Question about dealing with pressure transducer offset voltage

Discussion in 'The Projects Forum' started by mxw8950, Sep 28, 2011.

  1. mxw8950

    Thread Starter New Member

    Aug 19, 2010
    7
    0
    Hello,

    I am using a 1 INCH-G-BASIC pressure transducer in order to measure very small pressure changes. Looking at the datasheet the sensor can have an offset voltage @ zero differential pressure of +/- 10mV. The total output span of the sensor is nomially 7.0mV.

    The outputs of the sensor are input to an AD623 instrument amp. I would like to be able to calibrate the sensor so that at zero differential pressure reads 0mV. I have connected potentiometers between the output pins of the sensor and Vcc in order to adjust the input to the AD623 to read 0V difference between the two pins.

    I have attached a portion of the schematic which illustrates what I have tested.

    This seems to solve my problem but I was just wondering if this was the best way to approach the problem.

    Thanks,
    Matt
     
  2. #12

    Expert

    Nov 30, 2010
    16,257
    6,757
    OK. That will work. After all, you said it worked, but...

    You can get better sensitivity if you diminish the effect of the pot with a fairly large series resistor. If you worked this with both polarities of power supply, you could get + and - correction from a single pot connected to only one input. If you keep it mono-polar, you need the 2 pots that you illustrated, or add a switch to select which input needs compensation to zero it. Depends on how hard you want to work and whether this is one of a kind or a production design.
     
    mxw8950 likes this.
  3. mxw8950

    Thread Starter New Member

    Aug 19, 2010
    7
    0
    Thanks so much for your response.

    I am definitely looking at sticking with the single sided supply for simplicity. I had tested the current setup with 100k pots but I definitely like the idea of throwing a series resistor in there to diminish the effect from the pot. Since it is a limited run production item I'd think I'd rather deal with two pots rather than adding a switch component in there.

    Thanks again for your help,
    Matt
     
  4. #12

    Expert

    Nov 30, 2010
    16,257
    6,757
    One more idea...it would be cheaper to use a fixed resistor on one input to force it as high as the sum of all possible errors, which is 10mv plus whatever the AD623 is guaranteed for, then set the sensitivity of a single pot on the other side to compensate from zero to 2x the max offset.
     
Loading...