Flatpack - 13.6kg BEV

I cleared out a lot of stuff (CNC router project, couple of printers, my airbrush) and so naturally had to fill that hole with more junk.

In my previous build spaces I’ve always had a drill press, mill or both. Bopping holes was never really a problem. Since I live in a flat now with no other workspace my options were limited. I could have bought another traditional bench top drill for ~£100, drive to Surrey and rescue my large rusty childhood one (and have bits fall off my small silly car) or buy a cute PCB drill named after a curry and press gang it into service for robot use.

Naturally I went with C.

It has so far been a good investment and has let me pump holes quite happily in the aluminium arm and the steel gears. The arm has had some work done to it. I have slightly bastardized the Ranglebox hubs. Instead of holding a wheel or gear to a shaft they’re holding in the 6mm pin that is driven by the front link arm.

The little teeny tiny bubbie drill did super well with some serious abuse straight out of the gate. This thing is meant to put 2mm holes in fiberglass sheet and I forced it to pepper my dubious Chinese gears full of 5mm holes and 8mm counterbores. It wasn’t pleasant but it jolly well worked.

The shoulderbolts arrived and I was able to properly mount up all the wheels. I drew and printed these cute little mounts which just support and space the driven wheels.

It was around this point (1 week to go! ) I managed to shatter one of the gears in the servo. Surprisingly it was not the plastic gear that went. It was the metal one just off the output.

Through a bit of spit and guesswork I worked out it was a 10 tooth MOD 0.6 gear. Luckily I’m a dab hand at bodging this sort of problem so I quickly had tracked down an RC car pinion exactly the size I needed. I bonded this and the other gear to the shaft with a liberal dash of loctite 638.

Dual servos were back on the menu! I had a huge headache trying to get them both matched in speed and travel. Though only purchased a couple months apart one was masses faster. Luckily they had adjustment knobs for travel and speed but it was time lost fussing about over them so they didn’t fight each other and eat more gears.

I also printed out a little covering cap that protected the bare control board and bolted up the locking bar mount tab thing. All the holes in this bent internal brace were countersunk at this point.

The wiring wasn’t too much of an ordeal but it wasn’t the most trivial task to keep it relatively neat. Wiring up two escs, a BEC and the servos somehow took an entire afternoon.

Batteries of shockingly dubious origin were acquired at this point. These are the absolute cheapest 5s packs I could find locally. On Amazon of all places! For £22 I got a pair of them. Low on capacity but sufficiently juicy with regards to current.

That is honestly pretty much it with regards to the build stages. Once it was wired and buttoned down it mostly behaved itself and I wasn’t having to solve awkward problems. I ran out of time a little but not too badly in the face of things. I was still drilling holes and such the morning of the competition which was fairly amusing.

On to the event!

I had a really genuinely chilled time which was fantastic. Weather held as well so I wasn’t having mild panic about water damage. Nice bit of chat with people too and there were some overly generous compliments about the robot (mainly the wheels!)

My first fight was against Barrog Doom in which I found myself stunningly outclassed. Having the drive in low was something done intentionally with the aim of reigning in oversteer and keeping it controllable for my dull reactions. It maybe was a bit too slow as I was outdriven at each turn and could never really get anywhere with the Irish behemoth. I felt that it was less of a fight and more Eoin picking the manner in which he won. Ultimately ended in a pitting.

CATastophe was next on the list which went the distance but ended again in a loss. Again feeling a little sluggish and that meaning the equally glacial weapon didn’t get to come into play. I think I was missing the momentum to push or bash and was losing out by being controlled and badgered. I was just unable to perform. Stupidly lost a wheel in this one due to the ole LH screw backing out under duress. It totally deformed the washer so it was done up with a certain degree of tightness. As it was the driven wheel that knocked the side of the drive out completely. Silly thing but easy enough not to replicate!

Third was supposed to be a rematch with Chris after our damp squib of a match from last year and we’ll be waiting again. He had unfortunately designed an automatic tyre removing device that doubled up as a robot and had run out of both wheels and inclination to continue. Instead I clashed with the wonderful Hover Bother which was about 70% motor by weight and about 60% batteries by volume. It is wonderful and quick and slightly erratic as a result. I had a bit more success getting in with the lifter but it was quickly struck with hammer blows from HB and something jammed. With an obscene amount of luck I was able to get HB beached on the ball pit (common phrase in combat robots) though I can’t say how much of this was down to me and how much was down to Hover Bother’s bombastic wall to wall arena tour.

Deservedly out of the running now I just went in for a whiteboard. The lifter was in a bit of a crunch state so I just blanked it from my mind and went in not expecting to use it. I put the drive into ‘high’ just to see how it faired. I was in against Pizza Time and Easy Over. Pretty well to be honest. It was a little bit chaotic from my end but it was more in keeping with what I want going forward. I tackled, slammed and missed with Pizza Time but felt much more involved in the fight than just a token trundling alsoran. I ended the fight upside down and smoking which is how I often end a Saturday night and begin Sunday morning. The more “drive it like you stole” attitude I adopted was not kind to the arena wall or two old, hardworking overvolted 775’s and I finished up the fight with one making a very suspect smell and smoking rather ominously. Cards on the table I am surprised this didn’t happen sooner as they did not have an easy time of it. Harsh overvolt, large wheels and a relatively narrow wheelbase.

Post match inspection revealed incredibly sad servo gearboxes (though the nature of the sadness is yet undetermined), a rather pungent 775 placed at a rather jaunty angle split off from it’s gearbox and a concerningly puffy over discharged lipo.

Overall I am pretty pleased with this build as it has been my first real scratch feather since 2017 I suppose so it was nice to get the feel back for it with my current year brain. The folded chassis worked super well and with the previously mentioned drive notes I am happy to continue down that road. New motors are a must and I will be revising how they mount to the gearboxes. I am toying with a printed clamp mount that an aluminium motor mounting tab interfaces with. Better than a few millimetres of rubbishy injection moulded ABS. The wheels were great and a lovely gimmick. I will be redoing these with a wider pulley so I can have a dual belt system to actually drive those pesky centre wheels and keep that beautiful 6wd look.

The weapon was a bit of a non starter but it was mechanically way more solid than before. Somewhere between the versions I think I have my sweet spot. Those servos will be junked as a stock item. I have some ideas as to how to retool the control boards but they are fairly limited to mild 550s. This is not a death sentence just relatively easy to outgrow.

Onwards and upwards I suppose.

5 Likes

So I’m rebuilding the CAD model as it’s actually a pretty good one to have as a show and tell alongside the lump of a robot. The steel frame makes it pretty industrial so it’s a nice example to wave under people’s noses.

It’s given me a decent push as I can look back and realize it’s not as awful as I remember. I really don’t have a good answer to the gearbox situation yet. There’s options for sure but they’re all going to require mega big brain solutions to get something slopped in there. I’m leaning towards some grungy power tool rejigged to fit one side. The battery placement sucked in this so not having dual motors frees up room for that.

Eagle eyed readers may notice the wheels are different. Thinner tyre with wider pulley to accommodate 2 shorter run belts so the middle wheels are directly driven instead of passively rolling. Tried my best to keep the vibe!

While the model and assembly drawings are more important right now (#Portfolio) I also dug out the boys to refresh the insides. The older slimline version is actually looking sweet. I had not cracked the lid since FarmFest '23 and was pleasantly surprised aside from the pair of gloves jammed in as shock absorbsion. That lifter mechanism is still not done justice. Honestly part of me thinks a sturdier gearbox housing and getting the Tiny Silly Pulleys welded to the damn shaft would help a lot.

The lid and the wedges are very poor and do the robot a massive disservice I think. Looking at it now with a single sheet of 5mm polycarbonate cut properly it could be a real winner. Getting a bit of the old classic Flatpack style back with bent steel could get an attractive wedge and break up the boxiness.

1 Like

I could not leave well enough alone and this is the result. I want to solve the problem on 6Pack because, quite frankly it is easier than trying to sort out the (decade) old lifter system on the brushless Flatpack. I wanted more meat and more speed than the 550 servos could afford me so naturally my mind drifted towards repurposing powertools.

Drills cop an unfair amount of derision. I get the bottom of the barrel single speed 36:1’s are pretty weedy next to anything but as soon as you get into the higher quality name brand stuff (that is, before cost cutting and MFG styles ruined them) I trust them implicitly. I’d trust a 90’s/00’s DeWalt more than anything in the banebots catalogue. This is from an AEG (?) I found in a skip alongside a similar but more slimlined 550 based chappie. It had a reasonable 775 as stock which I ruined taking the pinion off and had to replace with another from the scrapheap I call my life. It will be locked in low gear, which, with the additional stage on the arm will put the lift at about 110rpm. Faster and with a lot more beans behind it.

Now, if we ignore the slightly concerning rear pattern (the Freemasons/Musk/any available fascist would be gut wrenchingly proud) I picked it for its cost (£0) and because it had an intact spring clutch. This is a huge bonus as it will help the shock loading that the arm often sees and potentially be a form of limit switching. The output shaft is also super standard at 1/2" UNF so I can just drill and tap out the lifter gear and slap it straight on.

Amusingly I ran straight into the problem I always do with my Robortz which is extremely inefficient use of space. Long and the short of it, I can’t just slap the motor in without losing a wheel and cutting and ugly hole in the side of the robot to sling the drill in and out of. Coming swiftly to the rescue is RC pinion gears and 3d printed schlock.

Two 20t and one 45T MOD1 gears are used giving me a nice bit of reduction off the bat. This means too that I can dog leg the motor round, keeping them packaged in a less straightforward but more immediately useful way. Basically a less awful way of doing what I did with FP before. That used 6mm chain which was a huge pain (though fractionally cheaper). The use of gears and a much chunkier, more supported gearbox unit mean I’m very confident it will perform much better.

Lovely. It does actually fit. The centreline of this gearbox is 2.5mm higher than the previous one which is going to be a bit of an ugly ducking to retrofit but can be done without re-cutting and bending the chassis part.

4 Likes