Electret Microphone pick up 50Hz noise

Discussion in 'The Projects Forum' started by mhe, Feb 24, 2012.

  1. mhe

    Thread Starter New Member

    Aug 10, 2011
    10
    0
    Hi

    I am debugging on a microphone-preamplifier and discovered some 50Hz hum when the microphone is connected. When the microphone is disconnected the RMS value is halved. When I short-circuit the inputs the noise disappear and I am down to less than 1mV at the output which correspond to the noise reported in the datasheet by the opamp manufacture and noise in resistors etc.


    The system is powered by batteries (3.3V).
    the gain in the active filters are 270 times and have a 3dB cutoff FREQ at 1250Hz.
    The amplifier is build on a 4layer print, area 1.8 square cm. GND in hidden layer, Vdd in outer layers.
    The Noise are measured to be around 5.5mV RMS at the output from the amplifier or (5.5/270) = 20μV RMS at the input, when shielded by alu-paper.
    When un-shielded the noise are measured to be around 8.5mV RMS at the output from the amplifier or (8.5/270) = 31μV RMS if I move closer to my desk with Monitor/PC etc the RMS value Double 16mV RMS.
    When the Mic are removed and the OpenLoop RMS noise is 3mV at the output or 11μ at the input.

    The attached frequency spectrums covers 1000Hz. The Wall-supply in Denmark is 50Hz.


    Is it normal for a Electret Condensor Microphone to capture surrounding noise from fluorescent lamps/monitors etc. and how much should I expect ?

    Are there some optimization to the circuit, I have tried to change the Capasitor C7 on the schematic to 12.2μF but it did not seem to enhance the performance. ?


    Thanks.
     
    Last edited: Feb 24, 2012
  2. bertus

    Administrator

    Apr 5, 2008
    15,649
    2,348
    Hello,

    Are you using a shielded cable to connect the microphone?
    If you use normal wires you might expect some pickup.

    Bertus
     
  3. mhe

    Thread Starter New Member

    Aug 10, 2011
    10
    0
    Bonus Info

    I have now changed the Mic both with another lowcost electret condenser Mic and a hearing aid mic (50$).

    When measuring with the other lowcost electret condenser Mic there was no difference.

    With the hearing aid mic (3 pole) Mic the 50Hz noise is gone and it has a flat frequeccy response. The RMS noise is 7mV at the output, although there is 7dB more gain, because of the specified intensity on the hearing aid mic (33dB vs 40dB). In addition to that there is no change in RMS when I move the Mic closer or fare away from my other electronics, eg. monitor.

    I have also tried to disassemble a lowcost Mic and connect it to the amplifier and it did only measure 50Hz noise, huge amount of 50Hz noise.
     
  4. bertus

    Administrator

    Apr 5, 2008
    15,649
    2,348
  5. mhe

    Thread Starter New Member

    Aug 10, 2011
    10
    0
    Hi Bertus

    Thanks for the reply, and the nice link, looks very interesting, I will study them in more detail in the weekend.

    >>Shielded cable..
    Yes i have tried that, but it do not make any difference. Then I tried placing a finger on the microphone-house and this work on the cheap ones, on the expensive hearing-aid microphone the sound is just fine all the time.
     
  6. mhe

    Thread Starter New Member

    Aug 10, 2011
    10
    0
    Actually most of the noise is gone when I touch the microphone, the noise fall down with 20dB.
     
  7. bertus

    Administrator

    Apr 5, 2008
    15,649
    2,348
    Hello,

    The cheaper mic should have problems with grounding.
    The expensive one will have better grounding.

    Did you also see the modification of the (cheap) 2 lead mic to a 3 lead mic on the page I gave you?

    Bertus
     
  8. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    The metal housing of a cheap electret mic is its also its ground pin that should be connected to the shield of the shielded audio cable. The shield is connected to the circuit's ground at the circuit board. Touching the shield or mic housing should not change the amount of hum.
    The signal wire of the shielded audio cable should be very short at both ends so that most of its length is shielded.

    I am talking about co-axial shielded cable, not a twisted pair of wires.

    If you can hear your florescent lights buzzing (at 100Hz, not 50Hz) then the mic will also hear them.
     
  9. mhe

    Thread Starter New Member

    Aug 10, 2011
    10
    0
    Hi Bertus and Audioguru Thanks for reply..


    >>(Bertus) Did you also see the modification of the (cheap) 2 lead mic to a 3 lead mic on the page I gave you?
    Did you mean this: Electret microphone Integrated Circuits for High Performance Electret Microphones, pdf file
    Here they argument that audio microphones need highpass filtering with a cut-off frequency around 100Hz to attenuate large low frequency signals...


    >>(Audioguru) The signal wire of the shielded audio cable should be very short at both ends so that most of its length is shielded. I am talking about co-axial shielded cable, not a twisted pair of wires.
    The wire is very short, only unshielded by the displacement of the MIC connections 2mm. The cable is 3.5cm long. I have tried both shielded cable, twisted pair and parallel wires but it makes no difference or the difference is very small compared to the noise picked up by the microphone-house.
    >>If you can hear your florescent lights buzzing (at 100Hz, not 50Hz) then the mic will also hear them.
    When using the hearing-aid microphone I can hear florescent lights buzzing.
     
  10. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    The article does not describe an "audio" microphone.
    Instead it uses a highpass filter at 100Hz to reduce wind noise for a mobile telephone where there is no audio fidelity. Did you notice that they also chop frequencies above only 3kHz? No lows and no highs.

    Your mic is picking up the buzzing sound from fluorescent lights. Replace them with quiet lights.
    My compact fluorescent light bulbs operate at about 40kHz and are extremely quiet.

    I made a VU meter for my living room. The room has 4 compact fluorescent light bulbs.
    The VU meter uses a cheap 2-wires electret mic. The preamp IC is an MC33172 which is not low noise so it produces some hiss. The circuit board is in a plastic case.
    If I hold my breath and the furnace and refrigerator are not running then the VU meter shows an extremely low sound level of 40dB. No hum and no buzzing.
     
  11. mhe

    Thread Starter New Member

    Aug 10, 2011
    10
    0
    Hi Audioguro

    Thanks again.

    >>Your mic is picking up the buzzing sound from fluorescent lights. Replace them with quiet lights.
    That was also what my thought, but when I turn-off the light there is no change in the 50Hz part while the 100Hz part is reduced.

    On the other hand, if I move the microphone away from my LCD-Monitor or speaker the 50Hz hum is reduced and when I put a hand on the side of the mic-house the 50Hz signal is reduced also.

    So I'm pretty sure that the Microphones pick's up 50Hz hum and it is not because of sound.

    The PCB without microphones have been evaluated as well and there is no noise when the microphone is removed.
     
  12. Robyn

    New Member

    May 1, 2013
    4
    0
    This is an old thread but I am upping as I have the same problem and the same exact symptoms.

    I built myself a pair of earbud microphones to record my street wanderings. I bought some $2 earbuds, replaced the speakers with some cheap electret microphones from the electronics shop, and hooked them facing outwards. I just recorded my trip to the supermarket and it's awesome. But,

    I'm getting hell of a lot of hum when I'm close to anything electric. I can even hear it from the automatic cash registers and the neon lights.

    Now I've experimented a bit and if I hold the cable close to my computer I don't get hum. It seems to be only the capsule that catches the electric field.

    If I grab the earbud (as in the plastic, doesn't even need to be the actual shell of the mike) in my hand the buzz gets way fainter.

    If I touch the aluminium body of my computer while holding the earbud it goes down even more.

    Because the issue seems to come from the actual capsule I doubt that a shielded cable will help. Some circuits involve cutting the trace that links one of the pins of the capsule to the shell and a moderately complex circuit.

    I have only connected my two electrets to a stereo jack that goes in the back of the H4N. The recorder provides about 2.5V of phantom power through the tip (left signal) and the ring (right signal).

    Anyone has a theory?
     
  13. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,754
    760
    Balanced input ckt might help
     
  14. radiohead

    Active Member

    May 28, 2009
    474
    31
    If you are using an AC to DC adapter, you will have to filter the input voltage to eliminate the hum. Try wrapping the voltage input wire a few turns around a ferrite bead or ferrite donut. Do a google search for power line noise filters.
     
Loading...