Just on the tire thing. I know you already have tires, but craft and ride in the past have sent me XR tires for about $20-30US (can’t remember exactly cost) to NSW. They currently have a promotion (which may have ended in the past few days) offering free worldwide shipping and they have that a few times a year. I think shipping costs may have gone up since USPS stopped shipping to Australia and they’ve had to use UPS and FedEx
1 Like
Good to know. I’ll keep an eye out for this next time around if things don’t work out with the current tire setup.
In my last build, I made the enclosure before the battery, and then predictably the battery didn’t fit without some workarounds. Was keen to avoid that this time around so had the battery made up before building the chassis which needs to contain the battery on a one wheel. Specs wise I wanted to go above what you’d find in a OneWheel and settled on a Molicel p42a 12s3p pack, of course then the GT was announced with a 18s2p pack. Same number of cells, different layout 
.
Getting the pack built has been the easiest part of all this. Sent the following pic to @glyphiks .
And got this back just under a week later. Thanks again!
It then sat in a box for six months, but at least I knew how big it was and could start designing the rest of the board around it. Got lucky and found a FlexiBMS Lite in Australia of all places, and finished up the pack.
The pack was already built when I decided that I needed a wider wheel. This meant the board had to be wider and the battery pack width would have never been a problem. In fact, I think a 12s4p would probably fit, just not sure where the BMS would go. I’m really not sure a one wheel would ever need a 12s4p, but damn I’m curious now 
6 Likes
I’ll go with “bigger is better”
Usually I’d agree, but the board is very close to balanced right now. As in even with it turned off you could probably get it to balance. Really trying to avoid sticking more weight in the front where the battery is 
Don’t think it’s too much of a big deal if the front is heavier, you’d just have to push the rear down with your foot. More that’s it’s not ‘normal’ for them to sit that way.
It’s kind of important that the belt doesn’t skip in one wheel applications. I’ve skipped a HTD 5m belt before during heavy braking (and that was with two belts), so thought an upgrade to HTD 8m was in order.
Interestingly the specs for a 5m belt seem fine based on this chart (src). Assume motor makes 4kw, with an extreme’ish load factor of two (4kw motor power * 2 load factor = 8kw design power), this still puts us just within the HTD5m band at 2500rpm. Just.
Gearing is a real problem with higher pitch belts/pulleys. The 22t HTD8m motor pulley looks roughly similar in size to the old 56t HTD5m wheel pulleys I ran on my esk8. Even for a kart wheel HTD8m pulleys above 60t get impractical (they’d hit the ground when turning sharp). Found a very nice looking pulley from Jones Racing Products in 49t that was originally intended for a quarter midget race car. Idea with this pulley was that it came in a variety of sizes which would allow be to change ratios in the future, however over the last few years I see they’ve discontinued these pulleys. There are other options, but nothing with the same off-the-shelf bolt pattern that will mount straight up.
Motor pulleys are much easier to come by, in this case I’m using a rather industrial taper lock pulley. I don’t think these are an option for HTD5m pulleys or smaller bores, but they do have one neat benefit. The bushing locks into the shaft when the two grub screws are tightened. To remove the pulley you take them both out and screw one in to the third hole that just pops the pulley off.
So with a large wheel diameter (approx 280mm), and not really being able to get much gear reduction I needed a low kv motor to keep the top speed in a reasonable range. Still wanted an excess of power though, so settled on an APS 80100 50kv, this should absorb as much power as the battery can throw at it. If we plug all these numbers into the esk8news calculator this works out to be a loaded top speed of 46km/h, and that’s just fine for now.
I’d like to have a smaller motor as I felt packaging an 80mm motor somewhere in a one wheel lead to compromises, more on that later. Just not sure if they’ll have the performance I want. It’d also be great to find a better/more available solution for the wheel pulley.
There are obviously other ways to do this. Chains would be off the shelf from kart stuff, give more reduction, and are quite a bit slimmer, but I ruled them out because I just find them too dirty. Gears would work, but with a bunch of extra complexity and need for precision. So belt it was.
7 Likes
Held together with some 3D printed parts, zip ties and tape. Just wanted to see it do its thing.
It’ll balance when turned off, but you obviously can’t push it around like that. Played round with the PID settings, quite interesting seeing the effect they have.
12 Likes
What happens with this when it reaches motor capacity? With the normal OW it just shuts off and nosedives, what would happen with this one when it reaches its limits?
I think nosedives are inevitable when any one wheeled vehicle reaches its limit. Even an EUC will if you hit it’s top speed. What EUCs do well is make that top speed so high that it’s unlikely to be reached.
My gearing setup will give a top speed of 46km/h. If this works out ok I’d like to bump it up, just so I’ll never get near it’s top speed.
VESC apparently does field weakening now, could be an option to bump up top speed.
5 Likes
I know onewheels usually use high voltage Batts 12-16s, to get some decent torque on the hub motors.
Could you get away with using a 10 or 12s battery in this scenario ? As gearing could help out with torque.
1 Like
I’m sure they have their reasons, but OneWheels and the hubs used by floatwheel/etc have low kv motors. There’s not really a range of kv motors you can choose from like with esk8. You need to go high voltage for speed, not so much torque.
Anyway, you’re right. Having a reduction step gives a lot more flexibility in pretty much everything; battery voltage, motor kv, wheel diameter. You just need to run the numbers through an esk8 calculator to make sure you get the desired speed.
2 Likes
The bottom part of the battery enclosure was this rather large 3D print. This was only ever a temporary solution, just had no confidence it would survive long at all.
Plan is to use a clear sheet of polycarbonate; it’s strong, I can put lights/oled screen behind it, and I grew up when those clear case Game Boys were cool (or any electronic device for that matter).
Drilling polycarbonate is pretty easy, it’s acrylic you need to watch out for. Both look very similar but have very different material properties. You can actually cold work polycarbonate and maybe that’s what I should have done.
Next step was to bend the two 45 degree angles. Ideally, I’d use a sheet metal bender or a nichrome wire plastic bender but have neither. I do have a heat gun though. Used some bits of wood to limit how much of the plastic would heat up.
Waved the heat gun up and down that gap till the plastic was soft enough to bend. The first problem was some bubbling on the surface. This I don’t care too much about; it’ll be that scratched up the first time this end hits the road that these bubbles won’t be noticeable.
Now for some irony. The idea behind bending it on the frame itself was to get the bend to match up with the frame perfectly, but as you can see that didn’t happen at all 
. I believe the plastics contact with the aluminum meant that it didn’t heat up enough (I’d basically given it a huge heatsink), and as a result the only bit that got hot enough to bend was away from the corner I wanted it to bend around.
Not sure I’m ok with this. Next attempt I might try to insulate with a strip of masonite between the frame and plastic.
Anyway, still had to bolt it together to see what it looked like.
11 Likes
Spent way too long figuring out how I was going to accurately cut up the aluminum for the rails. Do I buy a drop saw, milling machine, hand router, etc. The thing is I didn’t really want to buy any of those tools, so chopped it up with a hacksaw (helped by some 3D printed miter boxes) and got the lengths down to a good tolerance/square with a crappy little disk sander.
It worked really well in the end; both rails were within 0.1mm of each other.
What didn’t go so well was drilling holes where I wanted them to go, but this did improve over time. Tried center punching first, and that was ok’ish but not great. Then moved on to what I believed was the correct approach of using a center drill first then the actual drill bit. This worked better, but I often found the small tip would snap within the part you were drilling. You then get stuck with a bit of HSS drill bit stuck down the hole you want to drill. I destroyed so many bits.
The center drill (left) is to be used for use on lathes, what I should have been using is a spotting drill bit (right) which is also sometimes referred to as a center drill bit 
. Picked up a few cheap countersink and counterbore bits from aliexpress, the power button counterbore was 20mm in diameter which my small drill press handled surprisingly well.
Tapping threads is now my least favorite thing in this build. There are about 40 m4 threads that needed to be done. The little 3D printed part did help keep the tap perpendicular, still sucks.
I did leave some of the tapping till after the frame was welded, but only because I was so over it. Marked up the different parts and dropped it off at a local fab shop that specialised in car intercoolers/catch cans/etc.
Same day service is not what I was expecting, but sure enough I got a call back that same day saying the job was done.
Welds look good, pity they’ll all be covered up.
In hindsight, I wouldn’t go for a welded frame again. Cost, makes it harder to work on, can’t change the width of the board (for wider/slimmer wheels).
13 Likes
Closer to a test ride?
2 Likes
This build inspired anyone else to make serious plans?
No . Life stuff just kind of gets in the way sometimes. I have started working on this again, and there’s nothing else I need for the build, only time. Just spent the last few days getting up to date on what’s changed VESC balancing app and it seems like a good time to get this build actually finished.
Designed and printed a little cover for the motor/VESC a while back.
Obviously not going for waterproof, even splashproof, so figured I’d include some louvers in the design. I’ve always liked them as a feature on race cars, and here they’ve even functional as this is covering up both the motor and ESC.
Also made a forged carbon fiber footpad (next update post) which turned out nicely. This is actually what got me bogged down with the build, the 3d printed molds took some 6 days of solid printing (big and high % infill) and while the carbon part survived the mold did not. Went through all the effort of making one, but just lost all motivation to repeat the whole process.
7 Likes
The process can be a drag sometimes definitely, but keep going dude! Would be super cool to see
2 Likes
Be careful with polycarbonate it has a tendency to stress crack/craze when exposed to some solvents or to UV then you are left with a lot of microcracks along the bend ^^ (already saw it on some things like ticket machine covers and such)
3 Likes
Threadlocker will also absolutely destroy it.
6 Likes
This is good to know. Chances of me getting threadlocker on this were high, and I likely wouldn’t have cared too much.
2 Likes