• State TI Thinks Resolved
  • Locked Locked
  • Replies 6 replies
  • Answers 2 answers
  • Subscribers 50 subscribers
  • Views 234 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.

PGA309: Sensor sensitivity

Cherry Zhou
Mastermind 22235 points
Part Number: PGA309

Hi team,

1) Use the PGA309 to amplify the differential signal of the pressure sensor, but the theoretically calculated amplification does not match the amplification achieved by the measurement:

The PGA309's VSA, VSD, are all 5-V DC powered with a common mode input voltage of 1.450 V and a differential input voltage of 10.1 mV. The PGA309 operates in test mode.

The internal settings are as follows: Internal Reference 2.5 V, Coarse Offset adjust set to 0 V, ZeroADC set to 0, GainADC set to 1.0. The previous PGA has a magnification of 32, followed by a magnification of 2. 

The theoretically calculated voltage value is [(10.1+0)*32+0]*1.0*2=646.4mv, but the actual measured VOUT value is 746mv. Why's this?

Are the following 3 parameters affecting the amplified output value? What do these parameters mean? 

2) External Sensor Output Sensitivity: 1-245mv/v on the datasheet, is this referring to the sensitivity of the PGA309 external sensor? If the PGA309 is continued with a sensitivity of 16uV/v for the external sensor, can a valid voltage value be measured by the magnification factor? 

Could you help check this case? Thanks.

Best Regards,

Cherry

  • 0
    TI__Genius 17757 points

    Cherry,

    1. Your calculation is correct, and your measured result is reasonable for this device.  The error may seem large, but it is important to understand that the gain and offset accuracy of the PGA309 is not great without calibration.  There are many different gain, offset, and reference blocks in the device and the errors compound.  The way the PGA309 achieves good accuracy is through calibration.  During calibration the output is measured relative to a minimum and maximum stimulus and the gain and offset are adjusted to achieve the desired output swing.  This is done in a few iteration so that the errors in the internal gain and offset controls of the PGA are calibrated out.  The values you circled in the tables are the various errors associated with the gain and offset blocks.  Yes.  These errors are large, but after calibration the errors will be corrected for and the overall accuracy will be quite good (typical errors less than 0.1% for PGA+Sensor.
    2. The "External Sensor Output Sensitivity" is an indication of the sensor span ranges that can be used with the PGA309.  This means that the PGA309 will work well with sensors from 1mV/V to 245mV/V.  Sensors with sensitivity less than 1mV/V will not achieve a full 5V output swing.  You can test the output swing by multiplying the sensor sensitivity by the excitation voltage and the maximum gain.  So, for 1mV/V the max output range is: Vexc*Vsensor_sen*Gin*Gout = 5V*1mV/V*128*9=5.76V.  If you try the same for a 16uV/V sensor you get: 5V*16uV/V*128*9=0.092V.  Thus, the PGA309 will not work for a 16uV/V sensor as it does not have enough gain.  
      1. Your sensor has a very low sensitivity.  Are you sure that you have this specification correct?  For the PGA309 when we say "sensor sensitivity" or "sensor span", we mean that the bridge exaction voltage multiplied by the sensitivity is the maximum output of the sensor.  So for your sensor applying a 5V excitation will give an 80uV output (5V*16uV/V = 80uV).  Is this how you understand your sensor specification?  If so, the PGA309 will not work for your application.
      2. Initially you mentioned a 10mV input to the PGA309.  That doesn't match the sensitivity of 16uV/V.
    3. Do you have the PGA309EVM?  If you do you should be able to do the calibration automatically.  The calibration will not work for sensors with spans outside the 1mV/V to 245mV/V range but otherwise it should work automatically with a low post calibration error as the internal errors of the PGA309 are correct by the calibration procedure.

    I hope this helps.

    Best regards,

    Art

  • 0 in reply to Art Kay
    Prodigy 20 points

    Thank you!

    I am the questioner. I already know what sensor sensitivity means. This is the specification of the pressure sensor. The sensitivity of the sensor is 16uv/v/mmHg. We used a pressure range of 300mmHg, so the sensitivity of the sensor is 16uv/v/mmHg*300mmHg=4.8mv/v. If the accuracy of measurement is 1mmHg, when 5V excitation voltage is used, the variation of voltage should be accurately measured as 16*5=80uv. I am worried that the error of front-end PGA is too large, leading to insufficient accuracy.

  • 0 in reply to Art Kay
    TI__Mastermind 22235 points

    Hi Art,

    Thanks for your support and lio is the customer I mentioned, you could keep discuss on this issue here and please let me know if any questions.

    Thanks and regards,

    Cherry

  • 0 in reply to lio
    TI__Genius 17757 points

    Lio,

    A sensitivity of 4.8mV/V is a fairly typical use case for PGA309.  With this level of signal the expected post calibration accuracy of the sensor module is 0.1% of full scale.  In this case, full scale is 300mmHg, so the expected worst case post calibration error is 0.1% of 300mmHg, or 0.3mmHg.  This leaves significant margin in your target accuracy of 1mmHg.  The key thing to remember here is that the internal PGA errors will be calibrated out as well as the external sensor errors during the calibration process.  You can test this system using the PGA309EVM.  The EVM includes built in calibration software so that you can test the calibration process.

    Best regards, 

    Art Kay

  • 0 in reply to Art Kay
    Prodigy 20 points

    Hi Art,

    Thank you for your careful answer. I will purchase a PGA309EVM for verification. A special place in my application scenario is that the operating temperature of PGA309 and the operating temperature of the external sensor have a large temperature difference, and they do not work in the same ambient temperature. Therefore, I think in addition to calibrating the temperature coefficient of PGA309 external sensor, I may also need additional MCU to calibrate the temperature coefficient of PGA309. Is this correct?

    Best regards, 

    Lio

  • 0 in reply to lio
    TI__Genius 17757 points

    Lio,

    You are correct that when the PGA309 is kept in a separate environment from the sensor, that the PGA309 drift will not be calibrated out.  In most customer applications the PGA is kept in the same environment as the sensor so the PGA drift is calibrated out, but your case is different.  The PGA309 does have an on board temperature sensor that could be used to facilitate a calibration of gain and offset over temperature.  To do the calibration you would have to apply different temperatures to the PGA309 and at each temperature apply two different inputs for a gain and offset calibration.  You may consider evaluating the system without the PGA calibration to see if the error is significant enough to need calibration.

    Art