0
\$\begingroup\$

I have a micro-controller board, driving small DC servo motors and I control the angle via a PWM signal (50 Hz) and everything is alright.

However, there is an oil gas motor, running next to the board, causes the DC servo motor to oscillate and a noise enters to my board.

I decided to to add a low-pass filter (L-C-L) after the DC regulated voltage (5v). I do not know the frequencies of the noise which are coming from that motor. So I am thinking to put a low-pass filter with 10 Hz as a cut-off frequency.

Would my new filter attenuate that PWM 50 Hz signals that drive my DC servo motor. Furthermore, the controller is running on a 20 MHz, would the filter degrade my controller performance ?

If the filter would not work, I guess I have to know the absolute range of frequencies I should band and then use EMI beads which block only a specific range of frequencies. But the problem is that the range of noise frequencies can not be determined.

\$\endgroup\$
8
  • 1
    \$\begingroup\$ If you put the filter in front of your regulator, you won't have to worry about any of your concerns. Any reason why you put it after? Also be aware a LCL filter can resonate easily. \$\endgroup\$
    – MadHatter
    Commented Oct 10, 2019 at 12:30
  • \$\begingroup\$ If I put the filter in front of my regulator, can it reject the noise of the motor from entering the PWM signals? The motor is a gas-type and the EMI can enter to any part of the circuit. That's why I was thinking to filter the 5v power supply. \$\endgroup\$
    – mmyname
    Commented Oct 10, 2019 at 12:39
  • \$\begingroup\$ You can dampen the filter, with an R in parallel with the inductor. Use R o value sqrt( L / C ). Thus 1 milliHenry and 1,000 uF is sqrt(0.001 / 0.001) needing 1 ohm. You can experiment with the value; 3.3, 4.7, 6.8 10. The resistor might get hot, if lots of trash to absorb. Make sure the inductor is large enough (core volume) to not saturate with the DC current needed. \$\endgroup\$ Commented Oct 10, 2019 at 12:44
  • \$\begingroup\$ The spark noise is > 10kV impulse at low frequency RPM but fast rise time > 10MHz noise fill. Driver is possible 50Hz 1~2ms pulse servo with some excess rise time <1us that can be < 1ms -- with say 100 ohm driver that would be a 0.1uF shunt cap on servo signal. Try 0.01uF \$\endgroup\$ Commented Oct 10, 2019 at 12:51
  • \$\begingroup\$ Do you mean that the noise of the motor is > 10 MHz, so an RC low pass filter with 160 Hz cut-off frequency on servo signal can solve it? What if I want to put a filter for the whole micro-controller circuit, there are other driving signals who should avoid noise also. \$\endgroup\$
    – mmyname
    Commented Oct 10, 2019 at 13:07

0

Your Answer

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