Just got into electric planes after flying glow many years ago. I am really pleased with capabilities of new computer transmitters, especially Exponential settings. This is a wonderful feature but soon realized it is a necessity. Think about it, your standard rotary servo driving a pushrod to your elevator for example is actually an inverse-exponential thing. Around servo center the pushrod travel has an essentially linear relationship with transmitter stick movement and that's good. But at extreme servo arm rotation the pushrod travel is less for a given stick movement. So the effect is negative-exponential and you need to add some exponential in you transmitter just to make the stick-to-pushrod travel a linear relationship.

My point is that Exponential is a necessity, not a luxury. I don't understand why they don't make linear servos instead of rotary servos to fix this problem. I remember years ago the first digital proportional radio systems had linear servos. It seems like a no-brainer to me.

Cost. It always comes down to cost, but if there's enough demand by folks who are willing to pay it could happen.

Pete

Once upon a time (the story of my life:-))))))) there were marketed "kits" for converting rotary to linear motion. Since they are no longer available, it appears there was no real demand.

Les

Speed, torque, cost are all factors. Plus, if I want more throw with a given servo, I just put a longer arm on. How do you increase throw on a linear servo?

Buzz.

Thank you Buzz.  I haven't given much thought to internals but I suspect proper gear ratio might satisfy speed and torque (force) issues.  As for throw the linear design would have sufficiently long travel, as much as you would get with a long arm on rotary servo.  Then use transmitter EPA (end point adjustment) to set the desired pushrod travel.  Might call for a longer servo case but would eliminate clearance zone required around servo for long rotary arm.

thebutterfly

Well I can get almost 3 inches of travel with the extended aluminum arm. That would make the servo 3 1/4 to 3 1/2 inches long.

Think of the linear servo trying to keep up with a tail rotor servo. They have speeds of .05 sec for 60 degrees.

http://www.servocity.com/html/hsg-5084mg_servo.html

Now, if you are worried about the differential that is imparted from moving around a circle, set the rotation of the servo 90 degrees to the control surface. An example from the pattern planes, the elevator servo is on its side so the trow for the split elevators is equal.

I just do not see any way of getting the speed and torque out of a linear as compared to a rotatory servo.

Buzz.