Closed loop for X and Y

Danal

@doctrucker said in Closed loop for X and Y:

BLDC motors can run sensor less, however yes, I think it is far easier to get the timing of energising the coils correct and at an appropriate advance or retard with an encoder on the motor.

The keyword being "run". Sensorless is based on back EMF from the un-energized phase. No motion, no back EMF. Small motion, not enough EMF to (properly) sense. Not a great thing for start/stop tracking of rotor position. Very un-suitable for stepper alternative, regardless of "closed" or "open" tracking of position over time.

DocTrucker

This project looks interesting:

https://odriverobotics.com/

...and it's statement about most industrial motion being done by brushless servos.

Neatly it looks like this is likely to have CAN support soon, which will mean not masses of work to get it to play ball with duet3, or maybe with earleir duets with a bit more hardware since CAN is on Arduino Due.

@danal more-or-less an expansion of what I said? Hence 'can run'.

The rotor position encoder doesn't have to be particually high resolution to achieve decent start control. You could also send signals through one of the coils which won't move the rotor, but allow you to locate it.

The sensorless techniques result in a vague knowledge of true rotor position when in use, much like only knowing a stepper rotors position is within two physical steps of that assumed. This is acceleration/instant speed tuning come in. Essentially a patch on open loop control.

The downside of using BLDC like this is a 'lost step' would result in a larger angular displacement, but equally you'd need a greater torque to cause this. You could also slap an encoder on a stepper, but by that point you've lost most of the cost benefits of a stepper system.

dc42

@danal said in Closed loop for X and Y:

@doctrucker said in Closed loop for X and Y:

BLDC motors can run sensor less, however yes, I think it is far easier to get the timing of energising the coils correct and at an appropriate advance or retard with an encoder on the motor.

The keyword being "run". Sensorless is based on back EMF from the un-energized phase. No motion, no back EMF. Small motion, not enough EMF to (properly) sense. Not a great thing for start/stop tracking of rotor position. Very un-suitable for stepper alternative, regardless of "closed" or "open" tracking of position over time.

Exactly. Running BLDC motor sensorless is great in drones and similar applications where the motors always run fast, but in 3D printers and similar applications you must have a position sensor that works at speeds down to zero. You could possibly use a combination of a BLDC using sensorless commutation, gearbox (because BLDCs are suited to higher speeds than we need in 3D printers), and an encoder after the gearbox.

Mabover

I always cringe a little bit when I see people using hobby/rc bldc motors for moto in control, as dc42 points out they are designed for high rpm operation, and not the low rpm position control a cnc machine needs, but I will say I am impressed people make it work as I have seen with the odrive.
There are low cost industrial options emerging for bldc motors from some of the same suppliers as stepper motors, they even cone in standard NEMA motor frame sizes.
Here is an example motor
https://www.omc-stepperonline.com/bldc-motor/24v-4000rpm-0185nm-775w-48a-42x42x80mm-brushless-dc-motor-42bls80-24-01.html
combined with an encoder such as
https://www.omc-stepperonline.com/encoder/1000-cpr-optical-rotary-encoder-ab-2-channel-id-5mm-w-shielded-cable-hkt30.html
Or
https://www.cui.com/product/motion/rotary-encoders/incremental/modular/amt10-series
And all you need is a drive of some sort.

If the demand grows for these motors due to support by a future duet 3 expansion boards and support from other boards in the market the price will likely drop just like they have with steppers.
Perhaps vendors will even sell the motors with an encoder attached saving the trouble of sourcing and mounting one.

dc42

On Duet 3 we're also looking at supporting stepper drivers with encoders, such as these https://www.aliexpress.com/store/group/stepper-servo-motor/5053149_515433734.html?spm=a2g0o.detail.0.0.133274b9HAWpmu. We think we can drive them from the existing 3-drive expansion board, using a small adapter PCB to connect the encoder outputs of 3 motors to 3 of the I/O connectors on the expansion board.

Mabover

@dc42 that would be awsome, I am guessing that the encoder will be used primarily for position/movement verification.
I have a stepper driver that supports the use of an encoder for position verification and commutating the stepper motor as high pole count bldc to run a control loop on the curent of being applied to maintain rotor position. It can help quite the stepper down versus the standard noisier position verification mode but I don't think there is much to be gained over the existing silent nature of the trinamic drivers using that approach. One advantage it does provide is the ability to momentarily over drive the motor to overcome a high torque demand that would normally cause a stepped to miss a step.

All in all really exciting news