Microphone interference problem - Stumped!

Discussion in 'The Projects Forum' started by bluewave, Feb 7, 2012.

  1. bluewave

    Thread Starter New Member

    Feb 4, 2012
    8
    0
    I have a simple circuit that converts sound level to a DC voltage - attached - for reading with a PIC ADC.

    My circuit works fine on battery power. But if I use a double-insulated AC->DC adapter (wall wart) without earthing the circuit, the microphone picks up 50Hz interference (mains AC is 50Hz here in New Zealand). The opamp then amplifies the 50Hz.... If I earth the circuit, via my scope for example, the interference disappears and it works perfectly on the plug pack.

    I've attached a couple of scope outputs - one with the circuit earthed and the other unearthed - green is the opamp input (x1), red is the opamp output (x10). The 50hz signal is apparent when earthed, but the extra mV or two when unearthed is causing the problem. Plus the level of interference varies a lot when unearthed.

    I am a newbie, but I have looked and looked and can't find a similar problem anywhere? Common wall warts without an earth pin mean connecting signal ground to earth isn't so easy either, if that is a solution?

    Thanks for any suggestions!!
     
  2. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,750
    759
    First add 2 caps 100nf at the input of the regulator and also at the output. This will help decouple it. Plus you should have around 470uf cap or more as the input side of the regulator.

    Next the mic cable should be a shielded type. This ground shielding will eliminate noise pick up from the cable u are using to connect mic to the opamp. Just a cable won't do in these kinda application. Low level audio must be connected using properly shielded microphone cable.
     
  3. bluewave

    Thread Starter New Member

    Feb 4, 2012
    8
    0
    Thanks very much for the suggestions! Really appreciated it.

    I have added the extra bypass caps as you suggested (including 1000uF on the input). Unfortunately this didn't improve things. So I am guessing the problem isn't a lack of decoupling in the local power supply.

    I haven't used a shielded microphone lead - but the lead from the microphone to the PCB is only 2 inches long, and is a tightly twisted pair.

    I think I am seeing some type of 'antenna action' in the microphone, which I don't understand at all. When I move my hand close to the microphone, the 50Hz interference gets worse. If I remove just the microphone from the circuit, the interference disappears (so is not simply passing through the power supply to the opamp). As mentioned before, when the circuit is connected to earth, all problems disappear. Does this ring any bells? It is a real puzzle to me!

    Thanks!
     
  4. bluewave

    Thread Starter New Member

    Feb 4, 2012
    8
    0
    ...I should add - I'm not sure how a ground shield would work here, because with a double-insulated plugpack, I don't have access to earth.

    There must be a way - because lots of things use wall warts and have microphones (e.g. cell phones!). What am I doing wrong?

    :confused:
     
  5. R!f@@

    AAC Fanatic!

    Apr 2, 2009
    8,750
    759
    No..Use a a shielded microphone cable to connect the mic.

    get an old CDROM audio cable and use it.

    As for plug pack, I am not sure about it's filtering capabilities. So I cannot comment on that. If you have added the caps I recommended then it shud work. I think u left something out.
     
  6. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    Of course your wires from the mic to pin 3 of the opamp is an antenna for 50Hz mains hum because the voltage gain is very high at 456.

    ALL microphones need a shielded audio cable from the mic to the opamp input pin.
    If you wired the circuit on a breadboard the its contact strips and connecting wires are also antennas for mains hum.

    The shield in a shielded audio cable connects to the grounded case of the electret mic and to 0V in the circuit, not to earth. The signal wire must be covered by the shield as much as possible at each end.

    Maybe you need such a high voltage gain because the output diode is in the wrong place. Make an Active Rectifier Circuit like in my modification:
     
  7. bluewave

    Thread Starter New Member

    Feb 4, 2012
    8
    0
    Thanks very much Audioguru.

    On your and R!f@@'s advice I have now installed a shielded cable to the electret insert as you describe, but that didn't help unfortunately.

    You are right that the gain is high, and the active rectifier is a good suggestion to bring it down. The 2.2k resistor is actually now a pot, and the gain is set around 300. Problem is, when the circuit is unearthed, the microphone signal and the 50Hz noise are similar magnitude - so I don't think I can fix things by reducing the gain - I also lose my signal.

    The problem has something to do with using a double-insulated AC->DC plugpack while the circuit is unearthed. When I earth the circuit, the 50Hz noise goes completely to zero and everything works great. I have tried several plugpacks, so is not specific to the one I am using. I am surprised this is not a more common problem, so I must be doing something very odd. Thanks very much for taking the time to make suggestions.

    I have made another discovery - I can reduce the 50Hz quite a lot by connecting circuit ground to a 1" x 4" x 1/4" piece of aluminium sheet. The sheet is completely unearthed, but still reduces the 50Hz a lot - not as much as actually earthing the circuit, but by enough to be helpful. I have no idea why???

    Thanks again!
     
  8. KJ6EAD

    Senior Member

    Apr 30, 2011
    1,425
    363
    I think you need a filter capacitor on the electret element power input (bottom end of the 4.7kΩ resistor in AG's schematic) to ground. It would have to be a largish value like 10µF to filter out such a low frequency.
     
    Last edited: Feb 8, 2012
  9. mhe

    New Member

    Aug 10, 2011
    10
    0
    Hi BlueWave

    I have the same problem, same symptoms, and almost the same setup, so I would like to hear if you got any progress removing the 50Hz noise.
     
  10. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    If it filters out 50Hz then don't you think it will also filter out ALL sounds from the microphone?
     
  11. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    I have made many mic preamps and none produce any hum.

    The 4.7k resistor that powers the mic is fed from +5V. But the +5V comes from an AC-DC power supply that might have lots of 50Hz hum.
    So filter the +5V that feeds the 4.7k resistor.
     
  12. KJ6EAD

    Senior Member

    Apr 30, 2011
    1,425
    363
    Good point.
     
  13. bluewave

    Thread Starter New Member

    Feb 4, 2012
    8
    0
    Hi MHE,

    I'm not sure I have the answer...but I do have some ideas and a solution of sorts! What puzzled me is that there is heaps of consumer audio gear (e.g. cellphones...) that run off wall warts and have microphones...without 50hz interference.

    My solution has been to mount the circuit on a 100 x 100mm 3mm aluminium plate, that is connected to signal ground. This has got rid of most of the problem - enough that the circuit works. I can still see a little 50hz noise on my scope, but far less.

    I am almost certain the problem isn't simply 50hz noise passing through my power supply (the noise is completely absent from the supply rail when the microphone is disconnected) - I think there is some antenna action or capacitive coupling happening, which is not something I understand properly. I am guessing the solution in commercial appliances is proper shielding and the ground plane - of course on a prototype board that could be lacking.

    Did you fix your problem? Good luck!
     
  14. Audioguru

    New Member

    Dec 20, 2007
    9,411
    896
    A microphone has a tiny output voltage. If shielded audio cable is not used to connect the mic to the preamp then of course ordinary wires are antennas that pick up mains hum from the air.
     
Loading...