• State Resolved
  • Locked Locked
  • Replies 7 replies
  • Subscribers 47 subscribers
  • Views 6833 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.

ADS1248 - Pt100 measurement with bipolar power

Dragan Zivanovic
Prodigy 40 points

Other Parts Discussed in Thread: ADS1248

Please find proposition of schematic for universal analog input (one at the time),from TC, RTD and 20mA based on ADS1248.
Application is with bipolar power, optional grounded sensor, slow industrial measurement with lot of disturbances. 
If pt100 sensor, results is AIN0-AIN1, with boths current sources switch on.
For voltage input, resutls is AIN4-AIN5, and for 20mA input, results is AIN6-AIN5.
GPIO0,1, and AIN2,3,7 can be used for other functions.
Questions:
1) Whether this configuration for measurement of RTD is possible at all? Common ground pin is used for all optional sensor,
and also, mismatch of current sources do not influence to measurement of pt100 itself, only to compensation of parasite resistances.
2) Whether it is good, and enough  in order to decrease ESD sensitivity of chip, to insert 2k2 resistors in the lines with inputs, according to schematic?

Thank you and Best Regards,

Dragan 

 

  • 0
    TI__Guru 55466 points

    Hi Dragan,

    The schematic looks great overall.

    1)  The RTD configuration will work.  The only thing to consider is that the ADS1248 offers an IDAC current setting of 250uA (but not 200uA).  Nevertheless, this still will work on the design.  One possible suggestion (not completely necessary) is to increase slightly the reference voltage to a ~1.8V  reference for better noise performance (provided enough headroom is available to the AVDD supply when the RTD is at the maximum temperature range).  Yes, the IDAC mismatch does not affect the measurement in this topology.

    2)  If the device could be exposed to a 24V or 30V supply due to a bad sensor connection,  a possible suggestion is to increase the protection resistors to >2.4KOhm to limit the current to 10mA during a fault condition. The ADS1248 can withstand up to 10mA continuous currents at the inputs per the Absolute Maximum Ratings Table.

    Thank you and Best regards,

    Luis 

     

  • 0 in reply to Luis Chioye
    Prodigy 40 points

    Hi Luis, 

    Thank you very much for really fast answer and very good words.

    1) Yes, I sow that 250uA can not be set :). It is good to increase reference voltage, but for Pt1000 range, currents must be decreased, or R1 must be decreased, not to exceed maximal output voltage of IECX1. It should be under VDD-0.7V according to data-sheet, or smaller? This voltage is not exact defined in data sheet

    2) Is it good to increase protection resistors up to 10k? How big resistors can be not to make bigger error than 10uV? 10uV or 20uV looks small enough for pt100, and also for low voltage thermocouples. Is those resistors really enough to protect ADS1248 in industrial environment? With serial resistors of 100k, even 230V can not hurt it :). If varistors or additional protection diodes are added in the front of chip, then currents through those diodes will make bigger error for Pt100 than any other in the chip, is it true?

    Best regards,

    Dragan 

  • 0 in reply to Dragan Zivanovic
    Mastermind 7395 points

    From my experience with applications for industrial environments, since analog inputs have very low input impedance, you have to add not only ESD protections but also lightning arrestors (for example, discharge gas protections).

    Depending the application you may also need to isolate the sensor side from the main processor side. In many projects I did the A/D front end was on the non isolated side, the data from the A/D was sent serially via optocouplers to the main processor. P/S for the A/D also had isolation. All this was because it was mandated to isolate all I/O (digital and analog) from control board which can be in contact with human operator.

    So, it is a question of requirements, you have to ask your system and marketing engineer what are the requirements for standards compliancy.

  • 0 in reply to Avi Chami1
    Prodigy 40 points

    For non grounded intelligent transmitter application with plastic housing, you can isolate input side, or maybe easier to isolate from uP to output side, and of course power of both sides .

    You are very right for requirements for standards. No matter, it is good to know  what can be done with input configuration of ADS1248 itself. Any additional varistor or ESD diodes in input will decrease accuracy.  

  • 0 in reply to Dragan Zivanovic
    Mastermind 7395 points

    I am not working for PLC designs for a long time, but even with the technology of more than 10 years ago we were able to achieve real precission of 11-12 bits.

  • 0 in reply to Avi Chami1
    TI__Guru 55466 points

    Hi Dragan,

    1)  The typical IDAC compliance at 250uA (or 100uA) setting is around 0.7V.  This will be an issue for the PT1000 sensor where you will run out of headroom to the AVDD supply.   A quick solution would be to adjust the IDAC to 100uA, and scale the VREF resistor to obtain the VREF=1V accordingly.    This may also allow you a little more headroom to increase the IEXC pin protection resistor.  The PGA gain could adjusted for the smaller differential voltage produced by the sensor. Atttached is the figure showing the typical IDAC compliance voltages.

     

    2)  The input bias current is in the order of ~10nA; and the input bias current mismatch is in the order of 100pA.  Using 10kOhm 0.1% resistors at the inputs should produce errors in the order of 1uV considering the input bias current mismatch. The design below shows a possible filter/protection circuit that could be adapted into your design.  A similar filter/protection circuit can be used at the inputs in your design.    This design could also be used with bipolar supplies by referring VREFN to AVSS.  The filter time constant across VREF and the ADC inputs of the device are also matched for good ratiometric cancellation.

    Thank you and Best Regards,

    Luis

     

  • 0 in reply to Luis Chioye
    Prodigy 40 points

    Hi Luis,

    Thank you very much,

    All of those will help me a lot. I am little not sure about transil with inverse current of 1uA  connected to  lines where current is 200uA, but it is not so hard to test transfer characteristic with and without transil.

    Best regards,

    Dragan