Interesting way to tune. I prefer to find the P&D balance first. Raise P until there is an overshoot then take it back down a bit. Once you’ve found that balance, raise P and D together until almost all prop wash is gone or until motors get hot. Tune I term then ramp up FF. Anyway, thank you for this series Chris.
I believe it is genuily easyer now because eif the way the values change... All we remember how hard it was to simplify when the sliders where reversed...
Yes, you can certainly tune PD balance first. The reason I tune master first is I find it easier to hear the bounceback and overshoot with higher PID gains because its a sharper sound.
I'd agree that this method is fundamentally better, but putting master in the ballpark is still something I recommend if you deviate from a standard 5" type setup, for all the reasons Chris has mentioned.
I was always hesitant to start pid tuning so I've been flying stock for a year. Your series has made it more approachable even though I'm still learning black box and the pid software you show. My quad feels great. I hope you have more on blackbox and such. Thank you so much
Chris - thank you so much for putting this series together! I've just spent just two hours doing my first tune, and now my quad is like a beast. It literally goes where I point it. You're a legend, and I genuinely cannot thank you enough!
First thank you @Chris Rosser for doing these videos. A week ago I tried to tune the D term lowpass 1 dynamic min cutoff frequency and went up to 145 and still couldn't tell the difference. Today I went to the field to tune the PIDs step-by-step by the instructions presented in this video. I raised my master multiplier slider up do the max (2.0) and I promise I didn't feel any difference. Didn't see any shootover in my goggles. Same with P&I gain. No noticable problem at 2.0. With the I-gains there were some overshoots at 2.0, so at the end I reduced all three sliders to 1.8. I really can't say I accomplished anything... I guess I'm not Pro enough to notice the difference. I mean, changing the parameters (sliders) via LUA is easy and quick, but I cannot do blackbox logging and analyzing in the fields... It's way too complicated. Since the steps in this video are so straight-forward so to speak, this process should really be somehow automated... it really makes no sense to do it all manually by the feeling. We use computer these days to do that. And drones happen to have one :) I know it takes a lot of programming but I hope I'll see that day somewhere in the near future, because I'm pretty sure it's totally doable.
Interesting. Last year I have been flying my rates in LiftOff for month or so and than when I went to tune a quad (tho I tune with blackbox), when I told myself "so this blackbox looks good, it should be tuned now" and I stopped doing tuning flights and stopped thinking about tuning, suddenly, it felt great and I got into the flow. My explanation is, that I found my favorite rates in LiftOff and when I tuned the quad it started following the rates properly so it flew as I like and muscle memory kicked in.
@@SukSukulent Well, I never said it doesn't feel tuned. It actually does. And that's the problem. No matter what I do, it still feels the same. And that's certainly not ok :) What I mean is that I must be doing something wrong...
Wow, thanks! Glad it was helpful. I think the BF 4.3 sliders make tuning even easier than before. I got the master slider I requested (in expert mode)😁
So well explained! That is the clearest explanation I have come across about PID's and how they work together. The visual explanation added by the slides is really nice. I think it helps with better memory retention of the information by adding the visual component. Thanks, that obviously took some real work to prepare and create.
Thank you, my iflight evoque f5x is a beast now with the vista and GoPro9 on top, I didn’t think it was capable of moving like this. TH-cam guys talked it up but this is miles above the stock tune on 4.2.11.
I followed this how guise without using the black box data but I got curious and checked because it flew so well I had to know how close I was to perfection and it was spot on I can’t believe it I must thank you again Chris. I defiantly got value out of this video but I’m a broke boy barely affording this hobby. But I will pick up one of your frames on my next build. Thanks again.
I have to admit man these 4.3 videos are much better than the 4.2 videos, I found these much easier for me to understand and help me tune my quads, even the ones tuning 4.2
Thanks a lot. PLEASE do some manual about - how can we track I gain (integral) with blackbox logs or PID toolbox. If we want to tune this parameter more accurately. Impossible to find any tutorial into whole TH-cam - except this video, but need more information about - how to handle with integral by blackbox logs. Awesome channel
Nice step by step explanation of tuning. Thank you Chris. I was thinking to re-tune my quad. Will try different approaches and this of course will be one my first to try out ! As always you rock 🤘
Chris, I've been flying the JB Xilo bundle (a F4 fc) with ELRS (R9 modules) for about 6 or 7 months on BF 4.2. I just did the BF 4.3 update (also applied EdgeTX and more recent ELRS fw). I applied the bf preset for ELRS 150Hz (I tweaked the values slightly since I'm at 200Hz) and I overlaid your Karate filter settings you recommended in the last video. My god, the thing is so stable in the air. I was able to hit gaps through the rungs of an extension ladder....quite a feat for this novice pilot. I haven't updated my spare Xilo yet, there is no comparison between the 2...the new FW & settings make it a completely different quad.
Hey, great Guide! I was just wondering, if Dynamic Damping - D Max is supposed to stay on 0 for after having tuned the other stuff, or if it is supposed to be set back to 1 , or tuned on its own... Many greetings from Munich
Chris: just a heads up, the Feed Forward Tuning section is different now. I was on a buddies computer and they have added Feedforward jitter reduction to the GUI. It sits just above where Feedforard Transition is. It isn't *super important* because you suggest leaving the Transition at 0 anyways, however a big change is that Feedforward jitter reduction does NOT have a default of zero, the default is 7. per the betaflight 4.3 tuning notes (which are unfinished/incomplete): >>> The default value of 7 works best for most quads with normal radio links, and is suitable for racing, ensuring that tight turns, and high-speed straights, are both smooth and responsive. Experienced race pilots with higher speed links will find that the straight-line sensitivity will be a bit less than normal, because for very small inputs, their feedforward 'push' is attenuated. They may prefer a value of 5. After a few flights we find that the quad 'stays on the intended line' a fair bit better than before, especially during tighter turns, which feel more 'on rails' than before. A value of 12-15 pretty much removes all feedforward from slow smooth inputs, and is what we suggest for freestyle/cinematic. The quad will be equally smooth in straight centred flight and in a sustained tight turns. However, if you flick the sticks, the response will be immediate. >>>
@@nellson.angelloFPV Feedforward had a lot of jitter in BF 4.2 and before. Some people complained about FF being weird and turned it off or dialed in a lot of FF transition. I'm sure it has to do with this fact. Now FF is working at its full (or near full) potential
Yeah I'm going to do a deep dive on FF now all the relevant PRs are merged and cover kitter reduction (which wasn't previously exposed in configurator)
@@TimeFadesMemoryLasts CTZ's code updates, particularly the FF Jitter and duplicate packet management.. is absolutely phenomenal. Honestly, that alone would have justified its own update... but there were lots of other simultaneous improvements that make it work even better.
Amazing tuning guide! One thing I would suggest to folks is try running VERY LOW 'Auto Factor' for RC Smoothing. Then you can tune your fingers and if you want smoothness add that in post production. Having lowest latency in the the FULL control loop from fingers to motors is a amazing feeling once your fingers stop jittering. =)
@@ChrisRosser Crossfire was good enough, Tracer is better. I'm sure by looking at the publicly available tests ELRS would be even cleaner signal, but it doesn't matter. What's more important is that connected feeling you get with the quad. Its not going to affect the tune to the point where you need to turn down the PIDS.(If your build is clean) Like I said, if you want stabilized footage, adding latency to the system is not the answer. Post production is. Also makes betaflight kind of feel like KISS???
As far as antenna mounting, most people would be very well served with a stubby antenna (or two) on a basic TPU mount. The first resonant mode of those setups is plenty high, the range achievable at 600mW is still well over a mile, and they actually can survive crashes quite well. Otherwise, non-stubby antenna affixing to a top plate extra heat shrink, adding VHB isolation, double zip ties, and taking up as much slack as possible with some light adjustment on the coax gets it done just time. The smaller antenna element options, they can go just about anywhere, physically isolating them with TPU mounts are pretty brilliant.
💯agree. A stubby on a good TPU mount is an OK solution and very durable. That said, I don't use SMA antennas anymore. I much prefer the all in one antenna that plugs straight into the VTX by mmcx or UFL. They are cheaper, lighter and so much more durable and easy to mount.
@@ChrisRosser For freestyle use, especially in 3.5" birds, I have to agree. From the racer mindset, serviceability is going to trump outright performance, so quickly changeable pigtails, and swappable antennas will tend to carry the day, hence that flexibility. The key aspect about them being cheaper is definitely not lost on me, although the fatigue-related damage I tend to experience with direct MMCX attachment is unusual, things will just arbitrarily wear out
First of all thank you for the tutorial. I'm following it step by step. One question though, settings, just as the "Dynamic Damping" that you say to set to 0 while tuning the PD master slide, should we set them back to their original value after tuning? Or do require tuning of its own? (sorry if the question seems dumb)
Watch out for the fine tuning video, I think he'll mention it in there. From my understanding and experience, you can leave it and fly without this feature. The point of D-max is, that while cruising, you don't need much damping, so you can decrease D and keep motors cooler and happier with less amplified noise. But, in sharp turns, dives and stuff, where propwash is encountered you need higher D to keep the quad under control. So, while cruising, the D is at the D value and boosts up to the D-max value for aggressive maneuvers. If I have understood it correctly. When thinking about it, if I would want this feature, I would set the D-max to the max D found in tuning, while lowering D a bit for normal flight. You could even setting the D-max a bit higher for better propwash performance, but than check motors for temperature and bad noises, just like in tuning D.
Hello, Chris. We really like your video, and we hope that we could share your video in our Facebook to help our clients to resolve the problems. Do you mind that?
Yep. I will be doing a practical PID tuning video to show what everything sounds like when you're tuning and when you go too far 👍 This will be for filters and PIDs
Really nice and clear guide. One problem I'm having it I don't really know what to listen for when tuning the master slider. I increased it to 1.7 and it sounds maybe a bit fluttery but definitely not like the motors are grinding. Motors are stone-cold but it's 3 degrees C outside so I don't know how much to rely on that.
That sounds like you are getting close to the limit. If you go a bit further you should confirm that it gets worse then you can back off until the motors sound smooth again.
@@ChrisRosser you're right. I increased it to 1.8 but then the quad wanted to take off immediately when I armed and the motors sounded a bit rough. I lowered it to 1.65 and now it flies well. I hope I don't smoke the motors when it is hot in the summer.
@@ChrisRosser on another quad I raised the master slider all the way to 2 and didn't notice any problems. I then moved the PI slider in steps, also all the way to 2 without any audible oscilations. Is it really possible I need to go this high?
@@franciswong26 when you have maxed out the master slider you can go further by moving the P&I and D sliders together. That keeps the same ratios between the PIDs. Only problem is you then have less room to move the P&I slider up to improve the tracking.
I followed every video and tutorials of yours When I start getting the perfect step response graph In Pid tool box After that when I tune the feed forward The whole step response gets messed up
I am at 1.8 master for D tuning, I think its better than 1.0 but scared just because the number is getting so high! But motors are cool and no funky sounds so I am gonna keep going, tomorrow I will test 1.9 and 2.0
Amazing tutorial. I have the geprc crown hd with 1408 motors at 2500kv. I want to tune the quad running a 1300 mAh 6s lipo and a full GoPro. Should I start tuning with the camera mounted or not? It flies well but the front motors are getting little warm to the touch not much but not that cool as the rear? Thank u.
Hi Chris! I'm following your great tutorial. When tuning D-gain, dynamic damping needs to be set 0 then move the master slider. So, after finding the optimal value, should dynamic damping stay at 0? Thanks!
There are a few of us already... Anyone figured it out yet? I guess I'll try to increase it until it gets worse. Same as with the other sliders hahaha :)
Amazing video, my tune is phenominal now! One thing I'm a little confused about that I didnt see touched here is do you use dynamic Dterm at all? and if so, would you adjust the sliders so Dmax would be where the tuned Dterm is from this guide, and how would you know how far down to push the Dmin?
I don't tend to use dynamic D term except when really fine tuning feedforward and stick tracking. If you want to see how I do it I have a mad scientist tuning video where I go through the steps.
I would however do not recommend dynamic idle on 13” if running Apd 120’s or 200’s as it conflicts with the inbuilt low end drive. Feature of the apd’s. They continue to drive down low there selves if you run dynamic craft will fall. Just thought id chuck that in.
Thanks for the video. I have no black box but I found a great tune. But when I try to land, the quad starts bouncing and I cannot get out of the bouncing other than just disarm. Is there anything else I can do that has minor impact on the flight performance but will stop the bouncing?
Hi Chris, i'm still very new to this hobby and find it very difficult to tune my quad. Your videos are very helpful and informative, just that i could not relate it exactly to my built as im using KISS gui. Will you be doing any KISS tuning videos soon? Thanks a bunch. Cheers...
I'm a little confused by your PID-F diagram. If the derivative term is acting purely on the gyro signal, you could have a quad that is perfectly following setpoint through some erratic motions, and still have additional motor signals that are uneeded.
I have a question here. Do we need to change the flight controller if I change my motors with a different model and size? I have a tuned flight controller for a 4 inch propeler but now want to go bigger so I replaced my motors but the question is for the new props (7 inches), do I need to retune it? In simple words how does the scaling affects the flight controller? Thank you
Hi Chris, first of all I want to congratulate you on your skill. I just didn't understand two things. The first is at 1.56 minutes of the video, after having set the 2 values of "PIDSUM LIMIT" to 1000, once all the tests are finished and the Tuning is finished, do we have to bring them back to the default value or are they left at 1000? The second thing I wanted to ask is: after having adjusted and found the correct value of the "Master Multipler" (for example 1.2) to proceed with adjusting the other sliders it is necessary to leave the Master Multipler at the value found (1.2) or for each adjustment should it reset the previous sliders to the value of 1? A thousand thanks
Does this still apply ?. My 5in quad has pitch oscillations at low throttle ( I figured this out through the sound of the quad and checking logs) and I’m trying to diagnose how to tune it out
Thanks Chris, I used your steps to tune my problematic quad by my senses and it flies great now. No significant propwash and the quad feels very locked. In the blackbox I see the gyro response abit noisy. But in the video the quad flies very smoothly and follows my stick commands very well. Should I continue to try adding filtering to reduce the gyro noise?
I can put my master slider all the way up and didn't notice a difference. Every time I try to tune. I can never feel much difference in PIDS and dont really feel like i ended up up much better than stock pids. using a 3.5 inch whoop this time
Chris, I find that default P-D ratio in Betaflight is great for 5" quads but not great for anything else. Would it be fair to say that your method makes sense for most 5" quads but for smaller/larger rigs it might be better to find the right P/D balance first? Thanks!
Great guide but please remind people what graphs you are looking at when you are looking at results/examples of blackbox logs (gyro filtered pitch etc).
Chris... I have a PAVO30 (HD) and tried to follow your guide... Even setting the pid limits to 1000 like you described I´m already maxed out on the MASTER slide and could not fell any noticeable improvement OR bad behavior... engines are still cool. What I´m trying to improve is a horrific wobble when I move the throttle to zero or going out of flips. What am I doing wrong ?? or this wont apply to small quads like this ? Thank you ...
I found it really hard to tune my quad by listening to the motors because it's a 3.5" with a 99g AUW and it's basically inaudible once it gets about 5-10m away. And it NEEDS a good tune because default Betaflight values don't work at all for it.
Hi Chris! Awesome guide, thank you so much! I have a question:ive been following your PID tuning process step by step, and i found that for many parameters(P:D, I term,…) i can raise the slide as much as 2 without noticing any adverse effect, so it is difficult to me knowing where is the performance point. Btw, i am applying tuning to a 65mm whoop. Is this normal for these small quads?
whenever i disable the feed foward, the quad becomes unflyable, it just wobbles almost uncrontrollably and then i crash land, it feels ok when armed and i pitch and roll a little bit but as soon as i take off it wobbles out of the sky, i feel like im being punked lol
I'm trying to find info on why I'm getting an arming osd warning flag that said "rc smoothing". I can't find any info on it. Does this mean the rc smoothing is not working? Too much? another issue? If I arm and get the error, I just have to disarm and arm again to fly. ADC on controller on or off does not change. Tried smoothing at 30, 50 and 100 no change to the osd warning.
Hi Chris, I recently came across the motor mixer weights. They are a bit hidden in betaflight but they can be used to correct for the different moment arms of the motors. On deadcat frames, the front motor are often further apart than the rear motors so the moment arms on the roll axis are unequal. Do you think it is worthwhile to change the weight? Iforce2d has a calculator to work them out.
On my Larva X HD BF 4.3 generates flyaway with default filters and what you recommended so far. It basically oscillates around 100 Hz when hovering and... goes up up up. That's very weird as my Mobula6 has no such issue. Given I've no blackblox on either it's a bit hard to dig into details. Increasing the master slider increases the oscillation frequency and but doesn't seem to fix it. No filter seem to fix it either. Reducing the Tracking (P&I Gains) down to 0.4 (instead of 1.0) seem to work (it's not oscillating anymore and seems not to flyaway, so far). However, should I now continue with what you're suggesting like increase the master multiplier etc., or is that nonsense I'm getting here due to something else?
When finding your max D gains using the master multiplier slider, shouldn't feed forward be turned off as well? Or, does it not really matter when initially finding max D gains, but only when dealing with the P:I ratio is when you'd need to turn feed forward off? Thanks!
hi Chriss, thanks for this tutorial, it really helps..there is ashadow point though.. after finding the correct Dgain what should we do with the dynamic damping slide? leave it at zero or increase? this is not shown in the explanation. thanks.
Thanks! But when D tuning first, do we leave sliders to positions we think are good? While whole time dynamic d slider is 0? And rest for other sliders?
Are tuning flights any different for cinewhoops? I just built a shendrones squirt and based on how unpredictable it is on slow casual flights right now, I could never imagine doing flips or rolls with it. I've been struggling to get it to a place where I have no yaw washout while eliminating fly away potential. I can't seem to find that happy medium and I keep flip flopping between having airmode on or off.
Hi Cris, wonderful job as usual. Thanks. What if is not a 5inch quad to be tuned, how to get the right P:D balance which is not default anymore for Betaflight... the master multiplier works great for 5 inch as you said... but for all others? have you any tips to where to start?
Pretty sure there is gonna be presets for all classes and you should be able to use that pd ratio and just slide the multipliers up till you get wierd shit to happen and then turn it down a notch.
Don't worry you can follow the same process for any quad. DO the master slider first, then increase the P&I gain slider until you get bounceback/oscillation on snap moves then back off a little. PID toolbox step response makes it very easy.
Amazing as always! One question tho: I applied the Frsky preset and watched UAVtech's video about RC links and he (and the presets) suggest the opposite setting of averaging: OFF for Frsky, higher for the different high-performance links (which I still don't have yet lol), so what's the matter?
@@ChrisRosser Thanks! Also, interestingly enough, I don't like anything over 35 auto smoothing, the quad feels really floaty and not locked in enough for me. And I think I have it tuned reasonably well - now that I'm thinking about that, I have D at 69 and 79 and everything is fine, while flying beat up mismatched pitch props lmao, only RPM and notch filter on gyro, no notch on D and D slider at 1.1.....and long boy TBS antenna, in a metal mount xD
Chris, do you tune in a bit of wind or wait for a perfect day? Does it make a difference if you tune in a light wind vs a quiet windless day? Thank you for all your videos!
I tune in all weathers. A bit of wind doesn't hurt when trying to tune a quad as it gives a bit of outside influence for the PID controller to fight against.
i never get over dampened. my quads seem to runaway from d gain way before that.. Now that its extremely cold.. my 3" oscillating sound like it wants to run up just a hair at certain point on throttle above idle.. but didnt run away. did full throttle burst was fine.. . DIdnt know cold would effect it.
Whenever I try to set my feedfoward averaging to the value 2 for crossfire at locked 150Hz, every time I try to save the change it turns averaging back off. Any idea why it's doing this?
what can i do after filtertuning on my 7 inch rig, while i have master slider is on 2.0 and everything is good. how can i find my perfect D? i need to rais D fist also till motors sounds noisy at arming. filter tuning is on default pids. i can not imaging the my d filter starts at 120hz minimum. motor is starting to get noisy at this minimum. i think the tutorial is not so good for 7 inch quads? 🤔
Been watching a few of your tuning vids today, great video! but this seems to not be for me, i`d rather just keep flying bad flying quads than learn all this
I know you don't specifically mention this in this video, but what software do you use to do finite element resonance analysis? I'm very curious. Thanks for all the great info, keep it up!
hello friend ....i just build 5" frame HSKRC 225MM motor 2306.5 2550kv/ esc BL Heli 30A / Flight controller SpeedyBee F7 AIO / gps .....just finish the the setup beta flight select auto 5 " ...still yest to test fly .....is my basic selection of parts correct ???....what will be the setting PID ??......Thank you
Nice video. Where is the part about D- min. You explain to turn it off when trying to find max D gain. But you never mention when to put it on and how to tune it properly ? - Would u still on 4.3 recommend to keep D-min off?
I would not use D-min. The difficulty with D-min is it takes time to ramp up which means its not good at tackling sudden propwash. Much better to just run the highest constant D-gain you can.
what are people doing to re calibrate their FC if they start fresh with these new versions of betaflight? how can you properly calibrate it when it's in the quad? I got a small level but it doesn't work very well.
I just hold the quad down flat on the desk to calibrate the accelerometer. If you don't use angle mode then I would just turn off the accelerometer to save CPU, black log space and avoid the need to cal 😁
I followed this tune completely on a cine whoop felt great got all the way to the end and noticed some wobbles when kind of just hovering very strange but even if I restart I still get the wobbles a little bit it’s not bad very minor just wondering if you had any insight
could it be adding feedforward ive changed everything back to normal except that and i felt like the wobbles only started then but at zero is fine im going to test this here when i can ill let you know and thanks for everything you do in the hobby
@@ChrisRosser sorry for commenting so much, what number do you think i should set smoothness to and do you think i should increase jitter to what betaflight suggest?
Yes, absolutely! RPM filtering is essential for best performance on 4.3. RPM filtering is more effective than dynamic notch for motor noise and has less delay. With RPM filtering a single dynamic notch is usually enough. Dynamic Idle is also essential for best performance on 4.3 as well so bi directional d-shot is a must have. 👍👍👍
@@ChrisRosser thx, I found that on my armattan badger 5" using ethix p4 props .... Dynamic idle of 28 makes the quad super-responsive coming out of sharp turns.... This is definitely not a gimmick... Took me awhile to find the correct value ... I appreciate your time my friend
Hi Stepan, in BF 4.3 D term is the derivative of Gyro not PID error and feedforward is the derivative from setpoint. This has been the behaviour for a long time. I made this mistake in my BF 4.2 video and was corrected by the devs!
Increasing pidsum limit allows the PID loop to push harder which can improve the tune for quads that don't have a lot of control authority. If you increase the limit you never need to turn it down.
@@ChrisRosser I upgraded my nazgul 5 V2 6S today.. Haven't tried it yet. What dynamic idle would you recommend for ethix s3 on 5in nazgul 5? I set to 20 based on the other video.. But maybe I need to go up to 25-30? I'm so overwhelmed by PID/filter tuning I'll have to do...I'm a perfectionist, so I can't just leave it default.
![เปิด อาณาจักร หมื่นล้าน "สาระตั้ม" l [Nickynachat]](http://i.ytimg.com/vi/3XE0nPF4YkQ/mqdefault.jpg)
Interesting way to tune. I prefer to find the P&D balance first. Raise P until there is an overshoot then take it back down a bit. Once you’ve found that balance, raise P and D together until almost all prop wash is gone or until motors get hot. Tune I term then ramp up FF. Anyway, thank you for this series Chris.
I believe it is genuily easyer now because eif the way the values change... All we remember how hard it was to simplify when the sliders where reversed...
Yes, you can certainly tune PD balance first. The reason I tune master first is I find it easier to hear the bounceback and overshoot with higher PID gains because its a sharper sound.
I'd agree that this method is fundamentally better, but putting master in the ballpark is still something I recommend if you deviate from a standard 5" type setup, for all the reasons Chris has mentioned.
I was always hesitant to start pid tuning so I've been flying stock for a year. Your series has made it more approachable even though I'm still learning black box and the pid software you show. My quad feels great. I hope you have more on blackbox and such. Thank you so much
Mark Spatz has some really good info for you also. UAV Tech is his channel.
Chris - thank you so much for putting this series together!
I've just spent just two hours doing my first tune, and now my quad is like a beast. It literally goes where I point it.
You're a legend, and I genuinely cannot thank you enough!
First thank you @Chris Rosser for doing these videos. A week ago I tried to tune the D term lowpass 1 dynamic min cutoff frequency and went up to 145 and still couldn't tell the difference. Today I went to the field to tune the PIDs step-by-step by the instructions presented in this video. I raised my master multiplier slider up do the max (2.0) and I promise I didn't feel any difference. Didn't see any shootover in my goggles. Same with P&I gain. No noticable problem at 2.0. With the I-gains there were some overshoots at 2.0, so at the end I reduced all three sliders to 1.8. I really can't say I accomplished anything... I guess I'm not Pro enough to notice the difference. I mean, changing the parameters (sliders) via LUA is easy and quick, but I cannot do blackbox logging and analyzing in the fields... It's way too complicated. Since the steps in this video are so straight-forward so to speak, this process should really be somehow automated... it really makes no sense to do it all manually by the feeling. We use computer these days to do that. And drones happen to have one :) I know it takes a lot of programming but I hope I'll see that day somewhere in the near future, because I'm pretty sure it's totally doable.
Interesting. Last year I have been flying my rates in LiftOff for month or so and than when I went to tune a quad (tho I tune with blackbox), when I told myself "so this blackbox looks good, it should be tuned now" and I stopped doing tuning flights and stopped thinking about tuning, suddenly, it felt great and I got into the flow.
My explanation is, that I found my favorite rates in LiftOff and when I tuned the quad it started following the rates properly so it flew as I like and muscle memory kicked in.
@@SukSukulent Well, I never said it doesn't feel tuned. It actually does. And that's the problem. No matter what I do, it still feels the same. And that's certainly not ok :) What I mean is that I must be doing something wrong...
Amazingly clear and concise! The best explanation of how the PIDs work that I've ever heard. You're a great teacher Mr. Rosser.
Wow, thanks! Glad it was helpful. I think the BF 4.3 sliders make tuning even easier than before. I got the master slider I requested (in expert mode)😁
So well explained! That is the clearest explanation I have come across about PID's and how they work together. The visual explanation added by the slides is really nice. I think it helps with better memory retention of the information by adding the visual component. Thanks, that obviously took some real work to prepare and create.
Thanks Conrad! I think the BF4.3 sliders make tuning PIDs a little easier. (especially in expert mode).
Thank you, my iflight evoque f5x is a beast now with the vista and GoPro9 on top, I didn’t think it was capable of moving like this. TH-cam guys talked it up but this is miles above the stock tune on 4.2.11.
Thanks Chris, I can see myself watching this series of vids again.
Glad you enjoyed it, hopefully its a useful reference for future tunes!
I followed this how guise without using the black box data but I got curious and checked because it flew so well I had to know how close I was to perfection and it was spot on I can’t believe it I must thank you again Chris. I defiantly got value out of this video but I’m a broke boy barely affording this hobby. But I will pick up one of your frames on my next build. Thanks again.
Thank you so much! I’ll tune my first quad and now it absolutely clear for me)
Not sure, but in general😅
You are the first guy that made PID tuning make sense for me, thankful !
Well done Chris, Thanks!
Thanks! Appreciate your support.
I have to admit man these 4.3 videos are much better than the 4.2 videos, I found these much easier for me to understand and help me tune my quads, even the ones tuning 4.2
Great work! You got my 3 inch build working great with BF 4.2.8 and now I will tune my Micro Apex 4 inch with this in BF 4.3 :D
Sounds great! Thanks for field testing the method 🤓
Thank you so much Chris! This is such a great guide. Amazing work mate.
My pleasure!
@@ChrisRosser purchased myself an aos 5.5 to support you. I'm really excited for it.
Thanks a lot. PLEASE do some manual about - how can we track I gain (integral) with blackbox logs or PID toolbox. If we want to tune this parameter more accurately. Impossible to find any tutorial into whole TH-cam - except this video, but need more information about - how to handle with integral by blackbox logs. Awesome channel
Finally! Thanks Chris, been looking forward to this video
Hope you enjoyed it!
Hello Chris, i was waiting for it, many thanks ! Wish you all great Christmas event. Greetings from France !
Thanks! I hope it lives up to your expectations 😁
@@ChrisRosser i have absolutely no doubt about that 😉👍
Nice step by step explanation of tuning. Thank you Chris. I was thinking to re-tune my quad. Will try different approaches and this of course will be one my first to try out ! As always you rock 🤘
Thanks, I hope you find the rest of the series helpful too. It goes all the way from zero to hero on BF 4.3 😜
This is like an FPV bible I keep returning to for proper guidance in tuning times of need... 😂😂 Thanks so much!
again AMAIZING presentation Mater Chris :)
Chris, I've been flying the JB Xilo bundle (a F4 fc) with ELRS (R9 modules) for about 6 or 7 months on BF 4.2. I just did the BF 4.3 update (also applied EdgeTX and more recent ELRS fw). I applied the bf preset for ELRS 150Hz (I tweaked the values slightly since I'm at 200Hz) and I overlaid your Karate filter settings you recommended in the last video. My god, the thing is so stable in the air. I was able to hit gaps through the rungs of an extension ladder....quite a feat for this novice pilot. I haven't updated my spare Xilo yet, there is no comparison between the 2...the new FW & settings make it a completely different quad.
Hey, great Guide!
I was just wondering, if Dynamic Damping - D Max is supposed to stay on 0 for after having tuned the other stuff, or if it is supposed to be set back to 1 , or tuned on its own...
Many greetings from Munich
ever figure this out
Chris: just a heads up, the Feed Forward Tuning section is different now. I was on a buddies computer and they have added Feedforward jitter reduction to the GUI. It sits just above where Feedforard Transition is. It isn't *super important* because you suggest leaving the Transition at 0 anyways, however a big change is that Feedforward jitter reduction does NOT have a default of zero, the default is 7. per the betaflight 4.3 tuning notes (which are unfinished/incomplete):
>>>
The default value of 7 works best for most quads with normal radio links, and is suitable for racing, ensuring that tight turns, and high-speed straights, are both smooth and responsive.
Experienced race pilots with higher speed links will find that the straight-line sensitivity will be a bit less than normal, because for very small inputs, their feedforward 'push' is attenuated. They may prefer a value of 5. After a few flights we find that the quad 'stays on the intended line' a fair bit better than before, especially during tighter turns, which feel more 'on rails' than before.
A value of 12-15 pretty much removes all feedforward from slow smooth inputs, and is what we suggest for freestyle/cinematic. The quad will be equally smooth in straight centred flight and in a sustained tight turns. However, if you flick the sticks, the response will be immediate.
>>>
I found this to be interesting to, it turns to be easyer to tune now comparing to the way it behaved before.
@@nellson.angelloFPV Feedforward had a lot of jitter in BF 4.2 and before. Some people complained about FF being weird and turned it off or dialed in a lot of FF transition. I'm sure it has to do with this fact. Now FF is working at its full (or near full) potential
Yeah I'm going to do a deep dive on FF now all the relevant PRs are merged and cover kitter reduction (which wasn't previously exposed in configurator)
@@TimeFadesMemoryLasts CTZ's code updates, particularly the FF Jitter and duplicate packet management.. is absolutely phenomenal. Honestly, that alone would have justified its own update... but there were lots of other simultaneous improvements that make it work even better.
@@ChrisRosser Can't wait! :)
Thanks Chris great video 📹 👍🏼 🥂
Thanks Tony!
Can't wait for the FF tuning guide!
Thank you!
Waiting lol
Amazing guide! Please, review Emuflight for us! Both Kalman and ABG filters are very good.
I will be looking at Emuflight as part of the 1.0 release 👍🤓
Amazing tuning guide!
One thing I would suggest to folks is try running VERY LOW 'Auto Factor' for RC Smoothing. Then you can tune your fingers and if you want smoothness add that in post production.
Having lowest latency in the the FULL control loop from fingers to motors is a amazing feeling once your fingers stop jittering. =)
Yes, you need a high quality radio link for that though. I guess you're running 500Hz ELRS?
@@ChrisRosser Crossfire was good enough, Tracer is better. I'm sure by looking at the publicly available tests ELRS would be even cleaner signal, but it doesn't matter.
What's more important is that connected feeling you get with the quad. Its not going to affect the tune to the point where you need to turn down the PIDS.(If your build is clean)
Like I said, if you want stabilized footage, adding latency to the system is not the answer. Post production is.
Also makes betaflight kind of feel like KISS???
As far as antenna mounting, most people would be very well served with a stubby antenna (or two) on a basic TPU mount. The first resonant mode of those setups is plenty high, the range achievable at 600mW is still well over a mile, and they actually can survive crashes quite well. Otherwise, non-stubby antenna affixing to a top plate extra heat shrink, adding VHB isolation, double zip ties, and taking up as much slack as possible with some light adjustment on the coax gets it done just time.
The smaller antenna element options, they can go just about anywhere, physically isolating them with TPU mounts are pretty brilliant.
💯agree. A stubby on a good TPU mount is an OK solution and very durable. That said, I don't use SMA antennas anymore. I much prefer the all in one antenna that plugs straight into the VTX by mmcx or UFL. They are cheaper, lighter and so much more durable and easy to mount.
@@ChrisRosser For freestyle use, especially in 3.5" birds, I have to agree. From the racer mindset, serviceability is going to trump outright performance, so quickly changeable pigtails, and swappable antennas will tend to carry the day, hence that flexibility. The key aspect about them being cheaper is definitely not lost on me, although the fatigue-related damage I tend to experience with direct MMCX attachment is unusual, things will just arbitrarily wear out
This literally came just in time! Can't wait to tune my Glide!
If you have a tune for 4.2 that you like you can transfer PIDs and start tuning from there. Might save you some time.
First of all thank you for the tutorial. I'm following it step by step. One question though, settings, just as the "Dynamic Damping" that you say to set to 0 while tuning the PD master slide, should we set them back to their original value after tuning? Or do require tuning of its own? (sorry if the question seems dumb)
Was wondering the same thing
Watch out for the fine tuning video, I think he'll mention it in there.
From my understanding and experience, you can leave it and fly without this feature. The point of D-max is, that while cruising, you don't need much damping, so you can decrease D and keep motors cooler and happier with less amplified noise. But, in sharp turns, dives and stuff, where propwash is encountered you need higher D to keep the quad under control. So, while cruising, the D is at the D value and boosts up to the D-max value for aggressive maneuvers. If I have understood it correctly.
When thinking about it, if I would want this feature, I would set the D-max to the max D found in tuning, while lowering D a bit for normal flight. You could even setting the D-max a bit higher for better propwash performance, but than check motors for temperature and bad noises, just like in tuning D.
He said in a comment to leave it usually 0, in the newer tuning like mad scientist video he goes through it he said, was confused too 😁
Hello, Chris. We really like your video, and we hope that we could share your video in our Facebook to help our clients to resolve the problems. Do you mind that?
Great video!
I love bf 4.3... ..is really easy to tune with this sliders.
...would be nice if you can record the real sounds you are mentioning too :)
Yep. I will be doing a practical PID tuning video to show what everything sounds like when you're tuning and when you go too far 👍 This will be for filters and PIDs
@@ChrisRosser the world is a better place with you!!!
@@akera9625 You too! Happy holidays 😁
Really nice and clear guide. One problem I'm having it I don't really know what to listen for when tuning the master slider. I increased it to 1.7 and it sounds maybe a bit fluttery but definitely not like the motors are grinding. Motors are stone-cold but it's 3 degrees C outside so I don't know how much to rely on that.
That sounds like you are getting close to the limit. If you go a bit further you should confirm that it gets worse then you can back off until the motors sound smooth again.
@@ChrisRosser you're right. I increased it to 1.8 but then the quad wanted to take off immediately when I armed and the motors sounded a bit rough. I lowered it to 1.65 and now it flies well. I hope I don't smoke the motors when it is hot in the summer.
@@ChrisRosser on another quad I raised the master slider all the way to 2 and didn't notice any problems. I then moved the PI slider in steps, also all the way to 2 without any audible oscilations. Is it really possible I need to go this high?
@@avsaase I found the same case as yours...
@@franciswong26 when you have maxed out the master slider you can go further by moving the P&I and D sliders together. That keeps the same ratios between the PIDs. Only problem is you then have less room to move the P&I slider up to improve the tracking.
I will need this video if in case I don't become lucky owning Kiss Ultra FCFC this month. I'll switch back to Betaflight if it happens. :)
Hi Chris, please do a video on pid tuning in betaflight 4.4
I followed every video and tutorials of yours
When I start getting the perfect step response graph In Pid tool box
After that when I tune the feed forward
The whole step response gets messed up
I am at 1.8 master for D tuning, I think its better than 1.0 but scared just because the number is getting so high! But motors are cool and no funky sounds so I am gonna keep going, tomorrow I will test 1.9 and 2.0
And what about the Dmax? You just leave it turned off or with the default values?
Same question :/.
Amazing tutorial. I have the geprc crown hd with 1408 motors at 2500kv. I want to tune the quad running a 1300 mAh 6s lipo and a full GoPro. Should I start tuning with the camera mounted or not? It flies well but the front motors are getting little warm to the touch not much but not that cool as the rear? Thank u.
Hi Chris! I'm following your great tutorial. When tuning D-gain, dynamic damping needs to be set 0 then move the master slider. So, after finding the optimal value, should dynamic damping stay at 0? Thanks!
i'm wondering the same.. @Chriss Ross
Same question here. Many thanks for the guide!
There are a few of us already... Anyone figured it out yet? I guess I'll try to increase it until it gets worse. Same as with the other sliders hahaha :)
@@alexdefelice2448 Was looking for the same but found the answer in a comment further down. Chris suggested to leave it at 0
Great,thanks :)
No worries!
Amazing video, my tune is phenominal now! One thing I'm a little confused about that I didnt see touched here is do you use dynamic Dterm at all? and if so, would you adjust the sliders so Dmax would be where the tuned Dterm is from this guide, and how would you know how far down to push the Dmin?
I don't tend to use dynamic D term except when really fine tuning feedforward and stick tracking. If you want to see how I do it I have a mad scientist tuning video where I go through the steps.
@@ChrisRosser I will do, thank you!
I would however do not recommend dynamic idle on 13” if running Apd 120’s or 200’s as it conflicts with the inbuilt low end drive.
Feature of the apd’s. They continue to drive down low there selves if you run dynamic craft will fall. Just thought id chuck that in.
Thanks for the video. I have no black box but I found a great tune. But when I try to land, the quad starts bouncing and I cannot get out of the bouncing other than just disarm. Is there anything else I can do that has minor impact on the flight performance but will stop the bouncing?
Hi Chris, i'm still very new to this hobby and find it very difficult to tune my quad. Your videos are very helpful and informative, just that i could not relate it exactly to my built as im using KISS gui. Will you be doing any KISS tuning videos soon? Thanks a bunch. Cheers...
I'm a little confused by your PID-F diagram. If the derivative term is acting purely on the gyro signal, you could have a quad that is perfectly following setpoint through some erratic motions, and still have additional motor signals that are uneeded.
I have a question here. Do we need to change the flight controller if I change my motors with a different model and size? I have a tuned flight controller for a 4 inch propeler but now want to go bigger so I replaced my motors but the question is for the new props (7 inches), do I need to retune it? In simple words how does the scaling affects the flight controller? Thank you
Hi Chris, first of all I want to congratulate you on your skill. I just didn't understand two things. The first is at 1.56 minutes of the video, after having set the 2 values of "PIDSUM LIMIT" to 1000, once all the tests are finished and the Tuning is finished, do we have to bring them back to the default value or are they left at 1000?
The second thing I wanted to ask is: after having adjusted and found the correct value of the "Master Multipler" (for example 1.2) to proceed with adjusting the other sliders it is necessary to leave the Master Multipler at the value found (1.2) or for each adjustment should it reset the previous sliders to the value of 1?
A thousand thanks
Does this still apply ?. My 5in quad has pitch oscillations at low throttle ( I figured this out through the sound of the quad and checking logs) and I’m trying to diagnose how to tune it out
First! Hello Chris 🤩🙌🙌
Nice one! You are first 🥇
Спасибо 💣💥
Thanks Chris, I used your steps to tune my problematic quad by my senses and it flies great now. No significant propwash and the quad feels very locked. In the blackbox I see the gyro response abit noisy. But in the video the quad flies very smoothly and follows my stick commands very well. Should I continue to try adding filtering to reduce the gyro noise?
First of all, great guide. I'm currently on step 5 but every time I throttle up to take off, my quad flips. Please advise why. Thank you!
I can put my master slider all the way up and didn't notice a difference. Every time I try to tune. I can never feel much difference in PIDS and dont really feel like i ended up up much better than stock pids. using a 3.5 inch whoop this time
Chris, I find that default P-D ratio in Betaflight is great for 5" quads but not great for anything else. Would it be fair to say that your method makes sense for most 5" quads but for smaller/larger rigs it might be better to find the right P/D balance first? Thanks!
Do I turn dynamic damping back to 1 once I'm finished with everything?
not sure why but i cant turn on RC_smoothing. it always goes to 50(auto factor) and the smoothing mode is empty. Any one a idea why?
Great guide but please remind people what graphs you are looking at when you are looking at results/examples of blackbox logs (gyro filtered pitch etc).
Chris... I have a PAVO30 (HD) and tried to follow your guide... Even setting the pid limits to 1000 like you described I´m already maxed out on the MASTER slide and could not fell any noticeable improvement OR bad behavior... engines are still cool. What I´m trying to improve is a horrific wobble when I move the throttle to zero or going out of flips. What am I doing wrong ?? or this wont apply to small quads like this ? Thank you ...
Interesting idea! If you take a before and after blackbox log you should be able to see if it makes any difference. Just set logging to gyro_scaled.
I found it really hard to tune my quad by listening to the motors because it's a 3.5" with a 99g AUW and it's basically inaudible once it gets about 5-10m away. And it NEEDS a good tune because default Betaflight values don't work at all for it.
Hi Chris! Awesome guide, thank you so much! I have a question:ive been following your PID tuning process step by step, and i found that for many parameters(P:D, I term,…) i can raise the slide as much as 2 without noticing any adverse effect, so it is difficult to me knowing where is the performance point. Btw, i am applying tuning to a 65mm whoop. Is this normal for these small quads?
whenever i disable the feed foward, the quad becomes unflyable, it just wobbles almost uncrontrollably and then i crash land, it feels ok when armed and i pitch and roll a little bit but as soon as i take off it wobbles out of the sky, i feel like im being punked lol
I'm trying to find info on why I'm getting an arming osd warning flag that said "rc smoothing". I can't find any info on it. Does this mean the rc smoothing is not working? Too much? another issue? If I arm and get the error, I just have to disarm and arm again to fly. ADC on controller on or off does not change. Tried smoothing at 30, 50 and 100 no change to the osd warning.
Not sure. Maybe ask in the support hanger on my discord?
How feedforward term related to rates?
Hi Chris, I recently came across the motor mixer weights. They are a bit hidden in betaflight but they can be used to correct for the different moment arms of the motors. On deadcat frames, the front motor are often further apart than the rear motors so the moment arms on the roll axis are unequal. Do you think it is worthwhile to change the weight? Iforce2d has a calculator to work them out.
Yes I think so. It can really help with yaw coupling on pitch and roll moves.
does that mean frsky and flysky are at 50hrz? I always wondered this.
Nice video!
(it says part 4 but there are only 3 videos in your BF4.3 playlist...)
Uh oh! I'll check and see which video I've missed. Thanks!
On my Larva X HD BF 4.3 generates flyaway with default filters and what you recommended so far. It basically oscillates around 100 Hz when hovering and... goes up up up. That's very weird as my Mobula6 has no such issue. Given I've no blackblox on either it's a bit hard to dig into details. Increasing the master slider increases the oscillation frequency and but doesn't seem to fix it. No filter seem to fix it either.
Reducing the Tracking (P&I Gains) down to 0.4 (instead of 1.0) seem to work (it's not oscillating anymore and seems not to flyaway, so far).
However, should I now continue with what you're suggesting like increase the master multiplier etc., or is that nonsense I'm getting here due to something else?
This sounds like you have a mechanical issue with your build. I would post in the support section of my discord.
When finding your max D gains using the master multiplier slider, shouldn't feed forward be turned off as well? Or, does it not really matter when initially finding max D gains, but only when dealing with the P:I ratio is when you'd need to turn feed forward off? Thanks!
FF doesn't affect max d gains as it is based of your stick movemement rather than the gyro.
hi Chriss, thanks for this tutorial, it really helps..there is ashadow point though..
after finding the correct Dgain what should we do with the dynamic damping slide? leave it at zero or increase?
this is not shown in the explanation.
thanks.
I would leave it at zero. I don't typically use dynamic damping.
Thanks! But when D tuning first, do we leave sliders to positions we think are good? While whole time dynamic d slider is 0? And rest for other sliders?
Are tuning flights any different for cinewhoops? I just built a shendrones squirt and based on how unpredictable it is on slow casual flights right now, I could never imagine doing flips or rolls with it.
I've been struggling to get it to a place where I have no yaw washout while eliminating fly away potential. I can't seem to find that happy medium and I keep flip flopping between having airmode on or off.
You'll always get better handling with airmode on. Head over to my discord and ask the Air Crew!
Hi Cris, wonderful job as usual. Thanks.
What if is not a 5inch quad to be tuned, how to get the right P:D balance which is not default anymore for Betaflight...
the master multiplier works great for 5 inch as you said... but for all others? have you any tips to where to start?
Pretty sure there is gonna be presets for all classes and you should be able to use that pd ratio and just slide the multipliers up till you get wierd shit to happen and then turn it down a notch.
Don't worry you can follow the same process for any quad. DO the master slider first, then increase the P&I gain slider until you get bounceback/oscillation on snap moves then back off a little. PID toolbox step response makes it very easy.
Amazing as always!
One question tho: I applied the Frsky preset and watched UAVtech's video about RC links and he (and the presets) suggest the opposite setting of averaging: OFF for Frsky, higher for the different high-performance links (which I still don't have yet lol), so what's the matter?
FF averaging is mostly set to off now due to improved RC smoothing in later versions of 4.3 so you are right to follow the presets.
@@ChrisRosser Thanks! Also, interestingly enough, I don't like anything over 35 auto smoothing, the quad feels really floaty and not locked in enough for me.
And I think I have it tuned reasonably well - now that I'm thinking about that, I have D at 69 and 79 and everything is fine, while flying beat up mismatched pitch props lmao, only RPM and notch filter on gyro, no notch on D and D slider at 1.1.....and long boy TBS antenna, in a metal mount xD
is there any tune available for the aos 5 on bf 4.3 ?
Chris, do you tune in a bit of wind or wait for a perfect day? Does it make a difference if you tune in a light wind vs a quiet windless day? Thank you for all your videos!
I tune in all weathers. A bit of wind doesn't hurt when trying to tune a quad as it gives a bit of outside influence for the PID controller to fight against.
Hi,
What does an oscillation sound like?
Also what does an overshoot sound like?
Oscillations give a nasty buzzing or trilling noise. Overshoot sounds like a double blip of the motors at the end of a sharp move.
What are the units on all these settings?
i never get over dampened. my quads seem to runaway from d gain way before that.. Now that its extremely cold.. my 3" oscillating sound like it wants to run up just a hair at certain point on throttle above idle.. but didnt run away. did full throttle burst was fine.. . DIdnt know cold would effect it.
HI Chris, should we leave the Dynamic Damping and Stick Response slider at zero after finishing the tuning?
Dynamic damping can be left at zero. Stick response should be returned to 1.0 then tuned according to my FF tuning guide.
Whenever I try to set my feedfoward averaging to the value 2 for crossfire at locked 150Hz, every time I try to save the change it turns averaging back off. Any idea why it's doing this?
Sounds like it could be a configurator bug. Try updating to the latest versions of configurator and BF 4.3 if you haven't already.
what can i do after filtertuning on my 7 inch rig, while i have master slider is on 2.0 and everything is good. how can i find my perfect D? i need to rais D fist also till motors sounds noisy at arming. filter tuning is on default pids. i can not imaging the my d filter starts at 120hz minimum. motor is starting to get noisy at this minimum. i think the tutorial is not so good for 7 inch quads?
🤔
You can find tuning support on my Discord server in the support hanger.
Been watching a few of your tuning vids today, great video! but this seems to not be for me, i`d rather just keep flying bad flying quads than learn all this
Hey no worries! You can have plenty of fun just flying stock settings. Tuning will always be there if you change your mind.
@@ChrisRosser i'll most definitely keep watching all your vids, it should stick eventually I hope, It just seems a bit daunting at times...haha
I know you don't specifically mention this in this video, but what software do you use to do finite element resonance analysis? I'm very curious. Thanks for all the great info, keep it up!
Could you do a low level explanation or point to some reading as I want to learn this from the bottom up.
You can watch my other PID tuning video series. That should take you from zero to expert.
@@ChrisRosser I think I'm getting it. I'm on my third year of physics and I'm wondering how or if I can apply that knowledge to this hobby.
hello friend ....i just build 5" frame HSKRC 225MM motor 2306.5 2550kv/ esc BL Heli 30A / Flight controller SpeedyBee F7 AIO / gps .....just finish the the setup beta flight select auto 5 " ...still yest to test fly .....is my basic selection of parts correct ???....what will be the setting PID ??......Thank you
Parts look OK to me. For help and support come over to my discord server :D
@@ChrisRosser thank you friend..... server???
@@muhammaddanial8497 discord.gg/z5nVeNE4KR
Nice video. Where is the part about D- min. You explain to turn it off when trying to find max D gain. But you never mention when to put it on and how to tune it properly ? - Would u still on 4.3 recommend to keep D-min off?
I would not use D-min. The difficulty with D-min is it takes time to ramp up which means its not good at tackling sudden propwash. Much better to just run the highest constant D-gain you can.
At what point do we turn dynamic D back on? Or do you recommend not using it?
I don't use it. I like my Dterm to stay still!
what are people doing to re calibrate their FC if they start fresh with these new versions of betaflight? how can you properly calibrate it when it's in the quad? I got a small level but it doesn't work very well.
I just hold the quad down flat on the desk to calibrate the accelerometer. If you don't use angle mode then I would just turn off the accelerometer to save CPU, black log space and avoid the need to cal 😁
What settings for DJI radio?
I would go auto smoothing of 50 and see how it feels
I followed this tune completely on a cine whoop felt great got all the way to the end and noticed some wobbles when kind of just hovering very strange but even if I restart I still get the wobbles a little bit it’s not bad very minor just wondering if you had any insight
could it be adding feedforward ive changed everything back to normal except that and i felt like the wobbles only started then but at zero is fine im going to test this here when i can ill let you know and thanks for everything you do in the hobby
If you think its feedforward try increasing RC smoothing.
@@ChrisRosser okay ill try that thank you i lowered feedforward and it got rid of it ill try to work them together till it feels perfect
@@ChrisRosser sorry for commenting so much, what number do you think i should set smoothness to and do you think i should increase jitter to what betaflight suggest?
@@EdturdEnfart Watch my latest video for more info on exactly this! Rates tuning in 4.3
They got rid of dynamic notch on the latest update... Wonder why?
It's still there, I think it moved to the filter tab?
Could u do next for cinewhooops
I'll do some practical pid tuning guides for different quads 👍
Where's the hat?! 🎩 Oh wait, that never happened.
Haha! Glad you liked that, I might slip that easter egg into future videos for you, watch out for it!
Are you still using RPM Filter on 4.3 Chris?
Yes, absolutely! RPM filtering is essential for best performance on 4.3. RPM filtering is more effective than dynamic notch for motor noise and has less delay. With RPM filtering a single dynamic notch is usually enough. Dynamic Idle is also essential for best performance on 4.3 as well so bi directional d-shot is a must have. 👍👍👍
@@ChrisRosser thx, I found that on my armattan badger 5" using ethix p4 props .... Dynamic idle of 28 makes the quad super-responsive coming out of sharp turns.... This is definitely not a gimmick... Took me awhile to find the correct value ... I appreciate your time my friend
I guess 3:34 IS NOT TRUE for D term. D is derivative from PID error, not from Gyro. It makes significant difference ! I checked it for BF 4.3.1
Hi Stepan, in BF 4.3 D term is the derivative of Gyro not PID error and feedforward is the derivative from setpoint. This has been the behaviour for a long time. I made this mistake in my BF 4.2 video and was corrected by the devs!
@@ChrisRosser Yes, you are right, I am wrong. I doublechecked my logs and found you were right :) Sorry for insuling you!
@@ChrisRosser And thank you for your videos!
I don't understand why we have to set Pidsum_Limit to 1000 and we can restore at 400 once the pid tuning process is terminated
Increasing pidsum limit allows the PID loop to push harder which can improve the tune for quads that don't have a lot of control authority. If you increase the limit you never need to turn it down.
@@ChrisRosser thank u so much...
Wait so is 4.3 out now? Or still have to wait
Release Candidate 1 is planned for Christmas. Final release early 2022.
It's available if you want to try it. The second video in this series will get you set up.
@@ChrisRosser I upgraded my nazgul 5 V2 6S today.. Haven't tried it yet. What dynamic idle would you recommend for ethix s3 on 5in nazgul 5? I set to 20 based on the other video.. But maybe I need to go up to 25-30? I'm so overwhelmed by PID/filter tuning I'll have to do...I'm a perfectionist, so I can't just leave it default.
@@PkwyDrive13 Start at 20 and increase gradually until you find the right balance of inverted hangtime vs propwash performance.
Why the dev's have set pid sum to 400 instead of 1000 ?
I think it's just a conservative setting from the past. I've always set it to 1000.