This week, we worked on further tuning the medium size of the two big motors with UPMC with FOC code, which took a lot of calibration and debugging. We were convinced that we needed a backup plan in case FOC does not work out, which is why we also started writing firmware for the far more simple six step zero crossing motor control algorithm as a backup. The dyno is assembled, after a whole lot more hours in the machine…
We got the boards and we flowed them this week! Most things seem to be alive, as we immediately tested our safety limiting operations such as overcurrent and overvoltage protection, both of which deemed safe and working. After some preliminary investigation with the old small motor with the UPMC, we have determined that sinusoidal FOC does not like to run on a motor that is not sinusoidal, which is our small motor. Therefore, we have moved onto PID and other…
This week we finished designing the power and logic boards for the UPMC. We performed a full review of the components, schematics, and board layout, then ordered both the PCBs and 2 BOMs for each one. The PCBs should get here sometime next week. We also got some more equipment for demo, including an electronic twist throttle for the bike, along with a chain and 2 sprockets to connect the motor for the tabletop dyno. This week we will work…
This week we took another major step in firmware bringup via the janktroller. One of the hardest challenges of this project is to get rotor estimation using only the current and voltage feedback from the motor, which involves using ADCs to sample values. After we found our firmware bug, we started debugging the ADC values we were seeing using Rotate, the GUI that Bolaji made last week and continued firmware debugging. We have also started to work on integrating our…
This week we were able to get the motor spinning slowly. This involved a number of different things. The first one was bringup of the microcontroller’s timer to output PWM. Another accomplishment this week was writing a rudimentary control loop to spin the motor. We were also able to send information about the voltages and currents over UART. Overall, the motor spinning validates our basic hardware architecture, and we are confident in moving forward both with further firmware development, and…
Right now, the biggest risks to our project staying on schedule is the Janktroller not working or being broken, which would hamstring firmware development. To minimize the consequences of any accidents, we have a stock of extras of the parts most likely to be damaged in our testing (MOSFETS and gate drivers), as well as extra boards. We haven’t had to make any design changes yet. We are proceeding to move into basic firmware development with timer and ADC bringup,…