Sensors, Electronics and Prototyping

Quaternion Output Drift

Welcome to Redshift Labs Forums UM7 Product Support Quaternion Output Drift

Viewing 10 posts - 1 through 10 (of 10 total)
  • Author
  • #638


    does my sensor UM-7 has a heavy drift problem?
    I’m reading out the Quaternion to get the 360° orientation of the yaw-axis.
    When I’m stopping the movement of the turn it still moves very slow like 30-45° away.
    In the tech specs you advertising with “0.5 degree angle repeatability”.

    Is this a Bug in the AHRS algorithm? I need a very stable 360° yaw-axis orientation RPY (I guess RPY it will not work 360° because you have the atan-problem) or Quaternion with +/-15deg accuracy.

    Is there a firmwareupdate available?



    Hi sven,

    To estimate yaw (or the yaw component of the quaternion), the magnetometer on the UM7 must be calibrated. You can use the Serial Interface to collect data, compute the calibration, and write the settings to the device. It can also help to issue a “zero rate gyros” command.



    I also met this problem. I did mag calibration of UM7, I also “zero gyro rate” and “set mag reference” at start up.
    At the beginning the yaw data is good, but when the sensor begin to move, yaw will drift slowly to about 30 degree away.

    Is there any methods to fix this?



    If the yaw angle isn’t remaining accurate, it would be caused by one of the following:

    1) The magnetic field is distorted badly enough that the sensor detects it and stops using it,
    2) Mag calibration is bad, or
    3) The magnetic field is distorted, but not in a detectable way and the magnetic field measurements are pulling the yaw estimate off.


    Could you please show me the link to the Calibration Tutorial?


    IS there a new Firmware? I don’t think its a magnetic field distortion that is affecting the Quart. output.
    Even my iphone compass or the Arudino is working fine.
    Maybe I’m doing the calibration wrong, is there a PDF?


    I am using the UM7 with arduino board and I also need the true quaternion values. Is it possible to calibrate the magnetometer without using USB expansion board & serial interface? If the answer is yes, then how can I do it?



    ..I mean is there any chance to use CHR Serial Interface via arduino connection? I tried but I can’t. I got “the port is using by another device” error. May be I am doing something wrong. If it is possible, could you explain in details about what I have to do?


    Hi Tanis,
    I cannot see any easy way to calibrate your unit without connecting to the serial interface software. To do so, you will need any USB to TTL converter, we recommend something with an FTDI chip set.
    like ours

    I would suggest buying a USB expansion board it will make your life so much easier, although it would be a fun project to do, it really depends on the time you have to commit to the implementation.

    The issue with calibrating and parsing the data with an Arduino is you would need to write a full parser and interface then pipe that through a VCOM port.

    With that said redshift labs will be releasing a full Arduino interface in the coming months, however that wont help you now, but keep it in mind.

    Michael Malloy


    (Awaiting moderation)

Viewing 10 posts - 1 through 10 (of 10 total)
  • You must be logged in to reply to this topic.