Sorry Libelium-Dev. But bear with me for a moment because I am going to go to calibrate the sensor soon and I would like to know if I am doing the right steps: -

I am trying to calibrate the TGS2611 Methane sensor as follows: -
[*] Find relationship from the Rs/Ro graph
[*] Estimate Rs when the concentration is equal to 5000 ppm
[*] Use this estimation as RL (right or wrong?)



The data-sheet from the manufacturer states the following: -

R(LOAD) - MINIMUM = 0.45 kOHMS
R(LOAD) - MAXIMUM = 100 kOHMS <<< specified by Libelium

Rs = 0.68k - 6.8k ohms (under stand. test conditions)

MY QUESTION IS:

Since I figured out a relationship for the graph. How am I suppose to estimate the value of Rs? I see many problems especially since: -
[*] Ro IS THE ACTUAL RESISTANCE of 5000 ppm concentration
[*] How is it possible to estimate Rs without getting V(LOAD) and R(LOAD) in the first place?
[*] Are you suppose to use a recommended R(LOAD) first ?

Please explain thoroughly how you would estimate Rs... I did the pleasure to post the relationship for you: -


-Thank you.

Hi AmroQuandour,

your math is correct, but not necessary for calibration.

For your case, calibration at 5000ppm, we have Rs/R0 = 1, with R0 between 0.68k and 6.8k. For the load resistor you should select a resistance in this range, for example 3k, and carry out the calibration.

When you get the output voltage, you can extract the real R0 value. If you want more accuracy, you can repeat the calibration process replacing the estimated value with the new calculated resistance.

If you repeat this process for different concentrations (for example, 1000ppm, 3000ppm, 7000ppm and 9000ppm) you will get a more accurate estimation of the sensitivity (the parameter you call m) and will have a better representation of the relationship between gas concentration and the resistance of the sensor.

Once you have R0 ant sensitivity, you can use them in your math to implement the direct conversion of the output voltage values you get into ppms.

Regards.

Beautiful. Thanks man :) This was much clearer !

Hi Libelium-dev,

1) Is the sensitivity of TGS 2442 will vary from 0.13 to 0.31 ?

2) from the sensitivity graph of the same sensor we can derive the equation
log(y) = m * log(x) + b
can we directly put the sensitivity value e.g 0.13 into m ?

3) from the sensitivity graph of TGS4161 the relation between x and y will be
y = m * log(x) +b
here also can we put the sensitivity value (varying from 44 to 72 mV) into m?

If not then shall we calculate the gradient from their respective graph ?

Regards

Hi abhi123,

the values of the sensitivities of both sensors you point are correct, they should move in those ranges (in fact, if you calculate the gradient from the graph, you will get a value inside them).

You may take a value in the middle of those ranges if you can't perform a calibration or don't need great precision (be careful about how those values are defined).

Regards.

but from the sensitivity graph of TGS2442 if we choose any two points e.g, (100, 1) and (30, 4) and then calculate the gradient in logarithm scale like below
we get

m = (log y2 - log y1) / (log x2 - log x1)

= - 1.1514

but that's not matching with the given range of sensitivity.
really confused, pls help me out...

Hi abhi123,

that is what I meant when I said you had to be careful with the way sensitivity had been defined.

The sensitivity value hasn't been given in a logarithmic scale, but the direct coefficient between the values at 300ppm and 100ppm, so you won't be able to use it directly as the sensitivity in that formula as you are doing. Given that, the sensitivity m would be:

m = (Log(0.13~0.31)/(Log(300)-Log(100));

With this, you get a result of logarithmic sensitivity between -1.85 and -1.06, so the one you got from the graph is within this range.

Regards.

thanks man.... much clearer now.. :-)

Hello,

The following document (from the Figaro) might be quite useful to determine the RL (Load Resistor) values for various applications. In fact, the manufacturer provides data on what values of load resistance should be applied for a particular alarm point, so that the resolution could be sufficient around the alarm point.

Here is the Figaro sensor (TGS2611) provides a list/table of load resistance values corresponding to serial serial Id.

www.figarosensor.com/products/2611app.pdf

Regards

Hi Amin,

thanks again, this is a very useful tip.

It must be taken into account that this option is only available, as far as I know, for the TGS2611 and TGS2610 sensors, so to calibrate other gas sensor the general method will have to be applied.

Regards.