Originally Posted by
yoko1
Could also be the result of attempt at using the manual trim wheel, as it was a relatively small movement. It is also possible that it was an anomaly in the data.
Effectively, when a pilot discovers that he is unable to trim manually his airplane, he may be tempted to try in the opposite direction for checking if the system is mecanically blocked. I read a description of such an action in a PC-6 accident report.
However, in such a case, the trim movement is brief.
The report states and shows in the diagrams that the stabilizer position moved gradually in the AND direction over a period of 2 and half minutes.
This gradual movement can't be imparted as a manual action.
And, as indicated by fdr, an error in DFDR data calibration would produce a constant offset
My best guess is that the brakes on the trim clutch are not totally effective in an extreme out-of-trim situation at high speed.