kriswiner
commented
6 years ago Well if you have more than one MPU9250 you will have to have separate q[4] functions, one for each, separate sensor data a[3], g[3], m[3], etc and also separate yaw, pitch and roll, one for each, no?
On Sun, Jul 8, 2018 at 6:52 AM, Constantinos Theophilou < notifications@github.com> wrote:
Hello Mr. Kris,
First of all let me thank you for your work regarding IMUs and how to get readings from them.
Right now I have an arduino sketch that takes the readings from a single sensor and performs either the mahony or the madgwick filter and prints the pitch, roll, yaw values on the screen as can be seen in image below [image: image] https://user-images.githubusercontent.com/10788292/42420219-62e4ea42-82ba-11e8-9fab-fd5326bf2005.png
Now my problems is that I want to connect more sensors. I have researched a great while before finding your post about the magnetometer address being the same even if the MPUs address changes when pulling AD0 HIGH or LOW. In that post you mentioned that a good way to connect more than a single IMU and get the magnetometer readings from the correct IMU is probably to use an I2C expander like the TCA9548A.
That I did and it worked like a charm. Nonetheless my problem is now arising from a separate problem. I have successfully connected up to 5 MPUs and get different raw readings for each individual MPU. Using your library (basically Sparkfun's but its all taken from your code) I get the problem again by getting all MPUs to affect the same quaternion instead of an individual one. The example serial window below shows that the RAW readings are independent of one another but the quaternion corresponding to each IMU is changing for all IMUs regardles of which one is moving or not:
[image: image] https://user-images.githubusercontent.com/10788292/42420373-9e2ebe00-82bc-11e8-9b3f-0de2210d4c7f.png
This example is from a test using 2 MPU9255 and can be distinguished from the id label on the very left of each serial output line. The raw readings are perfectly individual but the quaternions are exactly the same.
And here is an example with 2 MPU that can be compared to the first image above where only one is moved: [image: image] https://user-images.githubusercontent.com/10788292/42420490-81b1f222-82be-11e8-8997-085e86251718.png
I think my problem is developing from the getQ function (from quaternionFilters.ino which gets confused as to which quaternion to return. Am I right by thinking this?
Here is my code:
kris - Copy.zip https://github.com/kriswiner/MPU9250/files/2173777/kris.-.Copy.zip Once again thank you for everything and any help would be kindly appreciated as this is going to be part of my dissertation for my master's degree.
All the best, Constantinos.
— You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub https://github.com/kriswiner/MPU9250/issues/294, or mute the thread https://github.com/notifications/unsubscribe-auth/AGY1ql_1SDdNM0f03YV74zqCiX6NJ0Opks5uEg6vgaJpZM4VGu8t .
ct14872
Hello Mr. Kris,
First of all let me thank you for your work regarding IMUs and how to get readings from them.
Right now I have an arduino sketch that takes the readings from a single sensor and performs either the mahony or the madgwick filter and prints the pitch, roll, yaw values on the screen as can be seen in image below
Now my problems is that I want to connect more sensors. I have researched a great while before finding your post about the magnetometer address being the same even if the MPUs address changes when pulling AD0 HIGH or LOW. In that post you mentioned that a good way to connect more than a single IMU and get the magnetometer readings from the correct IMU is probably to use an I2C expander like the TCA9548A.
That I did and it worked like a charm. Nonetheless my problem is now arising from a separate problem. I have successfully connected up to 5 MPUs and get different raw readings for each individual MPU. Using your library (basically Sparkfun's but its all taken from your code) I get the problem again by getting all MPUs to affect the same quaternion instead of an individual one. The example serial window below shows that the RAW readings are independent of one another but the quaternion corresponding to each IMU is changing for all IMUs regardles of which one is moving or not:
This example is from a test using 2 MPU9255 and can be distinguished from the id label on the very left of each serial output line. The raw readings are perfectly individual but the quaternions are exactly the same.
And here is an example with 2 MPU that can be compared to the first image above where only one is moved:
I think my problem is developing from the getQ function (from quaternionFilters.ino which gets confused as to which quaternion to return. Am I right by thinking this?
Here is my code:
kris - Copy.zip Once again thank you for everything and any help would be kindly appreciated as this is going to be part of my dissertation for my master's degree.
All the best, Constantinos.