MGL Avionics User Forum

Since installing my efis lite there have been problems with the built in AHRS. The Artificial horizon cannot be calibrated to straight and level and once a turn is completed it is all over the place.
On further inspection during flight I notice that in the Sensor Diagnostics section it says

"internal gyro Failed" and "internal accelerometer failed"

Is this normal and if not what can be done to rectify this?

Newer firmware for your EFIS corrects the "fail" messages as these refereed to the internal gyro/accelerometer chip of the older lite systems. Newer systems have a plug-in modular AHRS that in effect is a SP-7 (Using similar gyro but a digital accelerometer chip).
Regardless of which - your description strongly points to vibration of your panel that is similar in frequency to the sample rates of the AHRS (most likely higher frequency harmonics of the vibration). Worst in our experience is a combination of vibrations caused by a geared prop drive and the engine itself.
Some rotor craft also can impose significant challenges to your inertial sensors.

If it is not possible to dampen the vibration the only viable solution is an external AHRS as that gives you the freedom of mounting it in a much better location and even if those locations are challenging it is easier to get rid of vibrations.

The AHRS wants to only "see" the actual movements of the air-frame. Vibrations are problematic as this is physical movement as well but much faster than gyros can follow so the system quickly builds up significant measurement errors. Worse - each new calculation is now based on a wrong assumption.

The errors tend to build up much faster when you do a turn compared to straight and level - that is because it is easier for the AHRS to correct for the errors in that attitude.

One of our competitors actually brings up a warning message when vibrations are detected that are too much to handle. Maybe we should do something similar.

Rainier, I am attempting to debug a similar issue and would like to know a little more regarding the raw sample rate of the AHRS? It sounds like these harmonics are affecting the filter. In more technical terms, are these harmonics above the nyquist frequency of the sampling and thus are being aliased so they can't be attenuated effectively by the digital filter?