This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TMAG5170-Q1: Temperature drift

Part Number: TMAG5170-Q1
Other Parts Discussed in Thread: TMAG5170

Dear Sir, Madam,

I have an other question about the temperature drift of the TMAG5170.

How does the temperature compensation coefficient work ?

For instance, I take a sensor set to the +/-50mT range and I detect 45mT maximum at 25°C. At -40°C, the magnetic field measured could be worth between 43,065mT (45 - 4.3%*45 = 45 - 1.935) and 46,935mT (45 + 4.3%*45 = 45 + 1.935). I saw on the datasheet the temperature compensation for NdBFe magnet is 0.12%/°C.

Is the temperature compensation worth 0.12%*(25 - (-40)) = 7.8% at -40°C ?

How do I have to apply this temperature compensation to the magnetic field values ? Can you provide me a formula in order to explain that please ?

Best regards.

Arnaud

  • Arnaud, 

    Thank you for reaching out again.  For neodymium magnets, the magnetic field weakens as the magnet heats.  This occurs at about -0.12% / C.  The temperature compensation in TMAG5170 is intended to counter this effect.  So, with the setting enabled at +0.12 % / C, the sensitivity gain will increase to balance out the negative effect the magnet experiences.  The device output would automatically be adjusted for this and your system should track more consistent output values from the sensor across temperature.  If the magnet is exposed to different temperatures than the sensor, however, it would be better to set compensation to 0 and then apply any needed corrections through alternate temperature sensing options.

    Thanks,

    Scott

  • Scott,

    Thank you for your answer.

    In fact I need to know how this value is applied during the correction process because I have to justify it for the customer of our project. It's for this reason I ask you to give me a formula which shows how use the correction value in order to compensate the temperature drift.

    Best regards.

    Arnaud

  • Arnaud,

    Lets assume:

    • Tcm = tempco of the magnet in %/C
    • Tcs = tempco of the sensor. in %/C
    • T = Temperature
    • B = mangetic flux density produced by the magnet at room temperature in mT
    • FSR = full scale output range of the sensor

    Then  the output will generally follow this behavior with temperature

    Output = 2^16 / FSR * B * (1 + Tcm*(T-25)/100) * (1 + Tcs*(T-25)/100)

  • Scott,

    I don't understand how to use this formula, can you provide me an example in order to illustrate this please ?

    Best regards.

    Arnaud

  • Arnaud,

    The default sensitivity range of TMAG5170A1 is +/- 50 mT.  This would be a total FSR of 100 mT.  Supposing an input of 20 mT at 25C we can estimate how the device corrects for this variation by entering the actual ambient temperature for T.  For example, we might be operating at 50 C.  This would result in a negative drift for the magnet and a positive drift for the sensor.  

    Since Tcm = -0.12%/C, the effective percentage change will be -3%.  This part of the equation is represented by (25C * -0.12%/C).  Likewise, the Tcs would produce a +3% change to sensitivity.

    Plugging this into the equation we have:

    Output = 2^16/100 mT * 20mT * (1.03)*(0.97)

    Thanks,

    Scott