If you’re getting feedback, to me that suggests your speakers are playing back what the mic is picking up. Why do you need to do that?
Disclaimer: I’ve never used sonarworks, but I’ve considered getting it, so maybe what you’re doing is correct… but that seems backwards. You need the mic to send a signal back into the computer but there’s no reason you need the speakers to be playing what the mic picks up, that will mess up calibration.