Joe replaces the spark plugs in his vehicle. Next day, the check engine light comes on. The plugs look fine, so he replaces the plug-wires. A brightly glowing indicator says the problem is still present. Cap and rotor replacement shows the same result. Joe is slowly learning a lesson about cause and effect. Proper diagnosis is simply far less expensive than other methods.
The human mind always looks for an order in things. Usually this serves us well and helps determine causes for the problems we have. Sometimes the same tendency can lead us astray. Joe had an evaporative emissions code causing the light, unrelated to the spark plug replacement. This is fallacy one in diagnosis.
1. Sequential progression does NOT always denote causality
Brushing things off as simple coincidence, is never wise, without considering the cause. Assuming the last thing that occurred is the cause, is equally troublesome.
A more logical approach would begin with checking the code, to see what area of the vehicle is causing the light. Depending on the system involved, testing would go on from there. For instance, a misfire code suggests testing the work with the spark plugs. A catalytic converter code or and evaporative emission's code would start with those systems. This leads to fallacy two in diagnosis.
2. "Put it on the machine"
No machine diagnoses automotive problems. Trained technicians use several pieces of equipment, depending on the type vehicle and the problem. Experience guides them to the most likely tests. A fuel pressure test might be the first choice, on a General Motors vehicle, that does not start. This is based on the high failure rate on GM fuel pumps. A Honda with the same symptom may get a different initial test. Honda has very little fuel-pump related problems.
Diagnosis may begin with scanning the computer to see if any codes are present. Diagnostic trouble codes or a DTC, can help identify the system in which the problem exists. This also brings up fallacy number three.
3. A trouble code does NOT tell what is wrong
Diagnostic trouble codes are information, displayed in response to things that read out of range. For example, the computer may expect a voltage between one and five volts on the oxygen sensor lead. The voltage should also vary, up and down, as the sensor reads. If it reads zero or six volts, the light comes on. A reading, which does not change, will also generate a code. The message is an oxygen sensor out of range. This does NOT mean the sensor is bad.
An injector stuck open, will flood the cylinder with fuel. This will make the oxygen sensor read a constant voltage. A bad mass air flow sensor can also cause the oxygen sensor to read out of range as can several other things. Replacing the sensor will correct none of these problems.
Another example could be a crankshaft-sensor code, on a vehicle that dies. Running out of fuel will cause the engine to die. The engine quits rotating so the crank signal stops. No signal from the sensor sets the code. Replacing the sensor is futile or worse. This leads to fallacy number four.
4. Replacing parts will NOT solve the problem
Unfortunately, many replacement parts today are not as good as the old components they replace. People create problems, removing a part, and replacing it with a defective component. Often this leads to replacement of several other good parts, assuming the first part was not the problem. Without a means of verifying the replacement part, this becomes an expensive circle and brings us to the final fallacy.
5. New parts are NOT known-good
Many new replacement parts are bad and off specification. The owner sends a vehicle to the repair shop, because it will not start. Several new parts are apparent. No ignition or pulse for the injectors can be found. The power control module or PCM, is bad. Fortunately, the shop tests the inputs to the PCM. Battery voltage is present on the signal wire from the fuel-air sensor.
Someone has installed an aftermarket oxygen sensor, in place of the original part, splicing the wires. Replacing the PCM, repairing the harness and installing a proper fuel-air sensor is necessary, before the initial problem can be found. They can now diagnose the original problem, which is a bad injector relay. The owner has wasted over two-thousand dollars, creating problems and trying to solve a two-hundred dollar issue.
Diagnosis is the most difficult facet of auto repair, by a wide margin. Great shops have a culture of diagnosis and quickly and accurately pin point problems. Sound methods save a huge amount of money and frustration. A proper diagnosis is the least expensive way to repair a vehicle.