Re: Did anyone used linear hall sensors on BLDCs for position sensing?

Hi AndrewBCN,

Thank you for your kind words.

I have tested three motors I think could be used.

1. The cheapest one was $5 but it was a MABUCHI 370 brushed motor [www.mabuchi-motor.co.jp] with 448 CPR incremental encoder. And included a 14-tooth MXL pulley. It seemed to be a bit short of torque at 12V but considering I was using an L298 for the drive electronics I guess that could be improved with more efficient drive. With 4x encoder readings we have almost 1800 "steps" per revolution, and if the original pulley is kept that turns into a resolution of 0.016mm/step. It is with these motors that I did this video [www.youtube.com]
2. The second choice was a brushless motor, I later learned it is manufactured for Ricoh, that comes with built-in drive electronics (which makes our life simpler) and a 100 CPR encoder (that is the worst part, as it seems a bit poor for my taste). I have failed to obtain a datasheet but motor feels very powerful at 12V and impressive at 24V. There are two models a 10W and a 20W motor. But unfortunately they seem to be marketed by Shenzen sellers that somehow grab a bunch of them but have no idea what they are dealing with. I could buy both versions just under $15 each. But to the cost of the motor we still need to add the closed-loop processor. I managed to get two of these motors controlled by a single Arduino, that, if Pro Mini is chosen can be had at $1.5 each.
3. The third choice is a brushless motor from Nidec, the 12V one from 24H series [www.nidec.com] . Again eBay sellers have no info or idea about what they sell and my unit arrived rusted on the outside (but operational). Nidec is a B2B manufacturer and unit prices are only available for manufacturers. Nidec motor has room in the drivers PCB for an encoder but my $7 unit did not have it on board. Though the motor is light, it delivers way more torque than Mabuchi's so I quite like it too. But I kind of rule it out as sourcing it seems sketchy too.

So I have ended up in talks with a manufacturer that seems interested and we are in the process of developing a brushless motor that will work at 12V and it should provide an easy replacement path for X & Y axis motors (just pull the Pololu and plug the motor wires instead). These motors would include drive electronics so interface uses the step, direction and enable pins that are standard on any stepper controller.

It is when trying to get the best price possible when we face with the idea of dropping the encoder, but as I am quite happy with how encoders work and I am doubtful about replacing them with other technology, I ask here to see if I get encourament or warning signs.