Trim is modelled wrong for FFB sticks?

[zerO_crash]

@BIGNEWY @EvilBivol-1 @Wags @NineLine

I recently got myself the Brunner CLS-E joystick (FFB), been many years on spring joysticks, and before that on MSFFB2. Something that bothered me back with MSFFB2 as well as now with Brunner, is that while the stick is set up correct (FFB enabled, Default trim mode), there is a major issue with them that doesn't make sense IRL. What I'm talking about is the overcorrecting trim. This issue has been raised now with Mi-24P, however it has always been there with Ka50 for FFB-sticks.

 

Basically, when you trim, often banking right or left, or pitching up or down, upon releasing the trim (click-method), there is an overcorrection added. This both seems wrong, and counterintuitive. I know the AP very well in the Russian helicopters, and I do know how their trims work. The problem is that this just doesn't make sense. Now having a FFB-stick like Brunner, I have the exact same cyclic as a helicopter pilot has IRL, they way it works and the way it "should" behave. I would expect to not have any jinxing around when re-trimming with big changes in speed, attitude, altitude, etc... between. Rather, that I would have to trim multiple times in the new position, in order to remove errors and imprecision generated over time from before, and settle a stable trim.
 

Take a look at the video below and turn up the volume so you can hear the pilot using the trimmer. This stable behavior with such seldom click-trimming cannot be reproduced in DCS with a pro-grade FFB-stick. If you try that, you will not get as smooth transitions, and worse yet, you will have to trim minimum 2x as much just to correct the over-correction:

 

https://m.youtube.com/watch?v=eq1kkvxAsd0

 

For a visual representation of the problem, please look at this post:

 

https://forums.eagle.ru/topic/274554-trimap-operation/

 

I post this issue in DCS world because it's a Ka50/Mi24/Mi8 issue. At least from checking videos, and using my knowledge of the systems (trim and AP), I cannot see such behavior being correct for a stick that can trim out of center. It just doesn't make sense. Back in MSFFB2 times, I thought it was imprecision of trim, and otherwise errors and inaccuracies, like they naturally exist. However looking at it now, the oversteering is just too pronounced for that.

 

It would be really nice if someone from ED could tell me whether this is IRL correct behavior, or not. Primarily for the Ka50, but the others as well. Am I in the wrong, and this is actually correct behavior for a FFB stick? It feels like "Joysticks without spring and FFB" trim mode is more accurate, without the jinxing and overcorrecting (I have a steady hand, and know what delicate control is) that FFB-sticks experience. 

 

Thank you in advance!

 

zerO

Edited June 18, 2021 by zerO_crash

[zerO_crash]

Ok, I've done more testing, and finally figured it out. I can say that there is nothing wrong in general with the trim in DCS and FFB joysticks. After having a accidental keybinding reset (installed new simshaker version), the issue seemed to go away. It made me think that it must have someting to do with Brunner, and apparently it is the case.

 

So, since Brunner doesn't support DirectX FFB natively, it uses its own application, CLS2Sim, to generate FFB-effects based on the relevant DCS .lua. In other words, to get FFB effects, one has to use CLS2Sim. Since Brunner is fed data from CLS2Sim, it also means that it won't trim out of centre when using DCS-trim. That's why we have to use "hardware trim" from CLS2Sim in order to have the stick remain at the new trimmed position out of centre. So far so good, everything works. 
 

Now, knowing that the AP is tied to the trimmer-button, and references to it, I thought that I'll just double-bind the the trimmer in DCS to the same button that I have set as the "Hardware trimmer" in CLS2Sim. Sadly, this is what causes the overcorrection when trimming. I thought that the overraction would be gone when enabling "FFB" in DCS settings. Sadly it does not in this case.

 

The problem then is that:

 

a) Either I use FFB in DCS and thus have to enable Hardware trim in CLS2Sim in order to get FFB-effects and out of centre trim. However then I cannot provide updated AP-info when using the trimmer (otherwise if I double bind the trimmer in DCS, it gives me overcorrection). (Module trimmer option - Default)
 

or

 

b) I turn off FFB in DCS, enable Hardware trim in CLS2Sim, double bind the trimmer in DCS, but now use the trimmer-option "Stick with no spring and FFB". That way, my Brunner trims out of centre, the AP get's its updates and the helicopter isn't overcorrecting on trimming. Sadly though, I get no FFB in my Brunner, nor my Jetseat.

 

This is really a crazy issue. What we, with a Brunner-stick, need, is a trimmer-option that works precisely like "Stick with no springs or FFB", however with working with FFB. It's really crucial. I was just testing it today, and the trimming is natural and just superb with option b), sadly no FFB though... Another part of the problem is that turning on FFB in DCS invalidates trim-option.

 

Can anyone from ED chime in on this?

 

Thanks in advance!

zerO

Edited June 19, 2021 by zerO_crash

[RealDCSpilot]

There is nothing wrong with FFB in DCS, it's the problem with no native support for Directinput from Brunner like i already told you in the other thread. It's really sad that Brunner doesn't implement it, people are asking them for years. Your only solution will be the Arduino "FFB translator".
The overcorrecting issue may come from using curves, if stick input isn't linear - FFB comes out of sync. Always use linear input settings without curves and with default saturation.

Edited June 20, 2021 by RealDCSpilot

[zerO_crash]

On 6/20/2021 at 7:16 AM, RealDCSpilot said:

There is nothing wrong with FFB in DCS, it's the problem with no native support for Directinput from Brunner like i already told you in the other thread. It's really sad that Brunner doesn't implement it, people are asking them for years. Your only solution will be the Arduino "FFB translator".
The overcorrecting issue may come from using curves, if stick input isn't linear - FFB comes out of sync. Always use linear input settings without curves and with default saturation.

 

Read my second reply. It has nothing to do with Brunner and it's software at this point. When it comes to all the effects not being modelled, that's one thing, but not this. This is really not a big issue at all. It's a matter of adding a new trimmer option to DCS that allows, very much like "Stick with no spring and FFB", to only give the AP updates, without the actual software trim function (Brunner will handle that, but still with FFB in comparison to "Stick without springs and FFB"). Therefore, it's really not a problem. And this is done to compensate for CLS2Sim's inability to trim the stick out of center with software trim.

 

As to overcorrection, I'm not sure why this issue even occurs at all TBH, as long as FFB is on. I have never used curves in DCS in my life. Real aircraft use linear controls, and thus I do too. I checked my controls time and again, and it's not a setting issue on my part. That unless I am missing something.

[heloguy]

Nevermind, didn’t read all of your post…

 

I was going to suggest your option B, which is what I used before I set up Chuls Arduino method.   Not sure why your Jetseat isn’t working with FFB off.  I never experienced that.  Mine always worked fine with Sim Shaker for Aviators on (maroon icon) and FFB off with hardware trim on.

Edited June 21, 2021 by heloguy

[zerO_crash]

5 hours ago, heloguy said:

Nevermind, didn’t read all of your post…

 

I was going to suggest your option B, which is what I used before I set up Chuls Arduino method.   Not sure why your Jetseat isn’t working with FFB off.  I never experienced that.  Mine always worked fine with Sim Shaker for Aviators on (maroon icon) and FFB off with hardware trim on.

 

You know, I haven't tried it yet, I kinda assumed that with FFB off, I will neither get G-effects in DCS (WWII, Hydraulic-controls, buffeting, G-forces, etc...) nor the Jetseat to work. From what you write heloguy, I assume it must work then. Damn, I'll try it out right away. I'll be in heaven if it works 

 

Not sure why, but I thought that if you turn off FFB in DCS, that it will not output the FFB-values in some kind of LUA-file. Gosh...

[zerO_crash]

Ok, I tried it. With:

 

- FFB off

- Trimmer mode - Stick without springs or FFB

- Bound DCS trim-button (Mi8/Mi24/Ka50) to the same that I have Hardware trim on in CLS2Sim (they are not interfering with each other, tested it)

 

I get no FFB-effects on my Brunner, which is a must with WWII and planes with hydraulic-controls. I surprisingly still get FFB with my Jetseat, which is nice. However, with "Stick without springs and FFB", I get overcorrection when trimming. The overcorrection is a surprise considering FFB is off and the trimmer mode selected. This just doesn't work.

 

I really hope that ED can come up with a solution to have FFB on, and allow a trim mode that only updates AP without the overcorrection. The Brunner is such a fantastic piece of hardware, professional hardware, that it's a pleasure and joy to fly with it. 

Edited June 22, 2021 by zerO_crash

[heloguy]

3 hours ago, zerO_crash said:

Ok, I tried it. With:

 

- FFB off

- Trimmer mode - Stick without springs or FFB

- Bound DCS trim-button (Mi8/Mi24/Ka50) to the same that I have Hardware trim on in CLS2Sim (they are not interfering with each other, tested it)

 

I get no FFB-effects on my Brunner, which is a must with WWII and planes with hydraulic-controls. I surprisingly still get FFB with my Jetseat, which is nice. However, with "Stick without springs and FFB", I get overcorrection when trimming. The overcorrection is a surprise considering FFB is off and the trimmer mode selected. This just doesn't work.

 

I really hope that ED can come up with a solution to have FFB on, and allow a trim mode that only updates AP without the overcorrection. The Brunner is such a fantastic piece of hardware, professional hardware, that it's a pleasure and joy to fly with it. 

 

Strange indeed.   Just did my first flight in the Hind using the Arduino/Brunner combo, and standard ffb trim, and I didn’t notice any issues.  I’ll try the method I suggested to you and report back.

[RealDCSpilot]

Just for info, DCS uses two different systems for haptics. FFB and telemetry output, independent from each other. FFB is meant for Directinput compatible joysticks, telemetry output is for motion seats, simpits, haptic mats (Jetseat, Buttkickers etc. via Simshaker for Aviators) - in general everything that is able to process that data. As far as i know, CLS2Sim only uses telemetry output. Telemetry output needs to be defined with via Export.lua in your Savegame/Scripts folder.

Edited June 22, 2021 by RealDCSpilot

[zerO_crash]

9 hours ago, heloguy said:

Strange indeed.   Just did my first flight in the Hind using the Arduino/Brunner combo, and standard ffb trim, and I didn’t notice any issues.  I’ll try the method I suggested to you and report back.

Ah gotcha brother, if it worked, I’d be in heaven. Sadly it requires the Arduino fix. My two gripes with that are: a) It’s not vanilla, little maintenance, but not standard. Not a fan of mods for aestethics and sometimes unexplained bugs. b) Some of the effects in the original FFB were overdone. As much as I like feeling a cannon shoot, it isn’t supposed to vibrate the stick. Sadly on e.g. Ka50 it’s overdone. 
 

I suspect that might be the difference in our results. You use Arduino fix heloguy, I don’t. But the overcorrection kicking in when not using FFB in DCS and “Stick without springs and ffb” just doesn’t make sense… Atm. I cannot use trim in DCS. It works so far, but it is far from perfect.

 

5 hours ago, RealDCSpilot said:

Just for info, DCS uses two different systems for haptics. FFB and telemetry output, independent from each other. FFB is meant for Directinput compatible joysticks, telemetry output is for motion seats, simpits, haptic mats (Jetseat, Buttkickers etc. via Simshaker for Aviators) - in general everything that is able to process that data. As far as i know, CLS2Sim only uses telemetry output. Telemetry output needs to be defined with via Export.lua in your Savegame/Scripts folder.

 

That’s interesting. But with that in mind, I don’t get why the FFB-effects would be gone if Brunner uses telemetry (which I know it does)? I tested WWII aircraft and L-39/Mig21Bis. No luck there. My Jetseat did work though.

 

CLS2Sim has to be used before in the mode of HID-device. Now it has a beta DCS-mode. It should support DCS in the basic way for sure. Much has changed, improved, with them and DCS-support over the past 2-3 years.

Edited June 22, 2021 by zerO_crash

[RealDCSpilot]

I find it also strange that DCS FFB on/off has an effect on how CLS2Sim reacts. It shouldn't be affected.
Directinput FFB has 2 settings for each aircraft: Force and effects.
1. Force is basically the same for each module (intensity goes from 0-100%). On jets you have stick positioning for neutral control surface position in the center or offset mode for control surface position (like in A-10C, F-14, F-5 etc.) on top of that trimming also affects stick position (DCS moves the stick around depending on your trimming inputs or autopilot features (Mig-21)). For WW2 birds you also have force intensity depending on air pressure on control surfaces (on ground stick goes limb, the faster you get the more force builds up and the stick moves to it's neutral position). In helicopters (only the 4 good ones) force is applied like magnetic trim or force trim would work like in real life.
2. Effects (intensity 0-100%) is used for secondary effects on the airframe, like runway rumble, stalling, gun rumble etc. - but here each module acts different. Some have more, some have a less number of effects.

Telemetry output is used for all kind of secondary effects, some run parallel to effects on the stick (stall for instance, VRS on helicopters, G-force intensity and direction, runway rumble and such). Telemetry can be more widely used for certain stuff but it doesn't hold any stick position data, it only holds information of what the aircraft is doing. One pretty cool effect is main rotor oscillations on helicopters for instance.
In my setup, as personal preference, i have set effects to 0% on the stick and get all the secondary effects via Simshaker on my Forcefeel seatpad (this way i'm disabling the stalling shakes on the stick, but that is my personal preference).

Edited June 22, 2021 by RealDCSpilot

[heloguy]

3 hours ago, zerO_crash said:

Ah gotcha brother, if it worked, I’d be in solved. Sadly it requires the Arduino fix. My two gripes with that are: a) It’s not vanilla, little maintenance, but not standard. Not a fan of mods for aestethics and sometimes unexplained bugs. b) Some of the effects in the original FFB were overdone. As much as I like feeling a cannon shoot, it isn’t supposed to vibrate the stick. Sadly on e.g. Ka50 it’s overdone. 
 

I suspect that might be the difference in our results. You use Arduino fix heloguy, I don’t. But the overcorrection kicking in when not using FFB in DCS and “Stick without springs and ffb” just doesn’t make sense… Atm. I cannot use trim in DCS. It works so far, but it is far from perfect.

 

That’s interesting. But with that in mind, I don’t get why the FFB-effects would be gone if Brunner uses telemetry (which I know it does)? I tested WWII aircraft and L-39/Mig21Bis. No luck there. My Jetseat did work though.

 

CLS2Sim has to be used before in the mode of HID-device. Now it has a beta DCS-mode. It should support DCS in the basic way for sure. Much has changed, improved, with them and DCS-support over the past 2-3 years.

 

I just tried flying around using this method:

 

- FFB off

- Trimmer mode - Stick without springs or FFB

- Bound DCS trim-button

 

I don't see any over-correction problems.  I can trim the helicopter however I want, and fly smoothly.  I don't use the DCS CLS2 plugin, as I have always had bad luck getting it to work.  With helicopters, the FFB trim is not working anyway, and you have to use hardware trim.

 

You definitely will not get anything but rudimentary FFB effects using Brunner's plugin.  Things like helicopter trim, and force trim in modern jets don't work because the interface between DCS and the plugin can't physically move the Brunner stick.  The only way you can get DCS to be able to move the Brunner is by using an Arduino.

 

I suggest you join the HOTAS Discord channel.  There is a Brunner section with a lot of knowledgeable people who can help you tune the stick to how you like.

 

https://discord.gg/ 

[zerO_crash]

I got my Arduino and went on with BrunnerDX. Sadly, I still get the issue with overcorrection. Both with FFB on and default trim, and FFB off and “Stick without springs and FFB”. In order to ensure what I mean, try to see if you have the behaviour from the picture.

 

I am really not sure what does that, but it is an issue, both with FFB on and off. Regardless of the trim settings.

 

@BIGNEWY Is this issue that I’m talking about, specifically for all helicopters in DCS, something that is reported?

 

I am also missing some FFB effects, like the cannon rumble. I have tried Ka50 and Su27 and had no such rumble. Otherwise over-G effects work, with buffeting and otherwise shaking due to atmospheric- and weather-effects.

 

 

[zerO_crash]

Well, I got it to work, but it is definately not working as it should. One just has to live with the bumps, when using click-trimmer. Those should not be there when trimming with a FFB-stick. I really hope someone from ED could just give me a straight answer whether the “bumps” and “overcorrecting” that I describe, ocurring with click-trimming, are supposed to be there? If not, then if a fix for this will come? I would also kindly ask to consider helping support the Brunner, beacuse this is a fantastic stick that shines really only in DCS, with these flight- and system-models.

Edited June 25, 2021 by zerO_crash

[heloguy]

I see what you're talking about now in your pictures.  I probably never noticed this because when I fly, I hold the trim release while flying maneuvers, and then release the button when the aircraft is at the attitude I would like to capture.  I don't push against the force gradient, and then click the trimmer, as I've always found that to induce errors in the stick position, even in the helicopters I fly now that utilize force trim IRL.

 

However, with the controls indicator above, it seems obvious there is no movement of the controls, but the aircraft attitude has changed, which doesn't seem right.

 

This seems similar, but less pronounced than what you see in the Ka-50 if you use the 'Stick without springs and FFB' option:

 

[REPORTED]Pitch Excursions while using Trim Release - Page 2 - Bugs and Problems - ED Forums (eagle.ru)