• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Answers 1 answer
  • Subscribers 44 subscribers
  • Views 1236 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

ADS1220 input voltage range with PGA

Elder Costa
Intellectual 430 points
Other Parts Discussed in Thread: ADS1220

Hello, I just would like to clarify a point explained in the topic bellow:

http://e2e.ti.com/support/data_converters/precision_data_converters/f/73/p/358252/1258347.aspx

In my application the inputs have high value pull-up/down resistors so an input may go to a voltage near to VSS when the sensor is disconnected. Other than saturation, is there any risk of weird behavior of the PGA due to non linearity out of input dynamic range that could cause an unexpected reading instead of one near to zero when chosing an AINx-AVSS MUX channel?

  • 0
    Expert 2585 points

    Hi,

    According to the maximum "Analog input voltage" your input signal could be in the range  from AVSS – 0.3 to  AVDD + 0.3. So you could

    But you should be aware,  that for correct measurement you should  provide  correct differential and common mode range according to the  datasheet to avoid misinterpretation of your data.

  • 0 in reply to Oleksiy Pazyuk
    Intellectual 430 points

    Hi Oleksiy, thanks for your answer. I am aware I am out of the allowed range. My point is that under that circumstance (i.e. sensor disconnected) the input will be close to zero and the PGA may be saturated. If I read something close to zero or, more appropriately, far from the value the input exhibits when the sensor is connected, it will be fine (the exact value is irrelevant in this case).

    If however the PGA has a highly non linear behavior outside the common mode range and say saturates positively (as I observed to happen with some OpAmps used outside their input common mode input range) then I will need to change my circuit.

  • 0 in reply to Elder Costa
    Expert 2585 points

    Hi Elder,

    As I know the main limit factor for the PGA except absolute maximum input voltage is the fact that PGA is not Rail-to-Rail output amplifier. To linear measurement you should guarantee that PGA output signal is at least  0.2V from rail.

  • 0 in reply to Oleksiy Pazyuk
    TI__Guru** 113765 points

    Hi Elder,

    If you have the AINP connected to a pullup and AINN connected to a pull-down, then an open circuit will appear as positive full-scale while using the internal reference.  This will be true if the PGA is enabled or if the PGA is bypassed at all available gains.  This circuit is equivalent to Figure 68 of the ADS1220 datasheet, and is used to establish the correct common mode for the measurement when using gains greater than 4 with the PGA enabled.

    However, if you use AVDD/AVSS as the reference the outcome will be unpredictable with the PGA enabled at a gain of 1 as the PGA will limit the voltage to less than full-scale due to saturation.  So the main key point here is that if the reference voltage (full-scale) is less than the saturated voltage of the PGA then you will always read full-scale otherwise the output will be unreliable.

    Does this answer your question?

    Best regards,

    Bob B

  • 0 in reply to Bob Benjamin
    Intellectual 430 points

    Hi Bob,


    Not sure if it answers my question as the use case is a little different.

    Some extra info: I am sampling several channels in a sequence using the one-shot mode. Acquision rate is 1KSPS, ADS set to turbo mode and 2000SPS. I am using the internal reference.

    The sensor is differential and I sample the difference between its two outputs (AIN1-AIN0, gain 128). I also measure the voltage of each output referred to AVSS (AIN1-AVSS and AIN0-AVSS, gain 1) to determine if the sensor is connected and if there is any wiring fault. AIN0 has a 10M pull-down to AVSS and AIN1 has a 10M resistor to a 2V voltage source.

    Now, if the PGA simply saturates close to AVSS when I select the AIN0-AVSS channel then it is OK as I just need to know it is far from the baseline voltage of the sensor. On the other hand if it shows some bizarre behavior (e.g. saturates on the opposite rail than the "expected") then I will have to redesign the circuit.

    I found that bypassing the PGA to make the unipolar measures is not an option because it takes some time to stabilize aftwer power on, adding noise to the differential channel. I need the gain 128 for the differencial channel for its signal/noise ratio.


    Thanks in advance for your inputs.

    Elder.

  • 0 in reply to Elder Costa
    TI__Guru** 113765 points

    Hi Elder,

    In this case the AINP is within the correct common mode range but AINN is not.  This makes the differential measurement itself invalid.  However , if you are just looking for a gross error then you should be ok.  You will not see the output flip to the opposite rail if that is your main concern.

    As far as the single ended measurement is concerned you are in violation.  Neither measurment is valid when the PGA is enabled.  The AIN1 measurement may appear closer to 2V, but the AIN0 is more difficult to predict.

    I think you would be ok to actually just take the differential measurement.  An open at a gain of 1 should be very close to positive full-scale.  At a gain of 128 it should show saturated to positive full-scale.

    That leaves the shorted sensor case.  I'm not sure how the measurement at a gain of 1 will help you that much if the sensor itself is differential.  In my thinking a 0V output from the sensor would appear as shorted.  Am I missing something here?

    Best regards,

    Bob B

  • 0 in reply to Bob Benjamin
    Intellectual 430 points

    Bob, thank you for the input.

    One point that is worth mentioning is that the sensor is supplied with 5V (AVDD) and both of its outputs are close to 2.5V when in balance. Maximum output voltage is +/- 8.00mV in our application (thus our need for the 128 gain on the differential channel).