My onewheel build:
1515 8020 scrap I had laying around
Cheap Focer V2 ESC
2 6s2p samsung 30q packs
phub-188 36V 800W wheel
Milled nut in the 8020
Testrun with no footpads and 2 1300mah 6s battery
Footpad: Got a .8mm PCB ordered with velstat on top. Fiber tape holding it down and griptape on top
Quick and dirty wire routing
6s2p pack with no bms - drew up and 3d printed the holder
Spotwelding with 330F 16V supercap!
Super clean looking wrapjob
TODO:
Replace the 90 angle brackets -those diecast ones are junk.
Motor acting badly at low torque low speed. Can’t go over a bump at very low speed without the motor desyncing or something.
Cleanup wiring
Been doing a couple longer rides with comparing the XR and funwheel back-to-back. 8 miles one day starting at 80% battery and ending at 10%. And another 4 at lunch today.
Buddy trying it after him riding a couple miles on the XR.
Great build. Thanks for the write-up. What kind of Erpm are you reaching on the 800W Phub motor?
Just tested it out: 11450 erpm at .95% duty cycle
kv = eRPM * (100/95% dutycycle) / (polepairs*voltage)
kv = 11450erpm * 1.052 / (15*50v) = 16.06kv
Math sound about right?
(edited had the polepairs wrong)
Pei scooter has seen the light. That’s a great kV for Onewheel builds.
I wonder if they actually changed the number of windings or if they just changed the wire termination from star to delta.
Tempted to crack it open, but i just got it working pretty well!
Key for me was to reduce the motor current to ~50A and increase the FOC time constant to 2000-3000us. I had motor current at like 80A and I think the motor would saturate(?) and give me clicking and inconstant torque especially at hard braking or very slow speeds going over a bump.
Did you use the motor wizard? I have heard multiple instances of people having exactly those problem after using e-bike (6kg) hub motor for the Phub 188. Clicking during braking and overcurrent errors.
Solution was using the 2kg outrunner as base.
Personally I do the manual motor detection on my hoverboard motors. The wizard has given me 18wh consumption and bad performance. The manual motor detection gives me 11wh consumption and good performance.
I yeah true I did do the motor wizard with the ebike hub, but then later was trying the manual detection as well as messing with they params directly. I made it worse before it got better, vesc tool is a bit of a learning curve.
I built vesc code to include additional hall effect filter from here: https://github.com/vedderb/bldc/issues/182
It helped out a ton when running my motor current higher! Which helped out for going over bumps on a standstill and braking fast.
Great job man! Looks great.
Nice stats.
Whats your topspeed with load? And this is just stock pei?
Im in between different motor designs. Would you guys rather have a bittt more torque and sacrifice a little topspeed?
I hit tilt back at about 16mph with 75% duty cycle. About the same as the XR.
I think some extra torque would be better. My wheel still doesn’t do as good as the XR on grass or low torque areas.
Best would be lower KV and higher voltage, best of both worlds! Just need a good 15-16s ESC to use!
What kind of KV are you shooting for? Try and pack as thick of copper as you can around the motor for max efficiently!
I have not tried the HFI, as far as I could tell from reading, hall effect sensors still win over HFI for the balance applicaiton, but it is a excellent option for skateboards without any sensors.
Thats true. I read through the github hall effect filter. I thought if you have to add more filter samples, maybe you get less errors with HFI. There was also a statment that the sensor data only comes every 60°. Consider the revolutions of a big wheel, probably software solution is more accurate. But i will keep this vesc code filter in mind!
