Tip 50: When an "approved instrument" is calibrated at the factory, the manufacturer doesn't simply adjust the electronics in the instrument to one concentration of ethyl alcohol. The electronics in a new instrument must be tweaked (using a screwdriver pre-2000 and software post-2000) so that electrical signals coming from the detector and other circuitry are accurately and precisely read by the instrument's software to output values corresponding to ethyl alcohol values of many different concentrations (e.g. 50, 100, 150, 200, 250, 300 mg/100 mls) as well as interferent values (at 10 mg/100mls threshold), ambient ethyl values (at 10 mg/100mls threshold), internal standards, and RFI. Calibration, that is adjustment, verification, adjustment, verification, repeated again and again, has multiple targets for different parts of the alcohol analysis range (normally encountered in evidentiary breath testing) and for the instrument's automatic error detection systems. Calibration of the ethyl alcohol quantification system is not linear, it is a curve. Calibration of the interferent detection and other systems at the factory is necessary to determine that the particular instrument is performing in accordance with manufacturer's specifications.
Out in the field, any of these positions on the ethyl alcohol calibration curve or other types of calibration may drift. It is therefore imperative that the approved instrument be properly inspected and maintained at the factory or the Canadian Authorized Service Centre on a regular basis to ensure that the instrument continue to meet the manufacturer's specifications.
The Leaning Tower of Pisa depicted in any 2 dimensional photo has an obvious problem with its foundation. A flat 2 dimensional image taken by a tourist, however, fails to do justice to the wondrous experience of actually walking around the tower in 3 dimensions. The image above was an attempt to create an unusual picture using the Panoramic feature on a mobile phone.
What if we were to think of calibration of an approved instrument as a multi-dimensional curve? There are many variables in evidentiary breath testing that can create error in the analytical system (response to ethyl alcohol at various concentrations, interferents, ambient conditions, and RFI) and in the reliability system (SOPs, #duisimulator, alcohol standard, simulator thermometer, ambient conditions, RFI, jar, and tubing).
A single-point calibration check may be very helpful on a day-to-day basis to quickly field-check the ethyl alcohol calibration of the approved instrument at values close to 100 mg/100mls but it is not a replacement for regular inspection and periodic re-calibration at target values throughout the likely ethyl alcohol range and as required by error systems. Individuals responsible for these re-calibrations must be able to calculate the uncertainty associated with such re-calibrations. Calibration laboratories that perform such re-calibrations must do so in accordance with ISO 17025 and Canada's international treaty obligations.
An example of a hypothetical methodology for re-calibration of an Intoxilyzer® 8000C, ethyl alcohol concentration system only, follows. Notice that the instrument "learns" the meaning of various target values and then automatically calculates the calibration curve for ethyl alcohol across the various targets:
1. Prepare simulators with wet-bath solutions: Distilled Water, 40 mg/100 mls, 50 mg/100mls, 50 mg/100mls, 100 mg/100 mls, 300 mg/100mls.
2. Commence optical bench re-calibration.
3. Check simulator temperature during each standard's use.
4. Instrument takes 4 samples (discards 1) of water solution, analyzes % Abs and % Abs Ref for each of 3 micron and 9 micron filters. Calculates Avg % Abs, STD DEV, and REL STD DEV.
5. Repeat for 40 mg/100mls.
6. Repeat for 50 mg/100mls.
7. Repeat for 100 mg/100 mls.
8. Repeat for 300 mg/100 mls.
9. Instrument then calculates adjustment factors: Act, Fit, and Residual for target values 000, 040, 050, 100, and 300.
10. Conduct a separate re-calibration of the ITP system using the 100 target standard.
Calibration of the ambient threshold, the interferent threshold, and the RFI threshold (at various frequencies) will require other standards and a different methodology.
Software calibration of today's approved instruments, across the multi-dimensional curve, is every bit as important as screwdriver calibration "tweaking" of older approved instruments such as the 5000C.
Comments