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.

ADS8688 - SCLK Falling edge to next data valid delay

Other Parts Discussed in Thread: ADS8688, ADS8319

Hi all,

Would you mind if we ask ADS8688?

SCLK Falling edge to next data valid delay of ADS8319 is defined(24ns), but we could not find the same one on ADS8688.
So, will this value be 35ns?(tHT_CKDO + tSU_DOCK)
Could you let us know it?

We need your help.

Kind regards,

Hirotaka Matsumoto

  • Hello Hirotaka-san,

    The SCLK falling edge to next data valid delay should be <34ns.  If tSCLK min is 59ns, and tSU_DOCK min is 25ns; the SCLK falling edge to next valid delay is 

     59ns-25ns <34ns.

    Thank you and Regards,

    Luis

  • Luis san,

    Thank you for your reply.

    Kind regards,

    Hirotaka Matsumoto

  • Luis san

    We have three additional questions as followings;

    <Question1>
    Does the data valid delay time depend on SCLK?

    <Question2>
    Could we control the data valid time on SDO to have enough time to acquire data with slow back-end logic?

    <Question3>
    What does cause to issue the next data bit; falling edge, rising edge, or etc?

    Could you let us know these?

    Kind regards,

    Hirotaka Matsumoto

  • Hello Hirotaka-san

    Typically the host will read the data from the device in the subsequent falling edge of the SCLK clock signal. The tHT_CKDO (hold time SCLK falling to (previous) data valid) spec in the datasheet is 10nS minimum.   Therefore, the host can read data during the SCLK falling edge. The SDO is updated with a new value >10nS after SCLK falling edge.

    -Does the data valid delay time depend on SCLK? The min/max specs are given for temperature range -40C to 125C with SDO load=20pF. The timing specifications are valid provided that the max frequency (17MHz) is not exceeded. The minimum specification limits for tHT_CKDO (Hold time SCLK falling to (previous) data valid on SDO), and the minimum tSU_DOCK (Setup time SDO data valid to SCLK falling) are valid regardless of SCLK frequency.


    Could we control the data valid time on SDO to have enough time to acquire data with slow back-end logic? Since the SDO is updated anytime10nS after SCLK falling edge (tHT_CKDO), and since (tSU_DOCK) is a minimum of 24ns; it is possible to read data on subsequent SCLK falling edge.   Also, in applications where SCLK frequency is slow, since SDO is updated with new data sometime 10nS after SCLK falling edge, it may be also possible to read data on SCLK rising edge. What SCLK frequency is used in this application?
    What does cause to issue the next data bit; falling edge, rising edge, or etc.? The datasheet shows that the SCLK falling edge triggers the new SDO data. The previous data is held on SDO for 10nS after SCLK falling edge and the SDO updates new data 10nS after SCLK falling edge.

    -       Can you provide a detailed description of the problem? Please provide an oscilloscope plot of SPI signals (CS, SDO, SCLK, SDI) while reading conversion data. Please let us know what the expected conversion data is and the actual data that is read on the SDO pin.

    -       Please let us know what is the SCLK frequency is used.

    Best Regards,

    Luis

  • Luis san,

    Thank you for your excellent reply! 

    <Q1>
    What SCLK frequency is used in this application?

    <A1>
    Our cutomer's stage is in the development stage.
    That's why they don't decide yet.

    <Q2>
    Can you provide a detailed description of the problem?
    -Please provide an oscilloscope plot of SPI signals (CS, SDO, SCLK, SDI) while reading conversion data.
    -Please let us know what the expected conversion data is and the actual data that is read on the SDO pin.
    -Please let us know what is the SCLK frequency is used.

    <A2>
    As we mentioned before, when our customer uses the SPI(SCLK) which the tHT-CKDO spec of Master side is 40ns,
    they consider that they could not get ADC's data exactly using falling edges.
    They have to consider that they  get ADC's data using rising edges.

    Kind regards,

    Hirotaka Matsumoto