Lee2018 9level
Lee2018 9level
Lee2018 9level
fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2018.2805685, IEEE
Transactions on Power Electronics
0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2018.2805685, IEEE
Transactions on Power Electronics
analysis among cascaded MLI with the proposed S3CM, the SCM are kept within Vdc. Therefore, two series-connected switches must
topology [9], and the H-bridge is conducted in Fig. 4. For fair be considered for S5 and S7–S10 for the SCM topology in [9] (Fig.
comparison, voltage stress of all the switches in these topologies 1).
Fig. 1. Switched-capacitor module (SCM) topology presented in [9]. Fig. 2. The proposed single-stage switched-capacitor module (S3CM) topology for
cascaded MLI.
TABLE I
EQUATIONS FOR CASCADED MLI USING DIFFERENT MODULES WITH
ALL SWITCHES PIV WITHIN THE DC SOURCE, Vdc
Proposed S3CM SCM in [9] H-Bridge
Levels 8N+1 8N+1 2N+1
Switches 12N 15N 4N
Diodes - N -
DC sources N N N
Capacitors 2N 2N -
Voltage gain 2 2 1
Fig. 4. Comparative analysis for cascaded MLI using the proposed S3CM, SCM N = number of cascaded modules/H-bridges
[9] (Fig. 1) and H-bridge. Voltage gain = ratio of the maximum voltage level to the sum of all dc sources
0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2018.2805685, IEEE
Transactions on Power Electronics
With cascaded H-bridge as a benchmark, cascaded MLI using III. SIMULATION AND EXPERIMENTAL RESULTS
both the SCM topology in [9] and the proposed S3CM topology
illustrate less number of switches for a given voltage level, with the Simulations were conducted for both the steady-state and
latter results in more switch count reduction. Besides, both dynamic conditions. Fig. 6 shows the output voltage, load current
topologies demonstrate the same number of dc sources, which is and capacitors voltage for N=1. Seven voltage levels are clearly
also significantly less than the cascaded H-bridge for all voltage seen with both capacitors voltages balanced averagely about 100V.
levels. It is also worth emphasized that cascaded H-bridge and the Unequal capacitances for C1 and C2 were then simulated
proposed S3CM topology require no diode in their circuitry, while considering 20% tolerances in each capacitor. Result depicted in
N diodes are mandatory in the SCM topology [9]. All equations Fig.7 demonstrates that both the capacitor voltages are still
related to the compared modules for cascaded MLI are summarized oscillating around 100V albeit the different peak-to-peak
in Table I. magnitude. The voltage ripple across C1 is smaller than that across
Some detailed comparisons between the proposed S3CM C2 due to its higher capacitance. A step load change was also
topology and the SCM topology in [9] are depicted in Fig. 5. It is conducted with load current increased to twice its original
worth emphasize that the switching sequences of the capacitors in magnitude. Findings from Fig.8 show that the capacitors voltages
S3CM are identical to that in [9]. C1 discharges for a longer period are slightly unbalanced after the step load change and gradually
during the positive half-cycle while C2 discharges for a longer stabilizes at 100V, which once again confirming its capacitor
period in the negative half-cycle. Nonetheless, they have equal voltage balancing ability during operation.
discharging periods when symmetrical ac voltage is under
consideration. This implies their equal average voltage, as similar
to [9].
The number of conducting semiconductor devices and the
number of switching transitions between both topologies are also
considered for power conversion efficiency comparison purpose.
Considering PIV for all switches are within Vdc, the proposed S3CM
has less conducting switches for 1.5Vdc, 2Vdc, -1.5Vdc and -2Vdc,
indicating its lower conduction loss. Besides, it also presents less
number of switch/diode commutations for all voltage level
transitions, except when the transition is between Vdc and 1.5Vdc.
Its two zero states introduces more switching transitions when the Fig. 6. Simulated steady-state waveforms for single S3CM (9-level).
output voltage is 0V. The proposed S3CM topology exhibits a total
of 56 switching transitions over one fundamental cycle, as oppose
to a total of 88 switching transitions in the SCM topology [9],
which signifies its advantage of lower switching loss.
0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.
This article has been accepted for publication in a future issue of this journal, but has not been fully edited. Content may change prior to final publication. Citation information: DOI 10.1109/TPEL.2018.2805685, IEEE
Transactions on Power Electronics
Note that the simulation considers ideal capacitors (C1 and C2)
without internal series resistance, their voltage waveforms are thus
showing instant charging with approximately zero rise time, as
opposed to the slower rising time as shown in the experiment. In
both simulation and experimental waveforms, the longest
discharging period for C1 and C2 takes place when the output
voltage equals to {1.5Vdc, 2Vdc} and {–1.5Vdc, –2Vdc} respectively.
Good agreement between experimental and simulation waveforms
(a) further confirms the validity of the proposed S3CM topology.
IV. CONCLUSION
REFERENCES
Fig. 10. Experimental prototype. [1] H. Akagi, “Multilevel Converters: Fundamental Circuits and Systems,” Proc.
IEEE, vol. 105, no. 11, pp. 2048–2065, 2017.
[2] K. K. Gupta, A. Ranjan, P. Bhatnagar, L. K. Sahu, and S. Jain, “Multilevel
Inverter Topologies with Reduced Device Count: A Review,” IEEE Trans.
Power Electron., vol. 31, no. 1, pp. 135–151, 2016.
[3] E. Samadaei, S. A. Gholamian, A. Sheikholeslami, and J. Adabi, “An
Envelope Type (E-Type) Module: Asymmetric Multilevel Inverters With
Reduced Components,” IEEE Trans. Ind. Electron., vol. 63, no. 11, pp.
7148–7156, 2016.
[4] E. Samadaei, A. Sheikholeslami, S.-A. Gholamian, and J. Adabi, “A Square
T-Type (ST-Type) Module for Asymmetrical Multilevel Inverters,” IEEE
Trans. Power Electron., vol. 33, no. 2, pp. 987–996, 2018.
[5] A. S. Alishah, S. H. Hosseini, E. Babaei, and M. Sabahi, “Optimal Design of
New Cascaded Switch-Ladder Multilevel Inverter Structure,” IEEE Trans.
3
Fig. 11. Experimental and simulation waveforms of the proposed S CM topology Ind. Electron., vol. 64, no. 3, pp. 2072–2080, 2017.
(N=1, 9-level) operated under resistive-inductive load. [6] R. S. Alishah, S. H. Hosseini, E. Babaei, and M. Sabahi, “Optimization
Assessment of a New Extended Multilevel Converter Topology,” IEEE
For further verification, an experimental prototype depicted in Trans. Ind. Electron., vol. 64, no. 6, pp. 4530–4538, 2017.
Fig. 10 has been tested. With limited resources, 14V dc source and [7] J. Liu, K. W. E. Cheng, and Y. Ye, “A Cascaded Multilevel Inverter Based
on Switched-Capacitor for High-Frequency AC Power Distribution System,”
resistive-inductive load with power factor of 0.93 was considered. IEEE Trans. Power Electron., vol. 29, no. 8, pp. 4219–4230, 2014.
The efficiency recorded at 15.47W is 80.61%. The experiment [8] J. Liu, J. Wu, J. Zeng, and H. Guo, “A Novel Nine-Level Inverter Employing
waveforms depicted in Fig. 11 shows good agreement with the One Voltage Source and Reduced Components as High-Frequency AC
theoretical analysis, where nine voltage levels are clearly seen in Power Source,” IEEE Trans. Power Electron., vol. 32, no. 4, pp. 2939–2947,
2017.
the output voltage. The maximum voltage level is approximately [9] R. Barzegarkhoo, M. Moradzadeh, E. Zamiri, H. M. Kojabadi, and F.
28V, validating the voltage boosting capability of the proposed Blaabjerg, “A New Boost Switched-Capacitor Multilevel Converter with
S3CM topology. Two capacitor voltages are balanced with their Reduced Circuit Devices,” IEEE Trans. Power Electron., vol. PP, no. 99, pp.
average voltage approximately 7V, which is half of the dc source. 1–1, 2017.
0885-8993 (c) 2018 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission. See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.