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.

ADS1257: Recommended clock source for the ADS1257

Part Number: ADS1257

Hi Ti Support,

I was reading through the datasheet and I was not able to find at which page it discuss the way to change the sampling rate. Does it affect the external clock source selection?

Could you help me select a suitable clock source for ADS1257? The datasheet says it should be in the range of 0.1 to 10MHz. Which external clock generator is best for this IC and is it the smallest size we have?

Thank you,

Khoi Ly

  • Hi Khoi,

    The easiest way to change the ADC's output data rate is by configuring the "DRATE" register setting. The data rates shown in the datasheet are only valid when using the nominal 7.68 MHz ADC clock. I would stick with a 7.68 MHz clock source unless you have good reason for needing a different clock frequency.

    Generally, I like to use crystal oscillators for clocking the ADC since they are generally inexpensive and easy to use. For example:

    At the time I checked, I didn't see any 7.8 MHz crystal oscillators in stock (which would be a more than valid reason for using a different clock frequency). In that case, perhaps look for a 7.3728 MHz or 8 MHz crystal oscillator.

    NOTE: If you have a micro-controller in your system that can generate this clock for you, then you wouldn't necessarily need another component to create this clock signal for you.

    Best regards,
    Chris

  • Thank you for your reply Chris,

    So if we use 7.3728 or 8 Mhz crystal, do we have any problem with the data rates?
    "The data rates shown in the datasheet are only valid when using the nominal 7.68 MHz ADC clock."

    Also, unlike microcontroller on which you have 2 pins for the clock, this ADS IC has only one pin CLKIN for the crystal. So the other three electrodes of the crystal are connected to ground?

    Thank you,

    Khoi Ly
  • Hi Khoi,

    If you modify the ADC's clock frequency, then the output data rates will scale linearly. So for example, with an 8 MHz clock all of the (7.68 MHz specified) data rates will increase by a factor of 8/7.68 = 1.0416 ( 5 SPS => 5.208 SPS, 10 SPS => 10.416 SPS, etc.)

    NOTE: The clock input for the ADS1257 is not meant to drive a crystal! This ADC expects a buffered clock signal (i.e. the output of a "crystal oscillator", also sometimes just called an "oscillator"). If you wanted to use a crystal, then you would need to pair it with a crystal driver (for example, www.ti.com/.../SN74LVC1GX04) before providing the signal to the ADS1257.

    Best regards,
    Chris
  • Sorry,
    I mean the crystal oscillator (the one similar to that in your provided links). A crystal oscillator has 4 pads. For a microcontroller, I usually connect pad 1 to OC1 and Pad 3 to OC2 of the microcontroller. But the ADS1257 only has one pin CLKIN. Do you have a reference design or some application notes on how I could connect the crystal oscillator to the ADS1257?

    Thanks,

    Khoi Ly
  • Hi Khoi,

    The oscillator IC integrates the oscillator + crystal so that you don't have to design the clock circuit. You simply need to power the oscillator and you get a clock.

    Here is an example pinout of a crystal oscillator...

    The connection to the ADS1257 is simply a direct trace between the oscillator's OUT pin to the ADC's CLKIN pin. However, I would advise adding a series resistor on this connection in case you need to slow down the rising edge of the clock signal to reduce clock noise.

    Best regards,
    Chris

  • Hi Chris,

    I wonder if there is a faster way to setup the adc to talk to the microcontroller (ESP32). Generally, do we have to manually setup the timing, the register pulses according to the datasheet of an ADC? 

    Also, when I look at the datasheet for this part, there are a few pins that, even after I went through their descriptions, I still don't fully understand.

    1. What is the purpose of the SYNC pin. "The SYNC command synchronizes the analog-to-digital conversion." But, isn't SCLK used for that purpose? I assume that the SCLK synchronize the microcontroller and the adc

    2. If I have 3 ADS1257 for each microcontroller (I use ESP32 -Thing from Sparkfun), do I need to have 3 separate DRDY output pins that are capable of hardware interrupt?

    3. Since I have 3 ADS1257, how would I be able to modify the pseudo-code to accommodate for 2 more ads?

              a. I think from the beginning to the line before LOOP, we will do for all ads1257's

              b. If I switch between ads1257s, I don't need to power them down, right? I will go back to the LOOP and just take the CS pin of the next ads1257 LOW?

    4. Where would the DRATE and PGA gain set in this pseudo-code? It is the WREG command after we send the SDATAC command, right?

    5. Is there any differences in terms of performance if I measure the signal Single Ended as opposed to Differential mode? Each ads will collect samples from 2 measurements so we also have to MUX each adc as well

    6. This might be a naive question, usually digital differential signaling needs 2 pins, one P and one N (like the USB connector). The timing diagrams of DIN and DOUT suggest that they are differential signals, yet I only see one pin for DIN and one for DOUT. I wonder how this works?

    7. Finally another naive question, If I set the ADC sampling rate to 1kHz but I have 2 measurements, do I get 1kHz sampling rate for each measurement? Does the LOOP frequency in the pseudo-code have to match with the sampling rate of the ADC? In other words, how would I collect 6 measurements (2 measurements for each adc) at the rate of 1kHz to my microcontroller?

    Thank you very much for your support and I look forward to hearing from you soon,

    Khoi Ly

    Phd in Mechanical Engineering 

    University of Colorado Boulder

  • Hi Khoi,

    I'm going to go ahead and close this E2E thread. It looks like our discussion was continued over in a new thread: e2e.ti.com/.../2784439