Electric power conversion systems – Current conversion – Including d.c.-a.c.-d.c. converter
Reexamination Certificate
2000-09-29
2002-04-02
Vu, Bao Q. (Department: 2838)
Electric power conversion systems
Current conversion
Including d.c.-a.c.-d.c. converter
C363S016000, C363S089000, C323S266000
Reexamination Certificate
active
06366474
ABSTRACT:
FIELD OF THE INVENTION
The invention relates to a switching power supply (SPS) converting an AC or DC input voltage into an unregulated high frequency AC output voltage or a regulated DC output voltage. Conventional topologies are implemented, wherein power factor correction and/or resonant operation is accomplished.
BACKGROUND OF THE INVENTION
By definition, any switch operates in binary mode. It is either turned on or turned off. However, the switch operates in linear region during a transition interval. In particular, numerous parasitic capacitances and inductances appearing in a switched circuit must be recharged. These components often form resonant networks that generate parasitic oscillations. The oscillations contribute to energy loss and always-undesirable EMI/RFI. A soft switching is used to reduce these effects. The linear region of the switch is expanded; the parasitic components are recharged at a reduced rate while the switch acts like a snubbing resistor. However, except for specific low power applications, energy loss of the switch is unacceptable. A hard and thus uncontrollable switching has exactly the opposite effect as merely losses of the switch are minimized.
Generally, resonant SPSs are inherently inferior to square-wave SPSs. The resonant SPSs develop a sinusoidal voltage and/or current. However, this is associated with a considerable interval, usually near zero crossing of the respective sinusoidal signal, when energy transfer is minimal or none. Therefore, the resonant SPSs require power semiconductors having higher current and/or voltage ratings. An output capacitor carries a high ripple current. By contrast, in the square-wave SPSs a maximum energy transfer is accomplished instantaneously.
The resonant SPSs have other inherent flaws. The resonant or switching frequency is determined by additional LC components, or an additional capacitor and leakage inductance of the power transformer. These components are fixed and so is the turn-on or turn-off time. Pulse frequency modulation (PFM) is often used to accomplish the regulation. Implementation of pulse width modulation (PWM) or other preferred switching scheme is often impractical. A minimum load may be necessary to avoid large variations of the switching frequency. Moreover, the maximum switching frequency can be quite high if it is determined by the leakage inductance. The resonant SPSs are preferred in special applications, such a high voltage conversion. General purpose resonant SPSs are costly and constitute a small fraction of all SPSs manufactured today.
Linear power supplies have tremendous advantages over SPSs. Any SPS is basically a high power oscillator that is rich with switching harmonics. Moreover, if AC line is used as the input source, numerous line frequency harmonics are created. The SPSs are therefore excellent sources of both conducted and radiated EMI. Energy is transferred from the input to the output in quanta. A current applied to an output capacitor is zero or otherwise never constant, even when the load is fixed. Therefore, the output ripple voltage delivered to the load is significantly larger. Statements about obsolescence of linear power supplies in view of apparent superiority of SPSs are common and greatly exaggerated. The staggering popularity of the linear power supplies is a clear evidence of unsolved and severe deficiencies in the SPSs. For example, a two-stage regulation is commonly used. A low-dropout (LDO) regulator follows an SPS in order to combine relatively high efficiency with low noise and fast transient response. Newest LDO regulators operate without an output capacitor, which at least maintains the gap between the linear power supplies and the SPSs.
SPSs exist only because of smaller size and weight at given output power. Generally, switching frequency and efficiency of an SPS determine its size and weight. The output ripple voltage is reduced by increasing the switching frequency and enforcing post-filtering. However, the frequency may be severely limited by a switching scheme, in particular PWM, rather than switching capabilities of real components. Post-filtering increases size and power losses, and severely affects stability and transient response. Efficiency is most effectively boosted by optimizing losses during transition intervals and minimizing conduction of switching components. Therefore, MOSFETs and synchronous rectifiers are favored. Additional reactive components are used to minimize voltages and currents that the switches endure during transition intervals. Therefore, zero-voltage and zero-current switching schemes are favored.
The performance of the power transformer operating at a high switching frequency can be understood with the help of a simplified model. Primary and secondary leakage inductances are caused by incomplete magnetic coupling between primary and secondary windings of the transformer. Primary and secondary resistances reflect copper loss of the respective windings. Skin effects further increase the resistances. The leakage inductances and the resistances are effectively coupled in series with the respective windings, wherein transformed voltages are reduced. Primary and secondary intra-winding capacitances establish resonant networks. Moreover, rapid recharging of the transformer causes current spikes. Inter-winding capacitances further contribute to performance limitations. A magnetizing inductance is determined by permeability and crossectional area of the magnetic core, and by the number of turns. A magnetizing resistance represents core loss. Eddy-current loss increasing with the switching frequency, hysteresis loss increasing with flux density and residual loss due partially to gyromagnetic resonance contribute to this resistance.
The leakage inductance of the power transformer is often the key parasitic element and the largest single factor in degrading the performance of an off-line SPS. An ideal transformer has no leakage inductance because magnetic fields generated by the primary winding are entirely contained by the core and coupled completely to the secondary winding. The leakage inductance represents magnetic fields that do not couple with both the primary and secondary windings. The inductive leakage energy is transferred to a snubber capacitor. A diode is employed for charging the capacitor to a peak voltage. However, parasitic capacitances are charged to that voltage as well. As the diode ceases the conduction, the respective resonant circuit continues to oscillate at a very high frequency. Obviously, the oscillations appear directly at the input. They are also applied to the collector or drain of the power switch, usually to its case, and further to a heat sink. Moreover, the oscillations are transferred to all other windings and are thus applied to all components connected thereto. Numerous filters, snubbers, shields and beads redistribute and reduce somewhat leakage energy that is wasted in form of heat and EMI/RFI.
The value of the leakage inductance is determined primarily by physical dimensions. For example, safety regulations of various countries require specified amounts of spacing and insulation between the windings. Turns ratio of a typical off-line transformer makes it difficult for the secondary winding to uniformly cover the primary winding. Furthermore, economics may be an overriding consideration since some core shapes are easier and less expansive to wind than others are. Transformer design becomes a tradeoff between these and other factors that always result in a significant amount of the leakage inductance.
SUMMARY OF THE INVENTION
The present invention is intended to provide SPSs incorporating power factor correction and/or switching at resonant transition. Conventional topologies are implemented. These include flyback, forward, half-bridge, full-bridge and push-pull SPSs. The invention offers the perfect compromise between the soft and hard switching schemes. Any switch is driven hard, whereas voltages across critical passive components vary at a rate predetermined by a resonant network.
LandOfFree
Switching power supplies incorporating power factor... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Switching power supplies incorporating power factor..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Switching power supplies incorporating power factor... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2883464