PID Tuning Principles: Step #1 Follow-up - Finalize YOUR Filters

[@taterfpv]

Thanks man. Your videos are the first to help me get a grip on this. Videos telling me " You go by feel" weren't very helpful for the way my brain works I guess. Prior to watching your videos it was all black magic and voodoo and I essentially flew defaults. I actually look forward to tuning now. You are absolutely the number one source for this content.

[@d3c0deFPV]

This is often just how people who don't know how it actually works try to explain it. That's why we like people like mark, because real data is good :)

[@uavtech]

Thanks for the support!!😀

[@Dorkyand40]

Just great mark! This is what fpv needs. Not crap reviews. Beers on the way

[@airdronelongford2027]

I can't thank you enough. Changing from 48KHz to 24 made a huge difference. Had exactly this problem with noise on pitch. Thanks again , great job!

[@scrimdog3328]

waited all day to watch this after work. Makes me happy that your so happy to help us out mark as you could keep all this to yourself but dont you share the love. cheers buddy till next time.

[@uavtech]

Appreciated! Thanks much for the kind words.

[@michaelwilkes0]

my internet died right at that bomb clip. Thats one hell of a filter :) Had to go restart my router.

[@uavtech]

Nice!! 😂

[@О.Ш-ь1щ]

oh my freaking god! You just solved my two weeks headache!! There was no answer why have I had lot of noise on pitch axis from 20 to 150 hz! I was lowering down my filters down to 20 hz... D term just lowered to 0, but still had vibrations on pitch, and in the pid toolbox step resp tool there was a S-form line at 0.4-0.5 of setpoint... And there was no any examples which were looked like mine. Thanks to the universe that I decided to watch that video properly....

[@frankposterello1628]

Fantastic, thanks. It's remarkable how much of what you said I actually understood. And a few questions I had as you went along, you brought up later [dynamic filter was still active, just not shown on the filter tab, and how to run the spectrum analysis on a specific time segment]. It's a shame I had to wade through so many lame tuning videos that just regurgitate what they read somewhere else or just 'works for them' ('Just increase the P Luuuuke'). This is how it works, and why!

[@Gosuminer]

Very valuable information there. I especially like the scientific approach and all the good explanations. Many youtubers just give you a "magic recipe" that may or may not work for you.

[@mashtater1832]

Thanks for making this series. Looking forward to PID tuning. Im downloading the force !!!!!!!!!!!

[@nonplayercharacter9653]

Great job Mark. We really need this, especially with the complexity of 4.0.

[@2101Striper]

great vid. Loved the notch filter Nuke cut scene lol!! Thank u

[@beargrillz613]

hi, i am looking a lot of your vids trying to fix my quad. it wouble's some times when i hit full throttle. thx a lot for the time you take to make such informative vids..

[@uavtech]

Thanks! If you have full throttle shake, maybe up TPA breakpoint to 1650 and increase amount of TPA to 0.8. TPA in the latest few releases decreases D-term only which is probably what is causing your quad to.shake at 100% throttle because.it is amplifying noise.

[@akera9625]

Thanks for this great playlist, man! I just built my first 6s and i knew i was going to crash into a lot of problems... ...i never did black boxes but now i need it. ...It`s hard to get into it: watching your videos over and over again and again helps a lot! I download your trace setup and I`m trying to follow step by step what you are saying... ...hopefully i can address all the issues i have. Keep it up! Thanks again!!!

[@odyss160]

Distracted by that deer chasing the quad... Lmao

[@fpvclub7256]

NIce video.. I love this series.. Can't wait for the next one!.. That white-tail deer was awesome, those things can jump!

[@MouwDown]

10:16 was freaking epic. So great.

[@uavtech]

Still makes me laugh listing to it. Glad you like too.

[@BeyondGravityFPV]

10:17 got you an upvote, it was worth it

[@vasiliyklx450r]

Mark! You are doing great great job explainig this! Thumbsup!

[@FPVQuadModz]

Mark, this is the second time I've seen a white tail chasing a quad around, you get this often? Looking great so far. I'll be starting to tune the rest of my fleet this weekend so this was EXTREMELY helpful 😉

[@uavtech]

not a lot, but yes! She was chasing me and not the other way around. It was weird and unexpected. Thought maybe she had a baby hiding in the brush and was protecting it?

[@FPVQuadModz]

@@uavtech so when will you start looking at the additional filtering you mentioned? After pid tuning is complete or if the lack of additional filtering becomes a problem? Going to begin tuning the whole rest of my fleet this weekend, just replaced a lot of motor bearings. One other thing, do you do anything different when tuning a 6 inch quad? I know they are a bit of a different animal.

[@uavtech]

Will look at logs after P/D are ramped up to see how else we can make prop wash better. This 6s is seeming the same as 4s to me so far.

[@FPVQuadModz]

@@uavtech yes it's quite docile. The 6 inch with 1900kv F80's on the other hand... and if you want a challenge, I'm building the same quad I just sent you, with 2150kv f40's on 6s!

[@theronin48]

Really like your technical but ok to follow explanations. Would like to see a quick video on tuning nuances for brushless 3s/4s acro micros in the 2-3 inch range (not brushed indoor tiny whoops).

[@uavtech]

can do.

[@theronin48]

@@uavtech Awesome. Here is So Cal, people don't take well to drones. So micros has been an awesome alternative. I've been flying this T-Motor collab with Diatone the TM2419 HD. Tiny enough to look like a toy but packs a decent punch for acro. I've put a lot of work into it ... bit.ly/2LS9X2c

[@MCsCreations]

Fantastic job and tips, dude! Thanks a lot! 😃

[@balint133]

Nicely done, demystifying filtering and PID in general with examples what makes this series amazing. Thanks a lot for your work! Pun intended with placing the Mark point in BB viewer? :D

[@STEDDiEFPV]

Hey Mark, is the filter strategy you propose on these videos still valid on BF 4.2? From personal experience, I don't see much difference between having the gyro filter slider all the way to the right or turning off all gyro filters except the dynamic notch and RPM filters.

[@uavtech]

It still applies. But use either. There is about 0.5ms saved by turning off the gyro sliders. If you don't notice the difference, I would not bother though; many do. I need to double check, but moving them up from 1.0 to 2.0 save 0.5ms to 1ms.

[@mburrows8255]

Hi Mark. As always, thanks so much for lifting the veil on the mysteries of filtering, I always learn something new whenever you post. Can I ask what the filtering strategy should be for a racer who has to fly on bent props? Is your recommended setting still suitable for that scenario or should we be adding further filtering to protect motors from burning up?

[@Mateyhv1]

For racing you should't fly on bent props, that just doesn't make sense. You are filtering for the quad inherent noise when its in perfect conditions. Moreover adding filtres just to chase vibes will make your quad behave worse. As I said, change the props. Filtering is not for chasing bad props.

[@mburrows8255]

@@Mateyhv1 lol.😁 I'm afraid I crash during races... Often... I need to be able to fly with a bent prop to finish a race. And as far as I am aware, filters are there to enable that.

[@uavtech]

I would say the Dynamic Notch is what is protecting you in either case.

[@mburrows8255]

@@uavtech Great, will give it a whirl then. I guess I can do a log with some pretty banged up props on too... See what that looks like in the traces.

[@uavtech]

@@mburrows8255 , that would be a good comp. More importantly, motor temps.on banged up props. You jist need to finish the 2 min and get back, right? So do a whole race on the banges up props.

[@justinwalker8364]

funny, the deer out here like to chase my drone too lol .

[@heggy_fpv]

thank you! please do a 7inch example video too :)

[@rollofpv975]

BAM! Fantastic job, as usual. :)

[@getchucked1135]

What method is used in BF to calculate the first gyro derivative? If it is using a multi point method is there a way for us to alter and add more points in order to make it more immune to noise? I assume this would slightly increase delay but if BF is using a 2 point method adjusting to a 3 or 4 point method at 8kHz sample rate would only be adding microseconds of delay. If you are already doing this it shouldn't be too hard to add a editable field (points averaged) so people can quiet noisy D terms on noisy quads w/o filtering. Or am I just missing something?

[@uavtech]

The gyro gives data at 8k. At 8k pid loop, we use each same. At 4k pid loop, two samples are averaged. All the calcs run within 1 pid loop and a motor signal data point is sent.

[@alexk766]

Right on time, thank you.

[@edouardmalot51]

Love your videos, thanks a lot

[@marinehm]

This is a great explanation! Would you be willing to help tune my 3” Gecko with Emax 1606 4000KV motors? Or just tell me if they are too big?

[@wmnsriteslol]

Well done as usual. What's the Coles notes approach for 3 inchers? I get this approach and have followed it to great affect. Wondering how it differs

[@Mateyhv1]

Same as 5" you just have to set the filters for the vibe frequency of the 3" quad.

[@michaelfpvchiemgau8082]

How cool would it be if the betaflight configurator would show us the latency introduced by filtering. Could be an estimation based on your knowledge Mark.

[@uavtech]

Done. See theuavtech.com/filter-calc

[@benjaminschippers3536]

Amazing video again! How do you feel the rpm filtering works with or instead of the dynamic notch?

[@uavtech]

I will switch to that at the end of it all to see the difference. I feel the ESC firmware is still Beta so I don't want to push too much. I feel it is awesome if you can run without the dynamic notch or lowpass filters. Dyn Notch phase delay is ~0.8ms. RPM is less, but not if you have to add to the dynamic notch 0.8ms to get the same results. RPM does an awesome job at crushing the motor peak noise based on eRPM data, but it is blind to the actual peak vibration on the frame. So it would be nice for 4.1.0 if the dynamic notch could track actual vibration, but only activate if above a certain amplitude. I know some boyz were looking at that on the Slack. I think that was somewhat built-in to the dynamic notch code already as it use to park at 1,000hz - in the first round of the new re-boot code - if peaks were below a certain amplitude. Then it was changed to stay at the last detected peak unless a new peak wanted it to move to a new spot. I think that is the BEST way for it to work without RPM. But maybe with RPM the park idea is better. So maybe it could park if RPM is active but not park is RPM is not enabled. I need to talk to CTZ and Joe about that.

[@davidlhchong]

Great job Mark. I only fly 3inch quads and always use the default setting in BetaFlight and have never use Blackbox logs. With the release of BF 4.0.x, my 3inch quads fly horribly and have to go back to BF 3.5.7. I am trying to follow your PID Tuning guide to tune my 3" quads in BF4.0.x. I'm waiting for the Matek F722-Mini FC with 32MB onboard Flash to arrive as my current FC don't come with flash memory or SD-card for black box logs. While waiting for Matek FC, my next step is to setup BF Blackbox Explorer that is mentioned in your video. As a beginner to Blackbox explorer, may I know how do I go about setting it up? Do I need to setup "PID Toolbox" to follow your tuning guide? Thanks

[@uavtech]

Nope. Just.download BBE.

[@davidlhchong]

@@uavtech Ok. I just downloaded "2019.06.03-betaflight-blackbox-explorer" for Win32 from your Google Drive.

[@uavtech]

@@davidlhchong , right. Just put that somewhere and run EXE. Then load up JSON preset trace templates and you are set.

[@maggoffm7586]

Does turning off filters also apply for bf4.2 or would you recommend to just use the sliders and don't turn off some filters completely?

[@uavtech]

Either is fine. Has similar effect. That said, technically off is a little better, but more complicated than moving a slider.

[@PIDtoolbox]

I remember older versions of BF had static dterm notch on by default ~260Hz I think. Im curious, did you find less propwash doing this in your setup?

[@uavtech]

Eh. A little but it needs D-boost vs. default if D_min

[@PIDtoolbox]

@@uavtech besides the minor ps delay advantage using a dyn dterm lpf, can anyone say with confidence that there is a meaningful advantage over a static dterm lpf? I agree you might as well use it since it's there, but I wonder if a static biquad @100Hz might work just as well (or perhaps better since it's maintaining that cut from 100+). IDK...

[@uavtech]

@@PIDtoolbox , will give it a comp after PIDs are tuned. the list is growing on what to comp after tuned.

[@PIDtoolbox]

@@uavtech no worries. my guess is the differences would be too subtle to detect anyway. we r at the point of diminishing returns so detecting real differences will require more and more data, or a controlled propwash maker, whatever that is

[@ken-kamm]

Thanks again for doing this series. Does having a filter down at 60 hz create a lot of phase delay? I thought we generally tried to stay at 100 or above, but maybe that’s not the case these days.

[@uavtech]

It does, but it only goes down to 60hz at 0% throttle. At 175hz max, at cruise (40% throttle), it is at 100hz and at 60% throttle (prop wash zone) it is at 140hz.

[@ken-kamm]

Ahhhhh. Makes sense! Thanks very much.

[@STEDDiEFPV]

Mark, does the d term lowpass filter add any delay? More to the point, I don't get hot motors with with just one d term lowpass at 60hz or 140hz - I can see in the logs that the noise increases as I increase the cutoff frequency, but still not getting hot motors or that high pitch noise motors make when there's excess D or not enough d term filtering.

[@uavtech]

Yes, all filters add delay. D-term filtering just on the D-gain signal to the PID Sum. Gyro filtering delays all gains response. If more D-gain filtering is what works for you, go for it. But if you filter too much it delays the D-gain response and it starts to align with the P-gain and cause oscillation. Basically, the more you filter the D-gain, the less effective it becomes. But, it amplifies noise so much it is reasonable to have to filter a lot in some rigs.

[@STEDDiEFPV]

UAV Tech makes perfect sense when you explain it like this. After a lot of log analysis today, the best result I could achieve for my rig was all gyro lowpass filters off and only the first d term lowpass set to biquad at 90hz. I also had to add a d term notch filter at 160hz with a cutoff at 100hz as I had an unexplainable spike at 160hz on pitch and roll. I would have never dared to turn off all lowpasses on the gyro if I hadn’t come across your tuning guide video

[@STEDDiEFPV]

@@uavtech I've come across something interesting in Blackbox. In the last 2 days Ive been extensively testing BF 3.5.7 against BF 4.2 on a Radix, especially around noise attenuation, and after doing everything else in this thread (Only leaving Dynamic notch on 3.5.7 and Dynamic Notch + RPM filters on 4.2), the noise profile on all axis is very similar. I wonder if the delay added by the RPM filters isn't worth how much of a good job the dynamic notch is at tracking peak noise. Do you have any thoughts on this?

[@uavtech]

When turning on the RPM filter, are you adjusting the DN to raise Q? If so, the delay is the same. Which is good on a proper test!! Also absolutely key is it MUST be the same quad and same all other settings. On a fairly low noise quad, the results are likely to be similar. Where it gets interesting is a bent prop or a high noise quad. Or a quad that has a frame resonance below the motor peak noise.

[@STEDDiEFPV]

UAV Tech yes, I raised DN Q to 250, and the width to 0. This is the same quad, within a space of 15 minutes. Flashed one firmware then the other with matched settings. I had to turn off a bunch of stuff to match 3.5.7 but I wanted to test rpm + DN against DN notch only. I see what you mean by things getting interesting, and from my testing it looks like a wider DN should be able to handle frame resonance and bent props better than the same quad with rpm and higher Q and width of 0 for the DN without adding extra delay compared to using rpm and higher DN

[@MewoFPV]

So as far as I understand, the biggest performance impact from moving to 48khz should actually be more torque generated by the motor according to tests I've read. Wouldn't this likely imply that adjusting the pwm frequency just emphasizes or mitigates issues with the current PID tune you have on the rig? I'd think ideally 48khz should always be used and that if you have oscillations from it then it should be addressed via your PIDs

[@uavtech]

IDK. Ryan Harrel says 48kHz all the time. But some are now saying 24kHz is better for them, so I don't know wsup. Clearly 48kHz was better for this rig.

[@MewoFPV]

@@uavtech A buddy of mine did his master's thesis on esc commutation methods and covered some tests on pwm frequency you might find interesting. I'll post it in the discord! It reasonably convinced me that I should always be running 48khz haha Edit: Nevermind it's too big for discord...let me know if you want me to send it to you direct

[@uavtech]

@@MewoFPV , makes sense and you can see here, was better for this rig. Question is why are others better on 24kHz, or are they missing another change they made?

[@michaelwilkes0]

@@uavtech sometimes slower pwm frequency can be better for crappy slow fets. Fets mostly heat up not when they are on or off, but during transitioning, so if you go slower and let them stay on longer and let them stay off longer, they heat up less. But like I said, those are crummy (or slightly defective) fets.

[@jakeh3144]

Are these suggestions still relevant in BF 4.2 and with RPM filtering? Thanks!

[@uavtech]

Yes

[@jakeh3144]

@@uavtech wow, thanks for the quick response! I noticed in your slider tuning tutorial you ran different filter settings so I wasn't sure. Things with BF change so quick it's hard to keep up! Thanks for the all the helpful videos!

[@jonathanroberts4262]

I have same FC and also have same Pitch noise issue. Can the esc setting be changed if using the RPM filtering?

[@uavtech]

Yes.

[@blaisearnaud7857]

Cool video, but i expected you to talk a bit more about PID tuning, and whether you should increase or decrease it and why?

[@Mateyhv1]

Thats the next chapter

[@uavtech]

That is the next video now that filters are done.

[@tehllama42]

Mark is doing it right - until you have this sorted out, you're trying to find the 'right' scalar value on a PID controller for something that has unwanted frequency domain content... there isn't a truly right answer, whereas with all of this effort, stock PIDs are going to fly very well.

[@jamesparkerfpv435]

Thank you.

[@jackbisson9226]

In your last flight in this video, I noticed that the pitch was not smooth. It was bopping up and down. What did you to to clean that up? Did the change to 48khz fix it or did you tweak your PIDs? Thank you this great material.

[@uavtech]

Honestly, the gyro on that quad was screwed. It seems like was a 6S issue on the Lux F7 board. I sent the quad back to the owner and he swapped out to a CLracing F7 dual and now it is great. I continued on for this series with my 4S quad. His quad was having huge gyro spikes on pitch and roll that kept getting bigger and bigger.

[@ronniestepanek6150]

First! Just what I needed today! Thanks man!

[@IshamPadron]

Bam! You've made my day.

[@Freebird_67]

I’m having a nightmare of a time with a new build. It’s a 6” iflight frame with Omnibus F4 V6 4-1 BLHelis 35amp with 2306 1750kv motors I was on 4.X latest but it flew horrible with weird wobble and jitters. I’m running 3.5.7 now and if you watch my video on my channel you’ll see how bad it flies. It was way worse in 4.x. Though. I’m not the smartest man on the block and I’ve build plenty of large vehicles from 400mm to 800mm Ardupilot Autopilots but they have auto tune so that helped me greatly. Any suggestions?

[@Mateyhv1]

Look at the blackbox graphs, thats why you have that awesome tool. to chase and solve bad frequencies. Most of the problems are hardware related though. Flimsy arms, notchy motor bearings, bent props, etc.

[@uavtech]

What he said.

[@Freebird_67]

Mateyhv1 wish I had time and patience for that but don’t so I’m scraping quads for plane builds. Much less worries about tuning with planes they just fly no matter what. The frustration this has been last few weeks just isn’t worth it for someone who’s days are basically numbered. I really don’t have the time to sit in front of computer screen hours on end tuning a quad. I’m deal will serious health issues so that’s a no go. I just want to fly not sit on my ass

[@uavtech]

@@Freebird_67 , agreed. Stick and rudder are much easier. No FC, no auto-anything. I came from nitro planes but do enjoy quads. I don't have to be such a wind snob.

[@upplsuckimcool16]

That deer was so fuckin confused. That was sick. Like a deer in LED lights ;O

[@MrBirdshell666]

Could you please specify in which order the filters are being applied now? Including rpm notches.

[@uavtech]

It doesn't matter (like adding numbers), but it is: 1. RPM/Lowpasses/Static Notches 2. Dynamic Notch (peak detection and filtering) [PID Loop] 3. D-term filters

[@MrBirdshell666]

@@uavtech just wanted to know how rpm notches and lpfs affect peak detection of dynamic notch

[@uavtech]

@@MrBirdshell666 , they do (now). If they kill a peak, it will not be there for the notch to detect.

[@MrBirdshell666]

@@uavtech the reason i am aisking is these funny flyaways after arming with biquad on gyro. Without lpfs it flies fine. Seems like biquad on gyro breaks peak detection sometimes.

[@HomeDronen]

TFS The info!

[@umbratherios5614]

what can we do if we cannot use a blackbox? like on a tiny whoop which has no blackbox?

[@uavtech]

Tune by eye/ear: kzbin.info/www/bejne/npivhotrqryUhKs Also, presets -- theuavtech.com/presets

[@umbratherios5614]

@@uavtech That is a... way better response than I expected 0_0 Thanks!

[@umbratherios5614]

@@uavtech I watched your slider tuning video (rather liked it!) but it's for PID tuning, and I'm more worried about tuning my filters than my pids. I did check your presets too and applied them, anmy mobula flies a decent bit better! Motors get a tttiiinnnyyy bit warmer, but not much (maybe 2 to 4c warmer) I just wanna know how I can fine tune my filters without a blackbox now...

[@uavtech]

Just move up or down the slider. If moving up it flies better and motors are not hot, do it more. If not, keep at position or move down some.

[@Gespar]

Could this same filtering strategy used with some modification on falcox?

[@uavtech]

Falco does not have notch filters. It has 1st order - same as PT1 in BF called the Frequency Filter in Falco and other filters that basically have the same end result with different math - and 2nd order also called BiQUAD. Since 32k sampling, you need more lowpass filters; so that is good for 600hz plus noise that you get in 32k gyro sampling mode. But for 200hz to 600hz noise I wish they would add something like a dynamic notch. That said, if you cascade the lowpasses at low enough cutoffs it whacks down the peak noise. Not as effective as the dynamic notch or RPM for peak noise, but with a good mechanical/electrical setup it is enough to get the job done. For example, my RevoltOSD on BlackBird tune presets really flies well with very little prop wash. It is really nice. But i do see on their FB page that others are not as lucky? So, "not really" is the answer to your question. I would just try the BlackBird preset and if that doesn't fly well, see what you can work on in your mechanical. What you may need to do is flash BF so you can log the raw noise and see wsup. Then work it down and try flashing back to Falco. I don't know how effective the Anti-Aliasing filter is because Falco doesn't support logging (yet?). Try that too before the potential flash over to BF for logging to see if it helps. I have not turned it on for mine since BlackBird worked fine and it was off by default. Do keep in mind that in 32k sampling mode, the Gyro's internal DLPF is basically turned off, so it saves 0.98ms of Phase Delay right there. So adding some more lowpasses through the firmware is warranted. I think they just need to add a peak killing centric filter to the bundle to finish it out.

[@Gespar]

@@uavtech Thanks for detailed reply !! My falco flies really good with small tweak on balckbird tune. Basically i just lowered the d term cut from 200 to 170. Only what im issuing is super super small oscilllation on mid throttle what i would like to get rid off. Blackbird tune works great on daily flying but i just need a bit more filtering somwhere for my cinemawork were propwash and snappy response is not needed. Directors dont like jello 😅

[@uavtech]

@@Gespar , maybe cut D-gains with the TPA curve right at 40% throttle (or whatever is your cruising throttle). I would do that vs adding filtering.

[@Gespar]

@@uavtech good tip! I will try that one. Thanks! Keep up great videos m8 🤟

[@toaster5148]

how to set gyro scaled ?

[@uavtech]

at CLI type "get debug". That will show you debug commands. From there to set debug trace to record raw gyro, type "set debug_mode=gyro_scaled" [enter] then "save" [enter]. With that your next log will have debug traces included that record the raw gyro data.

[@marlin-fpv]

With RPM Filter enabled, the dynamic notch filter is kind of doing nothing anymore.

[@uavtech]

Whack-a-mole. But this approach is without RPM since ESC firmware is still Beta.

[@marlin-fpv]

UAV Tech Plus figured out that the dyn notch is usefull. Set to low cut 80 Hz it‘s very usefull! I hate noise and gave up on getting lowest phase delay for the cost of getting noise in the system that messes up the traces for PID tuning, plus also little noise I feel kinda introduce already visible vibrations.

[@christopherconkright1317]

This is old but it’s not beta flight upgrade firmware was betaflight

[@tehllama42]

In some regards, this makes me feel like an absolute hack the way I'm tuning my race quads - I wind up stopping with PT1 lowpass on D-Term still enabled, and a gyro lowpass present, with typically very low D-Term and reduced P term... end result is that I get higher amplitude propwash at lower frequencies, but I'm just used to it so I don't care, and it lets me get away with pretty dumb setups (6S 2250KV on 5149 props), I guess this will get me relying on the internal DLPF more , especially on my 3" rigs where I can start letting that Dterm DynLPF get all the way up to 225Hz. This kinda makes me want to slap together a TBS Source2 build with Emax ECO motors and cheap (Racerstar/RCharlance stack) and send it to you and show that a cheap racer tuned properly will smoke a high end build on defaults.

[@tehllama42]

Also, definitely feeling the vindication on recommending to people that they run 48kHz with startup/rampup power down at 0.125, and motor timing moved up.

[@Mateyhv1]

Lol, why is the deer chasing your kwad?

[@uavtech]

I guess baby hiding in brush? IDK.

[@Mateyhv1]

@@uavtech yep might be it. Where can I find the BBE with the waterfall graph, mine is official v3.3.1 and I can't see it anywhere. Thanks

[@upplsuckimcool16]

Hmmm JB ha a video with these two guys talking about PID tuning and it seemed everytime they suggested an answer to an issue they were spot on, and they adamantly asserted that PID tuning for prop wash is tuning for a symptom and not getting to the route of the issue. Getting to the route of the issue would fix the prop wash naturally. that PID tuning video so far is one of the very best I've seen on youtube so far. I have since found your videos and love them, but still the accuracy they had on answering errors then fixing them was so impressive I was seriously impressed xD I don't think JB likes not being the smartest guy in the room so he kinda... IDK.... shut down? He was kinda weird in that video being a little facetious... w.e.... idk..... Here is the video I'm talking about: kzbin.info/www/bejne/pYmVg3avi9OkbNU

[@uavtech]

Oh gosh. Tim Sweet. Thank gosh is gone now. They spend all afternoon trying to solve i-Term induced bounceback. They never fixed it and it is as simple as having i-Term relax in BF or using FF to have your quad track closer to Setpoint. Don't listen too close to the prop wash B.S. It is as simple as needing a good P/D ratio and strong gains.

[@upplsuckimcool16]

@@uavtech Come on man spending all after noon tuning is the reason I watch your channel xD If I can go out side do a flip a roll and a punch out and come back to geta perfect tune then I never have to worry about ruining my quad xD That's what I enjoy about this hobby. Sure I wanna get good at it Line of sight, but really that's onyl because I can't afford goggles lol.

[@uavtech]

Tim's arrogance is what rubbed me the wrong way. He would publically slander BF for "messing around with the PID controller too much", and then spent all afternoon tring to solve something that i-Term Relax would have prevented from the start. Plus the fact that i-Term Relax is just implimenting i-Term Clamping which is a well know PID controller fundamental. The other solution was adding FF or D-term kick though D-Setpoint weight to tack Setpoint better so i-term will not wind up, but there again, they removed those because "too much messing". Funny part is both are also well know PID control fundamentals too. He was an a-hole.

[@upplsuckimcool16]

@@uavtech I never met him so IDK, I'm also nto a computer programmer so I can't really comment on the validity of saying "too much messing". Seems like he was saying that you can fix the issue without these gimmicks. There ARE many ways that this process of thought could be born from arrogance though... eg. "get smart and learn what ur trying to do instead of relying on programmers to make it easier for you" or just jealousy that he didn't come up with it first. IDK, regardless of his arrogance the information in that video has been invaluable to me. My black box looked like a fuckin MESS and basically using what I learned from that video my black box looks sleek and tight and smooth and very predictable. However I have the one nagging issue of my roll axis. Everything in my black box lines up just about as perfect as possible with my motors being COLD even after hard piloting..... I just have to get this roll axis in line with everything else and I'll have an amazing tune. I'm starting to think it has somethign to do with the fact I'm using a stretched X frame on a 3'' Watching your video has helped as well with. I think I might just be low on D term now, however I don't even want my motors to be warm after a flight. I want to keep flying but if anythign else goes wrong I simply don't have the money for this hobby. I'm almost regretful I ever got into it. Tuning has given me the ability to enjoy the hobby without too much risk. I love fiddling with shit like this, I love OVERCLOCKING computers and doing stuff like that, and tuning is very similar. So I can still have fun with the hobby without ever risking crashing and breaking something xD.

[@uavtech]

@@upplsuckimcool16, they are not gimmicks. And they could not fix. That is the point. In either case I do admit, if i recall correctly, their advice leading up to that stumbling block was ok. And it if helped you, whelp that is all that mattered. Glad some of this is helping too. Don't worry about Fatshark goggles maybe and just get a $40 pair of box goggles. I used Eachine VRD2 goggles for 1.5 years when i got started. You can mount some diopter glasses inside to help with focus and I still have them. They work fine. Huge FOV.

[@daslolo]

Useless and dangerous. 4.0 NEEDS though filters to prevent fly off

[@uavtech]

Incorrect. All my quads fly this way and so do many others. You have issue Bro. Look at you raw noise sub-200hz and you will see.

[@daslolo]

Damaged props give you sub 200Hz noise. So ding your quad and your setup gives nice flyaways. Well done.

[@uavtech]

@@daslolo , also incorrect. Interesting how i have NEVER had a fly away on BF (nor F1 either for that matter.)