Fantastic! Thanks for the support, glad you enjoy the technical deep dives! 🤓

I appreciate that! Glad you enjoyed the video.

Glad it was helpful! Always love to read your comments Brandon!

Glad to help!

I'll be doing a quick tuning guide at the end of this series, should be just what you are after!

@@ChrisRosser amazing, thank you very much!

I'll look into doing a GN style build mat!

Hope you enjoyed it!

Thanks! I run these filters on quads from 1.8" props up to 7inch props. I agree that for larger quads additional filtering may be necessary, hopefully for smaller quads my recommendation should be OK for most.

BF 4.3 will be much better at handling a wide range of differently sized quads with similar filter strategies. The signals going into the PID controller are way cleaner than on previous versions.

I run most of my builds on EMU, but I also run a few birds BF nightly builds (whoop, 3.5", 5",7") and I'm really impressed with how good the dyn notch work without a lot of tuning but if you have blackbox available you can get rid of a lot of latency by using dumb filters at the wright spot.

@@volkergoe Yeah. BF 4.3 also shaves off a lot of latency off of the RPM filter because the notches get turned off at low RPMs (crossfading). Now you can get perfectly clean gyro signals with delays under 1ms.

@@volkergoe just to clarify that a dynamic notch at e.g. 150Hz creates the exact same delay as a static or dumb notch at 150Hz. There is no delay benefit switching from dynamic notch to static notch. The dynamic notch uses more CPU cycles to calculate but the filter delay is the same (it doesn't depend on CPU usage)

This method is applicable to any quad, and if you watch the video, there's usually little pointers added in specifically for AOS frames (or any frame with similar resonance properties.)

No problem! Hope you enjoyed the technical detail 🤓

Thanks! 👍 More to come to get your quad flying amazing on BF 4.3.

You're welcome! Hope you find it useful.

It's time for the deep dive!

@@ChrisRosser Deep dive it was! Karate tune and AOS tune filter outputs does not seem that different to me. And because of my choice of antenna I ll probably better use Karate. Maybe you can make an additional video about your filter comparison tool. It might be easy to use for who knows what he is doing but probably doesn't make sense for many. Looking forward for part 4 !

if they sound any thing but nice then you are to far...

Yep should just be super smooth sounding like a desk fan running at constant RPM.

I'm also stuck at this step because from slider to 1.0 to 2.0 i don't hear any difference.

@@LAM_FPV Which slider? If you are tuning filters following the guide you wouldn't be using the filter sliders...

@@ChrisRosser in fact i'm using the slider to move 5 by 5 the dterm biquad filter as you suggest isn't it ? It seems to work with the last nightly build

It's worth checking all 3 axes. I just showed roll here because it was representative of pitch as well and yaw had very little noise of interest. In some previous videos I've looked at all 3 axes but it's only worth it if there is something of value to show 😁

found the tickbox that says profile dependent filter settings and turned it off. that opened it up.

Yep I'm planning on doing some presets soon!

We're having some quality control issues with the arm carbon from the supplier (typical new product stuff). Hopefully it will be sorted out in a few days but we're not taking more orders in the meantime.

No BF4.3 makes BMI270=MPU6000 in terms of noise. If you're getting noisy motors on the AOS 5 V2 then I would suspect in order: VTX antenna mounting (long and wobbly), stack mounting (plastic not steel and/or no base nut). Drop some photos in the discord or send me an email.

I will do. I currently plan to do a tuning guide for Emu 1.0

@@ChrisRosserDo you mean 0.4.0? I suppose. So nice! Can't wait ! Not that I'm especially fan of Emuflight, but on my heavy 6s 3" drone (cinerat with propguard, approximately 485g auw), I never found a way to make it fly decently with Betaflight (I'm not good in tuning, just gettin' better 'cause of you and some other), but it fly like a "charm" with Emu. I would like to understand why/ how... The hobby needs more Chris Rosser... I mean people who help, share their knowledges inna scientifical way, accessibles, ... Well, I hope you'll show the way (as UAV tech also does) to a new (and better way) to do videos about fpv drones and tuning. And, cherry on the cake, you build very interesting frames!

@@marcalbert8980 Hi Marc, 1.0 should launch shortly after BF 4.3 and merge in a lot of new features.

Hi! Typically noise on yaw isn't as bad as on pitch and roll. The dynamic notch should take care of a noise peak if you're using it. As long as your motors are cool then it's Ok to continue.

Hi Gadgetman, I think the best person to ask is Joshua Bardwell. He is much more experienced with troubleshooting than I am!

Hi Robert, thanks for the message. I'll raise this on the BF slack right now. 3.5.0 works acceptably in the meantime.

@@ChrisRosser that's for the quick response, love your content, can't wait for the 4.3 pid tune video.

There is a profile dependent/indepent filter settings box now that you need to turn on/off to get the numbers to appear.

That and using dynamic idle with low dshot idle 👍 Fortunately these problems are non-existent with 4.3!

Yeah the difference is very small really. Having a dynamic filter lets you achieve a higher cutoff at high throttle taking advantage of the gyroscopic stability of the motors. A static 100Hz Biquad is also a great choice.

I would start by slightly lowering the D gain filter cutoffs and see if that helps. Otherwise you can use the master slider to lower the PIDs but that risks the quad becoming quite unresponsive.

@@ChrisRosser thanks Chris can’t find the d-term filter cut of did you mean the d-gain slider ..also thinking about your suggestion to increase the PWM from 24hz to 48hz 😊

@@flipfpv4539 In the filter tab the Dterm filter cutoffs (they are 75Hz and 150Hz by default) if you need more specific help I have a great team of people in my discord server support hanger 😁

warm is okay, they should not be hot. Not as you can not touch them!!

Typically you don't want notches coming down too low as they add delay. Supafly's request deals with frames with bad resonance performance that have bad resonance peaks below 150Hz. Typically these are bigger rigs that have poor performance from a resonance standpoint. I don't have a problem with making the defaults more conservative to avoid flyaways on badly designed frames. But if you've got a decent frame you should definitely raise the min to 150 Hz to avoid filter delay.

what have u done?

If you set profile independent settings off in the header of the filter menu, you can enter the numbers instead of using the slider.

@@MaWeFPV I understand that. I don't understand why he goes against the defaults or where he comes up with the numbers from? I felt like he did a good job explaining the rest of the filters and the information to find on the graph. But lost me on the D term.

It's exactly the same as for 5". I usually use the same filter settings on all quads 5" and smaller and when I show the PID tuning process that is the same for whoops and cinelifters!

@@ChrisRosser awesome, thanks for the info

i downloaded the dataflash chip file and it loaded, i guess thats what i needed to do

Keep going up until they stop sounding smooth. You may be able to go up quite a bit on a 3" as they are typically lower noise than 5". You should use the AOS tune as it sounds like you have a clean build.

Have you seen the 533 tune for 4.3 that Evan posted? Haven’t tried it yet, but should be really good, just like all of them.

Was on Facebook last time I checked. I’ll try and find it and post it.

@@camblescott1195 could u please post it when u find it :)

drive.google.com/drive/mobile/folders/1uMQmG7brb-j88otxiatIiKiD1tmR9VBl?usp=sharing&fbclid=IwAR1MR_gfcrXwgUAqKweM8xaRq2vUIUqkiw75WmJH-dFxOiWMr-Oqp9EObnY Here is the drive link to 4.3 dump.

You may need to turn profile dependent setting to off to be able to change the cutoffs manually!

@@ChrisRosser Ok it works now, thanks for the help!

I'm only going to address the yaw twitching, as everything else is covered in the videos, and it'd be tough for me to explain it any better as I'm very much still learning like you. One possible explanation for the yaw twitching: You may be experiencing a hardware issue with your radio. If you have BB logging, take look at Yaw Setpoint for a flight during which you experienced yaw twitching, and look for sharp spikes in the plot that correlate. If you confirm spikes in the YSP plot that correlate with the twitches, first thing to try is carefully open your radio, find where the sensor for the horizontal (yaw,) axis of the left gimbal connects to the main board, and simply reseat the connector. Inspect the signal path for anything obvious. Reassemble and test for twitching. If you still have twitching, you want to go back in and perform thorough inspection of the sensor. Hall Effect gimbals use magnets. Check if any stray metal filings have collected on the magnet. Sometimes the magnet itself will begin to corrode, and bits of the nickel plating, and magnet, will bubble up and crumble loose, but remain magnetized to the magnet. In this case you want to get a new gimbal, but cleaning the loose material from the magnet might be work enough to keep you flying until your gimbal arrives.I think another common sensor type is optical, in which case you want to ensure it is free of dust, lint, or other debris. Remember these sensors can be delicate. So yeah, I thinks it's rare, or perhaps unlikely, to get yaw twitching from a bad config, or bad tune. Anyway, I hope you find the problem. Good luck!4

I'm afraid it's not. I get so annoyed that manufacturers still make FCs without Blackbox chips! You can get an openlager serial logging device or try my recommended settings and hope it works but without Blackbox those are your options.

I'll raise this on the BF slack.

You need to use betaflight configurator to set your FC in mass storage mode under the blackbox tab. If your FC has blackbox storage it will then appear as a usb for you to copy to your desktop. Then you open the file with BBE.

@@ChrisRosser, thank you for replying so quickly. I knew I was missing something!

Typically caused by clipping an object or a bump when taking off or landing.

Once you click on the gyro scaled legend bar you'll get the spectrogram (green spiky plot) full screen that then go to the top left and select frequency vs throttle 👍

There is a drop down in the top left.

Thanks!

Make sure you have release and release candidate selected in configurator 10.8 or newer.

You can but 110Hz should be more than fine so only go further if you want to test the limits!

I'm also trying to do that on my 4''. I went up to 145 with the D term lowpass 1 dynamic min cutoff frequency and still nothing. It sound completely the same. Can't tell the difference. I don't know what I'm doing wrong here...

@@gregor.potrebujes same here i went to 150 and nothing and for my d term max i went up too 220 it was fine so idk if im doing something wrong too lol

Well... this is a case where analyzing Blackbox logs is absolutely critical in being able to give you a direct answer to you question, the statement of "blackbox logging is not required" there is NO way to achieve "the best possible flight experience", without logging you are doing nothing more than guessing. Sure you could get a decent PID tune without blackbox but there is no way possible other than logging to get an accurate setting for filters. And with an accurate setting of your filters within the limitations of your craft will help lower phase delay, so that you can finetune your PID's to achieve the best possible flight experience. Without black box that is impossible

@@QuadflyerX I nowhere stated „böackbox is not required“ ? Blackbox is essential

@@jas-FPV my apologies for the misunderstanding, I was stating what was in the video 👍

Join the uavtech discord. People there are incredibly helpful when it comes to analysing your bb logs. Also, the iNav Community might be helpful. Paweł Spychalski who's a dev there seems to focus on larger drones. Check out his channel as well. Good luck with your quad my friend!

@@Deserthacker Thx, I‘m in these groups ;-) But still, my point is, that the filter/PID balance is not addressed very often but I think very important