Hello everyone. I flew my drone again after enabling RPM filtering, and it feels much better. I no longer see the vibrations that I noticed during my first flights, and the parameters seem to be working well. I also increased ATC_ACC_P_MAX and ATC_ACC_R_MAX to 280, which made the drone feel more responsive. I think I could increase them a little more for an even better response, but I’m not sure yet. The RPM filter seems to be working well, and my vibrations have been reduced significantly. I no longer hear the motor oscillations. During my first flights, the motors sounded like a tractor, and I really didn’t like it because my GoPro picked up that horrible sound. With the new settings, the noise is gone, and the drone flies smoother. I have uploaded the log from my latest flight with RPM filtering enabled. Please keep in mind that my flight controller has limited logging memory, so I recorded as much data as I could. If I don’t need to make any further changes, thank you all for your help and support throughout this tuning process. I really appreciate your time. Happy flying!
https://drive.google.com/file/d/1xbFdjLhTXOeTtOUEKHdp_YLFQIb8Zqp9/view?usp=sharing
It’s odd…
The RPM looks correct and that would give a (motor) noise frequency around 228Hz, which is where the notch is centred - but there is still a peak around 80Hz. Is there a fan fitted to the VTX or something like that??
That 80Hz peak will be why you seemed to get better performance by lowering the gyro filter to 60Hz, which would normally not be required. Leave it there.
Aside from that, noise is still VERY low and there is no need for a notch filter at all.
The frequency response and phase is MUCH better with the notch filter disabled. Even the 80Hz peak is below the level we would normally be concerned about.
INS_HNTCH_ENABLE,0
If that 80Hz peak became an issue then you could set a static notch for it.
Attitude control looks great.
Personally I would probably still run Autotune which will give you the maximum accelerations and PIDs - you can adjust down a bit from there for the best feel. I prefer the result with AUTOTUNE_AGGR,0.1 and these small quads suit tuning the Yaw D term too.
Usually these small quads end up with ATC_THR_MIX_MAN higher than 0.5 This is not adjusted by Autotune, you have to adjust it yourself to suit requirements.
OK, thank you! I will do some more flights this week with a few additional changes. If I run into any problems, I’ll let you know. Just to mention it, my VTX does not have a cooling fan. Thank you very much for all your help and support throughout this tuning process. I really appreciate the time you’ve taken to help me. 
@jhon_mesa, @xfacta, @MaxBuzz,
I really wonder what changed to bring the vibration levels down. They were terrible in the first 4.7 log. Was it a hardware change or was it just tuning? I’m very surprised if it was just tuning.
In my case, I installed and configured 4.7 from scratch instead of copying my parameters from 4.6. It’s possible that, due to the parameter name changes, some parameters were not migrated correctly and may have remained at their default values. After upgrading from 4.6 to 4.7, the increase in oscillations was already noticeable in my goggles. After enabling RPM filtering, those oscillations became even worse, but I’m not sure if this was the cause or if it was related to something else. I did not make any changes to the copter itself, apart from cleaning the copter.
I am not surprised about it being just tuning. I’ve had so much of it when playing around my own machines. That said, most of time I knew the most likely reason, like missing a resonance or too much noise let go to the D-term.
However, sometimes the vibration profile improves a lot (or gets much worse) by a bit of wire manipulation, which could accidentally happen when cleaning the copter… 
I’d probably agree that it may have been tuning. When the PIDs are quite wrong on these small quads they can really shake fast to the point of it being like vibrations.
That’s not real noise, it’s missing data probably because batch and raw logging enabled simultaneously.
Need INS_RAW_LOG_OPT, 0
Also fast attitude instead of medium in LOG_BITMASK might be helpful to see noise getting to the PIDs.
@OmkarSarkar204 simultaneous Batch (INS_RAW_LOG_OPT = 9) and raw logging (INS_LOG_BAT_MASK != 0) is a classic and is invalid. Only:
- INS_RAW_LOG_OPT = 9 and INS_LOG_BAT_MASK = 0
or - INS_RAW_LOG_OPT = 0 and INS_LOG_BAT_MASK != 0
Are allowed. Please add this to your log_analyser
AMC enforces this, but not all users use AMC 
.. and I wonder why we swapped the order on the “RAW”/“BAT” and “LOG” portions of the parameter name. I might try and re-order them so they’re consistent. Either:
- INS_LOG_RAW_xxx and INS_LOG_BAT_xxx
OR - INS_RAW_LOG_xxx and INS_BAT_LOG_xxx
Yeah, I’d actually said: