• State Resolved
  • Locked Locked
  • Replies 2 replies
  • Answers 1 answer
  • Subscribers 46 subscribers
  • Views 315 views
  • Users 0 members are here
Support feedback
Options
Options
Related

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.

ADS1120: Software configurable 0-1V, 0-2.5V and 0-5V single ended, unipolar inputs?

Subir Bhaduri
Prodigy 10 points
Part Number: ADS1120
Other Parts Discussed in Thread: ADS114S08, , ADS1220

Hi. 

I need to design a common generic circuit for interfacing arbitrarily connected single-ended sensors which may have multiple ranges, such as 0-1V, or 0-2.5V or 0-5V. All signals will be single ended and unipolar.

The configuration of the registers can be controlled via SPI commands on a case to case basis. I do not want to make the ports rigid on the PCB end, like using opamps or jumpers, etc.

My AVDD=5V unipolar. This question is to check if i have understood the design correctly.

  1. First of all, the resolution will be limited to 15bits (1 bit lost due to unipolar PS and signals).
  2. PGA needs to be disabled for all these measurements.
  3. For, 0-5V signals i can just use AVDD-AVSS as the reference and a gain of 1.
  4. For V range between 0-2.5, i can use an external reference connected to REF0P and REF0N.. Can you recommend a 2.5 ref chip? or can i get away with a voltage divider?
  5. For a signal of 0-1V, I can use a gain of 2 and internal ref of 2.048, right?
  6. If I wish to measure a voltage swinging around 2.5V (1.5V lowest, to 3.5 highest), is there some shifting i can do to without using any opamps to optimize the measurement?

Finally, i was wondering if the ADS114S08 is a better IC as it has more ports and i can play with making the AINN part either connected to AVSS (for single ended unipolar measurements) or use the incoming signals refered to AINCOM which is further connected to the 2.5 Vref and thus achieve a single ended by bipolar measurement. Any recommendations will help me.

thanks!

  • +1
    TI__Mastermind 26808 points

    Subir,

    Below are comments to your 6 points.

    1. First of all, the resolution will be limited to 15bits (1 bit lost due to unipolar PS and signals). Correct.
    2. PGA needs to be disabled for all these measurements. The PGA has common mode limitations as described in the data sheet.  The common mode limits do not allow swing to swing the full range that you want, so I agree with your assumption. 
    3. For, 0-5V signals i can just use AVDD-AVSS as the reference and a gain of 1.  Correct.  However, keep in mind that the results here are dependent on the accuracy of AVDD-AVSS.  Generally this is intended for ratiometric measurements.  That is, measurements where the sensor also uses the same supply so that as the supply drifts the sensor drift cancels the reference drift.  You may want to use a precision external 5V voltage reference if your input signal is not ratiometric.
    4. For V range between 0-2.5, i can use an external reference connected to REF0P and REF0N.. Can you recommend a 2.5 ref chip? or can i get away with a voltage divider?  When you bypass the PGA the input is no longer high impedance.  So, using a voltage divider with the PGA bypased will introduce some error.  The data sheet shows input current vs voltage and temperature for both the PGA and bypased mode.  You can use this current to give you an idea of what errors you might expect.  So for best accuracy you may choose to avoid the voltage divider.  The REF5025 is a very good low noise voltage reference.
    5. For a signal of 0-1V, I can use a gain of 2 and internal ref of 2.048, right?  Correct
    6. If I wish to measure a voltage swinging around 2.5V (1.5V lowest, to 3.5 highest), is there some shifting i can do to without using any opamps to optimize the measurement?  The ADS1120 does not have any shifting built in.  I recommend using your 0V to 5V range for this operation.

    One general comment.  If you want a higher degree of accuracy you could switch to ADS1220 (24 BIT version).  Adjusting the reference  input for each of your input cases as you have planed may provide some accuracy improvement, but it also adds complexity to your firmware.  Using one fixed reference, and a higher resolution device would simplify your firmware and improve your overall accuracy (with some additional cost).  The ADS114S08 has comparable specifications to the ADS1120 and does offer additional channels so I think it is a good choice.  It also has an internal 2.5V reference as opposed to the 2.048V which may help your application.  Both of these devices have evaluation modules.  You could use these to confirm your choice.

    Let me know if you have additional quesitons.

    Art

  • 0
    Prodigy 10 points

    Thanks Art Kay for the quick and detailed response. I understand taking up a 24bit ADC can help get around the different scaling up issues i could face with the ADS1120.