FDC1004: Offset on CIN3 channel in liquid level sensor

Aidan,

Thank you for your post.
Can you share some more info on your design, like the sensor configuration/layout and schematic?
Just to confirm, is the reference you mentioned TIDA-00317?

Regards,
John

Hey John, Thank you for getting back to me. Yes, TIDA-00317 is the reference I used. I have been triggering single differential measurements between CINx and CIN4 (floating, no electrode). Please let me know if you have any thoughts.

I'm not sure how to upload files here, please see the imgur link below for schematic, layer, and code. Or let me know if there's another way I can send you the files, such as by email.

Imgur Link to Design Pictures

Thank you,

Aidan 

Aidan,

Please upload the files to this thread by clicking Insert/(Image/video/file). 

Regards,
John

Oops, I didn't see the grey "upload" button.

float CapacitiveSensor::singleMeasurement(int measurementNumber) 
{

  if (measurementNumber < 1 || measurementNumber > 4) 
    return -1000.0;

  // Trigger Measurement
  uint16_t sensorConfig = read16(FDC_CONF);
  sensorConfig |= 1 << (8 - measurementNumber); 
  write16(FDC_CONF, sensorConfig);

  long start_time = millis();
  while (millis() - start_time < TIMEOUT_MS) 
  {
    sensorConfig = read16(FDC_CONF);
    if (sensorConfig & (1 << (4 - measurementNumber))) // Check if measurement is completed
      return readMeasurement(measurementNumber);
      
  }
  return -3000.0;
}

float CapacitiveSensor::readMeasurement(int measurementNumber) 
{
  // Returns the capacitance of the measurmenet in pF
    if (measurementNumber < 1 || measurementNumber > 4) 
      return -2000.0;

    uint16_t sensorConfig = read16(FDC_CONF); 
    if (!(sensorConfig & (1 << (4 - measurementNumber)))) // Check if measurement is completed
        return -3000.0;

    uint8_t registerMSB = (measurementNumber-1) * 2;
    uint8_t registerLSB = registerMSB + 1;

    // Read 24-bit measurement. Capacitance has a fixed decimal point with 5 bits before and 19 bits after
    uint16_t msb = read16(registerMSB); // D[15:0] contains the most significant 16 bits of the measurement
    uint16_t lsb = read16(registerLSB) >> 8; // D[15:8] contains the least significant 8 bits of the measurement

    // Serial.println(msb, BIN);
    // Serial.println(lsb, BIN);

    int32_t rawValue = (((uint32_t)(msb)) << 8) | lsb;

    int sign = 1;
    if (rawValue & 0x800000) { // Negative number
      sign = -1;
      rawValue |= 0xFF000000;
      rawValue = ~rawValue + 1;
    }
    
    uint16_t left_bits = rawValue >> 19;
    uint32_t right_bits = rawValue & 0x7FFFF;

    float capacitance = sign * (left_bits + (float)right_bits/524288.0);
     
    return capacitance;
}

Aidan,

Thanks for the info. I can't see anything obvious.
Is it possible to disconnect the individual ENV sensor from the FDC1004 and see the effect on the capacitance reading?
Also, is there a ground plane located near the ENV sensor?

Regards,
John