Because of the holidays, TI E2E™ design support forum responses will be delayed from Dec. 25 through Jan. 2. Thank you for your patience.

ADS1220: Data Acquisition System Resolution Problem

Part Number: ADS1220
Other Parts Discussed in Thread: ADS1259, ADS125H01, ADS124S08

Tool/software:

In my project, I will develop a board: four channel datalogger for the strain, RTD sensors and also ratiometric measurement. Firstly I designed a board which I used two ADS1220 as an ADC.

0-5V measurement,

1-128 PGA gain,

120 ohm switch selected resistors for 4-20ma current measurement.

But I need help about the design.

1) When I writing code for the design I confused about the resolution of the ADS1220 is sufficient or not. In the datasheet the noise free bit is 13.42 bits in Turbo Mode and 660SPS datarate. I used digital filter: mean filter 12 reading for each channel, total readings are 24 for each ADC. The required time is about  1 / ((660/2)/12)= 36ms. The time is sufficient for my design but the noise free bit not. I need 16 bits resolution at least for each channel. Can the mean filter or other improve noise free bits?

2) Also I need 0-10V analog input for the industrial level measurement. Should I use the voltage divider and unity gain opamp? Any other suggestion.

3) Also I confused about using two or more ADC (ADS1220) each channel or multiplexing ADS1259 and PGA280-281. Is this a good choice for the 4 channel design.

But In this solution I have some question:

  • I need 1200SPS for scanning for channel and 18.4 noise free bits is enough for me but Do the gain and multiplexing stages affect this noise free bits. Also will I experience the crosstalk interference for channel to channel.

  • In ratiometric measurement, I will switch the internal reference to the external reference. For example I connect loadcell to the channel_1, RTD to the channel_2, thermocouple to the channel_3, ... Can I change the reference ( 2.5V internal ref. to 5V external ref. and vice versa) when switching from channel to channel. Does the settling time affect the total time or SPS time? 
  • For the ratiometric measurement I will use %0.2 LDO regulator for the excitation voltage (For example 350ohm strain gauge) and external reference input but for the thermocouple I want to use 2.5V internal reference. Any other suggestion?
  • Can I use ADS1259 internal reference to output from REFOUT pin as a sensor excitation with buffer and transistor like a linear regulator. Is this a good approach? If yes any suggestion about current limiting for the short circuit?
  • Hi Fatih,

    Can the mean filter or other improve noise free bits?

    The ADS1220 is a delta-sigma ADC, which means lower noise results are achieved with lower date rates due to higher over-sampling ratios. The following article helps explain in more detail How delta-sigma ADCs work, Part 1 (Rev. A)How delta-sigma ADCs work, Part 2.

    What are your measurements for the different channels? Not too sure what was meant here:

    The required time is about  1 / ((660/2)/12)= 36ms.

    2) Also I need 0-10V analog input for the industrial level measurement. Should I use the voltage divider and unity gain opamp? Any other suggestion.

    Using a voltage divider to make the signal be within the input range of the ADC is a solution. If you'd like an ADC that can directly measure 0-10V you could take a look at the ADS125H01/02: ADS125H01 data sheet, product information and support | TI.com

    3) Also I confused about using two or more ADC (ADS1220) each channel or multiplexing ADS1259 and PGA280-281. Is this a good choice for the 4 channel design.

    You can certainly use two or more ADCs for taking multiple measurements.

    If you'd like a device with a higher channel count for multiple measurements with the same ADC you could take a look at the ADS124S08 (ADS124S08 data sheet, product information and support | TI.com), which is an ADC with 12 input channels for similar sensor measurements.

    It is possible to interface different RTDs for example using the same reference resistor. Here is an example from our app note "A Basic Guide to RTD Measurements" (https://www.ti.com/lit/pdf/sbaa275): 

    Additional information on your system and all the intended sensor measurements would help in providing any additional suggestions.

    Best Regards,

    Angel

  • Hi Fatih,

    Please expect delayed responses on E2E the following week due to the Holidays in the US.

    Best Regards,

    Angel

  • Hi Angel,

    Thank you very much for your reply.
    The required time is about  1 / ((660/2)/12)= 36ms.
    Measuring sensors in noisy industrial enviroment is crucial in this application. So, I used bubble sort to eliminate max-min peak values and then taking average of remaning values. The above equation means: SPS=660 and 12: adc value readings for bubble sort, /2: for 2 channel of an ADC. 36ms is required for filtered final value for each channel. This is not necessary for Delta-Sigma ADC. These operations are already done within itself??
    3) Also I confused about using two or more ADC (ADS1220) each channel or multiplexing ADS1259 and PGA280-281. Is this a good choice for the 4 channel design.
    I want to measure 4 independent channel to take signal from sensors. The PT100 ratiometric reference connection issue is a bit more challenging for me than other sensors. Is this measurement possible with two ADS1220s? Or do I have to use 4? An example connection is as follows,
    1. Channel: PT100 (RTD ratiometric reference measurement)
    2. Channel: Loadcell (5V excitation and ratiometric reference)
    3. Channel: LVDT (0-5V ratiometric reference measurement)
    4. Channel: Pressure transducer (4-20mA)

    Also, If I use voltage divider resistors for 10V analog input range, Is there any problem with connecting PT100 to ADS1220 input? (Such as leakage current from resistor to ground.)

     

  • Hi Faith Cavdarci,

    Our apps support is limited due to the holiday here in the US. Please expect a response sometime next week. Thanks for your patience.

    -Bryan