Copyright © 2016  Grand Rapids Technologies, Inc.                      



Electronic Flight Instrumentation for Experimental Aircraft Since 1991

Technical Advisory- Autopilot & Magnetometer Improvements


Summary:


Two customers reported  “small roll oscillations” or “loose autopilot performance” recently.  An investigation revealed that changes to the AHRS calculations that affected yaw axis appear to be responsible. This may be more likely when certain (normal) magnetometer errors were present. The problem was eventually duplicated in the company test airplane (an RV-6A), which will be used to validate the fix.


A software fix is being pursued that will be made available by February 28, 2017. This fix will be identified on the website. Operation of the autopilot when exhibiting this roll oscillation is not dangerous, but could result in degraded ride quality.


Read the full advisory below:


Technical Advisory- Autopilot & Magnetometer Improvements


Affected Systems:


All Mini, Sport EX, Sport SX (which use a digital magnetometer), and Horizon systems using the Adaptive AHRS.


Autopilot Symptoms:


Autopilot: Affected systems may exhibit small roll oscillations or a sense of “loose” coupling with the autopilot is engaged. Changing autopilot or servo gains do not resolve the problem.


Magnetometer Symptoms:


Calibration of the magnetometer may show larger corrections than might be attributable to local magnetic disturbances. In some cases the magnetometer may not complete fine magnetometer calibration, especially when mounted on its side (such as to a rib).


Discussion:


Customers reported the above symptoms for the affected system. Since no changes have been made to autopilot software or servos, we assumed initial reports of this problem were attributable to significant lost motion (slop) in the servo connection, or gains that were too high, and may have advised them as such.  The problem was not initially observed in our company airplane.


An investigation revealed that changes in the AHRS calculations that affected yaw rate data could result in the reported autopilot problems. This may be more likely when certain (normal) magnetometer errors were present. The problem was eventually duplicated in the company test airplane (an RV-6A), although it was subtle enough that a non-pilot passenger did not notice.


Solution:


A software fix is being pursued that will be made available by February 28, 2017. This fix will be identified on the website. Operation of the autopilot when exhibiting this roll oscillation is not dangerous, but could result in degraded ride quality.


In addition, this investigation lead to an improved method of factory magnetometer calibration. The benefits of this improved calibration are typically not significant enough to justify any action by our customers. However,  if you suspect your magnetic headings are not accurate after having performed the fine magnetometer calibration, and your magnetometer is not mounted near significant sources of magnetic disturbances, such as current carrying wires, moving ferrous metal parts, etc., then the new magnetometer calibration may benefit you. GRT Avionics will re-calibrate your magnetometer at no charge by returning it to the factory.  Note that this improvement applies only to the digital magnetometers used with the above referenced systems. Digital magnetometers can be identified by the word “Digital Magnetometer” on the label. They can also identified by their wiring, as they use only 3 or 4 wires. (Analog magnetometers use 6 wires).