D-100 autopilot wandering in HDG

[morhall]

I've got the D-100 installed in a Glasair II-RG. The servos had just been replaced when we bought the plane, traded up for the larger servos from smaller ones.

This meant the autopilot was in need of reconfiguring. Over the course of a few flights the prior owner and I tuned the settings for torque and sensitivity. It seemed to be working great, athough I wasn't using the AP much.

On occasion, the AP while in HDG mode will wander substantially from the HDG bug. Overshooting by maybe 15-20 degrees and doing this immediately after the AP is turned on.

On a different flight, it might act perfectly stable. Turning up the sensitivity makes the stick very twitchy as the servo jumps back and forth.

So far a sensitivity of 10 and torque of 25% seems good. On NAV mode this produces a nice steady track and subtle adjustments.

When the autopilot is working in HDG mode, it seems fine but it is inconsistent in whether it works properly or not. I certainly don't trust that I know what is causing it and whether or not it is going to behave or not.

The last time I used it, worked perfectly. The time before, wandered like it was really drunk.

Relevant Info:

D100
HS34
Garmin 430W
no OAT - bought one, haven't had a chance to install it yet.

Dynon firmware version is not at the latest 5.5. I think it was 5.4.3, but I didn't write it down.

HDG mode only seems to be the issue.
NAV mode has always been stable. Changing sensitivity will cause a twitchy stick as well, but no wandering.
TRK Mode - I don't think I've tested this.

Typical test mode when I experience the problem.

AP - Alt Hold On
Push the Heading Button to set the bug on the current heading.
NAV Source - GPS
GPS CDI set to GPS
Set mode to HDG
Turn AP HDG on

AP begins to wander almost immediately. The first few times I experienced it, I disconnected pretty quickly because we were 30 degrees off heading. Almost like it was trying to go to the NAV line and not fly HDG.

Hopefully there is something I am doing wrong with a setup or config.

Morgan

 

[Dynon]

Knowing the performance of TRK mode is the one piece of data you really want to take. It's essentially "follow the bug", just like heading, but it uses GPS ground track as its target instead of magnetic heading. If TRK works but HDG doesn't, then there's likely a magnetic heading calibration issues that's preventing the AP from properly targeting a precise heading. If you also have the issues when in TRK mode, then there's a more fundamental tuning or installation problem on that axis. So my first step would be to test TRK mode and compare the performance.

Next, take a look at our SkyView autopilot in-flight tuning guide at http://docs.dynonavionics.com . It's written for SkyView, but the concepts and settings are mostly the same and can be adapted for your aircraft.

One more thing: Check your turn rate setting. If it's set to something non-zero, know that turning it down to zero enables "bank angle mode", where the autopilot uses bank angle instead of turn rate as it's primary goal. This "secret mode" (it'a actually documented in the manual, but we never gave it a discrete switch with a clear name) worked so well that in SkyView that's just the way the AP works, period.

 

[morhall]

I tested HDG and TRK modes on a flight yesterday. HDG started off ok, and then after 15 seconds or so started to wander significantly before I disconnected.

I switched to TRK mode and it seemed solid. Turning the bug 5 degrees it gradually rolled to that new course and locked on.

I was not prepared with the tuning guide to run through that in the air. Next time I have a chance to go up with a safety pilot I will give that a go as well.

 

[Dynon]

If TRK works well and HDG doesn't, then your magnetometer (ADAHRS) location and/or calibration isn't ideal. So I'd look at that first, then move on to the tuning guide.

 

[morhall]

Does the compass calibration need to be performed any time you disconnect it from the D100? Or just once per airplane/installation location.

My partner calibrated the compass last week. Over the weekend I finally got around to installing the OAT probe I had purchased. That of course requires removing the connector from the compass and adding in the new lines. The OAT works well, but the HDG mode of the AP is still wandering. Probably 5-10 degrees in a slow oscillation centered on the desired heading.

I need to try the troubleshooting defined in the article linked in an earlier post. I haven't tried a sideslip to see if the roll is causing a heading change indication.

 

[Dynon]

Generally only once per airplane, but you may need to redo it if the magnetic configuration around the compass changes (new toys, modifications to the airframe, wire rerouting to solve other issues, etc.

Did your partner get the magnetic inclination and total field intensity values from the NOAA website for your geographical location before he calibrated?

In general, does heading seem to settle immediately when you stop your turn, or does it lead or lag at all?