FLOW_FXSCALER: is the manual up-to-date, or I just need to set it to -800?

I have just connected a Matek 3901-l0x optical&lidar sensor and got stuck with its tuning.

The manual suggests equating OF.flowX and OF.bodyX by adjusting FLOW_FXSCALER. It appears that in my setup the parameters OF.flowX and OF.bodyX are something like 5x apart from each other, namely, OF.bodyX is five times smaller than OF.flowX. As per documentation of FLOW_FXSCALER, as long as I understand it right, the decrease by 5x shall be achieved by setting FLOW_FXSCALER = -800, which is way outside of the recommended range [-200; 200].

What bothers me more is that OF.flowX is radians/second, whereas OF.bodyX is meters/second. At least the APMPlanner tells me these dimensions. Shall that not mean that their ratio is to change when the altitude changes? And yet another thing is that IMU.GyrX is radians/second but is nearly equal to OF.flowX.

So I seem to misunderstand something fundamentally. Where am I wrong?

The dataflash log is attached: 22-02-06-19-XX-XX.bin.zip (520.7 KB)

Have you figured this out yet?

Well, not yet. I plan to use 4.2.0’s automatic calibration to find out the truth. The first attempt to use it showed to me that I need more battery (and tolerance) than I had at that time in order to collect the data. The next run is to come, hopefully this year :-E

For those reading/watching this post. My attempts to run the in-flight calibration failed, I guess for various reasons, the primary being the very short lidar range. I decided to set the scalers to -800 as initially thought, which makes flowX/bodyX and flowY/bodyY curves match as well as they can. That seems to work on the bench. Will try to test that in-flight in the coming days…

1 Like