Itās really a matter of systems design choice though. With uart that can just as soon be vesc or receiver side depending how itās implemented. Thereās an internal timeout on the uart throttle input that acts as a failsafe in addition to the uart comms timeout. They all do the same thing in the end.
Iāve tested this, because with VESC comms timeout set to 1000ms, I can accelerate and turn off my remote, and it will hit neutral in under 500ms.
Edit: Going back to test this further so I have better data on it. Enjoy the discussion as always Brad!
Motor detection should be done without any load or anything attached to motor but with motor on mount and wired in the same way it will be when loaded this way the BEMF measurement is accurate given the actual resistance etc. to the motor. If loaded or hooked to a wheel it will take more amps to turn but the calculated BEMF I believe would be the same.
Yeah I guess I should clarify, mounted to the pulley with belts and wheels attached but no load.
Thanks for confirming!
I guess a followup question is for FOC, do I need to recalibrate if I switch a phase wire to reverse motor direction? My gut tells me yes but I wanted to confirm
I think yes as well any changes to phase wire length which I guess includes swapping them would change the calibration results I believe not sure how big of change that would be but canāt hurt.
Hmm now if I changed it in the vesc tool is calibration still required? Or does the software take that into account?
Yep, totally understood here. The particular radios Iām using have their own network transport layer that runs underneath the transparent UART though, so itās thereās packet validation at that level + at vesc packet checksum level. These are pretty complex radios- you can even load micropython apps directly to them, fully programmable MCU onboard.
I actually agree with you here, and itās one of the reasons my 1.0 receiver hardware design has an additional MCU onboard to handle radio/vesc comms separately among a few other things (like PWM output options)
My original point was only that itās inaccurate to say that uart method lacks failsafes that do not apply to itsā technology subset. Iām not doubting RC/Drone implementation (I did industrial long-range fixed wing back inārāday, we should talk shop on it sometime).
Make sure your battery will fit in your intended enclosure - a 10s5p of 21700 cells will be substantially larger than a pack of 50 18650s.
thanks i changed my mind iām doing 30q s instead because the guy im splitting with needs 18650.
I decided to just install those AS150 antispark switch instead. So much simpler. Thanks for the suggest Didnāt bother to install the new LED switch or test the built-in antispark.
Yes, do the same with my boards now. Good call.
What enclosure is that?
Whats a good method to secure my loop key port? Id prefer a non permanent method as if the xt90 goes south Id rather not have to replace the entire box.
The big, chunky main battery and enclosure are whatās left over from my first build, a diyeboard 10s5p duel belt-drive kit. The ESC enclosure (for my Flipsky 6.6 duel plus) is from Backfire. Itās $15 bucks on their site and ships from California so I got it in 2 days!
I probably should have said hot glue failed me.
@gustavthemighty Then neutral cure silicone. You should be able to pull it off if needed. I used a box made of tape around pieces before filling with silicone to give it a nice shape on my Frankenstein board
Pretty sure changing direction just changes the sequence the gates are open to let power into the coils so changing it in VESC tool shouldnāt change anything electrically per coil so think not needed.
This generic 10s2 is only lasting me 2.5 miles. Iām loving this way more than that. Anyone know of a usa based single stack battery supplier?
Would a name brand battery last me a big longer ?
Is the math that a 10s4 would give me 2x the amount of juice?