XAircraft x650 and aerial photography



Once, a friend of mine called me a photographer and asked a very interesting question: “What if I put some photo soap dish on a radio-controlled helicopter and try to somehow remove this soap dish from the air?”
And, I remember, from early childhood I dreamed about radio-controlled aircraft, but I lived in the so-called “outback”, so I didn’t go further than children's drawings in this matter then. This time I couldn’t answer anything straight away, but since the topic suddenly interested me (and who will not be interested in?), I promised to deal with the question and phone later.



From the very beginning of the study of the issue, I switched from conventional helicopters to quadrocopters, or rather AR.Drone, because it’s much easier to manage. But then it turned out that specifically AR.Drone could raise something completely light, and, accordingly, not very high quality. We even thought then about android phones, for which we could write a program and manage it from another phone via wi-fi.
Further searches led me into the jungle of DIY copters, which turned out to be at least an order of magnitude cheaper than AR.Drone with similar characteristics, and with a small overpayment they allowed me to raise much more interesting things, like mirrorless ones, for example.

As a result, after a couple of months of studying all this, we decided to order the XAircraft x650v8. By the way, it is strange that then we lost sight of the DJI F550 + NAZA, which is a bit cheaper, and much more stable in terms of reviews.
But be that as it may, with the x650 we spent almost a year, gained experience, bumps, broke a bunch of details and, I think, I could tell a lot about it.

Looking ahead, I will immediately warn you: if you are planning to take off from the air, or even control more or less heavy copters with a load, just think about FPV and telemetry in full. Save a ton of money because understanding the details of what is happening in the air at this particular moment gives +100500 to controllability and, accordingly, will more likely save a crazy copter without breaking any details and without frightening grandmothers with UFOs falling from the sky.

So, XAircraft X650v8:
  • 8 engines on 4 axles
  • Net weight: 1260g
  • Maximum load: 940g (we lifted more, somewhere 1200-1500g)
  • Flight time at 25C 3300mAh - about 10 minutes with a Canon 1100d + Canon 35mm F2
  • Price: ~ $ 750


The standard set does not include: a control panel and a receiver for it, batteries and a charger, a camera mount, optional compass and a GPS receiver.
When you have the parcel, you will see a bunch of fashionable little black boxes, and a million bags with details inside. Everything is very neatly packed. All bags are signed. There is a small supply of bolts.



There are just two tips.

1. All bolts should be replaced with ordinary cross. The set includes hex keys and I could not find a single key that would not be erased / broken after a month of use. Anyway, they are uncomfortable. You need 2.5mm and 3mm bolts. Pieces of 50 to stock. Fortunately, they cost a penny. But 1.5mm and 2mm hex keys are still useful, as there are very special bolts that you are unlikely to find in stores, I advise them (keys) to buy at least a couple of pieces at once.

2. It is possible to replace 4 carbon tubes with an outer diameter of 10mm. Those that are on the feet are primarily because this stupid carbon is crumbling, crumbling and breaking just with a bang and still leaves unpleasant splinters for any reason. In any plumbing or locksmith shop you can find aluminum tubes with the same diameter. 2 meters of such a tube cost about 30 rubles. Saw off the pieces of the desired length, and your copter gets +1000 impact resistance, especially in the area of ​​the legs on which it lands. The weight of aluminum tubes is not much more carbon.

By the way, we also attached foam tips to the ends of the legs (in the photo below they are wrapped with black tape for beauty) and thus got a much softer fit on hard surfaces without unnecessary jumps:



I would also replace the axes on which the engines hang (there are 16mm outside and 14mm inside), but I could not find such aluminum tubes. But on ebay you can find such carbon pipes on request “carbon tube 14 16”, they are much cheaper than “official” spare parts, unless they have to be cut and drilled.

In general, after meeting this copter, I simply grew an insane disgust for carbon.

So, the first impression, with the exception of the ubiquitous hexagons, was very positive. But then China began to show its true face.

For unknown reasons, the axes on which the engines are held could not be installed strictly in the same plane:



There were parts with under-drilled holes:




It did not work to install the protective covers on the engine regulator (ESC) holders - all the latches were successfully broken, after which we decided to simply wind them with black electrical tape:



The official website did not work (it worked in the middle of summer), but, funny, the update worked in the program itself. Well and so on.

Assembly process: He also has very fragile legs. While we were learning to take off and land, we broke a couple. For those who replaced, these places were immediately strengthened with tin strips: Since then they have not been changed. Fortunately, all this was not so critical as to break our enthusiasm and the helicopter was assembled in a couple of evenings. But here we again began to stumble upon oddities.











The main strangeness that I could not explain was keeping a horizontal position. The situation is as follows:
  • The copter gyroscopes are calibrated with a spirit level.
  • The joysticks of the console are calibrated.
  • Tips found on various forums about balancing screws using pieces of electrical tape were tested.
  • Tips have been tested about putting all kinds of pieces under the "brains" of the copter to change its horizon "hardware".


Despite all this, in the remote control you have to configure the so-called trimming about 2-3% forward, 20% tilt to the right and sometimes (!) up to 30% turn left, i.e. it’s as if we are forcing the remote control to always send the state of the joysticks a little non-zero. Otherwise, the copter will not stand in place at all. And then it will turn over during takeoff. And as I did not try, to make him at least a couple of seconds just hang in the air (without wind), without touching the joysticks - I did not succeed. I admit, I did something wrong.

And yet, sometimes I noticed a strange glitch behind it: after a long period of inactivity, when I tried to take off immediately after switching on, the copter tried to roll over on its side even despite the trim in the remote control described above. It only helped “turn it off and on again”.

In general, at the very least, but by the spring we started to fly a little. While additional parts for controlling cameras came to us, we hooked all kinds of soap dishes and mirrorless mirrors to it, set a timeout of 20 seconds on them, quickly took off, waited, lowered them and ... saw that something was wrong again.

Then we finally went broke on the Canon 1100d and the old Vivitar 19mm F3.8 FD wide-angle fix, which at first turned out to be pretty good. But then either the vibrations affected it, or something else, but the picture became completely soapy and we switched to the Canon 35mm F2, which was in stock.
Around this time, servo drives came to us and we, using electrical tape, wound one of them directly to the camera above the shutter button:



It was especially fun to do this, because at that moment we were in a crowded morning train to Vyborg. But the design was surprisingly reliable and we parted with it far from immediately.

A little later, such a thing came to us :



This is a relay designed to switch the signal from one input to two outputs. In general, its passport purpose is to change cameras in the FPV system, but first we used it to close the shutter contact on the camera. Then, when they decided that everything wasn’t cool without FPV, it fulfilled its intended purpose, allowing us to select a signal from two different cameras, and the first servo drive that came to hand in such a simple construction using foil began to play the role of closure:



Here it’s worth to step back a little from the topic and tell how our “engineering” idea went in terms of camera control.

Initially, for some reason, FPV seemed to us too complicated and unreasonably expensive (although as it turned out later (and more than once): the avaricious pays twice).
So, Google then prompted us a couple of interesting solutions for photographers.
One of them was Pixel Expert . The main charm for us of those that the arbiter promised was the ability to view the image directly from the camera’s matrix in live-view mode. Those. we would see exactly what we were about to shoot. Passport 200 meters according to the plan would be enough for our eyes and in general everything would be fine if ...
As it turned out after the tests, the range of reliable reception did not exceed a couple of tens of meters, or even less. In addition, the 1100d, despite the manufacturer’s claims, turned out to be incompatible in terms of live view. Communication with tech support was very funny and insanely absurd. Fortunately, there was a built-in camera on the Pixel transmitter, which we used for the first time. But then again, at very short distances. For insurance, a servo was hanging on the camera all the time, with which I could press the shutter button, at times when the Pixel lost contact.

This is how our copter looked with Pixel on board:



Having decided that it was somehow completely sad, we ordered an almost complete set for FPV and another control panel with a receiver.

This kit included:


An additional control panel with a receiver was needed so that the photographer could control the camera himself. In the future, in addition to the shutter button and FPV camera switch, there was a control of the tilt suspension of the camera, but we did not manage to do it before the cold, and buying ready-made was very expensive, sometimes like half a copter. It is completely not clear why.

We did not order a screen and glasses, because we decided to use a laptop with a usb video capture device and an existing Pixel, which had an analog video input on the transmitter.

In this case, with Pixel, everything looked something like this:



But even here, this Pixel constantly let us down. The fact is that the remote control also works in the 2.4GHz band and somehow it turned out that their frequencies coincided too much - next to the remote control, the Pixel receiver picked up the signal from its transmitter, at best, two meters away from it. But in general, it was possible to shoot.

And here’s everything except the copter:



On the left of the photo and on the right of the camera’s handle are the taped cameras.

About the process.
As we imagined it:
  1. We arrive at the place
  2. Choose beautiful shooting points
  3. We connect, take off, look in Pixel on what the camera sees
  4. As soon as we see the frame - freeze in place, take pictures
  5. We get down, check the pictures
  6. If batteries remain, GOTO 2


How it looked in reality:
  1. Have come
  2. Found a place where the sky is not completely hung with wires
  3. Decomposed
  4. Found that the Pixel receiver does not see the transmitter
  5. Found the position in which the pixel works
  6. Set up the camera (we shot in manual mode)
  7. Soared 10 meters, Pixel started to fail again
  8. I keep the copter in place, comrade shaman with Pixel
  9. The pixel is defeated, comrade says where to turn and how much more to rise
  10. I follow the instructions
  11. Comrade says: turn a little to the right and freeze, now I’ll shoot
  12. I turn a little to the right, try to keep the direction (it’s not clear with the height, I just don’t touch the joystick)
  13. The copter doesn’t care, it either carries it further to the right, then it turns left
  14. I inform my comrade that in terms of retention I do everything I can
  15. He tries to seize the moment and take off
  16. Hang out in the air for about five minutes, trying to control either the wind or the copter
  17. Get down
  18. We look at the pictures, in half the horizon is almost vertical
  19. Goto 7
  20. For the second time, the photos seem to be better
  21. re going
  22. If batteries remain, GOTO 2


Here it is worth noting that the copter itself in terms of brains was in the minimum configuration, i.e. without a compass and GPS, which at the time of order were sold only together and cost about $ 300.

The compass promised us, in addition to improved stabilization, features such as Head Lock and Care Free. The first is the retention of the direction, the second is the ability to rotate the copter around its axis without changing the direction of movement. It would seem that two super-necessary and useful features in our case.
By the summer, a new version of the compass appeared in online stores, and even separately from GPS. He spared no money and ordered with the fastest delivery.
But ... the very first flight with Care Free led to a crazed copter who refused to respond to any signals from the remote control. This, of course, ended in an epic accident with victims in the face of one engine, 3 propellers, an ESC holder and even little things. It could have been better, but the test flight was at the stadium and the copter flew at full speed into the goal post, caught on to it, and climbed inside, tangled in the net. The score, by the way, has long been no longer in our favor. From the part, therefore, by that time we had already stocked up with spare parts and the repair was not a problem. But the sediment remained.
Head Lock, too, did not arouse confidence and kept the direction even worse than the constantly taxiing myself. I would even say that he didn’t hold him at all.
In general, flight stability, if increased, is imperceptible.
After that, they decided not to order a module with GPS.

By the way, with a large camera on board the copter behaved much more stable, although it was clear that it works to the limit. In particular, the wind didn’t blow him as much. You can not say about flights, for example, with GoPro.

A few more words about the software, which looks surprisingly nice.

After connecting, we can download one of the many models that are updated over the network into the brains of the copter. There is still the opportunity to create your own, but it's really hardcore.



Next is the panel for calibrating the joysticks of the remote control (their status in the General Input frame) and setting some switches for all sorts of different features (Gimbral and AUX Input):



On the next tab, we can adjust the copter's behavior:



On the Gimbral tab - camera suspension settings.
And at the last - calibration and verification of gyroscopes, accelerometers and compass (the graph shows how I shook the copter):



Software, probably, turned out to be the most pleasant part of this copter, with one small nuance. If you disconnect the usb wire without pressing the Disconnect button, then the program hangs tightly - only rebooting helps.

In the summer, in the midst of flights, we often thought about changing the platform to the DJI F550 with NAZA. Especially after this video (attention at 2:00!).

But it was getting colder, and after another epic accident on one very wacky day, we were completely demotivated and decided to postpone everything until next year, save up for a more powerful copter model, a full set of FPV in two copies and all that.

Now we are considering between two options, equal in price: DJI S800 + DJI WKM and MK Hexa XL + MK NaviCtrl + MK GPS.
The Internet is generally inclined to consider the German MK much more stable and resistant to wind, but the S800 beckons with its folding design, theoretically more stable carbon out-of-box propellers and a very open approach, which manifests itself for example like this .

As for the x650, I probably could surely advise it only to FPV-shnikam. The copter is powerful - you can hang a lot of junk on it, not very expensive for its capabilities, well, a killer feature - it folds up and becomes very compact when you need to take it somewhere. For aerial photography, IMHO, it is not very stable. Even the videos from the XAircraft themselves (where the x650 is stabilized with GPS) in my opinion look somewhat less interesting than the clips from the same DJI.

I almost forgot, I am not a “pilot” with experience, so I can be mistaken in some ways. Here we describe our experience, the size of a year, which accounted for about 50 sorties with flights in completely different conditions.
I will be glad to clarify, give advice and generally constructive discussion.
And yes, I could write a lot more, but I can’t add up the agenda, but I’ll be happy to answer what I can in the comments.

Resources on the topic

Manufacturers:
XAircraft: www.xaircraft.com
DJI: dji-innovation.com
MK: mikrokopter.de

Miscellaneous:
multicopterwiki.ru
igorkandaurov.ru/wiikii/doku.php
multicopter.org

And finally, a couple of our photos:




Also popular now: