reported Problem with slew axis in game using Thrustmaster Warthog throttle

[Hellbat]

Hi,

I am testing a brand new Warthog throttle and noticed something very peculiar with the slew control.  When I slew in an arbitrary x/y direction and I then slide the stick such that the sign of the vertical input changes, it ignores the left-right command which is being held.  When I play with the inputs in game on the controls screen it all looks fine, but it doesn't work in the cockpit while flying.

I performed a slow repair, no mods are installed and I demonstrate this behavior in an instant action mission.  I start by slewing the TGP cursor left and up/down and as I alternate between up and down you can see the cursor freezes unexpectedly.  I would expect that the cursor should not stop moving to the left or right as it is being held.

warthogThrottleIssue.trk

Edited September 27, 2022 by Hellbat

[Hellbat]

To add to this, I just tried the same test with my Cougar throttle and I can reproduce the same behavior; the Warthog throttle is fine, but this seems to be a limitation in the game how the axis commands are interpreted when the sign of one of the inputs changes.

[BIGNEWY]

Hi, 

thanks for the PM, 

I have tried to reproduce but it seems ok for me, unless I am misunderstanding the issue.

I will ask others to try also 

 

 

[Hellbat]

Hmm, does that mean that watching my track replay doesn't show the unexpected stop in left/right motion?  I watched it back on my system without any input devices attached and I saw the behavior.

The easiest way to reproduce on my system is to select TGP as SOI in A10C II, hold the slew ministick as far to the left as it goes.  The TGP will start slewing to the left.  Now, maintain maximum displacement on the slew ministick and move it up and down around the 3 o'clock (or 9 o'clock) position (depending on how you look at your sensor).  When you first slewed left the vertical sensor probably had a small nonzero value which resulted in a very slight up or down trend.  As soon as the vertical component changes sign from this initial input, I lose the horizontal motion and the TGP only slews up and down.  Horizontal motion only starts again if I release the ministick a little and then it seems the game gets updated with the new sensor value.  I originally had the autodetected TM Warthog settings, although playing with saturation and curves didn't resolve the issue.  It might be that throttle setups with older analog sensors might have enough noise from the potentiometers so that there are constant perturbations on the sensor values which masks the behavior described above.

[jaylw314]

I have noticed this using stick control on a TWCS, WinWing Orion, and the ministick on a VKB Gladiator.  It's most noticeable when you move the stick clockwise/counterclockwise at its limits.  Occasionally, when moving around one corner, it will fail to register the next axis.  So when moving right->right-aft->aft, the TGP goes right->right-aft->nothing until the stick is re-centered.  I've tried this with curves, desaturation and cutting off the corners in joysticks that support this, to no avail.  The TWCS uses pots, while the Orion and Gladiator have digital sensors, but the behavior appears identical

Edited September 27, 2022 by jaylw314

[Hellbat]

Hi,

I brought this issue up 9 months ago while testing the Warthog HOTAS, yet I don't think it was really considered a bug.  It has been observed on multiple hardware setups.

 

I recently decided to test the module again and it persists.  I believe this is a software issue and I am reporting it as such.

The issue only occurs when HOTAS slew is controlled with an axis; for those who use discrete buttons, you wouldn't notice this.  If I slew to the limits in left/right and make small adjustments up and down (around the 3 o'clock or 9 o'clock positions), it appears to ignore one of the axes of my controller as it crosses into the new quadrant.  I have attached a track file where I first demonstrate behaviour with the keyboard and then I do the same thing with the axis and you see the slewing just randomly stops, but I am holding my joystick at its limit.  It only recovers once I reset the mini joystick closer to the center.

I tried modifying the calibration of my hardware so that when I pull the axes to their physical limits it only registers as 96% in game.  This alleviated the problem when slewing left, but not right.  This suggests to me that it is in fact a software issue.  I'm not sure what resolution the input signals are converted to in the game world but I think this bug should be simple to find and correct and I don't think it is worth it sacrificing 10%+ of the resolution of my physical sensor to try and get a workaround.

warthogThrottleIssue.trk

[Hellbat]

I'm not sure how the decision was made so quickly that this is a hardware limitation without checking the track file.  My ministick has 10 bit magnetoresistive sensors and I don't experience issues in other modules.  The same result was obtained on a brand new Warthog HOTAS in September 2022 and while Thrustmaster stock sensors aren't the highest fidelity, if these are considered low quality, then I'm not sure what hardware passes as good quality?

 

Edit: I just mapped my Gunfighter Mk3 base to the TGP slew control.  Same behaviour!

Edited July 9, 2023 by Hellbat

[Yurgon]

Seeing the same on my TM Warthog with DeltaSimElectronics slew mod.

When slewing with maximum deflection (for example to the right) and then adding up or down motions to the mix, the TGP stops slewing to the right while the ministick is still in the rightmost position. As I move the ministick out of the rightmost position, slewing to the right eventually resumes when the ministick reaches circa 90% of maximum deflection. The same applies to all slew directions at maximum deflection.

Checking the ministick in DCS' Adjust Controls -> Axis Assign overlay, the ministick output looks perfectly fine, all deflections reflect exactly the inputs given.

Apparently, when one axis is in or near maximum, it can get blocked by movement of the other axes.

Is this maybe a workaround dating back to the days of Thrustmaster's sub-standard slew control?

Edit: Just tested, in the Viper this problem doesn't show for me with the same hardware.

Edited July 10, 2023 by Yurgon

[BIGNEWY]

threads merged, I will take another look, but as mentioned before I was not seeing any issue with my hardware. 

I will try to get others to reproduce again.

@Yurgon if you still see the issue please make a CBT report. 

thank you 

[Yurgon]

15 hours ago, BIGNEWY said:

 if you still see the issue please make a CBT report. 

Done.

[Radboy16]

Has there been any update on this? It's getting a bit annoying using the TGP because I can't slew around consistently. The TGP consistently just forgets i'm slewing one axis as soon as I try to do any mixture of X and Y axes. I think it happens whenever you center one of the axes? Not sure.  

I've gone in and checked in DCS control setup, and the axes work completely fine when I spin in a circle and everything. Same goes for in windows and other programs. It's just when slewing things around in the a-10c ii.

I have a video clearly showing this behavior. Tried editing it to be easy to see the issue, and overlayed it with a live recording of the joystick inputs at the same time.

If you look closely, as soon as one axis crosses the neutral position, the other axis also completely stops, and won't respond to input until that one then crosses the centerline.

 

Edited October 22, 2023 by Radboy16

[doright]

I was able to replicate this problem by assign two independent axes to the slew control. For example the Warthog controller throttle friction axis to the A-10C II vertical slew input, and the Warthog right throttle axis to the A-10CII horizontal slew input. This way I could be absolutely sure I wasn't blending controller inputs on accident. As Radboy16 pointed out the bug is that when both axes are deflected and then one input axis is moved to the zero position the game zeros both slew rates. If an input axis crosses zero the slew is zero'd for both directions, but the movement of that input past zero restarts the slew, however only in that one axis' direction.  An axis doesn't have to be at the maximum range for this to happen. This occurs on both TGP and Mav slewing. Hud cursor and HMDS cursor work correctly (although the circular cursor limit of the Hud and HMDS might make it seem like the bug is happening there too.)

Any tiny movement of the input axis which the in game slew has been incorrectly zero'd and the game starts slewing again in that direction at a rate appropriate to the axis' displacement. This is why the bug becomes very apparent when controller is held against the limit in one direction, there is no small change to the input value to get the in game movement to re-engage.

Hope this helps.

Open Beta
DCS/2.9.1.48335 (x86_64; Windows NT 10.0.22621)
Thrustmaster Warthog controllers, custom script profile

[Hellbat]

Thanks @Radboy16 and @doright for adding your experiences of this issue.

@BIGNEWY can you provide an update?  Has this been reproduced internally?

[BIGNEWY]

Hi, 

a team member has reproduced the issue and will report it. 

thank you