Man I've learned PID in school and have been exposed to it constantly over the years, but I've never understood it better than after watching this. THANK YOU
I hope i can eventually get my brain to process all this where it alows me to make the proper adjustments to the pid loop. Its all a geuss at this point for me still. It has got to be frustrating for you to keep explaining this to us!! Thats why your an engineer and i just fall in line!! Im so glad that Betaflight mostly gets it right or i would have quit long time ago. Thank you Mark!!
I watched sooo many videos including yours and they are all soooo good. But I am stupid...really stupid.... this video at 3:30: this finally made me ACUTALLY UNDERSTAND the P-D-Balance relationship. Thank you Mark :)
excellent vid! I love how you always point out new ways to look at PID control. Can never get enough of this. You're the only TH-cam guy talkin this stuff nowadays. Thanks and keep it up please!
I suddenly got this NEED to go tune my copters, even though they fly reasonably fine. :-) To bad it's blowing 30 knots. Good content, as always. Thanks!
Terrific explanation. Really helping me to understand PID tuning. My key takeaway is that the P&D terms work against each other during stick moves, but they work with each other during wind/external disturbances. I do wonder though whether PID is basically completely the wrong approach for RC flying.
Thanks so much for these videos, the sliders in betaflight are awesome, at first I didn’t use them but they let you dial in 95% of the changes needed in a couple packs. I always use RPM filtering now and adjust my P/D balance slider until I get a tiny bit of bounce back then raise it back up a notch. Then adjust P/D gains to make the quad feel more or less locked in. Then I leave PIDs alone and raise my filter sliders together while checking for hot motors, this usually makes propwash handling much better. Then I stop tuning and fly, I think this is where a lot of people come unstuck, unless you’re logging continuing to bring out problems to fix is a recipe for chasing your tail. Instead I just fly and make small tweaks if I notice a problem recurring, I might adjust antigravity or raise P term on yaw or I terms (I move away from sliders at this point). I also spend a lot more time flying rather than tuning this way too 👍
Thanks again, Mark. I get 3 quads now tuned how you described it in your videos and it's pure joy to fly them. But there are difficulties when it comes to toothpicks or other micro quads with no blackbox. Additionally some AIO FC's have much electrical noise on their supply rails that affects the gyro signal. May be you can make a video about toothpick tuning?
Great video! One question on the P term trace at time 4:07, why p is always positive when gyro signal bounce on both sides of the setpoint? I thought P just tracks error with a constant multipler, so it should map the polarity of the error as well?
P is the opposite sign of PID Error - proportional in shape - times a gain strength (p-gain). At that time stamp, P is in both signs from what I see. What am I missing?
@@uavtech Oh, sorry the time should be 8:39. I was watching from my phone and didn't notice the neutral dashline line is actually above the setpoint. I thought setpoint was at 0, but turned out setpoint is at a constant negative value. Now it's clear. Thanks for the reply!
Hello UAV Tech, this video is really good. Simple and full off examples. You are talking about BouceBacks, Propwash etc, but what about “jerky movements” during the straight flight?? How to deal with this?
This is good stuff. I realised I had both the PD ballance and the PD gain sliders increased to 1.1 and 1.3 respectively, and experienced quite a lot of propwash. The quad also seems to have mid throttle oscillations when I fly upwards, it's very visible in the gopro footage as jello. Could it be caused by too high P and D? RPM filters slider set to 2.0.
10:00 From what i have seen and from my personal experience, other firmwares (FL1 and KISS) handle better prop wash. So based on your recent videos about kiss what could make kiss handle better prop wash?
People don't push BF enough. They get stuck on defaults or don't push up P and D gain enough or reduce filter enough. KISS and F1 have low filtering. Turn off all LPFs on BF for gyro and just have only Notches on Gyro and lowpasses on D-term. Also people are not comparing on the same rig., so it could just easily be mechanical differences. Folks spending more $ to build their first special KISS or F1 rigs. In my comp of KISS vs. BF the wash performance seemed the same. KISS and F1 work exactly the same as described in this video.
In Bf slider PD balance they add together all pitch, roll and yaw.. And all using same pD ratio on all axis? Some long deck frame like alien, actually need higer value on pitch right? It also means the PD ratio on pitch may differ from roll and yaw right?
Yes. They don't add though; the sliders that is. Each axis has a default ration between the axis. They hold that. But on quads far off from a 5", you many need to deviate. Sliders only keep it basic. If you want to tune per axis, you have to break away and just.put in the numbers. Same principals though.
No ideally. They DO go the same direction for any push on the quad. So that's why we want to have them push hard (up P&D Gain slider) to counter any of those forces.
Did i understand correct that for stick input the ration between p and d is the most important? So it doesn’t matter if it’s 30:40 or 60:80. But for external influences the higher the numbers with that ratio the better. Correct?
Great video. How do you know when you've gone to far with the PD gain? Does having the motor traces at 100% at the end of sharp moves for extended periods of time indicate a too high PD gain?
Hey brother thank you for shedding some light! I have a question, in previous videos you started the tuning proces by setting the D term somewhere between 30 and 40 then start moving with P term until you reached the desired balance. With the Betaflight 4.2 slides, it seems you just move the PD balance slider to increase the P term but at the same time the D term decreases. If you keep moving the slider there's a point where D becomes really low. Does it matter if the D term gets pretty low? And in that case should I use the master multiplier to bring it back up?
So, the one thing I don't get is, why wouldn't you just max out pd gain after dialing in the balance? what is the downside to "more p and d"?. Is it just the thermal limits of the motors or something? when will it poorly affect flight performance?
If you keep upping P & D gain, you eventually start to get D-term oscillations. That's you limit then and you need to back down a couple notches. See th-cam.com/video/hgLUX5uWaNI/w-d-xo.html
Hi man, I was watching your videos becuase my quad does a weird thing were is flies perfect but in the hd camera there is a slight vibration sometimes when im just cruising
What props? Try some HQ tri-blades if not already on them. If it does it on default filters and PIDs, it is probably a mechanical issue. You can always try my presets too though just to check. If it does on defaults or my presets, look for mechanical / electrical. Log with gyro scaled and reach out to FB group or Discord for some advice.
@@uavtech I fly hq props lmao, I love them, Thanks for the advice, I checked and it seems than one motor works waaaay better than the other three, other than that I havent found anything else, I dunno if that can unbalance the quad or something
@@uavtech thanks for the response and appreciate your helpful videos . One last question, in your website you have presets for all type of quads except the 3 inch category? Can I use the toothpick presets for a 3 inch quad or the 5inch presets? Again thanks a lot
Got my new whoop board today WITH blackbox, was looking forward to tuning it but blackbox does not show my traces :( I Guess i'll have to wait for new firmware.
@@uavtech I think its because i flashed it 4.2 and there is no firmware for it yet, Its the new Jhemcu 411 6s with blackbox. It shows the motors on the pic quad in corner but nothing else. am using gyro scaled and tried it set raw but nothing different.
@@Rich2brokeN could be. But it's also in all the quads I help tune... The base tune I give people in uavtech discord which works 99% of the time is 1.3 pd balance.
UAV tech, without you would be devastating to the fpv hobby. Your more important than Bardwell. Thanks so much because tuning is truly important
I watched more than 10 PID tuning videos in the last weeks, but non of them explain the goal this clear. Thank you!
One of the best explanations of PD balance, I'm finally starting to understand.
Man I've learned PID in school and have been exposed to it constantly over the years, but I've never understood it better than after watching this. THANK YOU
I hope i can eventually get my brain to process all this where it alows me to make the proper adjustments to the pid loop. Its all a geuss at this point for me still. It has got to be frustrating for you to keep explaining this to us!! Thats why your an engineer and i just fall in line!! Im so glad that Betaflight mostly gets it right or i would have quit long time ago. Thank you Mark!!
I watched sooo many videos including yours and they are all soooo good. But I am stupid...really stupid.... this video at 3:30: this finally made me ACUTALLY UNDERSTAND the P-D-Balance relationship. Thank you Mark :)
Such a good tutorial. I like these hyper focused and precise videos.
This is a really good video.
I'm not sure about the background music, but it was a clear and simple explanation
for the first time the interplay of P I D makes sense..thank you
This is video I was waiting for!!! thanks :)
I exactly fix these kind of problems this morning and what you say confirms what i was trying so i'm happy to see you talk about that :)
Nice vid!!
excellent vid! I love how you always point out new ways to look at PID control. Can never get enough of this. You're the only TH-cam guy talkin this stuff nowadays. Thanks and keep it up please!
Very nice, simple and clear explanation! Thanks!
I suddenly got this NEED to go tune my copters, even though they fly reasonably fine. :-) To bad it's blowing 30 knots. Good content, as always. Thanks!
Well, I think I am starting to finally understand a little of what P term and D term is all about in apposing each other to achieve that balance.
Terrific explanation. Really helping me to understand PID tuning. My key takeaway is that the P&D terms work against each other during stick moves, but they work with each other during wind/external disturbances. I do wonder though whether PID is basically completely the wrong approach for RC flying.
Great. That is the key take away.
Lots of smart professionals in the hobby. All firmware uses a PID controller. Know of anything better?
Thanks so much for these videos, the sliders in betaflight are awesome, at first I didn’t use them but they let you dial in 95% of the changes needed in a couple packs.
I always use RPM filtering now and adjust my P/D balance slider until I get a tiny bit of bounce back then raise it back up a notch. Then adjust P/D gains to make the quad feel more or less locked in. Then I leave PIDs alone and raise my filter sliders together while checking for hot motors, this usually makes propwash handling much better.
Then I stop tuning and fly, I think this is where a lot of people come unstuck, unless you’re logging continuing to bring out problems to fix is a recipe for chasing your tail. Instead I just fly and make small tweaks if I notice a problem recurring, I might adjust antigravity or raise P term on yaw or I terms (I move away from sliders at this point). I also spend a lot more time flying rather than tuning this way too 👍
Very informative video. Thanks for explanation. I probably will still struggle a bit but it made it more clear.
Great video! thanks for taking time to explain that in detail with good example and clear evidence ie. blackbox.
Very helpful explanation. Thanks 🤓
Aarrrright, nice and smooth! 👌🏻
Many thanks for this tutorial! Would be nice to find out more about the 'stick response gain' please
th-cam.com/video/66svCI4jV34/w-d-xo.html
@@uavtech thank you ✌🏻☺️
Fantastic explanation, thanks for doing this video!
Very interesting! Great video !
Thanks again, Mark. I get 3 quads now tuned how you described it in your videos and it's pure joy to fly them. But there are difficulties when it comes to toothpicks or other micro quads with no blackbox. Additionally some AIO FC's have much electrical noise on their supply rails that affects the gyro signal. May be you can make a video about toothpick tuning?
Already done.
Great and simple tutorial!
Great video! One question on the P term trace at time 4:07, why p is always positive when gyro signal bounce on both sides of the setpoint? I thought P just tracks error with a constant multipler, so it should map the polarity of the error as well?
P is the opposite sign of PID Error - proportional in shape - times a gain strength (p-gain). At that time stamp, P is in both signs from what I see. What am I missing?
@@uavtech Oh, sorry the time should be 8:39. I was watching from my phone and didn't notice the neutral dashline line is actually above the setpoint. I thought setpoint was at 0, but turned out setpoint is at a constant negative value. Now it's clear. Thanks for the reply!
Hello UAV Tech, this video is really good. Simple and full off examples. You are talking about BouceBacks, Propwash etc, but what about “jerky movements” during the straight flight?? How to deal with this?
That can be a lot of thing. First step is to tune it in general.
This is good stuff. I realised I had both the PD ballance and the PD gain sliders increased to 1.1 and 1.3 respectively, and experienced quite a lot of propwash. The quad also seems to have mid throttle oscillations when I fly upwards, it's very visible in the gopro footage as jello. Could it be caused by too high P and D? RPM filters slider set to 2.0.
10:00 From what i have seen and from my personal experience, other firmwares (FL1 and KISS) handle better prop wash. So based on your recent videos about kiss what could make kiss handle better prop wash?
People don't push BF enough. They get stuck on defaults or don't push up P and D gain enough or reduce filter enough.
KISS and F1 have low filtering. Turn off all LPFs on BF for gyro and just have only Notches on Gyro and lowpasses on D-term.
Also people are not comparing on the same rig., so it could just easily be mechanical differences. Folks spending more $ to build their first special KISS or F1 rigs.
In my comp of KISS vs. BF the wash performance seemed the same.
KISS and F1 work exactly the same as described in this video.
@@uavtech thanks!!
Thank you for this tutorial
In Bf slider PD balance they add together all pitch, roll and yaw.. And all using same pD ratio on all axis? Some long deck frame like alien, actually need higer value on pitch right? It also means the PD ratio on pitch may differ from roll and yaw right?
Yes. They don't add though; the sliders that is. Each axis has a default ration between the axis. They hold that. But on quads far off from a 5", you many need to deviate. Sliders only keep it basic. If you want to tune per axis, you have to break away and just.put in the numbers. Same principals though.
Love the content... not sure about the background tunes.
I need to find something softer that doesn't build.
I learned something today.
Lol.. really? :X
@@Krunked The propwash bit, I never realised about the fact that the P and the D ideally need to go the same direction to fight propwash.
No ideally. They DO go the same direction for any push on the quad. So that's why we want to have them push hard (up P&D Gain slider) to counter any of those forces.
Did i understand correct that for stick input the ration between p and d is the most important? So it doesn’t matter if it’s 30:40 or 60:80. But for external influences the higher the numbers with that ratio the better. Correct?
I should really learn black box. I tune everything by ear.
Great video. How do you know when you've gone to far with the PD gain? Does having the motor traces at 100% at the end of sharp moves for extended periods of time indicate a too high PD gain?
You will hear a D-term oscillation sound. See th-cam.com/video/hgLUX5uWaNI/w-d-xo.html
@@uavtech Thanks
Hey brother thank you for shedding some light! I have a question, in previous videos you started the tuning proces by setting the D term somewhere between 30 and 40 then start moving with P term until you reached the desired balance. With the Betaflight 4.2 slides, it seems you just move the PD balance slider to increase the P term but at the same time the D term decreases. If you keep moving the slider there's a point where D becomes really low. Does it matter if the D term gets pretty low? And in that case should I use the master multiplier to bring it back up?
Use PD Gain to bring it back up.
It does matter.
So, the one thing I don't get is, why wouldn't you just max out pd gain after dialing in the balance? what is the downside to "more p and d"?. Is it just the thermal limits of the motors or something? when will it poorly affect flight performance?
If you keep upping P & D gain, you eventually start to get D-term oscillations. That's you limit then and you need to back down a couple notches. See th-cam.com/video/hgLUX5uWaNI/w-d-xo.html
Very very interesting 👍👍👍
Thank you!
thank you.
Hi man, I was watching your videos becuase my quad does a weird thing were is flies perfect but in the hd camera there is a slight vibration sometimes when im just cruising
What props? Try some HQ tri-blades if not already on them.
If it does it on default filters and PIDs, it is probably a mechanical issue. You can always try my presets too though just to check. If it does on defaults or my presets, look for mechanical / electrical. Log with gyro scaled and reach out to FB group or Discord for some advice.
@@uavtech I fly hq props lmao, I love them, Thanks for the advice, I checked and it seems than one motor works waaaay better than the other three, other than that I havent found anything else, I dunno if that can unbalance the quad or something
You are my hero
always wondered why this cant be autotuned, firmware has all data it needs to try to tune itself :/, something like INAV does for wings
do you have a writen tuning guide some thing i can print off to take in the field with me? kind of like a check list and in order of what to do? :-D
tiny.cc/pidtuning
@@uavtech awesome, thanks dude
very helpful 🙏👍🏻👌
What about motor heating due to high p and d gain? Do you have to lower it? And if you did lower it do you lower the PD Balance in parallel ?
No, just P and D gain. I would add filtering vs. lowering P and D Gain slider for hot motors.
@@uavtech thanks for the response and appreciate your helpful videos . One last question, in your website you have presets for all type of quads except the 3 inch category? Can I use the toothpick presets for a 3 inch quad or the 5inch presets? Again thanks a lot
how about the drone shake sometime? its the problem in PD? my quad have shake, until now i cnt fix it, to smoth fly again..
Again?
Defaults, preset or tuned? If defaults, try a preset at theuavtech.com/presets
thanks sir, i will tune back again...😀
Great infos 👍 thxxxx a Lot 💚 lg
Got my new whoop board today WITH blackbox, was looking forward to tuning it but blackbox does not show my traces :(
I Guess i'll have to wait for new firmware.
It should work.
@@uavtech I think its because i flashed it 4.2 and there is no firmware for it yet, Its the new Jhemcu 411 6s with blackbox. It shows the motors on the pic quad in corner but nothing else. am using gyro scaled and tried it set raw but nothing different.
Hi. Is it dangerous (for the motors) to push the P and D gain too high ?
Yes. Once you hear D-term oscillation, back off. See th-cam.com/video/hgLUX5uWaNI/w-d-xo.html
@@uavtech thank you 😉
GJ
Is it normal to have D ≈ P on 4s 5 inch quads?
No. P is usually 20 points higher.
@@uavtech setting it like that I always get bounce back. Where should I start looking at?
You typically need more D-term if the quad has a larger moment of inertia or weaker motors.
I would just.tune for no bounceback.
Why on any quad I own is that pd balance always over 1.3.. I've never had a quad under 1.0 pd balance.
lower power to weight? 48k?
@@uavtech maybe. Just interesting. I haven't been able to get any quads under 1.0 and fly without bounceback.
maybe because of all your kwads being basically the same rotational mass
@@Rich2brokeN could be. But it's also in all the quads I help tune... The base tune I give people in uavtech discord which works 99% of the time is 1.3 pd balance.
@@uavtech I presume 48k less powerful then? Would you still suggest 24K pwm?
👍👀🇭🇷
First?
Appreciate the content!
Not a fan of the background music.
This is such a good explanation. Thanks!
Thank you!