Technical Witts, Inc. (Flagstaff AZ), a supplier of intellectual property, simulation software, and educational and engineering services to the electronics industry, has filed suit against Skynet Ltd. (Taipei, Taiwan), charging that Skynet violated one of its patents. Technical Witts holds 19 patents in the United States for power conversion technology, which is used in a broad range of electronic system applications and electrical appliances.

Technical Witts is seeking damages from infringement as well as an injunction against the importation of infringing products being manufactured by Skynet and sold in the United States. The Technical Witts patent (US Patent 5,402,329) teaches the implementation and structure of low cost single-ended power converters which can operate at significantly higher switching frequencies while also achieving gains in active-on energy efficiency. These benefits are achieved through a zero voltage switching (ZVS) mechanism which leads to reduction or elimination of switching losses in the power semiconductors and elimination of rectifier reverse recovery effects. The ZVS mechanism suggests a power semiconductor design optimization process that enables significant reductions in power semiconductor conduction losses. In addition to reduced power semiconductor losses, lower EMI noise generation can be achieved without requiring the use of complex snubbing circuitry. Adding to this impressive list of power supply improvements are improved cross-regulation of multiple outputs and the ease of incorporating synchronous rectification for lower conduction losses in low-voltage outputs.

The Technical Witts patent (US Patent 5,402,329) teaches a flyback topology, illustrated in figure 1, and a forward, or coupled inductor buck, topology, illustrated in figure 2. Both of these topologies rely on the energy stored in a small inductor, L1, to provide the energy needed to drive the main switch, S2, to zero volts during the turn on transition of S2. The switches S1 and S2, shown together with their intrinsic capacitance and body diode, operate alternately (in anti-synchronization) with short dead times between the operation of the switches. During the time that S2 is conducting, the current in L1 flows in the same direction as magnetizing current in the transformer T1. During the time S1 is conducting, current in L1 falls, then reverses, and reaches a maximum current in opposition to the magnetizing current in T1 at the end of the conduction period of S1. The energy stored in L1 at the end of the conduction period of S1 is sufficient to drive S2 to zero volts for a wide range of line voltages and load currents. L1 also forces the current in the secondary circuit to drop gradually to zero amps, thereby eliminating reverse recovery effects that might otherwise afflict D3 and contribute to EMI and switch S2 turn on losses.

Zero voltage switching is accomplished by controlling the switches to provide a small delay between the turning off of one primary switch and the turning on of the other primary switch. During this "deadtime" when both switches are off, the current in L1 will force a resonant voltage transition between the two switches to effect "lossless" switching. This action can be described as follows: Assuming that S2 is conducting, opening S2 quickly will transfer the current from L1 into capacitor C2 with zero voltage turn off of S2, as the voltage on C2 cannot change instantaneously. The current from L1 will then flow into both C1 and C2 (effectively in parallel) to raise the voltage at their interconnecting node to the point where body diode D1 becomes forward biased. At that time switch, S1 can then be activated "losslessly" at zero voltage. During the time that S1 conducts, the current in L1 drops to zero and reverses to a maximum in the opposite direction. At that time, S1 is turned off quickly at zero volts, as C1 cannot change voltage instantaneously, and current from L1 flows out of capacitors C1 and C2 until the body diode D2 of S2 becomes forward biased, at which time S2 is turned on at zero volts, and the cycle repeats. An important and necessary element of this technology is the inductor L1. This component can be the intrinsic leakage inductance of T1, but there are significant advantages to using a small separate auxiliary inductor. The main advantage is the access provided to both terminals of the inductor, which enables clamping the voltage at the node which connects L1 and T1. Without the clamp diode connected between L1 and T1 the node connecting L1 and T1 will ring above and below ground driven by stored energy in L1 and C3.