OpenMarine
Issues connecting and calibrating IMU. - Printable Version

+- OpenMarine (https://forum.openmarine.net)
+-- Forum: Pypilot (https://forum.openmarine.net/forumdisplay.php?fid=17)
+--- Forum: General discussion (https://forum.openmarine.net/forumdisplay.php?fid=18)
+--- Thread: Issues connecting and calibrating IMU. (/showthread.php?tid=5251)



Issues connecting and calibrating IMU. - shinken - 2024-03-14

I decided to get an IMU to get heading and other data.
The first IMU I bought (IMU-9250) wouldn't be recognized by Pypilot.

[Image: attachment.php?attachmentid=287606&thumb=1&d=1710437203]

I tried different jumper cables, different breadboard. Nothing. So I returned it and got another one.

I tried it this AM and was happy to see that it was recognized as 925X.

[Image: attachment.php?attachmentid=287607&thumb=1&d=1710437437]

So I proceeded to calibrate it using the instructions in the manual for OpenPlotter. I am not sure if the sphere is supposed to rotate or not but that was not happening and I was getting an error of scale out of range.

[Image: attachment.php?attachmentid=287608&thumb=1&d=1710437592]

And the boat was taking a dive apparently.

[Image: attachment.php?attachmentid=287609&thumb=1&d=1710437664]

I left it there and went to lunch planning to try again in the PM. When I returned I noticed that the new IMU was now not being recognized.

[Image: attachment.php?attachmentid=287610&thumb=1&d=1710437783]

I rebooted and it recognized it again. I got a totally different yaw, pitch, roll and heading parameters from this AM and it seems like it only sent ONE set of readings and then froze and now it's not being recognized again.

[Image: attachment.php?attachmentid=287611&thumb=1&d=1710437965]

Is this IMU also a bad one? Or am I missing something?


RE: Issues connecting and calibrating IMU. - seandepagnier - 2024-03-15

yes, it is probably bad too..

I used to use the mpu9255. I threw away about 60% of the ones I got because often even if all the sensors worked (many did not) the accelerometer biases would be so huge that they could not be used in the normal +- 2g scale. In theory you could modify the software to support them with a +- 4 g scale but this would reduce resolution.

I cannot say for sure what the problem is, but only that I suggest icm20948 as so far the quality has been perfect for these.