@jandjfrench ,
What you described above sounded very much like a typical scenario related to that calibration procedure. It has happened many times since folks started doing this stuff, back in the ST10 controllers.
Here is the typical story:
1.) A problem is noted in controller performance.
2.) A check of the hardware monitor shows anomalies in controller response.
3.) No actual repair is performed. (Usually cleaning or exercising the controller would be all that is needed)
4.) Instead of actual repair, the calibration procedure is performed.
5.) Hardware Monitor is checked, and all looks good. (The program is now accommodating the anomaly)
6.) Maybe do several flights, no problems.
7.) Then, without warning, everything goes south.
What happened? The anomaly, which was never repaired, has now shifted. Maybe to something else bad, but more likely, it just cleaned itself up, and is sending correct signals. But the controller is acting up because you programmed the anomaly into it, and the anomaly is either gone, or has shifted to something else.