Jump to content


- - - - -

Force Feedback Revisited


  • Please log in to reply
57 replies to this topic

#21 Lee200

Lee200

    Denny Hulme

  • Administrators
  • PipPipPipPipPipPipPipPipPipPip
  • 1,678 posts
  • Gender:Male
  • Sim interest:GPL

Posted Jan 28 2014 - 11:47 AM

John, I don't know, but it sounds logical that the "caster" setting as was we call it might affect mechanical trail, because as you say, that affects the nose rise effect.  But Nigel said that "caster" was applied directly to the FFB curve.  The curve looks like pneumatic trail so perhaps it's a fudge factor that is applied to both?

I need to do some debugging to better understand the FFB code.  Have meant to do that for a long time and never gotten around to it.

The FFB curve is adjusted for vertical force, temperature, and pressure.

In real life, too much mechanical trail is bad because it can mask the pneumatic trail force which gives the driver warning about approaching max G slip angle.

#22 Lee200

Lee200

    Denny Hulme

  • Administrators
  • PipPipPipPipPipPipPipPipPipPip
  • 1,678 posts
  • Gender:Male
  • Sim interest:GPL

Posted Jan 28 2014 - 02:35 PM

I did some snooping in the code and couldn't determine how the caster angle was applied to either the FFB or nose rise code.  I know where the caster value resides, but it's used in some subroutines that are beyond my comprehension at this point. :(

The FFB code itself is much more complicated than I expected, but it's not too hard to follow along.

#23 Roo

Roo

    Jim Clark student

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 318 posts
  • Gender:Male
  • Location:France 77
  • Interests:climbing
  • Sim interest:GPL

Posted Jan 28 2014 - 04:50 PM

View PostClaudio Pablo Navonne, on Jan 26 2014 - 09:13 AM, said:

This is my profile setting, I hope it could be useful for you.

forgive me if this is stupid question but as this is an .xml file, should this go in gplsecrets/options & be selected in GEM?

#24 John Woods

John Woods

    Be Somebody

  • GPLLinks Team
  • PipPipPipPipPipPipPipPipPipPip
  • 2,657 posts
  • Gender:Male
  • Interests:Too Much Fun
  • Sim interest:GPL

Posted Jan 28 2014 - 07:26 PM

Are derivative coefficients fudge factors that modify specific variables that come from other equations?
(If anyone answers, all I will understand is "yes" or "no"). :)
Isn't mechanical trail constant?
And pneumatic trail is function of varying lateral force?
Maybe somewhere in the clutch.

Edited by John Woods, Jan 28 2014 - 07:35 PM.


#25 Lee200

Lee200

    Denny Hulme

  • Administrators
  • PipPipPipPipPipPipPipPipPipPip
  • 1,678 posts
  • Gender:Male
  • Sim interest:GPL

Posted Jan 28 2014 - 08:50 PM

1.  I've no idea what you're asking.  :)
2.  Yes
3.  No.  Yes.  The torque about the wheel is the product of lateral force and pneumatic trail distance.  And pneumatic trail distance varies with slip angle/lateral force.

Edited by Lee200, Jan 29 2014 - 11:32 AM.


#26 John Woods

John Woods

    Be Somebody

  • GPLLinks Team
  • PipPipPipPipPipPipPipPipPipPip
  • 2,657 posts
  • Gender:Male
  • Interests:Too Much Fun
  • Sim interest:GPL

Posted Jan 29 2014 - 07:23 AM

1. Me neither. Its a reference to a past in which I paid less attention. But here is a quote from somewhere: "Differential calculus or the concept of Derivative and Differential Coefficient was discovered by Isaac Newton (1642-1727) and Gottfried Wilhelm Leibnitz (1646-1716) in the process of solving two old problems one of finding slope of tangent drawn to a curve and another of finding instantaneous velocity of an object in non-uniform motion."
2. OK.
3. Pneumatic trail caused by varying lateral force is the distortion of the tire patch behind the wheel center, in proportion to slip angle, and that produces the torque which may or may not be felt in FFB?

Well I think I'm past my capability to sort of credibly think about this aspect too much more. Thank you again, Lee, in my case for your patience.

Edited by John Woods, Jan 29 2014 - 08:00 AM.


#27 MLGathome

MLGathome

    Denny Hulme

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 310 posts
  • Gender:Male
  • Location:Netherlands
  • Interests:simracing
  • Sim interest:GPL

Posted Jan 29 2014 - 07:34 AM

View PostRoo, on Jan 28 2014 - 04:50 PM, said:

View PostClaudio Pablo Navonne, on Jan 26 2014 - 09:13 AM, said:

This is my profile setting, I hope it could be useful for you.

forgive me if this is stupid question but as this is an .xml file, should this go in gplsecrets/options & be selected in GEM?

No Roo, I think it is his listing from Logitech profiler to use these settings for GPL, look at the values.

Greetings, Marcel

#28 Lee200

Lee200

    Denny Hulme

  • Administrators
  • PipPipPipPipPipPipPipPipPipPip
  • 1,678 posts
  • Gender:Male
  • Sim interest:GPL

Posted Jan 29 2014 - 11:30 AM

View PostJohn Woods, on Jan 29 2014 - 07:23 AM, said:

1. Me neither. Its a reference to a past in which I paid less attention. But here is a quote from somewhere: "Differential calculus or the concept of Derivative and Differential Coefficient was discovered by Isaac Newton (1642-1727) and Gottfried Wilhelm Leibnitz (1646-1716) in the process of solving two old problems one of finding slope of tangent drawn to a curve and another of finding instantaneous velocity of an object in non-uniform motion."
2. OK.
3. Pneumatic trail caused by varying lateral force is the distortion of the tire patch behind the wheel center, in proportion to slip angle, and that produces the torque which may or may not be felt in FFB?

Well I think I'm past my capability to sort of credibly think about this aspect too much more. Thank you again, Lee, in my case for your patience.

#3.  Yes.  :)

Edited by Lee200, Jan 29 2014 - 11:31 AM.


#29 Claudio Pablo Navonne

Claudio Pablo Navonne

    Denny Hulme

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 252 posts
  • Gender:Male
  • Location:Buenos Aires, Argentina
  • Sim interest:GPL

Posted Jan 30 2014 - 03:13 AM

View PostMLGathome, on Jan 29 2014 - 07:34 AM, said:

View PostRoo, on Jan 28 2014 - 04:50 PM, said:

View PostClaudio Pablo Navonne, on Jan 26 2014 - 09:13 AM, said:

This is my profile setting, I hope it could be useful for you.

forgive me if this is stupid question but as this is an .xml file, should this go in gplsecrets/options & be selected in GEM?

No Roo, I think it is his listing from Logitech profiler to use these settings for GPL, look at the values.

Greetings, Marcel

:thumbup:  Yes Sir.

#30 Roger_F

Roger_F

    Fiero NStar V8

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 67 posts
  • Gender:Male
  • Location:USA
  • Interests:Classic Sports Cars
  • Sim interest:GPL

Posted Dec 30 2019 - 08:37 PM

Gentleman,

That was a very good in-depth discussion!  But is there a practical thread to these concepts that will allow my new G29 Logitech wheel to behave better from a force feedback perspective?  Basic GPL and Logitech software do not seem to be able solve the problem; the FFB2 patch is a noticeable improvement for the G29, but still vague in the center area (seems to exhibit a lack of feel).  The best so far seems to be the FFB3test patch I found and tried today.  

Suggested parameters for FFB variables appear to be all over the map .... I keep trying various combinations hoping for a steady long term solution.  Many moons ago I campaigned an E-Production MGB in SCCA  and over the years got to sample other drivers's rides (both sports cars and formula cars); none were as hard to drive as our beloved GPL mod cars.

The biggest differences seem to be in transition;  GPL cars seem harder to control during weight shift transitions than real race cars. A good race car driver can catch a miss behaving car with a very high degree of success maybe 95% of the time or more; you rarely see an unassisted spin out in an SCCA event.  GPL catch success seems much lower than 95%!

I look forward to your thoughts and comments,

Roger

p.s. The G29 journey has been an experience; my old MOMO was much easier to tame but still not as catch-able as real race cars.

#31 marco lenarduzzi

marco lenarduzzi

    Denny Hulme

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 427 posts
  • Gender:Male
  • Location:brome lake Quebec, Canada
  • Interests:racing sims,cars, F1,cooking,skiing
  • Sim interest:GPL

Posted Dec 31 2019 - 01:42 AM

Hi Roger. Have you adjusted the Force feed back settings in Gem+. There is a "control" button and when opened you have settings. Try disabling the steering hack,make sure the FFB is checked and from what I have read in the forums set the FFB latency to 0 and the FFB dampening to 0 as well. I have the FFB Maximum force set at 290 but most find that high. Remember that the lower the number for Maximum force the stronger the forces are so you may want to raise that number to lessen the forces. See if that improves you FFB. You can also play with the steering linearity in game in the options menu. The more towards the linear side (left) the less the dead zone. I hope this helps.

#32 M Needforspeed

M Needforspeed

    Denny Hulme

  • GPLLinks Team
  • PipPipPipPipPipPipPipPipPipPip
  • 2,669 posts
  • Gender:Male
  • Interests:GPL - start a collection of GPL era die cast racing cars -Skiing
  • Sim interest:GPL

Posted Dec 31 2019 - 02:11 AM

View PostRoger_F, on Dec 30 2019 - 08:37 PM, said:

Gentleman,

That was a very good in-depth discussion!  But is there a practical thread to these concepts that will allow my new G29 Logitech wheel to behave better from a force feedback perspective?  Basic GPL and Logitech software do not seem to be able solve the problem; the FFB2 patch is a noticeable improvement for the G29, but still vague in the center area (seems to exhibit a lack of feel).  The best so far seems to be the FFB3test patch I found and tried today.  

Suggested parameters for FFB variables appear to be all over the map .... I keep trying various combinations hoping for a steady long term solution.  Many moons ago I campaigned an E-Production MGB in SCCA  and over the years got to sample other drivers's rides (both sports cars and formula cars); none were as hard to drive as our beloved GPL mod cars.

The biggest differences seem to be in transition;  GPL cars seem harder to control during weight shift transitions than real race cars. A good race car driver can catch a miss behaving car with a very high degree of success maybe 95% of the time or more; you rarely see an unassisted spin out in an SCCA event.  GPL catch success seems much lower than 95%!

I look forward to your thoughts and comments,

Roger

p.s. The G29 journey has been an experience; my old MOMO was much easier to tame but still not as catch-able as real race cars.
  
informative post Roger, and sum up what we though about FF being a feature better than nothing, but far from real car physic behaviour

Edited by M Needforspeed, Dec 31 2019 - 02:12 AM.


#33 Roger_F

Roger_F

    Fiero NStar V8

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 67 posts
  • Gender:Male
  • Location:USA
  • Interests:Classic Sports Cars
  • Sim interest:GPL

Posted Dec 31 2019 - 06:11 AM

View Postmarco lenarduzzi, on Dec 31 2019 - 01:42 AM, said:

Hi Roger. Have you adjusted the Force feed back settings in Gem+. There is a "control" button and when opened you have settings. Try disabling the steering hack,make sure the FFB is checked and from what I have read in the forums set the FFB latency to 0 and the FFB dampening to 0 as well. I have the FFB Maximum force set at 290 but most find that high. Remember that the lower the number for Maximum force the stronger the forces are so you may want to raise that number to lessen the forces. See if that improves you FFB. You can also play with the steering linearity in game in the options menu. The more towards the linear side (left) the less the dead zone. I hope this helps.

Marco,

Thank you very much for the detailed response,  I have tried most of the suggestions you have detailed above.  My current tact is to run pure linear with the steering hack turned off.  I have also customized each control panel setting by mod.  My general pattern is to set latency to 0;  FFB damping to a very small number like 9 or so;  and adjust each mod to it's own Maximum force number.

I generally run 60fps and left foot braking with paddle shifters; my home built gaming rig is fairly up to date with a relatively new 4 gig video card and a 25 inch gaming monitor that has a native resolution of 1920 by 1080.

Generally I'm experimenting with setting the "higher the weight and power of the car set" the lower the maximum force (i.e. the greater the maximum force number).  F1 1965 and F2 have the lowest maximum force numbers (around 450) and SCGT and the CANAM Mods the higher (around 630).  The G29 Logitech wheel produces a lot of FFB intensity with its dual motors; much more than my old MOMO; other wheels would probably need much lower numbers.  Also keep in mind the FFB3 test patch and it's characteristics.  I may be wondering into a parallel universe between the experimental patch and the G29 Logitech wheel.

This latest approach is very much in the experimental stage and truly a "set variables by feel" experiment.  I would love a more structured scientific approach but have not yet identified one.  Each physics model seems to have it's own built in FFB characteristics?

My forcing objective is to improve "catch-ability"  and thus reduce the probability of a single car spin outs in "qualifying like" situations.  i am making some progress but have a long way to go.  Best lap times are less important to me than consistency; I'm shooting for something approaching the 95% "catch-ability" of real race cars.

As I said earlier, I really appreciate your detailed response ... I am always looking for fresh ideas!

Best,   Roger

#34 Lalethor

Lalethor

    Alessandro Isacchini

  • Supporter
  • PipPipPipPipPipPipPipPipPipPip
  • 307 posts
  • Gender:Male
  • Location:Verona Italy
  • Interests:Racing & Gnagna
  • Sim interest:GPL and P&G

Posted Dec 31 2019 - 01:59 PM

Hi Roger, first of all I must to tell the G29 is a terrible wheel for GPL (IMHO of course) I buyed it last year, try it and re-try for a week and sold :lol:
Maybe the problem is "over filtered" that wheel by default and no possible to adjust via logitech software, works well in the long turns where the precision of wheel allows very clean and precision trajectory, but was terrible in the quickly maneuvers and recovery car after a oversteer for example
And I remember problems in the center of wheel, and no natural progressive from middle

I don't know if is your same problem but you can try this procedure also
http://forum.project...Gaming-Software


P.S. where is possible to find  FFB 3 test patch?

#35 Roger_F

Roger_F

    Fiero NStar V8

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 67 posts
  • Gender:Male
  • Location:USA
  • Interests:Classic Sports Cars
  • Sim interest:GPL

Posted Dec 31 2019 - 03:26 PM

Lalethor,

Try searching for "Force feedback update for GPL"; the subtitle is "Brr's Ffb V3 Test."   I'm not on the computer that has the patch or i would just send it to you.  If you have trouble finding it, let me know and I will send it to you from my Desktop.  Full linear steering and Ffb V3 Test made a big difference in the play-ability of G29; hope you made out alright on the sale of the G29 wheel.

I'll check out your suggestion site.

Roger

p.s.  Don't apologize for your English, I was born in the US but am Italian by heritage; at best I know a few words in Italian.

#36 marco lenarduzzi

marco lenarduzzi

    Denny Hulme

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 427 posts
  • Gender:Male
  • Location:brome lake Quebec, Canada
  • Interests:racing sims,cars, F1,cooking,skiing
  • Sim interest:GPL

Posted Jan 01 2020 - 02:45 AM

View PostLalethor, on Dec 31 2019 - 01:59 PM, said:

Hi Roger, first of all I must to tell the G29 is a terrible wheel for GPL (IMHO of course) I buyed it last year, try it and re-try for a week and sold :lol:
Maybe the problem is "over filtered" that wheel by default and no possible to adjust via logitech software, works well in the long turns where the precision of wheel allows very clean and precision trajectory, but was terrible in the quickly maneuvers and recovery car after a oversteer for example
And I remember problems in the center of wheel, and no natural progressive from middle

I don't know if is your same problem but you can try this procedure also
http://forum.project...Gaming-Software


P.S. where is possible to find  FFB 3 test patch?
Hi Lalethor.  Click on the link  http://srmz.net/inde...050#entry128564

#37 Lalethor

Lalethor

    Alessandro Isacchini

  • Supporter
  • PipPipPipPipPipPipPipPipPipPip
  • 307 posts
  • Gender:Male
  • Location:Verona Italy
  • Interests:Racing & Gnagna
  • Sim interest:GPL and P&G

Posted Jan 01 2020 - 09:29 AM

Thanks Marco

#38 marco lenarduzzi

marco lenarduzzi

    Denny Hulme

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 427 posts
  • Gender:Male
  • Location:brome lake Quebec, Canada
  • Interests:racing sims,cars, F1,cooking,skiing
  • Sim interest:GPL

Posted Jan 01 2020 - 06:18 PM

View PostLalethor, on Jan 01 2020 - 09:29 AM, said:

Thanks Marco
:up:

#39 Pete Gaimari

Pete Gaimari

    Jim Clark

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 3,749 posts
  • Gender:Male
  • Location:Colorado
  • Interests:Racing and Muzzleloader Hunting
  • Sim interest:GPL

Posted Jan 02 2020 - 08:26 AM

Roger...........I'm curious. Did you LF brake in your MGB?

#40 Roger_F

Roger_F

    Fiero NStar V8

  • Members
  • PipPipPipPipPipPipPipPipPipPip
  • 67 posts
  • Gender:Male
  • Location:USA
  • Interests:Classic Sports Cars
  • Sim interest:GPL

Posted Jan 02 2020 - 01:38 PM

Pete,

A very good question, and as a matter of fact ... no left foot braking.

My gas pedal had an addition welded on the bottom (left hand side) that extended under the brake pedal and facilitated heal / toe down shifts.  The gas pedal looked like a backwards  "L."  My left foot rested an a pedal rest just left of the clutch pedal when the clutch was not being engaged.  Another fact:  Due to the poor sincro-mesh units of the day, we also had to double clutch on all down shifts ... another hindrance to left foot braking and advantage to heal / toe downshifts.  This situation was pretty typical for professionally prepared cars in the 70s, 80s, and even some of the 90s.

Some other interesting facts on the car:  I bought my car from a very successful national guy that built the car just for SCCA.  It was converted to right hand drive to move 250 lbs plus from the outside of most turns to the inside of most turns;  steering column, pedals, driver seat, harness and driver are a fair amount of weight.   Most SCCA tracks are run clockwise.  Right hand drive also facilitated being able to exactly see the apex of all right hand turns instead of judging the where-abouts of the right side of the car from the left seat.  Anything to gain an advantage!

In my days of racing, left foot braking was not very prevalent; the steering column often separated the break pedal from the clutch pedal and made left foot breaking physically difficult if not impossible.  I believe this was even the case in formula one and two in the 60s and early 70s.  I occasionally drove a Formula Ford that belonged to a friend of mine and there was no way to left foot brake on his car.  How times and technology have changed!

Hope this wasn't too much information,

Roger




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users

Sim Racing Links