• State Resolved
  • Locked Locked
  • Replies 4 replies
  • Answers 1 answer
  • Subscribers 40 subscribers
  • Views 619 views
  • Users 0 members are here
Support feedback
Options
Options
Similar topics

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.

Original question:

WEBENCH® Tools/INA326: tidu969

WEBENCH® Tools/INA327: INA326EA

Sisto Zanier
Prodigy 70 points
Part Number: INA327
Other Parts Discussed in Thread: INA326, , INA333, ADS8866

Tool/software: WEBENCH® Design Tools

Hi everyone,
with reference to the TIDU969-May 2015, I performed a simulation with Multisim 13 from National Instruments, reporting exactly the same scheme, but the simulation results are absolutely different from what was expected and reported on the application note.

The output voltage value with RTD value 80.31ohm at -50 ° C is 0.117V as reported in the application note, but the RTD value at 150 ° C reported in table 3, of 149.94 ohm, is incorrect and should be 157.33 ohm. However, apart from this error, setting the wrong value of 149.94 ohm (which does not exist in the table of ITS-90 and alpha = 0.00385 values), the output voltage is not 4.82V, but close to the saturation of the amplifier , or 4.931V. By setting the RTD value 149.83ohm, i.e. at 130 ° C, similar to that of the TIDU969 application note, the output voltage is still not close to 4.82V, but 4.919.

Where is the problem?

Could it be that the Multisim Spice model for the INA326 component is wrong?

  • 0
    TI__Guru 51411 points

    Hi Sisto,

    The output voltage value with RTD value 80.31ohm at -50 ° C is 0.117V as reported in the application note, but the RTD value at 150 ° C reported in table 3, of 149.94 ohm, is incorrect and should be 157.33 ohm. 

    You are correct that RTD value should be approx 157 Ohm range at 150C. There is a typo in these figure.

    Assume -50C, R=80.3Ohm, 0C, R=100C. The slope will be (100-80.3)Ohm/50C = 0.394 ohm/C. If we assume a linear behavior, the temperature at 150C should be equal to 0.394Ohm/C*150C+100 = 159.1 Ohm. With R= 147.9Ohm, the sensor's temperature should be 121.6C (RTD/PT-100 sensor is fairly linear temperature sensor in the temperature range). 

    apart from this error, setting the wrong value of 149.94 ohm (which does not exist in the table of ITS-90 and alpha = 0.00385 values), the output voltage is not 4.82V, but close to the saturation of the amplifier , or 4.931V. By setting the RTD value 149.83ohm, i.e. at 130 ° C, similar to that of the TIDU969 application note, the output voltage is still not close to 4.82V, but 4.919.

    Yes, with Alpha=0.00392 PT-100 sensor, R=148.75C at 125C, with Alpha=0.00385 PT-100 sensor, R=147.95 Ohm at 125C. However, I think that the Vout vs. RTD Resistance plot is correct, also see the attached simulation. 

    https://www.pyromation.com/Downloads/Data/385_c.pdf

    https://technitemp.com/rtd-pt100-temperature-resistance-table/

    With RTD=149.83Ohm (130C in alpha = 0.0385 sensor), Vin+ - Vin- = 7.11mV, The Vout = gain*7.11mV = 4.9628V, where Gain = 2(698/2) = 698

    There are several typos in the application note. However, author did not say what RTD sensor is used in the article. It is saying that Vout vs. RTD resistance in the circuit behaves linearly from 80.3 Ohm to 147.9 Ohm between -50C to +125C. Of course, the circuit will be linear beyond 125C, but the INA326's supply voltage has to be increased to meet the Voltage output Swing requirements.  Or you have to lower the gain settings in the circuit, if you want to use 5V rail. 

    Below is the Vcm and Vout swing requirements in INA326T.

    Below is the circuit design requirements. 

    Enclosed is the above simulation in Tina. 

    /cfs-file/__key/communityserver-discussions-components-files/234/INA326-TIDU969-E2E-08212020.TSC

    If you have additional questions, please let us know. 

    Best,

    Raymond

  • 0 in reply to Raymond Zhang1
    Prodigy 70 points

    Raimond, thank you for your answer which resolves this question.
    However, in the context of my larger project, I need to integrate the management of a PT100, in a temperature range from -50°C to 140°C with a resolution of 0.1°C, but necessarily powered at 3, 3V like all the rest of the project.
    Now the question is: are INA326 (or INA327) suitable for this use and if so, what problems could I encounter?

    Multisim simulation would confirm this possibility, but as already pointed out it depends on how accurate the Spice model is.

    For this reason, I believe it is better to purchase the TIDU969 evaluation board and carry out real tests with 3.3V power supply. What do You think about it?

    As for the linearization, it would be my intention to use a lookup table of all the degree by degree values and to use the Callendar Van Dusen formula to calculate the tenths of a degree considering linear the interval between the two values 1 ° C apart.

    Thanks for Your support,
    Sisto

  • 0 in reply to Sisto Zanier
    TI__Guru 51411 points

    Hi Sisto,

    Yes, it is good ideas to get EVB for this type of accuracy. The simulation can only do so much. Please also take a look at INA333 for low voltage application, 

    Enclosed are some of our op amp and instrumentation amplifier selections and application notes, which you may find it helpful.

    https://e2e.ti.com/support/amplifiers/f/14/t/818894

    https://e2e.ti.com/support/amplifiers/f/14/t/818894

    You may need to use higher precision resistors to achieve 0.1C resolution from -50C to 140C, which is consider a relative wide temperature range. My recommendation is to extract the Temp and R values from a given PT100 sensor, and perform a curve fit within -50C to 140C. The instrumentation amplifier is accurate enough to convert temperature to voltage linearly. If you are able to curve fit Vout vs. R from -50C to 140C (or fit Vout vs. Temp), then you should be able to achieve the accuracy or exceed it. 

    RTD's Resistance vs. Temperature from -50C to 140C is fairly linear, but it is not perfect linear. TheTech note uses 3rd order polynomial to fit the curve with two curve fit equations for two different temperature range. You may try the same approach or you may work out your own curve fit equation to cover the -50C to 140C range in a signal conversion expression. 

    If you need further assistant, please let us know.

    Best,

    Raymond

  • 0 in reply to Sisto Zanier
    Prodigy 70 points

    ... finally, I tried to simulate the scheme using Tina 9, but changing the supply voltage from 5V to 3.3V and the gain resistor from 2K to 4,174.9K ohm (4.12K 1% 54.9 ohm 1%) to obtain 2.5V at the maximum temperature value at 140 ° C. For A/D conversion I will use an ADS8866

    The result is equal to the simulation with Multisim 12 wich tells me that the solution is applicable for my case.

    Thanks a lot for Your support and technical documentation links.
    I seen the RTD American curve table, alpha = 0,00392, but i will to use a RTD with European curve, alpha = 0,00385

    Thanks again and with this post, resolve my issue.