Re: [Paparazzi-devel] =?ISO-8859-1?Q? Paparazzi joystick c?= ontrol

[martin]

Hi Sami,

the joystick code worked and we flew it (some time ago). Think it was
the funjetmm that still contains the setup. It was intended only to be
an additional control through auto1 as it did not have a high update
rate and uses normal (acknowledged) XBee messages. You also need the
joystick ground station. That could be Gautiers in
ground_segment/joystick or the older one in sw/logalizer/ctrlstick.c.

There is some code in in_progress/rctx that should be finished...this
one would use a TWOG/XBee connected inside a stripped (no 35MHz module)
RC transmitter. It would use broadcasted (non acknowledged) messages
that do not fill the telemetry link if connection gets bad. The same
mechanism could be used for a gcs code in the laptop and a joystick if
the rate is increased. This was tested on the ground and worked without
much lag. But needs to be finished.

What is missing is the code that puts the data into ap_state->commands[]
in a safe way and does the "RC_LOST" checks.

Martin

Am 20.07.2009 um  Uhr haben Sie geschrieben:
> Hi,
>
> I recently ran into problems with my traditional 35MHz radio equipment
> and haven't been able to overcome these problems. Then it dawned to me
> that why not use the datalink for manual control as the data modem I
> have on my A/C is far superior in every respect to the 35MHz radio.
>
> I noticed that some code already exists for the joystick control but
> that the code has not been designed to fully replace the traditional
> radio transmitter control. Well, I'm OK with that; I can later on
change
> the code if I wish. However, I haven't been able to get even the
> existing code for joystick control to work at any level:
>
> The PPZ Center gives me errors while compiling the existing Joystick
> code (although the airborne code compiles OK): undeclared identifiers
> joystick_x, joystick_y. These are the variables in the settings that
are
> supposed to (?) store the joystick's attitude before it's sent to PPZ.
> It's unclear to me where I'm supposed to declare the joystick_x and _y
> variables and, as I'm, also, unfamiliar with Linux drivers and Linux
> programming in general, it's unclear to me how the GCS is supposed to
> interface with the joystick hardware so as to give some reasonable
> values to these variables. Besides, where did the throttle go; it's
not
> in the settings file (joystick.xml) although there are references to
it
> in the airborne code. It seems to me as if the code is only designed
for
> simulation not flying the real deal. I'm OK with that too if only
> someone could explain me how I can make the GCS send the joystick
input
> to the A/C. A nice addition would also be to be able to interface with
> the joystick buttons in order to change the flight mode through them.
>
> I (at least) consider myself to be a fairly experienced C/C++
programmer
> so I'm not asking for someone to do it for me but, as it is, the PPZ
> code is very poorly (if not at all) documented so it's very hard to
> understand what happens and where and understanding even little bits
of
> the code is laborious at it's best so I'd need some help on a
> theoretical level: how is it (the joystick control) supposed to work
and
> how am I able to interface with a joystick in the GCS/Linux in
general?
>
> I'm fairly confident that, if someone can help me over the problems
with
> the GCS end of the code, I can modify the airborne code to be
> independent of the traditional PPM radio control which would IMO be a
> step forward in the robustness of the PPZ (atleast in my case).
>
> Regards,
> Sami Finnilä/Safi
> address@hidden
>
>
>
> _______________________________________________
> Paparazzi-devel mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/paparazzi-devel
>
>
>