I have noticed a problem with LFS, whenever I have force-feedback enabled on the wheel, any car I drive begins to 'wobble', for lack of a better word, when driving on straits and under braking.
When I disable the force feedback on the wheel (my wheel has a button for it, Microsoft Sidewinder Force Feedback Wheel) these problems do not appear.

I've attached a replay showing this. Down the straight for the first time I let go of the wheel once I've placed the car correctly on the straight and don't touch it until I reset the car after crashing. After this I drive back on the straight and this time without FF, only turning the wheel to keep the car on the road. After this I turn around once again (after a nice flight) and go down the straight without the FF on and once I've gotten the speed up I turn on the FF and begin to brake, the car begins to wobble. Go back up the straight with FF off and brake again (without FF) and this time the car is stable.

I can feel this in the wheel with FF on and makes the car feel unstable and effects my laptimes. I can't fully trust the car with those problems, nor can I without FF (steeringwheel is to light).

Anyone know the cause of this problem?

no, but i get the same thing on my Momo

Afaik this is caused by the shortcomings of the currently available FF hardware. Either turn down the FF strength or simply don't remove your hands from the wheel (why would you want to do that anyway - while driving).

The steering in real cars naturally tends to self-center due to forces acting on the wheels. In LFS, when using a force feedback wheel these self-centering forces move the steering wheel back towards its center position. However, for some reason (and AndroidXP may be correct that it is a limitation of the technology), the wheel slightly overshoots the center and ends up steering slightly in the opposite direction. Self-centering forces then cause the wheel to move back towards the center in the other direction which results in another overshoot and so on. The wheel ends up oscillating around the center steering position.

These self-centering forces can be affected by such things such as toe-in, camber and especially caster.

Check the center spring force in windows game controllers is switched off and lower the main force strenth here too , in LFS use < > keys to reduce/increase force strenth until it feels better .
High camber FWD sets used to cause this for me until I realised I didnt need the forces so strong to still get a feel for the road .


Good luck

SD.

this happens to me also on my formula foce gp but i like it cuz it makes the car seem more real...(ex.braking hard rear end slides and i gotta correct...does this while im on blgp on gti...i love the feeling :wow

Check you FF settings in control panel. I have a dfp set to 720degree lock and 100% in CP and LFS and can let go of the wheel and not have it spaz out on me. I also have centre spring disabled.

Also look at your setup. I drive race_s or similar and these are stable. If you run high cambre & KPI you might find this happens as the car will not sit flat on the road and WILL pull on the slightest bump.

My old Saitek R440 had this problem though.

Yup, center spring is a no-no. And lower the in-game FF strength till the wheel is stable as said. To much FF makes the wheel feel "jaggy" when turning it slowly anyway

also up the FF steps in your cfg file from 128 to 256. This specifies the granularity of the FF, higher is better and 256 is max

256 steps is maximum resolution?
wow, i have no idea about the specs of the material but in model-builder scene even the cheapest digital steering instruments (for controlling planes/helicopters) use more than 512 or even up to 1024 steps.
ok, there are just a few really good-usable controllers anyway, but these are far more precise...
and even some servos use more than 512 steps to tune...

makes me think why noone complains about that.
and the new (a few months ago) wheel does anybody know about it's resolution?

regards

Quote from micha1980de :

256 steps is maximum resolution?
wow, i have no idea about the specs of the material but in model-builder scene even the cheapest digital steering instruments (for controlling planes/helicopters) use more than 512 or even up to 1024 steps.
ok, there are just a few really good-usable controllers anyway, but these are far more precise...
and even some servos use more than 512 steps to tune...

makes me think why noone complains about that.
and the new (a few months ago) wheel does anybody know about it's resolution?

regards

the 256 is the FF steps not the resolution that LFS accepts from the controller for input.

256 does feel better than the default 128, although not sure if you feel much difference unless you run a DFP tbh.

No, the resolution is of the steering axis (also pedals) is usually 0-65535, with the underlying hardware defining how small the steps are, in reality you get maybe 1/3rd of that, so you have about 20000 different steering states.

This however, is NOT what the FF steps are. The FF in LFS is not sent 1:1 to the wheel, because it would likely clog the buffers with too many instructions. Instead, the FF force is seperated in steps, and each time the calculated FF changed more then the size of one step, the new force is sent to the wheel. With a higher setting of those FF steps, each step is smaller, so even smaller force changes get sent to the wheel, increasing the resolution and making it a bit more detailed (atleast in theory).

This essentially means you have 256 different FF-forces and I'm not sure we even have good enough hardware so you could notice that.

Might sound silly but I'm not convinced the FF steps do anything on a DFP at least (and prolly all FF wheels)..
Force comes in big lumps, I doubt the wheels have as much as 16 steps of actual force. The static friction in a DFP is considerable so before the force has overcome the friction, you're certainly more than 1/128 or 1/256 of force away from the previous experienced force. 1/16 could well be too optimistic as well..

I think you're a bit pessimistic here. Also the FF steps are not really related to the hardware, but a LFS internal thing. If you (could) set them to 16, then you couldn't feel any fine FF details anymore, like when driving over the kerbs, because the change in force would be considered too small to send to the wheel at all. I'm not sure how many different forces the wheel hardware is able to produce, but my guess is a fair bit more than your 16.

Quote from Niels Heusinkveld :

Might sound silly but I'm not convinced the FF steps do anything on a DFP at least (and prolly all FF wheels)..
Force comes in big lumps, I doubt the wheels have as much as 16 steps of actual force. The static friction in a DFP is considerable so before the force has overcome the friction, you're certainly more than 1/128 or 1/256 of force away from the previous experienced force. 1/16 could well be too optimistic as well..

The static friction on my DFP is near 0 ? I can go lock to lock using the tip of my little finger to move the wheel. It is only when the FF is active that there is any resistance. It is only the FF motor that gives the resistance.

Also the steps, as already pointed out are to control updates to your wheel. If you set it to 16 steps you would only get updates to FF when the force crosses 1/16 of the range for FF strength. With it set to 256 you get more updates sent as smaller changes cause a FF update.

Its a bit like fps in that the faster the fps the slower everything feels as the transistion between frames is smaller. It does the same with the FF in that it feels smoother and more subtle.

What I found odd was.. I think you can test this by using the centring spring and just try without a game. The force required to go to a certain angle, and then I can let go a lot of my hand force while the wheel remains in the same position.. It seems only when some sort of friction is overcome does the wheel turn back.

With a 'real' spring, the wheel will move towards the center as soon as you let go. OF coruse you shouldn't use the centering spring for sims when you have an FF wheel but it does highlight issues in tests like this.

I think I've (well, Astro [ BJRL ]s, idea actually) found a way to overcome the problem, at least slightly, increase the range of the wheel turn. My wheel has about 270 degrees of wheel turn, and in LFS I've set it to about 400. Though I can still feel the wheel turning, it's no way close to what it was before.

Quote from Niels Heusinkveld :

What I found odd was.. I think you can test this by using the centring spring and just try without a game. The force required to go to a certain angle, and then I can let go a lot of my hand force while the wheel remains in the same position.. It seems only when some sort of friction is overcome does the wheel turn back.

With a 'real' spring, the wheel will move towards the center as soon as you let go. OF coruse you shouldn't use the centering spring for sims when you have an FF wheel but it does highlight issues in tests like this.

In LFS centre spring should be OFF. LFS creates the force feedback based on the forces acting on the front wheels. Centre spring is a constant canned effect and should be disabled so not to effect LFS FF.

Yes centre spring might be wrong but who cares, so are all the other effects wheels can play. They are not used.

And IRL if you let go of the wheel it does not spring to the centre. You have to be moving as it is the forces effecting the front wheel that cause the wheel to centre, not some magical spring.

^ You totally didn't get his point

maybe you should try another setup.
normally you have a negativ value for the front tread width. that causes an unstable handling on the front axes. you could also try to make the antirollbars harder and the dampers harder. that should give the result you want.. but thats nonsense for racing.

Quote from AndroidXP :

^ You totally didn't get his point

Not really. Centre spring is a canned effect. So using it to test anything means nothing. I could say then when I move the dpad on my DFP in CP to trigger canned effects I can test stuff with those effects, they still mean nothing.

We don't know how the spring force is applied so without that information any observation is just a guess, it might be programmed to act that way. Or did I miss something else?

Well okay, but also a "canned effect" does nothing else but give the FF motor some instructions. The only question that remains is if the centering spring "effect" is set up to be a constant or a gradually increasing force to the center.

I thought I would look here again to see if things have been sorted regarding FF but it looks as if it hasn't.

I should say I love the way it's done in GTR,GTL and NkPro and that will upset many here.
I know the forces in LFS are a physics simulation and are supposed to feel more real and it dosn't rely on canned effects.
I've also posted here and elsewhere before and people have tried to help but no amount of tweaking fixes the problem which is this:-

At 20-30 FF strength the wheel resistance is fine for me except in the center of the wheel where about the first 15% to 20% of the wheel has very little resistance.
If I increase the FF strength to 60 I can feel some resistance be it a small amount however after the wheel has turned more than 15%-20% the resistance increases way to much for me.
If I set ithe FF to 100 the center is perfect but and the resistance increases from the center and not 15%-20% but again it is way too strong.
I know some will say thats how it is in a real car, fine but this is a game and if you had RSI like I do you would find that it is unplayable.
I do like a light wheel and you carn't do that in LFS and get a a positive feel for the but.

Also I find having so much force trying to center the wheel at high settings causes the wheel to overshoot the center which make the car much harder to control.
Turning of FF in LFS makes it a much easier ride but I carn't live without FF as the sim feels one dimensionl to me.
It dosn't just happen on this pc or with this wheel I've tried others with the same result, I carn't find a setting to reduce the strength of the wheel or at least keep it even the more you turn it.
In other sim I've seen settings where you can set the strength curve but most don't seem to increse in strength or if they do it's much less than LFS.

The bottom line is I carn't play LFS and I've been trying since it was released and have tried every tweak I've seen posted.

May be your wheel - a slightly miscalibrated wheel may lead to these issues. You could try DX Tweak (not sure it works with a Side Winder) to calibrate your wheel more acurately. Also use some steer center reduction (it is under controls) to make the wheel less responsive near the center position. I run mine at 0.75 IIRC. I used to have these issues with my old FF GP. Mind you, I didn't know how to set it up at the time either.

Hi Folks

Since using a Momo or DFP i have always had a problem of wheel wobble and i am not sure what it is but i think it could be a bug in LFS

Try this. Set the BFI off as A1 on Aston National using your setup and watch it in cockpit view.

It does it on all cars but the BF1 shows it the most.

On my system the wheel goes mad in certain places. In this case its LFS thats driving and not me.

If someone can confirm this, if they cannot i will put up a replay.

Thanks