Bigger props resulting in a lot more instability?

[Winny the pooh]

Hey all-

I recently upgraded my 10x4.5 props to 11x4.7 props and my quadcopter has become a lot more unstable because of it. With the 10" (my first ever props) I barely had to do any tuning to keep the copter stable, but with the 11" APC Slow Flyers for 3DR I had to reduce the PID sensitivity all the way down to its lowest value (using a pixhawk). It helped a bit, but the copter is still very jerky and oversensitive. Especially when descending into it's own downdraft. I was never worried about vortex ring state before, but now I have to punch the throttle to the max just to get it to stabilize after a banked descent.

Anyone have any ideas on why this is happening? I thought that bigger props would equate to more stability, not less?

thanks!

 

[Av8Chuck]

There are other variables that effect stability other than prop size. What motors are you using? Generally changing from 10" to 11" props won't have that big of a difference, however depending on the KV of your motors, sometimes its necessary with higher 770+KV motors to reduce the voltage by going from 4S to 3S when you go up in prop size. Also when going up in prop size the MR usually gets less responsive and you need to increase your gains.

Bigger props means more thrust, not necessarily more stability. It depends on the battery and motors your using. Do your motors and ESC's run hotter?

 

[Winny the pooh]

There are other variables that effect stability other than prop size. What motors are you using? Generally changing from 10" to 11" props won't have that big of a difference, however depending on the KV of your motors, sometimes its necessary with higher 770+KV motors to reduce the voltage by going from 4S to 3S when you go up in prop size. Also when going up in prop size the MR usually gets less responsive and you need to increase your gains.

Bigger props means more thrust, not necessarily more stability. It depends on the battery and motors your using. Do your motors and ESC's run hotter?

I'm using 1000 KV motors with a 3S battery. That seems like a high KV no? I didn't feel my motors after my flight this morning unfortunately. My ESCs did feel kinda hot though...

 

[Bartman]

all of these flight controllers seem to have a sweet spot around 55 to 70% throttle where they fly the best. everything together makes up the success of the system so if the props have the ability to make a lot of excess lift then that has to be constrained by the tuning of the helicopter or it creates instability. If the flight control is overly sensitive then a light and floaty heli will lead to instability. in the 55 to 70% throttle range things just seem to fall into place better and the FC's we use seem to be better able to balance control, power, resisting outside forces like wind, etc.

you said it was stable with the 10" props so why did you change them?

 

[Winny the pooh]

you said it was stable with the 10" props so why did you change them?

My old props were getting a bit beat up and I wanted to increase the flight time. I guess I didn't appreciate the complexity and interconnectedness of thrust and stability..

I want to get a bit more flight time out of the my MR, and also start figuring really how different voltages, prop size, KVs.. etc affect flight characteristics- do you know of any good resources about this? Could I expect to see that much of an improvement with 11" props, and would that affect other flight characteristics?

cheers

 

[Rotorfreek]

I'll try and give you a quick 101 class on how rotors affect stability and thrust...

Multirotors work very differently to helis in that we don't use blade pitch to alter lift, we use rotor speed. Therefore, unlike a helicopter, we are constantly accelerating our rotors. This happens every time you make a sick input, or automatically by the flight controller to maintain stability.

Therefor inertia is critical to us. Inertia is like rotational momentum - it resists changing angular velocity. The more inertia a rotor blade has, the more power it will take to change its speed quickly. Now bigger blades have more inertia. Depending on the design, it can be a lot or a little. The more blade thickness or cord width there is towards the tip, the more inertia they will have.

Now your gain settings tell your FC how much power to use in response to instability. Higher gains send more power and vice versa. There is a minimum response speed that results in acceptable stability and control. With a given motor power, there is a maximum rotor acceleration that can be achieved - regardless of your settings. If this is less that the critical rotor reaction that results in stable flight then there is nothing you can do to fix it.

Now P = T x V were in this case T is motor torque and V is angular velocity, or motor rpm. Every motor has a max power output and a kV rating. Take 2 motors of the same power, one 500 kV and the other 1000kv. If you run both at maximum power, one will spin at double the speed of the other. Since both have the same power, the faster one has half the torque.

Now bigger props produce more thrust per rpm, so run slower that a smaller prop on the same model. So if your motors are capable of spinning (with your battery choice) faster that the props max rated speed, or its peak efficiency speed then there is a whole section of your motor that is worthless and all your doing is reducing your avaliable torque (and therefore response speed, therefore stability) in the section you are using. So what you should really do is match your motors to your props.

For 1000 kV, I reccomended 9-10". If you want more flight time, you don't need bigger props, you need more efficient, better designed props. try something like a DJI 9443 NST.

This is why you use big props with slow motors etc. It's also why aerobatics are done with smaller props - they respond faster. Now bigger can be more efficient, provided you have the power to accelerate them to gain stability and keep the tips an inch apart.

Any questions you know where I am!



Sent by my thumb, a trusty slave to my crazy mind. The above is the ramblings of a lunatic, and should be treated as such. Terms and conditions apply.

 

[Bartman]

Nicely done R-freek

The only thing I'd add is that props don't keep making more thrust for every increase in RPM, at some point they start to make less and less lift for each increase in RPM. For a given RPM change they will make more lift at lower RPM's than they will at higher RPM's. Spin them too fast and they won't make any more lift, more RPM's equals zero lift gain.

At lower RPM's where the increase in lift will be larger for a given RPM bump, the FC's can sometimes be tuned to handle it but that generous availability of lift seems to be too much for them to handle and they will be very sensitive to things like wind and changes in air density. It's like hammering the gas pedal for a second when you're in second gear, the car will lurch.....but.....

At higher RPM's you're props are working harder and things like wind blowing over them will have less of an effect. The motor can hammer the RPM's like your driver with a heavy foot and the response will be more subdued. This is where our flight controllers seem to be happier, they have the right amount of lift available for control but not so much that they're constantly trying to undo what they just did.

DJI flight controller especially seem to be happier in the 60 to 65% throttle range for hovering. Hovering in the 45 to 55% range makes them very floaty feeling and unstable on windy days, IMHO.

 

[R_Lefebvre]

Hey all-

I recently upgraded my 10x4.5 props to 11x4.7 props and my quadcopter has become a lot more unstable because of it. With the 10" (my first ever props) I barely had to do any tuning to keep the copter stable, but with the 11" APC Slow Flyers for 3DR I had to reduce the PID sensitivity all the way down to its lowest value (using a pixhawk). It helped a bit, but the copter is still very jerky and oversensitive. Especially when descending into it's own downdraft. I was never worried about vortex ring state before, but now I have to punch the throttle to the max just to get it to stabilize after a banked descent.

Anyone have any ideas on why this is happening? I thought that bigger props would equate to more stability, not less?

thanks!

Personally, I think you're on the wrong track. You should be able to go from a 10" to 11" prop without much trouble. You say it's "jerky and oversensitive", which is not the same thing as unstable. Are you saying that it's response to inputs is too "sharp". Or that it's unstable in wind? If you reduced PIDS to min, I would think you've gone the wrong way.

 

[Bartman]

yet another consideration is prop loading, known to full scale aircraft as wing loading.

a wing that is working hard to carry the aircraft will not have a lot of excess lift available and will plod through the air with less sensitivity to gusts and small pitch deviations. a very lightly loaded wing will have a lot of excess lift available and will dip and climb at every little gust making for a bumpy ride.

props work in similar ways as they are also airfoils. props that are working harder will plow through turbulence without allowing what the pilot will perceive as instability (movement of the helicopter away from level).

switching from a narrow, thin, heavily loaded ten inch prop to a wider chord, thicker, airfoiled, 11 inch prop might give enough of an increase in lift to make the helicopter bad mannered and the flight control may or may not have the flexibility to deal with it. MK seems to take these things in stride, DJI not so much. Different systems will adjust with more or less success.

also, tuning is usually appropriate for a narrow band of operation that is a little above and below hovering power. reduced throttle descents for example are often unstable for a number of reasons but if you wanted to tune to that flight regime then other things like a neutral hover would probably suffer.

food for thought.
bart

 

[bensid54]

Interesting read and everyone is right to a major degree. Bart you mentioned bigger blades and after so many RPM don't make a difference in thrust well your right as far as the flight controller is concerned but although it may be minimized and not nearly what they are at lower RPM they still do increase thrust. Keep going guys this is good and in one post by Rotorfreek you said "sick" input does that hard and quick stick inputs on your transmitter controls?

 

[Rotorfreek]

Yes sorry it should have been "stick inputs", so control inputs in your transmitter.

Sent by my thumb, a trusty slave to my crazy mind. The above is the ramblings of a lunatic, and should be treated as such. Terms and conditions apply.

 

[R_Lefebvre]

They I suggest you auto-tune it to get the controls working properly. And use Arducopter 3.2 with the new Acceleration Damping Parameters to make it fly smoother. Detuning PIDS is not what you want to do.

 

[Bartman]

Hey all-

I recently upgraded my 10x4.5 props to 11x4.7 props and my quadcopter has become a lot more unstable because of it. With the 10" (my first ever props) I barely had to do any tuning to keep the copter stable, but with the 11" APC Slow Flyers for 3DR I had to reduce the PID sensitivity all the way down to its lowest value (using a pixhawk). It helped a bit, but the copter is still very jerky and oversensitive. Especially when descending into it's own downdraft. I was never worried about vortex ring state before, but now I have to punch the throttle to the max just to get it to stabilize after a banked descent.

Anyone have any ideas on why this is happening? I thought that bigger props would equate to more stability, not less?

thanks!

Winny,

What model/brand were the 10" props?

 

[Scott Drysdale]

Bartman and R_Lefebvre are correct.......

Put another way.....
FC leveling control relies on the motors capability to respond instantaneously to complete PID loop feedback correction.
If the motor torque is adequate for the size of props, there should be little or no delay with motor speed changes and therefore excellent MR stability control.
If the motor torque is less than adequate, speed changes are too sluggish for good PID leveling control.

Contributing to the challenge is disc loading which is very similar to wing loading.
A super light aircraft with large wing area is more at the mercy of wind disturbance than is a heavy jet with stubby wings.
Likewise over sized props on an Multirotor make the MR more susceptible to rough air.

eCalc can be used to optimize airtime but cannot determine the limits of MR airframe stability relative to prop size/pitch.

 

[Scott Drysdale]

Are the following statements correct?

Factors affecting the ability of the Flight controller to maintain horizontal stability include:

Angular momentum = motor arm length + motor weight + vertical rotational speed...... http://en.wikipedia.org/wiki/Angular_momentum
Motor torque needed to overcome angular momentum as well as prop drag closely associated with (prop pitch and diameter)
Disc loading which is similar to wing loading such that MR stability can be exacerbated by rough air conditions.

eCalc can be used to help optimize airtime but can ultimately lead towards MR airframe instability.

 

[Winny the pooh]

Wow thanks for all the good info everyone! As a junior year mechanical engineer, the technical stuff like this is what I really love to sink my teeth into.

Winny,

What model/brand were the 10" props?

My 10" props were just cheap one from Hobby King. Slow Fly I believe..

 

[gtranquilla]

Both Mechanical Engineers and Electrical Engineers have to use calculus on an ongoing basis...... but almost predominantly PID loop control......and now all the modern day hobbyists and robotic systems developers.

originally conceived by Sir Isaac Newton in his book - Principa Mathmatica...... very interesting really......but not fast enough for: http://www.lockheedmartin.ca/us/100years/stories/missile-defense.html

Good stuff here.....
http://en.wikipedia.org/wiki/PID_controller

 

[gtranquilla]

Yes...... Mech and Electrical Engineers cannot get far without the calculus and PID loop control is the essence of calculus.

The higher the motor torque capability relative to the load, the faster the motor can change to the target speed.
But even PID loop controllers have finite time limitations so something better had to be developed for:
http://www.lockheedmartin.ca/us/100years/stories/missile-defense.html

Good stuff here.....
http://en.wikipedia.org/wiki/PID_controller

 

[Bartman]

anyone else notice that the site seems to be getting more and more blinky by the day? I'm not sure why but we're going offline Friday night at midnight and will return by late Sunday evening with an entirely new site which will be running the very awesome Xenforo forums software. We've got it going at giantscalenews.com and the guys lovvvvvvvvvve it.

just a few more days and we'll be there.
bart

 

[Winny the pooh]

Nope, been very impressed with the site so far! :friendly_wink: