Autopilot controls the rudder, and it shouldn't.
-
Given the Hawk has no AP IRL..... I think the way they manage to enable a 'psuedo AP' is irrelevant.
Maybe we should be grateful we have the option of an AP of sorts.
It's not a PMDG airliner. Does it really matter ?
I also believe there is an issue across the board with rudder/steering linkage and visual external rudder deflection, anyway.
@gabe777 said in Autopilot controls the rudder, and it shouldn't.:
It's not a PMDG airliner. Does it really matter ?
The autopilot has a noticeable effect on the rudder, which it should not have. For those of us that buy Just Flight aircraft for the realism (which is a selling point they promote) yes of course it matters.
-
@gabe777 said in Autopilot controls the rudder, and it shouldn't.:
It's not a PMDG airliner. Does it really matter ?
The autopilot has a noticeable effect on the rudder, which it should not have. For those of us that buy Just Flight aircraft for the realism (which is a selling point they promote) yes of course it matters.
So how does an autopilot perform a coordinated turn ?
Think about it.
For a start I believe all turns are auto-coordinated in MSFS anyway. There is no need for a rudder in normal flight. Rudder trim is useful though.
So bunging in a fake AP may be a bodge job... but not made easier by fundamental flaws in the sim.
-
So how does an autopilot perform a coordinated turn ?
Think about it.
For a start I believe all turns are auto-coordinated in MSFS anyway. There is no need for a rudder in normal flight. Rudder trim is useful though.
So bunging in a fake AP may be a bodge job... but not made easier by fundamental flaws in the sim.
@gabe777 said in Autopilot controls the rudder, and it shouldn't.:
So how does an autopilot perform a coordinated turn ?
Interesting and valid question, so I Googled "autopilot coordinated turn" and got this: "Planes can have three different types of autopilot software: one-axis, two-axis, and three-axis. The next-generation aircraft can be guided by improved three-axis autopilots. New generation autopilots can also direct the yaw by controlling the rudder along with rotation and reclining movements."
and this: "it just depends on whether the GA autopilot is 2 axis (eg. KAP140) or 3 axis (eg KFC225 w/yaw damper). A 3 axis unit will co-ordinate turns - a 2 axis unit will not, although this is rarely an issue in reasonably calm winds and at cruise speeds."
I'm pretty sure the Piper Arrow autopilot is not a next generation or new generation autopilot, in which case it doesn't perform coordinated turns.
-
@gabe777 said in Autopilot controls the rudder, and it shouldn't.:
So how does an autopilot perform a coordinated turn ?
Interesting and valid question, so I Googled "autopilot coordinated turn" and got this: "Planes can have three different types of autopilot software: one-axis, two-axis, and three-axis. The next-generation aircraft can be guided by improved three-axis autopilots. New generation autopilots can also direct the yaw by controlling the rudder along with rotation and reclining movements."
and this: "it just depends on whether the GA autopilot is 2 axis (eg. KAP140) or 3 axis (eg KFC225 w/yaw damper). A 3 axis unit will co-ordinate turns - a 2 axis unit will not, although this is rarely an issue in reasonably calm winds and at cruise speeds."
I'm pretty sure the Piper Arrow autopilot is not a next generation or new generation autopilot, in which case it doesn't perform coordinated turns.
@sender46 I agree that it's far from perfick.
But with the AP system in MSFS presently, I think that pseudo APs shoe-horned in by JF.... (like the entire thing on the Hawk.....obs ! ... and the altitude hold on the Arrows), are just an example of making compromises to make it work.
Far from accurate and not ideal, but better than nothing.
-
@sender46 I agree that it's far from perfick.
But with the AP system in MSFS presently, I think that pseudo APs shoe-horned in by JF.... (like the entire thing on the Hawk.....obs ! ... and the altitude hold on the Arrows), are just an example of making compromises to make it work.
Far from accurate and not ideal, but better than nothing.
-
@gabe777 Other sim quirks and limitations have been worked around. Rather than just settle for "better than nothing", let's let others more knowledgeable than you and me assess what can or cannot be done.
@sender46 I definitely agree this should be fixed. Have you created a ticket? If not I am happy to do some testing and create one
On the point of "how does an autopilot make a coordinated turn" without a yaw damper - I think it is just because the turns made by the autopilot are gentle enough that it isn't noticeable.
In the warriors and archers I fly IRL (haven't worked my way up to arrows just yet) if your turns are gentle enough you hardly need any rudder. The planes I fly (and I assume arrows are probably the same) have have frise ailerons which reduce adverse yaw
-
@sender46 I definitely agree this should be fixed. Have you created a ticket? If not I am happy to do some testing and create one
On the point of "how does an autopilot make a coordinated turn" without a yaw damper - I think it is just because the turns made by the autopilot are gentle enough that it isn't noticeable.
In the warriors and archers I fly IRL (haven't worked my way up to arrows just yet) if your turns are gentle enough you hardly need any rudder. The planes I fly (and I assume arrows are probably the same) have have frise ailerons which reduce adverse yaw
Well the latest update has changed all that Rudder confusion.
Haven't tested it yet, but the setting is now Auto Rudder as it was in FSX and P3D !
Under Accessibilty it now clearly says " Auto coordination of turns " ....AND "Reducing crosswind effect on takeoff."
No mention of landings though.
A lot of people with rudder pedals will be doing cartwheels.
I prefer seperate assistance TBH, splitting takeoff and in-air coordination... but massive improvement if it works.
No-one appears to have noticed yet.
I only noticed because the update reset ALL Accessibilty options to default.
-
@sender46 I definitely agree this should be fixed. Have you created a ticket? If not I am happy to do some testing and create one
On the point of "how does an autopilot make a coordinated turn" without a yaw damper - I think it is just because the turns made by the autopilot are gentle enough that it isn't noticeable.
In the warriors and archers I fly IRL (haven't worked my way up to arrows just yet) if your turns are gentle enough you hardly need any rudder. The planes I fly (and I assume arrows are probably the same) have have frise ailerons which reduce adverse yaw
@nickd27 said in Autopilot controls the rudder, and it shouldn't.:
@sender46 I definitely agree this should be fixed. Have you created a ticket? If not I am happy to do some testing and create one
I haven't because I assumed Stagefer would have when WalterBeech prompted him, but he may not have. If you are happy to do some testing and create one that's probably worth doing.
-
Given the Hawk has no AP IRL..... I think the way they manage to enable a 'psuedo AP' is irrelevant.
Maybe we should be grateful we have the option of an AP of sorts.
It's not a PMDG airliner. Does it really matter ?
I also believe there is an issue across the board with rudder/steering linkage and visual external rudder deflection, anyway.
@gabe777 said in Autopilot controls the rudder, and it shouldn't.:
Given the Hawk has no AP IRL..... I think the way they manage to enable a 'psuedo AP' is irrelevant.
We're not talking about the Hawk here, we're talking about the Turbo Arrow, that has a real single axis autopilot, not a pseudo one. And if you want realism in a flight sim this is absolutely relevant to me.
-
@sender46 I definitely agree this should be fixed. Have you created a ticket? If not I am happy to do some testing and create one
On the point of "how does an autopilot make a coordinated turn" without a yaw damper - I think it is just because the turns made by the autopilot are gentle enough that it isn't noticeable.
In the warriors and archers I fly IRL (haven't worked my way up to arrows just yet) if your turns are gentle enough you hardly need any rudder. The planes I fly (and I assume arrows are probably the same) have have frise ailerons which reduce adverse yaw
@nickd27 said in Autopilot controls the rudder, and it shouldn't.:
I definitely agree this should be fixed. Have you created a ticket? If not I am happy to do some testing and create one
Didn't create a ticket. Please do it if you will. Thanks.
-
Well the latest update has changed all that Rudder confusion.
Haven't tested it yet, but the setting is now Auto Rudder as it was in FSX and P3D !
Under Accessibilty it now clearly says " Auto coordination of turns " ....AND "Reducing crosswind effect on takeoff."
No mention of landings though.
A lot of people with rudder pedals will be doing cartwheels.
I prefer seperate assistance TBH, splitting takeoff and in-air coordination... but massive improvement if it works.
No-one appears to have noticed yet.
I only noticed because the update reset ALL Accessibilty options to default.
@gabe777 said in Autopilot controls the rudder, and it shouldn't.:
Under Accessibilty it now clearly says " Auto coordination of turns " ....AND "Reducing crosswind effect on takeoff."
I do not see this anywhere in accessibility settings, can you clarify where taht is?
-
So I was trying to reproduce this problem but for me the rudder didn't move by itself. I wonder if you have a bug in the sim itself where thee auto-rudder is enabled even though it shows as disabled? Maybe you should try enabling it, restarting the sim, and disabling it again.
During my testing I did find some unusual behaviour with the yoke when using the autopilot in roll mode (not heading mode)
https://youtu.be/XdIMYYLtihE -
@gabe777 said in Autopilot controls the rudder, and it shouldn't.:
Under Accessibilty it now clearly says " Auto coordination of turns " ....AND "Reducing crosswind effect on takeoff."
I do not see this anywhere in accessibility settings, can you clarify where taht is?
-
Just took this topic for a test drive, and while I agree that given the AP in this plane should be a single axis AP (minus the ALT hold convenience JF added) at no point did I see the rudder or the rudder pedals move under AP control. Be it in the 15kt crosswind or in heading mode or using the roll knob...or even using the rudder trim.
-
So I was trying to reproduce this problem but for me the rudder didn't move by itself. I wonder if you have a bug in the sim itself where thee auto-rudder is enabled even though it shows as disabled? Maybe you should try enabling it, restarting the sim, and disabling it again.
During my testing I did find some unusual behaviour with the yoke when using the autopilot in roll mode (not heading mode)
https://youtu.be/XdIMYYLtihE@nickd27 said in Autopilot controls the rudder, and it shouldn't.:
During my testing I did find some unusual behaviour with the yoke when using the autopilot in roll mode (not heading mode)
https://youtu.be/XdIMYYLtihEStrange, I'm not seeing that behaviour. The roll mode appears to work correctly for me left and right.
EDIT: That was in the Arrow III. Will check in the Turbo.
EDIT 2:
Just checked again and now I see what you're seeing (in both the Arrow III and the Turbo Arrow IV. I'm guessing the Warrior will be the same but I don't have that.)
- Autopilot ON and in ROLL MODE, straight and level.
- ROLL SWITCH to left (or right) and plane rolls and turns as expected.
- While still turning, change to HDG MODE, leaving ROLL SWITCH to the left (or right).
- Plane steers to heading bug and straightens up.
- Change to ROLL MODE (ROLL SWITCH still turned left or right) but it does not revert to ROLL MODE.
- Turn ROLL SWITCH back to centre and only then does ROLL MODE engage.
Also:
- ROLL SWITCH still turned left (or right).
- In HDG MODE and plane turning to the heading bug.
- Before reaching the heading bug, switch to ROLL MODE.
- The plane stays in HDG MODE, continues turning and straightens up at the heading bug (i.e. even though HDG MODE is switched off, it stays in HDG MODE).
- Turn the ROLL SWITCH to centre and only then does it come out of HDG MODE and engage ROLL MODE.
And there's more. After the step below (copied from above):
- The plane stays in HDG MODE, continues turning and straightens up at the heading bug (i.e. even though HDG MODE is switched off, it stays in HDG MODE).
- Then, without doing anything else, try to use the yoke to control the plane and boy, do you get a wild ride!! The plane becomes almost uncontrollable but this can be recovered by switching the autopilot off.
Before I raise a support ticket, can you confirm whether any of that is how it works in real life or not?
-
@nickd27 said in Autopilot controls the rudder, and it shouldn't.:
During my testing I did find some unusual behaviour with the yoke when using the autopilot in roll mode (not heading mode)
https://youtu.be/XdIMYYLtihEStrange, I'm not seeing that behaviour. The roll mode appears to work correctly for me left and right.
EDIT: That was in the Arrow III. Will check in the Turbo.
EDIT 2:
Just checked again and now I see what you're seeing (in both the Arrow III and the Turbo Arrow IV. I'm guessing the Warrior will be the same but I don't have that.)
- Autopilot ON and in ROLL MODE, straight and level.
- ROLL SWITCH to left (or right) and plane rolls and turns as expected.
- While still turning, change to HDG MODE, leaving ROLL SWITCH to the left (or right).
- Plane steers to heading bug and straightens up.
- Change to ROLL MODE (ROLL SWITCH still turned left or right) but it does not revert to ROLL MODE.
- Turn ROLL SWITCH back to centre and only then does ROLL MODE engage.
Also:
- ROLL SWITCH still turned left (or right).
- In HDG MODE and plane turning to the heading bug.
- Before reaching the heading bug, switch to ROLL MODE.
- The plane stays in HDG MODE, continues turning and straightens up at the heading bug (i.e. even though HDG MODE is switched off, it stays in HDG MODE).
- Turn the ROLL SWITCH to centre and only then does it come out of HDG MODE and engage ROLL MODE.
And there's more. After the step below (copied from above):
- The plane stays in HDG MODE, continues turning and straightens up at the heading bug (i.e. even though HDG MODE is switched off, it stays in HDG MODE).
- Then, without doing anything else, try to use the yoke to control the plane and boy, do you get a wild ride!! The plane becomes almost uncontrollable but this can be recovered by switching the autopilot off.
Before I raise a support ticket, can you confirm whether any of that is how it works in real life or not?
@sender46 I have not flown with the Autocontrol III autopilot in real life so I'm not sure exactly how it works. I do know that typically you would have the plane flying straight and level before activating the autopilot and typically you would already be flying the heading you have set it to (if in HDG mode) so it kind of makes sense that the roll mode doesn't activate until you set it to straight and level - but I'm not sure if that's how it would operate in real life.
Definitely the graphical glitch where the yoke seems to match the roll selector on the autopilot is a bug though
-
@sender46 I have not flown with the Autocontrol III autopilot in real life so I'm not sure exactly how it works. I do know that typically you would have the plane flying straight and level before activating the autopilot and typically you would already be flying the heading you have set it to (if in HDG mode) so it kind of makes sense that the roll mode doesn't activate until you set it to straight and level - but I'm not sure if that's how it would operate in real life.
Definitely the graphical glitch where the yoke seems to match the roll selector on the autopilot is a bug though
@nickd27 said in Autopilot controls the rudder, and it shouldn't.:
Definitely the graphical glitch where the yoke seems to match the roll selector on the autopilot is a bug though
The yoke and the ailerons are mechanically tied together with cables, so the two should always move together... But, you are correct that the yoke seems to be tied to the ROLL knob position instead of following the ailerons!
-
@sender46 I have not flown with the Autocontrol III autopilot in real life so I'm not sure exactly how it works. I do know that typically you would have the plane flying straight and level before activating the autopilot and typically you would already be flying the heading you have set it to (if in HDG mode) so it kind of makes sense that the roll mode doesn't activate until you set it to straight and level - but I'm not sure if that's how it would operate in real life.
Definitely the graphical glitch where the yoke seems to match the roll selector on the autopilot is a bug though
@Nickd27 and @RetiredMan93231 I thought it strange that it stays in heading mode even when the heading mode is switched off (until the roll knob is centred) but it appears we don't know if that is wrong or not.
So the only thing we're sure is wrong is the yoke tied to the roll knob instead of the ailerons? I'll let you raise a support ticket, Nick, as you picked it up.
-
@Nickd27 and @RetiredMan93231 I thought it strange that it stays in heading mode even when the heading mode is switched off (until the roll knob is centred) but it appears we don't know if that is wrong or not.
So the only thing we're sure is wrong is the yoke tied to the roll knob instead of the ailerons? I'll let you raise a support ticket, Nick, as you picked it up.
@sender46 said in Autopilot controls the rudder, and it shouldn't.:
So the only thing we're sure is wrong in is the yoke tied to the roll knob instead of the ailerons? I'll let you raise a support ticket, Nick, as you picked it up.
This logic was used as the PID used for roll control causes lots of small fluctuations in the position of the ailerons that is not very noticeable on the exterior model but looks odd and distracting when shown by the yoke position, which 'flickers' and doesn't necessarily show a logical position for the selected bank. I'll have to see if there is a way to add some 'smoothing' code for the yoke but either way hopefully this is a minor sacrifice in realism.
Roll mode only engaging when the roll knob is centred, the aircraft is banked less than 10 degrees and the roll rate is low is a mixture of real-world AP logic and attempts to tame the unpredictable MSFS AP system. Allowing roll mode to engage outside of those constraints causes even less predictable behaviour. We'll continue to tweak/improve it wherever possible.
The AP remaining in heading hold with the HDG button off has now been fixed.
Martyn - Development Manager
-
@sender46 said in Autopilot controls the rudder, and it shouldn't.:
So the only thing we're sure is wrong in is the yoke tied to the roll knob instead of the ailerons? I'll let you raise a support ticket, Nick, as you picked it up.
This logic was used as the PID used for roll control causes lots of small fluctuations in the position of the ailerons that is not very noticeable on the exterior model but looks odd and distracting when shown by the yoke position, which 'flickers' and doesn't necessarily show a logical position for the selected bank. I'll have to see if there is a way to add some 'smoothing' code for the yoke but either way hopefully this is a minor sacrifice in realism.
Roll mode only engaging when the roll knob is centred, the aircraft is banked less than 10 degrees and the roll rate is low is a mixture of real-world AP logic and attempts to tame the unpredictable MSFS AP system. Allowing roll mode to engage outside of those constraints causes even less predictable behaviour. We'll continue to tweak/improve it wherever possible.
The AP remaining in heading hold with the HDG button off has now been fixed.
