TLV320AIC3204 headphone delay, causes annoying echo when speaking into mic with headphones

Are you sure that the delay is through the codec and not through the rest of the system? Please explain the signal path in more detail.

Thank you for your quick reply!

Yes, the delay is through the IN1_L/IN1_R to HPL/HPR path of the codec .  The board has an analog bypass output (circumventing the codec entirely, taking the same signal presented to the IN_L/IN1R pins) which can be used to connect aux equipment.  Monitoring with headphones there results in no confusing echo.

Hi!  

I'm following up, hoping to get help from TI support on this ... Thanks!

Oh, and please ignore these addresses.  Here is what the new registers read.  I originally cut/pasted this register list from a register readout of a board that had an error (connection between analog/digital grounds missing).

Page  Addr   Old     New
0     44     0x00C0  0x0000
0     46     0x00C0  0x0000

Anyhow, looking forward to TIs help with this headphone delay issue, resulting in an annoying user echo experience ...

Attached is an audio file, so you can hear for yourself what I am talking about.

There is a microphone plugged into our board.  The signal goes through an op-amp preamplifier.  Then, the signal goes into IN1_L of the codec, and comes out HPL to headphones.  I have rested the headphones on the table, and have the mic of an external recorder next to the earcup.  The recorder mic picks up ambient (my direct voice) as well as the codec headphone output.  Playing the attached recording, you can hear the annoying echo experienced by the user (direct voice plus delayed voice through the codec).  This is a big problem for us.

We are certain that the delay path is though the codec.  When we take a signal directly from the opamp output (which subsequently feeds the codec IN1_L) and use an analog headphone amplifier, there is no echo/delay.  I have not attached an audio file of this, as it is not instructive.

TI must have some test bench of this codec.  Do you experience the same delay/echo?  What programming settings are you using?  If you use the register settings we are using, do you experience this delay/echo?

Thank you for your help ...

D. Hartl , and other TI apps people,

Are there any other questions that you would like me to answer, to convince you that this delay is through the TI codec?

I have provided an audio file, so you can hear the delay for yourself, and experience the very distracting 'echo' effect that is experienced by the user.  Have you listened to this?

Please help ...

Don Dapkus,

I've seen this video of yours:

http://www.youtube.com/watch?v=iJyWx57vCX0

For me, so far, E2E has not been a success story. It's great to have a searchable database.  I have searched, to no avail.  Now, this issue requires the attention of your apps team.  Please tell me how I can get a response to this matter?  Do you need more clarification?  Ask me more questions and I'm happy to oblige?  I've got to believe that there must be some registers that I have programmed that are not optimized, resulting in this delay.  Or, is there something fundamentally wrong with this chip and nobody at TI wants to talk about it?  That doesn't help me.  If there's something fundamentally wrong, then set me free and let me know, so I can go design in a Cirrus (or something else) chip.  But, right now, we've got this chip designed on our board, and it's just not an acceptable user experience with this delay.

Please guide me on what I need to do to get a response from you and your app team ... please help!