining two diodes have a reverse-direction bias and are turned off. Simultaneously, the lower part’s diode with the largest amount of voltage in the negative direction is turned on and the other two diodes are turned off. The phase voltages e_a, e_b, and e_c at this time are as follows:

To reduce the ripple that occurs during the rectification process, a high-capacity capacitor is installed on the rectifier’s output side.

An AFE rectifier is a power conversion device that controls the amount and phase of the input current. When used together with a switching element, such as IGBT connected in antiparallel with diodes, it allows current flow in both directions, (from the input unit to DC link unit and vice versa).

The output that goes from the input side to the DC link unit is created via the diodes, and the bidirectional flow from the DC link unit to the input side is controlled using the switching element. The switching element allows a quality current with low THD and high power factor to be supplied by controlling the flow, so that the power supply voltage’s waveform and the input current waveform are the same. It is essential to install a line inductor on the input side for the bidirectional energy flow.

Figure 2 shows a 3-phase AFE rectifier. The rectifier consists of six switches in three units. To maintain a fixed DC output voltage during even rapid voltage changes, the power supply’s output side is equipped with a capacitor and an inductor for adjusting the amount of rectifier input current during power conversion. At this time, the AC power supply e_a, e_b, and e_c maintain a 3-phase equilibrium.

Figure 2: 
Circuit diagram of an AFE rectifier

If the switches in the upper and lower parts are turned on at the same time, dv/dt increases rapidly and causes switch damage. Therefore, a time delay is used during switch operation to control the action of the switches in a mutually complementary way, so that two switches are turned on or off.

The rectifier’s voltage equations are as follows:

where e_a, e_b, and e_c are the power supply voltages for the a, b, and c phases, respectively. i_a, i_b, and i_c are the phase currents. V_a, V_b, and V_c are the rectifier input voltages.

This study created a topology for DC motor speed control using an AFE rectifier.

A DFE rectifier that uses a diode bridge was created, as shown in Figure 3, for comparison with the proposed AFE rectifier.

Figure 3: 
Circuit diagram of a conventional converter

The rectifier with the diode bridge has a capacitor installed in the DC unit to stabilize the DC voltage that is produced. The DC unit’s capacitor is equipped with a much larger capacity than a capacitor normally used in an AC unit.

Figure 4 shows an AFE rectifier composed of a diode and IGBT connected antiparallel to each other, and a rectifier unit. The inverter unit consists of a conventional DC motor speed control topology.

Figure 4: 
Circuit diagram of the proposed converter

Each IGBT element in the rectifier unit is switched through PWM control. The DC unit is equipped with a 1/50-size small-capacity capacitor instead of a high-capacity capacitor, which was a drawback of the existing method.

Figure 5 shows the PSIM block diagram of a DC motor drive simulation using an existing DFE rectifier. A PI controller was used to control the speed and current.

Figure 5: 
PSIM diagram of a conventional converter

Figure 6 shows a graph of the DC motor’s speed response characteristics when speed commands of 0 to 1500 rpm were given.

Figure 6: 
Simulation responses for step change of speed setting (0→1,500 [rpm])

Figure 6 (a) shows the change in speed and Figure 6 (b) shows the change in load proportional to the square of the speed. Figure 6 (c) shows the input unit’s phase current waveform, and it includes a lot of harmonics.

In this case, the measured THD was very high at 743%. Figure 6 (d) shows the motor side load current. In a normal state, it was in the range of 27.1–29.3 A with a mean of 28.09 A. Figure 6 (e) shows the input unit’s voltage and current waveforms.

Figure 7 shows graphs of the response when a load torque of 5 N·m was applied during a normal operation at 300 rpm.

Figure 7: 
Simulation responses for step change of load torque (300 [rpm], 5 [N·m])

Figure 7 (a) shows the changes in the motor’s speed and Figure 7 (b) shows the changes in torque before and after the load torque was applied.

Figure 7 (c) shows the changes in the input unit’s phase current before and after applying the load torque, and the input unit current’s THD was 689% in the normal state. This means that too many harmonics were included in the input unit current. Figure 7 (d) shows the motor load current before and after the load torque was applied. In the normal state, it was in the range of 4.43–6.21 A and had a mean of 5.22 A.

Figure 8 shows the PSIM block diagram of a DC motor simulation using an AFE rectifier, as proposed in this study. The same PI controller was used as the existing DFE rectifier to control the speed and current, and sine pulse width modulation was used to control the voltage.

Figure 8: 
PSIM diagram for the proposed AFE converter topology

Figure 9 shows a graph of the DC motor’s speed response characteristics when speed commands from 0 to 1500 rpm were applied. Figure 9 (a) shows the changes in the motor’s speed. The stable control results were similar to those of the existing method. Figure 9 (b) shows the changes in load proportional to the square of the speed. The load was similar to that of the existing method in transient states. Figure 9 (c) shows the input unit phase current. The THD was 3.51%, and the waveform was similar to a sine wave, unlike the existing method. Figure 9 (d) shows the motor load current. In a normal state, it was in the range of 27.4–29.1 A with a mean of 28.4 A. Figure 9 (e) shows the input unit’s voltage and current wave forms. The power factor was 0.86.

Figure 9: 
Simulation responses for step change of speed setting (0→1,500 [rpm])

Figure 10 shows graphs of the response when a load torque of 5 N·m was applied during a normal operation at 300 rpm. Figure 10 (a) shows the changes in the motor speed. The stable control results in transient states were similar to the existing method. Figure 10 (b) shows the change in torque before and after the load torque was applied.

Figure 10: 
Simulation responses for step change of the load torque (300 [rpm], 5 [N·m])

Figure 10 (c) shows the changes in the input unit phase current before and after load torque was applied. After the load became stable, the input unit current showed a form that was closer to a sine wave, unlike that in the existing method. The THD was 5.84%. Furthermore, the change in the amount of current because of the load torque was smaller than that in the existing method. Figure 10 (d) shows the motor load current before and after the load torque was applied. After the load became stable, it showed a flow within a range of 4.43–6.51 A with a mean of 5.24 A. Even in the transient states, it achieved stable control similar to the existing method.

By comparing Figure 6 (a) and Figure 9 (a) as well as Figure 6 (b) and Figure 9 (b), it can be seen that when speed commands from 0 rpm to a high-speed region of 1500 rpm were used, the DC motor control method that used an AFE rectifier (as proposed in this study) and the DC motor control method that used the existing DFE method were able to control the speed and torque in a similar stable manner during the transient state intervals.

When Figure 6 (c) and Figure 9 (c) are compared, it can be seen that the existing method’s input unit current waveform had a THD of 743% and included too many harmonics to be compared to a sine wave. The input unit current waveform of the proposed method had a THD of 3.41% and was similar to a sine wave. It was confirmed that the input current was significantly improved.

When Figure 6 (d) and Figure 9 (d) are compared, it can be seen that the mean values of the load currents during a normal state in the existing DFE method and the proposed AFE method were not very different at 28.09 and 28.04 A, respectively. Even in the transient state intervals, the proposed method controlled the current in a stable manner, similar to the existing method.

By comparing Figure 7 (a) and Figure 10 (a) as well as Figure 7 (b) and Figure 10 (b), it can be seen that when speed commands from 0 rpm to a low-speed region of 300 rpm were used, the proposed AFE method controlled the speed and torque in a stable manner during the transient states, similar to the DFE method.

When Figure 7 (c) and Figure 10 (c) are compared, it can be seen that the existing method’s input unit current waveform contained a large number of harmonics with a TDH of 689%. The input unit current waveform of the proposed method shown in Figure 10 (c) has a waveform similar to a sine wave with a THD of 5.84%, thereby highlighting that the input current quality was improved.

When Figure 7 (d) and Figure 10 (d) are compared, it can be seen that the mean values of the load currents during a normal state after the load torque was applied in the existing method and the proposed method were not very different at 5.22 and 5.24 A, respectively. Even in transient intervals, the proposed method was able to perform stable control, similar to the existing method.

When Figure 6 (e) and Figure 9 (e) are compared, it can be seen that the proposed method made a great improvement so that the power factor was 0.86.