Jump to content

Un-commanded inputs when trimming Mi-24


Go to solution Solved by AeriaGloria,

Recommended Posts

Posted (edited)

This affects Mi-24 and Black Shark 2 (don't know about BS3 - or other helicopters). The Mi-8 does not exhibit this behaviour. All three helicopters are using 'Central Trimmer Mode'.

When you hit the trim button the helicopter will 'bump' or 'lunge' in the direction of your previous input. It's like an echo of your original stick input the moment when trim was engaged.

Here is a video comparing the Mi-24 trim behaviour to the Mi-8.

In particular compare the T indicators in the top-right. In the Mi-24 when the trim button is pressed the T bar resets to zero then 'rebounds' which seems to be the cause of the 'bumping'. Compare this to the Mi-8 - when you engage trim the pitch bar stays in place and the attitude of the helicopter remains as desired - no bumping/lunging.

Here's a Black Shark 2 video showing similar behaviour. In the video I push the nose down and hold attitude. I engage trim and the nose pitches down further. I pull the nose up to level with the horizon and engage trim - the nose then pitches upwards into the sky. Again, compare this to Mi-8 behaviour.

I've attached the track for the Mi-8 test; I used the same mission for each helicopter.

trim test.trk

Edited by Lixma 06
I boosted the audio around the trim clicks
Posted (edited)
5 hours ago, Lixma 06 said:

This affects Mi-24 and Black Shark 2 (don't know about BS3 - or other helicopters). The Mi-8 does not exhibit this behaviour. All three helicopters are using 'Central Trimmer Mode'.

When you hit the trim button the helicopter will 'bump' or 'lunge' in the direction of your previous input. It's like an echo of your original stick input the moment when trim was engaged.

Here is a video comparing the Mi-24 trim behaviour to the Mi-8.

In particular compare the T indicators in the top-right. In the Mi-24 when the trim button is pressed the T bar resets to zero then 'rebounds' which seems to be the cause of the 'bumping'. Compare this to the Mi-8 - when you engage trim the pitch bar stays in place and the attitude of the helicopter remains as desired - no bumping/lunging.

Here's a Black Shark 2 video showing similar behaviour. In the video I push the nose down and hold attitude. I engage trim and the nose pitches down further. I pull the nose up to level with the horizon and engage trim - the nose then pitches upwards into the sky. Again, compare this to Mi-8 behaviour.

I've attached the track for the Mi-8 test; I used the same mission for each helicopter.

trim test.trk 515.28 kB · 4 downloads

 

I made an autopilot guide that explains this 

The trim in Mi-24/Ka-50 resets the attitude hold. So let’s say it’s holding 0 degrees pitch. You pitch down 5 degrees to fly 250 kmh. Pitch AP uses almost full authority to fight you during that 5 degrees. When you trim, the attitude hold trying to pitch you up to 0 degrees goes away, and now tries to hold you at your current attitude of -5 degrees 

So it’s a matter of your trim re aligning the attitude hold and removing its input 

As designed and not a bug

 

Ways to get around this are told trim button for longer, or trim 2-3x in quick succession. The more you trim, the less this will manifest also. I usually trim 2-3x just for precision anyways 

This is exact same for Black shark when not in FD 

 

Mi-8 doesn’t do this because trim doesn’t re align the attitude hold, it is only telling autopilot what your new center position is so that it can “help” you when you move the stick from that position, which Mi-24 also does at the same time 

 

You can see this clearly by holding trim, which will make AP center as if off. And then will stay centered if you stay at the attitude you release trim at 

Edited by AeriaGloria
  • Like 1
  • Thanks 1

Black Shark Den Squadron Member: We are open to new recruits, click here to check us out or apply to join! https://blacksharkden.com

E3FFFC01-584A-411C-8AFB-B02A23157EB6.jpeg

Posted

Thanks for the info!

22 minutes ago, AeriaGloria said:

I made an autopilot guide that explains this 

Do you have a link?

23 minutes ago, AeriaGloria said:

As designed and not a bug

It seems inferior compared to Mi-8. Is there a reason it's designed this way?

Posted (edited)
4 hours ago, Lixma 06 said:

Thanks for the info!

Do you have a link?

It seems inferior compared to Mi-8. Is there a reason it's designed this way?

You want your AP channels to stay centered as much as possible, to give them biggest authority. Mi-8 is doing almost the same, just not the aircraft. Flight engineer is constantly tunning AP channels. For obvious reasons, Mi-24 must do it differently. AH-64 SCAS is quite similar in its behaviour.

One way to work around this is to use trim hat. It will change aircraft's attitude without affecting autopilot.

Edited by admiki
  • Like 1
  • Thanks 1
  • Solution
Posted (edited)
13 hours ago, Lixma 06 said:

Thanks for the info!

Do you have a link?

It seems inferior compared to Mi-8. Is there a reason it's designed this way?

First page 

 

it may seem harder then Mi-8, but it is superior. As Miki said, with AP only having 18% authority in Mi-24, you want it close to center as possible so it can dampen and stabilize attitude. If I trim at nose level, and go to -5 degrees pitch for cruise, my pitch AP will be stuck 100% trying to pitch me up, unable to have any authority left to dampen oscillations, rapid movements, or stabilize attitude. At this point it is merely offsetting your command, and the trimming ( or in Mi-8 case, re adjusting the AP dials to center AP deflection) will set the AP back to inertial so it can both hold your current attitude and have room to stabilize and dampen you 

It’s designed this way so that you can, easier them Mi-8, not only keep AP centered and have its most authority but also make it easier to fly. This way, if I fly at 150 kmh, I just trim nose and wings level and it will hold it. 
 

If I want the autopilot to hold me at cruise, or return to level when I release the controls on cruise, I just trim when flying level at 250-270 kmh, which is usually -5-7 degrees pitch down, and then AP will hold me there and return to level when I release stick pressure in any other attitude 

There are other reasons, as I mentioned, both Mi-8/24 are taking into account where the stick is when you trim so that the AP knows what you want “neutral” to be. The reason for this is that AP will deflect in same direction you move the stick to give you extra authority in pitch and roll. This can’t work if the AP is stuck at full deflection trying to return you to an attitude far away from where you are currently.
 

When trimmed and AP is near center, you will see especially roll channel deflect a lot with lateral stick movement and increase roll rate, but then return you to trimmed attitude when you neutralize stick. this way AP doesn’t fight you, but helps you as long as you have trimmed reasonably close to the flighty speed/attitude you are at.

Another example is, not only does pitch attitude need to decrease as you increase speed, but stick needs to be deflected forward to maintain this speed. Same with roll, you usually need slight right stick to hover (10-25%,), but often left stick to fly at cruise or above. If you don’t trim, but only turn on AP on the ground, the roll AP will be expecting your neutral stick position to be slightly right, and since it sees the stick deflected left, (which you need to maintain cruise speed or above), it will try to make you roll left. So trimming will re center it and solve al these problems for you 

 

Mi-8 is exact same in this respect, it’s just that since the attitude hold isn’t being reset, it is making it smoother for you to “fight” the autopilot stuck at 100% deflection during trimming 

So for example, you have three flight regimes you want to be primarily trim for. Hover, where you are usually 3.5 degree pitch nose up. 140-150 kmh, which is your best climb and loiter speed, where you would trim about nose level from there to 200 kmh. And above that you need nose down to maintain speed, to the point at 250-270 kmh cruise you are -5-7 degrees nose down, so you want to trim there as well. 
 

If you pitch attitude changes a lot from change in speed, say 3-5 degrees, or AP is close to maximum deflection, you want to trim 

It can be useful for other things, for example if I want to fly straight and takes my hands/eyes off, a short trim will make sure the AP attitude hold keeps my current attitude. If I want to stay in one place while I take my hands off the controls or eye off the screen, I can trim while turning, and the AP will keep me in that turn until I grab the controls again and trim level 

If you watch cockpit videos of Mi-8 or Ka-50, you will see that they are pressing trim near constantly. I find myself trimming alot when I’m level because I like to make my autopilot assist me as much as possible. No that you are aware of it, you might be able to get a good feel for it, for how it feels when AP is at 100% deflection limit and you are less stable, And how nice it is to trim to cruise and have it hold it, or trim in hover and suddenly have it help you stay there. 
 

TLDR: Imagine that trim button turns off AP when pushed and back on when released. Play with trimming level for a certain speed, and you might find the benefits 

Edited by AeriaGloria
  • Like 1
  • Thanks 1

Black Shark Den Squadron Member: We are open to new recruits, click here to check us out or apply to join! https://blacksharkden.com

E3FFFC01-584A-411C-8AFB-B02A23157EB6.jpeg

Posted
3 minutes ago, Lixma 06 said:

Great explanation, thanks AG.

Thank you. Also forgot to say, this is in my guide but short version is 

Pitch Roll AP is mixing 3 things. Attitude hold, dampening movement, and copying your stick movements to exaggerate them. The farther any of these are from 0, the more they override the other inputs. So close to the attitude you trim/turn on AP, your primarily have the dampening and stick copying to help you, with attitude hold just keeping you close. 
 

But as attitude changes, within a limit of usually around 5-10 degrees, it will override all other inputs and only try to force you back to original attitude. You are basically loosing the other ways it can help while having to deflect your stick 18% to fight it. You will notice then at this 5-10 degree region or above, when pitch AP is at max deflection, it’s here that your uncommanded input (many previously have called it “nodding”) is highest. 
 

The numbers for roll channel are different, it’s much less aggressive in attitude hold and much more aggressively copies your stick to increase roll authority, but we are lucky here with the transmission tilt, ED Has modeled that it doesn’t need any roll or drift to fly straight at any speed 
 

So yeah, trim maybe 2-3x, hold the trim button for longer and let it stabilize, or pressing trim then maneuvering and only releasing once you are where you want to be are different ways people use it. Sorry if I’ve made things complicated! 

  • Like 1
  • Thanks 1

Black Shark Den Squadron Member: We are open to new recruits, click here to check us out or apply to join! https://blacksharkden.com

E3FFFC01-584A-411C-8AFB-B02A23157EB6.jpeg

Posted

So the thread just took a dramatic twist (!)

Anyway I made this video last night but didn't bother posting since I thought it was obsolete but after @shwed posted it might be relevant.

In this video I trim Mi-24 nose level. I then nose down a couple of degrees hit trim and release pressure on stick. I then keep pressing only the trim button and the nose pitches down further and further with each click.

I do the same after pulling up and trimming for level. I release the stick and press only trim button. The nose rises with each successive click.

If this is normal behaviour as @AeriaGloria has indicated then I've no problem. I'll just fly it differently.

Posted
4 hours ago, shwed said:

Developers said this helicopter behavior is a bug.

 

I read through and a bit unclear what is the bug. The behavior of the trim resetting the pitch/roll AP is described in many manuals, is there some other issue going on with the trim? I have noticed some issues when I switch to front seat and back as if there are ghost inputs 

4 hours ago, Lixma 06 said:

So the thread just took a dramatic twist (!)

Anyway I made this video last night but didn't bother posting since I thought it was obsolete but after @shwed posted it might be relevant.

In this video I trim Mi-24 nose level. I then nose down a couple of degrees hit trim and release pressure on stick. I then keep pressing only the trim button and the nose pitches down further and further with each click.

I do the same after pulling up and trimming for level. I release the stick and press only trim button. The nose rises with each successive click.

If this is normal behaviour as @AeriaGloria has indicated then I've no problem. I'll just fly it differently.

I watched your video a few times. I don’t know for certain what is happening, but looking at your inputs I might have a guess 

You give for example forward stick to pitch down. The attitude hold fights the pitch forward and your pitch rate slows down. You then press trim without moving the stick, which removes the autopilot trying to pitch you back up. Since you had enough stick to fight the autopilot trying to pitch you up before, and the autopilot is no longer trying to pitch you up; it causes further nose down. E.I, you are giving enough stick that autopilot cannot fight it (it only has 9% authority of your stick in each direction), and when you press trim it increases the rate and then pitch AP fights it slowing pitch rate down then you trim making the cycle continue 

What you often have to do, and my experience is different since I use FFB, is when you trim after nosing down, you have to move stick slightly back to make up for the 1-9% that the autopilot was fighting you before trimming. For pitching up reverse, you often have to give forward stick after trimming to make up for the Input of the AP that is now removed by trimming 

This is the big reason I usually recommend trimming 2-3x while holding your desired attitude to really make the system work for you 

If you hold trim longer, the process is often easier as you can then compensate for the autopilot input being removed before it re engages. 

  • Like 1

Black Shark Den Squadron Member: We are open to new recruits, click here to check us out or apply to join! https://blacksharkden.com

E3FFFC01-584A-411C-8AFB-B02A23157EB6.jpeg

Posted (edited)
16 minutes ago, AeriaGloria said:

Since you had enough stick to fight the autopilot trying to pitch you up before, and the autopilot is no longer trying to pitch you up; it causes further nose down. E.I, you are giving enough stick that autopilot cannot fight it (it only has 9% authority of your stick in each direction), and when you press trim it increases the rate and then pitch AP fights it slowing pitch rate down then you trim making the cycle continue 

After the initial pitch down and trim I release all pressure from the stick. After that my only input is pressing the trim button repeatedly which causes the nose to 'ratchet' downwards/upwards.

Edited by Lixma 06
Posted
1 hour ago, Lixma 06 said:

After the initial pitch down and trim I release all pressure from the stick. After that my only input is pressing the trim button repeatedly which causes the nose to 'ratchet' downwards/upwards.

 

And you have return to center trim with spring stick? It just sounds like to me you were giving more input then the autopilot could stabilize when you trimmed the first time, and since trimming kept the stick forward, it was too much for the AP. It only has 9% authority in one direction. So 1/10th stick movement from what you need to fly stable is enough to override it and stop it from being able to hold an attitude 

Black Shark Den Squadron Member: We are open to new recruits, click here to check us out or apply to join! https://blacksharkden.com

E3FFFC01-584A-411C-8AFB-B02A23157EB6.jpeg

Posted
38 minutes ago, AeriaGloria said:

And you have return to center trim with spring stick?

Yes.

In the video I zoom out a bit to keep the cyclic in view to show there was no further directional input from me after the initial pitch down.

Posted (edited)
2 hours ago, Lixma 06 said:

Yes.

In the video I zoom out a bit to keep the cyclic in view to show there was no further directional input from me after the initial pitch down.

Yes that’s what I mean, your initial pitch down was too much for the autopilot to stop. After you trim; the stick is stuck too far forward unless you were to move stick back 

And then the pitch rate changes as your toggling AP on/off with trim 

Edited by AeriaGloria

Black Shark Den Squadron Member: We are open to new recruits, click here to check us out or apply to join! https://blacksharkden.com

E3FFFC01-584A-411C-8AFB-B02A23157EB6.jpeg

Posted (edited)

For the record, I have spoken to real life Mi-24 pilot. This is exactly how trim works in Mi-24.

Lixma, issues you are having are easy to solve.

As AG said, AP is fighting you when you push nose down from trimmed position. Once you press trim release that AP input is gone and your nose starts to drop. You have only nose down input now. If you take your time and observe what AP is doing, nose will stabilise at some lower position after you let go of trim release because AP is fighting it again to regain new trimmed position. If you press trim release, nose goes down again. And it goes on and on. Why is that?

This is a problem of a flight simulator (not just DCS, but any simulator) in combination center spring joystick. Your joystick is the cause here. Let me explain.

There is a misconception that with center trim option in DCS, stick will stop working as soon as you press trim release, until you return it to center. That's wrong. It will work until you let go of trim release.

At the begining of the flight, your stick is in the center position, same as your in-game stick. Let's say that for forward flight with no AP, you would need to push your stick 5 mm forward. But because of the AP fighting you, you need to push it 8 mm. So, it's 3 mm too much. You press trim release, let go of trim release and return your physical stick back because you have to, AP resets. But in-game stick is still at 8 mm forward position. That's still 3 mm too much and what is causing nose down moment again and again. Try holding trim release instead until you are stabilised in your new position, release, return stick to center and try pressing trim again. If you AP is neutral, you will get no movement.

Listen how much pilot is pressing trim release in Mi-8. They are doing the same in Mi-24.

As for the link shwed posted, I would not take that Ka-50 video as a reference, since we don't know what mode it's SCAS was in. It works differently in certain modes.

 

Edited by admiki
  • Like 2
Posted

Bug in the logic of AP hydraulic mechanism. There are many real Mi-24 pilots in Russia 🙂 such behavior does not exist in a real helicopter, it is not correct. ED found a bug and promise to change it.

Posted (edited)
7 hours ago, admiki said:

There is a misconception that with center trim option in DCS, stick will stop working as soon as you press trim release, until you return it to center. That's wrong. It will work until you let go of trim release.

@admiki

There's no misconception - that's exactly how 'Central' mode is supposed to work; and exactly how it works with the Mi-8.

image.jpeg

That's the whole point of Central mode - to allow player to return stick to neutral; to prevent un-commanded/extra input that we see in Mi-24 and Ka-50.

So while I trust the Mi-24 autopilot is correctly modelled, the Central trim mode in the Mi-24 and Ka-50 is apparently broken.

The Mi-8's Central mode works fine.

I've thrown up a post in the Bugs section - hopefully they'll take a look at it.

Edited by Lixma 06
Posted
23 minutes ago, Lixma 06 said:

@admiki

There's no misconception - that's exactly how 'Central' mode is supposed to work; and exactly how it works with the Mi-8.

image.jpeg

That's the whole point of Central mode - to allow player to return stick to neutral; to prevent un-commanded/extra input that we see in Mi-24 and Ka-50.

So while I trust the Mi-24 autopilot is correctly modelled, the Central trim mode in the Mi-24 and Ka-50 is apparently broken.

The Mi-8's Central mode works fine.

I've thrown up a post in the Bugs section - hopefully they'll take a look at it.

 

I think this part is just written bad. It should say :" After presing and releasing trimmer button............After that, any further deflection of the joystick will be added........."

Just try it, I did and that's how I know it works as I said. 

You are confusing center trim option in DCS and Hind AP. You are not the first one, AG and me had this same discussion with a bunch of people over time.

@BIGNEWYmaybe you guys should do a bit of a cleanup of manuals?

Posted

Thanks @admiki and @AeriaGloria for your patience!

Still....

I remain skeptical that the real life Mi-24 behaves like ours.

It seems insane that any trim system that throws the aircraft further into a turn, or a climb, or a dive would make it past the drawing board.

Posted

It becomes a muscle memory to hold trim when doing anything.

Problem a lot of people have is that spring joystick is a lousy control for a helicopter.

Real stick is somewhat hard to move with trim engaged (it should be called brake) and easy with trim release. So you learn quite quickly to use it.

  • Like 1
Posted
43 minutes ago, admiki said:

I think this part is just written bad. It should say :" After presing and releasing trimmer button............After that, any further deflection of the joystick will be added........."

Even this doesn't cover it. Watch this video - After the first pitch down and trim I release the joystick; there is no further deflection; it remains centred. All I do is repeatedly press trim and the nose ratchets its way down. I then level off, trim, release joystick, and keep pressing trim button to 'walk' the nose up the ladder. 

 

Posted (edited)

We already covered this in another thread.

Watch what your autopilot is doing and what it is doing when you press trim release.

And read this again:

At the begining of the flight, your stick is in the center position, same as your in-game stick. For sake of this discussion, let's say in hover your nose is 5 degs up from the horizont. Let's say that for forward flight with no AP (nose is now level with the horizont), you would need to push your stick 5 mm forward. But because of the AP fighting you, you need to push it 8 mm. So, it's 3 mm too much. You press trim release, let go of trim release and return your physical stick back because you have to, AP resets. But in-game stick is still at 8 mm forward position. That's still 3 mm too much(because AP is not fighting it) and what is causing nose down(let's say it is now 3 degs below horizont, AP is now fighting it to bring nose back to level, that AP input is what stops nose from dropping further) moment again and again. Try holding trim release instead until you are stabilised in your new position, release, return stick to center and try pressing trim again. If you AP is neutral, you will get no movement.

Edited by admiki
  • Like 1
Posted

Speaking of trim release....

Here I am cruising and trimmed for level flight. The AP doesn't appear overly taxed.

image.jpeg

Question: What happens when I hit 'Trim Reset'?

Posted

Your nose will still drop a small amount since T channel is still slightly below neutral. It means AP is inputing nose up to keep your trimmed attitude.

Like I said, hold trim release until you are stable at attitude you want to keep. Once stable, let go of the trim and press it again and tell me what happens?

  • Like 1
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...