Mic input to pc

Discussion in 'PC Hardware' started by steve marchant, May 9, 2007.

  1. Want to do some processing of the mic signal before it goes into the mic
    input, but know nothing about the input/output conditions presented. Had no
    luck so far in "googling" for technical info about the circuitry related to
    the mic input area of a pc. Anyone out there with a clue?
    steve marchant, May 9, 2007
    1. Advertisements

  2. steve marchant

    Conor Guest

    It's a line level input conforming to normal audio standards so if what
    you want to do would work with a normal amplifier input, it'll work
    with a PC soundcard.
    Conor, May 9, 2007
    1. Advertisements

  3. steve marchant

    Paul Guest

    You can research how AC'97 sound chips work and extrapolate from there.

    Analog Devices are more clever about providing real specs at the
    chip pin. If you look on PDF page 4, input impedance is 20K ohms
    in parallel with 5pF. That would be 20K between input and ground.
    If you were feeding it from a 600 ohm device, there would be
    little loading from the 20K number. If you had a high impedance
    output of some sort, then YMMV.


    A popular AC'97 at Realtek, is the ALC650. The reason for
    downloading the packages here, is a reference schematic
    is provided. Realtek uses FTP servers for their files.


    The second package is a ZIP, with some separate ZIP files
    inside. If you unzip everything, you should see
    alc650_demo_circuit_ver_11.pdf . The thing to note there,
    is analog I/O is all capacitively coupled. The input pin
    has some kind of bias network, so the A.C. coupling is used
    to prevent upset to the chip bias. Or at least that is my best
    guess as to why they do it. Otherwise, a D.C. connection
    would be desirable from the user's perspective.

    Note that the jack sharing scheme shown in the ref schematic
    would be present on a motherboard that had only three 1/8"
    jacks on the rear panel. If a motherboard has five or
    six jacks, then there is no need for a sharing circuit.

    So based on that, you have a 1uF series cap, feeding a 20K
    ohm or so resistor to ground. Roughly a high pass, with
    8Hz corner frequency, using 1/2*PI*R*C.


    Now, another issue, is MIC_BIAS. The 1/8" jack has three
    contacts, tip, ring, and sleeve. Tip should be 1uF and
    20K to ground. Ring can have a MIC_BIAS signal on it. The
    purpose of MIC_BIAS, is to provide DC power to electret
    microphones. The voltage is 3.3 to 5V perhaps, with a
    series limiting resistor of 2K ohms.

    On a monophonic microphone input, DC bias is only on
    the Ring contact. If a microphone input is stereo, then DC
    bias *could* be provided on both Tip and Ring. Each would
    have their own separate 2K or whatever, limiting resistor.
    This would be handy for things like the Andrea Superbeam
    stereo microphone. (The following page is mainly for a mono


    This is the Andrea Superbeam. I think it needs the equivalent
    of mic bias on both Tip and Ring. The Superbeam is stereo,
    and clever software on the computer can use the stereo signals,
    to remove some background noise.


    The reason I mentioned all of that, is if you have any
    concerns about the Microphone jack, use the Line-In instead :)

    No matter what jack you use, you should test for the
    presence of DC on Tip or Ring. (At least, if your proposed
    circuit would be affected by a stray couple milliamps of

    To test, first find a male to male, 1/8" cable, like the
    kind that comes with a TV tuner card. Plug the 1/8" extender
    cable into the Microphone jack. Using your multimeter, see if
    Tip or Ring has a DC microphone bias voltage present. If there
    is no DC present on the jack, then chances are you are safe to
    assume just the series 1uF, and 20K to ground. You could then
    use the Microphone input for your experiment.

    Note - when doing the above DC check, go into the sound card
    software, and tell it a microphone is connected to the
    microphone jack. Some chips have the ability to turn the
    bias on and off, and the DC bias may only appear when you
    tell the software that a microphone is present. The DC
    bias can come from the +5VA rail, but on some chips, the
    bias actually comes from a regulated source on the sound
    chip itself.

    Paul, May 9, 2007
  4. Hi Paul
    This is extremely helpful stuff, and I'll be taking a close look. It's a bit
    of an over-kill for what I immediately have in mind, though, which is to
    take what comes out of the pc speaker output(s), mix it with voice from the
    mic, and push the composite back into Windows Sound Recorder via the mic
    input. You also sorted out the mystery I found when looking for what pole of
    the jack carried the mic bias. I was getting 2.7 volts on both the tip and
    ring. Now I see this is for a stereo mic.
    Thanks again. Great stuff.
    steve marchant, May 10, 2007
  5. steve marchant

    Franc Zabkar Guest

    Can you connect the line output to the line input and the mike to the
    mike input, and then enable both recording devices in your sound
    card's mixer?

    - Franc Zabkar
    Franc Zabkar, May 11, 2007
  6. Didn't think of it Franc. Been spending a while setting it up. It works, but
    the line input sound level is very low compared with the mic input it's
    being mixed with.. I'm wondering if the signal from the speaker output needs
    boosting. I'll try this next.
    steve marchant, May 11, 2007
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.