servo motor pulse width

[tinto]

what should be the minimum delay between two successive pulses given to a servo motor? Or, what should be the maximum duty cycle of the pulse train given to a servo motor, if the required pulse width to bring the servo motor to middle position is 1.5ms?

 

[klystron]

Refer to https://www.edaboard.com/threads/88773/

 

[tinto]

I had referred that. But I am not clear. I need to know the minimum value for t2. How short it can be? Please refer the figure.

 

[vikas_verma]

generally a servo works with t1 around 1 to 2 ms.
and if T= t1 + t2
then T shuld be around 18 to 20 ms.
means a servo requires 50Hz refreshing rate!

 

[tinto]

can t2 be small as 5ms? Will servo motor work with higher refreshing rate?

 

[FvM]

can t2 be small as 5ms? Will servo motor work with higher refreshing rate?

It depends on the servo design. With some analog servos, the control loop gain may increase, possibly causing instability. Some won't accept the control signal at all, others will. 50 Hz repetition rate is preferred, because all servos can work with it.

 

[jpanhalt]

can t2 be small as 5ms? Will servo motor work with higher refreshing rate?

FvM is absolutely correct and this comment is not meant to disagree. However, using a Microchip MCU, one may have have problems using the built in PWM to drive a hobby servo with a 50Hz refresh rate, depending on your clock speed. I tested some Hitec analog servos a few years ago, and they will work up to a refresh of about 100 to 120 Hz. Above about 140 Hz, they became unstable. For that particular project, I used a higher refresh rate. For other projects, I have just generated the servo PWM by software.

John

 

[vikas_verma]

Yes FvM is right and i would suggest you to do some practical with servo!

 

[tinto]

I did some experiments with my servo (TowerPro SG 5010) by giving 50Hz refresh rate as well as 100Hz. Both works with my servo. If so how can I understand whether my servo is working properly with 100Hz refreshing rate or my servo is neglecting extra pulses? I don't need to spend time giving extra pulses in my program. Will higher refreshing rate increases the power consumption of the servo?

 

[jpanhalt]

What are you using to test your servo? That is, is it something like one of the many NE555 clones or a microcontrller? If you can change the repeat rate easily (e.g., with the 555), you could try something like an 80 Hz repeat rate. If it is dropping pulses, then it would be seeing only 40 Hz and its response -- rate of movement -- should be slower. Slowing the repeat rate is one way to slow servos down, but it is not very effective, and there are better ways available now to do that.

You could also go up in repeat rate until it becomes unstable. Let's say that is 140 Hz. Then test at half that rate. If it is stable there, then it probably is not dropping pulses. I suspect it is not dropping pulses.

John

 

[tinto]

I am now using pic microcontroller to give pulses to the servo motor. Let me try giving higher pulse rates too as you said.

 

[vikas_verma]

50Hz refreshing rate is sufficient and power consumption doesn't depend on refreshing rate,actually motor's motion causes it so!