The Michigan BAC Datamaster suffers from several problems. The DMT, which is Michigan's newer breath testing instrument, has increased processor speed, greater memory, and an improved method of measuring breath flow, but it is essentially the same device. Years ago, Michigan ordered that the BAC Datamaster's data collecting abilities be turned off, which is the feature that set the Datamaster apart from all other breath testing instruments of its day. For all of the DMT Datamaster's new functionalities, which include a graphing interface for viewing the breath flow, Michigan has followed its past practice and ordered that all of the DMT's improved functions be disabled. This is simply astonishing. Michigan employs breath testing devices that have been intentionally crippled and find no use anywhere else in the United States.
Although the Datamaster is supposedly accurate to 0.004, that accuracy is premised on laboratory conditions using controlled samples. Introducing a police officer as the breath test operator and a live person as the test subject, test results can reveal wildly deviating results. The Michigan Breath Test Operator Training Manual provides an example of three test results produced registering a 0.17, followed by 0.11, followed again by 0.17. That's a range of inaccuracy of 0.06, nearly the legal limit itself.
Police officers are required to begin a 15 minute observation of a breath test subject before administering the breath test. They are specifically instructed to look inside the subject's mouth and record the start of the observation period off of the Datamaster's clock. This appears in big, bold, underlined print. Frequently, officers fail to follow this extremely simple rule because they cannot be bothered to take the few extra minutes to properly administer a breath test. The risk of a wrongful conviction is too great, and officers do a grave injustice when they fail to follow these rudimentary rules.
The BAC Datamaster can also display several error messages, and the police are specifically instructed not to take the unit out of service unless the company instructs them to remove it from service. This includes "testing the machine into compliance," where simulator tests are run repeatedly until the device passes a simulator test. It also includes temperature error, which can dramatically alter a breath test score, calibration errors that obviously undermine the integrity of the breath test results, and even an error message "FATAL SYSTEM ERROR." (If a fatal system error occurs, the police have no choice but to remove the device from service because it simply stops working.) Without knowing whether error messages were produced on the date of a breath test, judges, prosecutors and most defense attorneys accept the Datamaster results as reliable.
Frequently, we see a common error messages that states "INVALID SAMPLE." If an invalid sample is detected, this means that a negative going value was detected when a positive value was expected. In simple terms, this means that an alcohol spike has been detected. A valid breath test will always rapidly increase and then plateau, with slightly higher values always occurring once the plateau has been reached. (Incidentally, this also means that a longer blow will always produce a higher result. Hence, a very long blow producing a .08 would have likely produced a .07 with a shorter blow.)
Using a sip of beer in our office to demonstrate, I have taken a sip of beer and spit it out. The trace amounts of alcohol in that sip are enough to produce results in the range of .23 to .35. Liquor will easily produce an overflow value exceeding .70. On a long blow, the Datamaster will return the result of "INVALID SAMPLE." But here is the really scary thing: on a short blow, this will produce a valid test result because the machine has not had enough time to measure the negative going value, even though I am completely sober!
An invalid sample requires an officer to begin a new 15 minute observation period, but if the officer violates this rule, Michigan courts have held that this result is still admissible. This decision was not based upon the science of breath testing, which would require suppression. Instead, this decision was made based upon the language of the Michigan Breath Test Operator Training Manual which happened to use the phrase, "should begin a new 15 minute observation period," as opposed to "shall begin a new 15 minute observation period." Incredibly, as a result of this typo which was made by a single police sergeant at the Michigan State Police years ago, the jury must decide what weight to give to the test result.