• State Resolved
  • Locked Locked
  • Replies 6 replies
  • Subscribers 30 subscribers
  • Views 644 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

TLV320AIC3204: Headset button with fully-differential microphone

Redhatter
Prodigy 120 points
Other Parts Discussed in Thread: TLV320AIC3204

Hi all...

Well, on my earlier post (which turned out to be my fault) I've got headset detection working nicely, however we're dealing with a mono handset, and hence the designers have opted for fully-differential audio I/O.  So handset earpiece hooks across HPL and HPR, handset microphone hooks across IN3L and IN3R, with DC blocking capacitors in the appropriate places, 1kOhm pull up to MICBIAS on MIC+, and a 1kOhm pull-down to 0V on MIC-.

For the push-button, a resistor is depicted in the example topologies given in page 96 of the TLV320AIC3204 datasheet, however no suggested resistance is given for the resistor in series with the switch.  The way I understand it, is that the CODEC is watching for a pull-down on SCLK, a weak-ish pull-down is most likely an electret microphone, and hence that allows the headset detection, a stronger pull-down being the button.  What is not discussed in the datasheet, is the maximum recommended resistance between SCLK and 0V when this push button is down, for the CODEC to detect a button-press.

We have a 1kOhm resistor between MIC- and 0V, and the switch connects between MIC+ and MIC- without any series resistance, so effectively it pulls SCLK to 0V via a 1kOhm resistance.  Combined with a 1kOhm pull-up on MIC+, the CODEC likely sees a ~0.9V voltage.  Is this suitable for detecting a button "press" or does it need to be lower?  Or am I completely on the wrong track?

Regards,

  • 0
    TI__Mastermind 46195 points

    The switch is expected to ~100 ohms or less. Microphone is expected to be around ~1k to ~2.2kohm.

    Once a speaker+mic type headset is inserted and detected, if the mic to ground connection is shorted afterwards, or has low impedance (i.e. ~100ohms) afterwards then it is registered as button press. AIC3254 CS software has a diagram for the recommended headset jack type in the headset detection panel, for 4-conductor type plugs.

    Regards,

    J-

  • 0 in reply to J-
    Prodigy 120 points

    J Arbona said:

    The switch is expected to ~100 ohms or less. Microphone is expected to be around ~1k to ~2.2kohm.

    Once a speaker+mic type headset is inserted and detected, if the mic to ground connection is shorted afterwards, or has low impedance (i.e. ~100ohms) afterwards then it is registered as button press. AIC3254 CS software has a diagram for the recommended headset jack type in the headset detection panel, for 4-conductor type plugs.

    Regards,

    J-

    So is that 100ohms as seen between SCLK and 0V or SCLK and one of the input pins?  Our microphone connects between IN3L and IN3R.

    Where do I obtain this "AIC3254 CS" software, and is it really necessary to download an entire software suite (for a different CODEC) to obtain a single (presumably static) diagram?  Is it different to what's shown in the TLV320AIC3204 datasheet on page 96?

    Thanks in advance.  Regards,

  • 0 in reply to Redhatter
    TI__Mastermind 46195 points

    Sorry. I meant AIC3204 CS.

    Typically, a four conductor jack is used. The tip on the jack is connected to the left channel, the ring1 to the right, ring2 to the mic and sleeve to the ground.

    The PCB is routed such that MICBIAS is connected to MICDET (or SCLK) with 2.2kohm and the MICDET node is also connected to the MIC in the jack (at the codec side of AC-coupling cap).

    See figure 11 of the EVM User's Guide in http://focus.ti.com/docs/toolsw/folders/print/tlv320aic3204evm-k.html along with the EVM schematic.

    In a nutshell, MICDET will monitor its voltage. If no plug is inserted, it sees the MICBIAS voltage because it has a pullup to micbias and nothing is pulling in the jack. When a stereo headset is inserted, MICDET will see ground, because the stereo jack shorts ring2 with the sleeve (ground). when a stereo+mic is inserted, MICDET sees ~2k (the mic impedance), which creates a divider between the MICBIAS 2.2kohm resistor and the 2kohm mic). When a button is pressed (in parallel with the microphone), it will see from 100ohm to ground while the button is pressed, depending on the impedance on the button path.

    Regards,

    J-

  • 0 in reply to J-
    Prodigy 120 points

    Ahh right, so in other words, it relies on the microphone being wired up in a single-ended configuration, not a differential configuration that we have.  Is this correct?

    Also, on the button... is there a way to detect button-up events?  At the moment I get an interrupt the moment the button is pressed down (i.e. the button-down event) but nothing on release.

    Regards,

    Stuart Longland

  • 0 in reply to Redhatter
    TI__Mastermind 46195 points

    Stuart,

    It only interrupts upon button press event. The button press flag could be polled periodically if button release routines are desired.

    Regards,

    J-

  • 0 in reply to J-
    Prodigy 120 points

    Right, sounds like we've got some re-designing to do.  Thanks for your assistance.