This invention pertains generally to AC to AC electrical power converters and particularly to dual bridge matrix converters.
A variety of circuit topologies have been developed using solid-state switches for conversion of AC power at one frequency to AC power at another frequency. Among the many circuit designs are conventional AC to DC to AC converters, in which the AC power is rectified to a DC voltage applied across DC bus lines and the DC voltage is then converted to AC by an inverter, and matrix converters, in which the input AC power is not rectified but is directly converted using a matrix of bidirectional switching elements (conventionally formed of pairs of transistors). The main advantages of matrix converters are adjustable power factor (including unity), bi-directional power flow, high quality power output waveforms, and the possibility of a more compact product because a large energy storage component (such as a DC bus capacitor) is not needed. However, the matrix converter has not been widely adopted. One reason is that the conventional modulation algorithm for such converters requires an involved and difficult pulse width modulation (PWM) switching strategy. A complicated commutation scheme and an elaborate multi-diode clamp circuit typically must be used for safe operation. See P. Nielsen, et al., “New Protection Issues of the Matrix Converter: Design Considerations for Adjustable Speed Drives,” IEEE Trans. on Industry Applications, Vol. 35, No. 5, 1999, pp. 1150–1161.
A relatively new converter topology is the dual bridge matrix converter. See, L. Wei, et al., “A Novel Matrix Converter with Simple Commutation,” Proceedings of 36th IEEE Industry Applications Society Conference (IAS '2001), Chicago, Ill., USA, 2001, Vol. 3, pp. 1749–1754. The reason this topology is also referred to as a matrix converter is that it shows the same input/output performance as conventional matrix converters, and can also be described by switching matrices similar to the conventional matrix converter. The dual bridge matrix converter also has many of the advantages of the conventional matrix converter, including near sinusoidal input/output waveforms, adjustable input power factor, and a compact physical package because no large energy storage components are required. The dual bridge matrix converter has several advantages over the conventional matrix converter, including reduced difficulty of commutation since all line-side switches turn on and off at zero current and all load-side switches commutate similarly to a conventional DC/AC inverter, and the number of switches required can be reduced under certain constraints. A nine-switch dual bridge matrix converter has been developed that has the least number of switches while still providing high quality input and output waveforms. Three switches are utilized on the input side and six switches are utilized on the output side for three-phase operation. However, a disadvantage of this converter configuration is that its DC link current must be non-negative to guarantee safe operation. If the DC link current becomes negative, some high voltage spikes can be generated because there are no reverse current paths in the line-side converter, and the converter may be damaged by these spikes. It has been suggested that the output power factor should always be higher than 0.866 to guarantee safe operation of the converter. See J. W. Kolar, et al., “Novel Three-Phase AC/DC/AC Sparse Matrix Converter,” Proceedings of 17th IEEE Applied Power Electronics Conference and Exposition, APEC 2002, Vol. 2, 2002, pp. 777–791, and L. Wei, et al., “Matrix Converter with Reduced Number of Switches,” Proceedings of IEEE Power Electronics Specialists Conference, PESC '02, 2002, pp. 57–63.
In accordance with the invention, a dual bridge matrix converter has an input (line-side) converter with controllable switches that receives AC power and provides unidirectional power to high and low DC link lines, and a full bridge output (load-side) converter which receives the power from the DC link lines and provides AC power to output lines. For three-phase operation, the input converter preferably has three switches, each of which is connected by two diodes to one of the input lines and by diodes to the DC link lines, and the output converter has six switches, a pair for each phase leg, which are controlled to provide PWM output waveforms on output lines connected to junctions between each of the pairs of switches. A DC link clamp having a series connected diode and capacitor is connected across the DC link lines. Any negative DC link current flows into the clamp capacitor through the clamp diode for a short period of time to reduce or eliminate high voltage spikes. Because the clamp capacitor is not required to store large amounts of energy to supply the output inverter (as is required in conventional AC to DC to AC converters), the clamp capacitor can be relatively small, inexpensive, and low rated. The DC link clamp may also include an additional switch connected in parallel with the clamp diode to allow exchange of power between the clamp capacitor and the output converter, allowing operation under low power factor conditions.
In operation of the invention utilizing a clamp switch in parallel with the clamp diode, when the converter is first started, all of the switches on the line-side converter turn on initially and the clamp capacitor voltage is charged up to the maximum peak-to-peak line voltage. Upon occurrence of a fault state, all switches in the converter are turned off immediately, and the energy stored in an inductive load will flow into the clamp capacitor to avoid high voltage spikes. During normal operation, if the clamp capacitor voltage is above a threshold voltage that is higher than the peak-to-peak input voltage, the clamp switch is turned on. Under these conditions, the line-side converter is anti-biased because the clamp capacitor voltage is higher than the maximum amplitude of the input line voltage. In this condition, the voltage of the clamp capacitor begins to reduce because it provides power to the load. During normal operations, if the clamp voltage is lower than the threshold voltage, the clamp switch is turned off and the line-side converter switches operate normally to supply power to the DC link.
Further objects, features and advantages of the invention will be apparent from the following detailed description when taken in conjunction with the accompanying drawings.
In the drawings:
With reference to the drawings, a dual bridge matrix converter with a DC link clamp in accordance with the invention is shown in schematic form generally at 20 in
With reference to
The load-side converter 33 is preferably a full bridge converter, as illustrated in
In accordance with the invention, the clamp circuit 38 is connected across the DC link lines 24 and 25 to suppress voltage spikes as a result of back current flow from the load-side converter 33 toward the line-side converter 21, which is blocked from flowing through the line-side converter 21 by the diodes 48 and 49. The clamp circuit 38 includes a series connected diode 55 and a capacitor 56, with the diode 55 arranged to conduct current from the DC link high line 24 through the capacitor 56 to the DC link low line 25. In contrast to clamp circuits for conventional matrix converters, which generally require a full bridge composed of several diodes, only one diode is required for the clamp 38, a considerable savings in cost and simplicity of circuitry. Because the clamp capacitor 56 is not required to store energy for operation of the converter, it may be relatively small, inexpensive, and low rated as compared with the large and potentially failure prone electrolytic capacitors typically required for the DC link capacitors of conventional AC to DC to AC converters.
Under certain conditions, particularly with low output power factor, current may flow for a sustained period of time from the load-side converter 33 on the DC link lines toward the line-side converter 21. This current is diverted through the clamp diode 55 to the clamp capacitor 56, which continuously charges as long as current is flowing through the diode 55. As shown in
The following describes a space vector PWM control that can be carried out by the controller 40 under normal conditions with high output power factor, and further describes the effect of the clamp circuit 38 on the converter control under low output power factor conditions.
Under normal conditions for the converter of
In order to simplify the analysis, it is assumed that there is no input filter on the line side. Referring to the notation used in FIGS. 2 and 3:
Lf=0; Cf=0; Vsx=Vx; isx=ix; x=a, b, c
It is assumed that the input source voltages are described by
and the output currents are
where ωi and ωo are the input and output angular frequencies, θ0 is the angle of the expected output voltage vector and φ0 is the output power factor angle. Vm and I0 are the amplitudes of input voltage and output current respectively.
To help determine the conditions for safe commutation, six intervals of a switching cycle can be identified, based on detection of the input voltage synchronization angle as shown in
Two portions are also identified in each switching cycle interval. In each portion, some appropriate switching behavior can be analyzed on different switches according to the number of intervals to replace the double bridge matrix converter topology as a DC/AC inverter.
For example, Vsc in interval 2 has the largest absolute voltage, the two largest positive line voltages are Vsa−Vsc and Vsb−Vsc, respectively. The line-side switching states in each portion can be determined by the following:
In portion 1, Sbm and Scm remain turned on; Sam remains turned off. The DC side voltage Vdc is then equal to Vsb−Vsc, the DC side current idc equals isb and −isc, and isa equals zero. The duty cycle of this portion is defined as dbc.
In portion 2, Sam and Scm remains turned on; Sbm remains turned off. The DC side voltage Vdc equals Vab−Vsc, the DC side current idc equals isa and −isc, and isb equals zero. The duty cycle of this portion is defined by dac.
In the two portions of interval 2, the converter can be considered as an equivalent DC/AC inverter with different DC voltages during each of the two portions.
Initially, it is useful to consider the conventional voltage source inverter with three-phase output voltage Vsu, Vsv, and Vsw supplied by a DC voltage source Vdc=3Vm/2. In complex form, the space vector of the desired output voltages is
Where 0<k<√{square root over (3)}/2 is a constant.
Assuming 0<θ0<π/3 and that the system operates in interval 2, this vector can be approximated by its two adjacent voltage vectors (V1 and V2) and the zero voltage vector V0, as shown in
The DC current of the inverter as voltage vectors V0, V1 and V2 can be expressed as 0, isu, and −isw respectively. The average DC current of the inverter with the above duty cycles is determined as
idc=k·I0·cos(θ0−θ0i)=Iim (5)
Because there are two portions during each switching cycle, the duty cycles V1, V2, and V0 are also distributed to each portion. During the first portion, they are:
d1bc=d1·|cos θb|; d2bc=d2·|cos θb|
d0bc=d0/2; dbc=d1bc+d2bc+d0bc (6)
During the second portion,
d1ac=d1·|cos θa|; d2ac=d2·|cos θa|
d0ac=d0/2; dac=d1ac+d2ac+d0ac (7)
Combining from Equations (3) to (7), the actual average output voltage vector and the input current can finally be obtained as
This result demonstrates that the space vector PWM control method generates the same actual output voltage as the reference voltage and that the line-side power factor can inherently remain at unity.
When the system operates during the other intervals or when θ0>π/3, the same results can be obtained.
Since the DC side current equals to isu while the output voltage is V1, then from equation (2)
idc=I0 cos(θ0−φ0)≧0 (9)
Thus we have
Because 0<θ0<π/3, the following equations can be derived from (10)
On the other hand, the DC side current equals to −isw while the output voltage is V2, then from equation (2)
Thus we have
Because 0<θ0<π/3, the following equations can be derived from (13)
Thus, to apply the space vector PWM control method to the 9-switch converter topology of the converter circuits of
This is strict limitation for the 9-switch topology if there are no additional circuits in the DC link to provide paths for the negative current. Consequently, the application of this topology is somewhat limited; without the clamp circuit 38, for instance, it cannot serve as an induction motor drive. A clamp circuit as shown in
The clamp circuit of
When the converter is started, all of the switches in the line side converter 21 turn on initially and the clamp capacitor voltage is charged up to the maximum peak line voltage.
Under a fault state, all switches in the converter 20 are turned off immediately. The energy stored in the inductive load flows into the clamp capacitor 56 to avoid high voltage spikes.
During normal operation, if the clamp voltage Vcl is higher than the threshold voltage Vth, the clamp switch Sc is turned on. Then the line-side converter is anti-biased because the threshold voltage is higher than the maximum amplitude of the input line voltage. In this condition, the voltage of the clamp capacitor begins to reduce because it provides power for the load.
During normal operation, if the clamp voltage Vcl is lower than the threshold voltage, the clamp switch Sc is turned off. The line-side switches start to operate again.
From the analysis, it can be found the converter can operate safely when the output power factor is lower than 0.866. However, since the space vector PWM is not applicable in this condition, some low order harmonics will be generated when the output power factor is lower than 0.866.
The duty ratio calculations are then carried out at 73 for each space vector, and the PWM sequences are determined at 75.
It is understood that the invention is not confined to the particular embodiments set forth herein as illustrative, but embraces all such forms thereof as come within the scope of the following claims.
This invention was made with United States government support awarded by the following agency: NSF 9731677. The United States government has certain rights in this invention.
Number | Name | Date | Kind |
---|---|---|---|
4628425 | Venturini et al. | Dec 1986 | A |
4864483 | Divan | Sep 1989 | A |
6185115 | Sul et al. | Feb 2001 | B1 |
6330170 | Wang et al. | Dec 2001 | B1 |
6762947 | Hammond | Jul 2004 | B2 |
6850424 | Baudelot et al. | Feb 2005 | B2 |
6856038 | Rebsdorf et al. | Feb 2005 | B2 |
20040136210 | Oh | Jul 2004 | A1 |
Number | Date | Country |
---|---|---|
1 280 263 | Jan 2003 | EP |
1 289 112 | Mar 2003 | EP |
Number | Date | Country | |
---|---|---|---|
20050099829 A1 | May 2005 | US |