Thanks for making it clear. I thought INS_HNTCH_REF has more authority over this.
Does this notch filter tune looks good to you?
Before:
This is the result with following:
INS_HNTCH_FREQ 48
INS_HNTCH_BW 24
INS_HNTCH_REF 0.45
I could not find any info on INS_HNTCH_ATT and what exactly this one does. I left it at 40 by default.
Hello,
I tried the Tuning today with a new model, this is the before results:
than with the following changes:
INS_HNTCH_FREQ 108
INS_HNTCH_BW 54
INS_HNTCH_REF 0,2051521
Would be great if someone could take a look and let me know if this is okay.
You could instead try
INS_HNTCH_FREQ 54
INS_HNTCH_BW 26
as I think you may have a first harmonic there
Thank you, will try. I forgot to mention the type of drone, it is a flat octocopter with 118cm dimension from motor to motor and 10kg total weight.
Andy, is there any word I don’t read correctly in INS_HNTCH_BW ? Because I see alot of _FREQ/2, especially from Shawn @xfacta I have the main peak at 120 Hz, starts rising at 110 Hz, falls to zero at 130 Hz. First harmonic at 240 is very shy with less width, but it’s there. I went for 20 Hz _BW, not 60 Hz
What matters is the Q factor which is a function of the ratio of freq/bw. Even if you have a very spikey noise peak it’s unlikely that we can track it accurately enough to make such a narrow notch work. So FREQ/2 is a good rule of thumb. You can try higher Q but you should look at an FFT for dynamic flight after to see how well the noise is being attenuated.
I´m trying to tune a 10" quad, but it seems, that the onboard FFT tracks the second harmonic. Sometimes it even switches to the first one. Any suggestions?
Try setting FFT_HMNC_PEAK
I tried to set FFT_HMNC_PEAK to 2, but that only makes it worse.
I ended up using the throttle based harmonic notch, which seems to work the best in my case.
Do you have a log you can share?
Hi Everyone,
I tried to do Harmonic notch filtering following the @andyp1per guidelines and noticed the first harmonic at 150Hz and second at 300 Hz, there is also one more notch at 74 Hz which I believe is due to Frame vibration (Correct me if I am wrong)
Also Attached are the Pre and Post Notch FFT Data. Kindly Go through it.
It is a quad plane with 18-inch prop and has performed dynamic notch using throttle based setting
Parameters Set
INS_HNTCH_ATT 40
INS_HNTCH_REF 0.35 (that is my Throttle hover value got through Motor throttle Hover)
INS_HNTCH_FREQ 150HZ
INS_HNTCH_BW 75Hz
INS_HNTCH_HMNCS 3
INS_NOTCH_ENABLE 1
INS_NOTCH_FREQ 74Hz
INS_NOTCH_BW 20Hz
and when I have changed
INS_HNTCH_HMNCS 7 for1st,2nd and 3 rd harmonics (not 11 because it was showing only 4th harmonics on my CUAV H7 pro FC I don’t remember now)
I got this
Let me know is it correct and still I am getting vibrations at 150 Hz. Is it normal or I need to do something else.
Just post a link to the log file.
Sure.
here it is.
https://drive.google.com/file/d/1O8NAS7MqMeI5ejSgjxGEUhsEjVXrX6Ha/view?usp=sharing
That link has an access requirement…
https://drive.google.com/file/d/1O8NAS7MqMeI5ejSgjxGEUhsEjVXrX6Ha/view?usp=sharing
kindly check this one
Do you have your test flight log where INS_LOG_BAT_OPT was set to 0?
And you don’t need to collect data on all IMU’s:
INS_LOG_BAT_MASK,3 1 is good.
yes, Here it is.
https://drive.google.com/file/d/111p5egXskBDkog4yrcZcQtif7G39xLuE/view?usp=sharing
I was just testing so collected two imu data to get more precise data.
Is it a problem?
Also after this notch filtering, I did autotune which resulted in a crash. Pretty bad day for me 