Thursday, 27 August 2015
Saturday, 15 August 2015
This video shows a bit of how I've been debugging this during development - as the soft UART takes up most of the CPU time, there's no time left to do anything sophisticated for logging... however right on my desk I have a high-speed data-capture unit, aka an oscilloscope! :)
Friday, 14 August 2015
I recorded these videos so I could more easily explain what I was talking about when I contacted the supplier:
Tuesday, 11 August 2015
Total damage: one carbon fibre arm, 3 props, a cable ripped from the motor, and a very bent LED bar. A sad day. Compounded by the fact that I bought this model because it was supposed to be easy to get spares, but seemingly all the UK stockists have no stock of any of the components.
Fortunately, I discovered that the manufacturers website had the parts available at a reasonable price, so now I just have to wait 7-25 days to begin my repair job. Oh well, I guess I need to hurry up and order the bits for other build that I'd always planned to do in the background...
The case seems to just be plastic and the PCB inside seemed fine, so I put the two pieces of case back together and taped them back together. I went out again and after about a minute of flying (on a new battery) the quad started diving left and down and then seemed to recover a bit and then down and left again before crashing and breaking another prop. I put it down to just a rookie mistake (after all, it was only my second crash).
I took it back home, recalibrated all the sensors (as the first crash was 200 miles away where I was on holiday) and took it out, again with a fresh set of batteries. Almost the exact same thing happened, which got me thinking that maybe the antenna was actually broken. I spent most of the day trying out different settings on the receiver for loss of signal, but each time they'd work the first time I switched off the receiver and then not again. It turns out that the receiver overrides the loss-of-signal settings in the receiver, so there's no point in following the instructions that came with the recceiver.
To change these settings, on the model setup screen (2/12 on the model), at the bottom in "Failsafe mode", select "Custom" and then "[Set]". These then work consistently over multiple power cycles of both transmitter and receiver. I set my defaults to turn on stabilised mode, turn on the beeper (not that I've got one fitted yet) and throttle at -20, which seems to be just below hover on a fresh set of batteries. I figure that will allow a gentle descent and hopefully let me get closer or a better signal somehow and regain control rather than having it plummeting to the ground. Of course, I guess with the props still going when it lands, it'll probably destroy the props and/or motors in the process but that's probably better than a half kilogram falling uncontrolled from the sky!
And perhaps the root issue was that the battery voltage on the transmitter was down to 7.6V. It feels like it should be OK at that charge, but after topping up the charge AND setting the failsafe I managed about another 30 minutes yesterday without any incident at all.
Additionally, the battery voltage telemetry is limited to voltage the receiver gets (which needs to be between 0-5V supplied by the flight control board) and without the external port isn't really of any practical use for determining the voltage level on the battery pack.
S-BUS does look interesting and there are SBUS-to-PPM converters, and the CC3D board does appear to natively support SBUS so I'll probably use that as the starting point when I make my own quad from scratch.
SPORT is the new telemetry interface. Getting my own telemetry data back to the received will be the first homebrew project I attempt (and I might as well start with a level indicator!)
I borrowed an old ATX power supply from James which got me by initially, but it's 11.3V with no load can easily dip below 11.0V under only 1.5A load and cause the charging process to abort. However, I've manaed to find a better power supply in another old PC in the loft that didn't go below 11.7V even when sucking 3.2A, although it's still not close to the 6A the charger is supposed to be able to deliver (and yes, the batteries claim to be able to be charged at up to 8A!) I'll try upping the power setting from 50W to the 60W maximum for the next charge.
Still, at least I know my charger isn't fake (I got a bit worried as it looks different to most of the B6 chargers available on ebay and hobbyking etc), as I checked the code on the holographic security sticker at SkyRC's website.
I've gone and ordered myself a nice ARTF (almost ready to fly) FPV (first person view) quadcopter and a FrSky transmitter. I''m still finding my feet and although practising with my Hubsan X4 has served me well, FPV is still a bit scary yet!