The power supply filter. Fighting the interference.

Discussion in 'DIY' started by Zampotech, Jan 27, 2019.

  1. Zampotech

    Zampotech Friend

    Pyrate
    Joined:
    Oct 15, 2018
    Likes Received:
    4,195
    Trophy Points:
    93
    Location:
    Moskou
    Home Page:
    NOTE: Please be indulgent to my English

    I live in a big city (Moscow). We don't have a very clean power grid. There's a lot of interference in the power grid. This is probably a common trouble of big cities.

    Sometimes that noise gets right into your ears when you listen to music.

    It made me struggle with this unpleasant phenomenon.

    For a start, we decided to determine the spectrum of the voltage in the power supply.

    To do this, I downloaded the program SpectraPLUS on an old computer.

    Through the step-down transformer supplied voltage from the mains to the sound card.

    And saw a terrible picture.

    [​IMG]

    A lot of noise in the sound range.

    With low frequency noise rectifier power probably cope. High-frequency noise penetrates the ears probably due to the parasitic capacitance.

    I decided to check the industrial EMI filter.

    [​IMG]

    Here is the spectrum at the output of the industrial EMI filter.

    [​IMG]

    The result is close to useless.

    The use of such a filter for audio devices is equivalent to the use of a navel warmer.



    My next step was to use a self-made simple filter of inductance and capacitance.

    I used a choke from an old computer power supply and a capacitor from fluorescent lamps.

    [​IMG]

    [​IMG]

    [​IMG]

    Calculations have shown that these two elements form a low-pass filter with a boundary frequency of 367 Hertz. That's enough.

    [​IMG]

    After a quick build, I checked the spectrum after a homemade filter.

    [​IMG]

    Result. A self-made filter of two simple elements works more efficiently for audio devices.

    Research will continue. Now I have to put the filter in the case and listen for a while.
     
  2. Zampotech

    Zampotech Friend

    Pyrate
    Joined:
    Oct 15, 2018
    Likes Received:
    4,195
    Trophy Points:
    93
    Location:
    Moskou
    Home Page:
    Continuation of research.
    I assembled a simple integrator from a 56K resistor and a capacitor of 1 uF Measured the voltage at the output of the integrator.


    [​IMG]


    [​IMG]

    It was found that the power supply contains direct current and infra-low frequencies.
    Constant voltage has reached up to 110 mV. An average of 20-40 mV
    This is a very large value for transformer power supplies.
    Toroidal transformers are very sensitive to direct current through the winding

    Now I understand why my little amp sometimes gets very hot.

    We will look for a cure for this disease...

    PS I do not know how to upload a video with full-scale measurements
     
    Last edited: Jan 29, 2019
  3. Priidik

    Priidik MOT: Estelon

    Pyrate BWC
    Joined:
    Sep 27, 2015
    Likes Received:
    2,153
    Trophy Points:
    113
    Location:
    Estonia
    That can be relevant topic also when somebody is using an angle grinder (or any universal motor with variable load) on the same phase with you and the line impedance is high-ish. I could literally hear the motor through headphones, hahaha.
     
  4. Zampotech

    Zampotech Friend

    Pyrate
    Joined:
    Oct 15, 2018
    Likes Received:
    4,195
    Trophy Points:
    93
    Location:
    Moskou
    Home Page:
    In pauses I hear rustling, crackling, howling. Interference is quiet, but sometimes annoying. Perhaps the source of interference is a children's clinic, located in the neighborhood. They have a large number of different medical devices.

    In the interference spectrum I sometimes saw some strange frequencies, for example 375 Hz, 1220 Hz with a rather large amplitude.
     
    Last edited: Jan 29, 2019
  5. k1arg

    k1arg Facebook Friend

    Contributor
    Joined:
    Oct 12, 2015
    Likes Received:
    153
    Trophy Points:
    43
    Location:
    London, UK
    It might be quick to try something along the lines of this for DC.
     
  6. Zampotech

    Zampotech Friend

    Pyrate
    Joined:
    Oct 15, 2018
    Likes Received:
    4,195
    Trophy Points:
    93
    Location:
    Moskou
    Home Page:
    Unexpected continuation

    Prehistory
    I was too lazy to make a case for my simplest filter and assembled it on a piece of plywood. I gave the filter to a friend so that he could tell his opinion as objectively as possible. My friend has "Golden ears" and he meets once or twice a month with headphone lovers in an artistic cafe.
    Its conclusions: after application the sound became more volume, the amplifier works "blacker than black". It's probably improved the transmission of micro-sounds.
    He asked me to make him the same filter in a more civilized form for demonstration at the next meeting.
    My friend by his own admission "rukozop". It's slang. Means a man whose hands are growing out of his ass.

    [​IMG]

    I made him a new filter with additional features.
    The new filter has a standard EMI filter, overvoltage protection and a low frequency filter.

    A complete filter circuit

    [​IMG]



    PCB


    [​IMG]

    The filter turned out at low power, up to about 100 watts. But the headphone amp be enough power.


    Experimentally found that the most evil interference starts from 800 Hertz and above.
    Filter interference below 800 Hz does not make sense. they have no effect on anything. Unless, of course, they have enormous power.

    Now I'll wait for the collective opinion

    PS This filter will only be suitable for transformer power supplies. For an amplifier with pulsed power supplies, this filter will be inefficient.

    PPS If you're interested, I can show you how the filter is made.
     
    Last edited: Feb 5, 2019

Share This Page