Inav: Recommended PIDs For Large Hex?

[CradyLab]

I replaced the aging Pixhawk on my 790mm hexacopter with a Matek F765-WING and was wondering if anyone had some tips for the PID settings in iNAV (the largest prop size mentioned in the configurator 12 inches and I'm running 15-inch props)

Thanks!

[issue-label-bot[bot]]

Issue-Label Bot is automatically applying the label Support to this issue, with a confidence of 0.89. Please mark this comment with :thumbsup: or :thumbsdown: to give our bot feedback!

Links: app homepage, dashboard and code for this bot.

[Vissie26]

Dear Cradylab,

I am flying a 690 size hexa and my quite good flying pids are:

Roll : P 155 I 35 D 45
Pitch: P 165 I 35 D 65
Yaw: P 145 I 55 D 0

I am currently on version 2.2.1

This should fly reasonable for a first try.

I hope you find this info useful.

[CradyLab]

Sweet! Thanks Vissie! Much appreciated

[pDR8iLQ4go]

@Vissie26: May you please post your complete dump? I have a 650 size hexacopter and would like to compare the parameterization. Which kind of props and motors do you use?

[Vissie26]

here you go.

hexa dump 2.3.0.txt

Props: https://hobbyking.com/en_us/multirotor-carbon-fiber-t-style-propeller-12x5-5-black-cw-ccw-2pcs.html

Motors: https://hobbyking.com/en_us/turnigy-multistar-4225-610kv-16pole-multi-rotor-outrunner.html

Let me know if you need more info.

[pDR8iLQ4go]

I'm using a Matek F405 CTR and cheaper stuff from AliExpress:

Gartt ML3508 700KV Motors: https://de.aliexpress.com/item/32751865371.html?spm=a2g0o.productlist.0.0.ebc25ddaBaUuAo&algo_pvid=e105fdd5-14fa-4e8c-9c03-5c23b005f712&algo_expid=e105fdd5-14fa-4e8c-9c03-5c23b005f712-0&btsid=42319929-d8ec-4b3b-a563-b61a6b45168f&ws_ab_test=searchweb0_0,searchweb201602_9,searchweb201603_55

12x55 Carbon Props: https://de.aliexpress.com/item/32917571592.html?spm=a2g0s.8937460.0.0.3bc22e0ehCftce

[Vissie26]

That combination should work in my opinion. What kind of battery are you using? I'm flying with a 4 cell 10.000 mAh.

[pDR8iLQ4go]

Yes, it works fine. But I am using an older version of INAV, here is the dump:

2019-07-19_1255_Carbon.txt

I have a 4S 5500mAh, flight time is about 15min.

[Vissie26]

if you drag the file onto the screen where you can reply it adds it to your message.

[pDR8iLQ4go]

Thanks!

[Vissie26]

your D values for pitch and roll seem a bit high. I had mine up that high and it gave worse results the higher i set them. My current values are around 60 and a P gain of 140-160 but my frame is 690mm.
You could try (P 120 I 35 D 55) for roll and (P 130 I 35 D 60) and see how it flies.

[pDR8iLQ4go]

I think, that I‘m going to update INAV first. Getting a sufficient filter setup was a nightmare. I hope that the new dynamic filters do a good job. On my smaller Q380 size frame, it seems to work.

[Vissie26]

If you activate dynamic filters the video from Paweł Spychalski might be useful if you haven't seen it already.
https://www.youtube.com/watch?v=2YFkDsnPnqE

[pDR8iLQ4go]

I know this video. Pawel told me, that the old notch filters have to be deactivated, in case of using dynamic filters. But in your dump, the notch filters are still parameterized. But you use the dynamic filters as well. Does that work in combination with the old notch filters?

[Vissie26]

as far as I know the old values are showimg but not used by the flightcontroller. I have followed pawels video step by step and he doesn't say anything about removing the old values.

[pDR8iLQ4go]

Yes, nowhere is explained what to do with the old/existing parameters. Maybe they will write it down sometime.

[Vissie26]

I'll send pawel a message asking about this. curious about what he has to say about this.

[pDR8iLQ4go]

Keep me informed. See you...

[pDR8iLQ4go]

I am using high D values to improve the stability when going down in the copters own turbulence (= propwhash?). But as mentioned, I’ll update INAV and try your D term.

[Vissie26]

I just got a reply from pawel.

''You might try disabling static gyro notches but one by one and verify
if motors are not overheating''

looks like we have to check blackbox data to verify if the dynamic filters are working or not.

[pDR8iLQ4go]

Have you planned to disable the „old“ (static) notch filters? If yes, please inform us regarding the results.

[Vissie26]

I am planning to disable them but the weather forecast isn't too good for the following days. I will report back when the weather gets better (less wind)

[Vissie26]

Just flown without static notch filters and it seems fine (No hot motors and weird sounds).

What I did was:

set gyro_notch1_hz = 0
Allowed range: 0 - 500

set gyro_notch1_cutoff = 1
Allowed range: 1 - 500

set gyro_notch2_hz = 0
Allowed range: 0 - 500

set gyro_notch2_cutoff = 1
Allowed range: 1 - 500

set dyn_notch range = LOW
Allowed range: HIGH MEDIUM LOW

set dyn_notch_min_hz = 75
Allowed range: 60 - 1000

I hope this is enough info.

[CradyLab]

Thanks!

[pDR8iLQ4go]

I've been updating my hexacopter as well. Still flying. Motors don't get warm at all.

@Vissie26: May you please post your dump once again?

[Vissie26]

I have upgraded to 2.4.0 in the meantime.

This is my current dump file:

dump 2.4.0.txt

[pDR8iLQ4go]

Most important settings are more or less the same:
Compare.pdf

I like "set inav_reset_home = EACH_ARM" to ensure that the home position is where I armed last time. And "set debug_mode = GYRO" to be able to use the Plasmatree's PID-Analyzer.

[Vissie26]

The reason I set ''set inav_reset_home = FIRST ARM'' is that if i accidentally flip the arm switch in flight, the home position isn't reset. Quite a rare occasion but i'd rather be safe than sorry.

Saw the debug mode parameter also. Was wondering how to use it exactly. Not much info on the wiki.

Would you be so kind to explain a bit about the function if you can? especially VIBE and Gyro?

[pDR8iLQ4go]

I'm not a real expert in that. But "set debug_mode = GYRO" seems to add some more details to the blackbox file. This setting is necessarry using Plasmatree's PID-Analyzer. See https://www.youtube.com/watch?v=VaqlLSQlJ4M and this hint:

[stale[bot]]

This issue / pull request has been automatically marked as stale because it has not had any activity in 60 days. The resources of the INAV team are limited, and so we are asking for your help.
This issue / pull request will be closed if no further activity occurs within two weeks.