IMU sensor position questions

Hardware, Parts and Tools
1

First post. Wanted to start by saying thank you to all the contributors here. I’m working on a VESC build using the Little FOCer v3.1. I found that all the information needed to get one of these controllers working is already posted somewhere if you look for it. Still testing everything out and playing around with settings but am currently using ATR24 with dado’s XML. I found this configuration to be a really nice starting point to tinker from.

When first configuring the controller I was able to figure out the yaw offset issue on my own. The board didn’t seem to be responding correctly and when I looked at the realtime data I was able to see that when I moved the board around manually the pitch and roll readings were reversed.

So, this is where my questions come in. If a 90 degree offset in the software is needed, I’m led to believe that the controller board is mounted inside the box without consideration for the exact position of the IMU sensor itself. This surprises me. Why not design the hardware placement so that an offset configuration isn’t needed? To go a step further, hypothetically speaking, what would the perfect placement of the IMU sensor be if you could put it anywhere? Wouldn’t it be at or directly above the center of gravity? Wouldn’t that be directly above the axle when the board is level and not in the front controller box?

If I at least wanted to try putting the sensor on the center line of the board (front to back) and turning it 90 degrees so that the yaw offset setting isn’t needed, how would I do that? Does anyone know where specifically the IMU sensor is on the Little FOCer v3.1 controller board?

2

the answer is simple: different controllers, different IMU orientations

even the phsyically identical LFOCv3.0 and LFOCv3.1 have different IMUs and different orientations.

Also, different use cases will require different controller boxes. So there’s no 1-size-fits-all solution here…

Hopefully someone will improve the IMU Wizard in the VESC Tool soon to make this less difficult in the future, but it’s a necessary evil…

1 Like
3

If accounting for the IMU orientation through software works totally fine, why should we pose hardware restrictions upon ourselves by limiting how we can physically orient the controller? There really isn’t any downside to the need for an IMU Yaw Offset outside of the confusion that can be caused for people that aren’t aware of the need for it.

4

Definitely a fair point about hardware restrictions. I was envisioning a bracket that attaches to the fender mounts and positions an external IMU sensor directly above the center of the tire. I was curious if it would make any noticeable difference in the way the board feels while riding. A lot of hassle if it isn’t a significant difference though.

5

Gyro position doesn’t matter. The rotational rates at the fender are the same as the rates at the controller box. This may help: https://youtu.be/XJAeaFzDwko

6

That is helpful! For gyroscopic readings the placement of the sensor doesn’t matter. That makes sense.

For accelerometer readings, it does matter where the sensor is. But, does the balance app use accelerometer readings for anything?