Today is mostly about trying to plan the layout of all the small boards that we'll be using, so lots of hot melt glue will be in use and lots of placing components only to find that they wont fit and then having to relocated them.
Attempt one of who knows how many can be seen below.
The target for today is to get everything fixed down, albeit temporarily fixed down with hot melt glue. Then knock up a quick battery box so we can hopefully get this moving under it's own power again.
One of the trickiest parts to setup on our 2019 Pi Wars robot was the articulated steering, there are a lot of aspects to consider and very often if you adjust one aspect, it has a knock-on effect to rest of the steering components. It took a lot longer that I would have liked to get the steering dialed in, but we did manage to get it to a level that I was relatively happy with. So for 2020 we're not changing a lot in this area (if it aint broke etc), but there's always room for a little improvement.
We're sticking with the same size gears as we tried a number of different options before and these seemed to work out best for us, but we are moving the steering servo from the back to the front and now mounting it on same section as the wheels (the bogies?).
Relocating the steering servo has presented a few extra challenges though, it has altered the steering geometry and there seems to be a little extra play in it all, but we're hoping that the change in geometry will be sorted in the software tuning and the extra play will be solved with a redesigned servo mount.
When designing components, I always try to consider how these will effect neighbouring components and also consider if it's beneficial to integrate them in any way. By moving the servo mount to the front of the robot it makes sense to also use it as the front accessory mount too, that way accessories such as the Pi Noon adaptor or TOF sensors can be mounted and removed with ease.
Unfortunately the first servo mount I made for the 2020 robot was located in a way that seemed to allow the acrylic sheet to flex a little and it looks like this is where most of the extra play was, so the mount has been extended and reinforced in the hope that this tightens it all up.
Above is a picture of the 3D model and below is it being printed, in fact it's being printed as I write this.
We'll find out in a future post if this worked out or not.
Update: It printed just fine and looks like it will work quite well.
And here is it fitted into the prototype chassis.
It's solid now and has given a really firm base to fix all the front mounted accessories to.
The bulk of our code is Python and we'll use a number of different applications to write it, depending on what we're coding and what device we're on, but generally we use Visual Studio Code and if we're in the CLI then we'll often make code amendments in Nano.
As is often the case with writing software, there are many ways to achieve the same (or very similar) results and I believe this is often the case for hardware too. Just take a look at how vastly different the Pi Wars robots have been in previous years, but all setting out to solve the same challenges. For us, it's about sticking to what we know and what we can achieve with the tools we have on hand. We can't list every tool we use, well we could but you'd soon get bored of scrolling down the list, so below we've listed the main ones.
- Wanhao Duplicator 6 3D Printer
- Precision Gold A55KJ Soldering Iron
- Heat Gun, used for heat shrink tubing and heating up acrylic sheet for bending
- Vernier calipers
- Handheld acrylic sheet scribe
- Wire strippers
- Dupont terminal crimper
- Various craft knifes
- Side cutters
- Long nose pliers
- Hot melt glue gun
- Various files and saws
- Various screwdrivers, spanners and sockets
A small selection of the tools we use.
And so it begins...
The first step was to get our application form submitted before getting down to some serious planning.
The three of us have sat down and discussed what we've done well and not so well previously, and in doing so we've identified a number of areas that we need to improve on.
The big plan (at this stage, likely to change a number of times during the build)
- Testing, we need to make a lot more time for testing.
- Never take for granted how much (little) time we have left to finish.
- Ditch the idea of tracks and use bigger wheels, tracks gave us lots of grip when we used them in 2018, but things tend to get stuck in between the tracks and the drive wheels, causing the tracks to become detached.
- Increase ground clearance but without pushing the center of gravity up too much. Our plan has always been to build a robot as compact as possible, so simply piling components on top of each other is something we aim to avoid.
- Stick with the articulated steering, it takes a while to get it tuned correctly, but when you do it does make the robot handle pretty well at high speed.
- Make the wiring harness even more bespoke, we bought the tool for crimping Dupont terminals last year, to enable us to get connecting cables just the right length, but the wiring was still not as neat as I would have liked it.
- More testing, then GOTO line 8
- More testing, then GOTO line 7
So, where are we so far in the 2020 Pi Wars build?
Well, at this stage we're picking over 'neatly organised' pile of components and parts from the 2019 build and deciding what gets reused, redesigned or redistributed to a storage box. *Disclaimer: It will almost certainly change a lot and many times during the build, so it's extremely unlikely to look like this when "finished".
We'll continue to keep you all updated as the build progresses.
One of my favourite photos from Pi Wars 2019, we only missed one target here and that was 100% down to human error.