Subscribe to download
[1x2] Ackermann Steering

Flipping still happes, only on the rearmost wheels and only when steering to the right, at ~0.4 steering input:

https://steamhost.cn/steamcommunity_com/sharedfiles/filedetails/?id=2215429636

On certain configurations the flipping happens even earlier, with the wheels ~-20 blocks for the y coordinate.

Yeah I don't see much point in using an inverter when I can just place the wheels right.

The controller here is updated btw. so the flipping around shouldn't happen anymore.

Guess I solved it, in your vehicle example you're using the old wheels which don't need to invert the steering axis for the left hand wheels, this got me confused so I assumed the formula took care of inverting the axis on its own, depending on sign of the x input.

Simply add an inverter for the left hand wheels, solved it.
Rear wheel axle still flipping around at full inputs.

The example vehicle is setup correctly but I see what you mean with the rear wheels flipping around.

That is actually caused by the steering limitation the devs put in, I'll fix it.
Aside from that though, I see no issue there.

I'm confused about how to set this up.
In your other workshop upload (the one with the vehicle) it says:

X position of the wheel relative to the desired center of rotation (+ right, - left)
Center of rotation? Your comment recommends the center of mass, but the results from using COM were weird.

So I assumed you meant a point outside of the vehicle, an extended axis from the COM to the vehicles side, when that desired center of rotation is 9 blocks left, 3 block behind the wheel I enter 9 for x and -3 for y?

If I set it that way I still have to invert the signal going to the left wheels to make the result somewhat symmetrical, while the steering seems to be off to one side, having a wider radius than when steering in the opposite direction.

Even with your default workshop example vehicle the steering seems wonky at best with rear wheels flipping at opposite sides with full steering to either side

thank you for this

Makes no difference really.

After doing some testing, I believe the devs clamped steering angle to one radian or some approximation of it in each direction. I'm not sure if this affects your formula's calculations.

That looks beautiful. Another microcontroller is suggesting 70 degrees, ,and I'm not entirely sure if or how his formula may be different from yours.

It was actually almost unlimited for a while, thats how this thing came to be: https://steamhost.cn/steamcommunity_com/sharedfiles/filedetails/?id=1478976292

Not sure how much it is now but that should be easy to test.