Possible to remove 777's automatic stabilizer/elevator trim?

[OE-LML]

Hallo!

First of all I fairly like the FG-777 and I've already spent some time flying it.

I do a lot of manual flying, below TA/TL I hardly use autothrust, FD or autopilot.
But the automatically acting stabilizer/elevator trim makes that damn difficult and drives me crazy sometimes.

It's moving even when on ground and sometimes changes to crazy values. Every time short before take off you have to recheck if take off trim is still set.
During climb or descend it's moving and you have to overrule it by trimming and elevator inputs to keep your desired vertical speed.
Levelling off takes lots of time and trimming inputs. Finally once airspeed is steady, vertical speed is 0 and the plane is trimmed, auto trim starts again and you have to overrule it permanently every few seconds.

I don't know much about programming and I'm quite happy if my Hello World html page is functioning.
In FSX it was not too difficult to change some parameters in an aircraft's file to make it fly more realistic.
But looking into FG-777's files just confuses me.
Is it possible and could someone tell me how to remove this "feature"?

As always looking forward to your replies.
Thx in advance.

[Octal450]

Yeah, its rubbish imo, there is no FlyByWire simulation, so the autotrim is useless and unnatural. I reworked the Autopilot, but didn't care to touch the autotrim code, so I just left it in at the bottom.

Just open up the 777-autopilot.xml, and find the autotrim section, (ctrl+f will find it)
Find this:

Code: Select all

         <Ki>
            <condition>
               <and>
                  <not><property>instrumentation/afds/inputs/AP</property></not>
                  <greater-than>
                     <property>/orientation/roll-deg</property>
                     <value>-5.0</value>
                  </greater-than>
                  <less-than>
                     <property>/orientation/roll-deg</property>
                     <value>5.0</value>
                  </less-than>
               </and>
            </condition>
            <value>-0.02</value>
         </Ki>


Remove that entirely, and replace it with this:

Code: Select all

         <Ki>0</Ki>


That should stop it from autotrimming unless the AP is turned on.

Kind Regards,
Josh

[Isaak]

Thanks for pointing this out, itp. I found this thread on how it should behave. I'll remove the current autotrim from the 777 in fgaddon and dig into some more docs to see if I can make it work as expected.

[Octal450]

Hi Bel Isaak,
I made proper A/P autotrim then, and brought the AP up to date with my ITAF System Standard, it's quite more stable in my testing. I've sent you a mail with the details.

Kind Regards,
Josh

PS: A basic FBW system can be made with YASim, but don't bother unless you plan to simulate at least the C*U logic, or it will just react weird. Let me know if you want more details.
PSS: I wouldn't know if a FBW simulation is even worth it, you can't get proper simulation of trims since YASim is really not very realistic at all, especially in the pitch axis, but maybe I could make a "simplified" FBW that still uses basic C*U logic and autotrim, but isn't perhaps as perfect as it could be.

[Soitanen]

There was jsbsim branch on gitlab repo, i have started using real data, but didn't finish it due to lack of time. But pitch axis in that version was good enough, you can use it like a stub.

[Octal450]

I didn't bother with it since much code has to be rewrite do for it to work properly.

I don't have the time to commit to doing a JSB maintainer, and since currently no one is working on that version, it's left in a in-dev state. The good news is property rules can easily translate to jsb. I moved the A320 FBW to JSB in 2 hours, and that is with refactoring it. If someone were to pick up it's work, it wouldn't take me long to convert.

Kind Regards
Josh

[OE-LML]

Hallo!

@ it0uchpods:
Setting that code to "0" does some improvement. Yoke inputs (push/pull) don't change trim setting anymore and it doesn't change to weird values on the ground. But during manual flight (climb/descend/levelled off) stabilizer trim slowly moves to 0.0 units (I'm talking 'bout the stab trim gauge on the display just above the thrust levers, if you change it to FCTL). (Btw: Shouldn't it show the same values as the gauge beside the parking brake?)

Nevertheless thanks for your help and maybe you have another solution/work around!!??

@ all 777-developers:
As I said before, I'm glad if I can get a Hello-World- html-page to work and I don't wanna tell anybody to do anything but:
Is it really necessary to simulate the FBW-system of a plane that is behaving mostly like a conventional plane?
And I don't think there is any great difference in a simulation between a selected trim speed and a directly moving stabilizer.
Wouldn't it be easier to replace the auto trim totally with a conventional trim system?
Just my personal opinion ...

As always thx in advance.
And I'm looking forward to your help and replies.

[legoboyvdlp]

If we're going to simulate aircraft we might as well do it properly

[Octal450]

Hi OE-LML
I made a new trim system that fixes it entirely. It can be a seen soon in FGAddon.

No, sorry, that is not a good. It aims to fly like a conventional aircraft, but it has many improvements that make it "easier" to fly. That would be more realistic to simulate. Further YASim is weird and doesn't act like how an airliner should. The FBW will help to hide these flaws.

Josh

[Isaak]

Hi all, for those eager to have the improvement: I 'm currently on a business trip without access to my dev PC with FG etc. (work policy logically forbids me from installing FlightGear on my work laptop ). Normally I will commit it0uchpods' improvements this weekend, after I 've reviewed and tested them (I consider itp a far more experienced developer than me, but it's always good to have someone else test the system). I 'm sorry that I have to let you wait, but normally you will have the improved FBW systems this Saturday.

Thanks for you patience

[OE-LML]

Hallo!

@it0uchpods:
I definitely couldn't wait so I took the recent 777 from your hangar/depository to do some test flights today (flight time about 3 hours).
Wow ... that's some great work!!!
I really like the behaviour of stab trim and the moving spoilers.
Do the spoilers now affect roll rate or is it just a visual effect?

But there is some issue now with rudder trim and rudder sensitivity/effectiveness. This also seems to affect heading and track mode of AP (I didn't set up a flight plan, so I don't know about LNAV).
But I guess you are aware of that problem.

If you wanna know more about my impressions and all other things I've discovered, just tell me and I'm gonna tell you.

[Octal450]

Hi OE-LML:
Great. Yes, spoilers are factored with ailerons in a custom calculation, then sent to YASim. So yes, they will affect roll

I do not have any rudder issues here, it is a bit too sensitive, I toned it down for my next push.

Kind Regards,
Josh

[OE-LML]

Hallo!

@it0uchpods:
I'm now talking about the version from yesterday.

The rudder issues: Beside the sensitivity of rudder and rudder trim (which is in my opinion far to high), rudder trim sometimes acts automatically.

- When starting in cold and dark cockpit, rudder trim is not centered.
- Mostly it changes values when engaging or disengaging AP. And sometimes during manual flight.
- Sometimes the centered position is not 0.0 units on the display-gauge of FCTL.

About the next 2 issues I'm not sure, they could also have occurred on the FGAddon-version:

- When AP and autothrottle is both on and then you disengage AP (but autothrottle still on), sometimes autothrottle accelerates engines far above selected speed and causes the plane to overspeed. Mostly that happens in FLCH mode, but also happend when levelled off.
- Sometimes AP skipps waypoints in LNAV mode. - That's great fun when you are on RNAV transitions with short legs between tight turns (e.g. LOWW).

There are still some minor things which are also on FGAddon-version, but I don't wanna bother you with that.
Keep doing your great job!

As always: Thx in advance for the replies from all of you.

PS: I got one question about the ailerons: On your version and on FGAddon-version they only move when flaps are down at least at pos. 1.
In 777-CBT-videos on a "very popular video platform" it is said that ailerons only move when airspeed is low. Is there in real life a speed threshold where ailerons move or is it right that they only move when flaps down???

[Octal450]

Hi,
I have NO rudder issues, in fact, NONE of my code touches the rudder trim! So I have NO idea what you are talking about, this fault must come from your side.

The AP works perfectly. The A/THR is not always speed by adjust thrust, in FLCH mode, even with the AP off, the speed mode doesn't change. If the FD and AP are all off, it might revert, but I didn't program the logic of the AP, so I have no idea. Basically, if you want SPD mode, just select another pitch mode and it should be fine.
AP and LNAV is because the 777 uses some weird LNAV system. If you want I can switch it to my LNAV system.

Keep in mind the FGAddon version has NO FCS right now. My git is the ONLY copy of the 777 with FCS simulated. The others only have animations. I will check the outer aileron.

Kind Regards,
Josh

[legoboyvdlp]

No -- it does not come from his side: I am seeing the very sensitive rudder trim as well.

I saw it on the -300.

By any chance are you only only testing the -200?