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.

ADC128S102QML-SP: early conversion ending

Part Number: ADC128S102QML-SP

Tool/software:

Hi, in a few words I would like to use the ADC in a lower resolution mode.

From the datasheet I understand that a frame is delimited by CS but also that it must be 16 clock cycles long:

"A serial frame is initiated on the falling edge of CS and ends on the rising edge of CS. Each frame must contain an integer multiple of 16 rising SCLK edges."

But from the following sentence it also seems that for the control register only 8 clock cycles are sufficient:

"During each conversion, data is clocked into a control register through the DIN pin on the first 8 rising edges of SCLK after the fall of CS."

Can I end a conversion early with CS if I don't need the full ADC resolution?

In other words, can I use the ADC like in the following timing diagram? Here I'm only using 8 clock cycles thus 4 output bits as an example. SDI and SCLK have an high idle state.

And for example if my SCLK frequency is 800kHz and I use 4 channels at 8 bits of resolution (12 SCLK cycles + 1 CS cycle) does that mean that my conversion frequency is 61.538kHz and my sampling frequency is 15.384kHz per channel?

Best regards

  • Hi Luca,

    Apologies for the late response. Let me know if you have determined anything new since your initial post.

    The ADC128S102QML-SP does not necessarily support resolutions under 12-bit. What you will be doing in this case will be truncating the least significant bits of the conversion, but in theory it is possible.

    The SDI pin loads the control register for the next conversion on the first 8 SCLKs.

    The input signal is sampled during the first 3 SCLKs.

    The first conversion bit is available on DOUT starting on the 5th rising edge of SCLK.

    These are the critical points in timing that should not be interrupted. Might I ask why attempt this approach rather than keeping the device in continuous sampling mode? Is there a need for a lower resolution, lower cost, space-rated ADC?

    Regards,
    Joel

  • Hi Joel, thanks for getting back to me.

    This approach would be essential to apply some optimizations on the logic side. I have yet to try it on actual hardware.

    And yes, probably a lower resolution/cost space-rated ADC with the same number of channels would be better, since it also has the problem of sparkling codes (although it has already been mitigated in the logic).

    Best regards

  • Hi Luca,

    Understood, and thank you for the input. This will be possible to do with the device. Please let me know if you have any additional questions.

    Regards,
    Joel