[simpits-tech] Need a plan

Matt Bailey mattb at rtccom.net
Wed Jan 19 23:41:52 PST 2005



Bubba wrote:
> 
> Matt Bailey wrote:
> >
> > I'd look into Phidgets. In fact the only real reason I haven't played
> > with them yet is I'm using Win98SE and refuse to "upgrade" Windows.
> > There is some third-party work being done for Linux support,
> > however.......if I can use Python in Linux to program a servo board, I
> > might have to try for analog instruments. :)
> 
> Oh yeah, forgot about those. I have XP sp2 on this machine (quite happy with
> it, too. It's by far the best version of Win I've used).

	I've been forced to use XP when I was at college (just about all the
computers were XP), but have since made a full recovery. ;) Win2k would
work, but I hate the thought of having to buy yet another Windows. And I
really don't want to get bogged down in VB coding. Python is wonderful,
I'd love to stick with that. From what I read on the Phidgets forum and
some related sites, it seems the Linux drivers allow you to program with
a number of languages, including Python. Very limited support of devices
though, at least for now. Frankly I'm kinda surprised Chester doesn't
handle the Linux support himself.......I guess he's just not a Linux
guy.

 
> I'm planning on using the same LCD for mine, and I may very well have to
> make my own faceplate/enclosure. If so I'll be doing it "right" and making a
> mold for either vacuformed ABS or laid up fiberglass. Either way I can make
> at least a dozen if you (or anyone else) needs one. I'm just going to send
> the video from an undocked GPS window straight to the LCD. I'm hoping the
> panel SDK will let me take the GPS display out of it's faceplate and send it
> to another monitor in it's own window. If that works, no problem. If not I
> think I can fake it by enlarging/recentering the display so that the bezel
> isn't displayed.

	Yeah, you should be able to spawn a view with an enlarged version of
the GPS. Wideview is your friend, I'd say. :) I'm in X-Plane though,
which has a very limited GPS, and only a handful of generic MFD's. I
actually had the idea of writing an MFD before considering the PS1
LCD......the thought came to me after learning some stuff in Pygame
(game-related Python modules, but suitable for other stuff as
well....it's actually a wrapper for SDL). It would just be really cool
to roll my own detailed, customized MFD, plus it gives me all sorts of
options on interfacing (switches, knobs, etc). I'll just have to have
X-Plane spit out engine parameters, XYZ coords, etc. What I may do is
set up my main machine for multiple displays, and run the desktop at
some insane res. My main screen will be at X-Plane's default res of
1024x768, in the upper left of the desktop (this is where X-Plane gets
drawn when it doesn't take up the whole screen), with my SDL window next
to it on the desktop. The various displays will be allocated to specific
parts of the SDL window. Only issue would be if X-Plane throttles
performance in other apps, my SDL app will be unusable. I may have to
get a second machine for this (actually I have a second machine
networked, but its video system has issues........with Win95 I can't use
3D apps, with Win98 the Windows GUI won't load). Anyway, for now I can
at least use the main screen on my secondary machine. I *may* be able to
do a multi-display setup with the old machine, in Linux, as it seems to
work fine there (I can use 3D apps and everything is fine). I'm thinking
for the MFD I might use a 7" screen. The 5" unit would be very nice for
a GPS (big moving map type, even), but perhaps a bit small for an
information-packed MFD.
	When I finally build a cockpit, this scheme should allow me to dedicate
a single screen to each glass instrument, as opposed to just placing a
large LCD screen behind the panel. This will be more flexible, and cool.
:)

> > But right now project #1 is my flight control setup:
> >
> > http://mattb.serverbob.org/pics/bungees2.jpg
> >
> > http://mattb.serverbob.org/pics/cockpit.jpg
> 
> Interesting idea with the bungees. Do you get much slop right around center?
> My plan is to use leadscrews and steppers to move the anchor points for the
> springs based on airspeed and a little math. The hard part will be the math
> to get extremely low airspeed right on the elevator. I want to be able to
> push the throttle forward and watch the yoke move back as the elevator sees
> some airspeed. The artificial feel units are easy. Just pneumatic cylinders
> with adjustable bleeds. No computer control since they're just simulating
> the friction from the cables, bellcranks, pushrods, etc.

	It doesn't really exhibit slop around center, but force does drop off a
lot. This is good.........it mimics the controls of a real plane
reasonably well in this respect, as the center is somewhat undefined.
You don't get the sharp, pronounced centering of typical PC rigs with
spring controls. I did make the bungees pretty tight between the bolts,
which helps. And control is positive, the linkages to the pots are
pretty tight. I'd actually say it's still tighter around center than the
average general aviation plane, but I like it.
	I thought about trying to mount the stop bolts onto some sort of
toothed rack, but I don't have any such device. But if I ran across one
(or made one myself) it should be simple to bolt it right onto the metal
case. This would be only for trim though. The system is too simple for
correct force feedback, and neither MSFS nor X-Plane even simulates the
needed stuff to really do it right. Adjusting force with airspeed would
be better than nothing....but for my simple setup, not worth the expense
& complexity. And I'd still be missing a lot of the effects that make
the controls feel the way they do. Even this bungee setup is still a far
cry from the complex feel of real controls for any given condition,
nevermind varying speeds and dynamic effects. But it's a WHOLE lot
better than a PC stick. :) I'll probably do the trim thing when I build
my pit, but doubt I'll work with force feedback any (unless experiments
with gas cylinders yield a sufficiently convincing feel to make it
worthwhile to adjust pressure based on aerodynamic conditions).
	Interested in hearing how your experiments go, however. :) The thing
with elevator weight shouldn't be too hard.......link a force with g
forces and apply that to the pitch force at all times. In the Decathlon
gravity actually causes the stick to remain aft when released at the top
of a loop! (Well, in my waaay slow lopsided loops anyway, hehe.)
	Overall I'm *highly* pleased with the stick, it really gets me into the
flying. I find myself once again enjoying just blasting around in a jet,
or doing aerobatics in an RV. I no longer have to be developing a plane
to have fun, and it doesn't have to be a really nice plane. I'll admit
after being in the hobby heavily for about 10 years now, I'm pretty
jaded......it takes a lot to interest me anymore. It's fun to be able to
take just the basic flight seriously again, on a regular basis. Hardware
development is definitely the next step for me. Trouble now is I'm
spending too much time flying the thing to get any more work done on it.
:)

 
> Gear is easy, I'm pretty sure FS9 has the ability to use individual gear up
> and down inputs. I'd really like to have the flaps on an analog input so
> it's possible to "goof" and put the lever between positions. I've been
> arguing with myself about the throttle quadrant. Levers vs. verniers. I
> personally HATE vernier throttles, though push-pull is ok. True verniers
> make rapid throttle changes a bloody nightmare to the point that I designed
> a gizmo that clamps onto the knob and keeps the button pushed for you. I'll
> probably wind up building the throttle quadrant as a module that can be
> swapped out for verniers/levers/single/multi/recip/turbine/etc. I'll
> definately have a 3 lever setup since I finally learned how to operate a
> constant speed prop ;)

	I originally was working on a lever style setup, but dropped that. The
complexity of the linkages was too much for the small box I'm using, so
I went the simple route. This way is probably better for a free-standing
unit anyway (I do have the unit Velcro'd to the stand, it's not really
usable if it's not secured). For my cockpit, I'll probably do the
modular route. Most of the time I figure I'll have one or two levers
alongside the left of the cockpit for throttle, and levers behind that
for other engine controls OR vernier controls for the non-throttle
stuff, on the panel somewhere. It is very cool to have all three engine
functions on analog axis, and handily adjustable with a quick push or
pull of the respective rod. I do need to make some handles, still.
	Actually for flaps I'd like to have a momentary up/down button in the
sim........i.e., flaps extend for as long as the button is held. This is
how Tigers are, I believe Merlins/Metros use this scheme as well. As it
is, X-Plane (and MSFS too, I think?) has discrete predefined detents for
the flaps. One tap of the down button moves the flaps down a notch. I'll
handle this with a 3-position switch (two positions plus a momentary
position) and "rapid fire" inputs. As long as you hold the down position
(or flip the switch into the up position), the Plasma will continuously
send button up - button down sequences. To move the flaps down a detent
you'll just tap the switch quickly. An analog control would be nice
too.......then you could do what you're talking about above, plus you
could simulate the mechanical flap systems in old 172's and other
planes. And you could use rotary switches to select specific
detents......the switch would be wired to an analog input, and each
setting would have a resistor on it, hence inputting a specific axis
position.

> Hehe, yeah, it would be fun, but Reno is 2 5 hr. days from here and Citabria
> and autopilot aren't words commonly used together ;) I'd really like the
> Navion we're considering since it's most likely an L-17B and even if it
> isn't it's still ok to paint it in military colors ;) I'm going to be
> instructing in whatever we decide on and I have a feeling I'll be able to
> gather a few more people with an extra line on my advertising: "Learn to fly
> in a warbird." :) Besides, the Navion meets the high performance/complex AC
> requirements since it has 225 hp, retractable gear, and a CS prop. And it
> has a sliding canopy <homer>mmmm... sliding canopy</homer>

	Oh yeeeaah, sliding canopies are kewl. I flew Tigers a lot in flight
school, fun plane with the canopy and stuff. It handled pretty nice too.
It would've been much better if it were a tandem two-seater,
taildragger, and had stick controls instead of yokes. Still not enough
to persuade me away from a Decathlon or similar, though. :) I remember
when I first flew the Tiger, thinking how light and responsive it was
compared to the 172's I'd become accustomed to. Man, the Decathlon makes
it looks like a DOG. :)

	-Matt Bailey


More information about the Simpits-tech mailing list