The ambient air around us contains 2% CO - this is thus the [correct] calibration setting for the Gunson reader.
This isn't quite right - I hope that the clean air we are breathing does not contain 2% CO.
The Gunson meter works by measuring changes in thermal conductivity, as the thermal conductivity of exhaust gas will change with CO levels. The 'clean air' setting process is basically you calibrating the instrument, i.e. telling the instrument that when it measures thermal conductivity of 'x' then that is equal to 2% CO in the exhaust gas. It then uses that as a reference when it is measuring your exhaust gas.
The Gunson documentation states that the thermal conductivity measured by the instrument in clean air is equivalent to what exhaust gas with 2% CO would be. What you are doing here is using clean air as a proxy for 2% CO exhaust - if Gunson are to be believed then the thermal conductivity measured is the same.
Instruments like this are prone to error (Gunson quotes an accuracy of +/- 0.5%) and other factors influence reading too, so one shouldn't take the absolute readout obtained as being spot on. They are useful, however, for watching the change as you alter fuel/air mix etc, so they do have their uses.
David
