Naughty autopilot on ecstasy..?

[nicky9499]

Hello guys,

I am facing a strange issue where the autopilot is constantly doing this head-bobbing-up-and-down thing like it's on drugs or something. The aircraft is a merge that one of my pilots made (Posky 77E model and PSS panel). It flies perfectly under manual control. Here is a short clip which best describes the issue. Note also the strange compass movements at top right.



Anyone knows what might be going on?

[propjob]

Try increasing the pitch stability value in the [flight_tuning] section of aircraft.cfg.

Art

[JSkorna]

Could be wind turbulence.

[mrzippy]

I remembered this pitch problem from last month with a CitationX and this was his final fix. Finally she works!!!!!

Had to edit the pitch_stability to 2.6 and the elevator trim effectiveness to 0.9 - flying well and no nasty handling characteristics - Finally success

[propjob]

I also thought at first that it might be turbulence, Jim, but it damped out. I've had that pitch problem occur with some autopilots.

Art

[jgf]

You can also try increasing the pitot static parameter, this increases the response time so the AP isn't "hunting" for the altitude. Default value is 2 but I've seen some large aircraft with this at 95.

[nicky9499]

You can also try increasing the pitot static parameter, this increases the response time so the AP isn't "hunting" for the altitude. Default value is 2 but I've seen some large aircraft with this at 95.

The one in this cfg is already at 999, is it okay to push this value past a thousand?
It's probably not turbulence as this occurs throughout the flight, even at straight-and-level cruise.

I have made the following changes so far, taxiing out as we speak will see how things go.

[flight_tuning]
pitch_stability=2.0 (from 1.0)

[autopilot]
max_pitch_acceleration=0.5 (from 1.0)

[jgf]

The SDK doesn't mention a limit for the pitot static setting, but I've never seen it set higher than 99. If tweaking this and pitch stability does little there may be basic design flaws in the aircraft (though I'd be surprised with a Posky ...perhaps the airfile is corrupt). Another thing you might try is tweaking the max_pitch_acceleration parameter in the autopilot section of the cfg; using a different autopilot might also help (don't know why but this has worked for me in the past).

[nicky9499]

Hi guys,

As this is a follow-up to the original issue I shan't start a new thread. The situation now is, well, still about the same. So far I observe that the oscillations only occur when the AP is changing the vertical speed (VNAV, for example) and does not occur when I have a constant VS set or established on the glideslope, you get the idea. Even in straight and level flight the oscillations often occur because of the wind/weather etc. Take for example, I am descending at -500fpm; the aircraft is steady. If I change the VS to -1000fpm, the oscillations go on for a couple seconds until the AP attains -1000fpm, then settles down again.

The following is the old CFG, double forward slashed (commented out) lines represent the original values.

[autopilot]
autopilot_available=1
flight_director_available=1
default_vertical_speed=800.0
autothrottle_available=1
autothrottle_arming_required= 1
autothrottle_max_rpm = 108.5
autothrottle_takeoff_ga= 1
pitch_takeoff_ga=10
max_pitch=10.0
//max_pitch_acceleration=1.0
max_pitch_acceleration=0.1
//max_pitch_velocity_lo_alt=0.7
max_pitch_velocity_lo_alt=0.1
//max_pitch_velocity_hi_alt=0.4
max_pitch_velocity_hi_alt=0.1
//max_pitch_velocity_lo_alt_breakpoint=11000.0
max_pitch_velocity_lo_alt_breakpoint=40000.0
//max_pitch_velocity_hi_alt_breakpoint=25000.0
max_pitch_velocity_hi_alt_breakpoint=43000.0
max_bank=30.0
//max_bank_acceleration=1.2
max_bank_acceleration=0.2
//max_bank_velocity=1.5
max_bank_velocity=0.5
max_throttle_rate=0.08
nav_proportional_control=11.00
nav_integrator_control=0.17
nav_derivative_control=0.00
nav_integrator_boundary=2.50
nav_derivative_boundary=0.00
gs_proportional_control=22.0
gs_integrator_control=0.530
gs_derivative_control=0.02
gs_integrator_boundary=0.70
gs_derivative_boundary=0.00
yaw_damper_gain = 1.0

I'm at quite a loss what to do now, since I've tried extreme min and max values and everything in between.
This morning I copied over the AIR file and its [autopilot] section from a similar Posky+Phoenix 772 merge (which flies excellent) but the issue still persists. I am uploading the entire 773 plane now, hopefully somebody in their spare time can take a look as to what may the problem?

Cheerio,
Nicholas

[jgf]

The proportional, derivative, and integrator parameters control how the autopilot reacts (the other parameters control how the aircraft reacts) but the SDK does not go into detail about the function and adjustment of these, and I've never had occasion to tinker with them. But i suspect you have some arcane interaction between the AP response and the aircraft response.