UC4H FC IV: First successful flight!
16. Mar. 2019
MILESTONE reached by UC4H project
This is really a MILESTONE for the UC4H project: I did the first flight with the UC4H Cube carrier, and it was fully successful.
So, why I’m so excited about this? Well, it accomplishes a vision I had for a long while: A copter which only uses dedicated CANbus components. „Dedicated“ here means that each CANbus component is really made with CANbus in mind, and not just a kind of add-on. The flight controller is the perfect example of this. In my opinion the currently available Pixhawk-type flight controllers are NOT made for CANbus (see e.g. here). The UC4H Cube carrier however is, and you’ll see that by two examples in the below. Furthermore, „only“ obviously means that CANbus is the sole communication backbone; no I2C, no SPI, no UART, no ADC, no PWM.
Let me show some pictures. First, the copter. It’s the flamewheel test frame I’m using for a long while now, and which you can see in many posts below too. The AUAV-X2 I’ve replaced however by the UC4H Cube carrier, carrying a Solo black cube. Since it’s a test frame it’s a quite messy build, sure. Yet, you may notice how clean the connections to the flight controller are. No spider-web of wires going to it.
The cube was loaded with BetaCopter-3.6.7-v020u. I think you really have no clue about what a rich set of features a full UC4H copter and BetaCopter gives to you. I could produce a long list, which however wouldn’t tell you much, since – as said – you have no clue . I just like to give one example: ESC telemetry. It reports live data such as RPM, voltage, current for each ESC/motor. The data are stored in flash and available via Mavlink, and it is also fed into a ESC BattMonitor, which you can use like a normal BattMonitor. Well, you know this and can get this all with a BLHeli ESC too. However, here it is all via ultra-reliable CANbus and not failure-prone DSHOT (DSHOT is great on racers, but really not for UAVs). Some data plots:
Another feature I always wanted to have is recording of the CANbus 5V voltage. For a simple reason: In a CANbus system the CANbus voltage is as critical as the 5V supply voltage for the flight controller and I2C/SPI/UART/ADC/etc sensors or the voltage on the servo rail. Not surprisingly, both voltages are thus recorded by ArduPilot. But not the CANbus 5V. The UC4H Cube carrier overcomes this, by a „trick“: On a UC4H copter there is no servo rail anymore, hence I simply could use the servo sensor voltage pin to measure the CANbus 5V voltage. Et voila:
Have you spotted the two examples of „dedicated“ I’ve promised? Yeah, no spider-web wiring and CANbus voltage recording . There would be really so much more to say.
Finally, a word about the cube. I’m not impressed. I’ve never spend much effort into mounting the AUAV-X2 flight controller, I just sticked it with double-sided tape to the copter. As strange as this may be, it really worked not too bad, sufficiently well to not change it (this is just a test frame, right). The cube carrier I’ve mounted essentially exactly the same way, but now the reported vibrations are huge. I really would have thought that the cube’s internal vibration damping would at least not yield much worse results. But the data are as they are. I surely have to investigate this further, but the bottom line is: The cube’s internal vibration damping isn’t always a holy grail.