PT100 circuitry(ADS1247/48)

Hi all!!!

I want to use ADS1247 for temperature sensing with PT100.

Either 3 wire or 4 wire circuit is fine, but please specify what is the advantage or disadvantage using any of them...?

In the document slyp160(http://www.ti.com/lit/ml/slyp160/slyp160.pdf) page18, the PT100 is used with the bridge circuit, Is it really required to use a bridge circuit with PT100..?

Thanks and regards

-Anand

  • Hello Anand,

    The ADS1247 is perfectly suited for use with PT100 RTDs. Three-wire or four-wire connections are simply ways of improving the accuracy of a resistance measurement by cancelling the resistance of the wires used to make the measurement. This is an issue because resistance measurements are performed by using wires to force a current through the resistance and measuring the resulting voltage drop. However, the wires themselves also have a resistance (called lead resistance), so the voltage drop that develops across the wires themselves can negatively affect the accuracy of the measurement.

    Three-wire and four-wire connections correct for this by measuring the voltage drop with additional wires (called sense wires) through which no current flows, so no additional voltage drop occurs. The "Operating Principle" section in the following article does a great job of explaining this concept: Four-wire Sensing. In general, a three-wire RTD connection is sufficient since the voltage drop from the "positive" force wire is the same as the voltage drop from the "negative" force wire or common wire. 

    A bridge circuit was used in the document you linked because the instrumentation amplifier used (INA114) did not have built-in current sources available. The ADS1247 has two built-in, well-matched excitation current sources which can be digitally programmed to certain values (50μA, 100μA, 250μA, etc.) so the additional bridge circuit is not required. If you follow the example of Figure 85 on the ADS1247 data sheet (shown below) you can get excellent performance with a three-wire PT-100 RTD.

    Best regards,

    Ian Williams
    Linear Applications Engineer
    Precision Linear and Sensing Products 

  • In reply to Ian Williams:

    Hi Ian!!!

    Thanks for you quick reply,

    So, shall I conclude that I need to use a bridge circuit with PT100 if built in current source is not available?

    Thanks and regards

    -Anand

  • In reply to Anand Kumar Rai:

    Hello Anand,

    You don't necessarily need a bridge circuit, but if you only have a voltage source available then you must use a resistance to generate a stable current source of some kind. Bruce Trump has an excellent article on how to achieve analog linearization of RTDs that uses a couple of resistors to generate the current source. The article also includes an Excel calculator and simulation circuits in TINA-TI which are very useful. Check it out here: http://www.ti.com/lit/an/slyt442/slyt442.pdf

    Best regards,

    Ian Williams

  • In reply to Ian Williams:

    Hi Anand,

    In addition to the methods that Ian and Bruce suggest, here are some easy ways to create current sources to excite your RTD.  Voltage excitation results in an inherently non-linear response because the excitation current will change through the RTD over the span or the RTD temperature measurement as the RTD resistance changes.  The circuits shown in the figure below can be used to create 100uA current sources and current sinks.  Please let us know if you're interested in one of these circuits and we can help tailor it to your desired current output.  If you choose to use one of the circuits in Bruce Trump's nice article then the non-linear behaviour of the voltage excitation is not an issue and cancelled with the circuit.

    Regards,
    Collin Wells
    Precision Linear Applications

  • In reply to Collin Wells:

    Hi Collin and Ian!!

    After going through TRUCE application notes suggested by Ian, I observed that my temperature range is very less (-50 deg C to +120 deg C) and for this range non-linearties of PT100 will not affect much and our BOM cost is not allowing us to use an extra Opamp to use as current source. We have decide to go with  ADS1015 for our application. So, please suggest a 3-wire or 4-wire circuit where RTD(PT100) can be excited by voltage source.

    Thanks and regards

    -Anand

  • In reply to Anand Kumar Rai:

    Hi Anand,

    The RTD nonlinearity will account for around 0.7% error over your temperature span in addition to the nonlinearity due to the voltage excitation. 

    4-wire measurements are not practical when using a voltage excitation source unless you're willing to measure the current flowing through the RTD which I'm going to guess you're not.

    See below for examples of 3-wire measurements with current and voltage excitation.  The "Rl" resistors represent the resistance due to the wiring and/or connections of the RTD to your circuit.  The 3-wire measurement with current sources enables a complete cancellation of the Rl resistors.  The 3-wire voltage measurement is better than a 2-wire circuit but does not completely cancel the lead resistors. 

    If you're curious as to how to set the resistors I suggest you download a copy of TINA-TI ( http://www.ti.com/tool/tina-ti ) and start having fun measuring the currents and voltages!

     

    Regards,
    Collin Wells
    Precision Linear Applications

  • In reply to Collin Wells:

    Hi Collin,

    Thanks for the details of the temperature sensing and nonlinearity accounting for my temperature range. We will go with the voltage circuit first, and will take care of nonlinearity and errors in the second iteration of hardware.

    Thanks n regards

    -Anand

  • In reply to Anand Kumar Rai:

    Hi Collin,

    Sorry to revoke this thread once again.

    We are now ready for using a current source and 4-wire RTD connection for exciting the RTD to be used with ADS1015. Below mentioned are the details:

    Temp: -50 to 120 degree = 170 degree x 0.1K = 1700 step

    Resistance at -50 degree = 80.31   R

    Resistance at 0 degree = 100.00 R

    Resistance at 120 degree = 146.07 R 

    Please suggest a cheap current source circuit that can be used for the above mentioned requirements ASAP.

    Thanks and regards

    -Anand

     

  • In reply to Anand Kumar Rai:

    Hello Anand,

    Sounds like a good application.  Out of curiosity, what are you measuring the temperature of?

    The REF200 is a very easy to use current source that can be used to source/sink 100, 200, 300, or 400uA.  These currents work very well for a high-performance 100Ohm RTD measurement circuit where self-heating of the RTD must be minimized.

    Otherwise, I would suggest one of the circuits I've shown a few posts up that only involve an OP-AMP and a MOSFET.  The cost of this solution will be heavily dependant on the precision of the current setting resistor (RSET) and the performance of the control amplifier.  An amplifier with low-offset-drift will help reduce the effects of age and temperature shift on the circuit's accuracy after an initial calibration.  A part with low-offset and low offset-drift may prevent the requirement for offset calibration all together. 

    Regards,
    Collin Wells
    Precision Linear

     

  • In reply to Collin Wells:

    Hi Collin,

    Thanks for you quick support,

    As I see the BOM cost, REF200 sounds a costly solution. So, I would prefer to go with OPA333 and MOSFET. But my doubt is which current source value 100, 200, 300, 400uA would best suit my requirement. And how shall I tailor the values of resistors to be used with opamp. Apart from that, MOSFET to be used should be any special functionality MOSFET or ordinary MOSFET is fine.

    Yup application point of view I can tell only one thing that design is going to be used in a mass production product.

    Thanks and regards

    -Anand