Redoing Things, Also Woot Hall Effect Sensors

So in the last post I was describing my special abomination. But looking at it makes my inner engineer feel bad  so in the near future it will be remade using a single chunk of aluminum that the sprocket can be mounted on. Also small motor is under going a few changes do to the fat assery of the bearings there was a bit of rubbing on some heat shrink so I turned on of the plates to seat the bearing around .125" farther out from the stator. This seems to have fixed the rubbing problem. On a side note all hot glue in the motor used to hold down wires will be removed from the motor after talking to Charles about fears of oozing thermoplastics getting all over the place.

Sexy red heat shrink
But on to something not about me messing up: HALL EFFECT SENSORS (read that again with enthusiasm and a disregard for the preceding paragraph) they tell where the magnets are in the motor which is *sometimes* essential for starting a brushless motor from a dead stop. I'm going to stick hall effect sensors in my motor soon but am currently out of town so I'm ti babble about hall effect sensors for a while. Now there are sensorless motor controllers and sensorless brushless motors, however to start from a complete stand still with a large amount of inertia you want sensors. The sensorless controllers work off of the back emf from the phases, the voltage from the changing magnetic flux tells them where the magnets are and when it is appropriate to turn on a particular phase buuuut to do this you need a bit of a spin to figure out the magnet placement. Therefore in vehicles where you want to be able to start from a stop, hall effect sensors are nifty things. Timing with hall effect sensors is analogous to the timing of cam shafts in a car engine; firing all 6 pistons (or 3 phases if its a motor) at once will get you nowhere but when things are done in the proper order you get rotation.
Generally in a 3 phase sensored motor you'll have 3 sensors. Each hall effect sensors i will work with (ATS177) are effectively flip-flops with some hysteresis and can tell the direction of the magnetic flux through the sensor. Because there are 3 phases and 3 sensors we want each hall effect sensor to be 120 electrical degrees apart (not necessarily physical degrees). The difference between electrical and physical degrees for the magnets is Eelectrical.deg=Physical.deg*polepairs (start at 0 after each 360) note: there are 20 pole pairs in my small motor, because of this there are several theoretical positions to place the hall effect sensors; really though there aren't many practical places you'd want to put them. Below my crappy illustration attempts to show this (for a prettier illustration of almost exactly the same thing see Amy's blog where she pretty much did exactly what I'm about to do), Tm represents 1 magnetic period/pole pair/360 electrical degrees/ the smallest repeatable section to get the proper magnetic pattern. Each colored dot represents a theoretical place where a hall effect sensor could reside. The purple dots represent the closest theoretical placement of the sensors at 120 electrical degrees apart...give or take a bit in the picture.

Motor diagram w/ magnets and hall effect placement
 Practically we want the placement of the hall effect sensors to line up with the stator slots because we physically can't shove them into the 1mm air gap (and the stator phase windings would mess up the readings) so that leaves us two possibilities where the stars align magnet and stator phases line up at spacings of 60 and 120 physical degrees; this is respectively represented by the orange and green dots. One important thing to note is the order of the phases with the placement of the sensors. By following the orange dots clockwise the phases will go ABC but doing the same for the green dots will give ACB. This difference in pattern will make the motor spin in backwards for the same sensor pinout, but that can be fixed by switching any pair of phase wires. Oh and just a note as long as you have 3 phases everything is independent of the winding pattern  but ummmm yeah that's hall effect sensors for now. I'll post pretty pictures soon.

No comments:

Post a Comment