It’s been a few weeks since I discovered SimpleFoc and I browse the forum, (great).
I rushed to order a (BLDC-SHIELD_IFX007T shield) and an AS5047 magnetic sensor, I already had a Multistar 9235 100 motor.
for a project.
As much as it didn’t take me long to get everything started, I was quickly disappointed with the hardware (I’m back from my little cloud!) and I have quite restrictive specifications,
noise (therefore switching frequency greater than 20Kz, automotive type battery power supply, hollow shaft motor, torque, slow speed (200 rpm max, phew).
So I have no choice, I design my hardware part, the cost is not an obstacle (to a certain extent), but I still have a problem.
I can’t position a position sensor in the center so I decided to use Hall effect sensors in analog mode and simulate an encoder,
I think I won’t have something ideal but it should be enough for my application, I want to use an ESP32-PICO-D4 with 6 pwm and the can bus,
Do you think it will be powerful enough for this function?
A PICO D4 is a very powerful MCU… I think it can handle it. You will have to be careful how you write the code to make sure CAN communications don’t interfere with the iteration speed of the FOC algorithm, but I think it is quite possible, esp. on a dual core MCU like the PICO D4…
For the hardware, by chance I have recently made a similar motor controller (no Hall effect sensors, just SPI) based on PICO D4 and CAN bus. Maybe you find it useful, or maybe you have some ideas for me. I will try to post the files to GitHub and share the link with you.
Regarding the Hall effect sensors for position sensing, I think it is a viable idea… are you on Discord? There is a recent thread on our discord server on exactly this topic, and user @nanoparticle has prepared some code for this, which we hope to integrate to the library soon! See this thread on our discord: Discord
So I think with some care and patience your project is very possible, and not so wacky at all! If you have the time, please keep us posted on how it goes!
I haven’t tested this much yet, and hence haven’t documented it… I have some other tasks in the queue I have to finish off first, but my first quick tests with the prototype I had produced were promising.