The ChainWheel | A Build Journal

[That_Jamie_S_Guy]

Hello everyone!

I am back, and this time bearing the birth of a new project. Before anyone gives me crap for failing to finish my all-terrain build, hear me out. As much as the idea of flying through the hills, shredding up muddy trails and bombing it through forest trails excites me, I know that unfortunately I just won’t be able to dedicate the time and energy to it.

A couple of years ago after meeting up with some ESK8’rs in London, I got a taste of the Onewheel and ever since then, the thought of building one has been on my mind. In fact, I struggled to make the decision to either build an all terrain ESK8 or a Onewheel, and despite initially settling on the first, I’m now going back the other way.

Anyone who has looked into doing a DIY Onewheel themselves on a budget will know, there are quite a few builds out there and some solid resources but nowhere near the level of documentation and support available for ESK8.

Almost every build I have seen uses hub motors of some sorts. Whether it’s a transplanted Hypercore, the SuperFlux, Phub-188, modified hoverboard motor, etc. Personally, I have a prejudice against hub motors. Sure, they earn some credits with regards to reliability, compactness, etc. But what if something goes wrong? Replacing an entire hub motor is an expensive ordeal compared to picking up another Flipshit motor. There aren’t as many options for hub motors out there either, so if I wanted to make any changes or upgrades down the line it would be difficult. I’m a real whore for modularity and repairability and being able to customise and tune things to my hearts content makes me happy.

So, how about the sensible option, i.e. belt drive? Well, @lock already blew that one out of the water with his awesome build that can be found here. I could follow in his footsteps but where’s the fun in that?

I have briefly considered designing a gear-drive system but let’s be honest, that would get bloody expensive and complicated far too quickly.

So, I’m opting to go with (as this thread’s title suggests) a chain-driven Onewheel! Sounds pretty fucking cool if you ask me. I am practically creaming myself over the mental image of riding on a glorified chainsaw.

Firstly, I’ll set a few things straight. This thing is going to be done on the cheap (where possible) and for that reason, I am limiting myself to re-using the 48 Samsung 40T cells, MakerX DV6 Pro and Flipshit 6354’s I ripped out of poor Nagini. Otherwise I’ll end up taking out a small mortgage for all the shiny new parts I want to buy.

Also, now that I’ve graduated and am officially a mechanical engineer, I’m now qualified to over-engineer the shit out of this thing. Stay tuned if you want to see enough numbers, symbols, statistics, equations, diagrams and graphs to send you into cardiac arrest.

I’ve started roughing up some concepts in CAD. Be warned - these aren’t pretty, dimensionally accurate, or complete by any means. But for the sake of keeping a log of my progress, here you go:

[trancejunkie]

bro your engineer now??? lol i dont think we can be friends anymore

[That_Jamie_S_Guy]

I will be starting this design off with the driveshaft. Over the last couple of days, I have done some preliminary research into the types of components I will be needing and trying to identify where I can cut costs.

There already exists quite a few different recommendations from people who’ve built their own OneWheels for what motor to use to get into the right ballpark of speed, torque, shape and size. But there’s practically nothing out there AFAIK for externally driven wheels.

OneWheels use the following dimensions:

It comes as a lovely coincidence that go-kart wheels have a very similar shape and size. This table from AMV Kart components sums up the common go-kart rim dimensions nicely:

From this, it is clear that the front wheels are typically in the region of 130mm wide (5.12 in for the hamburger guzzlers) which is about an inch shy of the OneWheel width. The diameter, though not shown here, is in the region of 4-6 inches and the tyre diameter boils down to whatever rubbers you choose to slap on the rim.

I’ve been chatting to a guy on the cesspit known as FB marketplace about a set of used OTK MXP’s he’s willing to sell for £10 a pop which is an absolute STEAL. These bad boys are made from a magnesium alloy, meaning they are lightweight and dissipate heat much more effectively than aluminium.

The keen-eyed among you may already be asking yourself how the hell these are mounted? Well, I present to you what is known as a hub. Rear wheel hubs look a bit like this, with some variation in the mounting hole pattern, bearing ID, and overall length:

They are designed to be restrained radially via a keyway, and the compliant clamping collar is torqued down to provide axial restraint.

Front wheel hubs are more like this:

They simply provide an interface between the shaft and rim through the bearings, with no radial or axial restraint. This is because most go-kart’s are rear wheel drive, so the front hub must be able to spin freely and independently of the axle.

This brings another architectural decision to the table - do I want to drive the wheel directly through the hub or rim like it is done on an ESK8, or should I go with a “live-axle” whereby the shaft itself is driven?

Driving the wheel directly would require a sprocket to be attached via. standoffs to the rim, which I anticipate may introduce stiffness related problems. Alternatively, it could be driven through the use of a custom-made hub but that would be expensive. Either way, all of the radial load transmitted through the wheel would be concentrated at the hub bearings, and the axial load at either the hub’s built-in clamp, separate shaft collars, or spacers filling the gap up to the rails.

On the other hand, driving the axle would mean mounting a sprocket to the axle and finding a robust method of mounting the axle at either end between the OneWheel rails. The added mass of the sprocket hub, any collars and spacers on the shaft, and the shaft itself would increase the inertia that the motor must overcome. Diagram courtesy of my epic skills using the drawing tool in google-sheets.

Although driving the wheel directly is objectively the better solution, given the space constraints I will be working with I anticipate it will be more complex and may even require custom machined parts. Quite frankly I don’t have the budget for that. I will go with the simpler option by driving the shaft. This means I can use COTS shaft collars, sprocket hubs, spacers, etc. which will keep the costs lower.

Provided I successfully procure the OTK MXP rim, my online sources tell me the pitch circle diameter (PCD) of the mounting holes is 58mm. To be honest, I’m hesitant to trust that until I get hold of the rims and measure them myself, so I’ll hold off on buying a hub until the rim job is complete .

While I wait for parts, I can at least do some preliminary analysis to get an idea of the forces I need to design for. Who’s ready to nerd the fuck out?

Let’s start by estimating the forces applied to the wheel. And to make life more difficult, I’m going to design this as if it was available to the general public.

In 2021, the NHS (National Health Service) released statistics in a report called “HSE 2021 Overweight and obesity tables” that details the self-reported mass of UK citizens of different age demographics. Using some basic statistics, we can work out the 95th percentile mass of our potential end-user here:

I don’t expect anyone over the age of 75 to be riding my Macgyvered chain driven death machine, but it’s good to know that they’ll be within the weight limit just in case. I prefer round numbers, so let’s take the rider mass to be 120kg.

Because I really can’t be arsed to read a bunch of literature on the mechanical properties of tyres and come up with a half-assed model that estimates how much force the tyre actually absorbs, we’re going to take the excessively safe approach and assume all force is transmitted through the tyre, into the rim, into the hub and onto the shaft. Realistically this is overkill as hell so if my loads look too high to handle, I can always estimate a factor to apply later down the line.

Stay tuned for some free body diagrams…

[That_Jamie_S_Guy]

Also, can someone talk me out of putting bindings on this before I end up back in hospital?

[Facko]

One remark from old engineer without dream​:face_with_peeking_eye:. Do not forget that chain will become longer after time. Consider this in Your concept. Idler will be needed from mechanical point of view. Not sure how big influence this will have on stabilization. Good luck and welcome in real world. Where managers would discuss with You about compromising solution missing logic and physical principles. Need change my job. Sorry

[zero_ads]

You can use those wing thingies for one wheels. Bindings not needed.

[Ash]

@That_Jamie_S_Guy

But there’s practically nothing out there AFAIK for externally driven wheels.

i think this is how future motion originally designed the proof of concept for the one wheel before they went to a hub motor

[That_Jamie_S_Guy]

@Facko

Do not forget that chain will become longer after time

That’s a good point, I’ll need some kind of tensioner. I’m already planning to have an idler so the chain sits lower and is less likely to shred my ankles.

[That_Jamie_S_Guy]

Yeah but I already have some bindings lying around haha

[zero_ads]

3D print?

[That_Jamie_S_Guy]

Not sure it would be strong enough? I suppose it’s worth a shot

[zero_ads]

A long time ago some dude created printable bindings when there was no stock of S2 bindings and TomiBoi toe hooks. Yo hook or something, probably can search it. But definitely doable.

[izzymonster]

I’m interested to see what you get worked up. I designed my own version on paper with the goal of reducing weight and increasing torque at the cost of speed for trial riding. I found it hard to find the right balance of gear ratio and kv. In the end, I didn’t think the weight and performance increase was going to be worth the cost of development.

I’m trail riding onewheel with bindings so it’s doable.

[Facko]

Or put motor on some linear guide with tension. Good luck .

[rafaelinmissouri]

Dude.
I really love my onewheels

Good luck on this adventure.

I hope your external Motor works well enough…

What are you using for an IMU?
Do you think it will run just like most Vesc onewheels off the Float Package?

[rafaelinmissouri]

The prototype from the Onewheel dude used a chain too!!!

We got the Original Onewheel to work again!

[CarelessAquarist]

It’s not realistically possible but it would be awesome to fit the motor inside the wheel using a large and offset rim, thin tyre and an internal gear drive.

How much power can OneWheels make before it’s more than anyone can use?
I have seen them with VESCs like a 30S Tronic X12 which seems ridiculous (126v >300A)
On a 1P pack as well it’s like voltage is god and they have no word for current but that’s a separate thing.
I’ve never ridden one but I feel like it must be hard to put down power when relatively on an esk8 the sky is the limit.

[ZachTetra]

That’s totally possible, you just need massive bearings

But it would be significantly more expensive and prone to damage

Also more current is better because it means the PID responds better, even if you only push 300A for a few ms at a time it will feel snappier than limiting to a lower current

[ShutterShock]

@That_Jamie_S_Guy

where I can cut costs

Ah. True engineering right there folks

[ShutterShock]

VOW has some printable bindings that he’s been working on for a while too