Hi Mark,
This is a very known bug and it was in the auto trim from memory. It has been the hardest bug to kill off in my long history of working with Dynon and for that matter my entire industrial automation career. The problem had been impossible to replicate on the work bench, and the diagnostic file only showed the result, and not what caused it.
About a year back, it was found. Lawrence who had been at Dynon since the beginning gave the task to an intern, and between them they managed to replicate the issue on the bench and when they did the problem in the code was staring right at them.
Next was fixing the code and testing. Given the code was fixed and I found this problem many years back, Lawrence gave me a version of software to fly and test. I can report to you the bug is 100% gone and I have now at least 200-250 hours of bug free flying across 4 airframes.
This code is in the next release, which is I am told not far away. That promise has been going all year, but recently I was told that it was now a serious effort to get it released prior to Christmas, and my guess is it will be this month some time.
All the best,