Continuing adventure with the micro:Bit and drones.
Beta Flight Software.
The flight control board has this open source software loaded in the flight controller. The drift I experienced can be due to the configuration settings of the software not being ideal. This proved to be the case.
The Pitch was off by almost 2 degrees and the Roll by almost 1 degree.
Also the CPU load was around 50%, the recommendation is to keep it below 30% or it could lead to instability in flight. I reduced the Gyro sample rate from 8kHz to 4kHz.
But I could not test any of this as one of the motors had a manufacturing problem.
Motor Wire not soldered. (see pics)
I was rerouting the path for the motor wires on the drone and one of the motors stopped working. Upon investigation the red wire easily pulled out of the motor. When I opened up the motor it was clear the metal connector was not soldered. There was also evidence of arcing on the plastic.
My repair job was unsuccessful, the delicate fingers that connect to the motor shaft had become bent. the motor measures 20mm x 10mm, the parts are really tiny. The motor now gets warm when connected but does not turn any more; it did turn once but not for very long.
The 3rd pic shows the connector and wire after my soldering job. You can see one finger is not aligned with the others.
I found what looks like the same motor on ebay; as a set of 4 ($14AUD) I could not find similar ones from Australian suppliers. Will take about a month to get here. Will update this post when I get them.
Update: Received the replacement motors last week.
They are the same as the ones in the project, so all good; essentially I bought the right ones.
I was concerned they would be slightly different requiring replacement of all 4 motors. After replacing only the faulty motor the drone flew ok.
In my opinion the faulty motor should be replaced by the manufacturer; but there is no facility for this through Core Electronics or the Manufacturer. Also, in my case I pulled the motor apart breaking any sort of warranty.
The default settings in the flight controller are not correct for level flight, run the CPU at too high a percentage and do not cater for the jitter present in the tilt control of the micro:Bit. To me this indicates NO testing of the product was done before it was shipped. (poor quality control)
Changing the settings allowed the drone to take off and fly level, sort of. It still drifts due to random air currents, especially when flying close to a heater. (rising air etc) It is also difficult to control; the tilt of the micro:Bit controller does not provide good human feed back. A joystick control would be much better allowing small adjustments to keep the drone where you wanted it. Trying to correct the flight can quickly get out of control resulting in a crash.
The micro:Bit programs provided by the manufacturer have a few minor errors and the code makes me think it is not as complete as the programmer intended. Probably it got to a basic working stage and they just left it at that.
The micro:Bit on the left is what came with the drone. Battery container replaced with one that has a on/off switch, plugging and unplugging the JST connector damages the connector and wiring eventually.
The micro:Bit on the right is one I have used with a Robot Cart, now modified to include a small joystick. I will be change the code to use this one and see how easy it is to control the drone.
Standby for future post.
Fantastic work as always! I’m disappointed to hear about the trouble you’re having with it, but keen to see whether your DIY physical control can do a better job!
Testing control via miniature joystick: Overall result, unsuccessful.
The drone moves too violently. To overcome the spring tension, the joystick moves further than you want which causes the drone to crash. Tilt control is more accurate. I suspect this is due to the type of joystick I am using. It is perfect for Robot Cart control and suits what I wanted on this particular controller.
Eventually I will buy a professional controller where the joysticks are designed for use with model aircraft and drones. The miniature joystick shown in the pic cost around $5, pretty cheap.
Drone Flight characteristics:
Further testing of the drone shows it is very susceptible to air currents and can easily get out of control if flown near a heater (rising air column). It is also affected by its own down draft when flying close to the ground, understandably. I probably need a large open space such as an indoor basketball court to better see how it performs.
The lounge room and study are not ideal, and I have been threatened with dire consequences if the drone damages any ornaments. LOL
Conclusion:
As a first programmable drone this has been a good learning experience. The device works and does what is expected compared to the sensors it has (none). But it is way over priced, if it came with 2 micro:Bits it might represent value for money.
PS The flight controller board and micro:Bit could be used in a different frame with sensors attached, and larger motors and props; I believe it is capable of being used this way. The current frame is fairly light weight, the props and motors are not very large. It would struggle to carry any more than itself. Might look into this.
This is awesome!! Have you had a chance to tune the controller? I have heard some good stuff about Joshua Bardwell in the quadcopter space.
Keen to see more!
