We've been gradually accumulating bits and ideas for antweights over the last few months, with little to show for it, so Becky, Simon and I decided to have a day of building this weekend to chuck together something resembling a functioning robot. We arrived at the Hackspace around lunchtime yesterday with a sheet of 3mm polycarb, a box of HDPE offcuts and various electrical gubbins, and set to work. Here's my attempt!

Weapon: servo-powered lifting arm
Name: Twisted Lifter (pun on Twisted Sister, in keeping with the vague metal theme of our team)

I had originally planned to 3D print a chassis, but figured it would be easier to bodge something together by hand for a first attempt, figuring that I can always just transplant the electronics and drive to a new chassis later if I like.

With no real design in mind, I picked out a triangular piece of 5mm HDPE to use as the side panels and started work to fit in the N20 gearmotors. I didn't really have much of a plan here, just made each bit and then made the next bit, so after making the two side pieces identical in size and shape I drilled an 8mm hole all the way through both then spent a long time with a coping saw and a file making it into a 10x12mm rectangle. That was obviously not sufficient to hole the motor so I then made a face mount with some 1mm thick aluminium sheet and bolted it in place:


Note the most mass efficient solution perhaps, and the weeny m1.6 screws interfere with the wheels when they're pushed all the way on (d'oh), but they're pretty solidly fixed in place.

At this point I had two triangles of HDPE with motors attached, so it was a fairly straightforward job to stick polycarb panels on two of those sides. This is already visible in the image above, because I am slack at taking photos so I built half a robot then tried to take pictures of each bit.

Originally I was thinking that I would spend yesterday doing the mechanical assembly of this machine, i.e., physically stick everything in place, then leave the electr(on)ics for another day. However, by about 5PM I was bored of drilling and sawing, and also had something that looked externally like a 2WD robot (albeit with no lid). Furthermore, without having some idea of where the wires were going to, it was tricky to visualise how the electrics would need to be laid out, so I figured I'd have a go at the slightly daunting task of soldering some wires to one of the DasMikro ESCs. This proved to be fiddly but not impossible.

We then headed home and I plugged everything in to a battery and Rx, and twisted the ESC power lines around the motor lugs without soldering, just to give it a test drive. With a strip of electrical tape to hold everything down, in true sparky style, that gave me Twisted Lifter Mk0.9:



After re-binding the Rx, it almost worked first time, to my not-inconsiderable surprise. One of the motors was wired in reverse polarity (or possibly I had the fwd/bk and L/R channels reversed), but that could be rectified by the extremely shoddy approach of holding the Tx sideways. I had a brief test drive around the lounge, which Simon filmed for posterity. I then handed the controls over for him to have a drive, and in keeping with Simon's recent driving record it immediately went haywire and caught fire. I'm kidding of course, it didn't immediately go haywire and catch fire, it went haywire after about 20 seconds and caught fire about 10 seconds after that, I shouldn't exaggerate.

After legging it outside to avoid burning the house down, I initially thought that one of the hastily twisted motor wires had come loose and shorted out. However, there was no sign of any such issues and it seems like the DasMikro just decided to spontaneously combust (I understand this is a common feature?). Fortunately, the ESC was well clear of the Rx and battery, so the damage didn't propagate to anything expensive.

Here's a charry ESC with some scorching to the motor:


And a bit more detail of the motor:


Still, I had fun driving this thing around for a couple of minutes, and the rest is still in decent shape. A few things I did note from driving and in general:
  • The 300 RPM motors with 35 mm wheels were a bit sluggish.
  • The machine is too long (12 cm) and not really the right shape (the long pointy wedge isn't really necessary for a lifter).
  • With all the components I need and no top panel I'm already close to 150g.
  • I was a bit concerned about the teeny lipo battery. Do people use fuses in antweights? I can't really find a suitably tiny fuse. Would a bit of fusewire somewhere be a suitable approach?
  • In a similar vein, do people usually bother with indicator lights in ants? I see that switches are allowed instead of links in the RW101 rules, and there's no mention of lights or fuses.


To make it faster, I've ordered some 500 RPM motors which is an easy fix. Sorting out the size and weight is less obvious, but having something to start with helps. From the current design, I'll cut the pointy end of the chassis down a bit and mount the servo for the lifter in what is currently the rear panel (i.e., what currently looks like the front becomes the rear). That saves about 20g, which I can use on some more body panels and a lifting arm cut from polycarb. I may yet have to drill some holes to save more weight, but this is really just a first try so I'm not too bothered about butchering it (eventually I'm hoping to transfer the overall general shape to a cooler design, i.e., with more of a hair metal theme).

In visual form, the red line is where I'll chop off the rear end, and the green line is where I need to cut a slot for a lifting arm to poke out:


Hopefully this will be good enough to have a go at some competitions until I can be bothered with 3D printing! If anyone has any recommendations for less flammable ESCs that would be handy (we still have 3 DasMikro 2s6a ones knocking around, but I'm not impressed so far).