• State
  • Locked Locked
  • Replies 3 replies
  • Subscribers 31 subscribers
  • Views 944 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.

TLV320AIC3106 Input Channel Crosstalk

Eric Hereford
Intellectual 355 points

Other Parts Discussed in Thread: TLV320AIC3106

I have a design in which I am using the LINE1LP and LINE1RP inputs on a TLV320AIC3106 CODEC.  My input signal is around 500mV and I am using single-ended input mode.  It appears that I am getting crosstalk between these two inputs.  Basically, I can inject a sine wave signal on either input and see lower level (sometimes distored depending on input volume) version on the output TDM stream of the other input.  I have made sure that both the LINE1R to Left ADC (21) and LINE1L to Right ADC (24) Control Registers are set to their default value of 0x78 which should isolate them from being mixed in.

What is causing this?  Is this a know issue or is there some type of Errata?  Is there some other register I could be setting improperly?

  • 0
    TI__Mastermind 28585 points

    Hi Eric,

     

    there is a channel to channel isolation spec in the datasheet. The spec for Line1 L to R is -75dB. Even if a channel is not selected, there will be some stray capacitance across the input switch. Channel isolation can sometimes be improved by going differential or choosing a different input pair ( you will see in the datasheet that channel 1 has the lowest spec). Layout can also be an issue.

     

     

  • 0 in reply to Converter Apps
    Intellectual 355 points

    Thanks for the quick reply.  I am not sure that the -75dB isolation is the root of our problem.  It appears that the crosstalk increases dramatically as the input begins to distort.  Basically, I am using a scope to view the input waveform before and after the input DC blocking capacitor (which on our board is 1uF).  As the input signal is increase it begins to distort on the side of the capacitor connected to the input of the CODEC.  This is the point in which I notice much more crosstalk.  The input level this is ocurring at is less than 707mV (which is the recommended full scale input voltage).

    Is this distortion expected behavior of the device at input levels less than 707mV?

    What is the expected behaviour at levels greater than 707mV?

    Is crosstalk expected to increase as the input level approaches and exceeds 707mV?

    Any advice would be greatly appreciated.

  • 0 in reply to Eric Hereford
    TI__Mastermind 28585 points

    Once the inputs go non-linear, you can run into a host of issues that could certainly cause more crosstalk.

     

    You will see that all of our distortion testing is done at -2 dBFS to avoid any clipping due to offset or gain error in the analog portion of the input circuits.