Newsgroups: comp.robotics
Path: brunix!news.Brown.EDU!noc.near.net!howland.reston.ans.net!gatech!swrinde!cs.utexas.edu!csc.ti.com!tilde.csc.ti.com!mksol!strohm
From: strohm@mksol.dseg.ti.com (john r strohm)
Subject: Re: Servo control Question
Message-ID: <1993Nov8.151332.25472@mksol.dseg.ti.com>
Keywords: Binky wubba Boink
Organization: Texas Instruments, Inc
References: <1993Nov7.210018.11792@msuvx1.memst.edu>
Date: Mon, 8 Nov 1993 15:13:32 GMT
Lines: 26

In article <1993Nov7.210018.11792@msuvx1.memst.edu> umlangston@msuvx1.memst.edu (Mark C. Langston) writes:
>  This may or may not be appropos for this group, but here goes:
>I understand in a general way what needs to be done to control a servo.
>However, I do not understand _why_.  Although I could wire something up
>to turn one, I have no idea _why_ a pulsed signal turns the shaft one direction
>or another.  Could someone please explain this?

Well, the shortest answer is that the servo behaves that way because it was
designed to behave that way: R/C servos were designed to speak pulse-width
modulation.

As for WHY R/C servos speak PWM, that is a system design issue that was worked
back in the dark ages before microprocessors.

R/C servos were developed specifically for use in hobbyist radio control
systems.  Hobbyists typically have relatively limited budgets; the manufacturers
recognized early on that keeping the cost of the transmitter and receiver down
was a Good Idea.  This leads immediately to a multiplexed data stream of some
sort.

The next problem is system reliability in the presence of noise.  Earliest
radio control technology was all AM radio, which is notorious for poor noise
immunity.  Whatever scheme is used to put the control signal onto the radio
wave has to be very resistant to noise.

Pulse-width modulation is easily multiplexed and very noise-resistant.
