1. This site uses cookies. By continuing to use this site, you are agreeing to our use of cookies. Learn More.
Dismiss Notice
Like RaceDepartment on Facebook.

The Project CARS FFB guide thread

Discussion in 'Project CARS' started by ermo, May 10, 2015.

  1. Status: WiP

    Just a quick explanation of what is going on with the FFB while we're waiting for Andrew 'AJ' Weber, the resident pCARS STM and FFB specialist to finish his FFB guide, which I'll be sure to attach here once it is available.

    Introduction

    The FFB UI design currently in the game was pushed through relatively late in the development cycle, just to ensure that all the knobs were exposed in the UI for the initial console release (on PC, it was possible to configure the FFB via text files during development). The current set of knobs represent v4 of the FFB topology, which was developed after AJ took a long hard look at how various FFB wheels on the market (mis)behave. A subset of the knobs, for example, were added to compensate for consumer FFB wheels not having an entirely linear response to FFB signal inputs.

    Physics-related FFB knobs


    First of all, you have to understand the FFB subsystem in pCARS as a set of 'sources' or 'channels' which can be adjusted individually to both change and augment the FFB signal.

    Definitions

    Fx = longitudinal force component (braking)
    Fy = lateral force component (sideways force when cornering)
    Fz = vertical force component (weight transfer, bumps)
    Mz = torque around the vertical (Z) axis due to e.g. caster, King Pin Inclination, scrub radius etc.

    The numbers represent percent, so 100 = 100% = 1.0

    Essentially, the FFB starts where the forces acting on the contact patch through the suspension geometry reach the steering rack. If you leave the relative Fx/Fy/Fz/Mz at 100/100/100/100 respectively, you effectively haven't transformed the FFB matrix at all and you get EXACTLY what is transferred through the suspension geometry from the contact patch. You can consider this the 'pure' FFB signal.

    The Spindle Master Scale simply scales the Fx/Fy/Fz/Mz transformation matrix by a scalar (0 - 200 or 0.0 - 2.0) on a per car basis. If you leave it at 100 (1.0), no scaling is introduced since 1.0 is the multiplicative identy. To give an example, having Fx = Fy = Fz = Mz = 50 and Master Spindle Scale = 200 is the same as having Fx = Fy = Fz = Mz = 100 and Master Spindle Scale = 100. Please take a moment to reflect on that.

    Splitting up Fx/Fy/Fz/Mz like that is useful when you want to make a 'recipe' where you favour one 'flavour' of the FFB over the others. If you want to emphasize the torque difference between the front wheels under braking, you would need to add more parts of the longitudinal component Fx compared to the other parts.

    Fx examples:

    a) Fx = 200, Fy = 100, Fz = 100, Mz = 100, Spindle Master Scale = 100
    b) Fx = 100, Fy = 50, Fz = 50, Mz = 50, Spindle Master Scale = 200

    Mathematically speaking a) and b) represent the exact same thing and will result in the exact same FFB mixing in terms of relative and overall output. This is because (as hinted at above) the Spindle Master Scale (Sms) is distributive such that Sms * (Fx, Fy, Fz, Mz) = (Sms * Fx, Sms * Fy, Sms * Fz, Sms * Mz). Again, take a moment to reflect on that.

    Fy, Fz, and Mz all behave in the same way from a mathematical perspective.

    It is worth noting that Tire Force is the global version of Spindle Master Scale. The global Tire Force default is 100, which again means 100% or simply 1.0. The reason there is both a global and a per car knob that do the exact same kind of scaling, is that different cars produce different forces as measured at the steering rack. So if you like the quality of the FFB overall, but think the wheel jerks around a bit too much in all the cars, try dialling down Tire Force. If you like the quality of the FFB overall, but you think one or two of the cars have too strong FFB, leave the global Tire Force at 100 and tweak the Spindle Master Scale for those two cars.

    In mathematical terms it looks like this:

    Tire Force * Spindle Master Scale * (Fx, Fy, Fz, Mz)

    Wheel-related FFB knobs

    Many of the other FFB knobs in Project CARS are there to deal with the fact that on consumer FFB wheels, the dynamic range (torque capability) of the FFB motors is much, much smaller than the actual torque acting on the simulated steering column. So the FFB signal will generally need to be massaged or 'mixed' to avoid simply losing the parts of the signal with torque values above what the FFB wheel can produce.

    When we talk about 'clipping' we're talking about sending torque values from the pCARS FFB subsystem that are beyond the amount of torque that the FFB wheel can produce. This is one of the reasons why it is generally a good idea to leave the in-game FFB knob at 100 -- it ensures that your FFB wheel is configured to deliver the maximum possible amount of torque into your hands.

    Using the Logitech G27 as an example, it can be shown that the torque motors do not necessarily respond to the weak-ish torque signal that are produced by the pCARS FFB subsystem around Top Dead Center (this is physically correct b.t.w.). Simracers have gotten around this by configuring the G27 profiler to an Overall Strength of 107%, but in pCARS a different knob for achieving the same thing was introduced: Deadzone Range Removal.

    You can think of DRR as essentially raising the floor (the lowest values) of the FFB signal to the point where the FFB motors react to them. This will tend to give a firmer feel around TDC on the G27. Better, more expensive wheels will typically not need as large a value and e.g. the CSW v2 might do just fine with a DRR of 0.

    So: If you own a CSW v2 and you keep the FFB at 100 and think it is too violent, I would suggest you adjust the global Tire Force down to a lower number to avoid losing out on the dynamic range your CSW v2 is capable of producing.

    (more to follow)
     
    Last edited: May 10, 2015
    • Beer Beer x 6
    • Winner Winner x 6
    • Like Like x 5
  2. (reserved)
     
  3. (reserved)
     
  4. Thank you for this thread. This 1st post already helped me a lot to find G27 set up to my taste and also get rid of annoying clipping.
    Actually I realized there weren't many things to do :
    - I decreased tyre force to 90
    - DDR to 0.1
    - Smoothers to 0 in car set up
    - Introduced slight damper

    I also use FFB file from Jack Spade.
    And I think that's almost all.
     
    Last edited: May 11, 2015
    • Like Like x 1
  5. Lazarou

    Lazarou
    Premium Member

    Great guide ermo.
     
    • Like Like x 1
  6. thanks ermo :) your guide is amazing. I managed to set up my wheel on PS4 the way i like and the game is completely different.
     
    • Like Like x 1
  7. Hi Ermo,

    i did a lot of research in the official pCars forum and around the web, but i couldn't find an answer to a question that puzzles me for long. When i stummbled across your post here, hope occured to me you may know...

    "Mz = torque around the vertical (Z) axis due to e.g. caster, King Pin Inclination, scrub radius etc."

    Fully understood. BUT does pCars' FFB system "weights" that Mz force as a part of the Fxyzm matrix in terms of being also one of those tireforces happening at the tire contact patch (i.e., e.g. acc. to Pacejka...), OR does Mz (in pCars) represent the torque at the (imaginary) steering axis, that is very depending on caster and KPI?

    I'm aware that the scrub radius as a result of caster and kingpin inclination has an impact to the tireforce Mz at tire contact patch as well, but as far as i know the absolute Mz force at tire patch is much less strong than the Mz force at the steering axis resulting from force translations where mainly Fy and Fx translate into a torque by the geometyr force effects due to the scrub radius.

    In other words: When i dial in Mz settings in the in-car FFB matrix, does the FFB system computates and weights into the mix the "weak" Mz (aligning moment) forces at the contact tire patch, or the "strong" Mz forces that happen at the steering axis, eventually transformed trough rod, rack and pinion to the final torque a the steering wheel rack?
     
    Last edited: Feb 7, 2016
  8. That's a very good question! I honestly don't know, sorry. If I get around to it, I'll try to poke AJ and see what he says. :)
     
    • Like Like x 1
  9. thanks a lot, Ermo!
    Also for being so fast to answer.
    your reply tells me that there is reason why that question kept me busy ;-)
     
  10. Per definition, Mz = (mechanical trail + pneumatic trail) * cos(caster angle) * Fy

    I'm fairly certain that the pneumatic trail torque is mixed in, but I'm also fairly certain that you get the stronger forces from the mechanical trail and Fy too. Does that answer your question?

    EDIT: Ok, got in touch with AJ (the tyre guru) and Casey Ringley (Vehicle Lead) and this is what they had to say on the matter:

    Hope that helps? :)
     
    Last edited: Feb 13, 2016
  11. Ermo thanks! very good and helpful answer. much appreciated.

    The Fx differential torque insight is great! It didn't come to my mind. But now your statement makes so much sense: It's the difference of the Fx right front tire and Fx left front tire that results in a torque moment due to scrub radius.

    Though one thing well probably stay a mystery to me for ever: Since with the game as in reality Fz almost doesn't do anything at the steering geometry that could lead to a torque, why is it that Jack Spade and most others with reasonable FFB setting are mixing so much Fz into the mix.

    My assumption is that this is the part that has not much to do with reality, but Fz is used as kind of differentiator (due its inverse torque direction with most cars) which adds just another effect to th eFFB feeling. Not really "canned" but kind of artificial... ?

    Anyway your post filled a blind spot, since the nature of Mz in the game is of principal meaning when you're looking for deeper understanding how pCars FFB really works...
     
    • Like Like x 1