I think that this lockloss was caused by DTT suddenly outputting a large excitation when I don't think it should have, at 1182033600. I increased the amplitude of the noise excitation in the attached template from 10000 to 20000, the template has a 5 second ramp up time, but for some reason the excitation reached an amplitude of 45,000 counts very suddenly as shown in the attached screenshots.
Since the input to the filter was off, the only thing coming into this filter should have been the excitation.