paparazzi-devel
[Top][All Lists]
Advanced

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Paparazzi-devel] Long Range Flying Wing


From: Stephen Dwyer
Subject: Re: [Paparazzi-devel] Long Range Flying Wing
Date: Thu, 10 May 2012 10:06:01 -0600

Hello Reto,

Thanks for the link. I added it to the Modem page. Keep in mind that
config isn't really optimized for low latency comms on UAS, but it
does explain most of the important commands.

In addition, I think the default modem setup is fine for communication
between two radios working on the assumption that usually only one is
going to attempt to transmit at a time, rather unlike the paparazzi
datalink situation. But yes, a few more example situations and setups
wouldn't hurt in the docs...

Thanks,
-Stephen



On Thu, May 10, 2012 at 5:44 AM, Reto Büttner <address@hidden> wrote:
> Hi Stephen
>
> Thanks for the research, the info here and on the wiki. Looking at all the
> work needed, the guys from MaxStream resp. Digi probably should improve the
> "plug-and-play" capabilities of the 9XTend ...
>
> Here are alternative instructions to configure the polling mode:
>
> http://www.digi.com/support/kbase/kbaseresultdetl?id=2178
>
> We intend our next tests with the new configurations by end of next week. I
> will let you know the results.
>
> Cheers,
> Reto
>
> Am 10.05.2012 um 02:15 schrieb Stephen Dwyer <address@hidden>:
>
>
> Hello Reto,
>
>
> We use 9XTend modems as well and used to have exactly the issues you
>
> describe regarding the downlink working (mostly) fine but not being
>
> able to change settings properly. I had discussed this with Chris
>
> previously, and he recommended polling as he did here.
>
>
> After playing with the various settings randomly, we found it
>
> difficult to find a good setup. So, we looked into the issue further.
>
> After closely reading the manual and studying the state flowcharts, we
>
> hooked up a logic analyzer to the tx and rx lines on the ground and
>
> airborne sides as well as to the rx led and tx power pins on the
>
> modems.
>
>
> We discovered that the default settings for the modems do not work at
>
> all. While in operation it appears as though the settings don't
>
> change. Looking at the logic output, one observes that more often than
>
> not, the groundside modem receives a command packet at about the same
>
> time the airborne side receives a telemetry packet. Each modem goes
>
> into transmit mode and attempts to send the packet, resulting in a
>
> collision. Under streaming or basic reliable mode (no
>
> acknowledgements), both packets are sent but not received because each
>
> modem is transmitting at the same time, and thus cannot receive. In
>
> acknowledge mode, the default settings have the same retries with no
>
> delay, so both sides continuously attempt to transmit but never
>
> receive an acknowledgement. Each modem receiving a packet at the same
>
> time seems to happen very often (almost always).
>
>
> Instead, each modem must be configured a little bit differently. We
>
> found the both acknowledged mode and polling acknowledged mode work
>
> quite well. Polling mode essentially eliminates all packet collisions
>
> but induces a bit more latency in one direction (from remote to base).
>
> Normal acknowledged mode has some collisions, but these collisions
>
> seem always to be resolved.
>
>
> I have posted some sample configurations here:
>
> http://paparazzi.enac.fr/wiki/Modems#Configuration
>
>
> I am sure other configurations work, and we haven't flown with the
>
> above configs, but it seems to work rather nicely in lab testing with
>
> a TWOG + Aspirin. Thanks to Chris for pointing us in the polling
>
> direction initially.
>
>
> Hope that helps anyone having issues.
>
>
> Thanks,
>
> -Stephen Dwyer
>
>
> On Tue, May 1, 2012 at 12:59 AM, Reto Büttner <address@hidden>
> wrote:
>
> Hi Chris
>
>
> We have problems over 25 km with 9600 baud. Something must be wrong in
>
> our setup. We use a dipol antenna on the plane as well, very similar
>
> to yours.
>
>
> Thanks for the tip to keep the modem body and the antenna in a
>
> straight line. In our current configuration we have a 90 degree bend.
>
> We will check on mounting it in a straight line.
>
>
> Cheers, Reto
>
>
> 2012/5/1 Chris <address@hidden>:
>
> I get over 25km with 115200 (haven't tried further) and a 8dbi panel antenna
>
> without any problems.
>
> At 9600 i have gone as much as 40km out and the signal was rock solid both
>
> ways so i bet it could
>
> probably go further than that.
>
> One thing to watch for is the airborne antenna, don't use the normal 900Mhz
>
> l/4 whip antenna,
>
> use a dipole antenna (l/2) or a J pole.
>
> If you use an antenna that can be rotated (swivel) keep the modem body and
>
> the antenna
>
> in a straight line, it makes a big difference.
>
> I use this antenna
>
> http://www.dpcav.com/xcart/900MHz-Antenna-Compact-Dipole.html
>
> You may also try a V dipole but not made from pcb (pcb antennas have a lot
>
> of losses).
>
> I haven't tried 9600bps and joystick control though only at 115200.
>
> As i said i had up link problems at normal mode but when i switched to the
>
> polling mode
>
> all problems went away, instant response to GCS and joystick commands with
>
> no visible latency.
>
> Nice plane you got i am sure that it must have almost no drag as it looks
>
> really slick.
>
>
> Chris
>
>
>
> _______________________________________________
>
> Paparazzi-devel mailing list
>
> address@hidden
>
> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
>
>
> _______________________________________________
>
> Paparazzi-devel mailing list
>
> address@hidden
>
> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
>
>
>
> _______________________________________________
>
> Paparazzi-devel mailing list
>
> address@hidden
>
> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
>
>
> _______________________________________________
> Paparazzi-devel mailing list
> address@hidden
> https://lists.nongnu.org/mailman/listinfo/paparazzi-devel
>



reply via email to

[Prev in Thread] Current Thread [Next in Thread]