This subset of trials looks fine. Do you get the problem if you run just this subset? If so, could I see the data log file that is produced?