HoverflyGIMBAL 1.8 rev 0
posted this on Dec 04 08:48 pm
HoverflyGIMBAL Firmware (version 1.8 rev0)
HoverflyGIMBAL Setup Client (version 1.8 rev0)
This version of the firmware corrects the roll drift (sometimes described as a yaw axis coupling) error that was present in the previous version of the firmware. It also has a couple features that allow the user to select behaviors that were present in earlier versions of the firmware. It is recommended to first try this version with default settings for the new featues (see below - mode 0 angle and mode 0 rate calculations).
Gimbal Angle Calculation Mode
The angle calculation is used to determine the orientation of the HoverflyGIMBAL. The calculation mode can be set on the Gimbal Mode configuration page in the setup client. The default value is ‘mode 0’.
Mode 0 is a more accurate calculation of the orientation of the HoverflyGIMBAL. It is particularly important to use mode 0 when attached to a gimbal with continuous rotation servos (a gimbal where the sensor is attached to the camera tray).
Mode 1 is designed to allow users to mimic the behavior of the system in earlier versions of the firmware. It is not our recommended mode, but a number of users have expressed a preference for this behavior.
Gimbal Rate Calculation Mode
The rate calculation is used in conjunction with the ‘rate gain’ factor on a servo. The calculation mode can be set on the Gimbal Mode configuration page in the setup client. The default value is ‘mode 0’.
Mode 0 is designed to eliminate axis coupling issues that can arise when the HoverflyGIMBAL is mounted far away from the center of rotation. However it may produce some jitteriness if there is a lot of vibration in the system.
Mode 1 produces a smoother value, which allows for higher rate gains. However if the HoverflyGIMBAL is mounted far away from the center of rotation, some coupling between axes may be seen while performing more extreme maneuvers.
Roll Horizon Stabilization / Vibration Dampening
Some users have been experiencing drift on the roll axis. This appears to have been due to vibration during flight. Additional filtering on the accelerometer data as well as the calculation of the rotational rate has been added to counter this effect.