Hello,
I set up my Alpha Datalink V2 according to the instructions:
I’d like to use the Datalink V2 as a tool to set up a Notch Filter, but I can’t find instructions on how to do this?
The datalink will provide ESC telemetry information. Just follow the Notch filter documentation for ESC telemetry.
I’m switching from a throttle-based notch filter. In this filter, I have the parameters set to INS_HNTCH_FREQ,52 and INS_HNTCH_BW,15
Can I use the exact same INS_HNTCH_FREQ,52 and INS_HNTCH_BW,15 value in the new ESC telemetry-based filter for the first flight?
Do you recommend using the multi-source INS_HNTCH_OPTS.2, or does its effectiveness depend on, for example, the size of the drone (in my case, a Hexa X, 22" with a 15kg MTOM)?
Yes, those do not change just because the data source changes.
Yes, that is always an improvement, but reduce the INS_HNTCH_BW to 8 if you do so.
I performed a clean test flight with the hotch filter disabled and INS_RAW_LOG_OPT,9. I set the filter parameters based on ESC telemetry, but I don’t understand why I have to use every other harmonic frequency value. If I select 1 and 2, they overlap. Am I doing something wrong?
Bin file: https://drive.google.com/file/d/1HyZm-4kCIB73Jxz04myVDZtUrx5NViYE/view?usp=sharing
It doesn’t appear to be tracking correctly. Perhaps the pole # is off?
I have no direct experience with these data links but the results look similar to when the pole count is wrong. Does 1500 RPM make sense with the Motor kV and voltage?
I have these motors: https://store.tmotor.com/product/mn605-s-kv170-motor-navigator-type.html?srsltid=AfmBOoo76MOKuPOaKW0icUjxTO4FfslHP-sST7VXQqZtx5ezYoC6Ao-1
The manufacturer specifies the configuration: 24N28P, so I set 28 poles.
Definitely not, there should be twice as many RPM.
Check the RPM with a tach. I haven’t used them but I see they are available for IOS and Android. Or, just adjust the Pole count in the script and see what happens.
In the tutorials from T-motor I found information that the number P should be divided by two, so this would be correct.
Read the file that you linked, and come back with specific questions 
1 Like
I performed another test flight, this time with new HobbyWing propellers and the correct ESC_HW_POLES,14 value. This time, the RPMs are being read correctly. I’ve attached the notch filter values I set. Are these sufficient?
Regarding Phase Lag, should I aim for the gray field to align as closely as possible with the graph line:
or should I aim for the field to extend far beyond the graph line?
In the frequency response graph, there should be a radio button that selects 180, your goal with the filters is to make sure that the place where the phase delay touches 180° is as large as possible while still rejecting an adequate amount of noise.
The gray shadows are simply how much the notch filters have moved through the part of the log you’re looking at and are just an indication to help you see how the phase delay and gain change throughout the flight. They’re neither good nor bad, but should be used as a check to see if the notch filters are tracking too low for your liking.
The notch filter settings you’ve shown in the graph above seem fine enough to start with, but I haven’t poured over the log yet. Did you post that somewhere already?
Thank you for the clarification. I’ll need some time to analyze your explanation and prepare questions.
In the meantime, here are the .bin files. One flight was performed before setting the filter:
and the second after setting the filter:
After setting the filter, I ran Autotune, so the second flight already takes into account the PID values set in completely calm weather.
Your notch filters seem fine, and you could probably get away with raising the gyro lowpass somewhat, but if the system is flying fine as is, then keep going!
I’ll test slightly increasing the INS_GYRO_FILTER,23-25 value.
What difference could I expect in the drone’s behavior by increasing the gyro filter value?
Is it possible to eliminate or reduce vertical drift while hovering? I recorded the drone’s behavior in calm weather, and the video shows that significant up-down correction occurs despite the absence of external factors: https://drive.google.com/file/d/1a5W7wP6-wAmITeBUoDSFDCgSLASmICl4/view?usp=sharing
Have you tuned the altitude controller?
Have you placed foam over your baro?
In the altitude controller, I adjusted these parameters according to the instructions:
ATC_THR_MIX_MAN,0.5
PSC_ACCZ_P,0.26 = MOT_THST_HOVER
PSC_ACCZ_I,0.52 = 2xMOT_THST_HOVER
Regarding the foam on the barometer, I haven’t modified anything on my Orange Cube Plus. Has anyone successfully done this with this FC
Good, that is correct
Generally, being able to increase the gyro lowpass will allow you to tune the vehicle tighter.
Do you have a log of this flight?