I've been in the hobby quite a long time and watched my fair share of tuning videos. You sir have created (IMO) the best, most easy to follow along series of videos on the subject I've ever seen. Thank you!
Hey Chris. Just wanted to say thank you for all of your tuning/analysis videos. They remind me that old chinese proverb: "Give a poor man a fish and you feed him for a day, teach him how to fish and you feed him for his life time."
This is becoming my favourite FPV channel, your content and method of delivery is exactly what I like. This complete tuning guide is just excellent, you've covered things in this particular video that I haven't seen anywhere else. My drone was doing the warbling and I didn't know how to fix it.
Glad you enjoyed it Stuart! Even this guide is not yet complete but I'll keep making videos hopefully some of these video in the future will also help you.
I mentioned this to UAV tech. We aren't learning "new" information. Its just being presented in a more tangible way with loads of graphics, info, and better explanation. I am not saying one is better than the other. I am saying that we need both because cavemen like me have to watch UAVtechs videos 5-6 times in an hour to comprehend.
Amazing! Yet another one! Your PID tuning videos have brought this subject from the realm of magic to the realm of artisanship. I need to watch it several times more to gain a full understanding of the subject and I will create an Excel/Open Offce spread sheet incorporating the steps of fine tuning you discuss in this video, so that tuning of P:D and I:P balance is easier than using the sliders in BF. The I:P:D default values/proportions in BF may be close to perfect on a typical 5-inch quad, but when I tried following your previous videos with my super-light 3-inch quads I was not able to continue with the sliders - I had to enter my own values - which caused the sliders to vanish. A well-designed spreadsheet, where each step is e.g. in a separate tab (I'm brainstorming as I write this comment 😄), that is fed data from the previos step/tab, may be a useful tool for many quad pilots - particularly those outside the 5-inch territory (2 - 4-inch on one side and 6+ inches on the other). When I'm ready, I'd like to share it with you (for the purpose of evaluation ) before making it public
@@ChrisRosser I have made that spreadsheet - and even tested it (with quite a lot of success) tuning a 3- and a 2.5-inch toothpicks, but all-in-all - it is a fairly simple and underwhelming tool, I'm not sure if it is worth even talking about... It certainly structurises the process of turning a quad and leaves a record of how you got from the base PIDs to the final result. Honestly - on a tiny quad with no blackbox - you can rely entirely on this schematic while turning (provided you took care of the filters first). I'm happy to share it - if you - or anyone else wants to check it out.
@@ChrisRosser drive.google.com/file/d/1-K8T_9_6aVjatYJONLtJObdRg9FCPosP/view?usp=drivesdk I am sure it can be developed further into a much more complete and comprehensive tool. However simple in its current firm - it was enough for me to achieve a decent tune of my micros. I only had to slightly modify the base PIDs to reflect the fact that my quads utilise true X frames - so I set the roll and pitch values as equal before I started tuning.
I've been PID tuning for about 2 years now, most of my understanding of PIDs comes from Bardwell's PID Tuning Masterclass series of videos. So by now, I can say with confidence that I have gotten good at tuning just from my own understanding of what to change to fix what issue. HOWEVER, I will say that your videos on filter tuning really put me 3 notches up as far as the quality of the tunes I'm able to achieve.
Another great video! I particularly enjoyed the D_min section. I agree with you; thinking of it as a D boost is way better than thinking of it as D min and D max. Thanks, Chris!
Thanks Chris!! I think you made the best video series about quad tuning I've ever seen on the web!! One thing I cannot properly understand... What does exactly "your quad is not stable in the air at zero throttle" mean ? What should I look at to figure out if this condition occurs ?
If you go to zero throttle (after a punch out) and the quad wobbles around when you don't move the sticks. That's low throttle instability. Best solution is to increase min throttle and dynamic idle rpm.
What a fantastic series of video sir! Super informative and super interesting to fully understand what we have on hands. If you can make more about the BB Explorer analysis and how to interpret all the values, that will be very much appreciated! Thanks once again for this big contribution to the community! Cheers
Also good thing about feed_forward is to play with ff_boost and Spike_limit. Helps to push gyro to set point closer at start of the flip, without increasing ff values so much, that will get overshoot at end of the flip :)
@@ChrisRosser Hard to explain, but high P+D work well enough to let I-term stay inactive. Maybe because good weight balance. But on sharp half front flip + look back after fast throttle cut, high I-term start to drag quad somewhere slowly without any reason. Feels like gyro slowly walking somewhere around setpoint. Then more I add to I-term, then more strange its become. Lowering I-term for one tick of master slider and incrace anti-gravity to 6.5 gived good enough fast throttle cut handling, not perfect. But in sharp flips after throttle cut its doing perfect setpoint tracking, like falkoX usually doing. Maybe because P/I balance, i dont know) My tune looks like completely wrong, but i like how its flies more than my Fettec quad fly))
@@LizzzardRUS Ok. You like how it flies so that's the most important thing! If the quad nose is wandering that could be I term oscillation (its much slower than P oscillation).
Yeah, it’s very counterintuitive and I wish the logic of “reduced to x% of original PID values” rather than “reduced by x% of original values” was implemented. It’s also worth mentioning that by BF default, TPA only affects D values. If you want it to also affect P values, it must be changed in the CLI to tpa_mode = PD
So you are saying increase TPA to reduce warbling on high throttle? How high can it go? Also how does TPA affect mid throttle oscillations? Thanks for all you do Chris.
Hi! Amazing video and explanation, anyway , I'm a rookie and I'm trying to "learn something" as the other guy on RUclips says. I have a Flywoo Explorer LR 4" quad , I installed Betaflight 5.0 and since there are no pid settings available for that, I need to PID tune manually. I followed Your videos and I'll watch them again, my question is the following: Saying that my quad already flies with some pid settings, I understand by looking at the "effects of dumping" chart at minute 9:18 that the right P/D ratio should be always around 0.75, independently of the initial P value, is this true? So having to start from scratch, should I set , for example intial PIDs P 40 , D 0 , I 0, increase gradually P until I see underdumping, then calculate D=0.75*P and after that start increasing both until I see oscillations or vibrations or motors heat? What about I ? Is there an ideal P/I ratio? How do I find the right starting value at the first iteration of the process? Thank You for Your help. Luca
Another tuning lecture, yay! I actually learned quite a bit here about TPA, dynamic idle (never knew about that) and especially D-min boost. Had to watch that d-min section mulitple times to really understand it and it's actually quite a nice tool if you think of it. Thanks for exmplaining how BF actually calculates those values! One question that I got after watching your vid: I learned that "v-bat sag compensation (in CLI) is the new v-bat PID-Compensation... so use that instead and leave the v-bat PID comp untouched." What's your take on this? And I keep thumbs pressed that BF implements those ratio sliders including IP as you hinted. This will make things so much easier to set up (especially on mobile!)
Two comments/questions about TPA. First, in the most recent version of BF the default TPA behavior is to attenuate only D with increasing throttle (tpa_mode = D). Any idea why this is the default? Wouldn't it affect the tune, step response, etc.? Should we change it to tpa_mode = PD for that reason? Second, the BF documentation says that the TPA setting represents "% of dampening that will occur at full throttle". Doesn't that mean that it has the opposite behavior of what you've described (see figure at 15:51)? If you set TPA = 0.25, wouldn’t the PIDs at max throttle be 75% of what they would be otherwise, based on the BF documentation? Once again, thanks for such a thorough and scientific approach to PID tuning! I’ll definitely be referring to this video frequently in the future.
Hi, thanks for your comment. I've added a correction for the behaviour of TPA. I haven't played with D attenuation vs PD so I'll need to look into that.
I mentioned I wished the sliders would work multiplicatively in the comments where we voted on the PD balance method the sliders took. Would be great if the sliders reset to 0 gains when you adjust the numbers manually then moving the sliders once again started multiplying.
Great video as always! There's a downside of crancking the I gain along with the P gain since in theory a greater I gain leads to a longer settling time. This is why I always kept my I gains around stock values. Maybe it's not the case with betaflight since iterm relax mitigate the iterm effect during step moves. I'll give your method a try 😊
Great instructional videos, been watching the whole series and trying to learn tuning following your footsteps. BTW quick question on Dynamic Idle, in order to derive the Dynamic Idle rpm value the test battery used should be at 3.8V per cell according to your presentation. So for a 3S quad, the test voltage would be 11.4V (3.8V x 3)? or is it to just connect a 1S battery at 3.8V? Thanks!
Chris could try to record how “warbling” sound actually sounds? All tuning by ear videos kind of assume “you’d know when you hear it”, but may be I’m just thick so I can’t. I’ve got 3 quads with different motors and props and they all have different sound. I have no idea if it’s “smooth” enough or not. So far I could only hear definite difference when prop is so bent that blade points upwards or resonance when I forgot to tighten frame bolts.
Its a hard to describe sound. Kind of sounds like a motor is vibrating while you're flying. "you know it when you hear it" as unhelpful as that sounds, is a really good way to describe it. I'm sorry lol.
It screams, like the drone is screaming... imagine a high resonance going through the carbon frame like a tuning fork and what that would sound like. For me, I had my drone tuned really well, but if I held the throttle open for about 3-4 seconds, I would get this noise right as it hit top speed, it was very loud and noticeable, if you're drone is doing it then you will know.
Just posted a correction comment. Increase TPA to reduce warbling at high throttle. If you don't have any warbling you could consider reducing TPA but I usually fly the default value.
Hi Chris! Your videos are great and more scientific than most fpv stuff, which operates on "feel" and "juicy sbang" whatever that may mean :) but still, at some place you give exact numeric recommendations like changing pids by 10%, but then in I term part, you say things like "if your I term because high in blackbox" without any reference to what is high :) would be great if you could give numeric definitions to all the terms, in absolute or relative values. Thanks!
If you open your log and use trace template 4. Stick tracking you can see your Iterm. Keep increasing I term until you start seeing I term oscillations in the logs then back off a bit.
I notice that when you reach motor saturation in the log at 26:42 you lose a little setpoint tracking on pitch and yaw at the end of the move. Trying to work around this on my quad I found I ended up with much higher roll and yaw values then pitch. I also tried things like the alternative mixer modes which were fun to experiment with. Ended up just settling on lower values over all. I could keep raising pitch gains without trilling but the loss of tracking on the other axis gets worse the longer motor/pidsum is saturated. Is this just the limit of prop/motor combo or is there a better way to keep authority on the other axis when a motor saturates?
Once a motor saturates you get what you get I think. The gyro will start to deviate from setpoint on all axes. The key is avoiding overshoot as the quad returns to the setpoint.
Hello Chris. I ve watched your tuning videos repeatedly. I have to say I'm having a blast , being able to tune my quads. I ve been thinking about d min . Thing is when we set master slider to get D values around 45 and + its not recommended to fly the quad back after crash because at that state its risky . Could d min value help returning the quad with no sharp inputs- slowly ?
Dmin doesn't work like that I'm afraid. It will boost your D gain in response to vibration not reduce it! You shouldn't really be using Dmin for freestyle. A good suggestion is to switch PID profiles and fly back on stock PIDs.
@@ChrisRosser Thanks for the suggestion that's a really good tip. I've read Mark's PID Tuning Principals PDF file and it says like "Once D_min is activated the “D_min” column adds to the PID gains grid (Figure 10). The D_min column is now your base D-gains in normal forward flight! The quad will spend most of the time flying at the D_min value. The “Derivative” column is the boosted D-gains in sharp stick input and during prop wash (aka “D_max”)." I got the idea from there. D min as conservative value and Dmax as the best feel I can get. But thinking further the transition between these values will cause some other performance issues.
This is great! Since TPA breakpoint is measured in us, I'm assuming this acts on the rc command, not the motor output? So if I have TPA tuned, and decide to increase my throttle limit, I'd need to then decrease TPA breakpoint to keep it kicking in at the same motor power, right?
Great question. I don't think TPA is affected by a throttle limit. A throttle limit in BF or on your radio will just prevent you from getting to full throttle. TPA will not get to its full reduction because you aren't getting to full throttle but that's OK. Betaflight still has an internal throttle scale that goes to 100% and that's what breakpoint and TPA refer to.. The 1000-2000 is just a thowback to PPM I think. With Sbus and Dshot I don't think there are us pulses any more but 1000 to 2000 is still the range of values used by convention.
Thank you for your vids!! They are really great and they will help fpv so much! One question please? If the idle and the rpms are too high? What are the disadvantages? Thank you!
Chris, Would you mind throwing your slides on the share drive. Like Mark's available PDF Tuning guide, I believe your slides will serve as a similar same reference guide. Thanks
I would love if you would do a video on black box login. For some reason the way you put things makes it easier for me to understand then U AV Tech. I think he gives a lot more technical details which confuses me. I also wanted to ask you is there any thing that you can’t do with PID toolbox that you can do with black box explorer? I am already learning blackbox explore.
Monumental. A true tuning masterclass. Chris did you ever had a look at Emuflight? They have some really interesting experimental feature in their dev branch (Smith predictor, alpha-beta-gamma filter, etc.) and also latest stable version of the FW flies really really good. I initially migrated to Emu from BF due to the GPS simply not working in BF, then sticked with it due to the performance.
yes agreed please mr takes your time to review it,,or maybe regarding th setup blheli32 such as rampup,motort timing, pwmHz,,hihihi,,,the world needs the knowledge in your point of view
This is the missing manual for what all the tuning values mean. The descriptions in the configurator are cryptic compared to this explanation. Thanks so much.
thanks for your videos! I followed you step by step, and this is the best way to scientifically tune pids that I have ever found. Anyway, I've got an exception: I have a 4" long range quad, with 1404 motors, and I think that the motors are not keeping up with this system. ex: I see overshoot on pidtoolbox, but it is a REEEALLY slow overshoot, with its consequent bouncebacks. if i reduce P term, the overshoot reduces, but i have a peak time of 180+ms and the oscillations are really excessive. so I tried to raise P term, and I have reached a peak time between 90-140 and reduced a bit the oscillations. is there a way to tune it better? do I have to manually raise a lot more the pids or the only solution is to change motors/propellers?
@@ChrisRosser thanks for the answer, Reducing the I term solved (partially) the problem. The thing with this quad is that in a straight cruise flight, you will always see some little bumps, hops, like you're driving a car in a messy road. they're very small, but no matter how much I raise my pids, I can't get rid of them. The method you propose in this videos is perfect for racing or acro (my other quads now are perfect), but I can't smooth the straight flight of this long range cruise quad. (it's now almost perfect on snap roll and flips though).
@@Morpheus981 Those little bobbles are caused by not having enough motor torque for the props. A 1404 is really too small for 4" in my opinion. To get rock steady footage you need to have big enough motors to have perfect control over the prop. 1806 is my pick for 4". You may be able to improve things by increasing your timing in BLHeli to 23deg (Medium High) and running lighter pitch and lighter weight props.
Hello Chris and others. What the issue, if motors sounds unhealthy at zero throttle (on the ground), but if I do just little more throttle its sound normally? After flying, motors not hot, everything seems ok. Master multiplier set at 0.5
That is pretty normal. I have motors that sound a little rough on the ground and when I take off it's totally fine. If the quad is on flat soft ground it's usually Ok and on hard ground it's worse. Not sure of the exact cause.
Can you share the slides from the videos? I find the visual references very useful. And thank you for the "reverse-alchemy" of turning black magic into rational (understandable) science!
I appreciate what you're doing. I don't really see the point of fine tuning your quad the 1st time you crash everything changes And I do a lot of crashing
No worries. The important thing is that pilots know the features exist and what they do and can decide if they need to use them for their style of flying 👍
hello，chris rosser，How do you fix the significant difference between PID curves with a rate greater than 500 and less than 500?Pid curves larger than 500 and smaller than 500 are severely separated and do not overlap at all.
If we want are d gain to be as high as possible that we are most comfortable with let's say 65 would we not just set both roll and pitch to have d gains of 65 why would we keep one less then the other?
First :D this tuning vids are really great, but... as beginner not so easy to understand. I think beginners would prefer a one click magical perfect tuning
I think this tuning tutorial is easy to follow, and it gives out the important points. UAVTech is actually very comprehensive but it's a very long series and you have to follow each video and all the blackbox analysis + the spreadsheet to understand
lol dang my idle is 3%, it will take off at 5, and it barely wants to decend at 4% lol, on my little crux35 (im relativly new to fpv and dont have a 5 inch yet)
@@ChrisRosser it’s actually even better than that, my good flying buddy calls you a genius and wants you for president.😂. And he used to be a Flight One fan boy.
I've been trying to fine tune my quad for 3 years already... This video as the other's doesn't help me at all. First of all I always see perfectly tuned graphs in pidtoolbox, but have never seen graphs I get from my blackbox. And every time I'm trying to get help no one from discord groups can help me because they see such a problems from me for the first time. I've changed 5 types of motors, 3 frames, 5 types of FC, and I'm still having the same issue like jello on hd footage. Also I tried different contractions and materials for hd cam holders and now I'm using tpu 70d. The problem I have is Jello on 1280-1320 throttle. In blackbox at this amount of throttle I see pitch dterm oscillation and the amplitude is rising when it goes up to 2000 throttle. I tried to decrease d, added tpa 0.2 on d(I'm now flying emuflight) but it doesn't help at all. Motors are new and I also balanced each one to have less than 0.05 amplitude in motor tab... Is there any ability to help me with tuning and troubleshooting for 1 evening for some donation? I have 8 packs, and could show you hd footage and blackbox files
@@ChrisRosser I watched those. I'm still working on understanding just how a pie controller works but I watched your videos and some other about what each term actually means and I'm starting to understand it.
@@noahmcelwey2598 Great to hear. It's a lot to learn but it sounds like you're on the road! I hope you get your quads flying even better as a result 👍😁
I'll maybe buy the second iteration of your quads, I'm sure they are technically wonderfull but they look kind of.... Dull to me. No offense, If you succeed to make them stylish the may well be the next gold standard of the industry.
@@ChrisRosser If you want to preserve the exact same shape, maybe hydrodiping could add some character to the frame, It could be a special edition. I particularly like the look of the Apex, the Babyhawk 2 also (Emax) it has "triangular" arms too and is known for having good flying characteristics. I think theses frames have in common "organic" curves, yours is really straight. But from what I've seen with deep learning engineering, if you keep optimising you will end up with organic looking shapes. It's not like the forces where equal at all points of the arm.
![Felix "Unfair" | [Stray Kids : SKZ-PLAYER]](http://i.ytimg.com/vi/Oswujxm2Ag0/mqdefault.jpg)
oh man, i am so glad this was in my reccomended videos, what an absolute WEALTH of knowledge on your channel!!
Welcome aboard!
I've been in the hobby quite a long time and watched my fair share of tuning videos. You sir have created (IMO) the best, most easy to follow along series of videos on the subject I've ever seen. Thank you!
👆🏻
You are very welcome! I'm glad you're enjoying the videos and finding them useful.
TRUTH
Hey Chris. Just wanted to say thank you for all of your tuning/analysis videos. They remind me that old chinese proverb: "Give a poor man a fish and you feed him for a day, teach him how to fish and you feed him for his life time."
Teach a man to tune and he flies happily for a lifetime!
This is becoming my favourite FPV channel, your content and method of delivery is exactly what I like. This complete tuning guide is just excellent, you've covered things in this particular video that I haven't seen anywhere else. My drone was doing the warbling and I didn't know how to fix it.
Glad you enjoyed it Stuart! Even this guide is not yet complete but I'll keep making videos hopefully some of these video in the future will also help you.
I mentioned this to UAV tech. We aren't learning "new" information. Its just being presented in a more tangible way with loads of graphics, info, and better explanation. I am not saying one is better than the other. I am saying that we need both because cavemen like me have to watch UAVtechs videos 5-6 times in an hour to comprehend.
Amazing! Yet another one! Your PID tuning videos have brought this subject from the realm of magic to the realm of artisanship. I need to watch it several times more to gain a full understanding of the subject and I will create an Excel/Open Offce spread sheet incorporating the steps of fine tuning you discuss in this video, so that tuning of P:D and I:P balance is easier than using the sliders in BF. The I:P:D default values/proportions in BF may be close to perfect on a typical 5-inch quad, but when I tried following your previous videos with my super-light 3-inch quads I was not able to continue with the sliders - I had to enter my own values - which caused the sliders to vanish.
A well-designed spreadsheet, where each step is e.g. in a separate tab (I'm brainstorming as I write this comment 😄), that is fed data from the previos step/tab, may be a useful tool for many quad pilots - particularly those outside the 5-inch territory (2 - 4-inch on one side and 6+ inches on the other). When I'm ready, I'd like to share it with you (for the purpose of evaluation ) before making it public
A spreadsheet would be a wonderful tool. When you finish it I'll check it over and then make a video for it!
@@ChrisRosser I have made that spreadsheet - and even tested it (with quite a lot of success) tuning a 3- and a 2.5-inch toothpicks, but all-in-all - it is a fairly simple and underwhelming tool, I'm not sure if it is worth even talking about...
It certainly structurises the process of turning a quad and leaves a record of how you got from the base PIDs to the final result. Honestly - on a tiny quad with no blackbox - you can rely entirely on this schematic while turning (provided you took care of the filters first). I'm happy to share it - if you - or anyone else wants to check it out.
@@onemanmob6756 I'd love to check it out. Can you link me to it?
@@ChrisRosser drive.google.com/file/d/1-K8T_9_6aVjatYJONLtJObdRg9FCPosP/view?usp=drivesdk
I am sure it can be developed further into a much more complete and comprehensive tool. However simple in its current firm - it was enough for me to achieve a decent tune of my micros. I only had to slightly modify the base PIDs to reflect the fact that my quads utilise true X frames - so I set the roll and pitch values as equal before I started tuning.
I've been PID tuning for about 2 years now, most of my understanding of PIDs comes from Bardwell's PID Tuning Masterclass series of videos. So by now, I can say with confidence that I have gotten good at tuning just from my own understanding of what to change to fix what issue. HOWEVER, I will say that your videos on filter tuning really put me 3 notches up as far as the quality of the tunes I'm able to achieve.
AMAZING Video! Continue keeping up the great work and providing helpful information.
Thank you! Will do!
Another great video! I particularly enjoyed the D_min section. I agree with you; thinking of it as a D boost is way better than thinking of it as D min and D max. Thanks, Chris!
Glad you enjoyed! I've still got a few more even more detailed topics to cover on ESCs and other stuff which I hope you also enjoy.
@@ChrisRosser Awesome! Other than changing pwm frequency, I know nothing about esc tuning. That video will be great!
Thanks Chris!! I think you made the best video series about quad tuning I've ever seen on the web!!
One thing I cannot properly understand... What does exactly "your quad is not stable in the air at zero throttle" mean ? What should I look at to figure out if this condition occurs ?
If you go to zero throttle (after a punch out) and the quad wobbles around when you don't move the sticks. That's low throttle instability. Best solution is to increase min throttle and dynamic idle rpm.
Thanks! Now it’s clear!
What a fantastic series of video sir! Super informative and super interesting to fully understand what we have on hands. If you can make more about the BB Explorer analysis and how to interpret all the values, that will be very much appreciated! Thanks once again for this big contribution to the community! Cheers
Happy to help. I'm going to keep pushing ahead with the educational content!
Also good thing about feed_forward is to play with ff_boost and Spike_limit. Helps to push gyro to set point closer at start of the flip, without increasing ff values so much, that will get overshoot at end of the flip :)
Yes, I think there may even need to be another video to go into some of the deeper functionality. I had to stop with this one at 30mins!
@@ChrisRosser Just tried higher I-gain values instead of high anti-gravity. No, for me lower I-gain works better with higher anti-gravity gain)
@@LizzzardRUS Interesting. Can you give more details as to what you mean by better?
@@ChrisRosser Hard to explain, but high P+D work well enough to let I-term stay inactive. Maybe because good weight balance. But on sharp half front flip + look back after fast throttle cut, high I-term start to drag quad somewhere slowly without any reason. Feels like gyro slowly walking somewhere around setpoint. Then more I add to I-term, then more strange its become. Lowering I-term for one tick of master slider and incrace anti-gravity to 6.5 gived good enough fast throttle cut handling, not perfect. But in sharp flips after throttle cut its doing perfect setpoint tracking, like falkoX usually doing. Maybe because P/I balance, i dont know) My tune looks like completely wrong, but i like how its flies more than my Fettec quad fly))
@@LizzzardRUS Ok. You like how it flies so that's the most important thing! If the quad nose is wandering that could be I term oscillation (its much slower than P oscillation).
Correction: the TPA value needs to be INCREASED to reduce high throttle oscillation rather than decreased.
Yeah, it’s very counterintuitive and I wish the logic of “reduced to x% of original PID values” rather than “reduced by x% of original values” was implemented. It’s also worth mentioning that by BF default, TPA only affects D values. If you want it to also affect P values, it must be changed in the CLI to tpa_mode = PD
So you are saying increase TPA to reduce warbling on high throttle? How high can it go? Also how does TPA affect mid throttle oscillations? Thanks for all you do Chris.
@@blazer5979 It can go up to 1.0. It may not affect mid throttle oscillation as that tends to be a resonance issue.
Excellent video again! -> you got so many topics covered it would be enough to fill 4 clips!
great vid Chris!
I'll be joining your patreon army as soon as I graduate from building a quad to custom programming the FC
so never?)
@@mikenomatter actually... just did. :P
Love the video! You should do a AOS 5 perfect build and tune so the rest of us can feel an amazing tuned quad :)
I've done an AOS 5 perfect build video. Hope you like it!
Hi! Amazing video and explanation, anyway , I'm a rookie and I'm trying to "learn something" as the other guy on RUclips says. I have a Flywoo Explorer LR 4" quad , I installed Betaflight 5.0 and since there are no pid settings available for that, I need to PID tune manually. I followed Your videos and I'll watch them again, my question is the following:
Saying that my quad already flies with some pid settings, I understand by looking at the "effects of dumping" chart at minute 9:18 that the right P/D ratio should be always around 0.75, independently of the initial P value, is this true? So having to start from scratch, should I set , for example intial PIDs P 40 , D 0 , I 0, increase gradually P until I see underdumping, then calculate D=0.75*P and after that start increasing both until I see oscillations or vibrations or motors heat? What about I ? Is there an ideal P/I ratio? How do I find the right starting value at the first iteration of the process? Thank You for Your help. Luca
Thanks for the video!!. Question... We must to do this before or after Filter Tuning?
Another tuning lecture, yay!
I actually learned quite a bit here about TPA, dynamic idle (never knew about that) and especially D-min boost. Had to watch that d-min section mulitple times to really understand it and it's actually quite a nice tool if you think of it. Thanks for exmplaining how BF actually calculates those values!
One question that I got after watching your vid:
I learned that "v-bat sag compensation (in CLI) is the new v-bat PID-Compensation... so use that instead and leave the v-bat PID comp untouched." What's your take on this?
And I keep thumbs pressed that BF implements those ratio sliders including IP as you hinted. This will make things so much easier to set up (especially on mobile!)
Your the man Chris. Thank you Sir 🙏🏼
Glad to help 😁
Another great video Mr.Rosser
Two comments/questions about TPA. First, in the most recent version of BF the default TPA behavior is to attenuate only D with increasing throttle (tpa_mode = D). Any idea why this is the default? Wouldn't it affect the tune, step response, etc.? Should we change it to tpa_mode = PD for that reason? Second, the BF documentation says that the TPA setting represents "% of dampening that will occur at full throttle". Doesn't that mean that it has the opposite behavior of what you've described (see figure at 15:51)? If you set TPA = 0.25, wouldn’t the PIDs at max throttle be 75% of what they would be otherwise, based on the BF documentation?
Once again, thanks for such a thorough and scientific approach to PID tuning! I’ll definitely be referring to this video frequently in the future.
Hi, thanks for your comment. I've added a correction for the behaviour of TPA. I haven't played with D attenuation vs PD so I'll need to look into that.
I mentioned I wished the sliders would work multiplicatively in the comments where we voted on the PD balance method the sliders took. Would be great if the sliders reset to 0 gains when you adjust the numbers manually then moving the sliders once again started multiplying.
Great video as always! There's a downside of crancking the I gain along with the P gain since in theory a greater I gain leads to a longer settling time. This is why I always kept my I gains around stock values. Maybe it's not the case with betaflight since iterm relax mitigate the iterm effect during step moves. I'll give your method a try 😊
Higher I gain will make the I term react faster. It can cause oscillations if it is too high.
Great instructional videos, been watching the whole series and trying to learn tuning following your footsteps. BTW quick question on Dynamic Idle, in order to derive the Dynamic Idle rpm value the test battery used should be at 3.8V per cell according to your presentation. So for a 3S quad, the test voltage would be 11.4V (3.8V x 3)? or is it to just connect a 1S battery at 3.8V? Thanks!
Yep 11.4V for 3S quad. 3.8 V per cell of the battery.
@@ChrisRosser Thank you for clarifying!
Very very informative thank you very much
Chris could try to record how “warbling” sound actually sounds? All tuning by ear videos kind of assume “you’d know when you hear it”, but may be I’m just thick so I can’t. I’ve got 3 quads with different motors and props and they all have different sound. I have no idea if it’s “smooth” enough or not. So far I could only hear definite difference when prop is so bent that blade points upwards or resonance when I forgot to tighten frame bolts.
Its a hard to describe sound. Kind of sounds like a motor is vibrating while you're flying. "you know it when you hear it" as unhelpful as that sounds, is a really good way to describe it. I'm sorry lol.
It screams, like the drone is screaming... imagine a high resonance going through the carbon frame like a tuning fork and what that would sound like. For me, I had my drone tuned really well, but if I held the throttle open for about 3-4 seconds, I would get this noise right as it hit top speed, it was very loud and noticeable, if you're drone is doing it then you will know.
It's like a goat scream 😉👍
lots of information,thanks,and great,Chris,,wish to see a flying footage at the end✌️
Thank You Professor 🙏
You are very welcome
Man you are legend.
Another great video Chris! Any tips on tuning TPA?
go to the PID toolbox channel - the have best description of that
Just posted a correction comment. Increase TPA to reduce warbling at high throttle. If you don't have any warbling you could consider reducing TPA but I usually fly the default value.
@@ChrisRosser Thank you, Chris!
Hi Chris! Your videos are great and more scientific than most fpv stuff, which operates on "feel" and "juicy sbang" whatever that may mean :) but still, at some place you give exact numeric recommendations like changing pids by 10%, but then in I term part, you say things like "if your I term because high in blackbox" without any reference to what is high :) would be great if you could give numeric definitions to all the terms, in absolute or relative values. Thanks!
If you open your log and use trace template 4. Stick tracking you can see your Iterm. Keep increasing I term until you start seeing I term oscillations in the logs then back off a bit.
the best fpv channel, question. In the second video we set pidsum_limit = 1000 so it stays at 1000?
Yes, that's right. You start this video at the end of video 2 so everything should be the same.
I notice that when you reach motor saturation in the log at 26:42 you lose a little setpoint tracking on pitch and yaw at the end of the move. Trying to work around this on my quad I found I ended up with much higher roll and yaw values then pitch. I also tried things like the alternative mixer modes which were fun to experiment with. Ended up just settling on lower values over all. I could keep raising pitch gains without trilling but the loss of tracking on the other axis gets worse the longer motor/pidsum is saturated. Is this just the limit of prop/motor combo or is there a better way to keep authority on the other axis when a motor saturates?
Once a motor saturates you get what you get I think. The gyro will start to deviate from setpoint on all axes. The key is avoiding overshoot as the quad returns to the setpoint.
Thanks again.
Hello Chris. I ve watched your tuning videos repeatedly. I have to say I'm having a blast , being able to tune my quads. I ve been thinking about d min . Thing is when we set master slider to get D values around 45 and + its not recommended to fly the quad back after crash because at that state its risky . Could d min value help returning the quad with no sharp inputs- slowly ?
Dmin doesn't work like that I'm afraid. It will boost your D gain in response to vibration not reduce it! You shouldn't really be using Dmin for freestyle. A good suggestion is to switch PID profiles and fly back on stock PIDs.
@@ChrisRosser Thanks for the suggestion that's a really good tip. I've read Mark's PID Tuning Principals PDF file and it says like
"Once D_min is activated the “D_min” column adds to the
PID gains grid (Figure 10). The D_min column is now
your base D-gains in normal forward flight! The quad
will spend most of the time flying at the D_min value.
The “Derivative” column is the boosted D-gains in sharp stick input and during prop wash (aka “D_max”)."
I got the idea from there. D min as conservative value and Dmax as the best feel I can get. But thinking further the transition between these values will cause some other performance issues.
This is great! Since TPA breakpoint is measured in us, I'm assuming this acts on the rc command, not the motor output? So if I have TPA tuned, and decide to increase my throttle limit, I'd need to then decrease TPA breakpoint to keep it kicking in at the same motor power, right?
Great question. I don't think TPA is affected by a throttle limit. A throttle limit in BF or on your radio will just prevent you from getting to full throttle. TPA will not get to its full reduction because you aren't getting to full throttle but that's OK. Betaflight still has an internal throttle scale that goes to 100% and that's what breakpoint and TPA refer to.. The 1000-2000 is just a thowback to PPM I think. With Sbus and Dshot I don't think there are us pulses any more but 1000 to 2000 is still the range of values used by convention.
Thank you for your vids!! They are really great and they will help fpv so much! One question please? If the idle and the rpms are too high? What are the disadvantages? Thank you!
Chris, Would you mind throwing your slides on the share drive. Like Mark's available PDF Tuning guide, I believe your slides will serve as a similar same reference guide. Thanks
Sure. I'll do it when I get back home this weekend.
Very good videos, thank you 👍
I would love if you would do a video on black box login. For some reason the way you put things makes it easier for me to understand then U AV Tech.
I think he gives a lot more technical details which confuses me. I also wanted to ask you is there any thing that you can’t do with PID toolbox that you can do with black box explorer? I am already learning blackbox explore.
Monumental. A true tuning masterclass.
Chris did you ever had a look at Emuflight? They have some really interesting experimental feature in their dev branch (Smith predictor, alpha-beta-gamma filter, etc.) and also latest stable version of the FW flies really really good. I initially migrated to Emu from BF due to the GPS simply not working in BF, then sticked with it due to the performance.
yes mr chris,,emu sounds great
Emu has a better filter strategy than beta flight so yes, chris, look into it!
yes agreed please mr takes your time to review it,,or maybe regarding th setup blheli32 such as rampup,motort timing, pwmHz,,hihihi,,,the world needs the knowledge in your point of view
This is the missing manual for what all the tuning values mean. The descriptions in the configurator are cryptic compared to this explanation. Thanks so much.
You are welcome! No problem. 😁
thanks for your videos! I followed you step by step, and this is the best way to scientifically tune pids that I have ever found. Anyway, I've got an exception: I have a 4" long range quad, with 1404 motors, and I think that the motors are not keeping up with this system.
ex: I see overshoot on pidtoolbox, but it is a REEEALLY slow overshoot, with its consequent bouncebacks. if i reduce P term, the overshoot reduces, but i have a peak time of 180+ms and the oscillations are really excessive.
so I tried to raise P term, and I have reached a peak time between 90-140 and reduced a bit the oscillations.
is there a way to tune it better? do I have to manually raise a lot more the pids or the only solution is to change motors/propellers?
A slow bounce back could be I term. Have a look in the log to see which term is causing the bounce
@@ChrisRosser thanks for the answer, Reducing the I term solved (partially) the problem. The thing with this quad is that in a straight cruise flight, you will always see some little bumps, hops, like you're driving a car in a messy road. they're very small, but no matter how much I raise my pids, I can't get rid of them.
The method you propose in this videos is perfect for racing or acro (my other quads now are perfect), but I can't smooth the straight flight of this long range cruise quad.
(it's now almost perfect on snap roll and flips though).
@@Morpheus981 Those little bobbles are caused by not having enough motor torque for the props. A 1404 is really too small for 4" in my opinion. To get rock steady footage you need to have big enough motors to have perfect control over the prop. 1806 is my pick for 4". You may be able to improve things by increasing your timing in BLHeli to 23deg (Medium High) and running lighter pitch and lighter weight props.
UAVtech recommends 0.65 and 1750 for TPA settings.. Which I put on my 5in nazgul 5 V2. Is 1750 too high?
Hello Chris and others. What the issue, if motors sounds unhealthy at zero throttle (on the ground), but if I do just little more throttle its sound normally? After flying, motors not hot, everything seems ok. Master multiplier set at 0.5
That is pretty normal. I have motors that sound a little rough on the ground and when I take off it's totally fine. If the quad is on flat soft ground it's usually Ok and on hard ground it's worse. Not sure of the exact cause.
Can you share the slides from the videos? I find the visual references very useful. And thank you for the "reverse-alchemy" of turning black magic into rational (understandable) science!
I'll put the slides on the google drive this weekend.
@@ChrisRosser Fantastic! Thank you.
so to simplify, TPA Break is roughly 10-15% higher than Throttle MID?
Yeah that sounds about right to me.
I like this kind of video, good job ! But i miss subtitles xd
How to enable dyn idle?
Enter the value in the highlighted field. If you don't see the field check your BF version.
I appreciate what you're doing. I don't really see the point of fine tuning your quad the 1st time you crash everything changes And I do a lot of crashing
No worries. The important thing is that pilots know the features exist and what they do and can decide if they need to use them for their style of flying 👍
hello，chris rosser，How do you fix the significant difference between PID curves with a rate greater than 500 and less than 500?Pid curves larger than 500 and smaller than 500 are severely separated and do not overlap at all.
Hmm, that's interesting. Can you provide a log?
@@ChrisRosser I've had this problem for months, and I can't solve it.Could you tell me your email address?I'll send you the black box.
@@穿越机入坑指南 you can find it on aos-rc.com
This is the bible of fpv right here.
If we want are d gain to be as high as possible that we are most comfortable with let's say 65 would we not just set both roll and pitch to have d gains of 65 why would we keep one less then the other?
The quad might be more sensitive to noise (or have more noise) on roll vs pitch so you can have more D gain on one axis than the other.
First :D this tuning vids are really great, but... as beginner not so easy to understand. I think beginners would prefer a one click magical perfect tuning
Watch his previous PID tuning videos, those are catered towards beginners.
We have a problem. One click magical buttons don't exist 😜 ...ok: yet :)
Beginner tune = buy a ready to fly and get some stick time. Or at the very least if you want to build your own just fly Betaflight defaults.
... good luck in this hobby. Wait for Joshua Bardwells digest, or you'll have to invest some time..
I think this tuning tutorial is easy to follow, and it gives out the important points. UAVTech is actually very comprehensive but it's a very long series and you have to follow each video and all the blackbox analysis + the spreadsheet to understand
lol dang my idle is 3%, it will take off at 5, and it barely wants to decend at 4% lol, on my little crux35 (im relativly new to fpv and dont have a 5 inch yet)
Excellent p
Super interesting as usual, but tbh, my quads fly really good now with the help of your first three videos so I’ll just stop there.
Glad you enjoyed it! I'm happy to hear your quad flies how you want 😀
@@ChrisRosser it’s actually even better than that, my good flying buddy calls you a genius and wants you for president.😂. And he used to be a Flight One fan boy.
Min/Max time!
I've been trying to fine tune my quad for 3 years already... This video as the other's doesn't help me at all. First of all I always see perfectly tuned graphs in pidtoolbox, but have never seen graphs I get from my blackbox. And every time I'm trying to get help no one from discord groups can help me because they see such a problems from me for the first time. I've changed 5 types of motors, 3 frames, 5 types of FC, and I'm still having the same issue like jello on hd footage. Also I tried different contractions and materials for hd cam holders and now I'm using tpu 70d. The problem I have is Jello on 1280-1320 throttle. In blackbox at this amount of throttle I see pitch dterm oscillation and the amplitude is rising when it goes up to 2000 throttle. I tried to decrease d, added tpa 0.2 on d(I'm now flying emuflight) but it doesn't help at all. Motors are new and I also balanced each one to have less than 0.05 amplitude in motor tab... Is there any ability to help me with tuning and troubleshooting for 1 evening for some donation? I have 8 packs, and could show you hd footage and blackbox files
Is the HD footage coming from DVR? Or is it from an action camera like a GoPro?
@@kreativ_corals insta360 one r
@@О.Ш-ь1щ you using ND filters?
@@О.Ш-ь1щ the insta 360 is why you have jello lol get a GoPro
@@kreativ_coralsnot yet
"I hope you enjoyed this video and have a more n depth understanding" *restarts the entire video because I have gotten completely lost*
This is the most complicated one. I have some earlier filter and PID tuning videos to watch first which are less in depth.
@@ChrisRosser I watched those. I'm still working on understanding just how a pie controller works but I watched your videos and some other about what each term actually means and I'm starting to understand it.
@@noahmcelwey2598 Great to hear. It's a lot to learn but it sounds like you're on the road! I hope you get your quads flying even better as a result 👍😁
There’s one thumbs down.. thanks to jb
I'll maybe buy the second iteration of your quads, I'm sure they are technically wonderfull but they look kind of.... Dull to me.
No offense,
If you succeed to make them stylish the may well be the next gold standard of the industry.
@@ChrisRosser If you want to preserve the exact same shape, maybe hydrodiping could add some character to the frame, It could be a special edition.
I particularly like the look of the Apex, the Babyhawk 2 also (Emax) it has "triangular" arms too and is known for having good flying characteristics.
I think theses frames have in common "organic" curves, yours is really straight.
But from what I've seen with deep learning engineering, if you keep optimising you will end up with organic looking shapes. It's not like the forces where equal at all points of the arm.
👍👀🇭🇷
Can I ask this same series but now for kiss? 💋
KISS doesn't have many of these features
@@DriftaholiC it will be a good short series lol