At 3/30/03, you wrote:
Michael Selig writes:

 > I have noticed that in many fgfs joystick property files some axes
 > are "squared".  This makes the stick-position to control-surface
 > mapping nonlinear (squared) rather than one to one (linear).  I
 > think real airplanes w/ "reversible" controls are closer to being
 > linear rather than squared.

Yes, that is correct; however, real controls are also heavily loaded
(either naturally, or in the case of fly-by-wire, artificially), so
that they naturally hold their trimmed position and require a fair bit
of force to budge.  Linear control input on a regular spring-loaded
joystick or yoke tends to make a plane feel very jittery because of
the lack of control loading -- it requires almost negligible pressure
to pitch the nose up or down several degrees, for example, and seems
(to the user) as if the plane is insufficiently damped.  Squaring the
axis requires more physical input (rather than pressure) from the
user, but after a lot of experimentation, I find it's the closest I
can get to the feel of a real plane without building a force-feedback
system with big servos to load the controls.

The problem shows up especially when flying IFR -- even the slightest
jitter of the stick or yoke sends you flying outside IFR tolerances
without squaring the axes.

 > To model small nonlinearities that can occur w/ push rods and control
 > horns, I think an exponent smaller than 2 would be more appropriate, if
 > used at all. Instead of "squared", I suggest we add an exponent tag:
 >
 > <exponent type="double">1.2</exponent>
 >
 > which would lead to the effect:

We already have that, except that it's called "power". Take a look at

Input/Joysticks/CH/pro-yoke-usb.xml

for an example:

 <axis n="0">
  <desc>Aileron</desc>
  <binding>
   <command>property-scale</command>
   <property>/controls/aileron</property>
   <power>2.0</power>
  </binding>
 </axis>


Thanks. That's good to know. You're one step ahead of me!

I noticed the 'squared' effect when I started flying w/ my R/C joysticks. Using the actual joysticks exactly models what an R/C pilot feels, so no exponent fudging is desired in that case. Also, exponential rates and mixing if used are done onboard the R/C transmitter, so again nothing needs to be done on the fgfs side.

While the squared effect might work well for some airplanes as you describe, it might not be wanted for others. For instance, some fighter planes have side sticks that are very much like a simple PC joystick. In that case, the one-to-one linear effect would probably be the better approach.

In special cases (e.g. some airplanes in the current fgfs hanger), suppose one wanted to change a property that is set in the specific base package joystick file (or any other xml file). Can the related tags be added to the -set.xml file with the effect of over-riding the previous values? There have been instances where I have wanted to do this, but I don't think it worked.

Regards,
Michael

The "squared" element is still supported for legacy purposes.

Actually, to be most realistic, we'd add an ability to vary the value
of <power> based on the calibrated airspeed -- the lower the airspeed,
the sloppier the controls.


All the best,



David


--
David Megginson, [EMAIL PROTECTED], http://www.megginson.com/

_______________________________________________
Flightgear-devel mailing list
[EMAIL PROTECTED]
http://mail.flightgear.org/mailman/listinfo/flightgear-devel



************************************************** Prof. Michael S. Selig Dept. of Aero/Astro Engineering University of Illinois at Urbana-Champaign 306 Talbot Laboratory 104 South Wright Street Urbana, IL 61801-2935 (217) 244-5757 (o), (509) 691-1373 (fax) mailto:[EMAIL PROTECTED] http://www.uiuc.edu/ph/www/m-selig http://www.uiuc.edu/ph/www/m-selig/faq.html (FAQ) **************************************************


_______________________________________________ Flightgear-devel mailing list [EMAIL PROTECTED] http://mail.flightgear.org/mailman/listinfo/flightgear-devel

Reply via email to