Does anyone know how to figure out the minimum speed of a particular motor? What tricks can be used to have a lot of torque and a low as possible speed? What are the limiting factors? Is it worth looking at a gearbox on the motor to get a lower speed (or does that not make sense with an electric motor).
Questions, questions questions... Anyone investigated this or are we all a bunch of speed demons here :-) ?
What kind of application do you have?
If you want full torque at 0 speed for e.g. position control you need an encoder such as the AS5047. In sensorless mode the VESC can also run at very low speeds, position control at 0 speed is not easy without an encoder.
My application is comparable with a wirecam, but a bit different in some ways (I think i need a bit more torque for example).
In FOC i can get the motor to do about a rotation per 0.75 second, but torque doesn't seem amazing when i do that. It would be great if i could run at 2 or more second per rotation.
Reading up on the AS5047 sensor and might try to make things work with that. Not so comfortable with electronics but it's an interesting journey so far :-)
Terribly impressed with the VESC project.
Yes I would use an encoder at those speeds.
Have a Nice Day.
Thierry
Benjamin, there are a bunch of applications where an encoder for position control is not necessary. And it would be great if the VESC could operate in such a mode.
Suppose I take an 8mm threaded rod, Hook it up to an HX750 motor and use it in a CNC / 3D printer application.
The HX750 has 7 electrical revolutions per mechanical revolution. So it has 42 possible "enhanced cogging points". So, with 1.25mm of running distance per revolution on the rod, we get an accuracy of about 30 microns, if we can position to the BLDC-step that is requested. The only thing is that with lack-of-feedback, we would have to keep the "holding torque" current running. But simple BLDC-mode hallsensors would fix even that.
If Davideo is who I think he is, the 42 BLDC positions per revolution will result in a positioning accuracy much better than what he really needs in his application.
My 30 microns is "on the edge" for a 3D printer. For a 3D printer you'd maybe add a 3:1 reduction belt from the motor. Oh... and top speed for moving in that manner? Ehh 60000 ERPM, 8570 RPM on the rod, 142 Rotations per second, or 180mm/sec. Not extreme but quite acceptable for a 3D printer. But especially with hall sensors we can deliver 10 or 20 times more energy into the system than with stepper motors.
I think the position controller should work with hall sensors as it is, but you have to adjust the PID parameters to make it stable. The low sample rate close to 0 speed makes it a bit more tricky.
Interesting! Can anyone recommend a information resource on the concepts around PID so I can figure out how to tweak them for my motor and purpose?
After updating to VESC_tool 0.87 today I noticed much smoother action at low speeds, it might actually be good enough for my purpose. Will still have a play with the AS5047 sensor on my next prototype though :-).