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.

ADS1256: ADS1256 FILTER

Part Number: ADS1256
Other Parts Discussed in Thread: OPA333, OPA350, REF5025

Hi,

i'm revising a PCB board based on ADS1256, it has a very noisy digital conversion ~4mV, and i'm working in it's filter, cause it doesn't have the filter indicated by the datasheet (VrefP-VrefN and analog input filters).

I saw it had some noise generated by the ADS in the Vrefp, in the sampling frequency, so i made some wire-ups and tested the circuit with the 49,9ohms and capacitors between Vrefp and Vrefn, it had a great improve, reducing the noise by 50%, and even with those wire-ups-noise-antennas.

So, now i'm re projecting the circuit, it's ok for vrefp and vrefn filter, but it uses three single-ended analog input for conversion, the datasheet just show filter for differential, so, should i  make three filters (based on the 301ohms and capacitors (datasheet)) between AIN's and AINCON (4 resistors, 6 capacitors)? Is it advisable?

Should i re-project for differential? How should be the hardware in this case?

Thanks,

Rodrigo

  • Hi Rodrigo,

    I'd be glad to help with your question.

    Just to clarify, the ADS1256 measures all input signals differentially.... The only difference then between a single-end and differential signal has to do with whether or not the negative (or common) input is fixed to known voltage potential or if this voltage fluctuates with the signal amplitude.

    Therefore, regardless of which signaling implementation you decide to use (single-ended or differential), I would always advise filtering the input signal differentially, meaning that you place a differential filter capacitor between the two input pins (whether those be between AIN0-AIN1, or AIN0-AINCOM, etc).

    If you have the ability to measure a truly differential signal from your sensor, then yes, I would try to implement differential signalling to the ADC. However, if your sensor only provides a single-ended voltage output (referenced to a fixed potential), then I would stick to your current implementation. Building a single-ended to differential output amplifier circuit often adds additional cost and degrades the overall noise performance, when compared to measuring the signal directly with a high-resolution delta-sigma ADC.

    One other question I would have for you, is can your sensor output negative voltages? If so, then it might make sense to bias AINCOM to a mid-range voltage to allow the sensor to swing around that common voltage.

    I hope that helps,
    Chris

    P.S. We'd be happy to review your revised schematic and offer any suggestions to help you get the best performance out the ADC!

  • Hi Christoher,

    thanks for your reply.

    I'll post here my old schematic and the new one, i'll be glad if you could validate before i update my project.

    Old schematic:

    New Schematic, changes on VREFN, VREFP, AINCOM and AIN's input.

    Thanks,

    Rodrigo

  • Hi Rodrigo,

    Thanks for sharing your schematics!

    The analog input filtering looks okay. However, I do see some potential issues with the the power supplies and reference that might lead to poor noise performance...

    • I would recommend removing the L8 and L12 inductors.

      The AGND and DGND pins on the ADS1256 should be shorted together to the same ground reference. Any impedance between these grounds can degrade the ADC's noise performance. Also, if the voltage potential between these two grounds becomes too large the ADC may not operate correctly or may even get damaged.

      On the supplies, inductors often cause a few negative effects such as ringing (due to the LC resonators formed by the decoupling capacitors), voltage spikes (when current transients occur), and they also choke the IC's supply current which may affect device operation. In general, I would not recommend using inductors on the ADC supplies.


    • Be aware that the OPA333 can probably only support a capacitive load of around 100 pF. Excessive load capacitance will likely result in opamp instability and ringing. In the case of the REF5025 and the ADS1256, you can probably connect the REF5025's buffered output directly to the ADS1256's reference inputs. Otherwise, you would want to replace the OPA333 with the OPA350 which can drive larger capacitive loads. For additional information on opamp stability, I would recommend the following E2E thread:

      Tim Green's series on Op Amp Stability - Amplifiers forum - Amplifiers - TI E2E support forums

      e2e.ti.com
      Other Parts Discussed in Thread: OPA4188 Several TI documents reference Tim Green's 15-part series on Operational Amplifier Stability. I have done extensive web


      Additionally, I would recommend routing the ADS1256's VREFN signal back to the REF5025's GND pin. This creates a star connection at the REF5025's ground pin and avoids ground noise from coupling into the ADC's reference. This doesn't require any changes on your schematic; however, it is something that requires attention to when creating the PCB layout.
       

    Best regards,
    Chris