Throttle Based Notch Issue

Need some help please…
Attempting to set the throttle notch, following the documentation. After the 1st flight I obtained the required thrust and frequency (about 75 Hz on a DIY 450 quad).

After setting up and executing the 2nd test flight, there is essentially no reduction of the
offending frequency. I have now attempted this three times.

Attached are the .bins for each flight, as well as a screen shot of the 2nd verification flight. Any suggestions are welcome, TU all!

DIY F450 quad with the latest 4.2.2 firmware
Latest MP
3DR PIXHAWK fmuv2 (memory restriction)

Data gather - Flight #arducopter:copter-41
https://drive.google.com/file/d/1ziOY9pBojZeNOfjnH-sb7-cHPcnbQUPn/view?usp=sharing

Analysis flight - https://drive.google.com/file/d/14y9fo2cR3CWE0VPDJGBMENs3u2XLfi5j/view?usp=sharing

Those links have an Access requirement. Remove it for help.

If it’s a 1Mb board Pixhawk1-1M is the correct firmware.

Dave,
Sry. Links here again: Open access

https://drive.google.com/file/d/1ziOY9pBojZeNOfjnH-sb7-cHPcnbQUPn/view?usp=sharing

https://drive.google.com/file/d/14y9fo2cR3CWE0VPDJGBMENs3u2XLfi5j/view?usp=sharing

TY

The pre-filter flight log looks fine and your Notch Filters settings are OK (I would use 76Hz center and 38Hz BW) but your post filter log is not configured properly to see the result.
This:
INS_LOG_BAT_MASK,0
INS_LOG_BAT_OPT,1
Should be configured as this:
INS_LOG_BAT_MASK,1
INS_LOG_BAT_OPT,2

There is a mechanical yaw bias present and it looks to be a bit unpowered:

Dave, TU! I see that a reply must be > 20 characters…TY!

Hi Dave, I totally messed up yesterday, and after the permissions confusion, I submitted an incorrect post-notch-flight log. I am sorry to have wasted your time… I realized my mistake after seeing your RCOU screen shot, which is from a different quad, an IRIS+,
which yes, is underpowered, and yes, has a twisted arm. On the bench now…

If you can take another look, attached is the correct post notch test flight log, which did have the parameters set as indicated. THANK YOU VERY MUCH!

https://drive.google.com/file/d/1l7vFsMHRzsPXm5Hl0aWI9XrDNGmMm9qT/view?usp=sharing

OK, I suppose this looks OK but you have removed the pre-notch log so can’t compare. When you said there was no difference between pre and post were you looking at the Y-axis scale of the FFT graph? The signature might look similar but the magnitude is probably greatly reduced in the post filter data.

Dave, noise analysis flight here. Yes, looking at the Y (magnitude) axis.
Thanks again.

https://drive.google.com/file/d/1yCmz2JEfDZHLe1i8QAQSEdot33wbRq7L/view?usp=sharing

So > than an order of magnitude less post filter. I would say it’s working…

Pre filter FFT
Post filter FFT

Dave, thanks! I was just responding - I placed the graphs side by side and then noted the Y-axis scaling had changed. Off to fly, have a good one! --John

1 Like