Thanks for reading,
Stuart
-----------------------------
As a bit of backgound, I want to make around 4-6 antweights plus an arena for my son's 8th birthday/Christmas present. This is partly so he and his friends can have fun fighting their own robot wars and partly to get him even more hooked on electronics and engineering. I'm hoping that once this initial batch of bots is out of the way we'll then be able to work together on more competition-oriented antweights (and ideally get his school involved, since they have a great ICT dept). Eventually I will probably invest some time and money in a 3-d printer but for now I'll make do with what's in my garage/workshop, which is set up for general DIY and carpentry, rather than delicate model making and electronics. I'm a natural bodger - prone to making things that are highly effective but usually inelegant and over-engineered, so the antweight rules are potentially my worst nightmare!
However... after reading through myriad posts on here and elsewhere, I started assembling the bits for a first prototype . Picked up a few N20 motors, wheels, cabling, switches, NiMH batteries, etc from various sources and also a Sabretooth 2x5A ESC (I know - expensive and too big/heavy for an Ant - but it was an easy option to start with and I may try building something bigger with it one day.) Next step was the R/C transmitters. After a lot of lurking and some sharp bidding, managed to pick up a couple of Spektrum DX6i's on eBay for reasonable prices. One of those came with an Orange receiver, so as a 'Proof of Concept Bot', I plugged that into the Sabretooth, wired up 4 motors and a 6v battery, bent some aluminium sheet into a rough box shape and plopped everything in. To my utter amazement it all worked first time. Incredibly gratifying to see PoC-Bot trundling around the workshop in response to the joystick movements. This was the first time I'd ever used a proportional controller (rather than those very skittish little 'all or nothing' cheap r/c toys) and it's amazing how much more responsive it is. Unfortunately PoC-Bot was more SUV than antweight - but it was enough to make me believe I might actually be able do this.
So, initial rush of blood out of the way, it was time to start on some more practical builds. Vague plan for the fleet is two flippers, two pushers, plus maybe a spinner and a grabber. With three bots bound to each transmitter that makes for plenty of one-on-one combat options for the kids. Nothing too destructive at this stage as it all needs to be safe enough for sensible 8-year olds to play with on their own (which is also why, for now, I'm sticking with NiMH battery packs instead of liPOs.)
Found some nice little Redcon 4-channel receivers on Banggood which are cheap, small, lightweight and seem to work perfectly with the Spektrums, so I invested in half a dozen of those. ESC options are less simple. Although half a dozen NanoTwo's would be the dream solution, it would presumably exceed Shakespearean production capacity and hardly be fair on anyone else. Of the various alternative ESCs I've seen recommended here, it was a nightmare distinguishing between the different versions of what may or may not be the same product being offered by myriad on-line sellers at wildly different prices. First up I tried a couple of generic single-channel 10A reversible ESCs but these were the 'with braking' option and I found that going from forwards to reverse required an awkward down-up-pause-down stick maneuver, somewhat akin to double-declutching - so that was a non-starter. I have a pair of the switchable Bustophephedons coming on a slow boat from China so will have to see if I get on any better with those when the brakes are off.
For now however, I'm concentrating on the dreaded DasMicro 256A. First challenge was wiring it up. Although I was very into electronics as a kid, I drifted away from it in my teens and consequently I've not done any soldering since the days when ICs were a novelty and staying within the tracks of a Veroboard was seen as 'finesse' work (also a time when my eyes were a lot sharper and my hands a damn sight steadier than they are now!) Tiny surface mount soldering pads on a postage-stamp sized double-sided board are a whole new ball game! Might have helped if the croc clips on the blue plastic 'helping hands' tool I'd bought didn't spring back half an inch every time I positioned a wire for soldering (can anyone recommend a good make of *really* helpful hands which stay where they're put?) Eventually, after a few retries and a lot of swearing I got it all together - and although the soldering looks hideous and I got two of the receiver wires back to front (it's OK as long as I remember to reverse the plug at t'other end), there are no bad joints or crossed tracks. Plugged it all together, taped the motors to a cork block, bound the receiver and tested with the transmitter. Everything seems to be working perfectly with no glitching or overheating. Incidentally, I did find one use for the little capacitors that ship with the DasMicro, even if motor sparking is unlikely to be a problem at this level; if you thread the cap's leads through the tiny holes in the motor's solder tabs (so the cap sits in the middle), you can bend the leads into little 'bunny ears', which are much easier for a novice to solder the heavier ESC connection wires to.
As an aside - I have two practical questions at this point which I'm hoping one of you experts may be able to help with...
1) Given Das Micro's reputation for burn-outs, plus my fondness for changing designs around, I'd like to make the motor connections pluggable instead of hard-wired. I have some JST (XH) male/female connectors pre-wired to 22awg cable. I was thinking I could solder the female side on short leads to the DasMicro and the male-end wires to the motor terminals. That way I can easily swap out an ESC if it fails. Question is, are these connectors up to the task or do I risk overheating when driving flat out?
2) Given the obvious physical weakness of surface soldered wires (especially when soldered by an old bodger like me), would it be OK/advisable to smother the connections (or even the whole board) with Araldite? Sort of like a poor-man's potting compound...
Anyway, with the basic electronics sorted, it was time to move on to some bodywork. Although I can just about manage cutting and bending sheet metal, polycarb seems to be the lightest option - so ordered some cheap A4 sheets of 1.5mm Lexan and dug out the aviation snips. Decided to start off by making a prototype wedge flipper. Never intended as a serious contender - rather just a chance to experiment with construction techniques - and given my lack of experience with polycarbonate, it soon earned itself the name "Flippin' Ugly".
I started out with a cardboard template - and once I was happy with the shape, traced this onto the Lexan and cut it out with the snips. Didn't have much luck bending it cold but gentle warming with an el-cheapo small heat-gun made this much easier (when I try this for real I might make a wooden 'last' first to bend the plastic around and hold it secure when drilling.) Because I didn't have any nuts/bolts smaller than M3, I decided to use some old-fashioned pop-rivets I had laying around from a previous DIY project. Although these may be frowned upon in proper metal-working circles, pop-rivets turned out to be surprisingly good for joining thin polycarbonate. Quick, cheap, easy to use and very light (aound 0.3 grams per completed rivet, compared with 0.8 for my smallest M3 nut & bolt).
Although the end results are grotesque and very uneven, it did only take a couple of hours to cut, drill, bend and join my first flipper body. Just to make it even uglier, I used an old brass cabinet hinge for the flipper, again fixed with pop-rivets. Dropping in all the electronics for a weigh-in, "Flippin' Ugly" tipped the scales at 128g, which leaves me ~20g for the top-plate, flipper linkage and servo mount. Now all I need is a few hours of free time to finish it off...
Lessons learnt so far:
1) 150g is a LOT less than you think (I may relax the weight limit for the Christmas present bots, though I want to do at least two that are competition-legal)
2) 1.5mm Lexan, though easy to work, lacks rigidity. I'm worried about how much of a flipper/weapon's kinetic energy will be wasted on bending the bodywork to which it's attached. For my next experiment I may try making a frame from 3mm instead as I seem to be able to bend it quite precisely with careful heating - and it takes little self-tapping screws beautifully (hard-drive mounting screws work just right). Better still I have loads of old sheets and offcuts left over from various DIY projects. I can then use 1mm Lexan for the skin.
3) Battery packs REALLY don't like being short-circuited, even if you do it accidentally. Forgetting that lesson (twice) was painful, messy and (with 4xAA Duracells) rather smelly. This is why I will NEVER cut through NiMH battery pack leads like that again and NEVER leave bare ended leads connected to the battery holder when I drop a half-finished experiment back in the box.
4) Leads of any sort in an antweight are always too long - up until the point when they're too short. There is no happy medium.
5) Without this forum and the generous info-sharing of all its member, I wouldn't have stood a chance of getting even this far!