TMAG3001: Immunity to external magnetic fields

Michael,

Welcome to E2E, thanks for reaching out with your question.  TMAG3001 will measure the total magnetic field at the sensor location, so stray fields can pose a challenge.

• If the timing of the stray fields correlate with specific positions, then you might be able to create a look up table. 
• Another possibility is that the fields are faster transient events and you might be able to integrate this noise out by over-sampling.  You could either increase the averaging on chip or capture multiple samples and average on the MCU
• If neither of the above are workable options, you might be able to create a differential signal by placing a second sensor.  Suppose we place two sensors side-by-side for an on-axis angle measurement.  If rotated 180° from each other the result will produce sin+, cos+, sin-, cos- outputs from two devices.  Any stray field would generally be very similar between the two devices.  If we subtract sensor2 outputs from sensor1, we could double our measured input amplitude while cancelling out the common mode stray field.  I've put together an example of how this could work with our free simulation tool: www.ti.com/TIMSS
  
  I've attached a configuration file which you can import into the TI Magnetic Sense Simulator here:
  
  Fullscreen Differential TMAG3301.json Download

  {
    "version": "3.5.0",
    "design_name": "",
    "magnet_id": 2,
    "poles": 2,
    "material_id": 1,
    "grade_id": 1,
    "select_remanence": "br_average",
    "remanence": 1200,
    "temperature": 20,
    "temperature_coefficient": -0.12,
    "coercivity": 10.9,
    "function_id": 4,
    "magnet_geometry": {
      "outer_diameter": 10,
      "height": 2
    },
    "magnet_position": {
      "x_position": 0,
      "y_position": 0,
      "z_position": 0
    },
    "magnet_angle": {
      "x_angle": 0,
      "y_angle": 0,
      "z_angle": 0
    },
    "magnet_movement": {
      "arc_length": 360
    },
    "sim_setting": {
      "angular_step_size": 1
    },
    "sensor": [
      {
        "sensor_id": "TMAG3001",
        "sensor_position": {
          "x_position": 1,
          "y_position": 0,
          "z_position": -5
        },
        "sensor_angle": {
          "x_angle": 0,
          "y_angle": 0,
          "z_angle": 0
        },
        "custom_inputs": {
          "variant": "TMAG3001A2YBGR",
          "applied_vcc": 3.3,
          "temperature_compensation": 0,
          "averaging": 1,
          "maximum_input": 120
        },
        "id": -1
      },
      {
        "sensor_id": "TMAG3001",
        "sensor_position": {
          "x_position": -1,
          "y_position": 0,
          "z_position": -5
        },
        "sensor_angle": {
          "x_angle": 0,
          "y_angle": 0,
          "z_angle": 180
        },
        "custom_inputs": {
          "variant": "TMAG3001A2YBGR",
          "applied_vcc": 3.3,
          "temperature_compensation": 0,
          "averaging": 1,
          "maximum_input": 120
        },
        "id": -2
      }
    ]
  }

Thanks, 

Scott

Michael,

Magnet fields decrease in magnitude exponentially as distance from the source increases.  Having your sensors close to each other will best match the stray component they are exposed to.  However, if you knew what the source might be (like a long current carrying wire) you might be able to place the second sensor further from the signal magnet, but equidistant to the wire.

The goal with the setup I suggested was to reject the field component in the XY plane, particularly when it is aligned to the signal.  If our two sensors produce values of:

Out1 = X + stray

Out2 = -X + stray

Then:

Out1-Out2 = (X + stray) - (-X + stray) = 2X

This will not work exactly the same when shear causes the alignment of the magnet to shift. The worst case condition is when the shear is in the same direction as the sensors are spaced apart.  When this happens, the magnet will travel towards one sensor and away from the other.  Your relative amplitude of Out1 and Out2 will not match.  

It is likely possible to create a similar schema with 3 sensors in such a way to create better overall rejection, but it will require a bit more effort to process the signals.  I am not quite sure if it would work out better to have them all in a straight line or if you would do better with them spaced to be 120° rotated about the center of the magnet.  I suspect that if you are able to work out the signal processing, that the latter option might work out more cleanly.

We don't yet have any device with stray field rejection implemented.  To minimize size TMAG3001 is going to be your best option.

Thanks,

Scott