I’m hoping to assemble a workable flight controller specifically for lawn rovers using the simplest setup I can reasonably assemble.
I was looking at this: https://www.waveshare.com/esp32-s3-touch-lcd-1.69.htm
it has a decent sized 16mb flash (maybe can use part for logging), 8mb ram, IMU and a bunch of other nice things like capacitive touch screen display and GPIO ports.
It seems I would need RTK GPS module (UM982?), Sik or other radio and an RC receiver.
of course it would need the esp32 version of ardupilot.
Anyone else have any input on this? See any issues?
When you can get a fully functional F4 autopilot for $20, it’s hard to justify a project like this. And I agree with @geofrancis about ESP32 based stacks.
I also wouldn’t use either of those options on a mower of any size, as I think there are significant benefits to using a more powerful autopilot for safety critical applications.
Thanks, which FC would you suggest? I have been using the holybro h765 wing but I see they are unavailable now. Plus it has a lot of stuff I don’t use and connecting to it is a bit messy. I’d like to take this from the “science project” phase to something a bit more professional. If there is a flight controller that can plug onto an IO board that would be great.
Current preference for advanced applications are any of those with onboard Ethernet, including but not limited to:
Holybro 6X
CUAV v6X
CUAV 7-Nano
Not sure what you mean by “plug into an IO board,” but all of these options are enclosed and expose click-type connections instead of just headers and solder pads.
Right now, just all of my telemetry (over a long range wifi bridge). Future plans include Ethernet peripherals and/or companion computer. Having wifi as the transmission protocol also allows for easy use of cheap IP cameras for monitoring.
Strangely, after all these years of Ethernet existing alongside such protocols as TTL serial, RS282, RS422, RS485, MIL SPEC 1553, CAN, and to a lesser degree, I2C and SPI, Ethernet seems to be the way of the future for wired peripheral comms. I’m seeing it in big industry just as much as the hobbyist side of ArduPilot. And I can’t say I’m upset about it.
Thanks, the holybro 6X like all their products is “unavailable”
And the CUAV devices definitely aren’t $20 controllers. They also show “out of stock”.
Even so with the pixhawk style controllers it’s a tangle of wires to connect radios, GPSs, relays, ESCs etc. Just researching what to buy and configure is a pretty daunting task. It’s all so “fiddly”.
It just seems to me the basic mower function doesn’t need so much complexity. Every mower I see on here is done from scratch and all differently.
It’d be nice to offer a simple, possibly single board solution to handle the specific mower task. It only really needs a couple servo outputs, relays for PTO and maybe lights, RTK GPS, RC control and telemetry. If that telemetry is over a reliable long distance ethernet connection that’s a plus.
You don’t need triple redundant IMUs and a barometer, but those aren’t really adding a ton of cost. I think you SHOULD have some processing power and a suite of IO options. The $20 boards fail in that regard, but so does any ESP32 ArduPilot implementation I’ve seen.
My point about the cheap boards is that they easily outclass the ESP32 hacks, despite the faster processor on the ESP32. So you could spin your wheels for weeks trying to gin up an ESP32 based solution only to have it pale by comparison to a cheap F405 board. That juice isn’t worth the squeeze to me.
You can get the Holybro products directly from them or from resellers. They just aren’t stocking them for regular sale on their own site for whatever reason.
Anything that meets any recent Pixhawk standard uses GH style connectors. How does a different board alter the “tangle of wires” dilemma?