time. At last, the switch M1 is still off, however, either diode
D1 or D2 is turned off because the voltage across the coil is
finally below the output voltage. The split capacitor topology
proposed by [7] (CS1 and CS2 in Fig. 1) is adopted to boost
the input in both positive and negative half cycles. In the
positive half cycle, D2 is off and CS1 is being charged. In
the negative half cycle, D1 is off and CS2 is being charged.
The three capacitors, CL, CS1 and CS2, share energy through
charge redistribution [7].
A feed-forward control, instead of traditional feed-back
control [7], is used in this paper. The control circuit is
illustrated in the bottom half of Fig. 1. In the feed-forward
control, the control signal is generated by comparing the
voltage signal (vc1) that is proportional to the absolute value
of vin (or vaux) to a saw-tooth signal generated by TL494
[8]. vc1 is labelled in Fig. 1 and the measured vc1 is captured
in Fig. 2. The vc1 is originated from the induced voltage in
the auxiliary coil, vaux. Since the auxiliary coil is next to the
primary coil, vaux is able to track both the magnitude and the
phase of vin at the primary coil. vaux is first amplified, and
then rectified (flipping its negative cycles to positive ones)