Dragonfly Technical Page

Elevator Roll Mixer Concepts

The Dragonfly has a minor problem: Its pitch sensitivity is greatly different than its roll sensitivity. This comes in part from the fact that there are 40 sq ft of "dead" canard being "dragged" into every roll by the wing. Some of the builders have been discussing the idea of separating the elevators to make them independently controllable, and then mixing in a little differential between them to assist in rolls. Nothing on this page is meant to condone a departure from the Dragonfly plans or approved modifications. These diagrams are merely a concept, have not been tested, and are not to scale. They should be used only to stimulate conversation and critical thought.

Dave Morris' Mixer Concept

Top view of the mixer. A semi-circular slot is cut in a plate of aluminum. The plate is attached to the control stick torque tube so it can pivot about a joint (shown at the top of the plate incorrectly as a screw.) A vertical rod from the control stick torque tube sits in the slot (probably using a bearing so there is no slop). As the control stick is moved forward and backward, the entire plate moves with it, moving both the left and right elevators in unison up and down. But as the control stick is moved left and right, it causes rotation of the plate (as shown by the arrows), causing a differential up/down motion of the elevators in the proper direction.

Side view. The mixer plate will rotate horizontally, moving the elevators up and down slightly.

Patrick Panzera's Mixer Concept

This idea uses a cam. Since this is my first attempt at drawing it up, it's not fully thought through. The intent here is for the elevators to live in the un-bent portion of the cam during level flight, through some minor aileron deflection.

But once the ailerons are cranked, the cam will come in to play, reducing the lift in the direction of the turn. This is because I don't want to increase lift, in the event that the canard surface in question is close to stall at the moment of giving in more down elevator.

To limit the travel of the elevators, the stick will have stops on it in all directions. The elevator will be allowed to pass this limit when the cam causes it to, bit it will not try and go pass the elevator stops placed on the elevator itself. In other words, there will be 2 sets of stops. Elevator stop and Elevon stop.

Geoff Willis' Mixer Design

Geoff Willis' Mixer Design

V-Tail Mixer (for reference)

For reference, here is what a V-tail mixer looks like

Problems with these designs:

Dave - If the stick is pulled all the way back, and you attempt to roll, you would be forcing one of the elevators to move even farther down, beyond its stop. Obviously, there would have to be a stick limit and an elevator limit that are separate.
Dave - Don't know exactly how to affix the plate so it will rotate horizontally, but not roll with the tube.
Geoff Willis - Dave, at my office we make mechanisms like your mixer concept but we use a tube instead of a flat plate. You can screw machine the path of the control stick in a 6 inch diameter tube pretty easily.
Steve Thorn - Don't the simple aileron servo tabs detailed in DFBN correct the roll stiffness? Why is this better?
Bob Wilson - Mixer plate has too much slop ==> flutter problems.
Pat Panzera - Steve Thorn, the servo tabs are a band-aid for a poor design. This mixing is an attempt to fix the problem, not mask it.
David Richards - I'll be thrilled to just get mine in the air. I agree with Bob that this could lead to flutter from the deflection in the plate, and the machined tube would likely result in less slop. You could even use a set of ball bearing followers to reduce the friction and wear on the cam. For me, the urge to fly currently outweighs the urge to modify though. Build on!

Send me

E-mail

Return to Morris Family Page

Return to Dave's Page