Racing the Sunset…
Double Sun Action…
Excellent news – following last week’s ‘unpleasantness’, it’s nice to have a working gimbal again and a new Devo 10 TX with all its switches intact. Quick ! Resume the testings…
Yes, let’s not forget I’m still in the test phase of the new build of my hexacopter, and this is…
Ding Ding – Round 3
For Rounds 1 and 2 we were in my back garden / ground zero test ground, and my tremulous early flights there involved almost no risk and very minimal height.
Now that I trust the set-up a bit more not to fall out of the sky, it’s time to venture a bit further afield so that I can improve the scenery I’m pushing past the lens, and have the freedom of a bit more space to play with, meaning I can test fly a bit more thoroughly / aggressively. I’m still not going to take big risks – there will be no overflying water or anywhere where recovery would be more than slightly inconvenient, but I am up for a bit of height, a bit of range, and I want to test that gimbal under a range of flying and atmospheric conditions.
I will be testing movement in all directions, at lots of different speeds, seeing how the craft behaves with ascents, descents, loose turns, tight turns, RTH, looking at GPS stability, gimbal reaction, gimbal battery times, gimbal RC control, and of course at all sort of heights where wind conditions vary from ‘none’ to ‘lots’…
I have fitted my FPV gear, but something wasn’t working and I had no picture, so I resigned myself to LoS flights, and that worked out fine in the end. I’ll solve the FPV thing by next time out. I bet it’s that hopeless FPV video cable that connects to the camera.
So first up, Petersfield Heath, 8.00 pm – just as the sun is about an hour away from the horizon. Not much wind, but definitely some up higher.
Here I’m testing movement – lots of low forward runs, sideways drags across the treetops, sharp descents, steep climbs – that sort of thing. Reviewing the video, the gimbal seems to handle all of this well, with the exception of really sharp turns and descents which cause some minimal blurring and disturbance to the smoothness, though less so than with any other gimbal I’ve used.
In general it did brilliantly, and I have no doubt that if I was really trying for smooth flying and immaculate control, I’d get an incredibly rewarding ratio of good, usable footage to flight time.
Having burned 21000 maH of batteries there, I jumped back in the car to go somewhere higher, windier, and where I knew the sun should look really good as it bid its farewell for the night…
Get up that hill…
I love Harting Down, which is where I headed next. It’s not only a hill surrounded on all sides by epic Hampshire countryside, but also one that has a car park right at the top of it and involves zero walking to take off point. Perfect for lazy, unfit pilots who were out gigging most of last night…
The one thing I didn’t do here was go flying straight off the edge into the valley, as my cinematic brain immediately volunteered when I got there. No – if it does go down, I’m not walking across half of Hampshire in the dark trying to find it.
So whilst I did restrict my flyings to the circumference of the large open field to the right, I did give it some fairly serious height at points, and the sun and clouds are worth a look in themselves, even if you’re not lost in wonder at how smooth that gimbal is, given that we’re on its default settings, and no tweakings have yet taken place…
Anyway – flights there were most enjoyable, and I met some great people – hello Dave, Dave, Hannah and Reg !!
Whilst I am still nothing like qualified to give this upgrade a thorough appraisal; I’m still on those horrible 11 x 4.7 props, which even being carbon fibre, are still depressingly unresponsive, I can still give you the few things I have noticed so far.
Hobbyking need to get their arses in gear and send me my new 11 x 5’s ! The slow-fly 11 x 4.7s are good for when you literally only want to really slowly sail-plane it around on a windless day. In any other circumstances you want 11 x 5’s, preferably Graupner’s, but Gemfan CF’s will do OK if the Graupners prove out of stock everywhere (not unusual). Today I’m stuck with the slow flyers, but I’m going to fly aggressively with them anyway !
Return to Home
This still works, and I tested it 3 times today (once to the amazement of onlookers!), but descent seems much slower now – perhaps 1 or 2 m a second. I remember the V1 being faster than that.
GPS Direction Compensation Tracking
OK, I’ve only done under 20 flights on this V2, but I have so far seen no sign whatsoever of this new feature working. The tracking on mine is odd. Not awful, but not as good as it should be given that my bits and pieces are lined up correctly, and my software settings reflect this. Instead of moving in a specific direction – forward, left, back etc, now I need to just look at the craft, ignore the directions, and just alter its course based on real-time feedback from my eyes on contact with it – like that it goes where I want it to. If, however I push just forward, I can’t trust it to not skew left or right at the same time. Interesting. Props ? Gain settings ? Or perhaps, as my GPS unit has been in about 10 crashes now, there is a possibility it has been knocked internally ‘squiffy’ somehow. I wonder about the Naza itself sometimes, knowing how precariously its own internals are suspended on some delicate looking elastic type lining. I dare not open either for fear of making things worse. Pretty sure the Naza is good on reflection.
I really like this feature, now that I’m expecting it ! Just raise throttle to 60 %, hold, and drop to 50% when it’s at the height you want. Easy.
Now this really might be down to prop wash off 11 x 4.7 slow-fly ultra-wide props, but the ground effect is massive, and descents very turbulent. I thought the Naza V2 would deal better with both these things. I expect it will when I’m back on wind-cutter blades.
General Control & GPS locking
Both of these things seem more unstable than with V1 Naza. It’s more like flying a large, dangerous verison of a QR ladybird, whereas V1 was like a nice calm puppy that obeyed your every command. This V2 firmware is trying to obey my every command, but GPS seems to have trouble keeping it in one place for long.
But not all the time. Sometimes it’s rock solid. GPS lock was attained in under a minute at both sites today, so the concerns I had in the back garden tests were unsupported. It also seemed to lose GPS in the air less than previous sessions, but also seemed to let it drift off course arbitrarily, sometimes requiring quite extreme correction from me to stop it entering ‘danger areas’…
Aeroxcraft in Action
Well – what can I say – still loving almost everything about this magic bit of gear. I could have made a video 20 mins long with the amount of usable footage I had left over after making this one ! It’s a lovely result.
Today I tried out the TX pitch control (default settings: +/-20% shift on Aux 4 rotary), and got to find out how the gimbal behaves when subjected to adverse conditions, namely a lot of wind, and purposely shoddy piloting. The gimbal setup certainly isn’t immune to shitty piloting – if you drag your craft to a sudden halt, or drop it fast through loads of prop wash, it’ll do a great job at correcting, but you will notice. If I want great results, I will have to fly smoothly and cinematically, which is fine with me.
During these tests I discovered that to my delight, the gimbal has a really slow roll recovery by default, which gives you lovely banking on the turns, even if the craft isn’t banking itself, and sometimes in reverse ! I love all that – it’s like filming from a plane, albeit one that defies physics when it banks the opposite way to the turn. No doubt I could eliminate it if I needed to in the gimbal controls, but I’m leaving it for now. The pitch correction remains lightning fast.
Before you can control the gains of your gimbal, you must first gain control of it ! This turns out to be an intensely tedious and drawn-out process, fraught with problems, and is perhaps the only annoying feature of this rig. You have to setup Arduino, the project programming interface, solve a plethora of driver issues to get the board to talk to it, then update the firmware, then finally you have access to the settings, once ! I have yet to see how much of that you have to repeat to get to tweak it again… needless to say – you need to be a tech-head to sort this out.
A bit too low…
Several times today in reasonably short grass, the camera was noticeably coinciding with the ground as it adjusted before launches and after landings. I must elevate the importance of getting skids sorted out, or make a landing box or similar, or my take-off areas will remain limited.
With my old XAircraft servo-driven gimbal I used to get about 1 min of usable footage out of every 8 minutes in the air.
With the new Aeroxcraft gimbal, even flying really badly with an unbalanced craft, I’m getting 6-7 minutes of usable footage per 8 in the air.
That is quite some quality leap and improvement to workflow. Also, thank goodness, there is no need to ever use Google’s de-shake ever again. Yay.
The sun sets on another day…
Well, that was 7 successful, rewarding and massively enjoyable flights, and I am beginning to regain faith that my machine won’t tumble from the skies in a repeat of last time’s power fail nightmare. The re-build does seem to have eliminated whatever was the problem factor, and I am now certain that all my motors survived that crash, as did the FC, receiver and all but 1 ESC.
Hopefully, this might be the end of the serial crashings…