I'm working on a project using a Savox SB-2290SG servo motor. The datasheet I received is only 3 pages long, and includes limited information on the control interface. I've posted the relevant datasheet snippet below.

While the datasheet does list the acceptable range of pulse widths, it never mentions the overall duty cycle or PWM frequency. For example, if I want to command this servo into the neutral position, I know I need a pulse of 1500 microseconds, but how long should I wait between pulses? Another 1500 microseconds? 3000 microseconds? 10,000 microseconds? It doesn't appear that that PWM wave I need is fully defined by this datasheet.

Does 'neutral position' imply a 50% duty cycle? Is the datasheet missing key information? Or am I misunderstanding something about how to control servo motors?

enter image description here

Update: I reached out to the manufacturer, and they informed me that they use a 333 Hz frequency during their testing.


1 Answer 1


Does 'neutral position' imply a 50% duty cycle?

No. The RC servo standard is not duty cycle despite being called PWM. It is the actual time length of the pulse and not its duty cycle that they care about. 1.5ms is center. 1 and 2ms are the extremes.

That's what this says:

enter image description here

Most servos expect a pulse at least every 20ms (50Hz). For analog servos, the pulse is not remembered and is directly used to drive the servo. The motor is only moved whenever the pulse comes in so this directly affects response and torque. Too low a frequency and analog servo will be sloppy and weak. Too high and it overheats.

Digital servos can remember the pulse and independently update the motor at a different frequency so the possibility exists that they will work with frequencies lower than 50Hz until it becomes so low the servo thinks the signal has been lost and goes into fail-safe. Some digital servos can handle as high as 333Hz. I expect yours can handle 333Hz. This was not listed.

  • 1
    \$\begingroup\$ While that is true, it's also critically important to realize that 1.5 ms at best an average, not a standard. Different servo models have different neutral points, different examples of the same servo have different neutral points, the same servo may even have different neutral points depending on temperature and age. These come from a thoroughly analog tradition of usage: if you didn't like where the control surfaces were during preflight, you moved the trims. If you ran out of trim range, you popped the clevis off and gave it a turn or two or rotated the servo horn on the spline. \$\endgroup\$ Aug 5, 2020 at 16:44
  • 1
    \$\begingroup\$ @ChrisStratton Yeah. it's not well controlled or calibrated. \$\endgroup\$
    – DKNguyen
    Aug 5, 2020 at 18:13
  • \$\begingroup\$ "The RC servo standard is not duty cycle despite being called PWM. It is the pulse length (actual time) not width (duty cycle) that they care about, not duty cycle." makes no sense. While many PWM systems may only care about duty cycle, it is modulation of the pulse width, not modulation of the duty cycle, that defines PWM. You can equally well build a 1-bit DAC (which is the technical description of the system caring about duty cycle) holding any one of on-time, off-time, or interval constant. Or you can do dithering. Those are often called PWM, but this servo control actually is. \$\endgroup\$
    – Ben Voigt
    Jun 5, 2023 at 18:42
  • \$\begingroup\$ @BenVoigt Changed the phrasing. \$\endgroup\$
    – DKNguyen
    Jun 5, 2023 at 18:53

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.