So with a week to go before the 2022 BBB Beetle Brawl, I’ve decided to get myself together enough and actually post about some of the upgrades I’m making to the bot before the event. I was mostly happy with how Babróg performed, but there was definitely room for improvement. Fortunately, Christmas saw me acquire probably the most useful tool I’ve gotten since my 3D printer: a 3018 Sainsmart desktop CNC machine. The working area is small, and it won’t do anything to most metals, but for working with plastic it would allow me to do things I’d never thought possible before, both for weight saving and general construction. With this in mind, my goals were to beef up the more vulnerable parts of the robot, and to find weight wherever possible to make those upgrades.


The CNC machine, with my "pragmatic" solution for cleaning swarf without an enclosure.

The majority of the weight saving was achieved by pocketing parts of the side armour, up high in the parts where the robot wouldn’t be hit normally, while also being an easy first job to get myself accustomed to the CAM workflow. From there I cut out new lifter cams and forks to replace the previously 3D printed parts, the latter using a new design entirely to be possible with a 2.5D operation. Finally, given how the old side bumpers had shattered after a single hit, I decided to try to make them from solid 20mm HDPE. The operation took a while to pull off and the finish is a little fuzzy, but the final product works, and should be far more durable.


On top is the new side armour, while the old hand-cut one is on the bottom. Both new sides save around 30g.


An assembly of all of the outer armour, including the side, top, and front pieces, along with the side bumpers.


The old 3D printed components next to their milled HDPE replacements.

Electronically, I began looking for ways to extend the range of my lifter servo. After not having any success with a servo extender, I decided to open the servo up again to get an idea of how I could manipulate it. Fortunately, I noticed that the potentiometer was only linked onto the gearbox (to get feedback on which position the servo was in) pretty loosely. This meant that I could simply position it into a position where it assumed the servo was at zero on the sticks, use some glue and tape to prevent it moving further, and leave it loose in the servo assembly, no longer connected to the main servo transmission. This way, it would be able to rotate infinitely in either way, and stay still when the stick was in a neutral position, essentially behaving like a normal motor, bypassing the internal servo ESC’s weirdness in how far the servo could travel (this also saved me from having to mess around with manually soldering resistors, something I’m very grateful for after my experiences with the original Léim).


The potentiometer. Interestingly enough, my 4mm motor d-shaft fits the potentiometer perfectly, which was handy when I was trying to get it centered to 0.

Finally and most importantly, I wanted to strengthen the wheels up to prevent the failures of November. My experiments with printing in tougher materials such as nylon would be costly and time consuming to implement, so instead I simply decided to look at ways I could improve the existing PLA design. As the problem last time was a result of the shaft digging into the plastic on the opposite side of the square nut, the simplest solution was to put another static square nut on the opposite side of the shaft. This one wouldn’t have a grub screw, but would mainly provide much more surface area and greater resistance in the most obvious direction the shaft would want to move towards. Prototyping was done on simple pulley assemblies before I felt confident enough to scale the design up further.


This is from the second batch of moulding I ended up doing, before this I had made four rear wheels for stress testing. This batch includes six rear wheels, two front wheels, and four lifter forks.

I finally gave the wheels a proper “torture test” last week, doing everything I could think of in my power to break them (pushing loads heavy enough to stall them, rapidly going back and forth, attempting to drift them, doing J-turns etc.). After about 5 minutes of hard running, the supporting square nut had actually managed to dig into the plastic enough to cause the wheel to free spin (22mm gearboxes and 1806 motors on 4S really are powerful!). This isn’t perfect, but I do feel better in the knowledge that:



  • They survived longer than the old wheels, and longer than a typical 2 minute BW fight


  • They were being pushed far harder than I would be driving normally, I’ve tended to use a driving style that “nurses” the wheels after my experiences with the Barróg’s fragile wheels


  • They had only three or four walls on the print before the gridded infill, since then I’ve made wheels with totally solid pulleys


  • This time I’m bringing spares, and lots of them. Even after losing this set, I’ve got four spare pairs of rear wheels, and a spare set of front wheels. This should give me enough scope for a deep run in the competition (provided I actually win fights this time, gotta break that 6 loss streak eventually)



You can actually see where the right shaft-side nut dug into the plastic. Given the surface area of the square nut this is really quite remarkable!

Hopefully these new wheels prove to be a success, if the design holds up better I may consider selling them in the future. My main reliability issues have been with the sheer power of the motors, but this should be mitigated with lower voltages or the more standard BBB brushed sets, though not without further redesign. I actually ended up purchasing a set of the BBB brushed sets with the intention of being replacements, and was surprised to find that the flat on the shaft was significantly smaller than on the Rotalin gearboxes (the flat starts roughly 0.5mm down from the diameter of the shaft on the BBB gearboxes, compared to 1mm on the Rotalin gearboxes. This is only really a concern for press fit wheels such as the design I use)


The new horizontal config ready to go, along with the standard setup in the background.


Bonus picture: this is how I've been adjusting my forks to be flush with the ground, using washers with base-mounted mounts. Keeping a flat edge on the floor reduces the chances of getting caught in seams on the floor.