Been doing a lot of thinking and tinkering in my head with not dine before drive transmission solutions for esk8 for new limits of handling on the track. Anybody with thoughts in the mstter, would love to hear on it.
For context, torque vectoring is super young but with potential for some of us on here who’s pursuing it either diy (@tuckjohn and @MacKeeper28 ) or production brand (radium). Already on track to tackle the handling challenge of optimized different wheel spin during corners.
Say now theres a hypothetically perfect plan: funding is ready, design is ready to send out to manufacture, and plenty time after to tortue test prototype batch to dial in a bulletproof, single motor differential drive for esk8 (to which then I look at @Ean.esk8 ‘s work and realize hes already ahead of me there lol)
What exact racing aspects would see improvement from its integration, if at all? And given them all, are said improvement goals worth pursuing in spirit of esk8 racing innovation?
You’re thinking in the right direction but straight up, a single motor diff setup won’t outperform torque vectoring in racing, especially on tight or technical tracks. Torque vectoring wins because it actively manages traction per wheel, while a diff just reacts, so you’ll still get inside wheel slip and less control on corner exit. Where your idea does make sense is weight distribution and reducing unsprung mass, which can improve stability, ride feel, and maybe consistency over rough surfaces. So it’s not useless, it just won’t be a top-tier racing advantage more like a cleaner, possibly more efficient setup. If your goal is pure performance, stick with torque vectoring; if it’s innovation and better overall board dynamics, then it’s worth building.
I can’t speak for @tuckjohn ’s efforts or any future Radium upgrades but I can speak for my experience with RTS and vesc traction control
I wouldn’t say RTS is optimised for addressing wheel spin through corners, or at all. I notice plenty of wheel spin with RTS and i’d probably go as far to say as i get MORE wheel spin with it than without it.
i’m not sure what if any traction control measures @Tony_Stark has implemented with RTS, but I can say that I believe it would be crazy difficult to implement in a way that works as it should, safely and reliably.
there’s so many factors to consider and testing would need to be rigorous.
i’ve come unstuck enough times using vesc traction control that i can comfortably advise anyone thinking about turning it on to just bin that idea.
i may be wrong but i don’t think any work has been done to the traction control side of the firmware since it was implemented in the first place and I do not believe it to be a safe or viable ‘feature’ to use on esk8.
TLDR: implementing a complete and thorough version of torque vectoring and traction control on esk8 is more than likely an extremely difficult project with many pitfalls. Good luck!
Torque vectoring can be used either for more traction in corners, or for help with steering. I think radium set it up to help with steering?
In theory if you use it for more cornering grip, it also has the effect of straightening the trucks.
My approach would 100% be to optimize cornering grip, so I’ve started with figuring out how to have the most leany divey and responsive, yet stable setup, which would still turn well even if the trucks were much tighter. I think I’m ahead of everyone else in that regard.
Proper traction control is possible to implement, I had a crack at it a few years ago with some friends, and got it working* it was actually doing it’s job well, except the microcontroller we used was unstable. We do have plans of remaking it properly on STM instead of ESP along with implementing traction increasing torque vectoring as well.
What we did is a separate canbus module that the receiver connects into, reads the PWM and ESC data, then sends canbus commands to each ESC.
The way vesc traction control works is by matching the speed of every motor to the speed of the slowest turning motor. So it does a reasonable job of keeping unwanted single wheel burnouts under control, but the biggest problem for me was brakes..
if you’re braking and one wheel locks up and skids, the other(s) will follow suit because they’re tracking the slowest motor, no exceptions. When you’re not expecting it and especially on a 4wd board, it gets pretty hectic.
I’m with @Dinnye on this one, the perfect world is an independent motor at each wheel. The only reason anyone ever developed transmissions at all (alert, sweeping generalization) is because gas engines don’t like to go below about 1200rpm or above about 6,000
And Vesc’s code was only ever developed to keep 1 wheel from spinning out. To do proper traction control and torque vectoring is a huge task worthy of a Ph.D thesis.
I did not know that traction control affected braking, that’s pretty scary. I’ve never used it, though on the mountain board I can see where it would be handy. But I’ll wait until we can turn it off for braking (which should be really easy)
For other applications? Yes. The biggest dev community I know of is the DIY onewheel folks. They have a lot of smart people contributing to the VESC project to make their application better, as well as writing “Packages” which is a VESC feature that completely missed the entire esk8 use-case because we have no devs.
Independent motors per wheel i agree. Is minimizing unsprung weight via moving wheels off them worth experimenting and testing? Helluva stretch to make possible via different drive, but if the benefits of lighter trucks are elbig enough, worth a shot
Right now I’m running WFBs; 93/95.5 tapered barrels front, 100/100 fatcones rear. This bushing setup isn’t available for sale now, so not sure how useful it actually is 40/17.5 angles if I remember correctly (front might be 37.5), on a flexy deck at around 114cm wheelbase.
Swingarm setup to replace flexy deck TBD very soon. Been also doing some real engineering work on the double stage drive as well in the meantime.
The canbus module didn’t have any torque vectoring yet, only inputs were receiver PWM and VESC data. I suggest you check out Tuckers TITS for angle measurement if you want to have a crack at that yourself. Not sure when exactly I’ll get started on a new revision of the TC/TV module, but probably not going to get much done on that front until autumn.
I don’t think you can achieve moving motors off the moving part of the trucks in any way. The amount of turn and lean we are dealing with is too much for any sort of coupling or U joint. Not to mention we are dealing with large torques on the motor shaft.
Also I don’t really agree with most; I think having light trucks is way too overhyped. I’d even go as far to say that it doesn’t matter.
Gotcha, definitely a take to favor less on committing tinkering energy to diff drive and more torque vectoring.
Kinda my hope for my original question: if anybody can imagine under their feet the kind of improvements they’d wish to push esk8 racing, i can gauge which direction of experimenting I’ll shoot for. Cuz atm, I can imagine something viable for diff drive or torque vectoring, but which one to hit up i want to depend on what yall think is worth favoring
I have tried torque vectoring on the Radium and I did not feel like it would serve me well. Riding styles vary enough that I can see it being hard to make a one size fits all type programming. I like to really throw my weight and push the board and on the Radium I tried it was like “OH SHIT HE TURNED LETS OVERCOMPENSATE TO CATCH UP!!! OOPS WE ARE SLIDING!” so idk if it would have to be a progressive learning type system or if it’s just better to not have it.
One thing I’d like to see in power transmission is better out-runners. Surely there is a size or design that won’t thermal out given the demands of our fastest heaviest riders on go kart set ups paired with these new high voltage Vescs. The Vescs are just about there. Semi Solid State Batteries are just around the corner, so I think now it’s about the motors. A question for @Tony_Stark I guess. Do we just double the size and run them at “low capacity” for their size or what’s the reason it’s mostly all 6485 6495 etc.
I remember when I was first playing with it and dialling in settings that I liked and I had thoughts quite similar to your own.
It’s a different vibe and you totally need to adjust your riding style to suit, it definitely took me some time to get used to the feeling and adapt to it.
Now that i’ve put many kms on it with RTS always on at the settings I landed on, I really should just turn it off for a ride and see how it feels without it, but my previous experience trying this had the board feeling less lively.
I don’t think it’s entirely fair to judge RTS in good confidence unless you’ve really spent a good amount of time with it… i’d be interested to know how much of a chance you gave it!
Yes it takes weeks to re-train your brain to control steering with the throttle. And the settings need to be adjusted to the rider and board setup. For example the acceleration expo setting alone only feels right for me within a 5% band on hard SR125s, and on pneumatic I had to drop it by 20% to go from feeling dead and like it won’t turn to feeling awesome.
40/17.5 angles if I remember correctly (front might be 37.5), on a flexy deck at around 114cm wheelbase.