Servo to Control Surface Linkages

[Kevin Wilson]

A bit of an academic question to gauge how other builders think.

If we consider a 1:1 ratio on the control linkage; ie the servo horn distance is the same as the control horn distance.

What difference (if any) would it make if both were 1/4" or both were 1" (metric equivalents accepted)

To kick start your brains; consider torque, friction, slop

I have my thoughts, but I know I am surrounded by greater minds with more experience.

Edited By Kevin Wilson on 19/03/2014 21:36:58

[Barnstormer 52]

Hi

My thoughts only:

Torque would not be affected if you were to keep the distance from the centre of the hinge line of the control surface equal to the distance from the centre of the servo arm as the ratio would be 1:1

Friction, the actual friction can't change, but with the linkage closer to the centre of the servo output arm the thrust must be greater. (if you want to cut something thick you put it close to the pivot of scissors or pliers, not at the tip) so I guess the servo would overcome more friction.

Slop, caused by misfit between the control horn and the clevis, or within the gears of the servo I would have thought must be magnified the closer you have the horn to the hinge line on the control surface end. But slop caused by a fault at the servo end would vary in that a misfit between the horn and hole in the servo arm would be less evident the further you moved away from the centre of the output shaft, but slop in the gears would remain constant as long as the ration of the distance from the centre of the servo arm as the ratio would was 1:1

I think!

Geoff

[Simon Chaddock]

Kevin

My own view is that the effective length of both the servo arm and control horn should be as long as practical.

1. It reduces the 'play' resulting from the bearing surfaces.

2. For any given torque at the control surface it reduces the load on the link itself and thus any deflection that the load might cause.

Using long arms and horns does require that they be rigid enough to handle the loads involved.

The above is particularly effective in the case of a direct rigid link.

With a snake (which I avoid using if at all possible) its inherent friction does alter the situation slightly in that the longer the servo arm the more of the available servo torque is absorbed in over coming its friction.

In this case the optimum arm length will depend on the level of friction in the snake and the 'flexibility' within the system.

But then I do tend to use very small servos!

[DG3]

Subscribed.

Think it is just a balancing act based on maximising servo torque (use short arm hole), maximising servo throw (to maximise resolution obtained), obtaining desired throws (this can vary hugely from ordinary sports model to 3D model) and avoiding any setups that don't promote mechanical advantage...?

**LINK**

[Plummet]

My gut feeling is that you should aim for the maximum possible pushrod/snake movement.

The system will usually have a tiny bit of slop in it - say 0.1mm.

If the movement of the pushrod for a typical non-violent control action is just 1 mm, then the slop is 10%. If the pushrod movement is 10mm, the slop is 1%.

Then, the greater the movement, the less the forces on the pushrod/snake. Less force means that it should distort or flex less.

Plummet

[Peter Miller]

Posted by Barnstormer 52 on 19/03/2014 22:19:03:

Friction, the actual friction can't change, but with the linkage closer to the centre of the servo output arm the thrust must be greater. (if you want to cut something thick you put it close to the pivot of scissors or pliers, not at the tip) so I guess the servo would overcome more friction.

I think!

Geoff

Yes, the shorter arm on the servo would increase the thrust but the shorter horn would reduce the leverage, you need a long lever. You know the old saying "give me a long enough lever and a fulcrum and I can move the world."

So the effecctive leveage would remain the same.

The main advantage of the longer arms is the reduction in slop

[Stevo]

I'm with Plummet here - I'm not sure if it does make any dfference on a 1:1 ratio, but I maximise the servo movement.

For initial setups, I aim for just a little over the textbook throws, and alter the linkages to give around 80-90% servo throw - as opposed to maximum control surface movement for 20-30% servo throw.

If the servo offers 1024 steps end to end, 90% of that is 921 steps. Using a small throw, it's around 256 steps, which could be a bit juddery and difficult to adjust accurately.

Oh - have I waffled on on and missed the point? Off for a lie down.

[Chuck Plains]

On the slop issue.

The amount of slop in a linkage does depend on what type of connection you use. On my foamies I am always using plain wire Z bends, which are most prone to slop. But, if I am bothered by it, I will just place a single drop of CA where the wire is through the arm, and when it hardens, just wiggle it a bit and you have a slop free pivot.

[Shaunie]

Posted by Peter Miller on 20/03/2014 08:34:31:

Posted by Barnstormer 52 on 19/03/2014 22:19:03:

Friction, the actual friction can't change, but with the linkage closer to the centre of the servo output arm the thrust must be greater. (if you want to cut something thick you put it close to the pivot of scissors or pliers, not at the tip) so I guess the servo would overcome more friction.

I think!

Geoff

Yes, the shorter arm on the servo would increase the thrust but the shorter horn would reduce the leverage, you need a long lever. You know the old saying "give me a long enough lever and a fulcrum and I can move the world."

So the effecctive leveage would remain the same.

The main advantage of the longer arms is the reduction in slop

The longer moment arm will also reduce the thrust in the system which will reduce wear in the linkage holes and the output bearings in the servo (particularly plain bearing servos).

Shaunie.

[Kevin Wilson]

I have deliberately held off commenting to see what the consensus was.

I was very careful in my original post to state a 1:1 ratio so we would not get sidetracked by differing ratios and mechanical advantages. ((and it being my normal starting point of using the full arm with a 1:1))

It was the length of arm and available servo torque that got me thinking and I initially fell into the trap of longer arms being a leverage advantage, but of course if the ratio is 1:1 then this cancels out.

Only Simon has expanded on my initial 3 factors; is this because I captured all (unlikely) or that my including 3 limited everyones horizons?

Of my three factors I decided, like Plummet, that slop would be the most significant. Although I had not considered Simon's one of stiction in snakes. Probably because I usually use rigid rods.

Gentlemen, thank you for joining me in some mental gymnastics.

Rgds, Kevin

[Shaunie]

Hi Kevin,

You mentioned torque friction and slop, I mentioned wear.

Shaunie.

[Kevin Wilson]

So you did.

I wonder if the ratio remains at 1:1 wil the length of moment arm affect the torque, therefore the pressure on the bearing points and therefore the wear?

There is a fixed pressure dictated by the air speed, control surface area. Intuitively a shorter horn will need more force to counter the control surface force.

[Chris Anthony]

Kevin,

As with all engineering, it's a compromise. This situation is similar to that of a steering system, on say a race car which is a more familiar scenario for me.

All things constant; overall motion ratio, control surface, hinges, load cases, "bearing types" (assuming by bearing, you mean connection of rods to levers at each end?) and servo.

The longer "lever arms" at each end will indeed result in less compressive/tensile force in the control rod. This is good for longevity of the rod (although I am sure they are up to it), but it also reduces any deflection caused by the bearings, as there is less force causing displacement at the bearing joints. What is more benefitial in this case though, is that any deflection in the bearings has less overall effect on angular displacement of your control surface. So the "lash/slop" amount at the bearing may be 0.1mm, but with a lever arm of 20mm at the flap, this only gives 1 deg of uncontrolled movement, whereas with a lever of 10mm, you may see 2 degrees.

Of course the limiting factor on lever length, is what is practical; what fits inside the fuselage, or what may be sticking out of the wing too much and resulting in aerodynamic performance issues (or becoming exposed to contact with the ground) etc.

Overall torque seen at the servo of course won't change unless you change the calculated motion ratio.

edit - friction sources will be at the control surface hinge I presume. Your lever mechanism comes before this, so no matter what your lever solution, it won't help to increase/decrease this friction value. The frictional force at the hinge will just be part of your load case, and added onto the aerodynamic force resisting flap movement. Unless of course you are considering the effects of friction in the joints between the control arms and the levers. In this case, the same theory above on joint deflection is valid, less force in the rod, means less joint friction (with a constant coefficient of friction in both cases).

I hope this is of some use.

Thanks,

Chris

Edited By Chris Anthony on 26/03/2014 08:42:35

Edited By Chris Anthony on 26/03/2014 08:43:28

Edited By Chris Anthony on 26/03/2014 09:08:28

[Bob Cotsford]

don't forget slop in the servo gearbox. The longer the horns, the more this play will be magnified, so it's another factor to be added into Chris's compromise criteria.

[Chris Anthony]

Actually Bob, from what I understand, this won't be in our control. If we have assumed a constant load case at the flap (let's say a torque around the flap hinge), and have also specified a constant motion ratio between flap and servo, then the servo torque required will be constant, no matter what our chosen lever lengths are.

What I think you are getting at Bob, is to adjust the torque required from the servo itself by changing the overall effective motion ratio, which can also considered to be a torque ratio between servo and flap hinge. This is possible, but the consequence of reducing torque required at the servo is a reduction of flap motion/travel for a given servo motion. This is a reduction in motion ratio, and as said before, this is not the situation the op wanted to consider.

Servo gearbox slop is not something that can be can be reduced (or amplified) in this situation, and is just one piece of the "compliance" puzzle that will have to be tolerated.

Chris

[Bob Cotsford]

Posted by Chris Anthony on 26/03/2014 14:47:03:

Actually Bob, from what I understand, this won't be in our control. If we have assumed a constant load case at the flap (let's say a torque around the flap hinge), and have also specified a constant motion ratio between flap and servo, then the servo torque required will be constant, no matter what our chosen lever lengths are.

What I think you are getting at Bob, is to adjust the torque required from the servo itself by changing the overall effective motion ratio, which can also considered to be a torque ratio between servo and flap hinge. This is possible, but the consequence of reducing torque required at the servo is a reduction of flap motion/travel for a given servo motion. This is a reduction in motion ratio, and as said before, this is not the situation the op wanted to consider.

No, I was just think in terms of lost precision, or increased slop, caused by the magnification factor of a long servo arm on any play in the gearset. At the control surface trailing edge it will be negated by the long control horn, but the movement in the pushrod will still be increased allowing for easier onset of flutter

Servo gearbox slop is not something that can be can be reduced (or amplified) in this situation, and is just one piece of the "compliance" puzzle that will have to be tolerated.

Chris

Yes, it is out of our control other than our choice of servo quality

 

Edited By Bob Cotsford on 26/03/2014 16:05:33

[Chris Anthony]

Posted by Bob Cotsford on 26/03/2014 16:04:20:

No, I was just think in terms of lost precision, or increased slop, caused by the magnification factor of a long servo arm on any play in the gearset. At the control surface trailing edge it will be negated by the long control horn, but the movement in the pushrod will still be increased allowing for easier onset of flutter

 

Sorry Bob, I'm not sure I follow you. Are you talking about the bending of the levers themselves, and the effects of which may be greater for a longer lever? If so, this is something that I had not considered, but is valid.

 

Posted by Bob Cotsford on 26/03/2014 16:04:20:

Yes, it is out of our control other than our choice of servo quality

 

Ah, well yes of course, agreed.

 

Edited By Chris Anthony on 26/03/2014 17:01:10

[Chuck Plains]

Posted by Kevin Wilson on 20/03/2014 23:09:26:

I have deliberately held off commenting to see what the consensus was.

I was very careful in my original post to state a 1:1 ratio so we would not get sidetracked by differing ratios and mechanical advantages. ((and it being my normal starting point of using the full arm with a 1:1))

It was the length of arm and available servo torque that got me thinking and I initially fell into the trap of longer arms being a leverage advantage, but of course if the ratio is 1:1 then this cancels out.

Only Simon has expanded on my initial 3 factors; is this because I captured all (unlikely) or that my including 3 limited everyones horizons?

Of my three factors I decided, like Plummet, that slop would be the most significant. Although I had not considered Simon's one of stiction in snakes. Probably because I usually use rigid rods.

Gentlemen, thank you for joining me in some mental gymnastics.

Rgds, Kevin

Don't forget the leverage points of the diameter of the servo's internal gears. If you had a pushrod position exactly at the meshing diameter of the final drive gear, you would have lots more torque than one at the end of a 1" arm.