ive got two hub motors that are rated at 30A. ive got the odrive limiting the coil current to 30A and they have significantly less torque than i would expect from the size of the motor,and the datasheet and supplier aren’t able/willing to clarify.
are motors typically rated by coil current or supply current? I’ve found contradictory information online and was hoping someone with some experience could give a reasonable idea.
They are rated by coil current.
How significant is significantly?
Also, if you are using Hall sensors, then the stall torque could be less than optimum because it is essentially an extremely low-resolution encoder, so the field-oriented control algorithm doesn’t work so well. If you have the possibility of using an encoder, e.g. one of the absolute SPI ones, then it might help.
its hauling ~1/3 the weight that i would expect on a given incline.
its a 144 count encoder and a 24 pole-pair motor so 3(?) encoder pulses per pole.does seem rather minimal information when put like that.
when stalled the motor does occasionally jump. like an overloaded stepper motor missing steps, should this be happening?
is this low torque just a quirk of odrive how do e-bike and other hub motor controllers deal with it?
it would be alot of work to add an encoder if theres even room for one. but its something to think about.
How are you calculating the torque constant of the motor? Do you know the velocity/back-EMF constant Kv?
It should not be jumping.
Generally, motors with Hall feedback have poor performance at low speeds or stall. Once the motor is running, the ODrive or any other drive should be able to commutate it properly and get full torque.
This effect is even worse with motors that have back-EMF feedback aka “sensorless”. Hall sensors are rubbish, but better than nothing.
That is not going to be sufficient. Per @madcowswe, you need at minimum 6 counts per pole-pair. Alternatively, you can also use normal 3-hall effect sensor mode, it would probably work better in this case.
The best thing you can do to improve performance is improve the encoder.
@Wetmelon I was going to say that, but @kragon had mixed terms between poles and pole-pairs.
3 counts per pole is normal, but normally we only talk in pole pairs.
144/24 = 6.
Nevertheless, I agree with your assessment. For better performance (especially at low speeds / stall), @kragon needs a better encoder.
yep @towen , sorry for mixing units.
is this ‘normal 3-hall effect sensor mode’ potentially still helpful?
how would i go about trying it?
It’s used when you have 3 hall effect sensors on the motor, see https://docs.odriverobotics.com/hoverboard for example.