[sic-en-2013-24] a high-power-density dc–dc converter for distributed pv architectures

IEEE JOURNAL OF PHOTOVOLTAICS, VOL. 3, NO. 2, APRIL 2013 791

A High-Power-Density DC–DC Converter forDistributed PV Architectures

Mohammed S. Agamy, Senior Member, IEEE, Song Chi, Member, IEEE, Ahmed Elasser, Senior Member, IEEE,Maja Harfman-Todorovic, Member, IEEE, Yan Jiang, Member, IEEE, Frank Mueller, and Fengfeng Tao

Abstract—In order to maximize the solar energy harvesting ca-pabilities, power converters for photovoltaic (PV) systems have tobe designed for high efficiency, accurate maximum power pointtracking (MPPT), and voltage/current performance. When manyconverters are used in distributed PV systems, power density alsobecomes an important factor since it allows for simpler system in-tegration. In this paper, a high-power-density string-level MPPTdc–dc converter suitable for distributed medium- to large-scalePV installations is presented. A simple partial power processingtopology that is implemented exclusively with silicon carbide de-vices provides high efficiency and high power density. A 3.5-kW,100-kHz converter is designed and tested to verify the proposedmethods.

Index Terms—DC–DC converters, distributed photovoltaic (PV)architectures, partial power processing, silicon carbide devices.

I. INTRODUCTION

D ISTRIBUTED photovoltaic (PV) architectures provideseveral benefits compared with the central inverter sys-

tems, including higher energy yield, higher system availabil-ity, design flexibility, and improved monitoring and diagnosticcapabilities. For medium- to large-scale commercial and util-ity PV systems, a string/multistring dc–dc converter topologywith distributed maximum power point tracking (MPPT), asshown in Fig. 1, provides the best cost/performance operatingpoint [1]–[3]. For a distributed system with string dc–dc con-

Manuscript received June 1, 2012; revised August 30, 2012; accepted Octo-ber 21, 2012. Date of publication December 20, 2012; date of current versionMarch 18, 2013. This work was supported in part by the U.S. Department ofEnergy under Grant DE-EE0000572. This paper was presented at the IEEEPhotovoltaic Specialists Conference, Austin, TX, June 3–8, 2012. This reportwas prepared as an account of work sponsored by an agency of the United StatesGovernment. Neither the United States Government nor any agency thereof, norany of their employees, makes any warranty, express or implied, or assumes anylegal liability or responsibility for the accuracy, completeness, or usefulness ofany information, apparatus, product, or process disclosed, or represents that itsuse would not infringe privately owned rights. References herein to any specificcommercial product, process, or service by trade name, trademark, manufactureror otherwise does not necessarily constitute or imply its endorsement, recom-mendation, or favoring by the United States Government or any agency thereof.The views and opinions of the authors expressed herein do not necessarily stateor reflect those of the United States government or any agency thereof.

M. S. Agamy was with the GE Global Research Center, Niskayuna, NY12309 USA. He is now with the School of Engineering, The Universityof British Columbia, Kelowna, BC V1Y 9W9, Canada (e-mail: [email protected]).

S. Chi, A. Elasser, M. Harfman-Todorovic, Y. Jiang, F. Mueller, and F. Taoare with the General Electric Global Research Center, Niskayuna, NY 12309USA (e-mail: [email protected]; [email protected]; [email protected];[email protected]; [email protected]; [email protected]).

Color versions of one or more of the figures in this paper are available onlineat http://ieeexplore.ieee.org.

Digital Object Identifier 10.1109/JPHOTOV.2012.2230217

Fig. 1. Distributed PV architecture with string/multistring dc–dc converters.

verters that are rated at (1.5–6 kW), the estimated gain in energyyield is in the range of 3–9% over a standard central invertersystems [3]. However, the implementation of such distributedsystem requires high-performance, high-efficiency dc–dc con-verters [4]–[6]. Because of the high-efficiency requirements,partial power processing converters are often used as a simpleway to improve the overall conversion efficiency by directlyfeeding forward a fraction of the input PV power to the outputDC-bus [7]–[16].

In this paper, a high-efficiency, high-power-density, partialpower processing string-level dc–dc converter topology is pre-sented. The proposed transformerless partial power converterfeeds a constant voltage dc-bus output, while the controller reg-ulates the input PV string voltage to achieve MPPT. The devel-opment process was started by designing, building, and testinga baseline 3.5-kW converter switching at 30 kHz, which wasbuilt using 1200-V Si IGBT devices and silicon carbide (SiC)Schottky diodes. This was followed by the design of a secondgeneration of 3.5-kW converters that operate at three times theswitching frequency (100 kHz) to improve the power densitywhile maintaining the high efficiency. To keep the same highefficiency as the 30-kHz version, the 100-kHz dc–dc convertertopology is built using state-of-the-art 1200 V SiC MOSFETsand SiC Schottky diodes. The impact of increasing the switch-ing frequency on the efficiency as well as on the size of passivecomponents is investigated by comparing the 100-kHz converterperformance to the baseline 30-kHz converter.

The converter efficiency is measured, and the weighted effi-ciency value that is based on the California Energy Commission(CEC) is used as an evaluation metric. Different SiC MOS-FETs have been tested, and their impact on the converter effi-ciency was compared. Furthermore, the effect of SiC MOSFETcost and the viability of its application for solar converters arediscussed.

2156-3381/$31.00 © 2012 IEEE

792 IEEE JOURNAL OF PHOTOVOLTAICS, VOL. 3, NO. 2, APRIL 2013

Fig. 2. (a) Full power versus (b) partial power processing structures.

Fig. 3. Fraction of total power processed versus voltage gain (Vs /Vin ) for apartial power converter.

II. CONVERTER TOPOLOGY

In order to increase the energy yield by distributing the dc–dcconverters with individual MPPT controllers, the dc–dc con-verter efficiency has to be very high (> 98%). A simple ap-proach to achieve this high efficiency is to use partial powerprocessing converters. These converters process only a part ofthe input PV power to generate the voltage differential betweenthe PV string and the output dc-link, while the rest of the poweris directly fed forward to the output. Fig. 2 shows the concept ofa partial power processing converter, and Fig. 3 shows what per-centage of power is being processed by the dc–dc converter fora given input/output voltage gain. As an example, for a nominalmaximum power point string voltage of ∼400 V and a dc busof 600 V, only one third of the power is processed by the con-verter, while two-third is directly fed forward at almost 100%efficiency. Therefore, this reduces the pressure on the converterblock design without compromising the overall conversion effi-ciency, which helps to reduce the system cost [17]–[21].

The proposed power converter is a nonisolated partial powerbuck-boost topology, as shown in Fig. 4. The output voltageVout is the sum of the input PV voltage Vin and the voltageacross capacitor Cs (Vs). The converter can be operated as asingle-channel or as a multichannel interleaved topology. Volt-age gain of the regulated voltage Vs at medium to heavy load-ing conditions [where the inductor current is in the continuousconduction mode (CCM)] is given in (1), which is the conversionratio of a noninverting buck-boost converter

Vs =d

1 − dVin . (1)

Fig. 4. Buck-boost partial power dc–dc converter with a SiC MOSFET anda SiC Schottky diode. One channel is rated at 1.75 kW. Input voltage: 200 Vto 600 Vdc . Output voltage: 600 Vdc regulated. (a) One channel and (b) twochannels.

Fig. 5. Stages of operation of the partial power processing dc–dc converter.

In the discontinuous conduction mode (DCM), the capacitorvoltage is given by

Vs = 12

(√1 +

2d2Rload

Linfsw− 1

)Vin (2)

where d is the duty ratio, fsw is the switching frequency, Lin isthe input inductance, and Rload is the equivalent load resistance.

The operation of the converter is similar to that of a simpleboost circuit. The stages of operation over a switching periodTs are shown in Fig. 5 and can be summarized as follows.

Stage 1 (0 < t < dTs): In this stage, the MOSFET S is turnedON, and the inductor current builds up. Capacitor Cs deliversenergy to the output.

Stage 2 (dTs < t < T2): The MOSFET is turned OFF, andthe inductor current is diverted to the diode D. The inductorenergy is consequently discharged into the capacitor Cs . Forcontinuous conduction mode, T2 = Ts , and the cycle ends atthis stage.

Stage 3 (T2 < t < Ts): This stage occurs in the case of theDCM. In this mode, the power is transferred from the input

AGAMY et al.: HIGH-POWER-DENSITY DC–DC CONVERTER FOR DISTRIBUTED PV ARCHITECTURES 793

Fig. 6. Timing diagram and idealized converter waveforms for the stages ofoperation shown in Fig. 5.

and output capacitors to the output. It is also worth noting thatduring this mode of operation, resonances can occur between theinput inductor and the device capacitance. The DCM operationleads to zero current turn-on of the MOSFET, thus reducing theturn-on losses at light loads.

Fig. 6 shows the timing diagram and key idealized converterwaveforms for a single-channel converter. The grid tied in-verter is used to regulate the dc-link voltage which is basedon its modulation index. Double-line frequency oscillations(100 Hz/120 Hz) are considered too slow, and thus, the dc-link voltage is treated as constant over the 100-kHz switchingcycle. Since the output voltage is held constant by the inverterstage, the input voltage and the voltage across capacitor Cs con-tinuously add up to equal the dc-link voltage. Using multipleinterleaved channels reduces the high-frequency voltage rippleon capacitor Cs , and, consequently, helps to fix the input voltageVin at the maximum power operating point.

A baseline 30-kHz design that uses Si IGBTs and SiC Schot-tky diodes has a weighted efficiency of 98.22%. The weightedefficiency is based on the CEC efficiency formula [22]

ηCEC−weighted = 0.04 η10% + 0.05 η20% + 0.12 η30%

+ 0.21 η50% + 0.53 η75% + 0.05 η100% (3)

where ηx% represents the efficiency at x% of the converter-ratedpower.

In order to reduce the current ripple at the input, interleavedconverter legs can be used as shown in Fig. 4(b), where the twochannels are switched 180◦ apart. Furthermore, the interleavedconverter can be used to improve the light-load efficiency byswitching only a sufficient number of channels (when topologieswith more than two input channels are used) that correspond tothe power level processed by the converter.

In a PV farm, multiple converters can be connected in par-allel to feed a grid tied dc–ac inverter as shown in Fig. 7. Inthis case, the distributed dc–dc converters are responsible forMPPT of their respective PV string(s), and the inverter maintains

Fig. 7. Distributed PV architecture with multiple dc–dc converters connectedin parallel feeding a grid tied dc–ac inverter.

a stiff dc bus and handles other grid requirements such as totalharmonic distortion compliance, ride through, etc. High-power-density converters can, therefore, be attractive for their simplic-ity of integration with the PV strings without the need for largeenclosures.

III. BENEFITS OF AN ALL-SILICONE CARBIDE CONVERTER

Many power converters are now designed with commerciallyavailable 600-, 1200-, and 1700-V SiC Schottky diodes to elim-inate losses due to reverse recovery currents and thus improvethe converter efficiency. Recent progress in SiC materials’ de-velopment, processes, and fabrication has led to the availabilityof 1200-V SiC MOSFETs with current rating on the order of10–50 A from different sources such as Cree, GE, and Rohm.Improvements in substrate and material quality coupled with theprocess improvements (especially in the oxide interface) haveimproved the device reliability and robustness. The low forwarddrop of these MOSFETs due to their low on-state resistanceRdson (∼120 mΩ Rdson at room temperature for 1200 V, 15 Aparts typical) and their very low switching losses (switchingtimes on the order of tens of nanoseconds) lead to significantreductions in converter losses and hence to significant increasesin efficiency. While these parts are currently enclosed in TO-247 packages, they can operate at much higher junction tem-peratures, hence reducing the need for active cooling. The lowswitching and conduction losses, higher energy bandgap, andhigher carrier mobility of the SiC material allow for a threefoldincrease of the switching frequency when compared with Si-IGBT-based converters of comparable voltage and power rat-ings (in the 30–100 kHz range) while maintaining the sameefficiency, if not slightly higher [23]–[27].

The use of SiC MOSFETs provides several benefits com-pared with ultrafast Si IGBTs. Therefore, replacing the Si IGBTswitch in the converter in Fig. 4 with a SiC MOSFET enablesa significant increase in converter power density by increasingthe switching frequency. Fig. 8 shows the calculated reduc-tion in passive component values compared with the baseline


Fig. 8. Normalized passive component size versus switching frequency.

Fig. 9. Loss breakdown at 30 kHz for baseline 3.5-kW partial power dc–dc SiIGBT-based converter versus an all SiC converter with SiC MOSFETs.

converter when the switching frequency is increased. Further-more, the loss breakdown for the proposed converter showsthat more than half of the losses are due to device the switch-ing losses. Therefore, replacing the Si IGBT with the SiCMOSFET significantly reduces total losses across all operat-ing voltages, and thus, the overall weighted efficiency can beimproved by more than 1% at the same switching frequency.Fig. 9 shows a comparison of the loss breakdown values for the3.5-kW partial power dc–dc converter switching at 30 kHz andbuilt using Si IGBTs versus the equally rated converter designedusing SiC MOSFETs. The values that are shown in Fig. 9 weresimulated using detailed device models and at different inputvoltage levels. The losses are reduced to less than half of theiroriginal value when SiC MOSFETs are used, due to the domi-nance of switching losses in the IGBT-based converter. Fig. 10presents losses for the 3.5-kW SiC-based partial power dc–dcconverter as the switching frequency is increased. It can be con-cluded that a switching frequency greater than 100 kHz canbe reached, while matching the same losses as in the baseline30-kHz Si-IGBT converter.

Based on Fig. 8, a 100-kHz design results in a 66% reduc-tion in inductor size and a 60% reduction in the output ca-pacitor size when compared with the 30-kHz design. Thus, thepower density can be more than double while achieving the same

Fig. 10. Losses of a 3.5-kW partial power dc–dc all SiC converter at differentswitching frequencies.

TABLE ICONVERTER COMPONENTS

efficiency and an improved control bandwidth due to the fasterswitching. Furthermore, the reduction in the size of passivecomponents leads to a 40–50% reduction in the passive com-ponent cost. While the SiC MOSFETs are still more expensivethan Si IGBTs, their cost is gradually coming down with theincreased market penetration. The 1200-V SiC MOSFET partsare now commercially available, and they cost $1/A to $1.5/A.Compared with Si IGBTs, the SiC MOSFETs cost is almostfour to five times higher. While this will impact the overall costof the converter, it is expected that the SiC parts cost will comedown substantially as the technology matures and gets adopted.A cost analysis of the 3.5-kW Si IGBT/SiC Schottky diode con-verter was performed and, based on the expected energy yieldincrease [3], the system net present value analysis is positive.While this may not be the case with the all-SiC version, it willreach it once the cost of SiC MOSFETs is on the order of twotimes the cost of Si IGBTs. As the SiC wafer technology ma-tures and as the device yield increases, it is expected that thecost of SiC MOSFETs will continue to decrease, reaching thetwo times Si IGBT cost mark by 2015 [28].

IV. EXPERIMENTAL RESULTS

A 3.5 kW, all SiC string-level MPPT dc–dc converter withcomponents that are listed in Table I was built and tested. Differ-ent values for passive components were used, and performanceat switching frequencies of 30, 60, and 100 kHz was studied.


Fig. 11. Converter prototypes for a 30-kHz hybrid Si/SiC topology (right) anda 100 kHz all SiC topology (left).

Fig. 12. Converter operational waveforms at 1.75 kW. Ch1 (yellow): inductorcurrent (5 A/div); Ch2 (blue): input PV current (5 A/div); Ch3 (magenta): SiCMOSFET voltage (100 V/div) and Ch4 (green): SiC Schottky diode voltage(100 V/div), time (2 μs/div).

Converter efficiency with 1200-V SiC MOSFETs from threedifferent sources was evaluated. The reduction in board size dueto the increase of the switching frequency from 30 to 100 kHzis shown in Fig. 11. The 100-kHz converter has twice the powerdensity (37 W/in3) compared with the 30-kHz board (18 W/in3).This is very beneficial in simplifying the installation, since sucha small size converter can be directly connected to the PV stringor on a dc rail, instead of being placed in a string combiner box.Figs. 12 and 13 show the switch, diode, and inductor currents at50% and 10% of full load, respectively. In Fig. 12, the inductoris operating in CCM, while in Fig. 13, at light load it is operat-ing in DCM, which explains the resonance between the inductorand device capacitance as can be observed by the oscillations incurrent and voltage waveforms.

Interleaved inductor currents are shown in Fig. 14 along withinput and output voltages for a converter operating at rated inputPV power.

Fig. 15 shows the measured efficiency of the converter atdifferent switching frequencies with both channels switching atpredefined input power levels [according to (3)]. An averageweighted CEC efficiency of 99.1% was obtained at 30 kHz (a1% average gain over the baseline topology), and an averageweighted efficiency of 98.34% was obtained at a switching fre-quency of 100 kHz. In Fig. 16, the switching of the interleavedconverter channels is controlled based on the power level input to

Fig. 13. Converter operational waveforms at 175 W. Ch1 (yellow): inductorcurrent (1 A/div); Ch2 (blue): input PV current (1 A/div); Ch3 (magenta): SiCMOSFET voltage (100 V/div) and Ch4 (green): SiC Schottky diode voltage(100 V/div), time (2 μs/div).

Fig. 14. All SiC Converter operational waveforms at 3.5 kW. Ch1 (yellow):inductor current channel 1(iL in1 ) (5 A/div); Ch2 (blue): inductor current chanel2 (iL in2 ) (5 A/div); Ch3 (magenta): input voltage (Vin ) (500 V/div) and Ch4(green): output voltage (Vout ) (500 V/div), time (4 μs/div).

Fig. 15. Weighted efficiency at different PV input voltages for an all SiCconverter switching at 30 kHz (blue square), 60 kHz (red triangle), and 100 kHz(orange circle).


Fig. 16. All-SiC converter efficiency versus converter input power at differentinput voltage levels with an output voltage of 600 V.

Fig. 17. Weighted efficiency comparison using MOSFETs from three differentsources.

the converter. For power levels less than 50% of the rated power,only one channel is activated; otherwise, both channels are al-lowed to switch. This improves the light-load efficiency shownby the dashed lines at different voltage levels. The weighted effi-ciency according to (3) is 98.34% if both channels are switchingall the time and this efficiency increases by∼0.19 – 98.53% withthe controlled switching that is based on input power level.

Finally, a comparison of the weighted efficiency of the con-verter using SiC MOSFETs from three different sources ispresented in Fig. 17. The efficiency measurements vary within0.13%, which is a good indication of the maturation and uni-formity of the device technology, and thus, the expected avail-ability of high-performance SiC MOSFETs at low price pointsin the near future, making them a viable option for PV con-verter/inverter designs.

V. CONCLUSION

A high-frequency, high-power-density all-SiC dc–dc partialpower converter is proposed for distributed PV architectures.The use of an all-SiC converter significantly reduces switch-ing losses, which enables the higher frequency operation. The

smaller converter is easier to connect directly to the string orto the mounting rail, and the capability of the SiC converterto operate at higher temperatures improves the reliability andmakes the converter more suitable for outdoor operation in harshenvironments.

As the SiC technology matures, the cost of SiC devices willprogressively drop, and the devices will become more readilyavailable from many sources, which makes this technology veryattractive for PV applications.

REFERENCES

[1] N. Kaushika and N. Gautam, “Energy yield simulations of interconnectedPV arrays,” IEEE Trans. Energy Convers., vol. 18, no. 1, pp. 127–134,May 2003.

[2] N. Femia, G. Lisi, G. Petrone, G. Spagnuolo, and M. Vitteli, “Distributedmaximum power point tracking of photovoltaic arrays: Novel approachand system analysis,” IEEE Trans. Ind. Electron., vol. 55, no. 7, pp. 2610–2621, Jul. 2008.

[3] A. Elasser, M. Agamy, J. Sabate, R. Steigerwald, R. Fisher, andM. Harfman-Todorovic, “A comparative study of central and distributedMPPT architectures for megawatt utility and large scale commercial pho-tovoltaic plants,” in Proc. Ind. Electron. Conf., 2010, pp. 2753–2758.

[4] G. Lijun, R. Dougal, L. Shengyi, and A. Lotova, “Parallel-connected solarPV system to address partial and rapidly fluctuating shadow conditions,”IEEE Trans. Ind. Electron., vol. 56, no. 5, pp. 1548–1556, May 2009.

[5] H. Patel and V. Agarwal, “MATLAB-based modeling to study the effects ofpartial shading on PV array characteristics,” IEEE Trans. Energy Convers.,vol. 23, no. 1, pp. 302–310, Mar. 2008.

[6] A. Chouder and S. Silvestre, “Analysis model of mismatch power lossesin PV systems,” J. Solar Eng., vol. 131, pp. 24504-1–24504-5, May 2009.

[7] N. Femia, G. Lisi, G. Petrone, G. Spagnuolo, and M. Vitteli, “Distributedmaximum power point tracking of photovoltaic arrays: Novel approachand system analysis,” IEEE Trans. Ind. Electron., vol. 55, no. 7, pp. 2610–2621, Jul. 2008.

[8] Y. Xue, L. Chang, S. Kjaer, J. Bordonau, and T. Shimizu, “Topolo-gies of single-phase inverters for small distributed power generators: Anoverview,” IEEE Trans. Power Electron., vol. 19, no. 5, pp. 1305–1314,Sep. 2004.

[9] J. Young-Hyok, K. Jun-Gu, P. Sang-Hoon, K. Jae-Hyung, and W. Chung-Yuen, “C-language based PV array simulation technique considering ef-fects of partial shading,” in Proc. Int. Conf., Ind. Technol., 2009, pp. 1–6.

[10] R. Bruendlinger, B. Bletterie, M. Milde, and H. Oldenkamp, “Maximumpower point tracking performance under partially shaded PV array condi-tions,” in Proc. 21st Eur. PVSEC, Dresden, 2006, pp. 2157–2160.

[11] R. Ramabadran and B. Mathur, “Effect of shading on series and parallelconnected solar PV modules,” Modern Appl. Sci., vol. 3, no. 10, pp. 32–41,Oct. 2009.

[12] M. Garcia, J. Maruri, L. Marroyo, E. Lorenzo, and M. Perez, “Partialshadowing, MPPT performance and inverter configurations: Observationsat tracking PV plants,” Prog. Photovoltaics: Res. Appl., pp. 529–536, Apr.2008.

[13] U. Schwabe and P. Jansson, “Performance measurement of amorphous andmonocrystalline silicon PV modules in eastern U.S. ENERGY productionversus ambient and module temperature,” in Proc. Int. Instrum. Meas.Technol. Conf., Singapore, May 5–7, 2009, pp. 1636–1641.

[14] S. Krauter, A. Preiss, N. Ferretti, and P. Grunow, “PV yield prediction forthin film technologies and the effect of input parameters inaccuracies,”in Proc. 23rd Eur. Photovoltaic Sol. Energy Conf., Valencia, Spain, Sep.1–5, 2008, pp. 740–743.

[15] (2010). [Online]. Available: http://www.steveransome.com/pubs.html[16] (2012). [Online]. Available: http://enphaseenergy.com/support/learning

center.cfm[17] R. Adler, “A new DC-DC switching regulator topology enhances effi-

ciency and power density,” Proc. Power Con, pp. F1-1–F1-4, 1984.[18] Min J. Lee, J. Kim, T. Kim, D. Yoo, and E. Song, “A new topology with

high efficiency throughout all load range for photovoltaic PCS,” IEEETrans. Ind. Electron., vol. 56, no. 11, pp. 4427–4435, Nov. 2009.

[19] H. Li, H. Chen, and L. Chang, “Analysis and design of a single-stage par-allel AC-to-DC converter,” IEEE Trans. Power Electron., vol. 24, no. 12,pp. 2989–3002, Dec. 2009.


[20] M. de Rooij, J. Glaser, and R. Steigerwald, “High efficiency photovoltaicinverter,” U.S. patent application # US2009/0323379A1.

[21] B. Min, J. Lee, J. Kim, T. Kim, D. Yoo, and E. Song, “A new topologywith high efficiency throughout all load range for photovoltaic PCS,” IEEETrans. Ind. Electron., vol. 56, no. 11, pp. 4427–4435, Nov. 2009.

[22] M. Agamy, M. Harfman-Todorovic, A. Elasser, J. Sabate, R. Steigerwald,Y. Jiang, and S. Essakiappan, “DC-DC converter topology assessment forlarge scale distributed photovoltaic plant architectures,” in Proc. EnergyConvers. Congress Expo., 2011, pp. 764–769.

[23] L. Stevanovic, K. Matocha, P. Losee, J. Glaser, J. Nasadoski, and S. Arthur,“Recent advances in silicon carbide MOSFET power devices,” in Proc.COMPEL, 2010, pp. 401–407.

[24] L. Stevanovic, K. Matocha, Z. Stum, P. Losee, A. Gowda, J. Glaser, andR. Beaupre, “Realizing the full potential of silicon carbide power devices,”in Proc. Appl. Power Electron. Conf., 2010, pp. 1–6.

[25] J. Glaser, J. Nasadowski, P. Losee, A. Kashyap, K. Matocha, J. Garrett, andL. Stevanovic, “Direct comparison of silicon and silicon carbide powertransistors in high frequency hard-switched applications,” in Proc. Appl.Power Electron. Conf., 2011, pp. 1049–1056.

[26] H. Zhang, L. Tolbert, and B. Ozpineci, “Impact of SiC devices on hybridelectric and plug-in hybrid electric vehicles,” IEEE Trans. Ind. Appl.,vol. 47, no. 2, pp. 912–921, Mar./Apr. 2011.

[27] “PowerSiC: Silicon carbide devices for power electronics market: Status& forecasts,” Yole Development Rep., 2006. Available: http://www.yole.fr

[28] “SiC 2010: How silicon carbide will impact electronics: A 10 year pro-jection,” Yole Development Rep., 2010. Available: http://www.yole.fr

Mohammed S. Agamy (S’01–M’08–SM’11) re-ceived the B.Sc. (Hons.) and M.Sc. degrees in electri-cal engineering from Alexandria University, Alexan-dria, Egypt, in 2000 and 2003, respectively, and thePh.D. degree in electrical engineering from Queen’sUniversity, Kingston, ON, Canada, in 2008.

He is currently an Assistant Professor with theSchool of Engineering, the University of BritishColumbia, Vancouver, BC, Canada. From 2008 toJuly 2012, he was a Lead Power Electronics Engineerwith the General Electric Global Research Center,

Niskayuna, NY, where his research interest included power supply technolo-gies for renewable energy sources and medical equipment. From May 2003 toOctober 2008, he was with the Energy and Power Electronics Applied ResearchLaboratory, Queen’s University, as a Research Assistant and then a Postdoc-toral Fellow. From September 2000 to April 2003, he was an Assistant Lecturerwith Alexandria University. His current research interests include resonant con-verters, power factor correction, soft switching techniques and modeling, andcontrol of power converters and electric machines. He holds one U.S. patentwith six others pending and has over 30 published technical papers in refereedjournals and conferences.

Dr. Agamy serves as a Reviewer for the IEEE TRANSACTIONS ON POWER

ELECTRONICS, THE IEEE TRANSACTIONS ON INDUSTRIAL ELECTRONICS, THE

IEEE TRANSACTIONS ON INDUSTRY APPLICATIONS, the International Journalof Electronics, and several IEEE conferences.

Song Chi (S’04–M’07) received the B.S. degree fromNortheastern University, Shenyang, China, in 1993,the M.S. degree from Tsinghua University, Beijing,China, in 2000, and the Ph.D. degree from The OhioState University, Columbus, in 2007, all in electricalengineering.

He is currently with the General Electric (GE)Global Research Center, Niskayuna, NY. Prior tojoining GE, he was a Senior Engineer with the Re-search and Engineering Center of Whirlpool. He hasauthored or coauthored 15 technical papers in IEEE

conferences and journals. He has two patents pending. His research interests in-clude sensorless control of ac drives, flux-weakening control of surface mountedpermanent magnet/interior permanent magnet (SPM/IPM) machines, and thecontrol of power conversion systems such as distributed solar systems and high-fidelity gradient amplifiers of magnetic resonance (MR) scanners.

Ahmed Elasser (S’92–M’96–SM’12) was born inDemnate, Morocco, in 1963. He received the Inge-nieur D’Etat- diploma in electric power engineeringfrom the Mohammadia School of Engineering, Ra-bat, Morocco, in 1985 and the M.S. and Ph.D. degreesin electric power engineering and power electronicsfrom Rensselaer Polytechnic Institute, Troy, NY, in1996.

He was an Electrical Maintenance Engineer andthen a Laboratory Engineer from 1986 to 1992 in Mo-rocco. Since 1996, he has been with General Electric

(GE) Global Research Center (GRC), Niskayuna, NY, where he is currently aSenior Professional. His previous research interests include the study, model-ing, and application of power semiconductor devices, systems modeling andsimulation, silicon carbide devices, six-sigma quality, and e-engineering. From2002 to 2007, he led new ideas and innovation within the GRC Micro and NanoStructures Technology (MNST) organization, and he was the Head of the MNSTDisruptive Technology Council. He worked across GRC with various technol-ogy councils to create a culture of innovation and growth. His current researchinterests include silicon carbide power device fabrication, design, modeling,testing, characterization, and applications. He has also been focusing on pho-tovoltaic systems, focusing on plant architectures, distributed maximum powerpoint tracking, and balance of systems work. He has authored or coauthoredmore than 20 papers, holds 12 patents, and has several patents pending.

Dr. Elasser received a GE Dushman team award in 1996; he also received nu-merous awards from the GE Research Center for his technical and organizationalcontributions. GE Industrial Systems also recognized him for his numerous con-tributions to circuit breaker modeling and design. He has presented many IEEEconferences on power electronics, power semiconductor devices, and photo-voltaic systems. He is a Regular Reviewer for various IEEE publications andconferences and has recently served as a Topic Chair for the energy conversionconference and expo (ECCE) 2011 sustainable energy track as well as a SessionChair.

Maja Harfman Todorovic (S’02–M’08) receivedthe Dipl. Ing. degree from the Faculty of Electri-cal Engineering, University of Belgrade, Belgrade,Serbia, in 2001 and the M.S. and Ph.D. degrees fromTexas A&M University, College Station, in 2004 and2008, respectively.

Since March 2008, she has been a Lead Engi-neer in the Utility Power Electronics Laboratory,General Electric Research Center, Niskayuna, NY.Her research interests include converters for photo-voltaic applications, subsea oil and gas applications,

switching-mode power supply design, uninterruptible power systems, energystorage devices, and digital control of power converters. She holds one U.S.patent with eight others pending and has over 30 published technical papers inrefereed journals and conferences.

Dr. Harfman-Todorovic serves as a Reviewer for the IEEE TRANSACTIONS

ON POWER ELECTRONICS, the IEEE TRANSACTIONS ON EDUCATION, the IEEETRANSACTIONS ON INDUSTRY APPLICATIONS, IEEE TRANSACTIONS ON INDUS-TRIAL ELECTRONICS, and several IEEE conferences.

Yan Jiang (M’11) received the B.S. and M.S. degreesin electrical engineering from Zhejiang University,Hangzhou, China, in 1999 and 2002, respectively,and the Ph.D degree in electrical engineering fromthe Center for Power Electronics Systems, VirginiaPolytechnic Institute and State University, Blacks-burg, in 2009.

She has was an Application Engineer with LinearTechnology Corporation, Milpitas, CA, from 2008to 2010. She is currently an Electrical Engineer withGeneral Electric Global Research Center, Niskayuna,

NY. She has authored or co-authored 14 technical papers in IEEE journals andconferences. Her research interests include photovoltaic inverters, power elec-tronics for healthcare applications, SiC devices, electronic ballasts, power factorcorrection techniques, power integration and packaging, electromagnetic inter-ference filter design, power converter modeling and control, and high-frequencydc/dc converters.


Frank Mueller attended Rensselaer Polytechnic In-stitute (RPI), Troy, NY, in 1977, pursuing a major inphysics.

In 1979, he enlisted in the U.S. Air Force where hecompleted a Navy ET-A school at Great Lakes NavalBase and then specialized in Meteorological Equip-ment training at Chanute Air Force Base (AFB), IL(he graduated at the top of his class). Afterwards, hewent to Blytheville AFB, AR, where he maintainedthe weather radar and other meteorological equip-ment for the base. In 1983, he joined the micropro-

cessor laboratory at RPI assisting in lab preparations and later also supported aPower Electronics Lab which included laying out gate drive circuits. In 1996 hejoined General Electric and has been active in Power Electronics testing and lay-out. He has experience in optics, ultra-high vacuum systems, and high-voltagework up to 160 kV and has been leading his organizations safety program since1999. He holds two U.S. patents and has two pending.

Fengfeng Tao received the B.S. and M.S. degreesfrom Tsinghua University, Beijing, China, in 1990and 1996, respectively, and the Ph.D. degree fromVirginia Polytechnic Institute and State University,Blacksburg, in 2001.

He joined the General Electric Global ResearchCenter, Niskayuna, NY, in 2002 and has since beena member of the power electronics team and, morerecently, a member of the high-frequency power con-version technology laboratory. His research inter-ests include high-frequency power conversion, novel

topologies, solar inverters, microinverters, and converters, as well as powerelectronics converters’ controls. He has authored or coauthored 16 technical pa-pers in IEEE conferences and journals. He holds seven patents and has severalpending.

[sic-en-2013-24] a high-power-density dc–dc converter for distributed pv architectures

Documents