Real ka50 trim

[Jumbik]

Ok here is something which I'm doing, maybe it's not the best practice, but it works wonders for me.

 

Do you guys know the Director control for autopilot? The ctrl+a shortcut?

 

What I do is hit this button, do whatever I want to do being it controlled turn, or combat move, and when i'm in the possition where I want to trimm out and stabilize, I hit the trimm and then enable the director mode again.

 

It was suggested to me by my colleague (he is working at army simulators in Czech rep.) as the proper way how to do maneuvering without the need to hold the trimm button which should not be done apparently.

[kilix]

as stated many times before, there is no good or bad way to do that. Just use the AP/trim however you like most.

I do most of my flying in FD mode. I only disengage FD during autohover, ENR mode or when I'm flying straight, then I use classic trimmed AP channels and FD off.

In this video, I think, that the pilot used FD mode on. There is no point fighting the 20% authority when doing crazy maneuvers.

[159th_Falcon]

How about everyone first starts calling things how there called and by what they mean?

AFCS stands for Automatic Flight Control System, and is made up out of several systems working together.

Auto Pilot

Stabilization Augmentation System

Trim System

Flight Director

and more.

 

Also, AFCS NEVER has authority over pilots inputs.

It can be used as an aid to fly hands off controls though.

However, the pilot ALWAYS has full authority over the flight controls.

 

At least in any helicopter that is designed in a sane way.

 

"So again, with AFCS on the authority over pilot´s inputs in real life are really so small, the amount of effort over the cyclic so little to be able to fly so clean and precisely OVER the AFCS control almost without trimming?"

 

Depends on what part of the AFCS.

SAS, yeah no problem, FD well that only provide visual cues to follow a pre progammed flightplan, Auto-pilot Route follow mode, not quite you would need to use "brute force" to override it if not using the trimmer.

 

To some other posters, you might actually wanna read what the discussion is about.

Its not what is better or how should i trim. It's debating about how the pilot in the video is likely flying his KA-50.

SAS on

FD off

Auto-pilot off

Trim system not sure, some helicopters have buttons/switches to decouple the trim motors.

Do not confuse those whit the SAS buttons in the KA-50 on the RH console though, completely different.

 

________________________________________________________________

See attached picture.

SAS actuators provide stability augmentation and work in parallel to the flight controls, have very limit authority. (5% orso). They provide quick, small precise corrections to the inputs in the flight control system from either the autopilot or the pilot.

They are controlled by the Attitiude Heading Reference System.

(gyro's basicly)

 

Trim motors, are controlled by either the autopilot or the pilot trough the trim system.

They are responsible for holding the controls in place if the pilot is trimming.

The auto pilot uses them to fly the chosen auto-pilot mode, during which you will actually see the flight controls moving in the cockpit.

(does not happen whit SAS actuators)

The trim motors can control the full 100% of flight control deflection, and as a fail safe can be overridden by brute force.

normally there decoupled by a clutch when you press the trim button, and couple again when you release it, holding the stick in the position you had it when you released the trim button.

[Esac_mirmidon]

i think you can understand me if you read my posts.

 

I´m talking about the three SAS channels. Yes my fault don´t call it SAS but AFCS but i´m talking about SAS channels on. So by your last post the pilot is flying WITH SAS three channels on?. That´s what i´m talking about. And if yes this is why i don´t understand why in DCS with SAS channels ON you are FIGHTING the flight system if you move the controls without trimming and it ´s impossible to fly so clean like the video without trimming a lot. Yes i´m able to move the controls and the Ka-50 moves in the direction i´m pushing but i notice a resistance to my movements if i don´t press the trim. Specially with yaw SAS channel on.

 

Maybe the pilot is flying wih pitch-roll SAS on and yaw SAS off. For me this option could be more logical.

Edited January 25, 2016 by Esac_mirmidon

[ShuRugal]

And if yes this is why i don´t understand why in DCS with SAS channels ON you are FIGHTING the flight system if you move the controls without trimming

 

In the DCS KA-50, the SAS buttons also turn on the Autopilot for that channel. The only way to disable Autopilot control is to either engage the Flight-Director mode, or to press and hold the trimmer.

 

This is the reason for always needing to fight the autopilot: When the SAS channels are engages and FD is disengaged, the default mode of the AP is "attitude and heading hold". Each channel will attempt to hold the attitude at which it was engaged, or at while it was oriented the last time the trimmer was released.

[Esac_mirmidon]

Yes, i know that. This was the info i had reading the dcs manual and i was flying the ka-50 that way from the begining.

 

And this is why i dont understand the difference with the real Ka-50 and that video. If in the real Ka-50 its possible to engage the SAS in the three channels without attitude hold AP why we cant do that in dcs?.

 

 

This is why i was assuming the pilot was flying in FD or with yaw SAS off. But if in the real thing its possible to fly with attitude hold off and SAS on this is new for me and its not possible in dcs. Except pressing the trim all the time or turning on the FD mode. And in the video and because others comments isnt the case. It seems.

[AlphaOneSix]

I bank, but the aircraft just does not start to turn - here the 20% AP authority become very aparent to me.

 

In this video, I think, that the pilot used FD mode on. There is no point fighting the 20% authority when doing crazy maneuvers.

 

And if yes this is why i don´t understand why in DCS with SAS channels ON you are FIGHTING the flight system if you move the controls without trimming and it ´s impossible to fly so clean like the video without trimming a lot..

 

Exactly. My entire point is that I don't believe that this happens in the real Ka-50. I just don't. Not only is it logically counter-intuitive, but no other helicopter autopilot (or SAS/SCAS) system on the planet does this.

 

The only way to disable Autopilot control is to either engage the Flight-Director mode, or to press and hold the trimmer.

 

While this is exactly true in-game, I'll reiterate that I don't believe it works this way on the real aircraft. Everyone seems to be assuming that it does because it works this way in-game and I'm saying please allow for the possibility that ED got it wrong on this particular (this very particular, specific) point.

 

When the SAS channels are engages and FD is disengaged, the default mode of the AP is "attitude and heading hold". Each channel will attempt to hold the attitude at which it was engaged, or at while it was oriented the last time the trimmer was released.

 

With the exception of having FD, the Mi-8 autopilot performs identically, yet nobody seems to be fighting the 20% authority on it. With the AP channels on, the Mi-8 autopilot will attempt to hold the attitude at which it was engaged.

[Esac_mirmidon]

For me the way the SAS channels in the Mi-8 works are a lot more logical. You have attitude and heading hold at the point on your last trim if you dont touch the controls after it. But if you press the rudder or the cyclic you are free to move anywhere and the attitude-heading hold is disengage with every control pressure without trimming.

 

Maybe the Ka-50 works the same way, controls pressure-attitute heading hold disengaged, plus extra AP options like route AP, autohover and FD mode. But for me it makes more sense that the real Ka-50 SAS works like the Mi-8. At least that way the video has more sense.

 

But the question remains. If this is true why in dcs is different?

[AlphaOneSix]

Maybe the Ka-50 works the same way, controls pressure-attitute heading hold disengaged, plus extra AP options like route AP, autohover and FD mode. But for me it makes more sense that the real Ka-50 SAS works like the Mi-8. At least that way the video has more sense.

 

Well, in the yaw channel, the heading control is disengaged when the feet are on the microswitches (simulated as movement in the pedals in-game). For the cyclic, the AP doesn't disengage at all, but rather there are "compensation transducers" that give the autopilot computer a signal that effectively negates interference from the autopilot when the cyclic is moved by the pilot. That is, the autopilot *wants* to counteract the pilot's inputs (i.e. fight) but the signal from the compensation transducers cancel it out.

 

But the question remains. If this is true why in dcs is different?

 

Brilliant question. I suspect we won't find out, however, and will have to just live with it and adapt to it like we have.

[Esac_mirmidon]

Thanks AlphaOneSix.

 

I really enjoy the Black Shark a lot in our squad. No matter the difference with the real thing. It´s only a surprise for me to find this divergence taking in count the high fidelity of DCS products.

 

 

If you ask me i love the way Mi-8 works. Is so intuitive and easy to understand, but the black shark.... well it needs time... XDDD

[159th_Falcon]

Understand it all now Esac-Mirmidon, and as AlphaOneSix stated.

Good question on why the KA-50 is so different from, lets say the MI-8.

 

And I'm whit him in thinking that ED got this particular thing wrong in the simulation.

 

While this is exactly true in-game, I'll reiterate that I don't believe it works this way on the real aircraft. Everyone seems to be assuming that it does because it works this way in-game and I'm saying please allow for the possibility that ED got it wrong on this particular (this very particular, specific) point.

 

 

Though, i also hope you understand now why I was so stubborn in saying i didn't see anything special in the video regarding AP and trimming.

 

Its because i compare it to what i know from real helicopters, and not from what i know about the KA-50 ingame.

[Esac_mirmidon]

No problem Falcon.

 

Sometimes its hard to understand ones thoughts only with some words writted in a post without a face behind. Moreover when my english is very rudimentary. If i had started more clear i was comparing DCS vs video and SAS channels instead AFCS all would run more easy.

 

Maybe ED is reading and would consider to talk with Belsimtek and take a look at the black shark AFCS.

Edited January 25, 2016 by Esac_mirmidon

[AlphaOneSix]

I guess I feel rather strongly about this not only from my general helicopter experience, but my experience with the Mi-8 specifically. The Mi-8 and Mi-6 have an autopilot system that went into production initially in 1970 (very similar systems between the two). In the documentation for that autopilot system, it calls out the problem of fighting the autopilot explicitly [my comments will be italicized in brackets]:

 

As soon as the pilot takes up the control [that is, as soon as the pilot manually moves the cyclic control] the helicopter control elements, at the very first moment, become deflected by the amount δ=δpilot [δ is for deflection] and the helicopter starts changing its attitude. Subsequently, signals ω and ɤ [roll angular velocity and roll angle, respectively] appear at the autopilot input. Influenced by these command signals the autopilot shifts the control elements by an amount δ=δautopilot striving to counteract the interference of the pilot [this should sound really familiar right about now]. Hence, the total displacement of the control elements becomes equal to δ=δpilot-δautopilot which means, therefore, that the efficiency of pilot's control is less than when the autopilot is de-energized.

 

Once the rod of the actuator minor cylinder [this rod is controlled by the autopilot] reaches the rest (i.e. δautopilot=δautopilotmax, =20%δ) the autopilot becomes hereafter inoperative because the helicopter control elements no more respond to autopilot commands. Henceforth the pilot's control of the helicopter is as efficient as when the autopilot is de-energized [in other words, the autopilot will input opposite control inputs until it hits its 20% limit of authority, and then it basically quits working. If you have flown the DCS Ka-50 for more than 5 minutes, this will ring a really big bell with you].

 

The necessity to maintain the efficiency of the pilot's control at an approximately constant level calls forth for introducing a compensation transducer.

 

The documentation goes on to describe how the compensation transducer is attached to the flight controls, so when the pilot moves the cyclic, it also moves an arm on the compensation transducer. The autopilot computer takes the signal from the compensation transducer and add it into the mix with the rest of the inputs it's getting from the aircraft instruments, effectively cancelling out what the autopilot would have tried to do if the compensation transducer wasn't there.

 

Here is an example. You move the controls to establish a 5 degrees bank angle to the left to initiate a turn. The autopilot recognizes this 5 degree left bank and tries to counter it (up to the limit of its authority) by moving the minor cylinder in the control actuator to force a 5 degree bank to the right. By adding in the compensation transducer, another input is made to the autopilot, in the same amount as the pilot's input. So the autopilot computer effectively provides no input to the actuator.

 

And finally...

 

As a first approximation, we can state that the role of the compensation transducer is to preserve, when the autopilot is running, the efficiency of controls to which the pilot is used to.

 

And this went into production on the Mi-8 and Mi-6 in January, 1970.

Edited January 25, 2016 by AlphaOneSix

[Migow]

autopilot authority

 

hi alphaonesix : what is the 20 % autopilot authority ?

 

EX: if i have a cyclic than have a Y (pitch)range [+100 ; -100 ]

the cyclic is trim to +20(cyclic range) to get -10(whatever doesn't matter) pitch attitude(not cyclic range) , the autopilot can move is own Y range [+20 ; - 20].

To keep attitude the total sum : pilot( cyclic in rest) + autopilot could move up to [+40; 0]?:)

Edited January 25, 2016 by Fifou265

[Flagrum]

AlphaOneSix, I can see how this works for a single channel, but how does this play out when more channels are involved?

 

My previous example: I give a bit roll input, but the BS keeps flying straight, maintaining the trimmed heading.

 

So, the roll input of the AP would be nullified by my input and the additional compensation transducer input. The helo banks to one side. But instead of the helo now starts to turn, the AP yaw channel gets no input from the pilot and therefore nothing gets "overcompensated", the AP is not cancelled out and maintains the heading.

 

IF i would use the pedals as well, their input could be overriding the AP input by the same mechanism, but if I don't, we have the behaviour I described, the "strange" behaviour we see with the DCS BS.

 

Somewhere, somehow there is still a gap in our theories, I think ...?

[AlphaOneSix]

All turns in the Mi-8 are made with the feet on the pedals as per the pilot's flight manual, resulting in the pedals microswitches being depressed and the yaw channel being "inhibited" for lack of a better word. If you bank without putting your feet on the pedals then yes, the yaw channel will try to hold your heading.

[AlphaOneSix]

hi alphaonesix : what is the 20 % autopilot authority ?

 

EX: if i have a cyclic than have a Y (pitch)range [+100 ; -100 ]

the cyclic is trim to +20(cyclic range) to get -10(whatever doesn't matter) pitch attitude(not cyclic range) , the autopilot can move is own Y range [+20 ; - 20].

To keep attitude the total sum : pilot( cyclic in rest) + autopilot could move up to [+40; 0]?:)

 

My apologies, I do not understand the question.

 

The 20% authority is authority over the hydraulic actuator. On the Mi-8, the swashplate tilt that can be made by the actuator is in the pitch axis is forward 7.5 degrees and rearward 5 degrees, for a total travel of 12.5 degrees. The autopilot has 20% control authority over this movement. Therefore, the autopilot can move the swashplate within a range of 2.5 degrees.

 

In the roll axis, swashplate movement is 4.2 degrees left and 4 degrees right, for a total of 8.2 degrees. Therefore, in the roll axis, the autopilot can move the swashplate within a range of 1.64 degrees.

[Flagrum]

All turns in the Mi-8 are made with the feet on the pedals as per the pilot's flight manual, resulting in the pedals microswitches being depressed and the yaw channel being "inhibited" for lack of a better word. If you bank without putting your feet on the pedals then yes, the yaw channel will try to hold your heading.

Yes, I know. But I meant, how would that work in a Blackshark, if the AP would work like in the Mi-8, but without the switches in the pedals. Assuming that the DCS BS AP implementation is not correct, but also assuming that at least nobody made such a huge mistake to miss the switches in the pedals - i.e. assuming there are realy none in the BS.

[AlphaOneSix]

I don't know, maybe the Ka-50 does have switches in the pedals? More likely, the Kamov bureau implemented an entirely different autopilot. You know, same in some ways, but different in others. It just seems to me that if the Mil bureau solved something like this at least as early as 1970, surely the Kamov guys didn't build helicopters where you had to fight the autopilot's inputs. I would guess that the Ka-50 autopilot is an upgrade/growth/evolution of the Ka-29 autopilot, but alas, I don't have data on it. But I'm sure it's available somewhere, somehow.

[Migow]

Autopilot Replacement

 

I don't know, maybe the Ka-50 does have switches in the pedals? More likely, the Kamov bureau implemented an entirely different autopilot. You know, same in some ways, but different in others. It just seems to me that if the Mil bureau solved something like this at least as early as 1970, surely the Kamov guys didn't build helicopters where you had to fight the autopilot's inputs. I would guess that the Ka-50 autopilot is an upgrade/growth/evolution of the Ka-29 autopilot, but alas, I don't have data on it. But I'm sure it's available somewhere, somehow.

 

i'm working on app to replace dcs faulty autopilot

:doh:

 

the app get input then talk to the game

so summary :

currently in the game

• the autopilot fight pilot hand on which should never happen ,irl there is no problem
  
• the autopilot add massive error on attitude after trim , irl there is no problem
  
• hover with 1 % autority ....
  

 

in my app:

the app simulate the rudder micro switch with axis: push the micro switch(the brake axis) (without the need to move the rudder axis) then the autopilot dont fight pilot like mi8 , it work very well :thumbup:, it has to be like this irl there is no other way for yaw autopilot?

 

for cyclic :when you move the autopilot don't fight the pilot (i tried to simulate the compensator to prevent fighting the autopilot) , the problem is when you move back at the center , the autopilot try to keep attitude in attitude mode so there is still some movement no so smooth unlike the real thing when you pass the center

so i'm questionning in the real thing did they implement a handon , handoff microswitch on the cyclic?:huh:

 

auto turn can't work , how can i get shkval's azimuth in export?

 

hover is buggy WIP:cry:

 

so i ask AlphaOneSix , or airtito ;) how handoff handon cyclic autopilot is working in kamov helicoper, compensator seems not enough, the real thing always move seamless

Edited July 29, 2016 by Fifou265