3 AC dropout technical challenges

The first challenge that I want to highlight is reverse-current generation when the AC input voltage disappears. Since all of the switches in the totem-pole PFC topology are bidirectional, it is essential that the FETs operating as synchronous rectifiers shut off as quickly as possible when removing AC. This shutoff prevents the generation of a negative current that will cause the output voltage to discharge and reduce the available holdup time. Figure 2 illustrates the path for generating this negative current for the synchronous conduction interval during the positive half cycle. In addition, any substantial delays in turning off the synchronous rectifiers can also result in a large current spike capable of activating overcurrent protection (OCP). For example, if the synchronous rectifier stays on when no input voltage is present, you can solve $V_{dc}=L_{b1}\cdot \frac{d{I}_1}{dt}$ for the amount of time it takes to generate 70 A of current, namely 2.5 µs. This short time presents a significant problem for the AC dropout detection to identify the problem and stop switching before the system hits OCP or causes damage.

The second challenge is resuming operation of the PFC after restoring AC. The central issue of this event comes from the fact that the bypass diodes on the PFC charges the output voltage to the peak of the input sine wave, which occurs most readily at high line when the output voltage has fallen well below this peak value. During these events, the converter has no mechanism to stop the current, making the surge current very large. Improper control of the switches during these events can make things much worse by saturating the inductors, creating OCP events and further discharging the output voltage. The need for a precise control algorithm during this time is again multiplied by the high-frequency operation point of the iTCM topology with the small-value inductors used for L_b1 and L_b2.