Autopilot trim wheel speed?

[seippg]

Hi all,

Is there any way, via either the air file or config file, to change the autopilot trim speed? In particular, I'm wanting to adjust the elevator trim speed.

Gregg

 

[Heretic]

Elevator_trim_effectiveness in the aircraft.cfg, if anything. But this will also affect manual trimming.

 

[seippg]

Elevator_trim_effectiveness in the aircraft.cfg, if anything. But this will also affect manual trimming.

Ah, okay. That should work. One thing I was noticing about the airplane I was working on the ILS on, that on autopilot, the airplane was lagging behind the Flight Director much more than I saw in real world videos of the real airplane. I had the thought that, if the plane would just follow the FD, it would be pretty solid. Also, if I could get it to respond more actively to the FD's directions, by increasing trim speed, that would also be an improvement. Does that make sense?

 

[WarpD]

These values in the aircraft.cfg [autopilot] section controls how quickly the aircraft responds on autopilot:

max_pitch
max_pitch_acceleration
max_pitch_velocity_lo_alt
max_pitch_velocity_hi_alt
max_pitch_velocity_lo_alt_breakpoint
max_pitch_velocity_hi_alt_breakpoint

 

[Roy Holmes]

Gregg,
Could you explain a bit more about what you mean by "trim speed".
Do you mean the rate at which the elevator trim setting changes? If so, this is a function of its angular range (aircraft.cfg) and trim rate Cm_dt (airfile). Each click of trim is 1/1024 of the range and Cm_dt is pitching moment per radian of trim movement. There are 57.3 degrees in a radian, the airfile uses radians, the aircraft.cfg uses degrees.

If you increase the value of trim range, (elevator_trim_limit in [airplane_geometry] section of aircraft.cfg) you get more degrees of trim angle per trim increment/click.
Similarly, if you increase the absolute value of Cm_dt you get more pitching moment per increment, that does not increase the rate of change of trim indication, but it does increase the effect.

The autopilot values WarpD covered control the maximum rate of response to the autopilot, so they are a limit to the .cfg and air file factors when using the autopilot.

A further secondary effect is the airplane basic stability. If it has strong pitch stability it will be slower to respond to pitch trim changes. Similarly if it has very high longitudinal moment of inertia (MOI), also an aircraft.cfg setting, it will be slower to change its pitch attitude which means it will overshoot the glideslope intersection and be less responsive to pitch trim changes. That would also give the appearance of slow trim.

Heretic's mentioned Elevator_trim_effectiveness in the aircraft.cfg. That setting changes the effectiveness of the Cm_dt setting in the air file. Same thing basically. I prefer to make all changes in the air file and leave the aircraft.cfg scalars at 1.0.

To sum up, if your wish is to have the airplane fly better on autopilot, you can alter the autopilot settings or the trim settings, but if it has sluggish flight response in general you need to really fix that first. That way you are not asking the autopilot to do things outside its control.

Roy

 

[seippg]

These values in the aircraft.cfg [autopilot] section controls how quickly the aircraft responds on autopilot:

max_pitch
max_pitch_acceleration
max_pitch_velocity_lo_alt
max_pitch_velocity_hi_alt
max_pitch_velocity_lo_alt_breakpoint
max_pitch_velocity_hi_alt_breakpoint

Thanks. Looking at these in the .cfg file I see...

max_pitch_acceleration = 1
max_pitch_velocity_lo_alt = 2
max_pitch_velocity_hi_alt = 1.5
max_pitch_velocity_lo_alt_breakpoint = 20000
max_pitch_velocity_hi_alt_breakpoint = 28000

I'll try playing with them.

Gregg,
Could you explain a bit more about what you mean by "trim speed".
Do you mean the rate at which the elevator trim setting changes? If so, this is a function of its angular range (aircraft.cfg) and trim rate Cm_dt (airfile). Each click of trim is 1/1024 of the range and Cm_dt is pitching moment per radian of trim movement. There are 57.3 degrees in a radian, the airfile uses radians, the aircraft.cfg uses degrees.

If you increase the value of trim range, (elevator_trim_limit in [airplane_geometry] section of aircraft.cfg) you get more degrees of trim angle per trim increment/click.
Similarly, if you increase the absolute value of Cm_dt you get more pitching moment per increment, that does not increase the rate of change of trim indication, but it does increase the effect.

The autopilot values WarpD covered control the maximum rate of response to the autopilot, so they are a limit to the .cfg and air file factors when using the autopilot.

A further secondary effect is the airplane basic stability. If it has strong pitch stability it will be slower to respond to pitch trim changes. Similarly if it has very high longitudinal moment of inertia (MOI), also an aircraft.cfg setting, it will be slower to change its pitch attitude which means it will overshoot the glideslope intersection and be less responsive to pitch trim changes. That would also give the appearance of slow trim.

Heretic's mentioned Elevator_trim_effectiveness in the aircraft.cfg. That setting changes the effectiveness of the Cm_dt setting in the air file. Same thing basically. I prefer to make all changes in the air file and leave the aircraft.cfg scalars at 1.0.

To sum up, if your wish is to have the airplane fly better on autopilot, you can alter the autopilot settings or the trim settings, but if it has sluggish flight response in general you need to really fix that first. That way you are not asking the autopilot to do things outside its control.

Hi Roy, I don't *think* the plane is sluggish except that the autopilot is slow to catch up to the Flight Director. In fact, the first thing the autopilot does right before intercepting the GS is a brief ascend up to the GS and then it has to chase it down which can take 10-15 seconds. If this "Autopilot guy" was a real guy in the cockpit with me, I'd think "come on already..." I'll try some things when I get back to the sim and report back.

 

[seippg]

These values in the aircraft.cfg [autopilot] section controls how quickly the aircraft responds on autopilot:

max_pitch
max_pitch_acceleration
max_pitch_velocity_lo_alt
max_pitch_velocity_hi_alt
max_pitch_velocity_lo_alt_breakpoint
max_pitch_velocity_hi_alt_breakpoint

Hmmm...I tried changing these values but don't see any effect. So far, haven't found anything that will appreciably change the sluggishness of the airplane following the FD.

 

[WarpD]

max_pitch would only set a maximum commandable pitch for the autopilot... say 20 degrees. max_pitch_acceleration and max_pitch_velocity_lo/hi_alt will control the actual speed at which the autopilot commands a pitch change.

If as Roy states it is strictly by how quickly the trim can change (which will be the same whether it's the autopilot or the pilot)... then as he stated above:

Do you mean the rate at which the elevator trim setting changes? If so, this is a function of its angular range (aircraft.cfg) and trim rate Cm_dt (airfile). Each click of trim is 1/1024 of the range and Cm_dt is pitching moment per radian of trim movement. There are 57.3 degrees in a radian, the airfile uses radians, the aircraft.cfg uses degrees.

At this point I don't think we can offer much more without knowing what all the values are, include the MOI entries.

 

[seippg]

Here's the MOIs

empty_weight_pitch_MOI = 3800
empty_weight_roll_MOI = 2700
empty_weight_yaw_MOI = 5000
eempty_weight_coupled_MOI = 0
empty_weight_coupled_MOI=0.000000

Here's the autopilot section:

[autopilot]
autopilot_available = 1
use_no_default_bank = 1
use_no_default_pitch = 1
flight_director_available = 1
default_vertical_speed = 0
autothrottle_available = 0
autothrottle_arming_required = 0
autothrottle_takeoff_ga = 1
autothrottle_max_rpm = 90
pitch_takeoff_ga = 10
default_pitch_mode =0
default_bank_mode =0
max_pitch = 12.65
max_pitch_acceleration = 16.0 // 1
max_pitch_velocity_lo_alt = 24.0 // 2
max_pitch_velocity_hi_alt = 1.5
max_pitch_velocity_lo_alt_breakpoint = 20000
max_pitch_velocity_hi_alt_breakpoint = 28000
max_bank= 25
max_bank_acceleration = 1.8
max_bank_velocity = 5.25
max_throttle_rate = 0.1
nav_proportional_control= 15.0
nav_integrator_control= 0.45
nav_derivative_control= 0.90
nav_integrator_boundary= 0.40
nav_derivative_boundary= 0.20
gs_proportional_control= 25.0
gs_integrator_control= 2.40
gs_derivative_control= 1.60
gs_integrator_boundary= 1.70
gs_derivative_boundary= 0.80

 

[Roy Holmes]

It seems the basis of your unhappiness with the autopilot is that it does not follow the FD commands as well as you have seen it do so in real world video. Try selecting the flight director availability in the autopilot section to "0" and see if it does any better in capturing the glide slope. I'm not sure the flight director does much in the sim these days, certainly if you program needles to follow its direction in a display it is totally unconvincing.
Maybe I'm wrong and if so I'm quite happy to be corrected but I have not seen a usable flight director display since FS9.
You have not told us anything about the airplane type, might be significant.
Roy

 

[seippg]

max_pitch would only set a maximum commandable pitch for the autopilot... say 20 degrees. max_pitch_acceleration and max_pitch_velocity_lo/hi_alt will control the actual speed at which the autopilot commands a pitch change.

If as Roy states it is strictly by how quickly the trim can change (which will be the same whether it's the autopilot or the pilot)... then as he stated above:

Ok, I am adjusting CM_dt and getting some results there. The airplane had a stock setting of -0.01 and I started doubling it all the way up to -16.0(!). It still lags the FD and overshoots the glideslope but not as bad.

It seems the basis of your unhappiness with the autopilot is that it does not follow the FD commands as well as you have seen it do so in real world video. Try selecting the flight director availability in the autopilot section to "0" and see if it does any better in capturing the glide slope. I'm not sure the flight director does much in the sim these days, certainly if you program needles to follow its direction in a display it is totally unconvincing.
Maybe I'm wrong and if so I'm quite happy to be corrected but I have not seen a usable flight director display since FS9.
You have not told us anything about the airplane type, might be significant.
Roy

Hi Roy,

I can try turning off the FD. To answer your question, it's the Flight1 Mustang (an airplane I've spent a lot of time tuning). Yes, watching some videos, the airplane seems very responsive, hugging the FD in the RW when on autopilot. I haven't gotten close to that though, now, at least flies a much better ILS than it did with a bit of overshooting. Visually, I'd say the FD looks like the airplane would fly a better ILS if I could get it to follow it better. Since it lags behind, it ends up having to overcorrect to get on the approach.

Here's one video where you can see how the airplane responds...seems very responsive. He stays on autopilot most of the way down the approach. Thanks again for your help!

[MEDIA]

[/MEDIA]

 

[WarpD]

OMG... ok... THAT addon...

First, I think your FD might be moving too quickly. I've seen that a lot in addons. Flight directors do not yank and bank... passengers really don't like how that feels, so there's always a much slower rate of change than one might think.

Having done an AATD for a C510... lesse what I can find...

Pitch Moment - Elevator (Control) CM_de in REC 1101 was set to -1.171875
Pitch Moment - Horizontal Stabilizer CM_h in REC 1101 was set to -2.294921975
Pitch Moment - Trim CM_dt in REC 1101 was set to -2.

max_pitch_acceleration = 3
max_pitch_velocity_lo_alt = 3
max_pitch_velocity_hi_alt = 1.65

max_bank_acceleration = 1.8
max_bank_velocity = 5.25

 

[seippg]

I applied WarpD's suggestions and I think it's about as close to the real world airplane as I'll be able to get....I settled on 1.6 for CM_dt for now to keep the trim just a bit less sensitive for now. In any case, I'm calling it done!

All this testing and trial and error has highlighted a problem with FLC on the Mustang. I'll open a separate thread on it since it appears to be unrelated to the topic at hand.

Thanks again for your help!