TMS320F28377S: Dead band while pwm sync happens

Depends on how you're generating the deadband. The TRM doesn't show any direct interaction between SYNCI and the deadband submodule. If you're generating from OUTA and OUTB from just one internal AQ output (ePWMxA or ePWMxB in figure 14-33 of the TRM), then there should be no problem, since the deadband is set up entirely within the DB submodule, which is unaffected by SYNCI. No matter what happens to the edges of ePWMxA or ePWMxB, the DB submodule will always produce the same delays relative to those edges.

On the other hand, if you are using both ePWMxA and ePWMxB to generate OUTA and OUTB, then it's plausible that SYNCI might affect the relative timings of their edges (since SYNCI can control the operation of the AQ submodule). This might affect the deadband of OUTA and OUTB. Maybe.

Also SYNCI feeds into the tripzone module as well, so that's something to check for.