What's new
Van's Air Force

Don't miss anything! Register now for full access to the definitive RV support community.

Autopilot elevator control 'porpoising' quite badly.

JackinMichigan

Well Known Member
I have an entirely Garmin control panel in my RV-10 and one of the features is an autopilot (roll and elevator, no rudder). The roll control works great, but when I engage altitude hold the plane will begin 'porpoising', and it will get exponentially worse till, after three up-down cycles, I have to disengage the autopilot. I've reduced the sensitivity, reaction speeds and reaction forces of the elevator servo but nothing seems to help.

One thing I've noticed is when I'm adjusting elevator trim, the trim indicator lags way behind the actual trim motion. Not sure why that is, but the entire electric trim assembly is buried deep in the tail with no good way to get to it. I suspect this has a lot to do with it.
 
If the tweaks don't work, you might have a problem with the static source - there can be a reservoir effect with too many gadgets hooked up and then the altitude hold does a lot of hunting and can't keep up. If that turns out to be the case, you'll want to consider a dedicated static line for the autopilot.
 
I have Dynon, but have experienced similar symptoms. Sounds to me like your servo shear screw has broken. There can still be some friction on the plate, especially if the screw broke but left a little chunk of itself behind that is scraping on the plate. It works a little bit but loses grip and keeps trying to correct itself.
 
I have an entirely Garmin control panel in my RV-10 and one of the features is an autopilot (roll and elevator, no rudder). The roll control works great, but when I engage altitude hold the plane will begin 'porpoising', and it will get exponentially worse till, after three up-down cycles, I have to disengage the autopilot. I've reduced the sensitivity, reaction speeds and reaction forces of the elevator servo but nothing seems to help.

One thing I've noticed is when I'm adjusting elevator trim, the trim indicator lags way behind the actual trim motion. Not sure why that is, but the entire electric trim assembly is buried deep in the tail with no good way to get to it. I suspect this has a lot to do with it.

Disengage auto trim and try it, sounds like maybe the auto trim is running backwards.
 
What Walt said,

or it could also be the pitch servo direction needs to be reversed. I had that problem when testing my autopilot. I had the auto trim off as recommended. Reversed the pitch servo direction and was good to go.


Edit: After thinking about it, it was after activating the auto trim that I had the issue. Reversing the auto trim direction corrected the problem.
 
Last edited:
It’s not the shear screw, Garmin servos use a clutch.

Go through the calibration routine described in the manual.
Check that all the surfaces are deflecting in the correct directions.
 
I agree with trying it without the auto trim to eliminate any issues on that system. On my 10 with G3X, I set up the AP with the recommended settings and found that I needed a good amount of change to the gain on the ptich axis to get it to be stable. Can't remember if it was porpoising or just twitchy. Either way, I ended up with a pitch gain that was quite a ways off from the recommended setting. Mine is rock steady.

Garmin has a very involved regiment for dialing in the AP in an appendix of their manual. Suggest reviewing that.

Larry
 
Adjusting

Do you have a link? I've got a slight porpoising problem (but only at high altitudes, 10K or so) that I need to troubleshoot. I also overshoot turns a bit.

Garmin has a very involved regiment for dialing in the AP in an appendix of their manual. Suggest reviewing that.

Larry
 
I don't have a G3X! :)

It might still work. They have you adjust gain unti getting a certain amount of oscillations before stabilizing and then back off a certain amount. That approach should work with most PID logic that I believe all AP's use, though I suspect some variability in the approach and the back off amount is not likely to be portable to other systems. That is likely unique to the Garmin code base.

Larry
 
I think I'm on the right track

I've got the Garmin autopilot running off of a G5 and a GMC 307, so presumably it's the same procedure. I just don't have the manual for the G3X.

Many thanks!

It might still work. They have you adjust gain unti getting a certain amount of oscillations before stabilizing and then back off a certain amount. That approach should work with most PID logic that I believe all AP's use, though I suspect some variability in the approach and the back off amount is not likely to be portable to other systems. That is likely unique to the Garmin code base.

Larry
 
AP Porpoise

I've got the Garmin autopilot running off of a G5 and a GMC 307, so presumably it's the same procedure. I just don't have the manual for the G3X.

Many thanks!

The G3X Installation Manual can be downloaded here, Appendix A contains the autopilot tuning outline.

Does your autopilot porpoise occur in ALT mode only?

You will want to first verify that PIT mode is working as it should, per the outline in Appendix A. If it is, you can move on to optimizing ALT hold mode. ALT mode is controlled by the Vertical Speed Gain. Instructions for tuning the Vertical Speed Gain are also included in Appendix A, but it is very important that you ensure PIT mode is working correctly before getting to it. The Vertical Speed gain relies on the standard pitch gain (what you change to optimize PIT mode) being correctly tuned for the airframe.

Please feel free to email us at [email protected] if you would like.

Thanks,

Justin
 
porpoising

Good question! I'll download the instructions and give it a whirl, in that order. Many thanks for the info.

The G3X Installation Manual can be downloaded here, Appendix A contains the autopilot tuning outline.

Does your autopilot porpoise occur in ALT mode only?

You will want to first verify that PIT mode is working as it should, per the outline in Appendix A. If it is, you can move on to optimizing ALT hold mode. ALT mode is controlled by the Vertical Speed Gain. Instructions for tuning the Vertical Speed Gain are also included in Appendix A, but it is very important that you ensure PIT mode is working correctly before getting to it. The Vertical Speed gain relies on the standard pitch gain (what you change to optimize PIT mode) being correctly tuned for the airframe.

Please feel free to email us at [email protected] if you would like.

Thanks,

Justin
 
porpoise after rain

If the tweaks don't work, you might have a problem with the static source - there can be a reservoir effect with too many gadgets hooked up and then the altitude hold does a lot of hunting and can't keep up. If that turns out to be the case, you'll want to consider a dedicated static line for the autopilot.

I realize this thread is relating to the Garmin, but bear with me here. My RV7A with dynon skyview ap (ver 16 firmware) and ray allen trim. The pitch, pitch trim and everything works great, EXCEPT. If its raining when I take off, it will develop a pitch porpoise. I have safe air static ports, one on each side, normal location with the tubes from each angling up so water should not be able to collect. I am using the Dynon non heated pitot tube. Dan, you mentioned the reservoir effect. If it did not work normally, I would accept that. The static line does go all the way forward to my D6 standby (which is rock solid)

Is there any possibility that the holes in the static port could be enlarged a bit?

Its almost like a drop of water gets in the static port and the autopilot senses that, however when I get back on the ground, the ports are clear and I can suck air from one side to the other.

I have tried slipping the aircraft trying to put a cross flow to clear it with no help.

I have also tweaked the pitch settings to no avail. remember it works great when the take of is dry. some of the older cessnas had a bulb near the static port. seems like that would be a problem as well, but of course most of those airplanes did not have an auto pilot.
 
I realize this thread is relating to the Garmin, but bear with me here. My RV7A with dynon skyview ap (ver 16 firmware) and ray allen trim. The pitch, pitch trim and everything works great, EXCEPT. If its raining when I take off, it will develop a pitch porpoise. I have safe air static ports, one on each side, normal location with the tubes from each angling up so water should not be able to collect. I am using the Dynon non heated pitot tube. Dan, you mentioned the reservoir effect. If it did not work normally, I would accept that. The static line does go all the way forward to my D6 standby (which is rock solid)

Is there any possibility that the holes in the static port could be enlarged a bit?

Its almost like a drop of water gets in the static port and the autopilot senses that, however when I get back on the ground, the ports are clear and I can suck air from one side to the other.

I have tried slipping the aircraft trying to put a cross flow to clear it with no help.

I have also tweaked the pitch settings to no avail. remember it works great when the take of is dry. some of the older cessnas had a bulb near the static port. seems like that would be a problem as well, but of course most of those airplanes did not have an auto pilot.

If it only happens when raining before T/O, I would assume that water is messing with something and the static ports are the most logical. I would look closer at the port setup to be sure you are not collecting water there. Might consider a tee, with a small stub going downward to let water move away.

Angling up is not the recommended approach. A 90* fitting is recommended. If the angle is too slight, the water can possibly collect.
 
Last edited:
Back
Top