Progress had been happening slowly since October last year. We had…
- Two, part-working, remote control systems.
- One, working, Pololu motor driver.
- One, in-progress, custom-built motor driver for encoded motors.
- One set of non-encoder motors glued to a piece of wood.
- Some code???
So, at the end of January, we decided to finally get the whole team together in the same room, and spend a couple of evenings making some proper progress!
Here’s what we got up to…
Two, part-working, remote control systems:
We had two competing systems: Paul’s, using a Bluetooth games controller, and Sarah’s, using a Taranis radio controller (via an Arduino, and already one Pi down…).
The Taranis system was working perfectly* in the morning, but as soon as the time came to demo it, the Raspberry Pi crashed. Then crashed again.
* this is definitely a lie, but it was mostly working!
Until the still-undiagnosed crashing: the Arduino took in information from the radio receiver, translated this into information the Raspberry Pi could understand, then the Raspberry Pi gave this information to the motor driver, and the wheels turned in the right direction.
Code can be found here: https://github.com/SkyFawkes/drive
It was a fairly long winded progress, that needed refining, but it ended up on the back-burner as Paul’s Bluetooth system was still operational, so sticking with that made more sense!
So we were down to one mostly-working remote control system.
One, working, Pololu motor driver:
Why use a thing that works when we can get experimental?! At least we have a back-up!
One, in-progress, custom-built motor driver for encoded motors:
Helpfully, the wiring diagrams were incorrect – the motors were wired backwards compared to the N20s, and nothing like how they were described online – but after a bit (a lot) of faffing around, we went from “is it working?” to “yay!”.
Project files & code:
One set of non-encoder motors glued to a piece of wood:
After the Taranis-controlled Raspberry Pi crashed, we were left with literally just motors…
Luckily, Sarah cannot stop buying robot parts, and so had some spare electronic speed controllers to attach to the motors. This meant they could be controlled without the Pi, giving us a base to start designing the robot on!
We started with:
— Sarah Townson (@Isioviel) January 17, 2019
(sorry for the Twitter link, we can’t add videos to a free WordPress account!)
Then, got some design inspiration from the stuff lying around the office:
Noelle gave Sarah some design tips, and we started work on the main design!
Jonathan and Paul had been working on camera control for the Raspberry Pi challenges. We discussed how everything should link together and made plans for our next steps.
We still have plenty of work to do, but we’re happy with the progress we made, and all left with jobs to do.
In the time between the meet-up and writing this blog post…
More than one nerf gun has been butchered, version one of the design has been built and driven off the table (to test the glue, obviously), and the code has begun to take shape.
Write-ups and code breakdowns will be blogged soon!