I spent last week on holiday with friends in the countryside. Naturally I took the EasyStar as I planned to use my recreation time to get it tuned, proven and dialled in.
The wind was a bit high in the day times so all my flying was done at sunset when it calmed down.
On the first evening I worked on getting roll and pitch PIDs in. After giving my friend Thom (flamboyant launch technique pictured) a crash course in PID controllers and a tour of Mission Planner I sat him down on a picnic blanket laptop in hand and took off.
I was happy with the roll performance but I was surprised how much P gain it could take without oscillating. The pitch was acting very strange, it would pitch down quickly and easily but struggled to get its nose above the horizon in Stab/FBWA. We put in some values until it started to shake and then called it a day.
Looking over the tlogs back in the cottage it was obvious the pitch wasn't tracking well at all so I reset it all and planned to set it again the next night. I think the problem was a centre of gravity issue.
After retuning the pitch I also attached the GoPro under the wing. This was stuck on with a self adhesive mount in the same way I got the video in my previous blog. Launching is challenging with all the extra weight of a GoPro2 in its plastic case (200g) and you really notice it in how much extra throttle is needed. It didn't achieve enough speed and glided straight down to the floor ripping the camera off in the process.
That was frustrating but I was still able to spend the time tuning the L1 Nav controller and recording some baseline data to plug into the TECS system. Mission Planner had the wind blowing 6-7 m/s and I wouldn't be comfortable flying in anything more. Flying a square it would whoosh downwind at 20+ m/s and crawl upwind at little over walking pace without overriding the throttle. The base leg was a nice arc but still, it did a remarkable job at hitting the waypoints I set. It managed this by flying in some very unlikely looking attitudes but its track was great.
On the final day the wind was dead calm, you couldn't really ask for a better evening to fly. I had some extra adhesive mounts delivered and gave them a good 48 hours to go hard. Looking forward to finally getting some footage we launched. Only for the plane to sink and rip off the camera.
Deeply frustrating but on reflection I would rather the camera came off 5 metres in front of me than in a field somewhere far flung.
Flying continued and I was able to put the final sink and climb rates into the TECS system. Loiter worked a treat with wonderfully satisfying round traces on the laptop, and flying a 3D circuit was even better. I set the WP radius to around 24 based on the "double your airspeed" tip from and the corners were sharp as a tack. I think I could lower the L1 period even further too.
It was so calm I even practiced some manual mode flying. This being my first ever RC plane I have jumped straight into the deep end and the ArduPilot does a great job of covering over my imperfections, but I do try and practice where I can as one day I might need to bring it in the hard way.
I configured some 20m low passes over our head to simulate the kind of flying I will be doing when gathering video, it also enabled my friend to snap some pictures of the plane in flight. I would have gone a bit lower for her but there were power lines in the field so I thought better of it.
At this point a man walked into the field and started striding towards us. He had the look of a farmer who wanted us out of his field so as he approached I asked "Is this your field?". He replied "No, I just saw a little plane flying over my house and wanted to know more".
Turns out he is a pilot himself (full size) and was fascinated by the little foam plane. He pointed out some of the cloud formations around us and helped me find some thermals! I explained the autopilot and ground station setup and offered to show him some of the capabilities. I now have a lot of confidence in the plane so it was an opportunity to try some new modes. We right-clicked the map and it put it in to guided mode and it duly obliged. I turned off the transmitter and it returned overhead (after a short delay, interestingly).
Light was fading so to finish off the last battery I sent the plane further off than I have before, with a circuit taking it out to 500m at its furthest point. Telemetry was rock solid despite me doubling the data rate this week to get more useful tlogs.
Next weekend i'm taking it out on its first real mission. I've been looking forward to videoing my girlfriends coastal holiday home for months and this weeks tests now give me the confidence to set the plane off to do its work. Hopefully with a camera attached this time!
The mission profile is (relatively) tricky due to the area of interest being slightly away from anywhere big enough to land, being perched right on the coast, and with reasonably high trees in between possibly giving me line of sight issues. I will post more on this later in the week as I would like to get peoples thoughts on the best way to proceed.
My tlog and final tuning params are available here for people to have a look at if interested, although the log file is pretty "messy" when viewed on drone share.
