Been working on this thing for a year need to share with somone.

My son and I have been working on this UAV for year. We started by designing a foam model and then glide tested it, then started adding to it, servos, power plant, APM 2.5, GPS, rudder yaw damper system. It's been a work in progress. The plane is a twin boom pusher we settled on this configuration because it seemed functional, looking around this site it looks as though we are not the only ones who came to this conclusion.

After what's been a year of development we now have a UAV that will autonomously fly nearly any flight path we give it. We have figured out auto takeoff, and it auto lands also. The auto landings are a bit firm because of my fear that setting a final waypoint to low may cause a CTAF situation. So I land firmly, I have some video of the last few auto landings here. https://www.dropbox.com/s/dtg6h1v3a20v8ty/Untitled.avi?dl=0 I'm not sure if anyone has a more experience with putting altitudes at the final waypoint such as 1 or 2m and been successful, if so I'd like to hear about it.

So we know have this aircraft that can fly a mission, be rerouted by the GCS during mission, take off, and land itself... The goal was to build this "thing" now that we have completed the goal, what can we do with it?

thanks for looking :-)

That would be great. I was able to fix up the plane and put it up tonight and splashed it again. This aircraft with the foam booms was 100% dependable now it flies well for a while, then I have not idea what it is doing. Last night the problems started when it registers a climb to 100m this was not commanded, but registers on the HUD, at the end it changes to manual mode and dives. Tonight it flew well then I lost communications with it, the drone does go into failsafe mode, but that didn't work properly either... it flew away and crashed. I am wondering if I'm experiencing brownouts on the board, you bring up the carbon fiber problem with radio waves that could be the problem also. I had 15 perfect flights in a row with this plane, I added the carbon fiber, 2 more batteries, and the power module that's when these current problems started. I have attached the the T-logs and I'll grab flash data and post in after I get it downloaded, I may only have flash data from tonight.

2015-04-01 19-46-39.tlog

2015-03-31 19-19-57.tlog

Replies

Scooter March 31, 2015 at 10:42pm

Update. To reduce weight so I could add another flight battery I removed the foam booms and used carbon fiber. All was going well. Little over 10 minuets in the air when the autopilot switched to manual, three seconds later this. I am not sure if it had to do with running the RDF900 using the 3DR Power module or if it was user error. Looking over the telemetry logs to see if they help pinpoint the problem.
- Developer
  
  Grant Morphett > Scooter April 1, 2015 at 5:12pm
  
  Any news? Your foam booms would of been both electrically and EMF inert i.e. cause no interference. Carbon Fibre is different and can mess with your radios. Did you have any antenna's mounted out on the boom? Feel free to post your logs (tlog and dataflash would be good) and I'll have a look.
  
  Thanks, Grant
- Scooter > Grant Morphett April 1, 2015 at 6:29pm
  
  That would be great. I was able to fix up the plane and put it up tonight and splashed it again. This aircraft with the foam booms was 100% dependable now it flies well for a while, then I have not idea what it is doing. Last night the problems started when it registers a climb to 100m this was not commanded, but registers on the HUD, at the end it changes to manual mode and dives. Tonight it flew well then I lost communications with it, the drone does go into failsafe mode, but that didn't work properly either... it flew away and crashed. I am wondering if I'm experiencing brownouts on the board, you bring up the carbon fiber problem with radio waves that could be the problem also. I had 15 perfect flights in a row with this plane, I added the carbon fiber, 2 more batteries, and the power module that's when these current problems started. I have attached the the T-logs and I'll grab flash data and post in after I get it downloaded, I may only have flash data from tonight.
  
  2015-04-01 19-46-39.tlog
  
  2015-03-31 19-19-57.tlog
- Developer
  
  Grant Morphett > Scooter April 15, 2015 at 7:24pm
- Developer
  
  Grant Morphett > Grant Morphett April 15, 2015 at 7:33pm
  
  Lots all my text due to inserting the graph!
  
  I had a look at your log and I can see from it you have FLTMODE_CH set to 8. So I graphed the raw input the APM see's from receiver for channel 8 along with your altitude (that's the GLOBAL_blah bit) above. You can ignore CHAN1, 2 and 3 for this. Green above is AUTO mode, pink is MANUAL. As you can CH8 spikes down and then goes back up to roughly PWM 1796. If you look at FLTMODE6 in:
  
  http://plane.ardupilot.com/wiki/arduplane-parameters/#flightmode_ch...
  
  you can see its in that PWM range (1750 to 2049). Again looking at your log file FLTMODE6 is set to 0 which is Manual.
  
  So, it looks like the mode switch on the transmitter was toggled. Hope that helps.
  
  Thanks, Grant.
- Developer
  
  Grant Morphett > Scooter April 1, 2015 at 6:54pm
  
  Hmmm, how do you have the batteries wired up? Might be worth bench testing. Plug everything in as if your going to fly then brace the plane and rev to max throttle for a few seconds and see if you get some odd behaviour like a brownout or something?
  
  BE CAREFUL WHEN YOU DO THIS - LIVE PROP!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
  
  Thanks, Grant
- Scooter > Grant Morphett April 1, 2015 at 7:17pm
  
  So I have tested the system live a few times. I have a habit of running a live prop often to see if a modification will work out, only been bit once.
  
  I didn't notice a brown out. As I think about when these flight start going downhill, it's when I start working in guided mode and sending fly here commands. It seems to fly a waypoint route well till I send it off the route, I then have a difficult time getting it back on the route.
  
  I did add additional carbon fiber in the form of two spars last night to support the wing. How could I tell if this was the problem?
  
  Flash data from both flights have been lost, not sure what happened to it.
- Developer
  
  Grant Morphett > Scooter April 1, 2015 at 9:40pm
  
  If you don't have problems at the start of the flight but do part way through it could be that the voltage on your battery has dropped so that when you apply full throttle you do get brown outs. In other words if your bench testing with a fully charged battery you might not see a problem. However if you bench test with a battery that is 3/4 drained as it would be at the end of a flight you might see very different behaviour when you go to full throttle.
  
  Do the issues occur immediately or only after you have been flying for a while?
  
  Thanks, Grant.
- julian vs > Grant Morphett April 2, 2015 at 10:33am
  
  if you look at the hardware status vcc in the logs its right around 4.8 and lower... that should be 5.2... I don't use APM anymore just pixhawk so I don't have a reference point... but I would try some flights using the old way to power your board (you mention that you had changed to the power module)
- Developer
  
  Grant Morphett > julian vs April 15, 2015 at 5:11pm
  
  The brown out voltage on an APM is very low - below 4.5volts. I know, I've tested in flight :-) Thankfully APM's are quick to restart. Pixhawk's however require a good steady voltage supply.
  
  Thanks, Grant

of 2

This reply was deleted.

ArduPlane User Group

Replies