Simple project based on the Raspberry Pi Pico and MicroPython to control a battery powered screwdriver, based on the current load of its motor. Also controls battery charging.
The goal is to use this setup to power a (high quality) manual coffee grinder until all beans are ground and therefore the motor power consumption drops – which is (hopefully) detectable by measuring the battery voltage.
Status: Working basically – Hardware has been built, but mechanical stability is problematic 😬. Software seems to work fine so far – but automatic grind stopping is WIP. Additional basic checks have been done using the rp2040js simulator.
To turn the motor on, the powered screwdriver to be used here uses a pushbutton to pull an N-channel MOSFET up. This button press will be automated via GPIO by using a P-channel MOSFET in parallel to the button.
Another P-channel MOSFET will be able to switch the power jack on or off via GPIO to control charging of the screwdriver's battery.
The grinding process will be started using an additional pushbutton connected to GPIO.
The Pico is powered from the battery via a switch and a Schottky diode.
For voltage sensing, the Pico's on-board VSYS measurement circuit connected to ADC3 will be used. This (hopefully) avoids some headaches regarding the Pico's ADC peculiarities.
See also the hw_design files in this repo. Note that the latest HW design (hw_design_v3.jpeg) does contain an error in the power jack disconnection circuitry. The P-Channel MOSFET disconnecting the "Rest" from the power jack has been replaced by the circuit described (and simulated) in transistortest.asc. This enables the high-side switch to be normally closed – so that the battery can be charged even when the Pico is unpowered.
This is probably the most interesting part of this project. The Pico's ADC noise properties are not particularly good according to its datasheet. The goal is to capture with a high sampling rate and do some averaging to counteract this.
The idea on how to do this is to leverage the possibility to use DMA for ADC capturing, and also use DMA sniffing to sum up the samples in hardware. This way, software only has to do a division as needed.
Released under the MIT license. Copyright (c) 2022 Tobias Modschiedler