I've been working on a design that will allow me to use an AS5047P encoder on a hover board motor. I can't attach the magnet directly onto hub motor because it is a single shaft so I've created a geared assembly. Here is a link to an image of the design I've created. I've printed it and installed it with no issues. The encoder and magnet just attach to the spaces on the middle of the small gear and back of the arm. The big gear has 64 teeth, the small gear has 25 teeth. The hub motor I'm using is a 14 pole motor. I've successful been able to connect the AS5047P to a Vesc6 and I get stable data. When I look at the position data it shows about 4 revolutions for every single revolution of the motor. I've tried to change the encoder ration and invert it but I've had no success in getting the position to read accurately. The encoder detection also doesn't accurately capture the encoder ratio.
I'm wondering if anyone has experience using the encoder ratio when it isn't turning 1:1 with the motor. Any insights or thoughts would be appreciated.
Update on this. I made a new design where the gear ratio is 1:1 but inverted. . I've been able to set everything up and the encoder now provides much more control.
I suppose I misunderstood what the encoder ratio is for. I thought this could be any ratio and the encoder didn't have to run in sync with the motor. If I'm wrong on this I'd love to know to use the encoder ratio to allow for a geared system that would be much more compact. The 1:1 ratio of the gears makes this quite a large addition to the motor.
The encoder ratio is just the number of electrical revolutions contained in one mechanical revolution of your motor. This is equal to the number of poles divided by two. Also, it can only ever be a whole number! You can use these simple rules to figure out some new gear ratios to try.
For example, I usually use motors where the number of poles is 8 so the encoder ratio is 4. If I was feeling creative, I could gear my encoder magnet to spin 2X or 4X faster than my motor and after running detection it would still commutate correctly. The 2X gear ratio would yield an encoder ratio of 2, and the 4X gear ratio would yield an encoder ratio of 1. Quite a few gear ratios are possible as long as dividing the motor's original encoder ratio by the encoder magnet gear ratio yields a whole number. Please note that some error will be introduced by the gearing itself, so motor efficiency will likely suffer.
Thank you for your sharing！
VESC XieMaster