Hej, just purchased LFS with S2 pass. I have a weird bug. First off the in car view starts pointing 45' degrees to the left, which i can adjust with the slider to be straight on. But the problem is really that forcefeedback is pulling the wheel to left. There is a pronounced notch in the center but with the wheel still wanting to pull left. If I let go it pulls left violently until it hits the lock. It does this all the time and unrelated to what the car is doing.

Nothing in settings seems to help and the wheel doesn't do this in other sims.

Wheel is VRS Directforce Pro (small mige).

Assistance greatly appreciated!

Update

The view pulled to the left has been an effect of game binding look to an nonexisting wheel button. When unbinding view returns to center.

Force feedback is still wrong:
-no FF when stationary -ok
-when pulling out there is a very pronounced notch in the center
-steering right feels like going against a strong rubber band, though car steers normally in game
-steering left immediately "unhooks" from the notch in the center and violently pulls/spins left and smashes (and keeps smashing) into the left steer-stop

My guess is that FF "thinks" center of the wheel is the right bump-stop and the left bump-stop is center, as if the whole range is shifted half the range to left. The steering input works as intended and matches that of the in-car view, just wrong FF.

Look here:

https://www.lfs.net/forum/thread/95152-Normal-FFB-with-wheel-turning-right-but-inverted-FFB-turning-left

Hej Lucas, yes this is it, thanks! Now the question is what to do about it?

You can see that "true" values range from 0-65000 with center at 32750 while raw data shows -32000 to 32000 with center at 0. I dont really know how to calibrate in DIView, seems to produce wierd resoults whatever I try.

You can recalibrate the axis in the DXView2 only. I would still try to check it with WheelCheck, put some spring force, and see if that works normally. If not, then it has to do something with the firmware of the wheel, which you can do nothing about. And about the wheel being reported by Windows as joystick or wheel, does not matter for LFS. If you can not find it, pm me, I can send you the exe file.

Did you try to set in LFS (patch V) options/controls/axis FF steps to 10000 and FF rate to 100Hz, turn of dead zone?

Hej Rane, hvala na pomoci brate. I just did a fresh instal yesterday, version 0.6V. Tried FF steps 10000 and FF rate to 100Hz, turn off deadzone off and on - makes no difference.. Turning right its a strong spring, turning left even a bit spins the wheel strongly to left and keeps smashing into bump stop.

I've looked a bit at wheelcheck, but to be fair i dont really understand it.

If i set any spring option in Wheelcheck (simple spring, force spring..) i just lose FF i LFS and instead get the selected effect.

Once, when I had a problem with MOMO Racing Force, for example, when the brake was released, it was still braking a bit or the gas pedal was not pressed "all the way", I used DXTweak2 to fix it. You could also change the linearity of the axis operation. Unfortunately, now I cannot attach the steering wheel, so I will not tell you exactly what and how. Enclosed is DXTweak2 in x86 and x64 versions. Good luck



PS
One more thing came back to me. I do not know about your steering wheel, but MOMO Racing Force had "reverse FB" you need to turn on the "inverted" option otherwise the steering wheel would repel from the inside.

PS2
try use Clear Calibration Utility

PS3
In LFS correct -1000 +1000 - try not center parameter e.g. -1500 -100

Thanks Lucas, I'll give it a go. Hopefully I dont mess things up

At this point, I'm afraid that you are only left with some diagnostics tools. Most probably, the fix needs to be in the firmware of the device itself. More accurately PID (physical interface device) part of the HID (human interface device) descriptor.

WheelCheck is really a great program. It was made by guys from iRacing. It can read device HID definitions, as well as PID - FFB effect definitions. You can basically see everything that is in the firmware of the device regarding FFB. Additionally, you can apply some FFB effects like a constant force to test your device. The only FFB effect LFS uses is constant force - it sends FF rate times per second a constant value of force to the wheel in the range of 0 to FF steps, giving the illusion of time-dependent force. How these values are then scaled to the signal which is sent to the motors is another part. Here you can only see the logical or software side of things, done by Microsoft in 1999.

Have a look at some of my print screens for where to look for important FFB things.

I see.. mine differ quite a bit? Can you see from my data what could be causing described behavior in lfs? Is there any way to really fix this or work around it, or is our only hope to plead VRS team to implement some changes in the FW?

In principle, everything looks good from the PID side I would say. There is a limit anyway in the granularity of the DirectX input of 10000. That is why instead of full 16bit values (+-32k, or 65k) they have put 10000 instead, that is all fine. Logical and physical values are reported and defined correctly. As I mentioned before, there is still a lot of processing done with those values before they get sent out to motor as some PWM driving signals. But it is fishy how other games run fine. This is very interesting since I did have kind of similar problems while developing my own firmware for an Arduino based FFB wheel. Without WheelCheck I would have not been able to scale effects and calculate forces correctly, it was a great help for troubleshooting FFB issues.

Have you tried playing with the wheel with force (PID) effect? Does it work correctly or you have it the same as in LFS? Try also those effects at the bottom, inertia, friction or damper. There is an offset for these effects, this will shift the X-axis range, but it is only for the WheelCheck program, see if that makes any difference. It might at least give us an idea of where the problem is. Before pinpointing the issue, I can not say for sure if the problem is in their firmware or elsewhere.

Maybe we can get on discord some time and try to do some tests in WheelCheck together.

Thanks, I'll run some tests today and report back. The other 2 games i tested, iRacing and AC work correctly out of the box. iRacing has very natural feeling FF out of the box with linear force option in iR checked. Assetto Corsa works correct but not that great feeling - too springy, hopefully that can get dialed in.

Talked to VRS team about LFS, they admitted they havent tested it, but got a similar report from another user (maybe the American lad from this forum). LFS is not on their priority list, but they said they will try to look at it. Maybe we can give them some good input on that so they can sort it out. I really feel LFS is a great and unique platform and would love to see it supported going forward.

You can add me on discord oldfashioned1912#9403