I am using the ADuM3223 gate driver to drive the gates of an IGBT H-Bridge. The control circuit is similar to the one shown in Figure 1 of the application note "Inside iCoupler Technology: Driving an H Bridge with ADuM3220 Isolated Gate Drivers" where two control signals are used to fire the four gates.
I have an issue with the gate pulse width. The attached document "ADuM3223 Anomaly" illustrates what I am seeing. If the gate pulse width is less than 4us, the gate doesn't appear to properly turn on the IGBT as seen in the top scope waveform of the ADuM3223 Anomaly document. If I increase the gate pulse width to about 4.5us, the gate control works as expected as seen in the bottom scope waveform.
I am suspecting the bootstrap charging of the high side gate driver. At present, I have 2.2uF in parallel with 0.1uF ceramic capacitors to charge the high side gate drivers with an isolated supply of +12V.
Could the bootstrap circuit be a likely cause of this problem?