Scheduled curtain control: opens and closes at specific times
- Fusion 360 link: https://a360.co/35GVD2y
- See release for STEP/STL files
- Add stiffness/ribs to reduce outer frame deflection
- smaller PCB mounting holes (changed diametr from 3mm to 2.7mm)
- larger chamfer to match curtain rail I have at home
- tighter tolerance on bearing holder
- choose a lighter/smaller motor bracket
- choos weaker/smaller mechanical end stops
Features:
- ESP32 WROOM32 module
- 1x button
- 1x programmable LED
- 1x VEML6030 ambient light sensor
- 1x CP2104 USB-UART bridge
- 1x DRV8825 stepper motor driver with controllable STEP, DIR, /SLEEP and /ENA
Power:
- 3S Lipo battery, charged externally - not by this board
- LDO for USB power during programming
- 12V-3.3V 1A switching regulator
- resistor divider to monitor battery voltage
find out how long it lasts on 1 full charge
| Signal | Pin | Notes |
|---|---|---|
| Stepper_step | 16 | |
| Stepper_direction | 17 | |
| Stepper_enable | 4 | active low |
| Stepper_sleep | 5 | active low |
| Limit_switch_1 | 14 | has external pullup (R23) |
| Limit_switch_2 | 12 | no external pullup (R22), enable internal pullup |
| I2C_SCL | 22 | |
| I2C_SDA | 23 | |
| Ambient_light_sensor_int | 21 | |
| LED | 13 | |
| Button | 15 | enable internal pullup |
| Battery_voltage | I35, A1_7 | ADC1 channel 7 |
- Remove R22: the external pullup messes with communication with flash when programming.
- Add 10k pull up resistor on the nENA line so that when the MCU is asleep the stepper is not enabled.
- Wake up from sleep
- configure pins
- check battery level: if it is low blink the light and get stuck in this mode until the battery is recharged
- else, connect to wifi
- get time from NTP server
- check if it time to open/close the curtains
- open/close the curtains
- turn Wi-Fi OFF
- set wake up timer
- set wake up button
- go into deep sleep
NOTE: this firmware version doesn't use the end stops to check the curtain status because the end stops I have are too strong and don't work. It will be updated once I find new end stops.
- have a seperate RTC chip with less drift ?
- you can retain some simple things in RTC memory
RTC_DATA_ATTR int bootCount = 0;- change open/close times over wifi (from M5 stack core maybe?)
- "wireless wake up"?
- gather data from ambient light sensor over a week or so and see if there are any interesting patterns