top of page
Search

# How to Calibrate - Teach an AI - Build the Inverse Logarithmic Calibration Curve

Updated: Apr 19, 2022

Purpose:

To educate the Court about the Analytical Theory of the 8000C Optical Bench including Emitter, Sample Chamber, Dual Detectors, Inverse Logarithmic relationship, and building the Calibration Curve based on the theory that differing ethanol concentrations have differing % absorption of IR light.

To lay the groundwork for the importance of calibration and re-calibration respecting the reliability of any approved instrument.

To explain the auto-calibration function of the instrument that is run at the factory or the Authorized service Centre.

To lay the groundwork for why changes to the relationship between electrical signals coming off the detector and true indication result in drift in accuracy and precision, i.e. loss of reliability.

Sample cross-examination of CFS expert follows:

Q. And the procedure would involve using that variety of solutions and the reason why they need to use that variety of solutions is because the instrument has to learn what each of those values means. A. Correct. Q. And that creates a calibration curve. A. Well, it creates the data that the instrument uses to generate a calibration curve. Q. I see. Right. Now, in that process of the

data being generated to create the calibration curve.... MR. BISS: And this Your Honour, incidentally, is an excerpt essentially from the same – the same manual, just different pages of the – of the C.M.I.

[From CMI 8000 Calibration Manual]

Q. The indication at page 2 of that document is that different ethanol concentrations are used and each of these has a different absorption percent, from the perspective of the instrument. A. Correct. It reduces the intensity of light that reaches the detector. Q. All right, so just – let’s talk a little bit about how an Intoxilyzer 8000 or Intoxilyzer 8000C works. At one end of the sample chamber there is a light source. In the 5000C it was a light bulb. In the 8000C it is an emitter that emits a pulse. A. Yes.

[See Ontario 8000C Training Aid 2013 pages 55-57 of 238]

Q. Of infrared light. A. Yes. Q. Inside the sample chamber there is either a breath sample or just air or perhaps the sample that’s come from the wet bath simulator for the purposes of a calibration check. A. Yes.

Q. At the other end of the sample chamber there is a detector and the detector reads an electrical signal of final intensity of infrared light coming out of the sample chamber. A. Yes. It measures the intensity – the decrease in the intensity of the signal and converts that to a voltage. Q. So if we know the initial intensity of the

light that is coming from the emitter and we know the decrease in that intensity of the light

at the other end of the sample chamber, then we can – an electrical signal is generated and

that electrical signal can be translated into a calculation of the – well, first of all, we will

know what the ultimate transmittance is of the light.

A. Yes.

Q. In the sense the transmittance as in the light that reaches the final end of the sample chamber, as opposed to the – some of the light is going to be absorbed in the sample chamber by the molecules that are in the sample chamber.

A. The molecules of alcohol, yes. So, if there’s no alcohol present the intensity of light that leaves the light source and that reaches the detector is the same.

Q. Yes.

A. So the relationship is inverse to what we usually expect to see an increase in signal. You actually see a decrease in signal. So the more alcohol that’s present in the sample chamber, the decrease in intensity of light that reaches the detector is the response.

Q. All right.

A. So the more alcohol the more the signal is decreased.

Q. Okay, and the more the alcohol....

A. The higher the concentration of alcohol that’s present, the higher the decrease in the intensity of the light that reaches the detector.

Q. But the signal that comes off of the detector actually has a logarithmic relationship to the concentration of the alcohol in the sample chamber.

A. Yes.

[Note: 1. Inverse relationship 2. Logarithmic]

Q. All right. That’s difficult to get our heads around unless we think in terms of the slide rule. The slide rule is something that some of us used a long time ago in high school. A slide rule uses logarithms. A. Yes. Q. And – anyway. So, the signal passes through the sample chamber and when the signal comes through the sample chamber there’s an electrical signal that’s produced and so during calibration the person who is doing the calibration is entering commands into the Intoxilyzer 8000 or 8000C which then, the software, during an auto calibration sequence in the instrument then generates a calibration curve. A. Essentially, yes. The specific procedure, again, I don’t know what is required, but that essentially is it, yes.