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.

ADS1258: Single-ended input resistance

Part Number: ADS1258

Hi team, 

Good day.

In the datasheet, it was stated that the differential input impedance is 65kohm. What is the typical ADC input resistance and charge capacitor value when the ADS1258 is configured to operate in single-ended mode?

Regards,
Carlo

  • Hi Carlo,

    Check on Figure 37 and the accompanying description about the ADC input structure on page 16 of the ADC datasheet. This shows a simplified capacitor network to model the ADC inputs, as well as the corresponding resistive equivalent at fCLK = 16 MHz. Note that the equivalent resistance will change with clock frequency as described in the datasheet. Also note that the ADC always takes differential measurements (ADCINP - ADCINN), even when measuring ground referenced signals. So differential input impedance is usually more appropriate here, and for high impedance sensors we would recommend a buffer.

    I have copied Figure 37 below for reference. Let me know if this helps answer your question.

    -Bryan

  • Thanks Bryan and Carlo,

    A few more questions:

    Q1: In single-ended channel mode, the two S1 and two S2 switches are all operated? or only the upper portion is operating while the lower portion stays opened, when ADCINP is in operation?

    Q2: If AVSS=0V (AVDD=5V), does it mean the ADC input is charged/discharged to 1.3V during S2 on pulse

    Q3: We use 33 kHz clock (into XTAL1), what will be the typical resistance values as shown on Figure 37? How about the capacitor values?

    Q4: Could you please send me the detailed timing information on S1 and S2 for sampling time of 100 ms (see Figure 36)? And for other sampling time?

     

    Thank you very much

  • Hi Daniel,

    Answers to your questions:

    1. As I mentioned to Carlo, the ADC always takes differential measurements (ADCINP - ADCINN), even when measuring ground referenced signals. So both switches are active for all measurements
    2. As the datasheet says, capacitor CB is discharged to 0V which is the voltage seen by the inputs (ADCINP - ADCINN)
    3. Are you using a 33 kHz crystal or clock oscillator? If you use a crystal it should be 32.768 kHz. If you are using an external clock oscillator we would recommend 16 MHz. You will see this as a reference value for most of the datasheet parameters that do not involve the crystal. The typical resistance values depend on fCLK, as described in the datasheet on pg. 16: "if fCLK is reduced by a factor of two, the impedances will double"
    4. This is also described in the datasheet on pg. 16: "This two-phase sample/discharge cycle repeats with a period of tSAMPLE = 2/fCLK".

    -Bryan

  • Thank you very much Bryan,

     

    Just to make sure I understand correctly.

     

    For question #2, Datasheet mentions “CA1 and CA2 discharge to approximately AVSS + 1.3V and CB discharges to 0V.” With the AVSS connected to 0V in my circuit, ADCINP – ADCINN = 0V, but ADCINP by itself, with respect to ground is at 1.3V (at the moment S2 changes from Close to Open at almost the end of tSAMPLE as shown on Figure 36)?

     

    For question #3, yes we are using 32.768 kHz +/- 2% clock. The factor from 16 MHz to 32.768 kHz is 488. That means the 3 typical resistance values in Figure 37 are to be multiplied by 488?

     

    For question #4, sampling time is 2/fCLK. With 32.768 kHz, sampling time is 61 us. Based on the timing diagram in Figure 36 (page 16), it looks like S1 and S2 On times are about 40% each of the tSample, with about 10% gap before and after S2 pulse. Do you happen to have more accurate data?

     

    Thank you very much,

    Daniel

  • Hi Daniel,

    Answers to your questions below:

    • Yes, you are correct, each input is at approximately AVSS+1.3V when S2 closes
    • fCLK is the master clock, and should not be confused with the crystal frequency. If you are using a 32 kHz crystal, this is applied to an internal PLL & oscillator circuit such that fCLK is 15.729 MHz. If you are using an external oscillator, we recommend a 16 MHz clock. This is described on pg. 17 in the datasheet. The values in Figure 17 are calculated using a 16 MHz fCLK frequency, as shown. If you are using a crystal, then your fCLK is 15.729 MHz, and the resistance values can be calculated accordingly based on the 16 MHz values. I would also recommend against using a crystal, since there are many startup issues other customers have seen. You can search E2E to get an overview of these challenges. Instead, I would recommend a 16 MHz clock oscillator
    • Again, fCLK is not the crystal frequency, as described above. At fCLK = 15.729 MHz, tSAMPLE = 127ns.

    It might help if you could explain what you are trying to do with the ADS1258. It is unclear why this information is necessary, so maybe if I understood how you were going to use this device I could provide better support

    -Bryan

  • Thanks Bryan,

    We try to measure a voltage (without OP Amp buffer) at about 1V. Due the effect of charging the sample capacitor, I expect the reading voltage is slightly lower than the input voltage. it is something similar to the inrush current where high demand of current occurs, and voltage reduced slightly. But on the contrary, I saw that the reading is slightly higher than th1 1V input. So I suspect that the 1.3V was not discharged fast enough due to high source impedance.

    In our circuit, 32.678 kHz clock is applied to pin 8 (XTAL1) and pin 12 (CLKSEL) is grounded. Is the fCLK =32.678 kHz in this case?

    We have around 12 kohm and a shunt capacitor f about 10 uF at input to the ADC.

    Thanks,

    Daniel

  • Hi Daniel,

    Yes, your results thus far indicate that buffering would likely be necessary for your circuit, especially with 12k impedance at the input. I would highly recommend considering adding a buffer

    fCLK is the master clock, and should not be confused with the crystal frequency. If you are using a 32 kHz crystal, this is applied to an internal PLL & oscillator circuit such that fCLK is 15.729 MHz.This is described on pg. 17 in the datasheet.

    -Bryan

  • Thank you very much, Bryan.

    Issue Resolved.

    Best Regards,

    Daniel