Flybarless mythbusters :D

[GravityKills]

Hello,

here it comes.
[ame="http://www.youtube.com/watch?v=HbqNoG8peho"]YouTube- flybarless mythbusters episode II.wmv[/ame]
I hope this gives some useful input to the discussion.

One key message: The dreaded "excessive forces" when using ultra-short servo horns do not matter, as long as push-pull linkage is used.

[jamiem]

If the linkages are the same length, and they are at different distances from the horn centre, then as the horn turns, the linkages will start to bind up. I've seen this first hand on a trex 600 which had the wrong size servo horns for the bell cranks; it made the whole front end of the frame twist as you gave large aileron movements.

[GravityKills]

ROFL
let's cover this topic in

Flybarless mythbusters: Episode III
"no it does not get stuck"

[ame="http://www.youtube.com/watch?v=RpZZTEFDtlY"]YouTube- flybarless mythbusters episode III.wmv[/ame]

What is crucial is that the lenghts of the links are equal. On that, I would wholeheartedly agree. Also, if both balls on the servo horn don't exactly match, it will jam. But this is easily avoided by using pre-drilled holes or working carefully.

[raptorheli2]

your videos are all well and good, but it does matter.

your linkages should be the same distance apart as the bellcranks. if i knew how to explain it i would. far better people than me have told me why it matters and tbh it went over my head or i switched off due to the bore of listening to mechanical engineering facts. however one of my mates is VERY knowledgeable on this so i'll ask him to chime in when he's on msn later.

i think there is a difference between being a little off the bellcrank size (hirobo) and then down to like 10mm of a difference such as this.

cheers

[0h5h1t]

If the linkages are the same length, and they are at different distances from the horn centre, then as the horn turns, the linkages will start to bind up. I've seen this first hand on a trex 600 which had the wrong size servo horns for the bell cranks; it made the whole front end of the frame twist as you gave large aileron movements.

i disagree, my 600, and two 700's have reduced throws setup on the servo horns and none of these bind in the slightest even at full cyclic..
in fact if you check the vbar head setup for the trex 600 and 700 it clearly shows that you need 17mm from centre of servo arm to centre of ball.

the 600 you saw bind up must have had other setup issues affecting the linkages.

[jamiem]

Nope, i saw it first hand. Changed from the wrong size servo horns to the correct align ones which have the same spacing, then the problem went away. Its an easy one to work out. If the bell crank is say 20mm out from the centre, and the horn is 10mm, then at mid stick everything will be fine, but at total 90 dgrees from this, the bell crank will have moved one linkage 20mm back and one forward (40mm total distance), and the horn 10mm forward and back, so there you go. if one end has moved back 20mm, but the other only 10mm, where has the other 10mm gone? stress in the linkages, frames, and servos. Simple maths. If yours isnt exibiting this, then its either only a very small differnce, or you cant see it, or your getting a bit of flex/movement in each component adding up to virtually zero net result at the servo.

[GravityKills]

... far better people than me have told me why it matters and tbh it went over my head or i switched off due to the bore of listening to mechanical engineering facts. ...

It's appropriate in many situations, but not in all of them.

The advantage of equal arm lengths is that both "wheels" (servo wheel and bellcrank) effectively turn as one, no matter the angle. I could attach the servos straight to the bellcrank, having them stick out of the canopy left and right, and it would work exactly in the same way as parallel linkages.

If the heli is designed appropriately - like a 700 flybarred - parallel links are a perfect solution. No second thoughts about this.

For a 700 flybarless, however it's probably the smallest evil to use non-parallel PP-links. It's not optimal, because they introduce nonlinearity: At the end of the travel range, the bell crank responds slower to servo movement than in the middle. That results for example in cyclic-collective interaction. It's all a compromise.

On the other hand, if I would continue to use the original (parallel) wheels, I get a pitch range of about 25 degrees instead of 14. Servo authority is effectively cut into half.
When it comes to damping, like recovery from elevator flips, the flybarless controller relies on the idea that it can predict what the servos are doing. It gives the correct control input, the servos move as commanded, and the heli stops crisply. Unfortunately, when the servos are overloaded, they don't always move as commanded, sometimes faster, sometimes slower, and the whole idea of damping the heli electronically falls to pieces.

To make a long story, short, it's all a compromise. Maybe we'll throw come conventional wisdom overboard and get a compromise that's better suited for FBL. The verdict on that isn't out yet.

[GravityKills]

...Its an easy one to work out. If the bell crank is say 20mm out from the centre, and the horn is 10mm, then at mid stick everything will be fine, but at total 90 dgrees from this, the bell crank will have moved one linkage 20mm back and one forward (40mm total distance), and the horn 10mm forward and back, so there you go. if one end has moved back 20mm, but the other only 10mm, where has the other 10mm gone?

the error in your thinking is that the bell crank cannot travel 90 degrees. Simple as that.
The servo can approach 90 degrees (the rods collide), the bell crank travels less. You can see it in the 2nd video.

PS: I fell into the same trap myself one year go :-)

[Gareth-71]

Hi All

...my 2cents, if you want it....

Mathematically you can have much much smaller servo horns than the cyclic horns, the result is simple, and can be easily seen by looking at a BMX chain gear.....each spocket gives as much as it takes, so binding will not happen IF the links are the same lengh, however, there are some things to consider:

1. If the servo arms are, say, half the length of the cyclic horns, then the servo will act like a servo with half the speed and twice the torque.
2. The travel is limited. the greater the difference the smaller the angle range. This can be demonstated with simple math, but I won't bore you all more than you already are.

[Dedcalm]

Remember this assumes (as I would hope is always the case) that if you drew a line through the 2 ball links on either horns it passes through the centre of the horn. Also no matter how you set up the linkages you will always get more movement around centre than at the extremes of movement because the servo rotates... the only way around this would be to use linear servos... and they went out with the ark (i think).

[RazzleDazzle]

If the linkages are the same length, and they are at different distances from the horn centre, then as the horn turns, the linkages will start to bind up. I've seen this first hand on a trex 600 which had the wrong size servo horns for the bell cranks; it made the whole front end of the frame twist as you gave large aileron movements.

Jamie... This problem rings a bell, i think i know what model you are talking about, mmmm??? Mine !!! lol i had ace servos on my t-rex 600 and the servo horns where to long of to short (cant remember) but because they had like 14kg torque or somthing like that. At full ailerons the whole front of my frame would bend and twist. Once i changed over to the align servos and fitted the align horns. Not a problem now

[moyesboy]

The attached diagrams show how with push pull the servo disc size must match the belcrank size or there will be some binding.

In the cad sketch the centre distance between the discs is fixed as are the disc sizes (at 30 one end and 20 the other).

The link rods are forced to be equal length but can stretch, and their result length is measured with the grey driven dimension.

This shows how with the arms at 90 degrees the rods need to be 60.21 long, but with the arms defelected 30 degrees, either way, they would need to stretch to 60.34.

So with different link centres on the servo to the belcrank the link should spring to centre - someting has to give to allow movement, or there needs to be a bit of clearance in the ball links (it'll develop if it wasn't there to start with).

The error is small and probably its practical to ignor it unless the difference in radii are extreme.

[GravityKills]

LOL OK there -is- a difference. If we look careful enough.

To put it into perspective, the difference is (60.34-60.21) = 0.13 mm. I can't even set the nominal length of the pushrods this accurately, even if I use half turns on ball links and put them on backwards. And the servo is, after all, mounted in rubber grommets.

>> The error is small and probably its practical to ignor it unless the difference in radii are extreme.

I'd tend to agree with that statement
On my 700, the reduction is quite extreme, and it still moves smooth as butter. There is no significant resistance when taking off the servo wheel and putting it back on (which would indicate a problematic mismatch in length).

[moyesboy]

There is no significant resistance when taking off the servo wheel and putting it back on (which would indicate a problematic mismatch in length).

You can tell pretty accurately by popping off one end of one link and offering it up to the ball to see if its centred.
To be sure the binding isn't significant you need to compare with the linkage centered, and with the linkage at full movment. Full movement in either direction should be the same.

[GravityKills]

now that I know what I'm looking for, I can even feel it
It takes about half a turn on one pushrod at the endpoint to get it even. But since I use only full turns (to put the links on with the wider end), I won't lose any sleep at night over this.

The pictures aren't too useful, since it's really only a fraction of a mm.