control for servo

Your program is the same as the official tutorial found at http://arduino.cc/en/Tutorial/Knob.

There are two issues that may be affecting your satisfaction of the servo performance.

1:   The Input is an analog value going into the ADC of the uC via A0.  It may be necessary to  ground the other analog inputs thru resistors so they are not 'floating'.

2:  The Output is driving a possibly high current drain servo motor that may need its own power supply in order not to overload the (correctly spelled) duemilanove's regulated power supply.

Please note that if the val (provided from the ADC when analogRead executes) is bouncing around more than +/- 5.7 (+/- 28 mV) then the 0 to 179 value is changing also and the servo will be instructed to move a 1 degree step.  If the circuit cannot achieve that level of stability then I may suggest that you use val=2*map(val,0,1023,0,89);  this will allow the random variations of ADC to swing +/- 11.4 (+/- 56 mV) before a 2 degree change in the servo position is generated.  IMO this method is superior to sampling multiple times and taking an average.  It trades granularity for stability.

If you live close to a commercial facility that is generating 1000 watts or more of rf energy it may be the cause of the instability of the value of the variable 'val'.

Don't forget to post complete feedback about what changes you made that improved the performance of your hardware.  

Happy Trails.

Your program is the same as the official tutorial found at http://arduino.cc/en/Tutorial/Knob.

There are two issues that may be affecting your satisfaction of the servo performance.

1:   The Input is an analog value going into the ADC of the uC via A0.  It may be necessary to  ground the other analog inputs thru resistors so they are not 'floating'.

2:  The Output is driving a possibly high current drain servo motor that may need its own power supply in order not to overload the (correctly spelled) duemilanove's regulated power supply.

Please note that if the val (provided from the ADC when analogRead executes) is bouncing around more than +/- 5.7 (+/- 28 mV) then the 0 to 179 value is changing also and the servo will be instructed to move a 1 degree step.  If the circuit cannot achieve that level of stability then I may suggest that you use val=2*map(val,0,1023,0,89);  this will allow the random variations of ADC to swing +/- 11.4 (+/- 56 mV) before a 2 degree change in the servo position is generated.  IMO this method is superior to sampling multiple times and taking an average.  It trades granularity for stability.

If you live close to a commercial facility that is generating 1000 watts or more of rf energy it may be the cause of the instability of the value of the variable 'val'.

Don't forget to post complete feedback about what changes you made that improved the performance of your hardware.  

Happy Trails.

I seem to recall we went through this previously...

Your dealy time is a little short (15mS) which is probably why you're getting strange reasons.

You can't possibly move the pot or expect the servo to reacte that quick, so try a delay(200) instead of the delay(15) to slow it down.

As you servo shifts, the motor current will have an influence on the supply (5v) and this will influence the analog reading, which influences the servo, which ....

As Billabott has pointed out, any small change will influence the result.

Your wiring can have some affect and the ground wire from the servo, and the ground wire from the pot should ideally be two seperate pieces of wire.

This stops the motor current travelling along the pot wire and influencing the reading.

You can also help by adding a 1000uF or larger across the 5v as close to the servo as possible.

This will help with the servo motor and the relatively large current drain at startup.

Mark