This is really a page to talk about how to measure the actual angle that a servo has reached, rather than the last position the servo was told to move to.  The aim is to be able to allow the servos to automatically configure their actual max/min angles.  In addition, it should allow for quicker response to the stop() method.

The stop() method stops the sweeper that is currently running for the servo.  However, this only tells it to stop the sweeper, the servo will continue to move to the last position it was told to go to.

The idea is to center tap the potentiometer and run that to an arduino pin.  Also, tap the positive voltage at the potentiometer.  (2 analog pins!)  (Why? because the supply voltage of the servo is likely going to be different than the supply voltage for the arduino boards, and we'll need to account for that in our calibration routines.)

So, the idea is the following  (Example using analog pins Vref and A0)

Vref = supply voltage of (V+ as measured at the potentiometer) of the servo  (hopefully this is lower than the supply voltage of arduino!!)

A0 = signal at the center tap of the potentiometer.

On Servo startup for calibration

1. move to rest position ( 90 degrees) , measure V0(90) = 550
2. move to rest position + 10 degrees , measure V0 (value between 0-1023)
3. This is a 10 degree difference, compute the change in voltage per degree of movement. (V0(110) - V0(90))/10 = 4.3 
4. TODO: MAX/MIN value auto compute
5. TODO: Move servo to 0 degrees  (wait for servo to finish moving)
6. TODO: measure V2, use calibrated scale to determine actual angle that servo is at  ( this computed angle is the servo min.   
7. TODO: move servo to 180 degrees, and repeat step 7 to compute the max angle.

Now, in addition to this, when we call stop on the servo, we'll need to compute the current angle of the servo based on the voltage read on the A0 pin.  When stop is called, measure that voltage, and tell the servo to move to that position so it stops moving.

[[home/kwatters/ServoFeedback.py]]

In my local testing, I've tried a few configuration to center tap the servo.  In all cases the servo is powered with an external power supply.  The ground of the arduino is connected to the ground of the external power supply.  For simplicity I'll refer to the potentiometer as pin 1, 2 and 3, for the plus, center, and minus ends of the potentiometer respectively.  In all tests, pin 2 was connected to A0 of the arduino.  These numbers below represent manually turning the potentiometer connected to the servo and the arduino board until the servo motor stopped turning.

Note: * These were not very stable values.  The motor did eventually actually stop, however it was very sentisitive to noise as the motor tended to jitter in this configuration.  Turning the angle down to about 170 degrees made it less sensitive and it was easier to make the servo stop.

It's clear from the data that the Vref pin is scaling the input.  This provides much better resolution for reading as is seen by the first two connection scenarios.  A few things to be concerned with.  

1. The Vref must be lower than the supply voltage of the arduino. (a higher ref voltage will damage the arduino.
2. Noise should be cleaned up (if possible from PIN2 on the pot to isolate it better from arduino noise.
3. The real range of values is not from 0-1024, but rather from about 167 - 839.  This is interesting because the potentiometers actually rotate though 270 degrees of rotation, but the servo only responds to the center 180 degrees of it.  This gives about 3.7 steps per degree for the servo motor.  This mapping will vary from servo to servo.
4. Variation on the servo supply voltage will properly be scaled on the A0 pin so long as the Vref pin is connected.  

References:

Here's a video about using the analog reference voltage pin to properly scale reading of the arduino analog input pins.  It was possible to get the servo to stop, but it was very sensitive in this range to noise.

https://www.youtube.com/watch?v=SpDXrhjwbVo