Autopilot tuning: Altitude Capture Procedure.

[cbretana]

I am seeing unsatisfactory autopilot behavior during Altitude Capture that I cannot find in the tuning guide. In my case the aircraft overshoots and undershoots the target, getting further away from the target on each iteration.

In figure 5 of the in-flight tuning guide, (page 5-4), one decision point is close to what I am seeing but is not exactly the same. There, it says, "... aircraft over and undershoot target, gradually getting closer". Is this the same as what I am seeing, (i. e., should I use the corrective action associated with this behavior - increase VSI Gain), in my case? Or should I decrease VSI Gain? Or is what I am seeing more appropriately addressed by some other point in the tuning procedure?

 

[Marc_J._Zeitlin]

I am seeing unsatisfactory autopilot behavior during Altitude Capture that I cannot find in the tuning guide. In my case the aircraft overshoots and undershoots the target, getting further away from the target on each iteration.

In figure 5 of the in-flight tuning guide, (page 5-4), one decision point is close to what I am seeing but is not exactly the same. There, it says, "... aircraft over and undershoot target, gradually getting closer". Is this the same as what I am seeing, (i. e., should I use the corrective action associated with this behavior - increase VSI Gain), in my case? Or should I decrease VSI Gain? Or is what I am seeing more appropriately addressed by some other point in the tuning procedure?

Charlie - if you look at the flowchart to which you're referring, the box that says "...gradually getting closer" has two exits - one says "yes", the other says "no". Your plane is doing "no". So the fix is to "Decrease Altitude Gain".

In a Long-EZ and COZY III in which I recently installed Garmin G5 based autopilot systems, when I first turned on the A/P with the default altitude gain, I quickly got the behavior you're seeing, but not so gradually . Loads of fun. I turned down the gain, and it got closer to reasonable, then turned it down some more and the behavior became wonderful.

I'll bet that turning down the altitude gain does wonders for this behavior - then you can continue the flowchart in the opposite direction .

 

[cbretana]

Marc,
HAH!! Thanks for embarrassing me!
Guess I should read the whole thing before posting! I assumed that the "No" meant it is not doing the bad thing described in the diamond. But both Yes and No branches are still faults. In fact. everything on the entire branch of the chart after "...pass through commanded target" is a bad fault with a specific corrective action.

As it is written, the Yes implies that Yes, the aircraft is over and under shooting the target, and is getting closer with each pass, and the No implies it is not over and undershooting the target (obviously contradictory given the choices prior to this point in the flow). Whereas what the "Yes/No" actually means is just whether the oscillations are getting bigger or smaller.

The chart would be clearer if they used different words for the outcome branches. That diamond should read "After passing through target altitude on each cycle, ...",
and the two result choice branches should read "... altitude oscillations/deviations increase ..." and the other branch should read " ... altitude oscillations/deviations decrease ... "


Charly