The purpose of the first fuel system on an automobile was simple: just dump enough fuel into the engine to make it run smooth under most conditions. Fifty years of driving gas-guzzling, high-pollution land yachts had left us with a legacy of ever-increasing fuel costs and crud in our air that made some type of exhaust emission controls eminent. A system was needed on board to monitor and control the fuel mixture, so the oxygen sensor was born.

Feedback fuel systems have been with us for a long time. I remember my first experience with a feedback fuel system in the 1980s; thinking how high-tech it was to have a computer running the carb through the mixture control (MC) solenoid. Feedback carburetor systems operate using the same basic principles on most cars: utilizing an O2 sensor to measure exhaust gas, then some type of solenoid to adjust the mixture by controlling either the fuel or the air to the main metering system in the carb. What once seemed like a complicated system of computerized fuel control back then seems relatively simple now compared with the technology we are using today in OBDII. We still see some of these cars occasionally for emission failures. Usually the cars are in the shop for high CO conditions. I like testing these cars using my lab scope and CRT current probe. This is the simplest way I have found to test these systems. The key is to monitor closed loop O2 voltage with one channel of a lab scope and then check to see what the computer is doing with the mixture control device by using the current probe on the other scope channel. Figure 1 is a typical picture of what I may encounter on a warm engine, at part throttle. This one came from a 1981 Corvette mixture control carb that failed the emissions test with more than 10 percent CO.

The ECM in this feedback system limits fuel flow by cycling the M/C solenoid. When the solenoid is energized, it pulls the metering rods down into the jets for minimum fuel flow. The longer period the current is high, the less fuel flows.

This picture makes the answers to a few basic questions obvious. The computer is trying to control the mixture based on the info it's getting from the O2 sensor. The low voltage signal with rich exhaust indicates the need for a new O2 sensor. Figure 2 shows how a new O2 sensor reacts on this system.

With a new O2 sensor the computer can control the mixture and satisfy the converter's O2 requirement while running the M/C solenoid in its mid range. The graduation from M/C solenoid-controlled carbs came in the form of throttle body fuel injection.

The principles remain unchanged. To control fuel mixture, the microprocessor operates some type of a fuel delivery solenoid, based on a signal from an O2 sensor. The typical TBI fuel solenoid is fairly low resistance and requires a lot of current to open but not a lot to keep it open. Typically, a GM TBI peaks at 4 amps and holds at 1 amp. TBI drivers are usually 4 to 1 or 2 to 1.

Although this system is technically more advanced, testing is actually less complicated than the M/C solenoid system. Figure 3 is showing the injector current and O2 voltage of a 1984 Chevy Cavalier that failed emissions testing with a 3.5 percent CO reading.

With a current probe reading of injector current on channel A of the lab scope, and O2 sensor voltage on channel B, at part throttle on a warm engine, Figure 3a is showing that this car has a narrow O2 operating range with what looks like a normal injector pulse width range. The right side of the picture (Figure 3b) shows the screen capture image, while the left is showing the waveform data.

Taking the readings again after replacing the O2 sensor (Figure 4) allows for a better comparison. Figures 4a and 4b reveal that the ECM is now able to get the O2 voltage to swing in a wide range, while controlling the injector at a narrower pulse width range.

Take a look next at the O2 sensor and injector readings from the 91 S10 shown in Figures 5a and 5b.

At first glance the O2 signal seems to be working in a normal range. Looking at the peak-and-hold TBI injector pulse though, you can see that the ECM has to vary it over a half a millisecond to get the O2 voltage to swing. Figure 5a shows the waveform data with the min-max waveforms. Figure 5b is the screen image capture showing the voltage and current ranges in shaded areas. When comparing these readings to the new O2 sensor and injector current readings in Figures 6a and b, you can see quite a difference.

With the new O2 sensor, the ECM is able to maintain a much tighter control loop, holding the injector pulse width range to under 100 uS. This is five times narrower than before.

This system served its purpose fairly well on many makes and models throughout a decade but along came tighter emission standards with higher fuel economy demands and fuel systems graduated to the next level: port fuel injection (PFI).

By injecting fuel into the intake at the valve, engineers were able to get an even tighter reign on fuel control, while lessening exhaust emissions and improving fuel economy. Most early PFI systems fired all the injectors at once or in banked pairs. Testing closed loop fuel control is still just a simple matter of monitoring fuel delivery and O2 signal.

PFI systems generally control the injectors using what's known as saturated switch driver circuits. Less current is required to get these injectors to open than TBI injectors, usually less than 1 amp. Once the injector opens, current is limited either internally by the PCM or externally by a resistor. The typical PFI injector current waveform is shown in Figures 7a and 7b.

These came from a '91 Acura that failed the emissions test with a high CO reading. On this car, the O2 sensor was dead and the ECU wasn't in closed loop control of the injector pulse. With a new O2 sensor installed (Figures 8a and 8b), fuel control is occurring and emissions went back to normal.

Graduating to the next level of fuel control, along came sequentially fired port fuel injection systems, which gave us fuel injection timed with intake valve opening, optimizing performance and fuel mileage while lowering emissions. This system uses a driver for each injector and times the pulses based on camshaft position signal input. The injector and O2 waveforms in Figure 9 came from a '93 Cadillac SDV with a slight rough idle condition. The injectors were delivering a little too much fuel @ 4.23 mS while the O2 voltage hovered around 342 mV.

Once a new O2 sensor was installed and the Block Learn reset, you can see from Figures 10a and 10b that the PCM had regained fuel control at a lower pulse width of 3.5 mS, which smoothed the idle out again.

Contrasting the vertical current waveform of the M/C device (whether it be a mixture control solenoid or a fuel injector) against the horizontal waveform pattern of the oxygen sensor voltage (at 200 to 500 microseconds for fuel injection or 10 to 20 milliseconds for a mixture control solenoid) gives a superb overall view of a vehicle's closed loop fuel control system.

Using this system allows a technician the opportunity to document and prove testing results. This is an excellent self-training technique, a superb confidence builder, and a great way to improve customer relations.

[an error occurred while processing this directive]

RATE THIS ARTICLE What do you think of this article? Your input will help AutoInc. develop additional articles on this subject. Share your thoughts! Your name Your e-mail address

MOST ACCESSED ARTICLES [an error occurred while processing this directive] MOST E-MAILED ARTICLES [an error occurred while processing this directive]

AutoInc. Web Site | ASA Web Site | Transmission Fluid, Pressure and Problems | E-commerce and Its Influence on the Collision Repair Industry | The Future of Shop Web Sites | AutoInc.'s Annual Top 10 Automotive Web Sites | Guest Editorial | Tech to Tech | Tech Tips | Shop Profile | Net Worth | Stat Corner | Chairman's Message