Not able to filter out properly imu data using notch filters

I am trying to set up notch filters to filter out noises of gyro data and I faced with the issue that not all noises are filtered out properly.
As an example, with this notch filter’s configuration:


I have the following results:

Fundamental and its first harmonic are filtered out as expected from the first notch filter setup. But I would also expect that my second-notch filter(based on FFT) should identify and filter out noises starting from 230 Hz, but as far as I can see that is not happening. So, where am I wrong? Why does the FFT notch filter not identify noises around 230 Hz?
This is the log file that I used for analysis: logs

Your BW is too big for multisource, that will cause a lot of lag at low frequencies.
Change it to 15 or 20

No need for a second notch filter -80dB means that it is irrelevant noise.

Thanks for your reply)
Could you please explain why BW equals to 30 is too big for multisource, because I did not find any information on Ardupilot wiki that could explain that? Also how do you know that -80dB is irrelevant noise? Is it always irrelevant noise for all copters or maybe, there are some criteria, that can be used to identify that?

If you want someone to review that log post a link to .bin log file.
But, your vibe levels are too high anyway (with clipping), address that 1st.

It would be great if you could look at my logs. I changed BW to 20 and disabled a second notch filter as amilcarlucas advised me, but noises around 365 Hz becomes greater than -80dB. Does that mean that now I need to add again a second notch to handle noise around 365 Hz? This is a log file that I used for analysis - logs

This is what you have pre-filter:

If you use these settings:

This is the estimated result:

Use the least amount of filter required to do the job. Make these settings and test it. Use Linear scaling and use the Estimated Post-Filter Function and you can fine tune this yourself.

1 Like