Thanks Bert....let me see - so it's the FAspread that sets how far the deviation from mean that can be accepted .... with the RejCount counting how many 'deviations from mean' there have been - once the Rejcount number is met, then the new level of incoming sample becomes a valid sample to average against?