• State Resolved
  • Locked Locked
  • Replies 3 replies
  • Subscribers 45 subscribers
  • Views 804 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.

ADS1247 with 3-wire RTD

styg
Prodigy 160 points
Other Parts Discussed in Thread: ADS1247

Hello, I intend to use the ads1247 in a project with 3-wire RTD to measure temperature from 0 to 500 ° C.
I want to use as reference voltage, the Rbias between pins and REFP Refn, the issue of noise.

Some questions:

1 - I have questions regarding the Rcomp in that case I should use a 100ohm resistor (resistance of PT100 at 0 ° C) to adjust my range between 0 and 500 ° C?

2 - And if due to resistor tolerance, the resistor value is 99ohm (for example), this difference will cause an error, how to fix this?

3 - Setting the current to 1mA of IDAC, the voltage generated by PT100 in the range of 000-500 ° C is 0-281mV, with a 1.2 Kohm Rbias, the reference value is 2.4 V (1.2 kohm x ( mA 1mA +1))?

4 - In this case, with PGA equal to 8x, the range will be + -2400mV, but only used the positive full scale, so are 23-bit resolution for a full scale of 300mV (~ 558 ° C ), resulting in a resolution of 0.036uV or 0.000067? ok?

Thanks in advance.

  • 0
    TI__Guru** 113765 points

    Styg,

    See my comments below.

    Best regards,

    Bob B

     

    styg said:

    Hello, I intend to use the ads1247 in a project with 3-wire RTD to measure temperature from 0 to 500 ° C.
    I want to use as reference voltage, the Rbias between pins and REFP Refn, the issue of noise.

    Some questions:

    1 - I have questions regarding the Rcomp in that case I should use a 100ohm resistor (resistance of PT100 at 0 ° C) to adjust my range between 0 and 500 ° C?
         [BB] You do not have to use Rcomp.  What Rcomp does is place the reading within a desired range for the differential input.  For example, if Rcomp is 100 ohms, and the RTD is 100 ohms, the voltage drop across them will be the same.  So a differential voltage of zero would be 0 C.  If the RTD goes below 100 ohms, then your reading would be below 0 C.  If the RTD goes above 100 ohms, the differential reading would be positive and the reading above 0 C.  In this way you have some +/- value referenced to a desired temperature.  If you do not use Rcomp, you will always have a positive differential reading.  The voltage drop read across the resistor is measured and will always be a positive differential input to the ADC.
    2 - And if due to resistor tolerance, the resistor value is 99ohm (for example), this difference will cause an error, how to fix this?
         [BB] You need to calibrate your system.  You can also use higher precision resistor, like 0.1 or maybe 0.5 %, but you should still calibrate your system to get a truly accurate reading.
    3- Setting the current to 1mA of IDAC, the voltage generated by PT100 in the range of 000-500 ° C is 0-281mV, with a 1.2 Kohm Rbias, the reference value is 2.4 V (1.2 kohm x ( mA 1mA +1))?
         [BB] Yes, 1mA from each IDAC would flow through Rbias resulting in a 2.4V drop.  2.4V would be your reference voltage.
    4 - In this case, with PGA equal to 8x, the range will be + -2400mV, but only used the positive full scale, so are 23-bit resolution for a full scale of 300mV (~ 558 ° C ), resulting in a resolution of 0.036uV or 0.000067? ok?

         [BB]  Your full scale input range is +/-Vref/PGA = _/-2.4V/8 = +/- 300mV.  1 LSB is Vref/PGA/(2^23-1) = 36 nV.

    Thanks in advance.

  • 0 in reply to Bob Benjamin
    Prodigy 160 points

    Hi Bob,

    in the case of using Rcomp would also enjoy my full range of the ADC, as with 0 ° C in I would have 100mV ADC (1 / 3 of my range).

    Moreover, as was done in an example in the datasheet, I could cause a fall of 1 / 2 of my range with Rcomp, making my range of 300mV, 0-150mV to stay negative in the AD and 150-300mV stay positive, taking advantage of the 24bits, right?

  • 0 in reply to styg
    TI__Guru** 113765 points

    Styg,

    The ADC input is differential.  This allows you to take a measurement where the result is two's complement.  This means that with a Gain of 8 and Vref of 2.4, one LSB is Vref/Gain/(2^23 - 1), which equates to about 36 nV.  The actual range is +/- 300mV (+/- Vref/Gain).  If you choose to use a Rcomp resistor value of 100 and you only take temperature measurements above 0 C, you will only be using 1/2 of your possible range which is 0-300mV.  Your LSB size is still a 24 bit value where one LSB is 36nV, but your range is reduced to 23 bits.

    If you set your 0V differential input point mid-scale of the range of temperature you will be reading (250 C), you would size the Rcomp resistor so that the resistance is equal to the RTD value for 250 C.  If the RTD is linear through the entire temperature range you could potentialy have readings that are +/- 140 mV (assuming a full scale range of 280mV).   Nothing to this point changes range or LSB size.  What changes is the readings are now +/- readings.

    The power of using the Rcomp resistor is that you can now reduce your range and LSB size if it is matched appropriately.  If you set it so that the input full-scale range is +/- 140 mV, you can change the Gain to 16.  This reduces the ADC full-scale range to +/- 150 mV, and one LSB to 18 nV and makes better use of the converter's range.

    A couple of things to keep in mind.  First of all you need to make sure that the voltage that is input to the ADC is within the specifications for the common mode input range at the Gain (PGA) levels you want to use.  Another very inportant thing to consider is noise.  Don't expect to get 24-bit performance.  You can get 24-bit resolution, but the ENOB will be closer to 21 assuming you have a very quiet system. 

    Best regards,

    Bob B