Adam noticed that his G-5500 was moving Clock-Wise every time the system was powering on. If the rotor were installed up on the tower, this would not have been so obvious, except that the rotor is still here in the shack.

Aside from the internal 300mS delay with the Teensy 3.2 start-up USB initialization delay, it appears the K3NG code is getting caught up doing things before the MCU pins are initialized and set.

The Teensy 3.x and LC USB delay is discussed here for those interested:

Making this simple edit to the k3ng_rotator_controller.ino file will help minimize unnecessary mechanical stresses on your antenna array.