rec.autos.simulators

>Ever since the arcade game HardDrivin came (mid 80,s ?)out with its force feedback wheel , I have been looking forward to having a
>driving game with force feedback that was as good as HardDrivin. With HardDrivin  you would feel more than just the shakes of off
>road and crashes, the steering effort would change with the weight on the front wheels and tire grip. Driving a race car has a lot
>to do with weight transfer and can be felt through g-forces or through steering wheel feedback or effort of steering. HardDrivin
>would change from a light wheel when hard acceleration to hard steering when braking. In a turn if the car understeers, the steering
>would get very light and you could modulate the throttle to where you felt the tires regain grip as the steering got harder. You did
>not have to wait to see the car understeer, you could feel it through the steering effort.  I thought HardDrivin simulated actual
>steering feel very well.
>Well when the first PC force feedback games and wheels came out, they were just with the simple jerks and rumble from road surface
>and crashing and no real feel of car handling. I also read that real handling feel for PC games and FF wheels could not be done
>correctly with affordable hardware. Now I read a lot of argument of the current games and FF wheels and how realistic they feel
>verses some who say they are not really that good for actual car feel. I also spent a lot of time making my own analog wheel and
>pedals (I would say quality almost as good a EECI wheel) so I am hesitant to run out and buy a new FF wheel.
>So if anyone that has tried new FF wheels and remembers how the HardDrivin game FF effect were, how do you compare them?

>Lutrell :-)

>DSR # 74
>http://www.racesimcentral.net/

>Mine will break your wrist if you set the force high enough.

>For me, there is no comparison at all.  Hard Drivin' & Race Drivin'
>actually run fast enough to be "realtime".

>>Mine will break your wrist if you set the force high enough.

>If so, you should either be working out or drinking a lot of milk.

>Here is one test for lag you might try, by running a simple frequency
>response test -- I haven't got all the hardware to try different
>combinations, etc.  Read it through a few times, I tried to cover all the
>bases but it's a little tricky to explain...

> On Wed, 20 Oct 1999 05:13:59 GMT, Doug Milliken

> >Here is one test for lag you might try, by running a simple frequency
> >response test -- I haven't got all the hardware to try different
> >combinations, etc.  Read it through a few times, I tried to cover all the
> >bases but it's a little tricky to explain...
> [snip]

> Are you saying that the steering wheel is eventually wobbling by
> itself?  Like with your hands off?

> Ah, I knew this wouldn't be properly explained, sorry...

> I meant that you have to force the wheel (with your hands) to steer
> back and forth.  I'm trying to get at the motion of the wheel vs the motion
> of the car.  Alternatively you can also look at the motion of the steering
> wheel shown on the screen (if there is one shown) and find the natural
> frequency for the wheel display..

> If the steering wheel oscillates back and forth on it's own, that's a
> different problem -- not likely to be found in race cars, but some
> production cars with heavy steering wheels will do this.  Years ago, there
> was a USA-built car called the Dodge Omni (based on VW Rabbit/Golf) that
> got in trouble with "Consumer Reports" magazine because it had this type of
> "hands-off" oscillation above some speed.  I think that the original Suzuki
> Samuri that was brought into the USA would also do this.

> -- Doug Milliken

> > On Wed, 20 Oct 1999 05:13:59 GMT, Doug Milliken

> > >Here is one test for lag you might try, by running a simple frequency
> > >response test -- I haven't got all the hardware to try different
> > >combinations, etc.  Read it through a few times, I tried to cover all the
> > >bases but it's a little tricky to explain...
> > [snip]

> > Are you saying that the steering wheel is eventually wobbling by
> > itself?  Like with your hands off?

> I know that if I let go in GPL...my MSFF starts wobbling from side to
> side BADLY...at maybe 4/5 hz

> > Ah, I knew this wouldn't be properly explained, sorry...

> > I meant that you have to force the wheel (with your hands) to steer
> > back and forth.  I'm trying to get at the motion of the wheel vs the motion
> > of the car.  Alternatively you can also look at the motion of the steering
> > wheel shown on the screen (if there is one shown) and find the natural
> > frequency for the wheel display..

> > If the steering wheel oscillates back and forth on it's own, that's a
> > different problem -- not likely to be found in race cars, but some
> > production cars with heavy steering wheels will do this.  Years ago, there
> > was a USA-built car called the Dodge Omni (based on VW Rabbit/Golf) that
> > got in trouble with "Consumer Reports" magazine because it had this type of
> > "hands-off" oscillation above some speed.  I think that the original Suzuki
> > Samuri that was brought into the USA would also do this.

> > -- Doug Milliken

> > > On Wed, 20 Oct 1999 05:13:59 GMT, Doug Milliken

> > > >Here is one test for lag you might try, by running a simple frequency
> > > >response test -- I haven't got all the hardware to try different
> > > >combinations, etc.  Read it through a few times, I tried to cover all the
> > > >bases but it's a little tricky to explain...
> > > [snip]

> > > Are you saying that the steering wheel is eventually wobbling by
> > > itself?  Like with your hands off?

>> >AFAIK, all the new PC FF wheels run through a Windows API, which gives a
>> >noticable (and variable) lag between car motion and steering force.

>> The lag is very low now with serial, or especially USB connections to
>> the wheel.  It really isn't an issue for me at this point.

>Hi Pat and all,

>Here is one test for lag you might try, by running a simple frequency
>response test -- I haven't got all the hardware to try different
>combinations, etc.  Read it through a few times, I tried to cover all the
>bases but it's a little tricky to explain...

>When driving straight ahead at some moderate speed (say 50-60 mph), turn
>the wheel back and forth in an approximate sine wave motion.  Kind of like
>race drivers do when they are trying to warm the tires behind a pace car,
>but much less steering wheel angle.  The idea is to keep the amplitude
>small enough that you stay in the linear range of the car tires, so there
>is no question of the car sliding around or getting loose.

>Now speed up the sine wave motion and watch the reaction of the car (and/or
>the steering wheel in the car).  At some point the motion that you see on
>the screen will be about 180 degrees out of phase -- the car (or picture of
>the steering wheel) will still be going left while you are turning the
>wheel to the right.

>Time several reversals of the wheel with a stop watch and determine the
>cycles/second (Hz) of the sine wave at the 180 degree point.  This is about
>the natural frequency of the car in yaw (rotation about a vertical axis)
>and it should be related only to the response of the car (without any
>additional lag from the steering wheel interface).

>You can also do this in a regular car (don't try it with passengers, they
>will think you have lost it...!!).  There are usually two natural
>frequencies on passenger cars, the roll frequency is low, maybe 1 Hz,
>ignore this one, it's just confusing the picture.

>The yaw frequency for smaller passenger cars is in the 2 Hz range, light
>race cars with big tires might be 6 Hz or more (I recall hearing about one
>F1 car from some years ago where the data channels all had 7 Hz
>oscillations superimposed on the car motions.)

>I just tried it with 2 different cars in "Race Drivin'", the red car (auto
>trans) was about 2.5 Hz, the other (the little Miata-like car) was a little
>over 3 Hz.  I started a stop watch, then got the steering wheel oscillating
>at the right frequency (at a constant 55 mph on a straight road) and then
>counted steering cycles (right/left/back-to-right) for 10 seconds.

>Please post your results -- include your hardware and software config,
>and what game you are running.

>I look forward to being proved wrong, but the little exposure to
>PC wheels that I've had makes me feel pretty safe...

>-- Doug Milliken

> GPL with USB ACT Labs wheel at 55 mph oscillates from side to side at 2.5/3

> of phase at all.

> When I let go of the wheel, the oscillations die out after 2 or three
> cycles, and the steering goes back to center.

> >> >AFAIK, all the new PC FF wheels run through a Windows API, which gives a
> >> >noticable (and variable) lag between car motion and steering force.

> >> The lag is very low now with serial, or especially USB connections to
> >> the wheel.  It really isn't an issue for me at this point.

> >Hi Pat and all,

> >Here is one test for lag you might try, by running a simple frequency
> >response test -- I haven't got all the hardware to try different
> >combinations, etc.  Read it through a few times, I tried to cover all the
> >bases but it's a little tricky to explain...

> >When driving straight ahead at some moderate speed (say 50-60 mph), turn
> >the wheel back and forth in an approximate sine wave motion.  Kind of like
> >race drivers do when they are trying to warm the tires behind a pace car,
> >but much less steering wheel angle.  The idea is to keep the amplitude
> >small enough that you stay in the linear range of the car tires, so there
> >is no question of the car sliding around or getting loose.

> >Now speed up the sine wave motion and watch the reaction of the car (and/or
> >the steering wheel in the car).  At some point the motion that you see on
> >the screen will be about 180 degrees out of phase -- the car (or picture of
> >the steering wheel) will still be going left while you are turning the
> >wheel to the right.

> >Time several reversals of the wheel with a stop watch and determine the
> >cycles/second (Hz) of the sine wave at the 180 degree point.  This is about
> >the natural frequency of the car in yaw (rotation about a vertical axis)
> >and it should be related only to the response of the car (without any
> >additional lag from the steering wheel interface).

> >You can also do this in a regular car (don't try it with passengers, they
> >will think you have lost it...!!).  There are usually two natural
> >frequencies on passenger cars, the roll frequency is low, maybe 1 Hz,
> >ignore this one, it's just confusing the picture.

> >The yaw frequency for smaller passenger cars is in the 2 Hz range, light
> >race cars with big tires might be 6 Hz or more (I recall hearing about one
> >F1 car from some years ago where the data channels all had 7 Hz
> >oscillations superimposed on the car motions.)

> >I just tried it with 2 different cars in "Race Drivin'", the red car (auto
> >trans) was about 2.5 Hz, the other (the little Miata-like car) was a little
> >over 3 Hz.  I started a stop watch, then got the steering wheel oscillating
> >at the right frequency (at a constant 55 mph on a straight road) and then
> >counted steering cycles (right/left/back-to-right) for 10 seconds.

> >Please post your results -- include your hardware and software config,
> >and what game you are running.

> >I look forward to being proved wrong, but the little exposure to
> >PC wheels that I've had makes me feel pretty safe...

> >-- Doug Milliken

>> GPL with USB ACT Labs wheel at 55 mph oscillates from side to side at
2.5/3

out
>> of phase at all.

>> When I let go of the wheel, the oscillations die out after 2 or three
>> cycles, and the steering goes back to center.

>Wow, more data, thanks!  This seems pretty good, I may have to change my
>position on PC FF in the near future.  As a wild guess, I'd say that those
>60's F1 cars on "medium" width tires might have been ~4 Hz.  The sports
>cars that are modeled in the old Atari Race Drivin' are probably in the
>2.5/3 Hz range so they were "easier" to model, in that sense.

>I'd think that the damping (cycles to damp out after you let go) would be
>better in a real F1 car, but 2-3 cycles isn't completely terrible.

> So Doug, have you tried the FF in GPL?  It has a damping setting that I keep
> as low as possible, since it tends to mask some of the other feedback.  I
> would imagine that cranking up the damping would make the wheel straighten out
> almost immediately.  I might do some experimenting with it to see.

> I'd like to hear your impressions of GPL with a Logitech or ACT Labs
> wheel.

> --
> Pat Dotson

> >> GPL with USB ACT Labs wheel at 55 mph oscillates from side to side at
> 2.5/3

> out
> >> of phase at all.

> >> When I let go of the wheel, the oscillations die out after 2 or three
> >> cycles, and the steering goes back to center.

> >Wow, more data, thanks!  This seems pretty good, I may have to change my
> >position on PC FF in the near future.  As a wild guess, I'd say that those
> >60's F1 cars on "medium" width tires might have been ~4 Hz.  The sports
> >cars that are modeled in the old Atari Race Drivin' are probably in the
> >2.5/3 Hz range so they were "easier" to model, in that sense.

> >I'd think that the damping (cycles to damp out after you let go) would be
> >better in a real F1 car, but 2-3 cycles isn't completely terrible.