Hi Dave @dkemxr , I just went quickly to the field for a few minutes…
log file
Gal
Cool. I believe Mr. Piper @andypiper would suggest an Hnotch frequency of 120Hz, start with a bandwidth of ~40Hz and an attenuation of 40db. 2 harmonics should do it. INS_HNTCH_MODE to 1 and INS_HNTCH_REF to .25.This will deal with the motor (RPM) generated mode.
Then use the fixed notch for what could be the frame mode at 72Hz with a narrow BW of ~15db and an attenuation of 40.
As discussed, vibe levels at hover look pretty good.
1 Like
Dear Dave @dkemxr though I speak English, reasonably (ESL), the only thing I understood was the mode and the ref, could you please direct me to the other configuration? I believe will learn it somehow 
Gal
Sure Gal!
INS_HNTCH_FREQ (120)
INS_HNTCH_BW (40)
INS_HNTCH_ATT (40) if MP complains set it anyway
INS_HNTCH_HMNCS (2)
INS_HNTCH_REF (.25)
INS_HNTCH_MODE (1)
INS_NOTCH_FREQ (72)
INS_NOTCH_BW (15)
INS_NOTCH_ATT (40)
After setting these also change INS_LOG_BAT_OPT= 2 before a test flight.
1 Like
Wow, so simple 
Thank you Dave @dkemxr, I will try this first thing in the morning. Here it’s already 10:16 pm.
I am truly grateful for all your help on this
Gal
This is fun stuff, happy to help!
1 Like
Dave @dkemxr, If I understand this graph correctly, in three words, you nailed it
flight log
Short video of my quad this morning
and a huge thank you to @davidae for noticing that something was totally wrong in the first place
Now, not to scare you all pros, this noob is still hunting for the perfect tune so I guess I shall keep on adding a little more to this thread.
This community is the best, thank you all.
You are welcome! And your drone is very stable at hover, so congratulations! But I think you can still improve vibes with an anti-vibration platform with hard dampers. I have a custom made platform with these gray dampers (https://www.dronestore.cl/zenmuse/703-zenmuse-zh3-3d-part-042-rubber-damper-6958265106590.html), the grays are the hardest and that I use, and they give me the best results in my drone. I didn’t buy them from that store, but they are the same.
1 Like
David @davidae, I understand. Actually I believe it is a good idea, however I would have some trouble with mounting any kind of system other than the double sided tape I currently use.
I don’t think I would drive my drone as hard as I did in these tests but I sure going to keep that in mind.
Thank you again for your insights
Gal
Good morning Gal-Wow, the notch filters are really doing the job!! It quite dramatic.
Here is the before:
And After:
Look at the magnitude change between them. Very nice. I would carefully try some more agressive flying now. I don’t think you should have a problem but when I used a static notch filter with a center frequency too low it effected the control. If anything feels sluggish this could be it. But so far it looks great!. Nice work my friend.
1 Like
Dave @dkemxr,
It is so simple to fix, when someone as knowledgeable as you gives the full recipe.
I just love this material. With just a few parameters in the front and who knows how many lines of code (@andyp1per) a serious issue is handled.
Amazing stuff
Thank you my friend
Gal
1 Like
Gal you may want to either run AT again or do a manual tuning using a knob on channel 6
1 Like
Hi Dave @smartdave, I was trying to figure out how to proceed with the tune process
So I am wondering, just do each axis in my last aggr = 0.1 or should I go the full path again from 0.6 aggr all the way up?
Thank you
Gal
I myself, always do one axis at a time. If after the autotune it still feels a little off you can manually tune using channel 6 and assigning it to one of the rate gains
1 Like
But go for only the last at ie 0.1 right?
Yes correct. And you can keep doing them at .1 to hone in the tune if your want as well
1 Like
Well how about that? all you good people of the AP community, what a sight to my soar eyes 
Yesterday I replaced my top 16" props with 15x5.5 because I felt the top motors were very hot in my last test.
So this morning I went out for Autotune, 0/S wind.
test flight log.
It seems like I gained stability all over, the quad was quiet and all in all good experience.
We can see the roll vs desRoll is followed closely and in the pitch vs desPitch followed closely but there is that gap between lines. Can someone share an insight on that?
Thanks
Gal
That gap is only because the Y-axis scale’s are a little different from Autoscaling which is default.
This is what it looks liked with the same scale (manually done). Looks great Gal, super job!
I see you have set INS_LOG_BAT_MASK back to 0. That’s good, now that the notch filters are set. Those files can get very large with the batch sampler running all the time and you don’t need it for general tuning and log analysis.
1 Like
Hi Dave @dkemxr,
I can’t thank you enough and to everyone from the community for all the help.
Amazing
Gal
1 Like