Post Go back to editing

ADXRS646-EP Questions

1) In datasheet 'Linear Acceleration Effect' for all axes is provided. Since it is a gyro which is mostly measuring the rotational rate, can you provide test data for 'Angular' acceleration effect for the remaining rotational axes?

2) May I know 'Vibration Rectification' parameter mentioned in the datasheet is for which axis?

3) Sensor Resonant Frequency parameter mentioned in the text is 17.5KHz whereas from the Figure 15, the same looks like it is 1.75KHz. Can you confirm which is right?

4) Can you provide the transfer function of the sensor? In Figure 15 of datasheet, magnitude and phase response for Cout = 470pF is provided. We want to determine the magnitude and phase response for a different value of Cout.

5) Do you have any test results that show the effect on measured Z axis gyro rate due to coupling from angular rotations in Y axis and X axis?

  • Hi

    1) The linear acceleration effect is seeking to quantify the response on the sensor (which measure rate of rotation), to a linear motion profile (defined in terms of acceleration). It seems like you might be trying to quantify something like “jerk rate” or the step response. Is that correct? 

    2) In this sensor, the response to vibration is greatest, when the linear acceleration vector is normal to the rate axis of rotation and is 45-degrees from the longitudinal and lateral axes, as defined in the following figure.  This provides 4 different “maxima” vectors, within the spherical space.  While not drawn for this purpose, the arrow associated with the A1 pin is pretty close to one of the maxima scenarios…

    3) That figure is not the resonant frequency of the MEMS structure.  Table 1, in the datasheet, offers the correct resonant frequency. 

     4) For a few hundred Hertz, the model is dominated by the two filter poles, which you can configure, using external capacitors.  That is fairly straight forward.  What frequency range are you looking to model the response?  What requirements do you have, for the frequency response? If the phase is not that important in your case, you can use the recommendations on the Setting Bandwidth section, page 9, ADXRS646 datasheet Rev.C.

    The “bump,” which you have identified, represents the separation of resonant frequency in the mems structure and the resonant drive frequency.  Modelling that would probably require help from one of our technologists, who are in demand and will probably ask for more details on your requirements and your sensitivity to the key attributes.  What should I offer to them, to get this help (if you need >~400Hz of insight)?   

    5) Generally speaking, we cannot separate cross-axis sensitivity, from the effect of physical alignment errors, which most account for, during their calibration routines.  In fact, I consulted with a colleague, on our IMU team (who uses similar gyroscope cores) and he confirmed this.  Is this what you were looking for, or an I missing something?



  • Thank you for your answers. Here are my follow-ups,

    1) What I am interested is to determine how the output of the sensor will be affected due to angular movement along longitudinal and lateral axes. How is the coupling between motion in lateral and longitudinal axes w.r.t. rate axis and whether is there any test data showing the sensor's capability in rejecting this coupling? (Assume that there is no misalignment errors)

    2) Thanks

    3) Thanks

    4) I am operating in the region below <20Hz, so I no longer need this info. Thanks

    5) It is now same as question 1. 

    I have an additional question as well,

    6) The datasheet specifies the null voltage varies from 2.7V to 3.3V. On what factors does it vary? Is it purely temperature dependent? What is the nature of this variation - Is it a DC error or can null vary from 2.7V-3.3V in the form of random AC noise?