Post Go back to editing

AD9375 SYNCOUT_B is from HI to Low, when change TXATT to 0dBFS from -2dBFS with 0dBFS digital input

Hi all,

Our customer evaluate AD9375 EVB + Aria 10 .

Evaluating TX performance with PA,DPD off,

SYNCOUT_B is from HI to Low, when change TXATT to 0dBFS from -2dBFS with 0dBFS digital input.

TXATT at -2 dBFS or less is normal operation.

What is happening ?

In the datasheet @ AD9375 Rev. 0 Page 10, following noted.

8 Note that the input signal power limit does not correspond to 0 dBFS at the digital output because of the nature of the continuous time Σ-Δ ADCs.

Unlike the hard clipping characteristic of pipeline ADCs, these converters exhibit a soft overload behavior when the input approaches the maximum level.

What is the " a soft overload behavior " ?

But we do not understand why require to reset the link with SYNCOUT_B ?

Please advices to us !

Best regards,

sss



additional
[edited by: sss@jpn at 9:05 AM (GMT 0) on 15 May 2019]
Parents
  • SYNCOUT_B is from HI to Low, when change TXATT to 0dBFS from -2dBFS with 0dBFS digital input.

    Do you mean JESD is getting Reset ?

    Is it custom board or evaluation board ?

    Soft overload means , you will not see sudden clipping.

  • Hi Vinod - san,


    Thank you for quick reply.


    >Do you mean JESD is getting Reset ?

    Yes, SYNCOUT_B status is change.

    They say changed the TXATT setting only.

    We wonder why it require the JESD reset.

    We think that this issue is not related to monitoring the JESD204B link for running disparity errors.

    Are there any factors that could be considered?

    >Is it custom board or evaluation board ?

    It's Evaliation boards.

    ADRV9375-W/PCBZ + Aria 10 Development Kit


    >Soft overload means , you will not see sudden clipping.

    In this time, I understand the note is no relation.

    Is it correct ?

    Best regards,

    sss

  • With attenuation change JESD will not reset. Only chance of this happening is variation in power supply when attenuation changes impacting JESD. I don't see this happening in evaluation board.

    >Soft overload means , you will not see sudden clipping.

    In this time, I understand the note is no relation.

    Is it correct ?

    Yes this has no relation to issue seen.

    Initial post mentions -2dBFS  , are they changing digital signal power at baseband. (FPGA) and not Tx attenuation ?

    Can they share more details like log files or steps to recreate the issue.

  • Hi Vinod - san,

    Here is customer evaluate details.

    1.Each evaluation boards is ,
    RFIC : ADRV9375-W/PCBZ
    FPGA: Intel (Arria V ST Soc or Arria 10 GX , both happend)。
    Setting to change is the TXATT only .

    2.Phenomenon when changing the TXATT
    note:The customer describes "SYNCOUTB0+/-" as "TXSYNC_B".
    "SYNCOUTB0" = "TXSYNC_B"

    When change to TXATT:-2dBFS >>>0dBFS, TXSYNC_B:H>>>L(repeat High and Low)
    then,
    When return to TXATT:0dBFS>>>-2dBFS, repeat High and Low continus.

    Please check the attached file.


    3.DAC (PA) output
    CW 960MHz, Tx signal
    TXATT = 0dB, IQ amplitude:-2dBFS, +26dBm output

    When the Digital FS data fix (0dBFS), and the TXATT changing (2~11dB,1dB step)

    When the TXATT fix (2dB), and the digital FS data changing (-2dBFS~-11dBFS,1dBFS step)

    In the above two patterns, the DAC (PA) operates in a linear manner.
    : If either TXATT or Digital FS data is not full scale.

    When the TXATT (0dB), and the digital FS DATA (0dBFS)
    Please refer to above 2.

    >With attenuation change JESD will not reset. Only chance of this

    >happening is variation in power supply when attenuation changes

    >impacting JESD. I don't see this happening in evaluation board.

    We realized that it was necessary to check the supply capability and transient response etc. of the power supply line at DAC FS output.
    Can you give me some advice on checkpoints and notes on power supply lines ?

    Best regards,
    sss

  • You may need to add some high value capacitors (220 UF) on power lanes especially 1.8 V lanes (VDDA1P8_TX) and 1.3 V (VDDA1P3_RF_SYNTH, VDDA1P3_RX_TX )

  • Hi Vinod - san,

    Thank you for very useful replly!

    Our customer checked the power supply line again.

    Then, when setting to the DAC FS output,

    they confirmed that the power line was unstable !

    Best regards,

    sss

Reply Children
No Data