Controlling speed of Servo motor

[p.gholap93]

i have controlled servo angle position by giving pulsed waveform through atmega 16,But now i want to control its speed too. Is it really possible?
For that what external circuitry is necessary? Is there any separate IC for the same?Or speed can be controlled through programming itself?

 

[Pommie]

Just change the pulse width at a slower rate and it will move slower.

Mike.

 

[p.gholap93]

Just change the pulse width at a slower rate and it will move slower.

Mike.

ya i have done the same but while doing so servo sinks more current.Though i have used high current regulator capable of providing 7A current it gets heated SLIGHTLY. So i thought there may be some other safe & reliable way to do so.
Is there any other alternative?
or that's full and final?

 

[jpanhalt]

There are two ways I know to slow a servo:
1) Decrease pulse repetition rate. I believe that method is limited in what it can accomplish.
2) Using an MCU, when a command to change is received, say from 1.5 mS to 1.8 mS, make the change in small increments separated by delays.

I am not sure which method Pommie meant in his post #2. I suspect it was the second method given above.

John

 

[dknguyen]

How are you able to measure how much current the servo sinks? Servo current tends to be very erratic and not reproduceable, even with the same load and motion. But yes, using PWM to slow the motor will increase current draw given the same amount of output torque. No way around that really.

If you want to reduce current but decrease speed and never need a faster speed, the best way is to increase the great ratio! But that tends to mean another servo which might not even exist. Depending on what you are using the servo for, you can decrease the length of the lever arms and linkages to get the same effect

 

[p.gholap93]

i measured servo current using DMM as ammeter in series with 6V power supply.Servo sinks around 2A current only at the instant when it rotates through its specific angle. So if i rotate it continuously through small angles to reduce its speed, naturally it's going to sink the same amount of current CONTINUOUSLY. this'll cost me my regulator circuit.That's why i asked for alternative option.

 

[p.gholap93]

ya it surely was the second one and i have tried it too and also faced the problem mentioned.
But will you be more elaborate about your first method?
i didn't get it

 

[jpanhalt]

Re. method #1: The pulse repeat rate for standard hobby servos is 50 Hz. That is, for a 1.5 mS centering pulse, the blank space is 18.5 mS. I read about the technique of lengthening that space (i.e., decreasing the pulse repeat rate) to slow down a servo about two years ago. I have not experimented with it, but it came from a university source. I have no idea what range of speed control you can attain with it, but the experiment with an MCU or 555-produced servo signal should be quite easy to do. I suspect the range of control is quite limited and don't really understand why it works.

John

 

[p.gholap93]

i also have tried that method and it also didn't work.
let it be i'm controlling speed by slowly changing pulse width as suggested by Pommie.

 

[dknguyen]

I don't think there's a way really...even if you are controlling the motor directly with your own drive electronics you still run into the problem of PWM reducing speed at the expense of decreased efficiency. It's a fundamental limitation for using PWM (or in your case pulsing) to control motor speed for a constant torque load, because PWM can only ever be on or off between two values.

If it's something like a fan where torque decreases with reduced speed you do reduce current consumption when you reduce speed with PWM...but not as much as if you reduced the input motor voltage due to the completely on/off nature of PWM.

 

[Russ Hensel]

I have been thinking about this a bit because I am planning a course on servo and stepper motors and was thinking that this might be one distinction between the two. I decided it is not really. To control the speed of a servo you simply slowly change the desired position. This has nothing to do with the voltage that powers the stepper or variations in the pulse rate, is is just the basic pwm with the duty cycle slowly changing. I may or may not write some of this up at: **broken link removed** which is meant to be a tutorial on servo control.