Comments

• 100KM

  Ben Dellar August 10, 2016 at 3:04am

  At West Coast UAV we are using the companion computer with conventional RFD900s from the support aircraft to the GCS but are using 433MHz radios to relay telemetry from the support AC to retrieval. This avoids the frequency interference issues others have been seeing.

  It does mean the support needs to stay within about 1km of the retrieval due to the low range of 433Mhz at legal power but works smoothly in our testing (several long range missions operating 2x aircraft under our belt now).

  We are using slip to create a IP bridge over the RFD900s which allows us to send 2x udp streams down the link that we then separate out to the GCS for each aircraft.

  If anyone is pursuing similar approaches would be happy to share our config in a bit more detail if people need help.
• Chris August 10, 2016 at 1:09am

  Hi Tridge,

  'that doesn't work as I tried to explain previously. The problem is the two radios in the support aircraft interfere badly with each other even if running on different frequencies.

  If you use LBT_RSSI as I described earlier it can be made to work quite a bit better, but still not great.'

  
  Apologies I missed that bit but reading back I see that now.  I'm surprised that if their given different frequencies to use that they still interfere that badly with each other.  What if the duty cycle was reduced, I know this would reduce the throughput.  No way around the packet loss though on the serial PPP link.

  I think it was good that we aren't ready, I didn't get to do the radio testing I wanted due to delays.

  Thanks for your efforts, good to see someone helping out for the greater good.  Good luck.
• Developer

  Andrew Tridgell August 10, 2016 at 12:40am

  I think you might be using more in hover than we do in the whole trip. ;-)

  very likely!
• 

  JB August 10, 2016 at 12:37am

  Lol Tridge. You know me, not much a fan of running fossil fuels! I think you might be using more in hover than we do in the whole trip. ;-)

  Our revised D3 Video is here btw: https://www.youtube.com/watch?v=5n6W8VV-bPk
• Developer

  Andrew Tridgell August 10, 2016 at 12:11am

  @JB, we're well below that current draw, but some might claim that using a petrol engine is cheating :-)

  Glad you got D3 in! I look fwd to seeing you in Dalby
• 

  JB August 10, 2016 at 12:09am

  @ Thx again Tridge. Might have to give Seppo a call.

  @Chris and Anthony

  Shame you guys can't make it, but I'm sure there will be more.

  @All

  PerthUAV managed to get our Deliverable 3 in today. We're pretty happy with it.

  Out of interest is anyone else managing 4Amp draw at 17m/s cruise in forward flight on a quadplane? Logs get really long with 10Ah onboard! At least we shouldn't have any range anxiety! ;-)
• Developer

  Andrew Tridgell August 9, 2016 at 9:23pm

  @JB, you'd have to ask RFDesign about the RFD900x. I hope it will be released soon!

  It would be possible to modify the older SiK firmware to do relaying. Not really easy, but quite possible.
• Developer

  Andrew Tridgell August 9, 2016 at 9:22pm

  @Chris,

  Hi Tridge, as a work around for people that arent lucky enough to have a RFD900x, surely running two telemetry radios off a companion PC in the support A/C, one to the retrieval a/c and the other to the base

  that doesn't work as I tried to explain previously. The problem is the two radios in the support aircraft interfere badly with each other even if running on different frequencies.

  If you use LBT_RSSI as I described earlier it can be made to work quite a bit better, but still not great.

  using a serial PPP link?

  PPP does extremely badly on a lossy packet oriented link. I think you will find your effective throughput is close to zero when you are over a long enough range to start getting significant packet loss.

  Cheers, Tridge
• Chris August 9, 2016 at 9:09pm

  Hi Tridge, as a work around for people that arent lucky enough to have a RFD900x, surely running two telemetry radios off a companion PC in the support A/C, one to the retrieval a/c and the other to the base.  The companion PC would be connected to the retrieval radio and the support A/C's wired telemetry link to MAVProxy and have it forward them over the Support to Base telemetry radio (using a serial PPP link?).  Two telemetry streams would be tight but possible from what we observed.  That was our fall back plan.

  Good luck, I wish we had a few more weeks but sadly life is what it is.
• Chris August 9, 2016 at 9:02pm

  Hey,

  Sadly SydneyRPA's is also not ready in time.  We had some motor delivery issues for our Support AC.  Sorry Aussie vendors you made us wait 6.5 weeks for a DLE 30.....  I got it in 3 days from an overseas vendor and it runs great.  We really need another 6 weeks to make it work in time.  Oh well, I hope you all do well and I hope that either this or a new challenge is presented next year.

  I look forward to hearing all about it.

  Hi Anthony, it'll buff out mate.....