50hz Notch filter ain't working...

Discussion in 'General Electronics Chat' started by chunkmartinez, Jun 17, 2013.

  1. chunkmartinez

    Thread Starter Senior Member

    Jan 6, 2007
    180
    1
    http://www.radio-electronics.com/info/circuits/opamp_notch_filter/opamp_notch_filter.php

    In the link above I used their 50hz example with 1% resistors, and 5% tolerance polyester capacitors. The calculation shows 49hz...which is pretty close to 50...I know the Q is supposed to be high on notch filters which is actually what I need but not too high. Does anyone have an idea of the Q of the posted circuit? And are my tolerances good enough?

    My issue is I need to cut out 50hz and only 50hz well I mainly just need there to be a good cutoff before about 100hz. Well, I tested the output on my oscope after running a 50hz and I get about 200mV starting and it keeps climbing with volume. I used a voltage divider of 8.6k to 1.5k to reduce the output of a 17VRMS(before clipping) swing from the amp. I am really trying to get the harmonics of 50hz to show on the scope. The harmonics will come in after beeing near/in the clipping region but I don't think it show any significant distance in amplitude from the fundamental, it just seemed proportional to volume turn.
     
    Last edited: Jun 17, 2013
  2. crutschow

    Expert

    Mar 14, 2008
    12,988
    3,226
    The notch voltage will be proportional to the input voltage, if that's what your are seeing. Did you expect otherwise?
     
  3. chunkmartinez

    Thread Starter Senior Member

    Jan 6, 2007
    180
    1

    Well you're right, I'm just thinking that once I turn the volume up into it's limits where harmonics are created, I should see a decent voltage jump right? The harmonics should be within the "passband" or not nearly as attenuated as the fundamental right? Maybe I need to run a sweep to see if my notch is a little off then I though?
     
  4. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    What is the application? Will the signal eventually go to an ADC and digital sampling?

    If so it's quite easy to digitally filter the 50Hz and keep the harmonics, and do it very effectively.
     
  5. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    Very close to 0.5.
     
  6. t06afre

    AAC Fanatic!

    May 11, 2009
    5,939
    1,222
    You should tell us why you need need this filter. And also about your setup. It is often much better to apply other solutions like proper cabling and shielding etc. Before applying a notch filter
     
  7. chunkmartinez

    Thread Starter Senior Member

    Jan 6, 2007
    180
    1
    It is a harmonic detector circuit. I do not need to filter out 50hz for performance but rather for detection which will show on an LED. It allows users to set a vaeriable gain knob on amplifiers so that if it is set wrong and the amp reaches clipping, the harmonics will be able to pass through only and emit an LED. A 50hz sinewave will be played through the audio setup, the volume on the preamp will be turned up, the adjustable gain on the power amp will be adjusted so that it can be set as high as possible without distortion.

    This project is not to help myself setup my audio system but it's a tool that I wanted to make for more people. My amp gives 17VRMS swing before clipping. Others can reach somewhere like 35VRMS at most...
     
  8. t_n_k

    AAC Fanatic!

    Mar 6, 2009
    5,448
    782
    I always believed audiophiles could detect audio distortion below the miniscule levels just by listening. They would be screaming blue murder with the distortion levels you are envisioning.:eek:
     
  9. THE_RB

    AAC Fanatic!

    Feb 11, 2008
    5,435
    1,305
    I'm doubting the viability of that system. You are better off detecting amp clipping by measuring the clipping inside the amp itself, there are lots of clipping detector circuits on the net.

    If you want to detect clipping by just testing at the speaker terminals on any amp (that's what you're doing?) then how will it work?

    If you inject a 50Hz sine into the amp, your clipping detector will only measure clipping of that exact amplitude signal. Once they connect their audio etc to the amp afterward it can still clip, as there's no guarantee the audio will have the EXACT same amplitude as your test sine did!
     
  10. Veracohr

    Well-Known Member

    Jan 3, 2011
    549
    75
    Harmonic distortion will generally be MUCH lower amplitude than the signal. If you have the test signal decently filtered out, you may have to amplify the output quite a bit before you start seeing noticeable voltage.
     
  11. vk6zgo

    Active Member

    Jul 21, 2012
    677
    85
    In Broadcasting a device called a Noise & Distortion test set is used.

    A low distortion tone is fed into the DUT,& the N&D set is fed from the output (or a low level sample of the output in some cases ).

    If the signal is at say 50Hz,the test set is adjusted to null out the component at that frequency.

    These are very high Q notches---the fundamental (in this case 50Hz ) is reduced by 70dB or better.
    What remains is the Total Harmonic Distortion at that frequency,plus noise.

    Before the notch is switched in,the meter scale is set to 100% or 0dB,depending on which convention is used.

    With the notch in circuit & nulled,the range switch is reduced until a readable value of Distortion is present on the meter.

    Noise is read in a similar manner,but with no input signal,& the input terminated in its correct input impedance.
    .

    It is some years since I worked in Broadcasting,so they may use audio spectrum analysers now instead.
     
    chunkmartinez likes this.
  12. nodistortion

    New Member

    Oct 30, 2013
    1
    0
Loading...