Home
Creators
Search Recent Random
Posts
Search DMs Popular Hash Lookup Tags Random
Importer
Import FAQ
Filehaus
Recent
Register Favorites Keys Review DMs Logout
Home Creators Posts Import Register Favorites Logout

PTB PRO support meeting #1 (Patreon)

Published:
2024-05-19 19:25:03
Imported:
2026-06
haven't archived this post yet. have a subscription? use the importer!

Videos

  • 309155152.mp4

Downloads

  • 309155152.mp4

Missing 1 video.

Comments

Simofly

Amazing service , sorry I couldn’t be present

Ivan

Brian, can i ask you one thing about phase delay? assuming that I need to apply a certain value of filters and that these cause me a delay in degrees almost close to 90°... is there a way to counteract the resulting problems? propwash or possible flyaway?

The PIDtoolbox Guy
>>138446765

First, to be clear, when you say 90deg of phase delay, this is wrt to one freq, just wanna clarify that. So if you mean 90deg phase shift at say your copter’s propwash freq (what ever that is), then it would most certainly be flying pretty bad. To avoid a flyaway you would have to either (a) reduce PIDs (especially D), or (b) optimize and/or relax filters, even if that means letting more noise through. Finding that balance is an art. Best to ask why you’re having to filter that much in the first place. Is there something mechanical or electrically that can be done so you don’t have to filter that much? Is there a way to optimize notches better so as to avoid so much low pass filtering? Am I optimizing filter Q? Those are the kinds of questions you wanna ask

Ivan

Let's see if I say well, with a delay of 2.4 ms and a propwash at 70hz ... I have a phase angle of about 60 degrees .. And now that I know it's 60 degrees? How should I behave? Do I have to assume that it is too much and therefore intervene as suggested? I'm sorry if I'm obtuse 😅😅😅

The PIDtoolbox Guy
>>138449164

This is not a “how to” discussion. That’s a level below what I’m talking about here. This is just a fundamental of filter delay. but, let’s say, if you’re sure it’s propwash and that it’s at 70hz then 2.4ms is not ideal. I’d try to relax filters. P term would be pushing 60deg out of phase with the wobble induced by propwash. D would be faster but still less than ideal. It’ll fly ok, but to shave off 0.5ms or so could have a big impact of flight. If I were to make a “rule of thumb”, I’d say you wanna be below ~40deg phase shift at propwash freqs, and if you’re really serious probably less than that. 70hz though? This sounds more like dterm flutter. Reducing filter delay can help this as well though

Ivan

BRIAN... it’s time for me to use your tuning service. 😅can you briefly explain to me how to organise? Do you usually concentrate everything in a couple of hours? I’m asking you so that I can organise it. At the moment I am ‘working’ on 8” which flies quite well but I would like to improve it.☺️

The PIDtoolbox Guy
>>138482322

Nothing to organize. When you book, we can get started over DM. Usually the tune can be completed over DM before the meeting and then the meeting may be used for questions. This helps when clients live many time zones away and frees them to to perform flight tests where and when they are able. It also leaves a “paper train” of communication notes for future reference

SuicideQuad_FPV

Meeting Overview - This was the first open zoom meeting for PTB PRO support. - Main topics included new features in PTB Pro v0.7. - Highlights covered the new Filter simulation and additions to the frequency-time spectrogram. New Features Discussed - Filter Simulation: Helps optimize filters for miniquads and cinelifters. - Frequency-Time Spectrogram: New enhancements were introduced. Q&A Session Highlights Phase Delay Concerns: 1) Explanation that a 90-degree phase delay relative to one frequency, for instance, the copter’s propwash frequency, can severely degrade flight performance. -Solutions proposed: Reducing PID values, especially the derivative (D) term. - Optimizing or relaxing filters even if it allows more noise, balancing the need for minimal delay against noise reduction. - Investigate mechanical or electrical changes to reduce the need for heavy filtering or optimize notch filters to minimize low-pass filtering. Specific Use Case: Example discussed with 2.4ms delay at 70Hz frequency, resulting in a 60-degree phase angle. Advice given included: - Reducing filter delay to improve flight performance. -'Rule of thumb' suggested to maintain phase shift below 40 degrees at critical frequencies to avoid significant impact on flight stability. - Mention that 70Hz propwash frequency might be more characteristic of D-term (derivative) flutter, and reducing filter delay could help mitigate this. Tuning Service: Explanation of how to organize tuning service sessions The continuation of weekly meetings with potential for changes based on participant feedback.

SuicideQuad_FPV

While most of this is still over my heads, I love that you are doing these and I get the chance to deep dive into tuning like this. Can you elaboate key takeaways of the topic on "Relevant Delay". As in what are action steps. You mentioned PT2 being more useful in most circumstances. Thank you!

kabakov

It would be great if you posted ZOOM meetings on YouTube with a private link. I don’t speak English and use the translation of videos in the Yandex browser. It allows me to watch YouTube videos with voice subtitles in my native language. Automatic translation of your videos helps me a lot in my hobby

The PIDtoolbox Guy
>>143120304

If you Click the little star shape cog in the upper right of the video window you can turn on subtitles. Not sure if it’s only English but subtitles certainly. You could also use sound to voice translator. I’m sure Google does this. iPhones do it.

vrjo

Super interesting Bryan, thank you! At min18 you talk about filterdelay in numbers. 3 inch should have less than 1ms and 5 inch not more than 2ms to fly good. Are you talking about Gyro delay, D-Term delay or both together? Thank you!

The PIDtoolbox Guy
>>151183786

gyro mainly. just roughly, you want filter delay 1-1.5ms or so for 3inch or less. for 5inch no more than about 2ms, preferrably less. I made a plot of this somewhere as a rough guide