Ok, I know I'm being a bit lazy here, I could just do some experiments to get my answer but I'm guessing someone here already knows the answer to this.
How far can a servo be from the microprocessor that's controlling it? Or to put it another way, how far can the PWM signal travel down wires without being degraded?
I've got a project that requires a number of servos to be controlled from a central point. The servos are all several meters (10-20 or more) away from each other and from the controller. The project itself should be quite simple if I can control all the servos from a single arduino at the center. If the PWM is going to degrade over the distance, then I may have to have an arduino next to every servo, if I have to do this then the project suddenly gets a lot more complicated.
The degradation of the control signal will depend on the conductivity of the wire and the pick-up of interference. With a sufficiently large gage of wire and shielding it should work at many meters. Will you provide power local to the servos? Running power down a long wire will require much heavier gage wires.
I was hoping to avoid shielding, the application is weight-critical, so I was hoping normal wires would be fine.
Running power down a long wire will require much heavier gage wires.
True, but at the relatively low voltages I'm using (and bearing in mind the servos only require occasional operation for a second or two) I'm hoping the wire loses will be minimal. The plan is to transmit 12 volts down the wire and use a 6v regulator to power the servo, that way even if the transmitted voltage drops to say, 8 or 9 volts, then the regulator should still provide enough oomph to power the servo.
The plan is to transmit 12 volts down the wire and use a 6v regulator to power the servo, that way even if the transmitted voltage drops to say, 8 or 9 volts, then the regulator should still provide enough oomph to power the servo.
I've done something somewhat similar using four conductor telephone wire, but the distance was maybe 15'. You might can put a large capacitor at the servo to hold a little power for the servo.
The problem we used to have with rc servos with long leads like say 3 metre glider aileron servos, was that under certain circumstances, the capacitance of the leads would cause the servo to shift position without any changes to the Tx input.
The fix was to make up and fit "servo buffers".
Check the reasoning as described in the attachment pdf.
I have experienced 'servo judder' myself but I had previous assumed that it was down to either poor quality servos, or a poor quality PPM signal coming from the controller. I had not considered line capacitance. This must be a common problem, you'd have thought that the manufacturers would have added built-in buffering.
This must be a common problem, you'd have thought that the manufacturers would have added built-in buffering.
It probably isn't a problem for 99.9% of hobby servo users. In typical RC setups the servo wiring is fairly short. If you want a specialty servo, be prepared to spend more than $10.
(and bearing in mind the servos only require occasional operation for a second or two)
This has nothing to do with the problem. It centers on two things
The rise/fall time of the control signal, the longer the wire the greater the the tim it takes to rise/fall but the servo will only see the signal when it passes a certain value the servo may allso require that it does so within a certain time frame,
and
The size of the signal when it gets to the servo, if the "hight" voltage of the signal is to small when it reaches the servo than it will not be seen at all.
If you can find good enough data sheets for your servos you may be able to work it out. But it will be far quicker to tryt it and see!