John,
Other micros I have worked with insist on an 0.1 cap right on the power supply pins of the chip and I notice there is none on the circuit just shown. Maybe it will help. Also, just wondered if you have tried "holding off" the power to the chip or the hall effect until the input regulator has settled?? I don't know, it may be something along those lines. I know that working with 555's you can get false triggering at startup and I use a simple "hold-off" on the reset there that works.