just add some slots for the Encoder in the 1mm thick Stainless Lid…
It didnt worked out… the space is so limited… not possible to adjust proberly
More then one worrie about the encoder gets into the wheel OR the magnet felt insinde the gears…
So i changed my mind for now… Will mount the Encoders from the Endbell …
More room for mounting, no grease, better adjusting automaticly with motor…
…This has gotten way out of hand…
Got no M2.5 nuts / treadcutter for the 2.6mm holes of the AS5047… do i tread the holes with M3 tread…
OMFG. That’s gonna be insane again. To be honest, each time I see that Jens post something… without opening the thread, I need to get prepared for the insanity that is going to punch me in the face ! That’s not even DIY, that’s craziness. All the times, it’s polished, finished. Crazy bro !!!
@squad:
But the AS5047 is technically the better solution, especially for FOC. It enables the VESC to read the EXACT rotor position and to calculate the EXACT current value for every 0.1° of a revolution.
Switching from hall to encoder ist just about the same evolution as changing from BLDC to FOC.
It’s not easy to adapt to an existing motorholder, but quite easy to incorporate into new future designs, where it sits on the side of the motor shaft, protected protected by the gearbox housing.
That’s good to know, I newer went into details of anything other than hall sensors, but what You wrote makes perfect sense, learning something new every day. But it does not change the fact that even though halls are kind of obsolete compared to encoder they do just fine in this application
there is not much info on AS5047 setups and principles.
? could someone explain how it actually works. is there any magic in the round magnet that you add to the shaft ? (i guess its not just a magnet cause how would the sensor measure the postion then)
does the AS5047 sensor have to be super sturdy an axact positioned like @Nowind did it or could a 3d-printed holder also work (cause 3d printed ones might have some shaking and bending and so slightly position changing to it)
@squad;
Of course halls do fine in this application, I’m also running most of my boards with them. But that phase switching every 30° always gives a current spike which limits the power you can pull out of a VESC before it fries. With FOC+encoder the switching get much cleaner. This all is just important for low rpms. At higher rpms the VESC relies on its observer and runs sensorless, so you may only feel a difference when pulling high currents at low speeds.
@notger:
The AS5047 has internally 4 analog hall sensors which measure the direction of the diametric magnet and uses this to calculate the angular position. It has to be centered quite good above the magnet and should not move under operation.
A 3d printed housing would work fine and yes, the magnet can be glued to the shaft. The manual says it should not be connected to ferromagnetic material but it works nevertheless.
GOD DAMMIT!!! Surely that’s the perfect setup?? What a beast! I’ve just finished my board (well nearly…for now!) but the cogging on the unsensored motors for startup is a pain in the ass! Bout time I learnt how to install sensors!