How I Think the Controls Should Work

[ARM505]

I can't see anyone as having grasped the point of the OP - he's stating that in the REAL Ka50, the AP does not fight the pilot because it senses cyclic input (even without trimmer button depressed), and doesn't attempt to correct bank/pitch during input, whereas the simulated Ka50 appears to fight back against cyclic inputs - he's not asking how everybody is using the trimmer. Maybe I've read this too fast?

Edited February 9, 2009 by ARM505

[connos]

No troubles flying the BS - holding trimmer while making changes, then release. No rubber-banding or wobbling for me either. This is after alot of practice though - as it should be.

 

I use a different approach but it seems it is different for everyone. I maneuver with all AP Channels on and then Trim(as i posted above). This is the recommended way by the way:pilotfly:.

[ruprecht]

I can't see anyone as having grasped the point of the OP - he's stating that in the REAL Ka50, the AP does not fight the pilot because it senses cyclic input (even without trimmer button depressed),

 

I would dispute that. The purpose of the autopilot is to hold an attitude/course/altitude depending on the channels selected. While the real KA-50 pilot may not feel much resistance per se due to the hydraulics, if they make a control input without retrimming, the aircraft would return to the set attitude/course/altitude when that input is released.

 

That is, as far as I can tell, the way the sim works. I believe it is currently as authentic as it can be given the limitations of the hardware.

[AlphaOneSix]

There are two things gong on with the controls. One is the force trim. In and of itself, force trim normally is not associated with the autopilot, as its purpose is to hold the cyclic and pedals in place through spring tension. When you press the trimmer button, you release the spring tension, making the controls easier to move. Releasing the trimmer button re-engages the spring tension, holding the cyclic and pedals in the new position. In the Ka-50, there is a second effect of the trim button: While the trimmer button is pressed, the "hold" functions of the autopilot get turned off temporarily, and releasing the trimmer button sets a new heading/attitude to hold for the autopilot.

 

I am not positive on this part, but it works this way in other Russian helicopter autopilots, so I have a sneaking suspicion that it works as follows on the Ka-50 as well... There are compensation transducers attached to the flight controls that tell the autopilot if the controls are moving. So let's say you are in straight and level flight and for whatever reason, the aircraft starts to roll to the left (let's say it was a gust of wind or something, just not pilot input)..the autopilot will sense this roll through a roll rate gyro (for rate of roll) and your ADI (for amount of roll) and will send an input the to the lateral booster to eliminate the roll. Nothing is felt in the controls, and the aircraft stays in straight and level flight and the pilot is none the wiser (except that he thinks to himself how easy it is to fly the Ka-50!).

 

Now let's take the same scenario, but the pilot inputs left cyclic to start a roll to the left. The autopilot still receives a signal from the roll rate gyro and the ADI telling it that the aircraft is rolling, but at the same time, the autopilot receives an opposite signal from the compensation transducer on the flight controls that essentially cancels out the signal form the gyro (or at least, it cancels out an amount equal to the input from the controls, it's still working to smooth out non-control related movements). Therefore, the autopilot won't offer any input and the aircraft will roll. Note that as soon as the controls stop moving, the autopilot will go back to holding the aircraft's attitude (the new one, not the old one) even without pressing the trimmer button (of course, you still need to hit the trimmer button, otherwise your controls will move back to where they were before you moved them, basically undoing your maneuver).

 

AGain, I will point out that I cannot guarantee that this is how it works on a Ka-50, so you may choose to ignore everything I've said. ;)

[DTWD]

so you may choose to ignore everything I've said.

 

Done! :D

 

No that all sounds like what it feels like it's doing. Well at least it seems like a good explanation of what it might be.

[ericinexile]

I can't see anyone as having grasped the point of the OP - he's stating that in the REAL Ka50, the AP does not fight the pilot because it senses cyclic input (even without trimmer button depressed), and doesn't attempt to correct bank/pitch during input, whereas the simulated Ka50 appears to fight back against cyclic inputs - he's not asking how everybody is using the trimmer. Maybe I've read this too fast?

 

Exactly!

 

Everyone posting seems covinced that the AP unlatches at pilot input exactly as AlphaOneSix describes. I'm starting to think I downloaded DCS v0.90.

 

SH

[ZaltysZ]

How this AP fight against pilot shows itself? I notice some kind of "rebel" only when AP heading hold is on, but this is normal as it is supposed to hold the heading. Can you describe the fighting by describing input and output in away like: if I apply deflection, the AP does this...?

[ericinexile]

Great question. Let's first agree on levels of automation:

 

1) All blue buttons off = no attitude/heading hold and no dampening.

2) FD ON = all axis dampening but still no attitude/heading hold.

3) FD OFF = dampening plus attitude/heading hold.

 

To keep things simple, let's forget "1" plus let's forget the option to press and hold trim to remove AP holds and neutral springs.

 

That leaves us with AP or FD. According to AlphaOneSix, any pilot cyclic input without FD will remove AP hold input. In other words, when the pilot makes an input the Ka50 behaves exactly as if the FD had been turned on. The AP relatches only once control pressures are romoved. Now with that knowledge, fly straight and level at a stable speed and turn. Note the amount of cyclic and ruddur input required to hold your desired bank angle and turn rate. (for the sake of comparison dont hold trim). Now, under identical conditions, do this again with FD ON and again note required amount of control input. Should be identical, right? Well the are not. That's my only point. I'm not looking for flying advice. I know the FD works and I know holding the trim works. I'm just convinced that the DCS Ka50 can more closely mirror the Real Ka50 without all the creative workaround some of us have discovered.

[ZaltysZ]

The thing with banking comparison is difficult, because of the difference how banking is done with AP on and AP off. In fact, I don't see how to make valid comparison. Lets assume that our input is deflection angle of cyclic (to left or right), then we get these outputs:

 

1. AP channels on, FD off: angle of bank;
   
2. AP channels on, FD on: rate at which banking angle changes;
   

I maybe simplified too much, but that is how these things look to me. Because of different output type, the amount and duration of input to achieve and hold the same banking angle is different of course.

 

Regarding rudder, it seems the same (with AP or without AP) for me when Heading/Course hold is in the middle position.

Edited February 9, 2009 by ZaltysZ

[ericinexile]

As I understand it, 1 and 2 should not be different. That's really my only point.

[DragonRR]

I think you are right eric. I spent some time flying the sim today and thinking about your suggestion.

 

Assuming Alphaonesix is correct then the way the AP currently fights the input from the joystick is wrong. As you originally state the AP should switch off whilst the joystick is moved away from dead center/deadzone and only switch back on when it sees the joystick hit center/deadzone again. The damping, imo, should be present as in FD.

 

Personally I would also like to have the ability to trim the bearing separate to the cyclic position (optional for non FFB users). Combined with the AP releasing on stick movement you wouldn't really need FD at all.

 

However! Assuming all this were done... what use would FD actually be?... and therefore this makes me question whether the real KA-50 actually does work in this way at all!

[Zorrin]

I think what it boils down to and this is of course just a theory...

 

Anyone with a reasonable amount of real flight experience finds that the AP fights them. I cannot get comfortable with the APs and find myself in a similar situation to OP.

 

Other squad members are just fine with the APs.

 

This could be a number of any issues, but the most likely culprit is real experience. Having never flown a Ka-50 nor an autopilot assisted helicopter, but having 150+ hours of fixed wing time with some heli time thrown in for good measure that is the root problem for me...

 

Just a theory...

[WynnTTr]

I think what it boils down to and this is of course just a theory...

 

Anyone with a reasonable amount of real flight experience finds that the AP fights them. I cannot get comfortable with the APs and find myself in a similar situation to OP.

 

Other squad members are just fine with the APs.

 

This could be a number of any issues, but the most likely culprit is real experience. Having never flown a Ka-50 nor an autopilot assisted helicopter, but having 150+ hours of fixed wing time with some heli time thrown in for good measure that is the root problem for me...

 

Just a theory...

 

Therein lies the problem. Some pilots that have actual flying experience have ingrained procedures that are hard to unlearn. The average gamer, specifically the sim enthusiast doesn't have this and are quick to learn what works.

 

HOWEVER, given the minute detail that went into this sim, I'm fairly confident to say that ED have replicated (not recreated) how the AP/trim handles.

Lots of people have figured out how the trim functions work and can fly the BS as is. I dare say those that exclusively use the FD the majority of their flying time are in the minority.

 

Lastly, has anyone actually flown a military grade helicopter? I remember reading earlier posts from ppl that have flown russian military helis commenting on how realistic this sim is.

 

There was even a post from a black hawk pilot commenting on the realistic application of the trimmer. Don't think he had any problems picking up how the ap/trim works.

[ruprecht]

just a theory...

Anyone with a reasonable amount of real flight experience finds that the AP fights them.

 

I know that one disproof is not conclusive, but I have a "reasonable amount of RL flight experience" including a lot of rotary, and I have no problem with the AP or force trim as modelled.

[ericinexile]

....However! Assuming all this were done... what use would FD actually be?... and therefore this makes me question whether the real KA-50 actually does work in this way at all!

 

If the real Ka50 works as I suggest in this thread, then the purpose of FD makes perfect sense. The FD gives the pilot the option of flying with the AP off regardless of whether or not he is moving the stick. Because this increases workload somewhat, command guidance is displayed on the HUD. There is little compelling reason to normally fly in FD other than that sometimes pilots like to hand-fly--and sometimes autopilots fail.

[Dugulus]

I have flown a Robinson 22 (piston helicopter)for about 50 hours and I flew an S76 (turbine helicopter) for a couple hours. In the R22, you move and manually hold the cyclic wherever you want it. However, when I got in the S76, I was all messed up because I would forget to depress and holdthe "Trim" switch on the Cyclic when I wanted to put a new control input in. The Stabilization system (a.k.a. the autopilot channels) did FIGHT me (with a good amount of force). I assume the KA50, although Russian, has a similar system. I feel the trim is modeled correctly in the KA50 with the only "non realistic" part being having to return my joystick to center (of course this has to be done this way for non FFB joysticks). But I dont have FFB, so I dont mind doing this.

 

Anyway, If people have found easier ways to trim, great! But holding the trim button in to make control inputs would be the most realistic to me.

 

BTW- I tried hovering the S-76 with the autopilot channels off....uuhh, not so nice, just like the KA50.

[ZaltysZ]

As I understand it, 1 and 2 should not be different. That's really my only point.

 

I think the difference between 1 and 2 (angle vs rate) is the point of Ka-50 AP stabilization.

 

Notice this is not the same as with rudder and heading hold. If you have heading hold on and move rudder, AP will try to compensate and you will have to apply even more rudder to overwhelm AP; after you neutralize the rudder, AP will try to apply opposite rudder and return the aircraft to previous heading which it is trying to hold. Doing that (changing heading without designating new heading to AP) is fighting with AP. This is also true if you try to change heading only with cyclic.

 

It may seem that requirement to hold cyclic deflected to keep banking angle is like fighting against AP as it returns to 0 bank when cyclic is neutralized. Especially it may seem so if you are used to aircrafts whose don't do so.

 

For most pilots the most common reaction to cyclic (or stick) inputs to sides is change to rate of roll (the more you deflect, the faster the aircraft rolls), however Ka-50 stabilization works differently: the more you deflect, the bigger angle of bank (not the rate) you get.

 

Lets compare.

 

We turn the FD on. We move cyclic to side by 15 degrees. Aircraft begins to roll slowly. We neutralize the stick when angle of bank reaches 20. We have aircraft flying with angle of bank 20. Now, we bring aircraft to angle of bank 0 to repeat the test. This time we deflect the cyclic by 30 degrees (twice the amount than before). Aircraft begins to roll again, but much quicker than before. We neutralize the stick when angle of bank reaches 20. So, with different deflection we get the same angle of bank (if we neutralize the stick when it is reached), however the difference in cyclic deflection makes the difference in time which was required to get the same result (angle of bank). In other words cyclic deflection to side is related to rate in which angle of bank changes.

 

Now, we turn the FD off. We move cyclic to side by 15 degrees. Aircraft begins to roll and stops rolling when angle of bank reaches (for simplicity sake lets say) 15 degrees. We neutralize the cyclic and aircraft returns to 0 angle of bank. Now we deflect the cyclic by 30 degrees. Aircraft rolls and stops at angle of 30 degrees. We deflect the cyclic by 20 degrees and aircraft stops rolling at 20 degrees and so on. In other words angle of cyclic deflection to sides is related to angle of bank and not to rate at which angle of bank changes.

 

It is not fighting with AP you encounter, but almost different control method when FD is off. AP isn't reversing your inputs to fight you, it just transforms your input to something else to what you are not used to.

Edited February 10, 2009 by ZaltysZ

[Zorrin]

I know that one disproof is not conclusive, but I have a "reasonable amount of RL flight experience" including a lot of rotary, and I have no problem with the AP or force trim as modelled.

 

You're probably right!

 

:)

[DragonRR]

EvilBivol-1..."Example: I input left cyclic. The AP begins to receive two signals - one from the controls, telling it that the cyclic has moved to the left and by how much and one from the instrumentation, telling it that the aircraft is banking to the left and deviating from the previously trimmed position. The AP wants to bank the helicopter to the right to bring it back to trim, but because the controls signal is proving it with left cyclic input, the result is no input from the AP onto the helicopter."

 

AlphaOneSix..."It has commonly been noted that in the normal autopilot mode with the three normal autopilot channels on, the autopilot will fight you if you move the controls. This is not true. It is true that absent any control inputs from the pilot, the autopilot will attempt to hold the aircraft's heading and attitude. However, if the pilot moves the cyclic or the pedals, the autopilot is aware of that and makes no move to interfere, although it continues to provide control input for the purpose of dampening (stability augmentation is the term typically used in the West)".

This suggests to me that the OP is correct, that when you move the stick the APs effectively turn off and only dampening (as in FD) is still in effect. Either Evilbol-1 and AlphaOneSix are right... or they were "mistaken" :) I have only limited experience flying a fixed wing and therefore don't pretend to know! Edited February 10, 2009 by DragonRR

[Ali Fish]

am i correct in assuming that for non FF stick users this sim will and can never function like the real helicopter,bold statement i know. but just a suggestion.

 

the work load for a FF stick is one thing but for non FF users like myself we have to instantly think in strange terms where this whole flight theory is concerned. I have to believe my stick is somewhere its not when trimmed and how to counter act problem when looking to apply a new input for the helicopter. the result is x2 workload for pilot. Is this realism ?

 

Heres how i think the controls should work.

 

2 options in game somehow. flight characteristics for non FF and FF sticks. even if it strays further from reality. Personally having a non FF stick has instantly taken this sim out of a complete reality anyway. So to have the sim adjusted for the fact we dont all have FF sticks would be welcomed by me immensly. and im sure other non FF stick users might think likewise. i know this means a simplification to the sim for us non ff users. The work load we have at present is one thing but adding to that for non FF stick users is just too much.

 

I propose that when a non FF stick is trimmed in flight, to change that cyclic input the user should match where the stick is trimmed (much like the FF users stick remains in position) before the cyclic engages to provide new input parameter. thus with a non FF stick we can imitate the use of the FF sticks where this sim is concerned.

 

Asides this i feel its necessary to include some documentation about the necessity to use a FF stick to expereince the true potential of this might fine piece of study simulation software.

Edited February 10, 2009 by Ali Fish

[ericinexile]

I think the difference between 1 and 2 (angle vs rate) is the point of Ka-50 AP stabilization...

 

If this is true, and I don't have the knowledge to dispute it, then it really needs to be explicitly explained by ED. Because what you are talking about is not simply one of the autopilot is on or off depending on whether or not the FD is engaged. You are talking about two unique sets of flight control laws dependent on the status of the Flight Director.

 

Assuming AlphaOneSix and EvilBivol are incorrect regarding cyclic stick pressure "unlatching" the autopilot, and you and others who've posted convinced that the ED modeled Ka50 FCS is as accurate as possible are correct, then I'd say the real Ka50 is flown in a manner which is in conflict with what we have been told is correct. Because if the real pilot is flying in a situation where constant maneuvering is required then he will almost certainly need to press and hold the trim to quickly repoint his helicopter. But we have been told that real Ka50 pilots do not normally press and hold the trim. So I am just as convinced as ever that we have either been misinformed about the proper use of trim and/or FD, or we have a simulated Ka50 with some errors in the flight control logic that extend beyond the limitations of desktop joysticks.

 

Smokin' Hole

Edited February 10, 2009 by ericinexile

[DragonRR]

@Ali Fish

 

Personally I don't think it's a necessity to have a FFB stick, it would be NICE to have one and if there was a stick like the X52 or Thrustmaster Cougar with FFB then I would seriously consider buying it. Currently I find using the current trimming system in normal flight with FD for combat and general maneuvers acceptable... just not perfect.

 

It would be really nice to hear from a dev about the true situation regarding the trimming. :smilewink:

[Maverick-GER-]

@ ali fish:

am i correct in assuming that for non FF stick users this sim will and can never function like the real helicopter,bold statement i know. but just a suggestion.

 

the work load for a FF stick is one thing but for non FF users like myself we have to instantly think in strange terms where this whole flight theory is concerned. I have to believe my stick is somewhere its not when trimmed and how to counter act problem when looking to apply a new input for the helicopter. the result is x2 workload for pilot. Is this realism ?

That's why the smarties from ED included a extra little window that will open with RCTRL+Enter i believe where you can instantly see where your controls are right now... So they actually did something for us non FFB joystick users!

But if you ask like that ... how realistic is it to sit in a heated room on your computer chair with cheap controllers on your desk staring into a monitor, flying a russian attack helicopter and if you crash you just hit on refly?

[WynnTTr]

If this is true, and I don't have the knowledge to dispute it, then it really needs to be explicitly explained by ED. Because what you are talking about is not simply one of the autopilot is on or off depending on whether or not the FD is engaged. You are talking about two unique sets of flight control laws dependent on the status of the Flight Director.

 

Assuming AlphaOneSix and EvilBivol are incorrect regarding cyclic stick pressure "unlatching" the autopilot, and you and others who've posted convinced that the ED modeled Ka50 FCS is as accurate as possible are correct, then I'd say the real Ka50 is flown in a manner which is in conflict with what we have been told is correct. Because if the real pilot is flying in a situation where constant maneuvering is required then he will almost certainly need to press and hold the trim to quickly repoint his helicopter. But we have been told that real Ka50 pilots do not normally press and hold the trim. So I am just as convinced as ever that we have either been misinformed about the proper use of trim and/or FD, or we have a simulated Ka50 with some errors in the flight control logic that extend beyond the limitations of desktop joysticks.

 

Smokin' Hole

 

hmm, who would I rather place my trust in? The people that have had actual personal correspondence, guidance and real data from Kamov and KA-50 pilots? Or the assumptions of a gamer?

 

Tough choice.

 

There was a video which showed a KA-50 doing some moves. it was split into 3 screens - outside, hud(?) and pilot. Someone posted much earlier, in it you can clearly see the pilot holding down the trimmer while maneuvering, then releasing.

In fact, I copied what he was doing when starting out.

 

And Eric, I like how you quickly discount ppl that have actual flight experience to aircraft that is similar, and to easily brush off EB or Alpha16's comments. EB is closer to the development team than you or I, and going from Alpha16's comments in the trim sticky, he's worked on Russian aircraft before (pretty sure that was him). And waht about Dugulus' comments earlier?

[ericinexile]

hmm, who would I rather place my trust in? The people that have had actual personal correspondence, guidance and real data from Kamov and KA-50 pilots? Or the assumptions of a gamer?

 

Tough choice.

 

There was a video which showed a KA-50 doing some moves. it was split into 3 screens - outside, hud(?) and pilot. Someone posted much earlier, in it you can clearly see the pilot holding down the trimmer while maneuvering, then releasing.

In fact, I copied what he was doing when starting out.

 

And Eric, I like how you quickly discount ppl that have actual flight experience to aircraft that is similar, and to easily brush off EB or Alpha16's comments. EB is closer to the development team than you or I, and going from Alpha16's comments in the trim sticky, he's worked on Russian aircraft before (pretty sure that was him). And waht about Dugulus' comments earlier?

 

Man! You could not have more greatly missed my point. My fault probably--sometimes I use 20 words where 3 would suffice. A16 and EB have both stated that they believed that, in general, the autopilots in many Russian helicopters ceased control input when pilot input was sensed and they further implied that perhaps the Ka50 used the same philosophy. That is definitely not what is happening in the sim. I personally think AlphaOneSix's assumptions intuitively sound right. (I will add that neither A16 nor EvilBivol ever stated that the game was inaccurate in this respect.) Again my point is only that what I've observed in the sim does not match their descriptions found on the Trim and AP stickies.

Edited February 11, 2009 by ericinexile