T500RS FFB

Discussion in 'Technical & Support' started by colinL, Jun 18, 2019.

  1. colinL

    colinL Registered

    Joined:
    Jul 2, 2016
    Messages:
    46
    Likes Received:
    0
    Hey, I've been trying to fiddle around with the FFB settings for a while now, but I just can't seem to get them right. I'm mainly driving the cars from the Enduracers Pack and the URD PX cars. As a reference track I'd be using the newly released laserscanned Le Mans track, which should be top notch for ffb I guess. So, the cars feel pretty floaty to me, I can't really feel brakes locking up, especially braking while turning is extremely hard. It just kind of lacks overall feel for me. And I don't know why. I've followed guides regarding oscilliation, to not turn the ffb up too much as I then get clipping etc. I've also manipulated the brake ffb setting in the .JSON from -10000 to -8000 as that is supposed to give a more intense feeling. Has anyone got good settings for the T500RS? And would be willing to share them with me? Also, I'm mainly using the F1 Wheel AddOn as I'm mainly driving modern cars.
    Any help would be greatly appreciated :)
     
  2. Juwe Lahann

    Juwe Lahann Registered

    Joined:
    Dec 24, 2017
    Messages:
    72
    Likes Received:
    45
    Have you already set the software wheel rotation?
    https://forum.studio-397.com/index....wheel-rotation-not-working.61750/#post-968519
    1080 degrees for t500.
    Then you should try to set "steering resistance coefficient" from 1 to -0,55 in controller.json (i've tested -0,45 to -0,65, -0,55 was best for me),
    and "max torque" to 6,5 Nm.
    In rf2 options set "min. force" to 2%-3%.
    In TM settings 60/100/100, Damper and spring to 0.
    Or you can try this: https://www.racedepartment.com/downloads/dodds-t500-force-feedback-tweaks.20201/
    But then I would recommend to lower "min. force" again to 2-3% in rf2 options (it's at 8% in these settings iirc).
    This helped me a lot with offical content and mods, but I haven't tried PX1 mod.
     
    Last edited: Jun 21, 2019
    Corti likes this.
  3. colinL

    colinL Registered

    Joined:
    Jul 2, 2016
    Messages:
    46
    Likes Received:
    0
    Ok, thanks for that, much of these settings do already apply for me but I'll reply back.
     
  4. colinL

    colinL Registered

    Joined:
    Jul 2, 2016
    Messages:
    46
    Likes Received:
    0
    I've tried it, but somehow it doesn't really help that much. Maybe I am too stupid in setting up a car? I dont know. The back seems to slide out ever so often when I'm braking and turning in. But the car also feels a bit sluggish in general.
     
  5. Juwe Lahann

    Juwe Lahann Registered

    Joined:
    Dec 24, 2017
    Messages:
    72
    Likes Received:
    45
    Must be setup then. What about other cars? FFB was already good out of the box for me with rf2. Did you try to reduce max brake force in the car setup? lower the arbs/rear springs/increase coast difflock may help, but I don't know about those cars and possible setup options. I have URD DTM '18 mod, and standard setups wasn't that bad.
     
  6. colinL

    colinL Registered

    Joined:
    Jul 2, 2016
    Messages:
    46
    Likes Received:
    0
    I've indeed made some changes to the setup and I can push the car better now. Still, I think the FFB could be more informative, but I can't really make out why and what I'd want changed. Mainly I can't really feel lockups.
     
  7. Akela_DE

    Akela_DE Registered

    Joined:
    Jan 23, 2018
    Messages:
    389
    Likes Received:
    268
    Default settings for the T500 are very good. I would suggest, go back to the default settings first and for tests stick with S397 content.
    I have multi on 75-80, Smoothing between 0-4, depending on car
    60/100/100/0/0 in TM Driver Panel. Dont change that.

    BTW: Mods can be a joke FFB wise
     
  8. Juwe Lahann

    Juwe Lahann Registered

    Joined:
    Dec 24, 2017
    Messages:
    72
    Likes Received:
    45
    Yes, default settings are already good, but with the settings i have posted, you can set the ffb multiplier to 1,0 and you don't need any smoothing. The negative "steering resistance coefficient" is to compensate the belt friction of the wheel, it's much easier to catch slides. Don't know if it's realistic, but some people here and @RD(T300 and T500 users) liked it very much.
    The following is from a private conversation I had with the creator of dodds t500 tweaks I posted above:
    "As you've probably figured out, having too high negative co-efficient means the wheel looses all turn in weight along with over pronounced occillations. Too low and the belt friction is not compensated for.

    The Thrustmaster default is 60% overall and 100% on everything else. 60% still allows the wheelbase to peak output at 100%, but without compressing the signal too much.

    Thrustmaster could easily write what I have done into the firmware, and the wheel would be excellent out of the box. Because it is no longer a production unit, there's no incentive to run it with optimum settings. If anything, allowing the user to crank up the force to 100% (which is really 150%) shortens the life of the wheel and increases the chance they need a new one
    ."
     
  9. Akela_DE

    Akela_DE Registered

    Joined:
    Jan 23, 2018
    Messages:
    389
    Likes Received:
    268
    yes, i have done something like that. no big difference for me, "perhaps" a little bit better understeer feeling. But i also seem to have a VERY good charge. Wednesday wheel ;)
    But for colin its more important to have a good base feeling. And thats delivered by defaults and S397 content, dont you think ?
    And why set 1080 ? in AC/PC2/R3E you have to set 900° ?
    And whats about firmware/drivers ?
    Oh and regarding PC2 ... i had to change the USB Port from frontpanel to mainboard to get a good FFB - from numb to perfect. Perhaps worth a try to change the USB-port. Dont use the blue ones :D
     
    Last edited: Jun 21, 2019
  10. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    12,345
    Likes Received:
    6,572
    In-game FFB Mult of 1.0 isn't a good idea, many cars will give you clipping at quite normal (albeit high) cornering loads. Steering wheel settings can't undo the clipping. Better to start at around 0.8 in most mods, but where possible check the FFB levels using whatever means (perhaps a [paid] third party HUD solution, or use telemetry).

    I've got a feeling this person likes the wheelcheck test, which I think is flawed in terms of raw output strength (static). Here's a proper output-strength test using a T300, which wouldn't be too dissimilar. Note that, as you'd expect, output force is pretty much linear with input force, because that's how electric motors work:

    https://www.racedepartment.com/thre...based-on-seriously-flawed-assumptions.124937/

    I think the wheelcheck results may have some bearing with an unloaded wheel, so if you do lots of rally/drifting it may be more relevant. Though I think somewhere in between is probably still better.

    I'll give the negative coefficient a go though, sounds interesting.


    *One note on what I've said: many games saturate the FFB near peak loads, to improve feel with cheap consumer wheels (more force -> better feeling). rF2's FFB isn't quite as bad but will still clip on 1.0 Mult as I've said. So I wouldn't increase from 60% overall strength for a T500 or similar unless you lower the Mult to avoid clipping. Otherwise you will be cranking the motor harder than it's really intended for.

    And one more note on the T500, consider throwing a Noctua or similar fan in to replace the original, it'll stay quieter and cool better.
     
    Last edited: Jun 21, 2019
    Juwe Lahann and Corti like this.
  11. doddynco

    doddynco Registered

    Joined:
    Dec 13, 2017
    Messages:
    574
    Likes Received:
    1,217
    Hi guys. The 'anti-friction' is setup to make the mechanism as transparent as possible and works in essentially the same way as Fanatec drift mode. A co-eff of -0.45 is something like fanatec lvl1, and a co-eff of -0.65 is fanatec lvl5. This is desirable because the T500 has a fairly decent amount of friction in its mechanism. The idea is to use the power of the motor to gently and discretely compensate for that friction.

    The wheel will feel more powerful because the force will be slightly assisted in the direction of user input, then when the games 'kicks' agaist that direction, the difference between the directions is decently higher.

    The T300 isn't really comparable because it has very little friction in comparison to the t500 - trying to use the same trick would be pointless.

    I understand the flaw with wheelcheck, and that distance is not a fair measure of motor output in the case ffb wheels. But it is quite clear from the feel and sound of the t500 that once it gets to about 60-70% of its theoretical hardware input level, it is in a hardware clipping state. I also believe (and this is entirely my subjective interpretation) that the t500 does 'compress' the signal to some extent, and wheelcheck, in its infinite wisdom, does seem to confirm visually what I am feeling. The linearity is therefore set to 0.8 - expanding the signal to offset that compression.

    The force is reduced in the ini down from -10000 to -8000 and the linearity set to 0.8. These two measures combined, if you graph them out, you will see that now @100% software input, hardware input is just about maximised.

    One final thing is that the minimum force should be set to 2-3% in game, official content at 100% multiplier. The Thrustmaster contoller settings should be set to default.
     
    Last edited: Jun 21, 2019
    Juwe Lahann likes this.
  12. Juwe Lahann

    Juwe Lahann Registered

    Joined:
    Dec 24, 2017
    Messages:
    72
    Likes Received:
    45
    Today I tried the recommended t300 settings linked by Lazza and rawly adapted it to t500. Changed Force/linearity back to -10000/1, steering resistance coefficient -0,45, software wheel rotation 900, TMcontroller strenght 100%, rotation 900.
    Wasn't a big difference, as expected. Had to fiddle around with minimum torque+ffb multipler. A bit more overall strength, but thats probably because I was using 1080 wheel rotation before(as mentioned in Lazzas link). I like it too. My neighbours maybe don't :D. The other setting is more friendly for the neighborhood (wood rig). Have to do further testing. Also have to try steering resistance coefficient at +1 to feel the difference again.
     
    Last edited: Jun 21, 2019
  13. Juwe Lahann

    Juwe Lahann Registered

    Joined:
    Dec 24, 2017
    Messages:
    72
    Likes Received:
    45
    and hello dodd :)
     
    doddynco likes this.
  14. Akela_DE

    Akela_DE Registered

    Joined:
    Jan 23, 2018
    Messages:
    389
    Likes Received:
    268
    just tried the tweaks from RD/Dodd. Feels very good, detailed and fast. maybe there could be a bit more understeer feeling.
    But for colinL: brake locking is there ;)
     
  15. doddynco

    doddynco Registered

    Joined:
    Dec 13, 2017
    Messages:
    574
    Likes Received:
    1,217
    Thanks for trying it out - I'm very happy that people are finding it to be a good improvement. If you like, try setting the steering resistance co-eff down to -0.6 for slightly more friction compensation - it is perhaps a better setting and will most likely be the setting for the next update. Your, erm, feedback... is much appreciated. Be sure to 'load' the file again when you're in rfactor if you try it.
     
  16. doddynco

    doddynco Registered

    Joined:
    Dec 13, 2017
    Messages:
    574
    Likes Received:
    1,217
    You have missed here that steering resistance saturation needs to be higher than zero - otherwise co-eff has no effect.
     
    Juwe Lahann likes this.
  17. Dady Cairo

    Dady Cairo Registered

    Joined:
    May 11, 2017
    Messages:
    243
    Likes Received:
    160
    Ähhem...i read for the T300 it is not useful to tweak it with doodynco's file at RD.
    Is there a good T300 setting you can recommend,at least a hint?
    Or is it good "out of the Box"
    Many ThanX
    DC
     
  18. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    12,345
    Likes Received:
    6,572
    There's some confusion here. Let me reiterate a couple of things.

    The link I posted showed the same compression on a T300 via wheelcheck, but linear force when actually measuring the force. The internal friction you're trying to reduce will surely have an effect on the rotation during normal wheelcheck tests. I'm inclined to believe the output force is linear (to 100%) rather than compressed, in the absence of an actual test.

    This is only ok because you're editing the .ini to 80% of full force. Otherwise a Mult of 1.0 in game will clip. Some cars will still clip at that level, so it's important to check.

    I didn't post any T300 settings. I have a post explaining the steps required to enable software rotation in rF2 (which I use myself, though I use 900 degrees (T500) as I don't drive cars requiring more lock and that figure is more compatible with other games). I also linked to the wheelcheck vs measured force test, which is using a T300, but if there are any other settings on there I wasn't referring to them - just highlighting the difference between wheelcheck wheel travel and force output linearity.
     
  19. doddynco

    doddynco Registered

    Joined:
    Dec 13, 2017
    Messages:
    574
    Likes Received:
    1,217
    I almost entirely agree however, I consider that when the belts pull tight around the pulleys, it effectively compresses the peaking forces from the (lets say linear) motor output. The lower power, sustained forces, are not going to be noticably changed by a 0.8 curve. And so that trade off i think is the right way to go about it. The t500 is a powerful but slightly crude wheel, so I just did what I could to iron out some of its imperfections with the suprisingly useful settings hidden in the ini.

    On that second point: it's still up to the player to set the right multiplier in game - especially with mods as they more offen than not need adjusting. I have just made it so official content will be as s397 intended, which is not clipping aside from rare occasions - a minor clip is really nothing to worry about.

    One other thing is that the tweaks won't show in wheelcheck because they only take effect once you're actually in RF. But i see what you're saying about the accumalative processes interacting. It was very much a feeling process. I had a bunch of other wheels to reference it to including a cswv2 so I wasn't just firing in the dark.
     
  20. Lazza

    Lazza Registered

    Joined:
    Oct 5, 2010
    Messages:
    12,345
    Likes Received:
    6,572
    Ok, if you're trying to 'flatten out' the response to jolts, changes of direction, more than the 100% sustained force, I can see the intention (we can see from the T300 scales test that sustained forces up to 100% remain linear, and the T500 isn't so much stronger that it should change markedly).

    I think this is separate to the overall strength setting of the wheel though, so it may need to be done whether the wheel is set to 60% or 100% overall (the wheel won't try to push more than 60% if it's set to 60%, so it can't compensate at moments of high 'impact'). I don't see 100% TM strength as detrimental in terms of feedback, and I think that's important to highlight as running the wheel at 100% but running cars lower than 0.8 Mult in-game (0.8@60% = 0.48@100%) allows that occasional clipping (which, as you rightly say, doesn't really matter in the scheme of things) to be further reduced - which isn't a bad thing. Actually it allows for a stronger wheel response overall, because you don't need to drop down to 0.48; 0.65 is a good figure for a lot of content, avoids clipping in nearly all normal driving situations, and immediately gives more strength than the default 60% wheel setting (obviously; 0.65 is already higher than 60%, while many previously-clipped forces are now preserved).
     

Share This Page