Mg811

lucacox

Hi all,

I'm using the mg811 co2 sensor to monitor office air, but I have some problems computing the ppm from the output voltage:
the data sheet says:
EMF = Ec - RT/(2F)*Ln(P(CO2))

where:
R = 8.314472 J/(mol K)
F = 96485 J
T = (i believe that is room temperature) 293.15°K

I've measured EMF at 500ppm and 20°C and it results 1.04V thus Ec = 1,12 V. So I've use the formula:

P(CO2) = e ^ ((Ec - EMF)*2F/(RT))

But this equation gives me results in the order of 10^(-6) that can't be the ppm of CO2 in an office!

What's wrong with my calculations?

Is P(CO2) expressed in ppm?

Thank you all.

Tracy Allen

Hanwei does provide a "typical" graph of mV output vs ppm, and at some point I had calculated the slope/intercept for that and came up with...

The formula is similar to yours, also taken from the data sheet, and if you convert volts to millivolts the slope is in the same ballpark (-0.0774 vs -0.0543). Where it differs is the ln(PPMo) term. It does take a two point calibration, I think.

There was an extended discussion of the MG811 in the Stamps in Class forum last year. One thing that came out is that there was considerable variability from one MG811 to the next. I don't recall if more of the variability was in the slope or the offset. There was a lot of confusion about it.

I'm puzzled by the units too. Partial pressure should be in units of say pascals, not ppm, unless you assume reasonably atmospheric pressure at sea level.

lucacox

Hi Tracy,

thank for your reply. Using your formula I get some better result... level of CO2 is in the order of 500ppm that is more suitable, even if it seems a bit too low for me...

If I came to new conclusions I will post here.

Tracy Allen

Oh, the value of ln(PPMo) and mVo were also determined by the example data, so the formula was,

ln(PPM) = -0.0543 * (mV - 336) + 5.29831737

The data sheet is not clear on what temperature to use in their Nerst equation, but it would probably be the temperature inside where the reaction occurring, and that is driven high by the heater, but how high, I don't know. The best recourse is individual calibration for slope and intercept of log(concentration) against mV potential. The normal sea level CO2 concentration is about 390ppm, 0.039 percent, so that is one readily available calibration point.