Logitech g940 'reversal bug' officially being worked on

[Dimebag1]

I thought I would post this here, as I think all g940 users should be aware of this.

 

The daunting 'Reversal bug' found on all axes of the g940 which basically makes the precision of the flight system useless by inducing a constant hunting for a desired input, has marred all it's users wanting precision from this setup, be it for precision gunning and bombing, or precision flying for formation purposes.

 

http://forums.logitech.com/t5/PC-Gaming/G940-quot-reversal-bug-quot-is-it-a-hardware-or-software-problem/td-p/445318/page/3

 

Charlesb on the logitech forums has confirmed that they have pooled technicians from other projects to work on a potential firmware fix for the reversal bug. He remains cautiously optimistic that they will solve the issue, and they will release more info on how it is progressing when Charles hears.

 

This is extremely good news for g940 users, as previously they have been rather silent on the issue, only once confirming the bugs existence and making no comment on the various reversal bug threads which have been around not long since it was released (over a year).

 

I am also cautiously optimistic, as they probably wouldn't release this information unless they were fairly sure they could squash it.

[Slayer]

Just like they were working on fixing the trim not releasing when you hold the button down or fixing the center slop in the FFB? eyeroll....

[Dimebag1]

Just like they were working on fixing the trim not releasing when you hold the button down or fixing the center slop in the FFB? eyeroll....

 

As far as I know the trim problem has been fixed in dcs bs and the slop is now negligable thanks to the last firmware update. I am holding out for this as there is a good chance they will crack this one. Other users have created c++ programs to remove the reversal bug successfully in il2, so it is possible, now we need a universal fix and it is looking like this is it. If you choose not to wait for this that's your choice. I understand your view, but this looks hopeful.

[Crash_Fistfight]

The "reversal bug fix team" is awsome news if they can fix it. Fingers and toes crossed for an end to the reversal bug. Once fixed a G940 may actually be a viable alternative to a CH setup but longevity still has to be tested.

 

As for the trimmer problem and slop in BS; ED with Logitechs assistance have been working on the problem and the miniscule centre slop in A10 Beta shows that the improvements have been made! This will of course be patched into BS when it recieves its compatibility patch so hooray for ED from the G940 community.

[mikey]

Seeing as I'm still having problems with FFB slop, I guess either my joystick sucks more or I'm just more critical, I doubt they'll be able to fix this problem fully as well.

 

 

And if logitech and ED were really working together, don't you think they would've made something already that moves up the deadzones and curves up to the newly trimmed position, so the stick isn't supersensitive when you've trimmed it?

[Crash_Fistfight]

Seeing as I'm still having problems with FFB slop, I guess either my joystick sucks more or I'm just more critical, I doubt they'll be able to fix this problem fully as well.

 

 

And if logitech and ED were really working together, don't you think they would've made something already that moves up the deadzones and curves up to the newly trimmed position, so the stick isn't supersensitive when you've trimmed it?

 

Are you playing the Beta? In FC2 the behaviour you describe is what i get but in the Beta the stick is solid and the response curve moves with the trimmed postion.

 

PS Have you installed the firmware update from 5.09?

[mikey]

I just tried for an hour to use the FFB implementation instead of relying only on the centering spring, which is how I have to play any ED game, and this is what I've found out:

- FFB slop is indeed less

- Deadzone and curves still don't move up. I setup 25% deadzones on the joystick's X and Y in the profiler, but those deadzones aren't there when I have trimmed into a new position (just the same as in BS and FC2).

 

Seeing as the deadzones and curves dont move up, I can't let go of my joystick so ED's FFB implementation is still not working for me.

 

I'm using the new firmware and Profiler 5.10 btw.

[power5]

What is this reversal bug? I have not had to reverse any axis in DCS series with my 940. What am i missing?

[Dimebag1]

What is this reversal bug? I have not had to reverse any axis in DCS series with my 940. What am i missing?

 

Ok, well read on at your own risk because ignorance really is bliss.

 

Firstly, go into the logitech profiler, then into the calibration section, of the throttles for instance. Now start slowly progressing the throttle forward 1% at a time. Aim to get 50% for instance. Now reverse direction of the throttle very slowly and try to get 49%; this is impossible, as you will see the input will pause even though you are moving the axis, until it finally moves about 3-4%, to about 47% aroundabout (amount varies person to person). So basically if you are aiming for a certain position and slightly pass it you can't reach it if you reverse the direction. This happends anywhere on ANY axis, including the stick and trim pots.

 

This makes it very difficult to rapidly find a correct throttle percentage when formation flying, or if you are trying to precicely gun down a target, your nose will constantly porpise up and down trying to hit paydirt.

 

Trimming works perfectly for me but I play fc1.12. Only 5% dead zone, and when I trim the aircraft the new stick position correctly responds to stick inputs, I have not noticed problems with the deadzone, but then maybe this is because it is so small. At any rate, a reversal bug fix would help the situation because this reduces the need for a deadzone.

[Crash_Fistfight]

I just tried for an hour to use the FFB implementation instead of relying only on the centering spring, which is how I have to play any ED game, and this is what I've found out:

- FFB slop is indeed less

- Deadzone and curves still don't move up. I setup 25% deadzones on the joystick's X and Y in the profiler, but those deadzones aren't there when I have trimmed into a new position (just the same as in BS and FC2).

 

Seeing as the deadzones and curves dont move up, I can't let go of my joystick so ED's FFB implementation is still not working for me.

 

I'm using the new firmware and Profiler 5.10 btw.

 

I've set my sensitivity in my Logitech profiler to 10% for the X and Y axes no deadzone. Then in the "axis tune" dialogue in game I've set deadzone of 3 and a curve of 20 for pitch and roll. I'd be interested what you think of the response using this setup. I maybe wrong but the deadzone and curve seem to follow the trimmed position maybe this is only true for deadzone's set within the game interface. I'm gonna go back and check the trim/response curve correlation with a really big deadzone ingame.

 

Edit: Your absolutely correct the response cuvre isn't following the trimmed position :cry:. I guess this might require moaning about but I don't know whom to moan at, sneaking suspicision it may be ED.

Edited November 3, 2010 by Crash_Fistfight

[mikey]

I dunno who to complain to either. A-10C is already feature frozen and logitech apparently only take serious problems seriously after 1 year. This is also why I don't believe that ED and logitech actually worked together at any point, since they would've definitely solved this problem since it affects all people with FFB joysticks.

 

Btw, there is a small chance it might get fixed in the open beta phase, since the open beta is only there to report hardware issues, and this might just fall into that...

[Slayer]

As far as I know the trim problem has been fixed in dcs bs and the slop is now negligable thanks to the last firmware update. I am holding out for this as there is a good chance they will crack this one. Other users have created c++ programs to remove the reversal bug successfully in il2, so it is possible, now we need a universal fix and it is looking like this is it. If you choose not to wait for this that's your choice. I understand your view, but this looks hopeful.

 

The trim issue isn't fixed for me there is still pressure on the stick when you hold in the trim button in DCS:BS are you saying it works like that on your setup? FFB slop is still bad and the only way to fix it is by adding a 10% deadzone on both axis.

[Dimebag1]

The trim issue isn't fixed for me there is still pressure on the stick when you hold in the trim button in DCS:BS are you saying it works like that on your setup? FFB slop is still bad and the only way to fix it is by adding a 10% deadzone on both axis.

 

I'm not saying that, because i don. Have bs, I'm only going off post I've read about the progress of trimming in dcs bs. Some people claim that it works for them. Now trimming in fc1.12 works this way; when I trim the aircraft all forces are removed from the new stick position. There is still a small centre slop area for me, but much improved from the update.

[Dimebag1]

New update from Charles: http://forums.logitech.com/t5/PC-Gaming/G940-quot-reversal-bug-quot-is-it-a-hardware-or-software-problem/m-p/519745#M6692

 

highlights are:

 

- Reversal bug can be fully removed, but engineer left slight amount to counteract ffb jerkiness

- will be available for testing for US users, contact Charles via the thread.

 

Sounds pretty good.

[mikey]

I guess we can all be happy for the transparency on this issue, but after 1 year of hearing "the next patch is gonna be the bomb!" I think it's safest to remain relatively skeptical. Also, am I wrong to believe that this reversal bug was actually intentional by the initial g940 developers seeing as they're still gonna keep a bit of it?

 

It's a shame I don't live in the US and have Win-7 x64 though...

Edited November 20, 2010 by mikey

[Dimebag1]

I guess we can all be happy for the transparency on this issue, but after 1 year of hearing "the next patch is gonna be the bomb!" I think it's safest to remain relatively skeptical. Also, am I wrong to believe that this reversal bug was actually intentional by the initial g940 developers seeing as they're still gonna keep a bit of it?

 

It's a shame I don't live in the US and have Win-7 x64 though...

 

You would be right in that assertion that it was an intentional inclusion. What I don't get is why it was even put in on non ffb axes. I completely understand your pessimism about the next software release. I'm not saying things will be significantly better, but to me on the reversal bug front we should be in luck it seems. I agree that the software itself is lacking as far as programmability, and I would still like to see native led programming, but well just have to wait and see.

[Redbow]

They released a firmware update yesterday. Says to drasticly reduce the reversal-bug.

 

http://forums.logitech.com/t5/PC-Gaming/G940-firmware-1-42-is-now-available/td-p/542496

[mikey]

Yup, it reduces the reversal bug for the joystick's X and Y axis by a lot. I just hope they're gonna "fix" all the other axes as well because it's still a broken product, although now with the new firmware a bit less.