Paralleling electret condenser microphones

Discussion in 'The Projects Forum' started by kenev, Jun 20, 2015.

  1. kenev

    Thread Starter New Member

    Jun 19, 2015
    10
    0
    The project: Installing a microphone system inside an accordion using a number of electret condenser microphones (ECMs). I have seen several commercial applications. They invariably use 3 to 10 (!) ECMs for the treble (right hand) part of the accordion and 1 to 3 ECMs for the bass (left hand) part of the instrument. I'm thinking of using 4 microphone capsules for the treble and 2 for the bass.
    My question regards the way the capsules are to be paralleled. Since I have not found a reliable source of information on the Internet to this issue, I have thought of two ways of implementing this: either paralleling all capsules together on each side with a common current limiting resistor and a common output capacitor or (more complicated solution) supplying each capsule with its own current limiting resistor and paralleling all outputs.
    A mini op-amp based preamplifier will be installed inside the accordion, together with the battery power supply, or the preamp will be housed in an external box.
    Any help is greatly appreciated.
    Evangelos
     
  2. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    Each capsule needs to have its own bias circuit.
     
  3. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,777
    1,103
    ....and its own coupling cap to the common output.
     
  4. AnalogKid

    Distinguished Member

    Aug 1, 2013
    4,518
    1,247
    An electret microphone element does not have a low output impedance, but it should work in a traditional analog summing circuit. Six individual *decoupled* bias resistors, six coupling capacitors, six series resistors into an inverting opamp summing amplifier. The summing resistors would like to be 20K or higher because an electret output impedance can be 2K-3K, but that will increase the summed signal noise. Start with 5K and expect to have to increase it 10K.

    ak
     
  5. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,777
    1,103
    Simulation shows it should be possible to connect electrets in series rather than in parallel, using just one bias resistor and one coupling cap.
    SeriesElectrets.gif
     
  6. ericgibbs

    AAC Fanatic!

    Jan 29, 2010
    2,501
    380
    hi alec,
    Just in case you do not have this model.:)
    E
     
  7. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    I caution against that. The probability that several electrets (with the built-in FET preamp) operate at the exact same bias point (same operating current) is very low, so some are going to be "saturated".
     
    Last edited: Jun 22, 2015
  8. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,777
    1,103
    Thanks for the model file, Eric.
    Point taken, Mike.

    Edit: Eric, I had an older version of that model file which had a bug. Your version works fine.
     
    Last edited: Jun 22, 2015
  9. kenev

    Thread Starter New Member

    Jun 19, 2015
    10
    0
    Sorry, but I don't understand what do you mean by the term "*decoupled*" for the bias resistors
     
  10. kenev

    Thread Starter New Member

    Jun 19, 2015
    10
    0
    So, that means that it would be safer to run them in parallel.
     
  11. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    I would. Like this:

    mic.gif
     
    Last edited: Jun 22, 2015
  12. kenev

    Thread Starter New Member

    Jun 19, 2015
    10
    0
    That's exactly what I was thinking of for the right side of the accordion. And the same configuration, with 2 capsules, for the left side (bass section).
    Each section will have its own preamp (with a potentiometer in place of the input resistor, acting as a volume control) and the outputs of the two preamps will be summed up in a second preamp stage .
     
  13. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    Do you mean that the bias resistor will be replaced by a pot, with the audio coming out the wiper?
     
  14. kenev

    Thread Starter New Member

    Jun 19, 2015
    10
    0
    No, I mean that, for the right hand (melody) section, I will put 3-4 ECMs in parallel, as per the configuration you posted above, and the combined outputs will be amplified by an op-amp based preamplifier, the input resistor of which will be a pot. The same for the left hand (bass) section, with 2 ECMs in parallel. The outputs of the two preamps will be summed up in another final preamp stage.
     
  15. MikeML

    AAC Fanatic!

    Oct 2, 2009
    5,450
    1,066
    If you did what I said, you could balance each pick up independently. No guarantee that every electret has the same sensitivity.
     
  16. kenev

    Thread Starter New Member

    Jun 19, 2015
    10
    0
    If I get it right, you mean using the configuration that you suggested in post #11 (biasing each capsule individually and paralleling their outputs) and adjusting each bias resistor to get equal output from each pick up. In that case, what is a practical method to adjust each pick up for equal output?
     
  17. Alec_t

    AAC Fanatic!

    Sep 17, 2013
    5,777
    1,103
    If you have access to a dual beam scope you could mount two capsules close together, facing a single sound source, and compare their outputs.
     
  18. kenev

    Thread Starter New Member

    Jun 19, 2015
    10
    0
    Actually, I do have a dual beam oscilloscope. So, I will put a speaker in front of 2 capsules side by side and feed the speaker with a sine wave signal, adjusting the bias resistors in order to get equal signal widths from the capsules.
    Thank you for the tip!
     
  19. kenev

    Thread Starter New Member

    Jun 19, 2015
    10
    0
    I'm returning to this older thread to ask a new question, as I'm finally in the process of designing the op-amp preamplifier for the microphone system. As I noted previously, it will include a double op-amp, one for the treble (4 ECMs) and one for the bass (3 ECMs). I'm thinking of using the circuit described in the attached pdf file, which is an application for the OPA172 opamp by TI. The potentiometers for adjusting the gain of each individual part (bass and treble) will replace the output resistor to ground (R6).
    My question refers to the calculation of the gain of the preamp for parallel ECMs. I can readily calculate the gain for 1 electret microphone, but what should that be when I parallel 2 ECMs.
    To make it more clear: the capsules I will use have a sensitivity of -45 dBV, measured at 94 dB SPL. That means that, in order to have an output of 1.228 Volts at a maximum SPL of 100 dB, I need a gain of ca. 40 dB (110X). If I parallel a second capsule, does this mean that i will have to half the available gain?
    Any suggestion is warmly welcomed.

    P.S. I tried a lot to upload the pdf file, but this proved to be impossible...
     
  20. kenev

    Thread Starter New Member

    Jun 19, 2015
    10
    0
    I had to make a zip file out of the pdf in order to be able to upload it.
     
Loading...