How to calibrate gas sensor?

Discussion in 'General Electronics Chat' started by presish, Nov 18, 2014.

  1. presish

    Thread Starter New Member

    Nov 18, 2014
    7
    0
    Does anyone know how to convert ADC output values to ppm readings for a gas sensor without exposing the sensor to a known concentration?

    such as, if ADC says 54 through a common conversion method we know it is 20ppm... is there such a method that can directly convert the ADC output to ppm?
     
  2. mcgyvr

    AAC Fanatic!

    Oct 15, 2009
    4,771
    971
    Nope.. You always need to calibrate from a known concentration.
    And if you are making more than one device they each need to be calibrated independently as variations will be present.
     
    presish likes this.
  3. MrChips

    Moderator

    Oct 2, 2009
    12,449
    3,365
    @persish

    1) please do not hijack someone else's thread
    2) please do not post your new question to an old thread
    3) please do not make multiple posts

    Now you have your own thread.
     
  4. presish

    Thread Starter New Member

    Nov 18, 2014
    7
    0
    thanks for the help! Ya sorry about that. I am new to the rules to forums
     
  5. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,699
    907
    What sensor are you using?

    John
     
  6. presish

    Thread Starter New Member

    Nov 18, 2014
    7
    0
    MQ7 and MQ2
     
  7. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,699
    907
  8. presish

    Thread Starter New Member

    Nov 18, 2014
    7
    0
    the graph is not detailed very well. do you know how to convert ADC readings to ppm? that would help a lot!!
     
  9. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,699
    907
    What other details do you want in the graph? It is a ratiometric sensor in which the resistance of the sensor varies with the concentration of the gas being measured. Its output cannot be fed directly to an ADC. From the datasheet:
    upload_2014-11-19_5-11-21.png

    Please post the circuit you intend to use. Your design of that circuit will determine the voltage range you will see and the "ADC counts." Be sure to read the last section (Sensitivity Adjustment) of the datasheet.

    John
     
  10. presish

    Thread Starter New Member

    Nov 18, 2014
    7
    0
    The circuit I am using is the same circuit listed in the datasheet. I am supplying 5VDC.... I read the entire datasheet multiple times but I do not have a fixed concentration of gas to calibrate.. thus my question on any alternative methods to calibrate these sensors.
     
  11. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,699
    907
    That is not the same question you asked in post#8. I will assume you know how to read the resistance of the sensor using the ADC and your test circuit.

    Is it that you don't have hydrogen to make standard concentrations, or you do have hydrogen and don't know how or don't have the equipment to make the mixtures? What level of accuracy do you need?

    John
     
  12. presish

    Thread Starter New Member

    Nov 18, 2014
    7
    0
    I do not have the means to make known concentrations that is why I wanted to find a general conversion method for my digital output
     
  13. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,699
    907
    If you can't calibrate it with known concentrations of hydrogen, then all you can do is get in the ball park, as shown in the datasheet.Whether that is 5% off or 30% or more off, I don't know.

    John
     
  14. presish

    Thread Starter New Member

    Nov 18, 2014
    7
    0
    I guess then the remaining question on calibration via data and ballpark measurements would be... how can I convert my data either using Linear Regression or some other method to best fit the graph in the datasheet? I am using a Sharp Dust sensor and I found an equation that places my ADC readings directly on the datasheet graph.. is there a method to do this with the MQ sensor?
     
  15. jpanhalt

    AAC Fanatic!

    Jan 18, 2008
    5,699
    907
    That is a log-log graph and looks pretty linear. (We used to joke that any data looks linear on a log-log graph.) I haven't looked at the dust sensor graph.

    John
     
    presish likes this.
  16. KL7AJ

    AAC Fanatic!

    Nov 4, 2008
    2,040
    287
    This is correct. We had to do this calibration with mass flow controllers and reference gases. Kind of a fun process, but you can't be sloppy!
     
    presish likes this.
  17. MrChips

    Moderator

    Oct 2, 2009
    12,449
    3,365
    I have designed many industrial gas sensors. You cannot calibrate your instrument properly if you do not have a calibrated standard gas or an instrument that is already calibrated. Yes, we also used volumetric gas mixtures to characterize the sensors from 0% to 100%.

    Note also that your sensor is also affected by environmental conditions such as temperature, pressure, humidity and presence of all the other gases.
    On top of that, most sensors tend to be non-linear.
     
    presish likes this.
Loading...