in your video above you mentioned about removing one winding for added efficiency and torque.
i am wondering would your controller be able to drive a motor with precise accuracy for say a camera gimbal ?
Thank you for stopping by, hexa. I have seen the gimbal videos and they are super cool! I'm pretty proud that one of the first people who built a BLDC-based gimball was able to use parts of my code, specifically the PWM part from here
. I have another project that includes positioning with a BLDC motor (though not a gimbal) , so hopefully will be able to answer your question from my own practical experience. But I'm not there yet.
However, I would like to point out that you may not need to rewire a large RC-style outrunner BLDC if you use a smaller motor instead, like the one on my video. It's holding torque us quite adequate for a smaller camera, something the size of a webcamera perhaps. As fas as positioning, the smaller motors are not terribly accurate - 10° steps. That's because most of them are 9 cogs / 12 poles which translates into 36 positions of maximum magnetic attraction for the full 360° circle. If you're using a differently constructed BLDC motor, you may have a slightly better positioning accuracy (delete the number of cogs by 3, then multiply by the number of poles - that's the number of stable positions in a 360° circle). It's not really the function of the controller per se, it's more of a function of how many steps are there in the particular BLDC motor. The controller needs to calculate the winding commutation for the next step and "ease" the rotor into that position using PWM (otherwise it'll just blow right past it) and hold that position
. That position holding is not currently a part of the software that runs the controller I'm building. I have a pretty good idea about how to implement it but, again, I'm not there yet. Sounds like a very interesting side project though, I may investigate that in better detail soon.
is your timing technique similar to this ?
Thank you for the video! Not sure how come I didn't come across it earlier but it explains running BLDCs in very fine detail, I wish I've seen it before. However, I would have to say that in order to build a BLDC gimbal, you'll have to pretty much forget all that he was saying
and do exact opposite. The way I understood InstaSPIN - it is an optimization technique of winding commutation for continuous
rotation - such as spinning a prop for example. In the case of a gimbal, the rotation is not continuous - it's interrupted at each of the stable positions (even though it may look to the eye as continuous - some of the stops may just be very brief). It's not even only not continuous but also reverses quite often. In fact, in gimbals BLDC is used exactly as a stepper
, which is what the InstaSPIN technique is designed to prevent. So, again, great video but for gimbals we need an exactly opposite approach.
By the way, there's one detail in design that makes DVD spindle motors even more suitable for BLDC gimbals than the RC-style outrunners (never mind the power for a moment) - most of them are equipped with Hall-effect sensors that can give you a very accurate positioning control.
So, yes, like I said, a very interesting project. I' need to take a closer look at this myself.