World Academy of Science, Engineering and Technology 42 2008

Single Phase to Three Phase Converter

Naung Cho Wynn, and Tun Lin naing

technologies. AbstractA new single phase to three phase converter topology

for small industries is presented in this paper: Phase converter, include this paper, is a new technology that supplies three phase power from a single phase source to power inductive, resistive and capacitive loads with distinct advantages over any existing converter technology. The converter consists of DC power supply, a MOSFET Hex-bridge, integrated gate drive IC, and a DSP to generate the switching signals. The switching signals generated are a unique version of selective harmonic elimination, which produces a consistent starting point for the switching functions, independent of the number of harmonics eliminated This converter covers the basis of induction motors and different types of other motors. They are ideal for farms, workshops, garages and large building etc.

II. PHASE CONVERTERS A. Static Converter The Static Converter is made up of two small components: A voltage sensitive relay and a standard capacitor (Cs) connected to your motor application (Red Box). The capacitor delays waveforms (or shifts the phase) during the start-up of your motor application. The relay disconnects this start capacitor after the motor has started. From this point, the motor will continue turning on the single phase supply. The performance of such a motor is fairly poor and can be compared to a car motor running on only a few cylinders Motors operated on a static converter will produce about 5060% of their name plate power. When you add another low cost run capacitor (Red Dotted Box) to the simple design, rated power goes up to around 70% of the motors name plate power. To help with understanding, the Start Capacitor (Cs) is used only to start the motor and then it is switched out completely. The Run Capacitor (Cr) is always in the circuit and is carefully sized to balance the voltages at one load rating (generally around 50% full load). Since Cr is fixed the voltage balancing at either end (0% and 100%) is quite poor.

KeywordsMOSFET Hex-bridge, DC power supply, gate drive IC, switching signal static converter, rotary converter. I. INTRODUCTION wide variety of commercial and industrial electrical equipment requires three-phase power. Electric utilities do not install three-phase power as a matter of course because it costs significantly more than single-phase installation. As an alternative to utility installed three-phase, rotary phase converters, static phase converters and phase converting variable frequency drives (VFD)have been used for decades to generate three-phase power from a single-phase source. Construction of three-phase power lines can cost as much as $50,000 per mile and can have an undesirable environmental impact. Even when three-phase lines are nearby, the cost of installation is considerable. Based on anticipated electricity demand for the three-phase application, the utility may or may not charge the customer for the cost of installation. Continuing monthly surcharges for the service are also common. Phase converters have historically been employed where utility three-phase power was unavailable, or where the electricity demand did not justify the cost of utility three-phase installation. Reduced motor life caused by voltage and current imbalance, harmonics that pollute the power grid and damage equipment, or the inability to operate sensitive equipment or multiple loads are just some of the problems that have limited the use of phase converters. This phase converter is a new, patented technology that supplies three-phase power from a single phase source to power inductive, resistive and capacitive loads with distinct advantages over any existing converter technology. There are various phase conversion
Authors are with the Mandalay Technological University, Myanmar (email: [email protected]., [email protected]).

Fig. 1 One line diagram of the static converter

B. Rotary Converter Add an idle running motor to a static converter; is called rotary converter. The added motor will compensate for some of the static converter weaknesses and help extend the range of motor sizes and loads. The internal motor is inactive at average load, but works hard when loads dont match the value of the chosen run capacitor (Cr). Rotary converters are clearly somewhat better than static converters. They can run several motors of different sizes. Large motors will produce up to 90% of their nameplate power, small motors (motor being much smaller than the converters idling motor or pilot motor) a bit more. If the manufacturers oversize these motors,

343

World Academy of Science, Engineering and Technology 42 2008

the output symmetry, start capability and capacity will all be increased. C. Document Modification

Where; Pmax = maximum power of secondary transformer side Vs = secondary voltage And, the turn ration is

Vs Vp

Pmax Vs

(2)

Where; Ns = number of secondary turns Np = number of secondary turns Vp = primary voltage And,

Vs Vp
Fig. 2 One line diagram of the rotary converter

Vp Vs

Np Ns

(3)

Both rotary and static converters have difficulty adjusting voltage balance to accommodate changing load conditions. Voltage regulation schemes for rotary converters are available which switch in different amounts of capacitance as the load changes. However, it is still difficult to get good control, and the high current pulses created in the system as the capacitors are switched in and out can be a problem. Digital Phase Converters are purely electronic and non-mechanical in nature. As such they are able to produce decently balanced three phase power. Another benefit of this type of phase converter is that they are very quiet. On the downside, this type of phase converter takes a sign wave, chops it up and spits out a somewhat choppy synthesized 3 phase sign waves their output harmonic distortion tends to be greater than a few other types of phase converters. D. Design calculation for power supply circuit
D1 220V AC MB351 D2 .0 D3 D4
Fig. 3 200V DC Power Supply for H-bridge Inverter

Where; Is = secondary current Ip = primary current For specifications Pmax = 500VA,Vs = 18V, Vp = 230V; By using equation 1, Imax = 28A By using equation 2, Np = 13 Ns By using equation 3, Is = 13 Ip The peak voltage of secondary transformer side is

Vp

2 Vrms

(4)

Where; Vrms = rms value of primary voltage The average voltage is

Vavg
Where; Vp = peak voltage

2Vp

(5)

And the ripple voltage yields

Vripple
+
C1 C2 C3

1 2CF

(6)

470 220 220

200V DC

Where; C = filter capacitance F = operating frequency The output DC voltage is obtained (7) VDC Vp Vripple By using above equation 4, 5, 6, and 7, the supply voltage 200V DC are calculated. E. Hex-bridge The heart of a standard three-phase inverter is the hexbridge. The hex-bridge takes a dc bus voltage and uses six switches arranged in three phase legs as shown in Fig. 4. From the middle of each phase leg comes the line which connects to the motor. The waveforms on these lines need to be a balanced three-phase sinusoidal waveform in order to drive the induction motor. This is achieved by applying carefully controlled switching waveforms to the gate of the switches.

There are two separate power supply used in this variable speed drive. The 200V DC power supply was used for Hbridge inverter and 12V regulated power supply was used for PWM frequency control circuit and driver circuit. This 200V DC power supply circuit is shown in Fig. 3. The maximum current of secondary transformer side is

Imax

Pmax Vs

(1)

344

World Academy of Science, Engineering and Technology 42 2008

Vdc

The HGTP12N60 is rated for 600V which is the very low end of IGBT ratings. IGBTs are usually used in higher power applications, so this part is not extremely common. The savings in losses and price prompted us to choose the FQP17N40 as the switch to be implemented in the hex-bridge. F. Gate Drive The gate drive circuit needs to provide an interface between the switching signals coming from the DSP or arbitrary waveform generator and the FETs in the hex-bridge. The DSP gives a 3.3V signal, while the waveform generators allow for a specified voltage level. The gate to source voltage needed for desired operation of the FQP17N40 FETs is on a 12 to 15V level. In addition, the high side FETs in the hexbridge do not have the source connected to ground, so the actual voltage needed to drive the gate depends on the varying voltage at the source. A single chip solution was found, the IR21362, which implements all six gate drives, including the circuitry that takes into account the issues with the high side FETs. There were other options, such as the IR2121, consisting of a single gate drive on a chip, but in the interest of lowering component count and cost, the IR21362 was chosen. Since six IR2121s would be needed, it is apparent that the single chip solution (IR21362) is more cost effective. G. Switching Signal Methodology Currently small motor drive systems are expensive and implement control schemes that use relatively high switching frequencies such as sine-triangle PWM, space vector PWM, or hysteresis current control. One drawback to the high switching frequencies is the decrease in efficiency that occurs from switching loss. However, the previously mentioned control methods do have their own merits. Each control scheme has been used widely and generates little acoustic noise since the switching frequency is on the upper end of the audible acoustic range ( 20 kHz). These control schemes also provide good dynamic performances. However, this application does not need good dynamic performance since there are no dynamic load and speed requirements. The switching signals implemented are a unique version of selective harmonic elimination, which is currently an active research area in power electronics. This technique lowers the switching frequency considerably compared to the previously mentioned methods. In addition, this switching scheme provides direct control over harmonics present in the waveform. The selective harmonic elimination method will be implemented in a switching waveform, while sine-triangle PWM will be used as a back-up. H. Switching Signal Implementation Two DSP development boards for motor control applications are available for proto-typing DSP code development. They are the eZdsp F2407 and the eZdsp F2812 boards. The F2407 has a lower cost then F2812. However, the F2812 has a faster clock at 150 MHz compared to the 30 MHz clock on the F2407. The F2812s analog-to-digital converter has a 12 bit resolution, whereas the F2407 only has 10 bit

A
Fig. 4 Hex-bridge

There are two kinds of switches that can be used for our fairly high power application, Insulated Gate Bipolar Transistors (IGBTs) or MOSFETs. To meet our specifications the switch needed minimum ratings of 400V and 10A. In addition we wanted a switch which minimized losses; therefore, switches with lower resistances are more desirable. The IGBT considered for this design was the HGTP12N60B3D, which has a small equivalent resistance of approximately 0.07 , but IGBTs also have a voltage drop at all times due to the collector emitter saturation voltage, which in this case is equal to 1.7V. The MOSFET investigated was the FQP17N40, which has an on-state resistance of 0.27 .
Average Conduction Based Power Loss

20 FQP17N40 HGTG12N60B3D 15

10

0 0

10

RMS current (A) [Single Frequency Sinusoid]

Fig. 5 Power losses in IGBT vs. FET

The power losses of the IGBT and the FET are plotted against the current. The IGBT has a fairly linear loss curve because the loss is mostly due to the IVsat, while the MOSFET losses are due to the I2RDS(on). The MOSFET losses increase faster with current, but the MOSFET losses do not surpass the IGBT losses until I is approximately greater than 9A. The motor used in our project is rated for 3A so the current should never reach 9A. In addition to having smaller losses in our operating region, the FQP17N40 is cheaper. It is $0.96 in a quantity of 1000, while the HGTP12N60B3D is $1.70 in the same quantity.

345

World Academy of Science, Engineering and Technology 42 2008

resolution. This makes the F2812 more desirable since it will give a better resolution for the speed commands. In addition, constructed harmonic elimination PWM implementation in the DSP requires the use of a large switching table that cannot fit into the F2407 FLASH. The eZdsp kit provides a complete development environment: DSP board, power supply, on-board JTAG compliant emulator, and an eZdsp specific Code Composer StudioTM full featured, with debugger IDE, and ANSI C and C++ compliant compiler. The DSP board has all peripheral signals needed on the board headers, making it easy to interface the board with the 3-phase inverter circuit. I. Inverter Control The control method implemented is open loop and based upon constant volts per hertz ratio. Although the inverter uses a 200V dc bus, the voltage is able to be varied by adjusting the modulation depth of the switching signals. Additionally, in order to meet the speed specifications over the given load range an efficient load compensation method will be used to determine optimal values of voltage and frequency to be applied to the machine at a given operating point. III. DESIGN DETAILS A. Hex-Bridge and Gate Drive (Component Selection) The FQP17N40 was chosen as the switch in the hex-bridge. Often, MOSFETs need external circuitry, called a snubber, in order to improve the switching trajectory during the commutation period. The switching trajectory of the FQP17N40 was measured (see Fig. 6) and appears to be fairly rectangular, as desired for an inductive load.

(PCB) that is mounted to the end of the motor. There will be a fan attached to the rotor which actively cools the FETs. These facts lead to the decision to forego the use of heatsinks. The IR21362 gate drive chip requires some external circuitry. The gate resistors (R11-R16), bootstrap capacitors (C4, C5, and C7), and decoupling capacitors (C3 and C6) were chosen based on information found in [2, 3]. A resistor and capacitor (R2 and C5) had to be chosen to provide a time constant which is used for the fault clearing time. A time constant of greater than 1.5s was desired so values of 33k and 47 F were chosen. The FAULT pin needs to be pulled high because it is active low. A LED was added in for visual output in the case of a fault. The protection circuitry, which includes R7, R9, R19, and C1, is chosen based on the fact that a fault occurs if the voltage on the ITRIP pin is larger than 0.5V. The variable resistor, R7, can be adjusted so that a fault is given at a desired current value. The values were chosen with the intention of initiating a fault at currents larger than 10A. Finally, three phase power for induction motor is got when DC power supply connected with Hex-bridge and gate drive.

Fig. 6 Switching trajectory

In addition, thermal issues needed to be taken into consideration. The FQP17N40 is rated for 10.1A continuous at 100C. Since the motor used is rated for 3A, the current should never be greater than 5A continuously. In addition, the final product is intended to be on a printed circuit board

346

World Academy of Science, Engineering and Technology 42 2008

B. Detailed Schematic
C_TO_MOTOR

ACKNOWLEDGMENT The First named author expresses sincere thanks to the university of Mandalay Technological University.

FQP17N40

FQP17N40

REFERENCES
[1]
R16 25

[2]
R11 25

[3]
FQP17N40

[4]
R15 25

[5]
R14 25

[6]
FQP17N40

[7]

[8]
R13 25

R12 25

1 u

1 u

V1

200V

VCC HIN1 HIN2 HIN3 LIN1 LIN2 LIN3 FAULT ITRIP EN RCIN VSS COM LO3

VB1 HO1 VS1 NC VB2 HO2 VS2 NC VB3 HO3 VS3 NC LO1 LO2

28 27 26 25 24 23 22 21 20 19 18 17 16 15

MUR160

MUR160

MUR160

1 2 3 4 5 6 7 8 9 10 11 12 13 14

IR21362

D1

D2

D3 U4

R7

500k POT

VCC

R9 100

C6 1 u

HIN1 HIN2 HIN3

FAULT

C3 1 u

Fig. 7 Hex-bridge and gate drive circuitry

C2 47u

33k R5

C1 0.022u

5V

D4 LED

R8 500

R19

0.05

C5

C4

C7

1 u

[9]

P. T. Krein, Elements of Power Electronics. New York and Oxford: Oxford University Press, 1998. International Rectifier, Appl. Note 978. International Rectifier, Appl. Note INT985. H. S. Patel and R. G. Hoft, Generalized techniques of harmonic elimination and voltage control in thyristor inverters: Part I Harmonic elimination, IEEE Trans. Ind. Applicat., vol. IA-9, no.3, pp. 310-317, May/June 1973. Hur Namho, Jung Junhwan, Nam K wanghee. Fast Dynamic DC-link Power balancing Scheme for a PWM converter-inverter System [A]. IECON 99 Proceedings[C]. The 25th Annual conference of the IEEE, 1999, 2:767~772 Jung Jinhwan, Lim Sunkyoung, Nam Kwanghee. A Feed-back Linearizing Control Scheme for a PWM Converter inverter Having a very Small DC-Link Capacitor[J].IEEE Trans. on Ind. Appl., 1999,35(5):1124~1131 Li Li, D. Czarkowski, Yaguang Liu and P. Pillay, Multilevel selective harmonic elimination PWM technique in series-connected voltage inverters, IEEE Trans. Ind. Applicat., vol. 36, pp. 160-170, Jan/Feb. 2000. J D Van Wyk. .Power Quality, Power Electronics and Control in Proceedings EPE.93, 1993, pp 17-32. Z Yang and P C Sen. Recent Developments in High Power Factor Switchmode Converters, IEEE Proceedings CCECE.98, 1998, pp 477480. H Akagi. .New Trends in Active Filters for Power Conditioning.. IEEE Transactions Industry Applications, vol 32, November/December 1996, pp 1312-1322. H Endo, T Yamashita and T Sugiura. .A High Power-factor Buck Converter, Proceedings IEEE PESC.92, 1992, pp 1071-1076.

M3

M2

FQP17N40

B_TO_MOTOR A_TO_MOTOR M4 M5

M1

FQP17N40

M6

347