Pitch Bucking during VNAV Approach

[steve m]

I was out doing some practice approaches in my RV10. I typically run 100 knots during approaches. I am running the latest software\firmware. Until today I have been quite satisfied with the A\P performance. I did a lot of tuning about 6 months ago and have been quite happy.

Today was extremely smooth weather. On two seperate GPS approaches, after VNAV was coupled I started getting "Bucking" in the pitch axis. The "Bucking" is cycling up\down about once per second. It was tracking the Glide Slope down but was bucking up\down as it tracked the glide slope.. At one point I applied pressure to dampen the cycles and it started to die out but then started up again. On this first approach I quickly got into the inflight AP setup screen and adjusted the Low Speed Pitch Sensitivity down by 2 but it did not seem to help.

I then shot another approach in opposite direction and had the same phenomena. Not sure if the perfectly calm day has any impact but I have not had this happen since I re-tuned about 6 months ago (probably 15 approaches ago?)

I am not certain if I should be looking at other parameters? Here is my current set of values.

Thoughts\comments\suggestions would be appreciated.

 

[RV8JD]

If you mean that you're on v16.5.2, I think I'll wait awhile before I update my system to this latest version.

 

[Rhino]

You might also try bumping your pitch trim a bit off center. This gives a constant force for the autopilot to work with.

 

[steve m]

If you mean that you're on v16.5.2, I think I'll wait awhile before I update my system to this latest version.

Yes, I did update to 16.5.2 prior to this flight but I did not see any Revision notes about updates to AP control algorithms?

 

[Marc_J._Zeitlin]

Check for new play between the pitch servo and the elevator pushrod. If there's substantial hysteresis in the system, this type of behavior can occur. There was also a new parameter for "play" in the system added a revision or two ago - don't remember the name - but you can look at the value for that, as well.

 

[RV8JD]

Yes, I did update to 16.5.2 prior to this flight but I did not see any Revision notes about updates to AP control algorithms?

Understand, but bugs can be introduced in SW revisions. As an example, when Dynon updated to v16.2.4 back in late 2021, many of us encountered very sharp roll inputs by the roll autopilot. Dynon initially came back and said there were no changes that should have affected the autopilot. After several of us complained about it, Dynon began looking and found the problem - From Don at Dynon at the time: “They introduced a bug that basically disabled the roll gyro feedback gains, (pitch too, it just didn't show up as much as there is also a vertical speed rate limiter in use)”. It took Dynon awhile to find the problems and design a fix. And we had to wait awhile for the next SW update which contained the fix.

The fact that you said in your post, "I am running the latest software\firmware. Until today I have been quite satisfied with the A\P performance. I did a lot of tuning about 6 months ago and have been quite happy.", made me suspicious of v16.5.2. Not saying your issue is necessarily a problem with v16.5.2, but I'll wait awhile before I update to it, just in case.

Remember, we Experimental folks are the Beta testers for the Certified folks!

 

[Carl_Froehlich]

By chance did you get the “Air Speed Low” alert during the approach? This will case the autopilot to pitch the nose down to gain airspeed.

The lowest you can set this is 1.3 times VSo1. So first make sure you have VSo1 set at something like 60 knots for the RV-10. This then lets you set the autopilot low speed limit at around 78 knots.

Carl

 

[steve m]

Yes, I did update to 16.5.2 prior to this flight but I did not see any Revision notes about updates to AP control algorithms?

Check for new play between the pitch servo and the elevator pushrod. If there's substantial hysteresis in the system, this type of behavior can occur. There was also a new parameter for "play" in the system added a revision or two ago - don't remember the name - but you can look at the value for that, as well.

I did think about this one. The parameter is called "Pitch Lost Motion Compensation". On the first approach when I tried to dampen the motion by hand, I did not feel any dead travel or clunking in the control links as the pitch "bucked" up\down. Also, the control rod system on RV's is pretty tight. I am thinking it may be more of "VDI Gain" or "VDI Dot Gain" but to me honest, I am not sure I know the difference between these two? Or possibly "Pull Rate". I think I will try to shoot few more approaches tomorrow and see if the problem is consistent.

 

[steve m]

Would any other RV10 drivers be willing to share their AutoPilot settings? I would than compare with mine and see if any of the big differences may be causing my issue.

 

[Solana]

pitch settings
o LS sensitivity 13
o HS sensitivity 15
o Lost Motion Compensation 0
o pitch gain 2
o altitude gain 0.7
o pull rate 1.0
o VSI gain 1.0
o G error gain 0.7
o G error limit 0.45
o GSI gain 1.0
o GSI dot gain 15
o VDI gain 4.0
o VDI dot gain 60

 

[Speeddog8]

I have a Glasair Super II with an HDX, AP panel and auto-trim. My aircraft has a Ray Allen servo controlling a movable tab on the elevator for trim. In general, the AP works great, but I often notice that when I go from fast to slow or slow to fast requiring the Dynon to re-trim the aircraft, for some reason, the Dynon needs to apply additional pitch trim, but elects not to do so. This results in the AP flying the aircraft out of trim and I can feel the servo fighting the air load slightly. There’s plenty of available trim, but the Dynon won’t command a change. I can manually add the trim typically to resolve the issue in most cases, but it’s annoying that it won’t trim it properly. I’ve been wondering if I have a faulty pitch servo that isn’t providing sufficient torque data to the AP panel for trim purposes. The second problem I have is very similar to the original post in this thread. Once the aircraft intercepts the glideslope on an LPV coupled approach, the aircraft immediately starts pitch bucking. Did anyone figure out what’s causing this, or what I should look for?

 

[steve m]

The "pitch bucking" has gotten better in terms of how often it occurs and has gotten less in intensity when it occurs...

I think I "tuned" this out via AP settings.

Seems to happen more in very calm conditions. I am still not 100 percent satisfied or confident with LPV performance.

 

[Solana]

The G- error gain is probably the main source of your pitching on the approach. It also is contributed to by your sensitivity and pitch gain. I had posted my auto pilot settings earlier but I’ve tweaked them since. Try all of these together because many of them interact with each other, and I think your problem will go away.



Trim Settings:
Pitch Roll
90 Kts – 70% 100%
130 Kts – 30% 57%

Autopilot
**roll settings
o LS sensitivity 19
o HS sensitivity 16
o Lost Motion Compensation 0
o roll gain 0
o CDI gain 3.8
o CDI dot gain 1.5
o XTE gain 0.00085
o XTE dot gain 0.0005
o bank angle limit 23
**pitch settings
o LS sensitivity 13
o HS sensitivity 15
o Lost Motion Compensation 0
o pitch gain 1.7
o altitude gain 0.7
o pull rate 1.0
o VSI gain 1.2
o G error gain 0.7
o G error limit 0.45
o GSI gain 0.8
o GSI dot gain 15
o VDI gain 3.8
o VDI dot gain 60