Active or passive low-pass filter?

Discussion in 'General Electronics Chat' started by exleonhart, Mar 15, 2009.

  1. exleonhart

    Thread Starter Member

    Nov 19, 2008
    10
    0
    Let say I need to filter out a noise at 100Hz. Which kind of low-pass should I choose? :confused: I heard that active low-pass is better for low frequency. Is this right?
     
  2. bertus

    Administrator

    Apr 5, 2008
    15,648
    2,347
    Hello,

    It also depends on the application.
    For low frequencies the passive filters might get big.
    With active filters it most times can be created smaller.

    Greetings,
    Bertus
     
  3. exleonhart

    Thread Starter Member

    Nov 19, 2008
    10
    0
    Thanks for the quick reply. Actually it was to filter out noise due to vibration.
     
  4. bertus

    Administrator

    Apr 5, 2008
    15,648
    2,347
  5. KL7AJ

    AAC Fanatic!

    Nov 4, 2008
    2,040
    287
    if the offending signal is EXACTLY 100 Hz, you might be better off with a twin-t notch filter...you can get infinite attenuation that way.
    eric
     
  6. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Maybe the mains is 50Hz and the main filter for the full wave rectifier capacitor is broken.
    Then instead of notching out the 100Hz ripple, why not fix the problem instead?
     
  7. Gus 67

    New Member

    Jan 21, 2009
    3
    0

    Sounds like a smart plan....I bet he's right..
     
  8. Gus 67

    New Member

    Jan 21, 2009
    3
    0
    Check your power supply first.Find the value of the filter cap and jump a known good cap of the same value in parralel and see if your problem goes away.
     
  9. exleonhart

    Thread Starter Member

    Nov 19, 2008
    10
    0
    Actually we have tried to apply some rubber pad to reduce the vibration but there is still a peak at 100Hz (taken by oscilloscope FFT function)
     
  10. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Vibration? What kind of equipment is having this noise problem?
     
  11. exleonhart

    Thread Starter Member

    Nov 19, 2008
    10
    0
    We were making a robot that use gyroscopic principles. The gyroscope is actually a fast flyingwheel. The problem of fast speed is that it cause the whole structure to vibrate. The angle-sensor (Crossbow CXTA01) mounted on the platform therefore gives out a lot of noises. Doing the FFT using oscilloscope I found that there is a peak at 100Hz and I want to get rid of this signal using a lowpass filter. Seem like this is the best way if not touching any mechanical modifications. I am just not sure which is the best filter to use.
     
  12. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Are you saying the gyro is vibrating? Is it out of balance, or is it bearing noise?
    Keep in mind that lowpass filters have delay, which might complicate any feedback loop that they are within.
     
  13. exleonhart

    Thread Starter Member

    Nov 19, 2008
    10
    0
    The gyro is spinning fast and as it attached to a frame, the frame vibrates, which causes the readings of the tilt sensors (measure angle) fluctuate.

    Thanks for the reminder about the phase lag. Is there a reasonable solution to suppress this vibration noise?
     
  14. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    A filter might work. It depends on your system. I think I would try to tackle the vibration problem. How much does this flywheel weigh, and how fast is it spinning? 6000 RPM?
     
  15. exleonhart

    Thread Starter Member

    Nov 19, 2008
    10
    0
    The flywheel weights 1.02kg and it max speed is 7000 RPM. We normally run at half the max (3500-4000RPM). Is there a direct relationship bw vibration freq and speed?
     
  16. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    Of course there a direct relationship between vibration freq and speed. If your noise frequency (100Hz) and amplitude are independent of flywheel RPMs, then I would think that the noise is coming from something else. What is your mains frequency?
     
  17. exleonhart

    Thread Starter Member

    Nov 19, 2008
    10
    0
    The test was done quite some time ago. I will re-do the FFT tomorrow and get the result back to you then. Thanks a lot for your help.
     
  18. bertus

    Administrator

    Apr 5, 2008
    15,648
    2,347
    Hello,

    When you divide the RPM's by 60, you will get the rotating frequency in Hertz.
    So 6000 RPM will give 100 Hz
    and 4500 RPM will give 75 Hz.

    Greetings,
    Bertus
     
  19. exleonhart

    Thread Starter Member

    Nov 19, 2008
    10
    0
    Hi, sorry for the late reply. Here is the FFT I have from osilloscope with 1kHz sampling rate, with the frequency spectrum spans from 0-500Hz, centre frequency is 250Hz.
    The max speed is 7000RPM ~ 116.6 Hz
    Running at 20% speed:
    [​IMG]
    Running at 40% speed:
    [​IMG]
    Running at 60% speed:
    [​IMG]
    Running at 80% speed:
    [​IMG]

    As you can observe, there is a peak running from 75 Hz (20% speed) to 200Hz(at 80% speed). I think this peak has significant contribution.
    But there are lots of other smaller peaks which span the whole spectrum as well. Can I play safe by designing a filter that has very low cut-off frequency, i.e. 10Hz?
     
  20. Ron H

    AAC Fanatic!

    Apr 14, 2005
    7,050
    657
    A 10Hz single pole filter has about 16mS delay. For every pole you add, the delay will increase. Is the filter in a feedback loop?You have to determine whether this will cause a problem or not.
     
Loading...