Wacky trimmer - bug or feature?

[zazz]

Quick question - forgive in advance if double posting

 

 

I've had this problem with both the old and the new trimmer implementation - whenever i press the trim button, the aircraft "jumps", around 10 degrees, sometimes up, sometimes down, sometimes rolls... Really quite unpredictable.

 

This makes it impossible to just click the trimmer button to set it, so i have to press and hold trim, compensate for the "jump", and then release. However, during the "jump", the virtual cyclic does not move a bit. Same thing happens when i use "t" on the keyboard.

 

 

Is this a depiction of a realistic trimmer, or is it my Logitech Extreme 3d Pro joystick? Thanks

[GGTharos]

I think it might be your stick or more specifically your stick settings ... though I can't begin to tell you where to look to get it fixed up. 10deg is quite a lot.

[Frederf]

The RCtrl-Enter virtual axes indicator is helpful in seeing what control inputs are being made.

[Chops]

Trim - Pitch Oscillations

 

Check out this thread. I have had similar issues.

 

 

http://forums.eagle.ru/showthread.php?t=41851

 

 

Good luck.

[sobek]

Is this a depiction of a realistic trimmer, or is it my Logitech Extreme 3d Pro joystick? Thanks

 

Here's what i think might be happening, correct me if i'm wrong: You might be trimming too seldom, so right before you press trim you are already fighting the autopilot's 20% authority. When you hold down trim, the autopilot recenters it's control input instantanously, thus causing a jerk away from where you had trimmed to previously. If this is the case, try trimming in smaller increments or holding the trim down BEFORE you change attitude, do the attitude change and then release the trim. See how this works out for you.

Edited December 15, 2009 by sobek

[EtherealN]

Well sobek, it shouldn't actually move your cyclic position - rather it should release the springs that held the stick in it's previous center position and, when released, will reapply the springs to cause the stick to have a new "center" position - thereby allowing the helicopter to have a stable attitude even if you release the stick.

 

On the question of whether this implementation is a depiction of a realistic trimmer or not - it both is and isn't. Basically, if you have force feedback then the trimmer can hold your stick's position through the force feedback just like it would a real stick. However, if you don't have FFB (which I believe is true for your case) then you will have to center your stick within 0.5 seconds or, using the "new" trim method, it'll hold the input until you have centered the controls. In either case a jumpy stick might cause jerky inputs (although it should then be visible in the cockpit too).

 

zazz, attach a track file - after flight, select "save track", you'll then find it in the Tracks directory under C:\Games\BlackShark\Missions or wherever you installed the simulator, then on this thread click "go advanced" and find the paperclip icon to upload that track file to the forum and have it attached.

 

That'll let us look at whether it's just a question of practice for you or whether you might want to change some joystick settings. I personally suspect that it's justd practice - I sometimes get jerky motions when trimming with my Saitek Cyborg Evo (which is a force stick, though I run with Force disabled due to incompatibilities) but most of the time I can trim without such issues simply through practice. But in your case there might perhaps be reason to increase deadzones on your stick.

[sobek]

Well sobek, it shouldn't actually move your cyclic position - rather it should release the springs that held the stick in it's previous center position and, when released, will reapply the springs to cause the stick to have a new "center" position - thereby allowing the helicopter to have a stable attitude even if you release the stick.

 

Yes thank you, i have an FFB stick and know how it works ;)

 

What can cause jerky movement is when you force the helo out of the attitude that the autopilot tries to hold and then press trim, because the AUTOPILOT CHANNELS will recenter, dunno how you came to think i meant that it would cause any input to the stick. I hope i made myself clearer this time.

[EtherealN]

Hm, yeah, so you are basically talking about a case where the AP is using all it's authority due to being out of trim, and pressing the trim causes this to cease? That could explain it I guess, but I don't really manage to get it to fit with my own experiences, since I've only ever really had those jumps on release of trimmer, not when first pressing it.

 

I'm gonna have to try to cause the same behaviour and see if I can make sense of it.

[sobek]

on release of trimmer, not when first pressing it

 

That is indeed weird and most likely not caused by what i described. :huh:

[EtherealN]

Well, in my case it's caused by not being fast enough with re-centering (I use the old method). But that can't explain the behaviour noted by the OP since it apparently is the same on both methods.

[Nate--IRL--]

Don't forget that the trim is not only moving the stick position centre (in the simulation), but also changes the autopilot stabilisation parameters. It is my beleif that the non FFB trim and the change in the autopilot stabilisation parameters, causes this jump.

 

Imagine a situation where the autopilot stabilisation might be using the 20% of it's 20% authority to maintain it's programmed parameters.

 

When you press the trim button the new centre position is established, this new position does not take into account the 20% input from the autopilot that was previously there, and that has now been removed as the Autopilot paremeters have been reset.

 

Another reason may be that in the real Ka-50 when you press trim, the Cyclic is in the position the autopilot has placed it, when you release it the autopilot has new parameters, but the stick is in the same position

because your hand is on it, nothing changes.

 

In the Sim, you effectively let go of the stick as you release the button, while returning your stick to centre, leaving the autopilot, which now has new perameters, to fend for itself, hence the jump.

 

This trim jump is a concession to simulating the trim/autopilot system without the real hardware. The solution is to trim much more often or use the Flight Director (Trim Jump is one of the main reasons I use it so much, but thats for another thread :) )

 

Edit:- just to make it clearer the ramblings above, there are 2 causes. First is the Difference between the simulated hardware and the hardware you plug into you computer, and the other is simulated trim system itself.

 

Edit2:- This post is still incomprehensible, and I wrote it. Ignore all of the above and just read the first edit. :)

 

Nate

Edited December 15, 2009 by Nate--IRL--

[sobek]

Another reason may be that in the real Ka-50 when you press trim, the Cyclic is in the position the autopilot has placed it, when you release it the autopilot has new parameters, but the stick is in the same position

because your hand is on it, nothing changes.

 

I do not think that the AP moves the stick in the real Ka-50, it's input goes just to the servos, so on the stick you feel nothing of the AP.

[EtherealN]

Actually, the new centre position is determined by your input when you release the button. However, my understanding is that while the button is pressed the AP will not fight you to keep the previously "programmed" attitude. Basically when you press the trim button you are telling it "new attitude coming, stand by" and when you release it you are saying "this is what I want, keep that".

 

In the real helicopter, if this involves a new stick position, this will cause the stick to (mechanically) stay in that position even if it is off-centre.

 

However, note that the OP has issues with both trim methods. This means that it is not caused by the release-part of it. With the new method there should be no twitch on release since the simulator will simply not accept any input from the stick until you have centered it - and that center position will be made to equal the input you had when you released the button.

 

Basically, if you get any jumpiness with the new method then something is acting weird, because there should be no input at all until physically centered.

 

And if you get jumpiness when pressing (as opposed to when releasing) the trim key while on the old method, I also think something is being weird.

 

Sobek has an interesting hypotheses for both cases though.

[EtherealN]

To clarify real Ka-50 trim as it applies to the cyclic and as far as I've understood it.

 

Starting with 0.0 (x and y) input, we press trim and apply some forward pressure, let's say we are placing the stick at a position equivalent to 0x and 10y (in degrees). Here we release the trim button and then release the stick, since we are happy with the attitude this achieves. This then causes the stick to stay at 0x and 10y. We then apply some pressure to the left, say -10x, giving us -10x and 10y - but without applying use of the trimmer - and then release. The stick will now move back to it's trimmed center - that is, 0x and 10y.

 

That make sense?

 

The issue with non-FFB sticks of course is that we cannot make the stick get that new "center" position, so the simulator has to cease reading commands either for a period of time (old method) or until the stick is centered (new method). Under the old method you can get jerkiness on release through not being centered when that time period is out, since if you trim for 10 degrees forward, release trim button, and then keep your stick physically at 10 degrees forward this will become 20 degrees after the time period allowed is over - causing an instantaneous "jump" from 10 to 20 degrees forward cyclic input.

 

This should not happen under the new method, and under neither method should it happen when pressing the key - only when releasing it.

 

Which is why I am intrigued by Sobek's idea and would love to have that confirmed.

[Nate--IRL--]

 

This should not happen under the new method, and under neither method should it happen when pressing the key - only when releasing it.

 

Which is why I am intrigued by Sobek's idea and would love to have that confirmed.

 

Indeed, in your example you have disregarded the effect on the Stabilisation system, and only concerned yourself with the stick position.

 

Having thought about it a bit more I think Sobek is dead on, It seems the Stabilisation syetem changes on pressing the trim button.

 

It changes from:-

Attitude + pitch and heading hold (Stick position + Damping + 20% authority) to

Stabilisation mode which is just Stick position + Damping (ie Flight Director)

 

If you have not trimmed often enough, while flying through different parts of the flight envelope (eg 150 IAS to 250 IAS) The autopilot could well be using its max 20% authority to maintain Attitude + pitch and heading hold. When the trim is pressed the system switches to Damping only, so in effect 20% of the input to the flight surfaces disappears instantly.

 

The 20% authority will reappear when the trim button is released.

 

Nate

[sobek]

If you have not trimmed often enough, while flying through different parts of the flight envelope (eg 150 IAS to 250 IAS) The autopilot could well be using its max 20% authority to maintain Attitude + pitch and heading hold. When the trim is pressed the system switches to Damping only, so in effect 20% of the input to the flight surfaces disappears instantly.

 

The 20% authority will reappear when the trim button is released.

 

Yes you got what i meant.

 

Note that this also happens naturally when you press trim, move the stick and release trim in fast succession, because the attitude does not stabilise but the AP still tries to hold the (more or less) old attitude. But now your stick deviates too much for the AP to hold that attitude so the chopper changes attitude. If you wait for the chopper to stabilise attitude and then hold down trim, for sure there will be a jerk from the AP channels centering. (All this just to illustrate the issue at hand).

 

What i have found to work best is to hold down trim until the chopper has stabilised in the wanted attitude and then release trim, this will give the AP maximum authority in all axes and all directions around the trimmed attitude. This way there will be hardly any if no jerk at all when the trim button is depressed the next time.

 

Edit: zazz, if the problem persists, please post a track of it happening. If not it would be nice to know what was wrong. :)

Edited December 16, 2009 by sobek

[EtherealN]

Well, that may be as it may, but I have been flying around like a maniac today in all kinds of ways, totally ignoring the trim - and when I finally do this I am unable to replicate the issue. Depression of the trim key has no effect whatsoever on stability.

 

So I'd really like to see a track file to help ensure that my way of attempting to replicate it is correct.

 

EDIT: All I had to do was post that. :) I think I have replicated it now through this method:

 

1) Take off, fly straight across the runway, come to a hover.

2) Enter auto-hover.

3) Trim-and-hold

(Hitting the trimmer real quick caused nothing weird to happen.)

 

That caused me to pitch up some 20 degrees. I have not been able to make this happen in flight, though.

Edited December 16, 2009 by EtherealN

[sobek]

Well, that may be as it may, but I have been flying around like a maniac today in all kinds of ways, totally ignoring the trim - and when I finally do this I am unable to replicate the issue. Depression of the trim key has no effect whatsoever on stability.

 

So I'd really like to see a track file to help ensure that my way of attempting to replicate it is correct.

 

EDIT: All I had to do was post that. :) I think I have replicated it now through this method:

 

1) Take off, fly straight across the runway, come to a hover.

2) Enter auto-hover.

3) Trim-and-hold

(Hitting the trimmer real quick caused nothing weird to happen.)

 

That caused me to pitch up some 20 degrees. I have not been able to make this happen in flight, though.

 

Let me set up my new HOTAS and i'll get back to you. :)

[zazz]

Really appreciate this guys, will have a track file in a few hours. Thanks

[zazz]

Attaching the track file. It was kinda hard to achieve the most noticable results.

 

Right after takeoff, i steady the machine, center the stick, and press and hold the trimmer a few times, and that causes the machine to pitch up. I do the same thing a few more times throughout the flight. Sorry for the sloppy flying.

 

Does this look normal, is this the AP channels re-setting? Oh, and does the rudder get trimmed in RL too? Thanks guys

trimmer_05.trk

[mdee]

Quick question - forgive in advance if double posting

 

 

I've had this problem with both the old and the new trimmer implementation - whenever i press the trim button, the aircraft "jumps", around 10 degrees, sometimes up, sometimes down, sometimes rolls... Really quite unpredictable.

 

This makes it impossible to just click the trimmer button to set it, so i have to press and hold trim, compensate for the "jump", and then release. However, during the "jump", the virtual cyclic does not move a bit. Same thing happens when i use "t" on the keyboard.

 

 

Is this a depiction of a realistic trimmer, or is it my Logitech Extreme 3d Pro joystick? Thanks

I have the same issue and it's not your joystick (I use X52). There is definitely something fishy with trim implementation in this game, I learned to live with it, but it's not working properly. Being forced to hold and release trim because of this jump is just wrong. You should be able to press-release (just as you do on your track, that's how you would trim a helicopter with this type of trim in RL) without such drastic change of flight parameters.

Anyway it's fine as it is, I don't mind, I understand that without having a proper force feedback it's really hard to implement, just don't try to press and hold method if you try to fly real helo ;).

[mdee]

Track, attached.

trim.trk

[EinsteinEP]

Nothing fishy going on here! All the cases outlined here sound like the Trimmer, autopilot, and the beautiful Kamov coaxial rotor systems are operating exactly as they should.

 

[edit: scenario doesn't demonstrate the concept as clearly as I had hoped.]

 

When you trim for nose down attidue, the cyclic is "centered" in that position and the AP attempts to hold that attitude per the PITCH autopilot channel. Initially, not a lot of "extra" control from the autopilot is required to hold this attitude.

As airspeed increases, the pitching moment created by the rotors changes, requiring more down cyclic to keep the nose from pitching up (google "Translational Lift" for more info on this behavior). The autopilot automatically adds this input as you accelerate to maintain the commanded pitch angle.

Once you press the Trimmer button, the autopilot control feedback responses are instantly canceled, which means the down cyclic that the autopilot has been adding is suddenly removed. To the rotors, it appears as if the cyclic just got yanked back. To the unawares pilot, it appears as if the helo is going crazy!

Edited February 2, 2010 by EinsteinEP
scenario didn't show what I wanted it to show

[AlphaOneSix]

When you press the trimmer button, the autopilot inputs to the servos are reset. That is, any attitude hold input from the autopilot is removed when you press the trimmer button. If you move the cyclic without holding the trimmer down, and then you finally do hit the trimmer, you'll get a jump as the autopilot inputs are removed (until you release the trimmer button, then the autopilot starts adding inputs again). This is exactly what happens in the actual aircraft, and I must say that it is modeled beautifully and is not a flaw in the game.

 

This is why you should either always hold the trimmer down while moving the controls, or make many trimmer presses during the control movement so you don't build up much so much control input against the autopilot. In real helicopters, the trimmer gets a workout, I'd say it's the most-pushed button on a helicopter, hands down (besides maybe the ICS switch, if the pilot is chatty).

Edited February 1, 2010 by AlphaOneSix

[mdee]

This is why you should either always hold the trimmer down while moving the controls, or make many trimmer presses during the control movement so you don't build up much so much control input against the autopilot. In real helicopters, the trimmer gets a workout, I'd say it's the most-pushed button on a helicopter, hands down (besides maybe the ICS switch, if the pilot is chatty).

 

Maybe.

So, you saying I should push the trimmer 3-4 times during 5-10 degrees pull up (that's 3-4 times a second, more or less) on cyclic when helo is hovering (ie. no speed)? If I do that, that jump would build up to around 30-40 degrees, I can demonstrate that and post a track, I've tried. By the way, the jump is happening on release, not on press.

 

I'd definitely not hold the trimmer for more than a second or 2, ever. I want to stay alive for a while ;). That's actually quite nicely modelled in DCS Blackshark, helo gets very unstable quick if there is any wind and trimmer pushed down.

 

Anyway, I don't mind, the game is great and definitely most realistic helicopter simulator which can be bought for less than 50$ (I don't expect military grade sim)