D is watching sudden change of position within the fraction time to give feedback. It does’t matter where the target is. So .. 1. Set P for the smooth and sluggish curve. 2. Set I to help add power to P if it is taking too much time to get to the target. 3. Set D to push the power back if I is doing too much work and excessive acceleration. P I’ll try to get to the target. I I’ll give you extra power because you are taking too much time. D I’ll slow you down because I is giving too much power and detected sudden movement in short period of time.
After watching countless PID videos to try and get understanding of PID separately, I can honestly say that your video was by far the most straight forward, easy to understand and straight to the point. Simplicity is key and you did that super well! Thank you so much!
P term is a rubber band with a resting length of zero. It needs distance to produce tension. If you increase the load on a rubber band, it needs to stretch further in order to produce the correct tension. This is insufficient for a regulator since you expect a regulator to compensate. The I term naturally complements the shortcoming of a rubber band. And since both pull towards the setpoint, you need the D term to compensate to avoid overshoot.
best discription i ever head out off all the videos i ever seen on tuning. i wii finally be able to tune my quad.. thanks a lot for breaing it down in a way anyone can understand and not just making stuff up....
This is just about the best explanation of PIDs ever! I've watched a lot of videos on the topic including Joshua Bardwell's PID tuning masterclass and as much as I appreciate his practical approach, You've just nailed it in theory! Understanding Your video gives one a clear view on the Anti-Gravity and Feed Forward in Betaflight. Thanks for allowing me to grasp the whole thing. Awesome job! Subscribing for more!
Thank you for making me understand the use of each one of the three terms. Before this explanations, everything was like XYZ in mathematics, but from now on i will have in mind what PID means when calculating. It will make a big difference, thanks to you.
I have to write and say thank you. I built a PID box to control oven temps and I have been screwing with it all day. I knew what I wanted it to do but I just didn't know how to get there. Your explanation of D solved my problems. I was getting a 20 degree over and 10 degree under oscillation. Now I am getting a .1 to .3 degree over oscillation and thats it. It holds at exactly 167.0 to 167.3. Perfect for killing covid-19 on masks gloves whatever. I am using 2 oven thermometers to verify the temp and it is dead nuts. Again thanks so much I hadn't heard D explained that way before and it all clicked when you said it. TAS
Awesome video this has already help me in this Awesome hobby/sport you are a excellent teacher this video will help every beginner like myself that’s intimidated by tuning cause some people in the sport are so advanced that they are far removed from knowing and understanding how a beginner thinks 🤔. Thanks so much this has shaved months maybe longer off understanding Fpv in its entirety 😃👍
One the best explanation I’ve come across watching quite a lot of PID vids. Still trial and error in tuning goes a long way. Especially if you got a very good updated trend to look at on a short time scale like inn15-30 min span or so. That allows you to see how things are playing out on a trend in adjustment as you try to smooth things out.
This makes more sense to me than any other PID explanation video I've seen, so thank you. I've been flying for about a year and have never really been able to visualize what each term does, and I just built a quad that is giving me more trouble to tune than my other two quads ever did, so I'm here to learn!
As an electrician, I sometimes had to tune PID for process temp. control. I was able to muddle through it but it was sometimes frustrating. Wish I could have viewed your channel years ago. Thankfully, most micro controllers now have auto-tune feature. Turn it on and let it do its thing. Your video can be applied to many different control application. Thanks for sharing.
I'm fixing an oven in our laboratory and it has a pid controller I tried to autone and offset the controller the pv and the sv is already equal unfortunately the actual temperature in the inside is very high
I'm surprised that I am just finding your channel... Absolutely worth the sub! Thanks for posting; I'll be checking out other content on your channel...!
Why do we use a PID controller if the only thing that it does is to follow the input. Why dont we directly fed our input to the system instead of passing it through PID.
Why is it that no one else could simplify what's going on. I've watched many many videos and read lots and I could never retain what was going on. You made this so simple to understand.
Have you ever tried zeroing out the kterm (the accumulating variable) when crossing the setpoint? I noticed it's always retaining it's value when overshooting, when it should really help to go the opposite direction.
Referring to the beginning of the explanation: P term, as i understand it. does not slow down or speed up. That is derivative. P term also does not move back and fourth. That is Integral. The P term tries to move to the target, but because of physics, it ends up being in the wrong place. It outputs proportional to the input. If the p is 2/1, it outputs 2 for every 1. It is one number, the desired destination. Integral and Derivative correct the errors.
Dude, i was here, because to know the meaning PID for tuning a stand alone ecu. But somehow i think its basicly the same and you explained it very good😅
I had a foreign professor teach a class on Control Systems. These concepts didn't make sense until I started manipulating objects in Godot. My cube is orbiting my target point.
Why do we use a PID controller if the only thing that it does is to follow the input. Why dont we directly fed our input to the system instead of passing it through PID.
I just don't understand. Whenever I see other examples, it shows a high rate of change (approaching set point quick, say a PV of 40degress and set point of 50 degrees)giving D a high value. If we are approaching a set point and want to avoid overshoot, is the D value inverse? So the D value gives a minus value to the PID controller to lower the power to avoid overshoot?
9:05 Just gonna let you know those arrows are backwards in size, as it curves away the arrows are bigger but in the center of the curve they're smaller because less change.
Awesome explanation. Your simple, colored lines and dots were a great visualization of your explanation. Looking at your dots made me think - hmmm....looks like iteration steps over time......which brings my question: How many times does this process iterate in a single second? I know my loop is set to 8kHz. Is that 8000 times a second? Or is it a higher magnitude, like 8 million? Thanks for the great video. 👍
I am not sure of the D term explanation. D term is base of the error slope (Target - current position), not calculated according to how far we are from setpoint. Anyway, I love yours videos, thanks a lot !
Why do we use a PID controller if the only thing that it does is to follow the input. Why dont we directly fed our input to the system instead of passing it through PID.
So, P = amount of correction signal, proportional to the deviation from set-point, and D= response rate of the 'P' signal (to prevent time delays by preventing any overshoots or undershoots) is my understanding correct?
D is independent of P, they just happen to work together. D also doesn't do anything to prevent undershoots ( I is what provides a slow nudge when you're not resolving on your exact set point ) it fights overshoots by virtue of it providing a corrective force when the error is increasing, which is what happens when P overshoots it's target
if the pid controller has settled but not exactly at it's target D isn't really doing anything. because even though it's not at it's target, the error is not changing, so D sees nothing happening
I term slowly nudges the solution towards the target over time. With no I term an undershoot is not corrected. Technically not even always an undershoot, you can have a small overshoot as well. You can resolve anywhere very close to, but not exactly at, the target. I term finishes the job.
D is watching sudden change of position within the fraction time to give feedback. It does’t matter where the target is.
So ..
1. Set P for the smooth and sluggish curve.
2. Set I to help add power to P if it is taking too much time to get to the target.
3. Set D to push the power back if I is doing too much work and excessive acceleration.
P I’ll try to get to the target.
I I’ll give you extra power because you are taking too much time.
D I’ll slow you down because I is giving too much power and detected sudden movement in short period of time.
Thanks for the brilliant comment!
This was so much better of an explanation of what each one does.
This made it really clear for me, thank you!
Nice breakdown!
Good explanation. Thx
After watching countless PID videos to try and get understanding of PID separately, I can honestly say that your video was by far the most straight forward, easy to understand and straight to the point. Simplicity is key and you did that super well! Thank you so much!
You wrote it already what I wanted to write.
I've seen a lot of pid videos, but this graph does a great job demonstrating what is going on. The graph is very informative
Thank you!
P term is a rubber band with a resting length of zero. It needs distance to produce tension. If you increase the load on a rubber band, it needs to stretch further in order to produce the correct tension. This is insufficient for a regulator since you expect a regulator to compensate. The I term naturally complements the shortcoming of a rubber band. And since both pull towards the setpoint, you need the D term to compensate to avoid overshoot.
best discription i ever head out off all the videos i ever seen on tuning. i wii finally be able to tune my quad.. thanks a lot for breaing it down in a way anyone can understand and not just making stuff up....
The best explanation of quad tuning I have seen thus far... You are the Khan Academy of quad-copters... Keep up the good work.
That’s a really great presentation, now I got some basic understanding of PID control! Thx a lot!
Great job man, simplified PID explanation.
Just what I wanted.
this is the best theoretical
explanation so far
Excellent way of explaining how PID effect the performance!
This is just about the best explanation of PIDs ever! I've watched a lot of videos on the topic including Joshua Bardwell's PID tuning masterclass and as much as I appreciate his practical approach, You've just nailed it in theory! Understanding Your video gives one a clear view on the Anti-Gravity and Feed Forward in Betaflight. Thanks for allowing me to grasp the whole thing. Awesome job! Subscribing for more!
Thank you for making me understand the use of each one of the three terms. Before this explanations, everything was like XYZ in mathematics, but from now on i will have in mind what PID means when calculating. It will make a big difference, thanks to you.
Kudos on his ingenuity!
Thx a lot!!! :-)
Your lecture helps me to survive final exam.
Very well explained ! it helps controling compressor speed for AC loop
I have to write and say thank you. I built a PID box to control oven temps and I have been screwing with it all day. I knew what I wanted it to do but I just didn't know how to get there. Your explanation of D solved my problems. I was getting a 20 degree over and 10 degree under oscillation. Now I am getting a .1 to .3 degree over oscillation and thats it. It holds at exactly 167.0 to 167.3. Perfect for killing covid-19 on masks gloves whatever. I am using 2 oven thermometers to verify the temp and it is dead nuts. Again thanks so much I hadn't heard D explained that way before and it all clicked when you said it. TAS
You sir are a great teacher.
Awesome video this has already help me in this Awesome hobby/sport you are a excellent teacher this video will help every beginner like myself that’s intimidated by tuning cause some people in the sport are so advanced that they are far removed from knowing and understanding how a beginner thinks 🤔. Thanks so much this has shaved months maybe longer off understanding Fpv in its entirety 😃👍
Have been struggling to understand PIDs for ages. This video helped a lot. Thank you. :)
This is the best explanation I have ever encountered, thank you!
your explanation about PID basic concept for drone is the best... easy to understand and simple....
This is a really nice look at PIDs. Good work.
Great explanation on PID's physical meaning!
Very good explanation dude! Now I know all I need to know about PID!
OMG finally someone to dumb down PIDs for me! I have been searching forever! Thank you soooooo much!!
One the best explanation I’ve come across watching quite a lot of PID vids. Still trial and error in tuning goes a long way. Especially if you got a very good updated trend to look at on a short time scale like inn15-30 min span or so. That allows you to see how things are playing out on a trend in adjustment as you try to smooth things out.
This makes more sense to me than any other PID explanation video I've seen, so thank you. I've been flying for about a year and have never really been able to visualize what each term does, and I just built a quad that is giving me more trouble to tune than my other two quads ever did, so I'm here to learn!
As an electrician, I sometimes had to tune PID for process temp. control. I was able to muddle through it but it was sometimes frustrating. Wish I could have viewed your channel years ago. Thankfully, most micro controllers now have auto-tune feature. Turn it on and let it do its thing. Your video can be applied to many different control application. Thanks for sharing.
What if auto tune doesn't fix the problem what would be the best way to fix it?
I'm fixing an oven in our laboratory and it has a pid controller I tried to autone and offset the controller the pv and the sv is already equal unfortunately the actual temperature in the inside is very high
OMG, YOU ARE THE BEST, I AM A BRAZILIAN STUDENT OF ELECTRICAL ENGINEERING AND YOU HELP ME SO MUCH, THANKS MAN
You've got a stable pulse on your mouse. Very helpful video. Thanks.
This video was so good and easy to understand!! Huge thanks!!
Straightforward and to the point. Thanks
That's the best explanation outhere! Thanks!!
This is a great explanation. Thanks for posting!
Perfect explanation! Thanks!
Really a great way you explained the most complicated part. Thanks
I'm surprised that I am just finding your channel... Absolutely worth the sub! Thanks for posting; I'll be checking out other content on your channel...!
This is the perfect visualization i needed. Thank you
Why do we use a PID controller if the only thing that it does is to follow the input. Why dont we directly fed our input to the system instead of passing it through PID.
Wow. that was the best explanation about PIDs ever!
Thank you for your explanation of pid I been trying to figure out how it works.
Best simplifying video on PID ever
Great video! When someone asks me about pid's I show them this video
Surprises me you don't have more subscribers, your channel is great. Keep it up!
Considering his video is completely wrong, I disagree. The comments, thankfully, saved me from this bad information.
Thank you, nice job.
Thank you for uploading. Well explained.
This is incredibly helpful thanks!
Thx. This is so easy to understand.
Why is it that no one else could simplify what's going on. I've watched many many videos and read lots and I could never retain what was going on. You made this so simple to understand.
Excellent video!
Great explanation. Kudos.
Very well explained in simple terms, well done, subbed you
Thanks for the simplified explanation!
Great explanation! 👏
the illustrations made it simple, I like it
great video, now I finally understand PIDs , thanks you!
Thanks, very helpful information
Have you ever tried zeroing out the kterm (the accumulating variable) when crossing the setpoint?
I noticed it's always retaining it's value when overshooting, when it should really help to go the opposite direction.
Referring to the beginning of the explanation: P term, as i understand it. does not slow down or speed up. That is derivative. P term also does not move back and fourth. That is Integral. The P term tries to move to the target, but because of physics, it ends up being in the wrong place. It outputs proportional to the input. If the p is 2/1, it outputs 2 for every 1. It is one number, the desired destination. Integral and Derivative correct the errors.
Awesome explanation!
Dude, i was here, because to know the meaning PID for tuning a stand alone ecu. But somehow i think its basicly the same and you explained it very good😅
Great explanation 🫡
really nice explanation. thank u
good explanation. Btw, what instrument do you use for drawing?! thanks
Well done 👍 thanks man
That's amazing
I had a foreign professor teach a class on Control Systems. These concepts didn't make sense until I started manipulating objects in Godot. My cube is orbiting my target point.
thank you for the wonderful video. Helped me understand better :)
Great explanation.Subscribing.
Great grea video and explanation thx
I got inspired to learn about these after watching the brick experiment channel build an automatic depth control Lego submarine.
Thanks for sharing
really insightful...
Awesome! Thanks a lot
You won the nerd contest. You are king nerd. Well done.
Thank you!
best explanation
Thanks man. I've now better understanding of PIDs.
Why do we use a PID controller if the only thing that it does is to follow the input. Why dont we directly fed our input to the system instead of passing it through PID.
really nice video
love yr class thank *1000 so much
I'm very surprised se how you chat with the control paraemters. That's kind of lovely relationship
Great writeup.
Is setpoint in blackbox explorer the target?
Thankyou SIR . Thankyou very much
I just don't understand. Whenever I see other examples, it shows a high rate of change (approaching set point quick, say a PV of 40degress and set point of 50 degrees)giving D a high value. If we are approaching a set point and want to avoid overshoot, is the D value inverse? So the D value gives a minus value to the PID controller to lower the power to avoid overshoot?
9:05 Just gonna let you know those arrows are backwards in size, as it curves away the arrows are bigger but in the center of the curve they're smaller because less change.
good catch, I'm drawing the value curve but thinking about the acceleration curve
wow, best explanation ever
did you see any code example like this explanation?, I mean basic codes
Awesome explanation. Your simple, colored lines and dots were a great visualization of your explanation. Looking at your dots made me think - hmmm....looks like iteration steps over time......which brings my question:
How many times does this process iterate in a single second? I know my loop is set to 8kHz. Is that 8000 times a second? Or is it a higher magnitude, like 8 million?
Thanks for the great video. 👍
Hz = per second, kilo = magnitude of 1000, so indeed 8000 times per second ;)
Is it possible to create a PID controler that won't allow overshoot like ' You need to stop this car right before it falls off a cliff '
This is knowledge mine.
Found really helpful
Thank you
nice job
Good video
.kool way to look at P.I.Ds
I am not sure of the D term explanation.
D term is base of the error slope (Target - current position), not calculated according to how far we are from setpoint.
Anyway, I love yours videos, thanks a lot !
Agree with you on both terms, the D term and the video as well.
+1👍 💪 best one, i refer to it often
Why do we use a PID controller if the only thing that it does is to follow the input. Why dont we directly fed our input to the system instead of passing it through PID.
Thanks a lot
Thx!
So,
P = amount of correction signal, proportional to the deviation from set-point, and
D= response rate of the 'P' signal (to prevent time delays by preventing any overshoots or undershoots)
is my understanding correct?
D is independent of P, they just happen to work together. D also doesn't do anything to prevent undershoots ( I is what provides a slow nudge when you're not resolving on your exact set point ) it fights overshoots by virtue of it providing a corrective force when the error is increasing, which is what happens when P overshoots it's target
@@SiieeFPV so when I is slow to react, exactly what role does D play in that situation?
if the pid controller has settled but not exactly at it's target D isn't really doing anything. because even though it's not at it's target, the error is not changing, so D sees nothing happening
@@SiieeFPV I see. So, when P is small and D plays no role (as it sees no error as you say), how exactly then an undershoot is corrected?
I term slowly nudges the solution towards the target over time. With no I term an undershoot is not corrected. Technically not even always an undershoot, you can have a small overshoot as well. You can resolve anywhere very close to, but not exactly at, the target. I term finishes the job.
What program did you use to draw this PID drawing explanation?
I just screen capped live drawing in TVPaint, it's 2d animation software, but it could just as easily be done in Photoshop or Krita
well explained