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.

TLV320ADC5140: large noise when connecting to customer's board

Part Number: TLV320ADC5140

Hello Sir, 

One of my customers connects TLV320ADC5140 EVM to their own board. They uses 4-CH  single-ended input. FSYNC is 32KHz, MCLK is 8.192MHz, SBCLK is 4.096MHz, the SBCLK/FSYNC is 128. The power up order is: Platform, AVDD, IOVDD and after they are stable, the CLK starts, then SHDNZ is driven to high level, after 10ms, REGISTER is set up with using PPC3. SBCLK and FSYNC works well. While the noise can be heard clearly and original input signal can be heard as well behind the noise.

 I thought this might cause by unjustified filters setting-up (decimation filter, HFP and Bi-quad filter). What and how do I need to check?

Thanks for your information!

Register setting-up:

// Page 0

{ 0x00, 0x00 },
// Book 0
{ 0x7f, 0x00 },
// Page 0
{ 0x00, 0x00 },
// Reset Device
{ 0x01, 0x01 },
// 1mS Delay
{ CFG_META_DELAY, 0x01 },
//
---------------------------------------------------------------------------
--
// Begin Device Memory
//
---------------------------------------------------------------------------
--
// Book 0 Page 0 (0x00) Dump
{ 0x00, 0x00 },
{ 0x7f, 0x00 },
{ 0x00, 0x00 },
{ 0x02, 0x81 },
{ CFG_META_DELAY, 0x01 },
{ 0x07, 0x30 },//0x30 32bit 0x20 24bit
{ 0x14, 0x36 },//add 32k 128 ratio
{ 0x21, 0x21 },
{ 0x3b, 0x60 },
{ 0x3c, 0xa1 },
{ 0x41, 0xa1 },
{ 0x46, 0xa1 },
{ 0x4b, 0xa1 },
{ 0x6b, 0x00 },
{ 0x6c, 0x02 },//AGC SELECT
{ 0x6d, 0x9b },
{ 0x74, 0xf0 },
#if 1
// Page 1 (0x01) Dump
{ 0x00, 0x00 },
{ 0x7f, 0x00 },
{ 0x00, 0x01 },
{ 0x05, 0x01 },
{ 0x07, 0x20 },
// Page 4 (0x04) Dump
{ 0x00, 0x00 },
{ 0x7f, 0x00 },
{ 0x00, 0x04 },
{ 0x49, 0xe6 },
{ 0x4a, 0x48 },
{ 0x4b, 0xc6 },
{ 0x4c, 0x80 },
{ 0x4d, 0x19 },
{ 0x4e, 0xb7 },
{ 0x4f, 0x3a },
{ 0x50, 0x7f },
{ 0x51, 0xcc },
{ 0x52, 0x91 },
{ 0x53, 0x8b },
{ 0x54, 0x24 },
{ 0x55, 0x46 },
{ 0x56, 0xc8 },
{ 0x57, 0x84 },
{ 0x58, 0x04 },
{ 0x5c, 0x04 },
{ 0x60, 0x04 },
{ 0x64, 0x04 },
{ 0x68, 0x04 },
{ 0x6c, 0x04 },
{ 0x70, 0x04 },
{ 0x74, 0x04 },
{ 0x78, 0x04 },
{ 0x79, 0x9e },
{ 0x7a, 0x7f },
{ 0x7b, 0x30 },
{ 0x7c, 0x04 },
{ 0x7d, 0x9e },
{ 0x7e, 0x7f },
{ 0x7f, 0x30 },
// Page 6 (0x06) Dump
{ 0x00, 0x00 },
{ 0x7f, 0x00 },
{ 0x00, 0x06 },
{ 0x1d, 0x04 },
{ 0x1e, 0xb0 },
#endif
// Power up/down register
{ 0x00, 0x00 },
{ 0x7f, 0x00 },
{ 0x00, 0x00 },
{ 0x75, 0x60 },

  • Wenting,

    It appears the your customer may be using an out of date version of the ADCx140 App.  Th page 1, 4 and 6 dumps should not be there.  If they have an updated app it may be that they created the .ppc3 file in an older version of the software.  If this is the case they will unfortunately need to create a new file.  as the page 1,4 and 6 registers are saved in the program. The most recent version of the ADCx140 Software is 3.0.5

    best regards,

    -Steve Wilson

  • Hello Steve,

    I've deleted page 1, 4, 6 while there still is a large noise even there is no audio signal input. I've send you .h and driven file through email. Do you mind having a view? If there still is an error in PPC3 output .h file, could you help to output a right .h file with using application needs I describe in email.

    Many  thanks for your support!

  • Hello Steve,

    Hope everything goes well. I've also sent you an email related to this case. If you are available , do you mind helping on having a view on it?

    Best regards,

    wenting 

  • Wenting, 

    Can you have the customer confirm that the PAGE 1, page 4 and page 6 registers are not being written to anymore?  

    please read page 0 register 0x15?

    You did provide a picture of your setup by email,  it looked like the EVM was still being powered up by the AC-MB correct? I saw that they connected MCLK but did not configure the GPIO to use MCLK.  is this intended?  the default use for the GPIO is as an interrupt.  They do not need MCLK, so I recommend just disconnecting this.  

    I also see that IOVDD is povided by your processor board which is good.  Note that the pull up resistors for I2C are on the ACMB so if that is not connected the processor board will need pull-up resistors. (I assume it has this) are you providing a solid ground connection to the EVM from the processor board?

    Can you provide a recording of the loud noise?

    best regards,

    -Steve Wilson

  • Hello Steve,

    Thanks for your reply. The customer is processing another urgent project now after 1 week debugging for TLV320ADC5140 but without progress. They will come back for this case in the the beginning of August. And I will debug together with customer them according to your recommendation. 

    Best regards,

    wenting