Linear?

IT’S THE GAS!I had a chemist the other day show me a response curve for a liquid analyzer that was challenged with three separate calibration standards. First off, he said that the analyzer had a linear response and did not use an algorithm to correct for a curve response. That turned out to be wrong but we’ll get to that. When I looked at the data it showed that two concentrations read high and one read low, so I told him it was not the analyzer but the standards were wrong. He didn’t really like hearing that since it was his lab that made the standards. So I plotted out the data and the curve showed a nice little dimple at the low reading point and I tried to explain that if that standard returned the same reading as another concentration then the standard was wrong. Never got my point across but we did find a curve correction polynomial in the software so at least we cleared that up.

Here’s the deal, if one of your standards reads high or low then they should all read high or low. There are exceptions to this but usually it occurs in analyzers that output is based on a secondary measurement. By this I mean analyzers who measure reaction and not actual transmittance or absorption. An example of reactive measurement would be something that most of you are used to like a Chemi analyzer. Chemi analyzers measure the reaction of Nitric Oxide and Ozone, when the two mix the reaction gives off light. That light is directly proportionate to the concentration, so really you are measuring the reaction and not the gas. Overall if your gas response to standards is questionable try and not make a snap judgement that it’s the analyzer because more times than not, you will be wrong.