miketsp:I understood that O2 sensors are sensitive to PP% but the construction of the sensor head is such that the reading is taken at very close to ambient pressure. The gas under test just flows across the sensor. So calibrating at ambient pressure is correct and sufficient. Altering the flow rate should have no significant effect.
The idea of having a cylinder of air to calibrate the sensor just adds another unknown variable.
While I don't disagree about the construction of the sensor, the flow rate does appear to make a difference.
When we teach nitrox, we show the students the effects of changing the flow. We can induce at least a +/- two percent difference by changing the flow.