HobbyKing.com New Products Flash Sale
Reply
Thread Tools
Old Nov 20, 2012, 08:44 PM
Registered User
Australia, QLD, Brisbane
Joined Jul 2012
82 Posts
Discussion
Why are servo motors instead of steppers used to actuate flight control surfaces?

I know I'm not Einstein and that I haven't come up with a brilliant idea nobody else has ever thought of. There's undoubtedly a good reason, but I just can't figure out why RC planes always use servos instead of stepper motors.

Ignoring the closed-loop versus open-loop for a minute, and just focusing on other aspects of the two types of motors:
I understand that stepper motors draw more current because they are constantly "on" (featuring high torque at low or zero rotational speeds) and that conversely, servo motors only draw current when they need to rotate to a new position. I would expect that a consequence of this is that servos have very little torque at standstill.

I would have thought that flight control surfaces would be under constant load, particularly those on the wing, and thus require constant torque? Is this assumption incorrect? If so, I can understand that the torque (and thus current draw) of a stepper motor would be wasted whilst the flight control surfaces were not being "used".

The reason I ask this question is because I am designing an unusual model aircraft where I think a stepper motor would be a more suitable choice, because the flight control surface is definitely under constant load. My concern is that the stepper will draw a lot of current from the batteries so I'm investigating other options such as using detent torque, worm gears or braking mechanisms to "lock" the current position of the control surface.

Would appreciate any responses to what may very well be a pretty stupid question.
robert_houdin is offline Find More Posts by robert_houdin
Reply With Quote
Sign up now
to remove ads between posts
Old Nov 20, 2012, 08:49 PM
I don't want to "Switch Now"
pmackenzie's Avatar
Toronto (Don Mills), Canada
Joined Dec 2002
17,106 Posts
I don't think you can ignore the open loop/close loop part. Servos are generally better in every way than steppers.

Only time I have seen steppers on a model was in a micro, a long time ago.
I will look for the link......

Pat MacKenzie
pmackenzie is online now Find More Posts by pmackenzie
Reply With Quote
Old Nov 20, 2012, 08:55 PM
I don't want to "Switch Now"
pmackenzie's Avatar
Toronto (Don Mills), Canada
Joined Dec 2002
17,106 Posts
Found the link

http://www.rcgroups.com/forums/showthread.php?t=141999
pmackenzie is online now Find More Posts by pmackenzie
Reply With Quote
Old Nov 20, 2012, 09:28 PM
War is over (if you want it)
vespa's Avatar
Thousand Oaks, CA
Joined Mar 2004
2,665 Posts
Stepper motors have two very significant advantages:
1. They need no feedback potentiometer or error feedback control loop. Steppers go exactly where you tell them with absolute precision.
2. Steppers have very high torque

Steppers have several very significant disadvantages:
1. If they meet excessive resistance even for an instant they can "lose step", completely eliminating benefit #1
2. The high torque design does not work well at high speeds, reducing the value of benefit #2.
3. They are hard to make on a small scale due to the number of poles/slots required for precision and the controller circuitry is somewhat more bulky than brushless.
4. They have extremely limited resolution based on the number of magnetic poles. "Micro-stepping" can interpolate between poles but with greatly reduced torque, efficiency, and accuracy. Resolution can be infinitely increased by way of gearing but steppers are limited to low RPM so gear ratios are limited by speed requirements. Thus for most applications you cannot achieve the desired speed and resolution, only one or the other.

So the bottom line is that steppers would require a potentiometer and error feedback loop for most servo applications, they have a higher cost, a lower power/weight ratio, and are more complicated to design.
vespa is online now Find More Posts by vespa
Reply With Quote
Old Nov 20, 2012, 09:46 PM
I don't want to "Switch Now"
pmackenzie's Avatar
Toronto (Don Mills), Canada
Joined Dec 2002
17,106 Posts
Just like you would in a motion control application you can get around the feedback problem by doing an origin return on power up.
But not sure I would want that for a control surface application.

Pat MacKenzie
pmackenzie is online now Find More Posts by pmackenzie
Reply With Quote
Old Nov 21, 2012, 01:41 AM
Registered User
Australia, QLD, Brisbane
Joined Jul 2012
82 Posts
The "flight control surface" to which I refer is also a wing. For this application, I need to be able to vary the AoA of the wing using direct-drive from a motor - but I also need to be able to "lock" the wing at a certain AoA. A motor with high torque at low and zero rotational speeds (but high load) is obviously crucial to this design, and is what steered me towards choosing to use a stepper motor - I was under the impression that a servo could not match the torque of a stepper motor whilst stationary under high load? I could very well be wrong about this though.
robert_houdin is offline Find More Posts by robert_houdin
Last edited by robert_houdin; Nov 21, 2012 at 01:49 AM.
Reply With Quote
Old Nov 21, 2012, 04:59 AM
An itch?. Scratch build.
eflightray's Avatar
South Wales U.K.
Joined Mar 2003
13,785 Posts
Have a look at this servo chart -

http://fatlion.com/sailplanes/servochart.html

Just click on one of the manufacturers/suppliers names further down to see the charts.

Then you can compare weight, size, and torque with any similar stepper motors you know of.

The advantage from a servo possibly comes from the internal gearbox that allows high resolution and high torque.

Though changing a wings incidence angle, (quite small movement ?), could probably be done with a lever system and quite a low powered servo.

But why change the wings incidence angle when changing the horizontal stabilizer, (elevator or all-flying stab), is so much simpler.
eflightray is offline Find More Posts by eflightray
Reply With Quote
Old Nov 21, 2012, 10:39 AM
agnotology
kcaldwel's Avatar
Joined Jan 2007
3,680 Posts
Quote:
Originally Posted by robert_houdin View Post
I was under the impression that a servo could not match the torque of a stepper motor whilst stationary under high load? I could very well be wrong about this though.
The newer digital servos have full torque to hold a position and for small movements.

http://www.futaba-rc.com/servos/digitalservos.pdf

On the other hand, the force required to hold a wing at a given AoA can be quite small if you chose the pivot location properly. A symmetrical airfoil will have basically zero moment around the 1/4MAC. Cambered airfoils will have a point just aft of the 1/4MAC where the moment will be very small across the speed range. And there are lots of mechanisms that are basically one-way, so the servo would not see the holding forces in any case.

http://zaerotech.com/design/Zaerotec...tml#!Zaerotech

Kevin
kcaldwel is offline Find More Posts by kcaldwel
Reply With Quote
Old Nov 21, 2012, 10:55 AM
An itch?. Scratch build.
eflightray's Avatar
South Wales U.K.
Joined Mar 2003
13,785 Posts
Is that for a 'twist wing' mechanism in the picture Kevin ?, where each wing panel can change incidence angle, instead or having ailerons.

I built an electric glider many years ago after seeing an article on twist wings in one of the model mags.

Did it fly ?.

Well it would have done I'm sure, but never just 'waggle the sticks' to see if the controls work, make sure they actually work the right way. But it did roll nicely before it disassembled itself.
eflightray is offline Find More Posts by eflightray
Reply With Quote
Old Nov 21, 2012, 11:44 AM
Registered User
Joined Oct 2004
2,913 Posts
Stepper motors require twice as many power transistors in the driving circuitry too. And if a step is missed, or the power is lost for a split second, or a bump strong enough has thrown a control surface away from the last known position, then there's no way to re-center it. Stepper circuits work by dead reckoning, that's why printers and scanners have to bring their carriages to a known starting position.
Brandano is offline Find More Posts by Brandano
Reply With Quote
Old Nov 21, 2012, 02:08 PM
B for Bruce
BMatthews's Avatar
The 'Wack, BC, Canada
Joined Oct 2002
11,629 Posts
You're also missing out on the simple fact that the servos we use will fight to remain in place.

You're right that once they stop that they do not lock in place. But if a load is put onto the output arm that is sufficient to displace the arm by some amount the feedback pot will sense the slight motion and power up the motor to resist that load and return the arm to the proper desired position and HOLD IT THERE against the load on the arm.

You can easily feel this for yourself. Power up a radio in or out of a model and reach in and try to push one of the servo arms. You'll feel the servo push back at you and you'll hear and feel the buzz of the pulses fed to the motor to restore and hold the position. This being done within the mechanical and electrical deadband limits of the servo.
BMatthews is offline Find More Posts by BMatthews
Reply With Quote
Old Nov 21, 2012, 02:10 PM
War is over (if you want it)
vespa's Avatar
Thousand Oaks, CA
Joined Mar 2004
2,665 Posts
Robert, a stepper is designed with magnetic pole pairs that align for absolute maximum holding torque and in practice most controllers will send full power to the motor at all times, even when completely resting at 0 RPM with no load so the "holding torque" is tremendous. A brushed/brushless servo by definition is the exact opposite -- motor power is completely disconnected anytime the output arm is in the commanded position so the holding torque is zero, by definition.

The above theory does not usually prove to be limiting in practice however, since potentiometers have very high resolution and digital servos can have very aggressive tuning, the net result being that most modern servos have adequate holding torque within a tight range of a fraction of a degree -- usually well below the range of slop of a modern linkage, especially something as inherently sloppy as a wingeron linkage.
vespa is online now Find More Posts by vespa
Reply With Quote
Old Nov 21, 2012, 06:52 PM
Registered User
Australia, QLD, Brisbane
Joined Jul 2012
82 Posts
Thank you very much to everyone for the responses. I now have a much better understanding of why servos are suited to RC aircraft applications.


Quote:
Originally Posted by vespa View Post
well below the range of slop of a modern linkage, especially something as inherently sloppy as a wingeron linkage.

Could you please elaborate on this? I'm not sure what you mean. I assume that you are referring to the linkage between a servo and a flight control surface in a conventional RC aircraft. Just wondering what you mean by "inherently sloppy"?

In this design, the entire wing will be rotated, with the servo shaft attached directly to the spar, so I probably don't need to worry about the sloppiness to which you refer, but I'm just curious.
robert_houdin is offline Find More Posts by robert_houdin
Reply With Quote
Old Nov 21, 2012, 07:32 PM
I don't want to "Switch Now"
pmackenzie's Avatar
Toronto (Don Mills), Canada
Joined Dec 2002
17,106 Posts
Quote:
Originally Posted by robert_houdin View Post
In this design, the entire wing will be rotated, with the servo shaft attached directly to the spar
Unless you need 30 to 45 degrees of travel in both directions, this is a poor choice.
Something indirect that allows for full servo travel over the desired range of motion of the wing would be better.
pmackenzie is online now Find More Posts by pmackenzie
Reply With Quote
Old Nov 21, 2012, 07:40 PM
War is over (if you want it)
vespa's Avatar
Thousand Oaks, CA
Joined Mar 2004
2,665 Posts
A wingeron has lots of places for slop to occur. First you typically need ball links which flex. Then there's some bellcrank that has slop and flex. Then some pins to drive the wings that must fit loosely for removal. Then the wings must fit loosely on the joiner -- even with ball bearings there is significant slop. Then the entire wing can roll L/R relative to the fuselage, producing a phantom input.

Compare that to a well done aileron linkage with two tight-fitting clevises and a skin hinge. I'm not advocating anything, just saying that if servos are precise enough for a solid aileron linkage they're precise enough for a wingeron.
vespa is online now Find More Posts by vespa
Reply With Quote
Reply


Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Using old servo control boards as actuator driver/motor ESC for PROPORTIONAL control EQMOD Electric Power Cars 12 Dec 20, 2012 10:05 AM
Discussion Why are servo motors instead of steppers used to actuate flight control surfaces? robert_houdin Radios 20 Nov 23, 2012 09:33 AM
Discussion Weakest servo, what control surface to use it for? ArizonaFun Electric Plane Talk 12 Sep 16, 2012 08:31 PM
Yippee! Stepper motor control using PWM signal from the apparatus RC Rumcajs Power Systems 5 Aug 07, 2012 05:28 PM
Discussion What Flight Controllers are YOU using on your Hex/Octo? .Joel Multirotor Talk 10 May 13, 2012 08:23 PM