4.3 Measurements and experimental results
4.3.2 Experimental results
The measured voltage waveforms of one inverter output phase (U) and one motor phase-to-phase voltage (U-V) without filtering are shown in Figures 4.28–4.32. The peak voltage level caused by the cable reflection is approximately twice the DC link voltage level (1100 V, 183 %UDC) without any filtering applied.
Operation of active du/dtfiltering without any load is shown in Figures 4.33–4.35. In addition to the perfectly timed charge pulses, the operation of the filter circuit without any control and with mismatched timing are also shown. Incorrect timing is not critical for the operation, but it can be seen that the control is nevertheless necessary, because of the strong LC resonance owing to the low damping factor.
When active du/dtis applied, Figures 4.36–4.44, the peak voltage at the motor end decreases considerably, and on the shorter, 30 meter and 100 meter cables, the oscillation is eliminated.
However, the slight inaccuracies in the charge pulse and the loading effect of the motor cable cause some error inducing oscillation in the output voltage of the LC circuit, even if the cable
A B B 5 . 5 k W 4 0 0 V / 1 1 . 3 0 A 1 4 3 0 r p m
T e k t r o n i x P 5 2 0 5
d i f f e r e n t i a l h i g h v o l t a g e p r o b e 0 - 1 0 0 M H z
A B B 7 . 5 k W 4 0 0 V / 1 5 . 2 A 1 4 4 0 r p m
M C M K t y p e p o w e r c a b l e 3 x 2 . 5 m m 2+ 2 . 5 m m 2 S
Figure 4.26. ABB 5.5 kW and 7.5 kW induction motors used in the measurements. The MCMK type power cable and the Tektronix differential voltage probe are also shown.
is left open ended. The effect of the motor current is also visible, since the current correction method was not implemented. However, since the load current is smaller compared with the filter charging current (75 A), the error is not significant. However, in designs where the load current is in the order of the filter current, correction pulses should be implemented;
otherwise, LC resonance up to twice the DC link will be induced. It is also shown in the mea-surements that the 300 meter cable is too long for the designed filter, and the cable oscillation is not eliminated.
4.3 Measurements and experimental results 89
F i l t e r c a p a c i t o r s 0 . 3 3 m F P l a s t i c - i n s u l a t e d p u l s e c a p a c i t o r s F i l t e r c o i l s 1 6 m H F e r r o x c u b e E T D 4 9 / 2 5 / 1 6 c o i l f o r m e r s
3 C 9 0 f e r r i t e c o r e w i t h a i r g a p
Figure 4.27. Active du/dtfilter circuit in more detail.
−4 −3 −2 −1 0 1 2 3 4
x 10−5
−500 0 500 1000
Inverter output voltage
Time [s]
a)
Voltage [V]
−4 −3 −2 −1 0 1 2 3 4
x 10−5
−500 0 500 1000
Voltage at 30 meter open−ended cable end
Time [s]
b)
Voltage [V]
Figure 4.28. Measured voltage waveforms without active du/dtfiltering applied. a) The inverter output voltage waveform, and b) the voltage at the open end of the 30 meter motor cable are shown. The overvoltage and oscillation caused by the cable reflection are clearly visible. Overvoltage 506 V, 84 %.
−4 −3 −2 −1 0 1 2 3 4
Voltage at 100 meter open−ended cable end
Time [s]
b)
Voltage [V]
Figure 4.29. Measured voltage waveforms without active du/dtfiltering applied. a) The inverter output voltage waveform and b) the voltage at the open end of the 100 meter motor cable are shown. The overvoltage and oscillation caused by the cable reflection are clearly visible. Overvoltage 507 V, 84 %.
−4 −3 −2 −1 0 1 2 3 4
Voltage at 100 meter 5,5 kW motor−ended cable end
Time [s]
b)
Voltage [V]
Figure 4.30. Measured voltage waveforms without active du/dtfiltering applied. a) The inverter output voltage waveform and b) the voltage at the motor end of the 100 meter cable are shown. The overvoltage and oscillation caused by the cable reflection are clearly visible. The effect of the 5.5 kW electric motor on the oscillation is minimal. Overvoltage 482 V, 80 %.
4.3 Measurements and experimental results 91
Voltage at 100 meter 7,5 kW motor−ended cable end
Time [s]
b)
Voltage [V]
Figure 4.31. Measured voltage waveforms without active du/dtfiltering applied. a) The inverter output voltage waveform and b) the voltage at the motor end of the 100 meter cable are shown. The overvoltage and oscillation caused by the cable reflection are clearly visible. The effect of the 7.5 kW electric motor on the oscillation is minimal. Overvoltage 473 V, 79 %.
−10 −8 −6 −4 −2 0 2 4 6
Voltage at 300 meter open−ended cable end
Time [s]
b)
Voltage [V]
Figure 4.32. Measured voltage waveforms without active du/dtfiltering applied. a) The inverter output voltage waveform and b) the voltage at the open end of the 300 meter motor cable are shown. The overvoltage and oscillation caused by the cable reflection are clearly visible. Overvoltage 498 V, 83 %.
−1 0 1
Inverter and active du/dt filter output voltage
Time [s]
Voltage [V]
Inverter output voltage Filtered output voltage
Figure 4.33. Measured voltage waveforms, when active du/dtfilter is attached to the output phases of the inverter, but no charge or discharge pulses are generated. No motor cable is connected. The filter is at full resonance, the frequency set by the filterLCconstant. The low damping factor (i.e. losses) of the active du/dtfilter is seen from the output voltage waveform, as the oscillation decays slowly, making the active du/dtfilter useless without the active control. Absence of the active du/dtsequence has caused approximately 500 V of the LC resonance overvoltage, 83 %.
−1 0 1 2 3 4 5
Inverter and active du/dt filter output voltage
Time [s]
Voltage [V]
Inverter output voltage Filtered output voltage
Figure 4.34. Measured voltage waveforms, when active du/dtfilter is attached to the output phases of the inverter. No motor cable is connected. When active control as presented in Chapter 3 is properly im-plemented, the filter circuit functions as predicted by the theory. A rising and falling slope is generated, and the du/dtis set by the filterLCconstant. No remaining oscillation of the LC circuit is visible.
4.3 Measurements and experimental results 93
Inverter and active du/dt filter output voltage
Time [s]
Voltage [V]
Inverter output voltage Filtered output voltage
Figure 4.35. Measured voltage waveforms, when active du/dtfilter is attached to the output phases of the inverter. No motor cable is connected. The effect of a faulty charge sequence is illustrated. The pulse width is over 50 %, causing the filter capacitor to overcharge above the DC link voltage. The transient induces filter resonance, the amplitude of the resonance being the difference between the DC link and filter voltages at the switching instant. An error in the active du/dtsequence has caused approximately 80 V of the LC resonance overvoltage, 13 %.
−10 −8 −6 −4 −2 0 2 4 6
Voltage at 30 meter open−ended cable end
Time [s]
b)
Voltage [V]
Figure 4.36. Measured voltage waveforms with active du/dt filtering applied. a) The filter output voltage waveform and b) the voltage at the open end of the 30 meter motor cable are shown. The overvoltage and oscillation caused by the cable reflection are eliminated. Slight resonance is shown in the waveforms resulting from the loading caused by the power cable to the filter, because the filter capacitor is not an ideal voltage source. Overvoltage 10 V, 1.6 %.
−1 −0.8 −0.6 −0.4 −0.2 0 0.2 0.4 0.6 0.8 1
Filter output current, motor voltage, 30 m cable, 5,5 kW motor
Time [s]
Figure 4.37. Measured voltage waveforms and filter output current with active du/dtfiltering applied.
a) The filter output voltage waveform and b) the voltage at the motor end of the 30 meter power cable are shown. The motor was a 5.5 kW induction motor. The motor current (towards the motor) causes faulty filter discharge during the falling slope. The motor current,≈5 A, has caused approximately 10 V, 1.6 % LC resonance overvoltage.
−0.03 −0.02 −0.01 0 0.01 0.02 0.03
Filter output voltage and current, 30 m cable, 5,5 kW motor
−0.03 −0.02 −0.01 0 0.01 0.02 0.03−10
Figure 4.38. Measured filter output voltage and current with active du/dtfiltering applied. The power cable was 30 meters long, and the motor was a 5.5 kW induction motor. The load current causes faulty filter discharge during the falling slope, and the negative motor current causes faulty filter charge during the rising slope. The resonance can be detected from the envelope of the filtered PWM voltage. The motor current, peak≈ ±8 A, has caused approximately 40 V, 7 % LC resonance overvoltage.
4.3 Measurements and experimental results 95
Voltage at 100 meter open−ended cable end
Time [s]
b)
Voltage [V]
Figure 4.39. Measured voltage waveforms with active du/dt filtering applied. a) The filter output voltage waveform and b) the voltage at the open end of the 100 meter motor cable are shown. The overvoltage and oscillation caused by the cable reflection are eliminated. Slight resonance is shown in the waveforms resulting from the loading caused by the power cable to the filter, because the filter capacitor is not an ideal voltage source. Overvoltage 6 V, 1.0 %.
−10 −8 −6 −4 −2 0 2 4 6
Voltage at 100 meter open−ended cable end
Time [s]
b)
Voltage [V]
Figure 4.40. Measured voltage waveforms with active du/dtfiltering applied. a) The filter output voltage waveform and b) the voltage at the open end of the 100 meter motor cable are shown. The effect of a faulty charge sequence is illustrated. Eventually, the oscillation in the filter output voltage will be visible in the open or motor end of the cable. An error in the active du/dtsequence has caused approximately 80 V, 13 % of the LC resonance overvoltage. The cable-reflection-induced overvoltage is 20 V, 2.9 %.
−10 −8 −6 −4 −2 0 2 4 6
Filter output current, motor voltage, 100 m cable, 7,5 kW motor
Time [s]
Figure 4.41. Measured voltage waveforms and filter output current with active du/dtfiltering applied. a) The filter output voltage waveform and b) the voltage at the motor end of the 100 meter power cable are shown. The motor was a 7.5 kW induction motor. The motor current (towards the motor) causes faulty filter discharge during the falling slope. The motor current, 6.5 A, has caused approximately 30 V, 5 % LC resonance overvoltage.
Filter output voltage and current, 100 m cable, 7,5 kW motor
−0.03 −0.02 −0.01 0 0.01 0.02 0.03−10
Figure 4.42. Measured voltage waveforms and filter output current with active du/dtfiltering applied.
The power cable was 100 meters long, and the motor was a 7.5 kW induction motor. The load current causes faulty filter discharge during the falling slope, and the negative motor current causes faulty filter charge during the rising slope. The resonance can be detected from the envelope of the filtered PWM voltage. The motor current, peak≈ ±8 A, has caused approximately 35 V, 6 % LC resonance overvoltage.
4.3 Measurements and experimental results 97
Voltage at 300 meter open−ended cable end
Time [s]
b)
Voltage [V]
Figure 4.43. Measured voltage waveforms with active du/dtfiltering applied. a) The filter output voltage waveform and b) the voltage at the open end of the 300 meter motor cable are shown. The overvoltage is approximately 180 V, 30 % ofUDC, because the du/dtof the designed filter is too high for the long cable. Furthermore, the operation of the filter is interfered by the cable resonance; the LC resonance overvoltage is approximately 120 V, 20 % and the cable-reflection-induced overvoltage 160 V, 23 %.
−10 −8 −6 −4 −2 0 2 4 6
Filter output current, motor voltage, 300 m cable, 5,5 kW motor
Time [s]
Figure 4.44. Measured waveforms with active du/dt filtering applied. a) The filter output voltage waveform and b) the voltage at the motor end of the 300 meter power cable are shown. The motor was a 5.5 kW induction motor. In addition, the current oscillation at the cable resonance frequency interferes the filter operation. The motor current≈5 A, the LC resonance overvoltage approximately 30 V, 13 %, and the cable-reflection-induced overvoltage 230 V, 37 %.