Getting an accurate yellow “Check Engine” or “Service Engine Soon” light diagnosis is often frustrating and expensive. Fortunately, if you happen to be technically inclined and have the basic tools, you can likely do the job yourself and avoid that expensive repair shop visit.
Designed to alert drivers to computer-monitored emissions problems, the “Check Engine” light is actually part of the vehicle’s emissions system. “Check Engine” lights became a standard equipment feature when automotive onboard computers proliferated in 1981. Starting with model-year 1996, the OB-II protocol became the standard for automakers. Since that time, every new car sold in the United States has required a “Check Engine” light, mainly to be compliant with Environmental Protection Agency regulations. But after all these years, “Check Engine” lights remain a mystery to many technicians and automotive do-it-yourselfers.
How the “Check Engine” light works
Here’s an ultra-simplified version of how the “Check Engine” light works. Vehicle computers use input signals from sensors to generate control signals for fuel and spark delivery, for transmission shifting and for other functions. The car’s computer continuously monitors all input signals that could affect emissions. If any of the monitored signals move outside government-mandated limits, the computer turns on the “Check Engine” light.
The computer also determines if the problem meets the criteria for setting a code. However, rather than identifying a part or system that has failed, these codes refer to the part or system that is being affected by what has failed, making them more confusing than helpful.
How oxygen sensors work
Because this code system can confound even experienced technicians, it often results in unnecessary repairs. Oxygen sensors, for example, are extremely reliable, yet millions are needlessly replaced every year, largely because of the wide array of problems that cause the computer to set an oxygen sensor code.
“Check Engine” light oxygen sensors examine the amount of oxygen in the exhaust gas leaving the engine. The sensor compares the oxygen inside the exhaust system to oxygen in the air outside the sensor. Rich exhaust has less oxygen; lean has more. The amount of oxygen in the exhaust is directly related to the fuel/air mixture entering the engine. A rich incoming fuel/air mix produces exhaust with less oxygen, while lean produces exhaust with more oxygen.
To understand how the “Check Engine” light oxygen sensor knows the difference between rich and lean, think of it like you would a battery. Oxygen sensors contain a crystal that is sensitive to oxygen. These crystals produce a small voltage when there is more oxygen on one side of the crystal than the other. When the mixture is lean, there is a small difference in oxygen content between the exhaust and the atmosphere, so the crystal produces a low voltage. The exhaust from a rich mixture has less oxygen than is in the air, so the crystal produces a higher voltage. Normal voltage ranges from 100 millivolts (lean) to 900 millivolts (rich).
Voltage produced by the oxygen sensor goes directly to the vehicle’s computer. When the mixture is rich and the sensor voltage is high, the computer reacts by commanding a lean mixture. The lean mixture produces more oxygen in the exhaust and the voltage drops, which causes the computer to command a rich fuel mixture. As a result, the fuel/air mixture constantly fluctuates between rich and lean. The rich-lean cycle repeats many times per second, producing an average fuel mixture halfway between the two extremes.
Oxygen sensor codes
What does this have to do with oxygen sensor codes? The oxygen-rich excess air can’t burn completely and leaves the engine with the exhaust. Excess oxygen in the exhaust causes the oxygen sensor to always produce the same voltage and eventually quit switching from rich to lean. Without fluctuating voltage from the oxygen sensor, the fuel mixture locks up. The computer monitors how much and how often the voltage changes, and if it doesn’t meet specifications, the computer turns on the light and sets a code. Unfortunately, the computer can’t determine why the sensor isn’t switching from rich to lean, so it is limited to setting a check engine oxygen sensor code.
This approach routinely creates expensive problems, because most people facing an oxygen sensor code install a new sensor, clear the code and assume the car is fixed. But the “fix” is usually short-lived, because nothing has really been fixed. The underlying problem still exists.
Sooner or later, the “Check Engine” light is activated again. Why? Clearing codes also erases the computer’s monitors. These subprograms in the computer constantly look at individual circuits, and unless all monitors are “Set and Ready,” the “Check Engine” light cannot come on. The monitors will not reset until a specific series of driving conditions has been completed.
Because some of the driving conditions necessary to reset monitors are not part of normal daily driving, it can take days, weeks or sometimes even months for a complete reset. As a result, “Check Engine” light repairs often work fine for a while, but eventually the light and the same code return. When the code resets, it’s usually presumed the new oxygen sensor is defective and it is replaced under warranty. But without proper testing and repair, the process repeats itself.
How to avoid “Check Engine” light problems
This may seem a lot to explain the nuts and bolts of dealing with codes and “Check Engine” lights, but unless you know how and why problems occur, you can’t know how to avoid them. Here’s the proper procedure.
The battery supplies electrical power to run the vehicle, but it also controls and stabilizes system voltage. Whenever you’re diagnosing a “Check Engine” light, it is important to start by having the battery tested. An automotive computer is like any other computer; proper input voltage is critical. A bad battery can cause problems in electrical and electronic parts, including the computer. It can even cause a “Check Engine” light.
Next, test the alternator and pay close attention to the condition of its diodes. An alternator produces alternating current (AC), but cars work on direct current (DC), making a conversion necessary. Converting AC to DC is done with a series of diodes inside the alternator. However, diodes sometimes fail to do a complete conversion, allowing AC current to pass into the car’s electrical system. The result is a confused computer. Garbage in, garbage out, as computer programmers are fond of saying.
Because some vehicle oxygen sensors normally produce AC signals for the computer, a false AC signal from a bad diode (“garbage in”) can make your computer produce faulty output signals (“garbage out”). The computer can’t distinguish between an AC signal from a sensor and a false one from the alternator. This can lead to rough shifting, performance problems, codes, and, yes, a “Check Engine” light. In this case, the “Check Engine” light and codes would probably be false. Therefore, if you don’t follow procedure you will almost certainly replace parts that are still good.
After verifying the condition of the battery and alternator, perform a thorough visual inspection of all hoses and wires. You’re looking for frayed or oil-soaked wires and vacuum hoses, and anything else that looks suspicious.
The next step requires critical test equipment you probably don’t have. Have the engine connected to an engine analyzer, which must not be confused with a computer scanner. An engine analyzer uses two primary testers to determine that all baseline parts are functioning properly. Baseline parts control fuel delivery and produce high voltage to fire the spark plugs. The two pieces of equipment are an oscilloscope to look at electricity as it moves through parts, and an exhaust gas analyzer to measure tailpipe emissions for diagnostic purposes. These are extremely important tests because there are many failures in baseline systems that cause “Check Engine” lights and codes.
Finally, it’s time to retrieve codes, which requires an inexpensive code reader from an auto parts store or online source. To fix code-related problems, you’ll need factory-test information from a source like Alldata or a factory service manual. You’ll also need a computer-safe test light and a Digital Volt Ohm Meter (DVOM).
Deciphering the codes
Just remember, codes do not directly tell what’s wrong, so you have to test to find the real culprit. This is where the true meaning of codes comes into play. The only thing a code really tells you is what test to perform to isolate the true problem.
Each numerical code has a matching-numbered test procedure. Following that test will direct you to the problem, providing you follow the rules. Each step of each test must be performed in absolute sequence. Skipping steps or performing steps out of sequence could make the entire test worthless. Tests may also give you voltage, resistance, temperature or time specifications, which are exact values. Close doesn’t count.
By following proper test procedures, you’ll get rid of the pesky “Check Engine” light without breaking the bank. If you can use simple test equipment and read and follow directions, you can do “Check Engine” light repair just like the pros.