Schematic review: I2C over 1.5m shielded cable (1.5m auger calibration device). ESP32 + PCA9515 + BMI160
Hi everyone,
I'm working on a device for calibrating an electric motor driving a 1.5-meter auger (single or dual-axis rotor mechanics). The sensors are mounted on the mechanics, while the motor itself could be located somewhat away from them (on top, bottom, or side of the assembly).
The environment isn't some heavy industrial setup with tons of high-power cables. The wires just run from the main control box to the mechanics. Still, I want my hardware design to be as robust as possible to avoid any unexpected EMI issues from the motor. I'd love a sanity check before I build it.
(attached the schematic image to this post).
The Setup:
- MCU: ESP32-C3 Super Mini.
- Sensors: 2x GY-BMI160 modules (addresses 0x68 and 0x69).
- Power: Will be powered by an 18650 battery or through the usb, completely isolated from the motor's power supply.
My approach to extending I2C (~1.5 meters):
- Capacitance buffering: The ESP32 is connected to two PCA9515DP I2C repeaters to physically isolate the bus capacitance.
- Cabling: I'm using ~1.5m of FTP (shielded twisted pair) cable for each sensor.
- Pair mapping: To avoid crosstalk, the pairs are strictly mapped:
Pair 1: SDA + GND
Pair 2: SCL + GND
Pair 3: 3V3 + GND
Pair 4: 3V3 + GND
Shielding: The FTP shield is grounded only at the MCU end to prevent ground loops.
Pull-ups: 3.3kOhm pull-up resistors on the ESP32 side, and another set of 3.3kOhm pull-ups after the repeaters (currently placed near the repeaters, but I'm not sure if I should moving them to the far end near the sensors or place two more pull-ups near accelerometers).
Tachometer (KY-003 Hall Sensor):
There is also a KY-003 sensor reading the RPM. To prevent false triggers from any occasional electrical noise on the 1.5m wire, I'm adding a hardware RC filter: a 10nF capacitor between the GPIO and GND, placed right near the ESP32 pin.
My Questions:
- Is this setup robust enough for this kind of application?
- Are the 3.3kOhm pull-ups sufficient for this cable length, or should I lower them to ~2.2kΩ(by adding 3.3kOhm pull-ups at the end of the cable near both accelerometers)?
- Does my approach with the 10nF capacitor on the Hall sensor line look correct?
Any feedback or suggestions for improvement would be greatly appreciated!