Rotary Drive System
                                            John Lightfoot

How often have you wished you could hide the pushrod and control horn on a flying surface?  Various systems have been proposed from time to time but few have proved really viable, all having a disadvantage somewhere or other.

One of the best has been a "closed loop" of steel fishing trace, from both sides of a servo arm to the outer edges of the rudder, but it only works well on a fairly thick rudder.  It's not much use on ailerons, which are usually too thin and elevators don't
need it as an all-flying stab has a fully hidden drive anyway.

Then there was the "Swinge" a clever design where a span-wise swinging arm converted movement to an aileron via a hinge at 45º.  I've never seen one installed, although the Valpergas have a couple, but it seems as if the push is not all that strong and the amount of slop a bit on the high side.  It would be interesting to inspect one installed to be able to assess it properly.

Then Brian M gave me an article he'd found on the Net the RDS or Rotary Drive System.  This intrigued me so much that I built a section of wing I could have used a piece of the dead Spectre wing, but Andrew decided that broken wings simply took up unnecessary space, smashed them down to bite-sized chunks and deposited them in the bin! and made up the pieces to see if it really worked as well as the article claimed.

It does comprehensively!    There is less slop than in most standard type linkages and movement is just as powerful.  It's perhaps not quite as easily adjusted as a clevis on the end of a push-rod, but how often do you need to adjust it once it's been set?  And there's nothing at all out in the airflow!  The only remaining uncertainty at present is how it would stand up to F3B type speeds . . . would it flutter?  The only way to find that out would be to try it and see.
It's going in the new Spectre wings and Brian seems to be interested in using it for both ailerons and flaps on Happy Again
(successor to Happiness Is but renamed Plug n Play).

It is commercially available in the States but there's no indication of cost and anyway, the only tricky part is the link between servo and torque rod, and I've solved that using one of the servo discs (which seem to be seldom if ever used) and an aluminium bush which is easy enough to turn on my little lathe.

The two really important points are to ensure that the elbow is exactly under/over the hinge line it should be as close to the actual hinge line as possible but it's not all that critical and that there is no slop worth mentioning between the final wire of the elbow and the "pocket" in the aileron.  Any slop is going to appear at the pocket and in the bearing just in front of the hinge line.

I used 4 mm brass tube for the drive shaft it's almost impossible to twist a tube and 2 mm steel wire for the elbow and drive arm.  A short piece of brass tube (4 mm inside) makes the bearing, and short lengths of "failed" carbon spar, both as the main strips and the spacers, make an excellent drive pocket.

The trick is to grind down the spacers until, when assembled, the pocket just grips the wire of the drive arm.  Glueing the parts together, even with cyano, will increase the spacing very slightly, giving a good fit.  Otherwise, a little delicate sanding, with fine emery paper on both sides of some 1,6 mm balsa or similar, will ease the clearance.  If it is at all sticky, a light dusting with some graphite powder makes an amazing difference in fact a small bottle of graphite is a very useful addition to any workshop!

The secret of the system lies in a good linking bush (coupler) between the servo and the drive shaft.  A small lathe is essential for this, but the main points are a 15 mm length of 10 mm diameter aluminium rod . . . with a 4 mm bore to take the brass torque tube . . . fitted with a 3 mm grub-screw to lock the drive . . . a 2 mm deep, 6 mm diameter hole in the servo end to clear the servo-arm-locking-screw (tightened with a screw-driver through the drive shaft bore) . . . and a couple of tiny (1,7 mm or 10 BA) screws to hold the aluminium bush onto the servo disc, which is turned down to the 10 mm diameter of the coupler.

Adjustment is by slacking off the grub-screw, aligning the aileron as required and tightening the grub-screw.  This means removal of the access hatch or, if you've taken the bold step of sealing the servo inside, a small hole through which to insert an Allen Key, closed by a small tape patch when done.

I've experimented with a range of crank angles, to show what sort of values to expect.  All figures are for a 32º servo movement, as for ailerons, but would need to be doubled for flap installations with "end-to-end" movement.
 

full stick movement (no trim)
moves servo 32º from centre
crank
angle
control
surface
movement
movement at
TE of surface with
50 mm chord
30º
15º
13 mm
45º
22º
20 mm
60º
29º
n/a
80º
32º
n/a

Observe that as the crank angle approaches 90º so the surface movement approaches the servo movement as one might expect.

The throw is physically non-adjustable.  It's decided by the combination of the angle of servo movement (32º on a Futaba S-133 and 45º on a Futaba S-148) and the angle of the crank 30º to 45º for ailerons and 80º for flaps.  Of course, computer radios allow servo movement to be adjusted, usually between zero and about 120%.

It is assumed that the drive shaft is set at the appropriate angle to allow the drive crank to work parallel to the chord.  This is not absolutely essential but the only situation I can see where a deviation might be required is for a flap installation of an 80º crank (to give greatest movement), where the shaft might need to be at 70 or even 60 degrees to the chord to get the servo into the thicker part of the section.  It appears to be unnecessary to go all the way to a 90º crank the increased movement beyond about 70º being negligible.

Spectre has now been flying with the RDS controlling ailerons for two years, with highly satisfactory results!

- - - o o O o o - - -