I developed a buck converter board driven by a ADuM7223. The high side power rail (VDDA) is charged by a charge pump from a 10V DC source which is also fed the VDDB (shown below, the VOA drives QA_H, VOB drives QA_L and GNDA is connected to the common node of LA,QA_H, QA_L).
When there was no load applied at the output of the buck converter and there was no PWM siganls sent to the VIA and VIB, I found there was a residual voltage (7~8V) exsiting at the output terminals (Vo shown below). Since the FETs didn't work yet, this voltage could only come from the 10V DC source. Then, I removed the charge pump (diode) and this residual voltage disappeared. I also check how this residual voltage genrates, this voltage climbs up slowly after the 10V DC is applied until it reaches the stable point. and if a 1K resistor connected at the Vo, the residual voltage can be 2V.
So, it seems like there is a quiescent current (I guess it's about 2mA) flowing the path 10V+ ->diode->VDDA->internal circuit of ADuM7223->GNDA->LA->Co->10V- once the VDDA is powered up.
It's just my guess based on the test results. So, please kindly help to correct me if there is anything I misunderstand.