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.

TLV320AIC34: low pass filter not working

Part Number: TLV320AIC34
Other Parts Discussed in Thread: CC8530,

I  am developing a system using 4 CC8530 based slave microphones with a CC8530 based master.  Both slave and master use the TLV320AIC34 codec. The slave codec is very under used.  The system pairs up all four channels and each microphone connects with it own output at the 4 channel master codec .  The four analogue signals pass into a mixer.

Image 1 shows the signal straight out of the Codec pin LEFT_LOP_A before the channel has paired with a microphone.

Image 2 shows the same signal when a microphone has paired with no acoustic input.

This image shows random noise of about 200mV p-p

Image 3 shows the same signal expanded to a 1us per division timebase

With the expanded timebase the noise bandwidth  is beyond acoustic frequencies.

 

Image 4 shows the  same signal with the paired microphone and acoustic microphone input from a whistle playing a 1.2 kHz note.

Texas recommend the use of a 20 kHz low pass filter to supress out of acoustic range signals  in their data sheet SLAS538B.

 

I have tried incorporating such a filter using the TI PurePath configurator and coefficients generated using TIBQ.exe.  I used the coefficients for filter 10 set to 20 KHz with a 48 kHz sampling speed

 

Custom setup option enabled.

 

The coefficients were entered into page 1 registers

 N0=  0x6F95, N1= 0x6F95, D1= A0D5.

 

The filters were enabled  in register 12 page 0

 

After reprogramming the CC8530 with the new configuration the output signals were unchanged.  It looked like the new coefficients had not taken effect

 

 

  • Hello Andy,

    your images are not showing up for me. Can you try reposting them.

    -Steve
  • Part Number: TLV320AIC34

    Hello Steve.  The first set of images were embedded in a Word document.  Attached are four jpeg files.  I hope you can make sense of them.

    Regards,  Andy.

  • Hello Andy,

    one thing to note, that the filter can be enabled with register 12, but if you are using coefficients that are not default, you must program register 107 as well.

    can you confirm that you have done this?

    best regards,
    -Steve Wilson
  • Hello Steve,

    Thanks for telling me about register 107.  I have looked at all my configuration files and I do not program register 107.  I will try this out and report the results to you.

    Regards,

    Andy

  • That is likely the problem. The default values are of course a HPF. with -3db point at .0045*fs, .0125 * fs or .025*fs depending on your selection

    -Steve
  • hi Andy,

    Were you able to test this? please let me know if you are still having issues and I would be happy to help you.
    best regards,
    -Steve Wilson
  • Spending time going over the AIC34 data sheet I now have a better idea of what it is capable of and how it is internally

    connected.  I have 4 slave microphones feeding 1 master.  The 4 outputs from the master  feed a mixer to give me a

    single signal with all 4 microphones mixed.  All 4 master outputs have 200mV p-p noise when the microphones are

    active.  This must be removed before mixing.  I have been trying to remove the noise using the effects filters

    configured as 20kHz low pass filters.  I did program register 107, as you suggested, to say the default coefficients are

    not being used. If a microphone channel is inactive, the DAC output for that channel is silent.  When in playback

    mode and the DACs are enabled, the effects filters are connected before the DACs.  I think the noise I am seeing is

    switching noise generated by the DAC.  I need the low pass filter AFTER the DAC.  Taking the 4 analogue outputs from

    the DACs and passing them through a simple single pole R/C network (1k ohm/ 10nF 10kHz 3dB)  the noise

    disappears leaving a clean audio signal.

    As I stated earlier the master output noise was only present on active microphones.  This led me to consider the

    noise could be originating in the slaves.  This noise is very high frequency (1MHz+) and with a sample rate of 48kHz

    should not be capable of transmission via RF.  Even so, I added a 20kHz low pass filter on the slaves with the DACs

    powered down in the record only mode.  I wouldn't say it cured the problem but the noise from each of the Master's 4

    analogue channels reduced from 200mV p-p to 150mV p-p.

    The board is made on a 4 layer PCB with a dedicated ground plane, 2 signal planes and a shared power plane.  There

    are 3 DC-DC converters for 3.3V 1.8V and 1.2V.  I would prefer to have no noise at all but with so much switching on

    such a small board I think 4 extra resistors and capacitors is a small price to pay to give me good results.