w.i.p Yaw very unstable on the Apache since the latest update

[BIGNEWY]

3 minutes ago, stonewall197922 said:

You must understand, that ED have obligations before buyers. Module was released two years ago and currently FM is incomplete and unstable. That is frighten off other potential buyers and make owners angry. If you ask me at current state module is very uncomfortable to use. I tried recently refund it with no luck. Unfortunately. Im  ok with EA, but two years with no light at the end of the tunnel is too much for me. You make fixes to module and reversing previous changes. Just to understand, current state of apache FM is returned to somewhere a year ago. My suggestion is to revert changes to the state before last patch 

I'm sorry you feel that way, the flight model is work in progress, but it is far from unstable. Early access is a long process and it certainly does not suit everyone, but as you can see work is continuing. I have passed on everyone's feedback here in the thread, but I am also getting others who say they are not noticing that much of a change since the last patch. 

Work from our last major patch:

https://www.digitalcombatsimulator.com/en/news/changelog/release/2.9.14.8222/

DCS: AH-64D by Eagle Dynamics

• Improved FСR synchronization. 
  Known synchronization bugs that we are working on:
  • FCR animation for all users (currently only displayed for helicopter host).
  • RMAP background (missing) on first client login if HOST scanned and stopped. With subsequent scan packets, everything is displayed for everyone.
• Fixed: Position Hold oscillation:
  • When initially engaged, Position Hold can cause some large oscillations in attitude and lateral velocity. At times, the Position Hold can generate a ground speed in the opposite direction well beyond the initial position and velocity when initially engaged.
  • When disengaged, Position Hold seems to generate an erroneous input through the SCAS. For example, if the helicopter was in a stable and trimmed hover and the Force Trim/Hold Mode switch is pressed to the aft position to disengage Position Hold, the green SCAS indicator may jump aft, incurring a nose up attitude when such a movement was neither commanded nor necessary.
• Fixed: RWR displays AI AH-64D FCR as "U" Unknown Fixed-Wing, even when AI AH-64D does not have FCR installed.
  • AH-64D FCR removed from AH-64D RWR classification library.
• Fixed: HAI - The left wheel sinks into the ground upon landing.
• Fixed: No fire zone NFZ desync in multicrew.
• Fixed: RFHO target data erased when FCR selected as sight on FCR-equipped AH-64D.
• Fixed: Firing RF missile using TADS is advancing FCR Next-To-Shoot (NTS) target.
• Fixed: NTS is not sequencing when firing RF missile while FCR is selected sight.
• Fixed: Cued LOS Reticle displaced downward in CPG HDU when FCR slaved to FXD.
• Fixed: Displaying COM page on right MPD causes C/S> error on left MPD.
• Fixed: De-selecting G/S on VIS page is enabling VSEL (T6) if a CPG underlay is selected.
• Fixed: George CPG unit IFF label errors. Labels updated in the following manner:
  • Inappropriate characters removed.
  • Prevent overflow of characters outside of Target List.
  • Revised confusing nomenclature of some unit types.
• Fixed: George CPG ranging behavior when employing gun.
  • When tracking a target with the gun actioned, George will continuously lase to ensure accurate ballistics calculations prior to firing.
  • George will only engage target within 2500 meters when in ROE FREE FIRE. May be overridden to engage out to full range of gun using Consent To Fire command.
• Fixed: George CPG gun burst interval when firing gun in ROE FREE FIRE.
  • George will pause between bursts to wait for rounds to impact target before firing another burst. May be ordered to fire another burst early using Consent To Fire command.

Changes to position hold were made, so I have asked the team to check if it could have affected yaw in some way. 

thank you 

[Leg2ion]

8 hours ago, BIGNEWY said:

Changes to position hold were made,

Oddly, before this latest patch - I don't think I suffered none any of the issues (that I was aware of anyway) that have been fixed?

Edited March 28 by Leg2ion

[BIGNEWY]

1 hour ago, Leg2ion said:

Oddly, before this latest patch - I don't think I suffered none of the issues (that I was aware of anyway) that have been fixed?

I think we often get used to a particular way of doing things or a feeling of flight, and when slight changes are made some people are more finely tuned and notice it more, I had people contact me to say things feel better now and others who say the yaw or something else feels different. I've asked the team to check to see if there is anything going on. 

thank you 

[BIGNEWY]

The behaviour of SCAS was changed in the most recent update to address oscillations in Position Hold, this may be causing some Yaw SCAS difficulty, but please be sure when using trim, to press and hold while moving controls and then release when you have finished and in position.

Simply tapping it does not allow adequate time for the SCAS servos to re-center within their range of authority. 

If at low speeds when Heading Hold is active, if you are fighting the Heading Hold without using trim and then exceed the force trim breakout value, the Heading Hold will disengage and amplify your pedal input. Reference the Flight Management Computer (FMC) section of the Early Access Guide for better understanding how trim and the hold modes function. Particularly Heading Hold, which is always enabled below 40 knots.

Can I please ask you all to try and let me know if you see a difference when pressing and holding the trim until you are in position. 

thank you 

[stonewall197922]

4 часа назад, BIGNEWY сказал:

The behaviour of SCAS was changed in the most recent update to address oscillations in Position Hold, this may be causing some Yaw SCAS difficulty, but please be sure when using trim, to press and hold while moving controls and then release when you have finished and in position.

Simply tapping it does not allow adequate time for the SCAS servos to re-center within their range of authority. 

If at low speeds when Heading Hold is active, if you are fighting the Heading Hold without using trim and then exceed the force trim breakout value, the Heading Hold will disengage and amplify your pedal input. Reference the Flight Management Computer (FMC) section of the Early Access Guide for better understanding how trim and the hold modes function. Particularly Heading Hold, which is always enabled below 40 knots.

Can I please ask you all to try and let me know if you see a difference when pressing and holding the trim until you are in position. 

thank you 

Managed to make helicopter more stable. I think main key to stabilize is fine-tune your equipment with correct settings on "special" tab in options menu.

I'm loosened clutch dampers on my VPC joystick to make it return to center and move easily. Pedals VPC leaved them with tightened clutch dampers. In settings Cyclic - "FFB friendly with springs", pedals - without springs. Used short trim tapping. Attitude hold engaged only when helicopter is hovering. Instability in yaw channel not gone. Much more difficult to control in comparing with other helicopter modules. Helicopter can be more stable with proper equipment and fine-tuning. 

[Leg2ion]

7 hours ago, BIGNEWY said:

Can I please ask you all to try and let me know if you see a difference when pressing and holding the trim until you are in position. 

So spent around an hour going hover/fwd flight/hover and it does help, but it still doesn't seem to be as smooth with the yaw movement being overamplified when moving from fwd flight into the hover. So zero stores in Sochi/Caucasus. Trying to maintain a <100 ft altitude whilst transitioning into a hover and maintaining heading from 90-100 kts fwd speed. I found that I dropped to between 23-30% torque in a 5-10^o nose up attitude to drop forward speed and whilst maintaining my desired altitude. As I decelerated through around 40-30kts I started to increase collective to smoothly maintain Nr and height, whilst still reducing speed. Due to the collective increases I was then compensating with (left pedal) yaw to counteract increased torque. However - at around 65-75% torque and around 5-10 kts there seemed to be further larger yaw input required as the aircraft continued turning to the right despite my attempt to reduce the movement. Tried both with the trim button pushed throughout the manoeuvre, and trim button pushed as starting to decelerate, released when in a 'steady state' of deceleration, then pushed again once starting to increase collective and counteract with yaw inputs - so as not to exceed any trim breakout limits. No real difference - maybe a bit better with trim pushed throughout but that may just be personal preference/subjective.

On top of this, once in a stable hover with both altitude and attitude hold on, if I blipped the trim release button fwd momentarily, whilst making no control inputs, a series of yaw/roll oscillations occurred that gradually dampened out at best, at worst causing me to transition forward to avoid 'chasing' the corrective inputs needed. In my mind (though could be wrong) by momentarily selecting trim release but without making any inputs there should be no movement of the controls to speak of - all I am doing is re-centering the trim to null against the current control positions to re-establish 'full authority' back to the trim system.

Hope the above makes sense. Trk file attached.

Hover Yaw.trk

Edited March 28 by Leg2ion

[[DE] T-Bone]

8 hours ago, BIGNEWY said:

The behaviour of SCAS was changed in the most recent update to address oscillations in Position Hold, this may be causing some Yaw SCAS difficulty, but please be sure when using trim, to press and hold while moving controls and then release when you have finished and in position.

Simply tapping it does not allow adequate time for the SCAS servos to re-center within their range of authority. 

If at low speeds when Heading Hold is active, if you are fighting the Heading Hold without using trim and then exceed the force trim breakout value, the Heading Hold will disengage and amplify your pedal input. Reference the Flight Management Computer (FMC) section of the Early Access Guide for better understanding how trim and the hold modes function. Particularly Heading Hold, which is always enabled below 40 knots.

Can I please ask you all to try and let me know if you see a difference when pressing and holding the trim until you are in position. 

thank you 

Thanks for the tip about pressing the FTR for a long time. This has weakened the oscillation somewhat, but it is still present.

In the attached tracks I was still able to reproduce the described oscillation behavior with long press FTR - both in hovering flight and in forward flight at low speed.

Yaw channel:

• You can clearly see that the yaw channel takes over the new zero position for a brief moment and damps it slightly, but then starts to oscillate left/right and only then levels off again.

ATT hold:

• I was also able to reproduce the forward and backward pitching of the ATT hold, even though the helicopter was well trimmed when the mode was activated and no control inputs were made.
• The behavior no longer occurred until the helicopter was aligned again using FTR.
• This phenomenon occurs for the first time when using the mode and then only very rarely during flight.

 

The yaw channel shows a control problem, as it swings up after a short damping.

The ATT hold mode occasionally exhibits faulty behavior, especially when first activated.

Deactivating the position hold has already brought a significant improvement, which must be mentioned positively . However, there is still a need for optimization, but that has already been communicated by you.

pitch_oszilate_ATT-Hold.trk yaw_oszilate_10-30kt.trk yaw_oszilate_hover-heading.trk

Edited March 28 by [DE] T-Bone

[Ignition]

The helicopter is very unstable now

[stonewall197922]

With training and fine-tuning helicopter can be somewhat more stable. But neverless stil quite unstabe at low speeds. I describe current flight model state is back like  it was a year ago.

[Ulfrun]

Almost 62 years old and 40 years experience with different flight simulators, I play DCS since more a decade and appreciate the huge work behind. Thank you for the great work, you make my day! It's an incredible progress since I flew the Gunship from Microprose to the real simulation DCS present us. Enough of the warm words...;-)

As a flight enthusiast I watch any videos of Apache helicopters I can find: Either I am to stupid, bloody unexperienced or both..... but when flying the Apache, there was - and still is a clear difference how accurate a real Apache pilot in a real Apache can control the "thing" and me in DCS.

Yes, I am very aware to be not a real pilot but my skill is not so bad as the actual spongy flight behaviour when hovering, make slightly or sharp corrections during landings/starts, hard braking from flight to hover, landings exact were you want to go...and not beside. I am looking forward to the next updates. 

[=475FG= Dawger]

1 hour ago, Ulfrun said:

Almost 62 years old and 40 years experience with different flight simulators, I play DCS since more a decade and appreciate the huge work behind. Thank you for the great work, you make my day! It's an incredible progress since I flew the Gunship from Microprose to the real simulation DCS present us. Enough of the warm words...;-)

As a flight enthusiast I watch any videos of Apache helicopters I can find: Either I am to stupid, bloody unexperienced or both..... but when flying the Apache, there was - and still is a clear difference how accurate a real Apache pilot in a real Apache can control the "thing" and me in DCS.

Yes, I am very aware to be not a real pilot but my skill is not so bad as the actual spongy flight behaviour when hovering, make slightly or sharp corrections during landings/starts, hard braking from flight to hover, landings exact were you want to go...and not beside. I am looking forward to the next updates. 

IRL, you can feel changes long before you can see them, making fine control much, much easier. Seat of the pants exists as a term for a very solid reason.

A simulator relying upon visual cues only will never be able to be controlled with the same precision as the real thing.

I used to explain this to my students in $20,000,000 full motion simulators which tried to replicate some of that seat of the pants sensory input when they complained that the simulator was harder to fly than the real thing. It isn't something that can ever be overcome, mechanically.

I flew the Apache after this update and found it pretty similar to the previous version.

[Raven (Elysian Angel)]

1 hour ago, =475FG= Dawger said:

A simulator relying upon visual cues only will never be able to be controlled with the same precision as the real thing.

Multiple RL helicopter pilots have stated on these very forums that flying helicopters in DCS is harder than IRL for the exact reasons you mention.

But I think that some liberties can be taken to make sure the end result that the user experiences is very close to the end result the RL pilot experiences - which is probably why the Roadmap has an entry for:

• Automatic Pedal-Input Assist Option

It is however striking how easy other helicopters in DCS are at low-speed control compared to the Apache, from Mi-8 to UH-1 to the recent OH-58, the latter having super care-free handling. In the Apache at sub-40kts speeds I feel like I am constantly fighting the FCS and need to pay extra attention to not have the tail swing wildly out of control…

[stonewall197922]

8 часов назад, =475FG= Dawger сказал:

IRL, you can feel changes long before you can see them, making fine control much, much easier. Seat of the pants exists as a term for a very solid reason.

A simulator relying upon visual cues only will never be able to be controlled with the same precision as the real thing.

I used to explain this to my students in $20,000,000 full motion simulators which tried to replicate some of that seat of the pants sensory input when they complained that the simulator was harder to fly than the real thing. It isn't something that can ever be overcome, mechanically.

I flew the Apache after this update and found it pretty similar to the previous version.

Don`t have such difficulties in control other owend DCS helicopters  (Mi-24, Ka-50,Kiowa). Flying Apache at speeds <40, i`m constantly fighting with unpredictable helicopter movements. We`re talking about game. This is not a military grade simulator for RL pilots training. So it must be comfortable to use by majority users without issues. But now many users struggle just to make it fly straight or hover. How long Apache is in EA? Three years? How long it will takes to finish it? If you ask me, this helicopter was uncomfortable to fly and  unstable at early start and remains in such state now.

Edited March 31 by stonewall197922

[bies]

4 hours ago, Raven (Elysian Angel) said:

Multiple RL helicopter pilots have stated on these very forums that flying helicopters in DCS is harder than IRL for the exact reasons you mention.

 

https://dofreality.com/product/heli/motion-heli-cockpit-6-axis-hero-h6/

[Raven (Elysian Angel)]

29 minutes ago, bies said:

 

https://dofreality.com/product/heli/motion-heli-cockpit-6-axis-hero-h6/

Perhaps I should have believed that Nigerian prince

[HPLsCat]

At the risk of starting a war here (and that's not my intention), the Apache, as it stands today, is one of, if not the, most nimble module in DCS at any speed. You have to fly ahead of it, and practice a lot, but the rewards are there to be had.

Everything it does is predictable, and it rewards hard work.

[bies]

13 hours ago, HPLsCat said:

At the risk of starting a war here (and that's not my intention), the Apache, as it stands today, is one of, if not the, most nimble module in DCS at any speed. You have to fly ahead of it, and practice a lot, but the rewards are there to be had.

Everything it does is predictable, and it rewards hard work.

That's true, ED's Apache has came a long way and now its flight control are working really well and predictable, compared to the first release.

Some fine tuning, especially the yaw control, and it will be as controlable and precise as that:

 

[HPLsCat]

It's that controllable and precise now, though. I do it all the time in a combat loadout, drop into little clearings, worm through the woods and along rivers.

At high torque, mashing left pedal can absolutely ruin your day, but provided that's accounted for, nothing in that video is out of the ordinary.

[admiki]

It's not that DCS 64 is nimble, it's random twitchiness.

Even bradmick is saying that FM needs improvements.

[HPLsCat]

Flight models always seem to be a work in progress, my disagreement comes with terms like random or unpredictable. It's neither of those things, IME.

[admiki]

1 hour ago, HPLsCat said:

my disagreement comes with terms like random or unpredictable. It's neither of those things, IME.

Agree to disagree.

That's what makes life fun.

[stonewall197922]

Look this discussion is pointless. Even ED agreed, that FM needs improvement and fixes. We all have different equipment. And equipment is tuned according to our own feelings. Some users i`m sure didn`t even want to notice such a problems(e.g. "i like it, and everything else i dont care"). People here just expressing their opinion, that they don`t like how helicopter is behaving right now. I agree with you, that with training and fine-tuning, some degree of stability can be achieved. But in my case nevertheless i found that FM is currently far from stability and not comfortable to use. Yes its nimble but at low speeds it needs improvement. Such a problems is have only with apache.

Edited April 1 by stonewall197922

[Floyd1212]

I think everyone can agree that the FM is still a WIP.  ED has said it, and those of us that fly it have said it.

Every so often we get a patch where ED has worked on the FM and made improvements, and we adjust to the new changes and life goes on.  Over time, the FM has gotten much better, and I found the Apache quite enjoyable to fly.  It had some quirks, but it was manageable if you put time in to practice and develop muscle memory.

I don't think I am out of place to say the reason this specific thread was created 10 days ago by DaveTC is that after the last patch there was a noticeable change in the behavior of the tail rotor or SCAS that seemed like a step in the wrong direction.  Instead of an improvement to the FM, this feels like a regression.  And because there was a patch around Christmas time that specifically improved the behavior of the tail rotor and SCAS in low-speed flight, and while in a hover, there was concern that whatever changes that were made for that patch got "undone" with this one.

[LorenLuke]

Firstly: while I don't think it's the case that Boeing would want this, I ultimately don't know what is actually the case in real life. That said, the point of the flight model is to make it realistic, not necessarily what's 'easier'. If the more difficult, regressive, twitchier yaw channel is more true to life in the real aircraft than what we had before, that's the one that should be implemented in the simulator.

I very much remember when the LMC ground stabilized when you turned it on and now it doesn't; it's not as easy to use anymore, but it is more realistic. This is how it should be.

Now, yes, everyone (including the experts!) agree that the flight model needs improvement, but just because 'They think it needs to be better too' doesn't necessarily mean the areas they think it needs improvement in (e.g., crabbing) are the same areas you do (unstable yaw channel).

My point is, I'm sure that they're trying to make the flight model as real to life as they can regardless of how 'easy' that ends up being. Because this game is ultimately meant to be a simulator as close as they can reasonably get, not some arcade airquake with simplified controls.

[HPLsCat]

Yes, and hyperbole doesn't help.