Rotavator - Half the mass is spinning mass

Ok so I know I still have more of my jackhammer write up to finish but I thought I’d make a start on this one so it can operate a bit more as a build blog as I work on it. Here goes…

This is my new beetle under the working name of Rotavator which I’m trying to make in time for the BBB new blood event in may. It may look like a pretty standard vert but I can assure you it is not. Lets start with the obvious stuff then. The drum weighs more than 800 grams and has a battery in it.

My initial concept was inspired by noob tube. I thought what if instead of having silly wheels on the ends you could just hold the spinning tube of death with a separate robot and maneuver it around like that. This would make it far more controllable whilst still keeping an insane amount of weight in the weapon. I thought this idea was pretty goofy and I hadn’t seen it done before so here I am making it.

PSA: I am absolutely aware that this will not be at all competitive but I’m trying to see it as more of a design challenge so I’m less sad when it’s awful.

Lets start with the weapon then, because I designed this before I even started thinking about what the rest of the robot would look like. Pretty early on I decided that it was going to be a hubmotor. Besides the fact that I’m going through a bit of a hubmotor phase at the moment I still reckon this is the only decently compact way of doing this design in the current form. That being said I still considered loads of other options including planetary gears and at one point I was also contemplating having a counter gyro spinning inside the drum itself. This is still something I might consider later (realistically it’s really not) but for now lets keep it simple (even though this is anything but).

I settled pretty early on having the hubmotor (which I still needed to design) on one end of the drum and then having the other end supported by some kind of bearing and shaft combo. I also gave myself my usual constraints of doing it on the tightest budget I could (given the whole student vibe), and doing as much in-house as possible. This limited me to my standard hdpe and 3dp along with 2D metal parts (no expensive china beaters) and for any cylinder-ish parts to be constrained by the size of my tiny hobby lathe (and my poor ability as a machinist). Discarding the repeat compacts I’d bought for Jackhammer and some of the electronics I had lying around I wanted to make this bot for under £100 (ambitious I know) including any spares I may need.

I experimented with loads of different designs for the drum, at first mainly pursuing a 2D beater inspired by the one on bulbaroar with the electronics housed in the centre. After that I moved onto one with discs which slid onto the 2d beater before settling on this current design. One disc houses the hubmotor and the other one is free spinning and turned by these 6mm shoulder bolts which pass through the whole drum. It sits on a 10mm shaft and a bearing. This leaves the centre void nice and free for my battery. The idea is that the outside shoulder bolts along with a bunch of tpu should protect the battery from anything too nasty.

So the hub. The hub was kinda tricky because on a brushless outrunner the magnet ring bit is on the outside and the electriccy wirey bit is on the inside. This is annoying because I want the electriccy wirey bit to be on the outside because that’s where the disc is (and the disc is moving with the battery) and the magnet bit to be on the inside because that’s where the shaft is (connected to the chassis for the rest of the robot). The obvious answer to this dilemma would be to use an inrunner motor but honestly I completely forgot they existed because I got so invested in designing myself out of this hole. Anyway this is what we ended up with as the design for the hub. We got this lovely 10mm shoulder bolt running through the centre and classic 2 part design but I’ve turned it kinda inside out on one end with a bearing on the outside instead of the inside. This means that the magnet ring side can sit in the robots framerail and the disc can sit directly over the top/around the big bearing which should make the load transfer better and not wonk the whole thing up in impacts. Words words whatever. Heres a picture.

The other problem I came across was that I couldn’t add any keying features into the hub since I was making it all on my hobby lathe not on a cnc. My solution was to attach an aluminium “transfer plate” to the back of the hub which would transfer rotation whilst the shock forces would still go through the bearing. Knowing this sport’s deep hatred for bolts in shear I aimed to use steel dowel pins to couple the motor and the transfer plate with a few m4 bolts with loose tolerances to stop the plate from popping off axially. I don’t love this solution but I really can’t think of anything better without getting the hub machined properly.


3Dp drill guide with tool steel drill bushings shown in background of second picture

I think now actually the limiting factor of the whole thing may be the donor motor I chose to design the whole thing around. It’s a 900kv 2812 which I now think may be too small for the huge drum. After seeing detonator using a pretty small motor on his MASSIVE drum I thought this would be ok, only thing is, he has a fat belt reduction and I don’t. Realistically though let’s be honest I was drawn in by the £8 Aliexpress price tag wasn’t I.

That’s pretty much all there is to say about the weapon for now other than the fact I’m currently procrastinating designing the tpu housing because it feels like a faff.


Here’s some nice pics of the hub components – note all the chatter in the part on the left – yum.

Turns out when your lathe is this size, it’s pretty hard to cut that mushroom bell profile thing. I had to position the tool on the other side of the workpiece and spin the lathe backwards in order to get the tools in the right place on the part.

Ok so time to start thinking about the actual robot! Since my predictions are that this drum is gonna weigh a metric butt ton I knew the rest of the robot was going to need to be pretty compact. Since I was pretty happy and excited with what I’d come up with for the drum I wanted to make a cool and somewhat visually interesting robot to go with it to try and do it justice. I spent a hot minute sketching some extremely generic 4wd boxy verts before I came up with this. I loved the simplicity and I loved the tractor wheels. I made a quick cad sketch of it and was sold. This was to be my new awesome robot.

features:

Tractor wheels – 65mm rear and 20mm front for that big disparity lol

Only 2 frame rails!!! (and they’re pretty small)

Chunky front side pod bits

Carbon fibre top and bottom for RIGIDITY

Tpu bum

2 completely separate control systems – drive and drum

Ignore the shelf of student things behind….

The most interesting part of this design to me is the drive system so I’m gonna start here. I loved how goofy these big wheels looked and so they were absolutely in. Due to my terrible history with 2wd stuff I really wanted 4 wheels and I love tractor style robots so enter tiny front roller wheels. These use some reject printer tyres I had left over from an antweight project last year and I’m pretty happy with them. The cool stuff all happens in the big wheel though. I wanted to gear drive these wheels because I had the repeat compacts, so putting a wheel the size of these big bois straight on the 4mm shaft would have been pretty structurally awful and probably simultaneously generated a land speed record given the speed the motors output. They needed to be gear driven but coming up with a solution that kept the robot compact was hard. That was until I remembered something cool I’d seen stratus do and use ring gears on the wheel to get the motor closer to the shaft. I played around with a bunch of variations and came up with this. I’m really happy with it.

The brilliant thing about having the gears on the inside of the wheel is that it doesn’t take up any more space width ways so the gears can be thick and fully protected without reducing the amount of tyre I can have on. It also means that I can have the belt on the outside nice and close to the frame, effectively occupying the same space in regards to the width of the robot to drive the other wheels. Finally, with the motor in this configuration, some of the weight of the robot should press down on the gears reducing the liklehood of them skipping. All in all I’m super happy with this setup as it’s sooooo space efficient. The only thing I don’t like is the possibility of debris ending up caught inside the wheel.

I know I should cast tyres but I really really can’t be bothered at the moment since these printer ones are so grippy and light. – I know that a shot from a spinner will likely cut the rubber and eject the tread but for now I think loads of superglue should be good enough.

The drive axels are somewhat experimental. I really don’t think I have the weight for shoulder bolts so I’ve gone for a carbon fibre tube for the wheel to run on. This is the same one I use for jackhammer’s shoulder joint and it’s held up pretty well there. The wheels are held on axially by a nylon bolt which runs through the centre of the carbon tube. I think this will likely not hold up to hits from horizontals ripping the wheels off but I’m hoping they should be strong enough to hold up against verts with the force going radially. The wheels are designed to be really squishy so hopefully they absorb lots of the shock. Each axel weighs like 3 grams so I’m saving the best part of 50 grams with this over all 4 wheels - definitely worth it in my books.


New robot frame Vs Jackhammer Frame


The drive motors slot up into the frame rails then lock sideways into place, held in by a little retaining wedge – again I’m really happy with this setup, as once the bottom plate is off, I don’t have to remove a single screw to get them out. A far cry from Jackhammer’s complete lack of serviceability.


In terms of electronics, I want to make the drive motors unpluggable so I can slide them into jackhammer, so the rest of the electronics for the main body are going to be held in this little brick thing that slots in the centre of the bot. Ill be able to just swap this whole unit out at events if anything goes wrong and then fix it at home afterwards. You can really see here just how compact everything is in here – you know it’s compact when your wires all have to be cut to the perfect length and precisely routed. Other than that, this houses everything but the battery. Escs, bec, receiver, switch and light. Pretty neat I think. Once that’s in position all the electronics fit up in here with this little battery previously run in stratus.

That’s pretty much all I have for now. Lots still to do on the drum itself and then obviously I’ve got to make it all in my signature black and yellow.

11 Likes

this makes yam look sensible in terms of weapon weight, love it! got any pics of the battery and wiring inside the drum? really interested to see how you crammed that all in there

the wheel solution is very neat too, really compact. id imagine you should be fine debris wise, would have to be seriously unlucky for anything to end up in there

2 Likes

WTF Rhys, this is AMAZING :heart_eyes:

That poor 2812 tho - jesus christ that is a lot of mass for a relatively small motor - really hope it works out for you tho :crossed_fingers:

Love this so much, the wheel design is so cool! id be intrested to see how long that battery will last big drums tend to have big draw!

1 Like

Rhys this is so cool!

Inside drum pics to follow! currently working on getting the switch and LEDs in with decent wire routing. As for the drive, I may be safe from arena debris but I’ve already dropped a few screws down there :rofl:

2 Likes

That battery runs drive only hence why its tiny. The one in the drum is gonna be a 4s 650 for now which I know is still small but I think it’s all I’ll ever fit in there. I think also this thing is unlikely to ever go a full fight length against anyone because it’ll probably explode itself by then. But only one way to find out…

1 Like

What kind of material do you use on your wheels? It does look really grippy.

It’s the part on a 2D printer that grabs and pulls the paper through. super cheap on Ali express and around 30A I reckon. I also talk a bit about them in my Jackhammer build log which uses the same things just a different size obviously :rofl:

Oh wow, I wasn’t expecting you could get them this big in diameter. Thanks!

For what it’s worth these are a really easy roller tyre option Mars High Grip 50mm Tyre (Pack of 2) – RANGLEBOX

Yep! sourcing my own just gives me much more control over the size and keeps my costs down

I’ve sent off for some Mild steel discs to be cut at uni to see if the little motor can hack it. I think if I move to a larger diameter motor I won’t be able to make my hubs on my lathe and I don’t really want to make the drum any much wider either. This means I’ve started considering a redesign for the hubmotor section. Anyone have any ideas as to the suitability of a 3115??


Frame rail left, Big aluminium housing/worlds fattest shaft surrounding the motor and finally bearings and disc.

This is what I’m thinking of now, idea is basically put the motor inside the shaft and have it sit pretty much unmodified in there. it will drive the drum via another “transfer plate” with some shock load distribution stuff to ensure everything goes through the bearings rather than the motor. Anyone have any initial feedback as to wether this would be a terrible idea?

Ok so it’s Thursday now and rotavator is supposed to be competing on Saturday. As of right now, the robot has never been put together and the weapon has not been spun up. Safe to say I have some concerns. That being said, this is pretty much on par with my usual levels of rushed-ness before an event. I assembled most of the drum last night and it’s coming in at about 850g of spinning mass at the moment. Lets hope we make weight!

In terms of the drum, yes I did completely redesign the whole thing because you guys said my hub motor wouldn’t be powerful enough. The new design I’ve come up with is definitely experimental (and maybe still underpowered) so I’ll give an explanation/teardown post event when I have the benefit of hindsight

Picture time:

Metal bits!! – have since been painted but only partially because I ran out

Non driven side of the drum (I know it has thin sections leave me alone I was running out of weight)

Absolute wire nest

Oo it actually looks kinda cool

Current plan is hope to god the replacement tap I ordered last night arrives and that everything just goes together and works (hmmm….)
In terms of spares I really don’t have much in the way of anything electrical (or most things mechanical) so I’m really hoping the drum doesn’t blow up instantly. If anyone competing at new blood is reading this, why don’t you go ahead and pit yourself before my super dialled in definitely working robot blows (you?) up.

7 Likes

Hi all, been a while since I updated this so will do a little rundown on what’s happened since the last post, what’s changed, and what I am currently changing.

This dumb thing did its debut at new bloods and really surpassed my expectations. Having only really had a hammersaw for the last couple of years roofing people at this scale was new to me. (I have previously partaken in this activity with my ant but it’s a whole different thing when the bots are bigger) I see why people never go back now lol.

My “most generic vert I could come up with ant HI-VIS and my less generic vert excavator can both roof stuff at the little scale but the little scale is LAME which Is why we need the bigger robots (foreshadowing for future hobbyweight rotavator??)

Ok so new blood this thing did ok it made the bracket I think idk but it got absolutely taken apart by k3. I think this was because

  1.   mark is a good driver and I am a bad driver
    
  2.   my robot weapon spin slow his robot weapon spin fast
    

Mainly I think it was down to reason 1 but in terms of actual robot upgrades that aren’t just driver upgrades I can only really do the second one. The weapon spinning slow was a good thing sometimes though because big bite = big roof and also means less gyro = better driving.

So for the new one I wanted to be able to spin slow when I wanted it spin slow and fast when I wanted it to spin fast.

Solution: bigger ESC and make it 5s

Next event I wanted

Better spinup

less vulnerable to horizontals

more weapon moi lol

2 Likes

Next event was the bbb pub one where I had to make the bloody thing out of plastic. Idk if anyones ever tried to rushedly make a 900g spinner designed to be made from metal out of hdpe and have it handle all the forces but apparently its hard. When I span the thing up in the arena during tech checks the thing span up really fast (good) but then lblew itself apart in the middle of the arena (bad). This was super loud and it gave everyone a bit of a fright. My hdpe counterweight had ripped itself straight out of the robot. Fixed it with some wood screws in a different orientation and the thing was fixed (kinda not really).

I mostly used the event to try out some stuff on the drive which worked terribly and then spent the rest of the event crying about my screw switches failing every fight.

(yummy hdpe bits) Heres some pictures of it for that event. You can see I’m transitioning to a different drum design instead of a drisc thing to like a two side bits, an impactor and a counterweight type of design.

3 Likes

After that one we had the most recent one which was capital punishment. I changed some stuff making the tooth harder to fly off since the counterweight came off at the last one didn’t change the counterweight since I thought it’d never hit anything so wouldn’t fall off (foreshadowing, it did.) the cool thing for this one was that I prepped 2 drums!! So when one shat itself after 2 fights I could put the other one on. heres the new drums they look cool asf don’t they.

Two small black and silver electric motors  AI-generated content may be incorrect.

Idea is the side bits are alu for weight and the the tooth and counterweight are hardox and are attached by super weak cheese steel bolts for extra fun factor.

I won my first fight against milo’s detonator (his link fell out) and lost my other 3 but they were all fun fights. Again really I was just trying to stress test stuff so even when the weapon went down I was just trying to ram face first into people.

I think the weapon proved effectiveness against dolos where I broke his AR600 weld thing that he was talking up beforehand so I have some cool forklets as a trophy (thanks morgan) but that was the fight where the counterweight broke off, cutting a slice through the rest of the robot as it span. After this, I hammered the counterweight back in and taped the robot back together only for it to do the exact same thing again against ice breaker. So backup drum it was for the final fight.

5 Likes

A black and yellow device  AI-generated content may be incorrect.

This is what the robot looked like btw

Final fight was against this overhead drill bot from a first time builder with magnet wheels which I thought was cool. He put a lifter on for this one to control me about which I think was the right strategy. For this fight I realised like 30s before I went into the arena that this was my backup drum and even though the two drums are meant to be the same they are in fact not. This one has all the parts with the rubbish tolerancing. Because of this, the weapon is not able to be friction driven and is instead driven by a single m5 nut which isn’t even a nyloc that I remembered to put on (fingertight) as I loaded in. this lasted like 1 spinup then came undone (who would have thought it) so i lost weapon. I ended up getting controlled pretty well and ended up on my nose. This is a failure mode that I have wanted to sort since the robots conception by getting the weapon to spin backwards but AM32 hates me and I just cant reconfigure my escs. I am have just ordered a new esc programmer we will see if this one works.

2 Likes

A small black and yellow machine  AI-generated content may be incorrect.

It does ‘the thing’ really bad because it has nearly a kilo of spinner on the front lol

Aside from these horrific problems, another one I had was that the drum less balanced than a brown bear riding a unicycle, so when you spin it up the robot just kinda vibrates and glides away in a random direction. (yay physics). Basically even though all the metalwork is balanced, the electronics and the battery are impossible to fully constrain so they just go off centre and then the whole your electronics are a centrifuge thing makes that problem way worse as it keeps spinning. Next version will need to address this.

Sorry ive done this in so many posts i think my images are too big…

2 Likes

Ok so were through to the present. I’m going to try and update this more regularly from now on but we’ll see how that goes.

New stuff that needs to change:

Stop it doing the thing

Make the drum actually mechanically driven you idiot

Make the drum not explode as much

Make the drum balanced

Magnets now that the drive actually works

Ok so lots of changes to the drum that we need to make. The hard one is balance. The obvious thing to do here would be to somehow prevent the electronics from moving completely and then rebalance the metalwork around it. You could probably do this either by filling the entire drum with resin which I really hate the idea of for so so many reasons. Other way is making some kind of integrated circuit board that does everything and you can constrain with bolts. That one sounds like electronics nerd shit and I am purely a mechanical guy so that’s not happening. Because of this, I will be attempting to implement some mechanical nerd shit instead.

The idea is to have some steel ball bearings sitting inside a circular channel thing on the centre of rotation. In theory, once the weapon gets up to a critical speed, these balls should align and automatically correct for any off centre the mass of the drum provided I have enough of them to counteract the off centre weight. The cool thing about this is that since this is dynamic self-centring, any damage to parts of the drum or further shifts in the electronics will automatically be corrected for.

If you’re interested, there is a steve mould video which does a pretty decent job of explaining the basic principle.

https://www.youtube.com/watch?v=T47s4L1Wje4

Theres not much space in the drum at the moment and I don’t really want to increase the size for weight reasons but thinking of implementing this here by the motor. I feel like there should be juuuust enough space for it in here around the motor. (annotation in red)

Either way this thing is gonna add weight so I need to find a way of taking some out as well as somehow making the counterweight more secure and less break-offable. I also have no idea as to how feasable this idea will be in practice…Wish me luck!

4 Likes