Theory of Operation
The early introduction of the oxygen sensor came about in
the late 1970’s. Since then Zirconium has been the material of choice for
its construction. The Zirconium O2 sensor, as we all know, produces its own
voltage, which makes it a type of generator. The generated varying voltage
shows up on the scope as the familiar 1 Hz sine wave, when in close loop.
The actual voltage that is generated is the difference between the O2
content of the exhaust and that of the surrounding ambient air. The
stoichiometric air/fuel ratio or the mixture of air-to-fuel equal to 14.7:1
is the best mixture ratio for gasoline engines. At this ratio, the
combustion process happens with the most power being generated and the least
amount of emissions being produced. At a stoichiometric air/fuel ratio
(14.7:1), the generated O2 sensor voltage is about 450 mV. The ECM
recognizes a rich condition above the 450 mV level and a lean condition
bellow it. Therefore, these sensors do not care about the air/fuel ratio
above or bellow stoichiometry or 14.7-parts-of-air to 1-part-of-fuel. It is
for this reason that the Zirconium O2 sensor is called a “narrow band”
The Titanium O2 sensor was used throughout the late 1980’s
and early 1990’s on a limited basis. This sensor’s semiconductor
construction makes its operation different than the Zirconium O2 sensor.
Instead of generating its own voltage, the Titanium O2 sensor’s electrical
resistance changes according to the exhaust oxygen content. When the
air/fuel ratio is rich, the resistance of the sensor is around 950 Ohms
and more than 21 K-Ohms when the mixture is lean. As with the
Zirconium sensor, the Titanium O2 sensor is also considered a narrow-band O2
As mentioned before, the main problem with any narrow band
O2 sensors is that the ECM only knows that the mixture is slightly richer or
leaner than 14.7:1. The ECM has absolutely no idea as to the operating A/F
ratio outside the stoichiometric range. In effect it only knows that the
mixture is richer or leaner then stoichiometry. An O2 sensor voltage that
goes lower than 450 mV will cause a widening of injector pulse and
vise-versa. The resulting changing or cycling fuel control (closed-loop) O2
signal is what the technician sees on the scope when probing at the O2
sensor signal wire.
The newer “wide band” O2 sensor solves the narrow
sensing problem of the previous Zirconium sensors. These sensors are often
called by different names such as, continuous lambda sensors, AFR (air fuel
ratio sensors), LAF (lean air fuel sensor) and wide range O2 sensor.
Regardless of the name, the principle is the same, which is to put the ECM
in a better position to control the air/fuel mixture. In effect, the wide
range O2 sensor can detect the exhaust’s O2 content way bellow or above the
perfect 14.7:1 air/fuel ratio. Such control is needed on new lean burning
engines with extremely low emission output levels. The tighter emission
regulations are actually driving this newer fuel control technology and in
the process making the systems much more complex and difficult to diagnose
The wide range O2 sensor looks similar in appearance to the
regular Zirconium O2 sensor. Its inner construction and operation are
totally different, however . The Wide band O2 sensor is composed of a dual
inner layer called “Reference cell” and “Pump cell”. The ECM’s
AFR sensor circuitry always tries to keep a perfect air/fuel ratio (14.7:1)
inside a special monitoring chamber (Diffusion Chamber or pump-cell
circuit) by way of controlling its current. The AFR sensor uses
dedicated electronic circuitry to set a pumping current in the sensor’s pump
cell. In other words, if the air/fuel mixture is lean, the pump cell circuit
voltage momentarily goes low and the ECM immediately regulates the current
going through it in order to maintain a set voltage value or stoichiometric
ratio inside the diffusion chamber. The pump cell then discharges the excess
oxygen through the diffusion gap by means of the current flow created
in the pump-cell circuit. The ECM senses the current flow and widens
injector pulsation accordingly to add fuel.
If on the other hand the air/fuel mixture goes rich, the
pump cell circuit voltage rapidly climbs high and the ECM immediately
reverses the current flow polarity to readjust the pump cell circuit voltage
to its set stable value. The pump-cell then pumps oxygen into the monitoring
chamber by way of the reversed current flow in the ECM’s AFR pump-cell
circuit. The ECM detects the reversed current flow and an injector
pulsation-reduction command is issued bringing the mixture back to lean.
Since the current flow in the pump cell circuit is also proportional to the
oxygen concentration or deficiency in the exhaust, it serves as an index of
the air/fuel ratio. The ECM is constantly monitoring the pump cell current
circuitry, which it always tries to keep at a set voltage. For this reason,
the techniques used to test and diagnose the regular Zirconium O2 sensor
can not be used to test the wide band AFR sensor. These sensors are
current devices and do not have a cycling voltage waveform. The testing
procedures, which we will go into further along, are quite different from
the older O2 sensors.
NOTE: Whenever the air/fuel mixture is exactly at stoichiometry
(14.7:1) there is no current flow through the AFR sensor. This is
precisely what the ECM tries to do with the AFR signal. A properly
operating engine will always have very close to 0.00 mA of current
flow. The ECM commands more or less injector open time to try and
keep the AFR sensor as close as possible to 0.00 mA. A rich mixture
will produce a negative current flow and a lean mixture a positive
current flow. The actual AFR current flow is extremely small and for
this reason, the AFR sensor signal should be monitored with a scan
The AFR sensor operation can be thought of as being similar
to the hot wire MAF sensor. But, instead of a MAF hot wire, the ECM tries to
keep a perfectly stoichiometric air/fuel ratio inside the monitoring chamber
by varying the pump cell circuit current. The sensing part, at the tip of
the sensor, is always held at a constant voltage (depending on
manufacturer). If the mixture goes rich, the ECM will adjust the current
flowing through the sensing tip or pump cell circuit until the constant
operating voltage level is achieved again. The voltage change actually
happens very fast. The current flow through the pump circuit also pushes
along the Oxygen atoms either into or out of the diffusion chamber
(monitoring chamber) which restores the monitoring chamber’s air/fuel ratio
to stoichiometry. Although the ECM varies the current, it tries to maintain
the pump circuit at a constant voltage potential. As the ECM monitors the
varying current, a special circuit (also inside the ECM) converts the
current flow into a voltage value and passes it on to the serial data stream
as a scanner PID. This is why the best way to test an AFR sensor’s signal
is by monitoring the voltage conversion circuitry, which the ECM sends out
as an AFR-voltage PID. It is possible to actually monitor the actual AFR
sensor varying current, but the changes are very small (in the low mA range)
and difficult to monitor. A second drawback to a manual AFR current test is
that the actual signal wire has to be cut or broken to connect the amp-meter
in series with the pump circuit. Today’s average clamp-on amp-meter is not
accurate enough at such a small scale. For this reason, the easiest (but not
the only) way to test an AFR sensor is with the scanner.
NOTE: Some diagnostics literatures suggest testing the AFR
sensor by goosing the throttle and monitoring the actual voltage. On
a good sensor the voltage will snap down-and-up and then go back to
its normal level because the ECM will immediately adjust the current
to maintain the constant operating voltage. DO NOT use a multi-meter
in voltage setting to test the AFR sensor. The only voltage reading
that should be used is the ECM’s interpreted voltage value that is
displayed as a scanner PID from the pump-current detection circuit.
Another major difference between the wide range AFR sensor
and a Zirconium O2 sensor is that it operates at above 1200 Deg. F (600 C).
On these units the temperature is very critical and for this reason a
special pulse-width controlled heater circuit is employed to precisely
control the heater temperature. The ECM controls the heater circuit.
The wide operating range coupled with the inherent fast
acting operation of the AFR sensor puts the system always at stoichiometry,
which reduces a great deal of emissions. With this type of fuel control, the
air/fuel ratio is always hovering close to 14.7:1. If the mixture goes
slightly rich the ECM adjusts the pump circuit’s current flow to maintain
the set operating voltage. The current flow is detected by the ECM’s
detection circuit, with the result of a command for a reduction in injector
pulsation being issued. As soon as the A/F mixture changes back to
stoichiometry, because of the reduction in injector pulsation, the ECM will
adjust the current respectively. The end result is NO current flow (0.00
Amps) at 14.7:1 A/F ratio. In this case a light negative hump is seen on the
Amp-meter with the reading returning to 0.00 almost immediately. The fuel
correction happens very quickly.
Toyota among others has always been a strong supporter of
wide-range AFR sensor technology. The OBD II regulation calls for an O2
sensor voltage range from 0.00 to 1.00 volt. In order to meet the OBD II
regulation, Toyota rearranged the AFR sensor PIDs (from the detection
circuitry) by dividing their original OEM PID value by 5. The newer generic
OBD II AFR sensor PID ranges between 0.48 (rich) and 0.80 (lean).
NOTE: The AFR’s pump-current detection circuit voltage
range is the opposite of the regular Zirconium O2 sensor. With the
AFR sensor, the lower the voltage value the richer the mixture, and
the higher the voltage value the leaner. The OBD II generic AFR PID
is called air/fuel ratio sensor and NOT O2 sensor.
The following table gives the values of the Toyota OEM PID,
generic OBD II and the actual air/fuel ratio value.
The following summarizes the wide-range AFR sensor
operation. The Toyota AFR sensor is used here as an example, since the
operating voltages change from one manufacturer to another.
• The AFR sensor operates at a much wider air/fuel ratio
detection range. Hence the name wide range.
• The AFR sensor provides the ECM with a signal value
throughout a broad (wide) range of air/fuel ratios.
• The ECM current detection circuit voltage (scanner PID) is
totally the opposite of a regular Zirconium O2 sensor. The higher the
voltage, the leaner the mixture and vise-versa.
• The detection circuit voltage signal (scanner PID) output
is proportional to the current flow applied by the ECM to the pump cell
circuit (to keep the operating voltage) and an indicator of the air/fuel
• In AFR sensor fuel control systems, the ECM can more
accurately measure the actual air/fuel ratio on a wider scale. This allows
the ECM to adjust to stoichiometry much faster.
• With AFR sensor systems, the ECM does not cycle
(rich/lean) as in the older Zirconium type O2 sensor. The output bias or
pump cell circuit current detection voltage is fairly stable.
• With the mixture at 14.7:1, the AFR sensor pump cell
circuit current flow is 0.00 mA.
• The pump cell circuit current flow changes
polarity (by polar).
• A rich mixture produces a negative current flow in the
pump cell circuit.
• A lean mixture produces a positive current flow in the
pump cell circuit.
• Because the current can flow in either direction, the
AFR’s ground is NOT chassis ground. The AFR sensor uses a floating or
ECM ground, which could be held at a specific voltage level above chassis
ground (according to the manufacturer). Some manufacturers call this circuit
• The actual pump cell circuit current flow pushes Oxygen
atoms into or out of the diffusion chamber, depending on the direction of
the current flow.
• The detection circuit always monitors the direction
of the current flow and how much of it is flowing.
• Toyota AFR systems show an AFR PID of 3.30 volts at 14.7:1
A/F ratio. Each manufacturer uses a different PID voltage value to signal
the stoichiometric point. Toyota also divides its OEM PID by 5 in order to
arrive at an OBD II compliant voltage value.
• The leaner the mixture, the higher the detection circuit
voltage value (scanner PID). The richer the mixture the lower the detection
circuit voltage value (scanner PID).
• The ECM tries to maintain a stable voltage level
across the AFR’s sensing tip or pump cell circuit.
• The AFR voltage reading on the scanner is not the actual
voltage across the AFR sensor pump cell. The AFR detection circuit (inside
the ECM) generates the scanner PID voltage data from the pump cell current
flow. The pump cell voltage is kept at a stable value by the ECM.
• Wide-range AFR sensors are current devices and do
not put out an actual voltage for their signal.
• The current output signal flowing through the AFR circuit
is in the mA range and can not be measured with a clamp on amp-meter.
• The same factors that affect the Zirconium O2 sensor also
affect the AFR sensor (contamination, vacuum leaks, EGR failure, heater
• The AFR’s heater operation is very critical to the sensor
operation. These sensors operate at a much higher temperature than Zirconium
• The AFR heater is pulse-width modulated by the ECM to
maintain a stable temperature.
• The AFR sensor heater is usually ON (pulsing) under normal
• The AFR heater carries more current because of the
higher temperatures necessary. For this reason the connections are more
critical so as to avoid resistance in the circuit.
• The AFR heater circuit carries up to 8 Amps compared to
the Zirconium O2 sensor at 1.5 to 2 Amps
These sensors also have the added advantage of being able to
have the fuel control system adjust to any desired air/fuel ratio other
than 14.7:1 (Stoichiometry) or lambda 1. This option is especially
important in new fuel control concepts such as lean-burn engines,
where the engine’s fuel control changes at cruising speeds from 14.7:1 to a
much leaner 19.0:1 or even higher. The result is tremendous reduction in
emissions and fuel consumption. It is also worth stating that these leaner
engines require special catalytic converter units capable of reducing the
considerable amounts of NOx generated at such leaner (high temperature)
The two more prevalent
wide-range AFR sensor system manufacturers are Honda and Toyota. We will use
Toyota in this section for explanation purposes. However, the testing
procedure is always similar. The only changes will be in the biasing-
voltage, which changes from one manufacturer to another. The basic operation
is the same. Always learn the system before proceeding to further
diagnostics. AFR sensors are also starting to appear in an increasing number
of makes and models. It is expected that within the next decade most systems
will be of this type. The best way to test the AFR sensor operation is with
a scan tool. With that in mind, the use of a graphing-software is highly
recommended. This will ensure the quick recognition of the sensor’s
operating parameters much faster than simply looking at the numbers. The
human brain can process graphical information much better than raw data
• First, determine that there are no mechanical or air/fuel
density problems (vacuum leaks, clogged air or fuel filter, ignition timing,
stuck EGR, etc).
• Determine that the AFR sensor bias voltages are within
specs. Using a voltmeter, disconnect the AFR sensor and back probe the
sensor’s signal wires. Probe between ground and one of the signal wires,
then to ground and the other signal wires. The bias voltages should be
measured to ground. Measure the bias voltages and compare to specs. The
signal + wire is usually the pump cell signal circuit while the signal – is
the ECM provided floating ground. (Note: Some manufacturers are using an AFR
sensor with 6 wires. Two heater wires, signal +, signal – and an extra two
set of wires, which are the pump cell current signal and the calibration
• Perform a W.O.T. snap test. The scan tool AFR voltage
should reach a full low (rich mixture) voltage potential (Toyota – 0.48
volts (OBD II) or 2.40 volts (OEM), then a full high (lean mixture) voltage
value (Toyota – 0.80 volts (OBD II) or 4.00 volts (OEM). (refer to fig. –
• If the AFR sensor did not pass the W.O.T. snap dynamic
test above, suspect an open or a short circuit to the AFR current circuit.
The AFR sensor has to react to the sudden W.O.T. snap test regardless of
engine A/F conditions. A non-changing detection circuit is a very strong
indication that the AFR circuit is down. Internal as well as external
circuit faults are possible. To verify the fault (whether the fault is
internal or external) disconnect the AFR sensor and either measure the
continuity between the ECM AFR wires and the AFR connector or apply a
varying voltage (O2 sensor simulator – 0.00 to 2.00 volts) to the AFR
circuit. Look for a changing voltage in the scan tool’s AFR data PID. The
actual amount of change is not important, since this test simply checks for
a changing response. This verifies that the circuit is not open or shorted,
which would be indicated by NO change on the scanner AFR PID display.
• To perform a current response test of the pump cell
circuit, simply break the pump cell circuit wire and connect a digital
low-amp meter in between the broken circuit. Start the engine and warm up to
operating temperature. Operate the engine at different conditions and
compare to table 2. (Table 2 is a general operating current value
table, which may differ slightly from one manufacturer to another).
• Perform a voltage reading at the AFR heater power feed
wire. Most AFR sensors a fed power through an ECM controlled power relay
while the other side of the heater circuit is a pulse modulated to ground.
(Determines if the power feed voltage relay is operating properly).
• Connect a low resistance lamp circuit (headlight) to the
AFR heater circuit and start the engine. Verify that the lamp turns ON and
OFF. Note: The use of a low resistance lamp (headlight) is needed due to the
fact that on some systems, the ECM constantly checks the heater resistance.
On these systems, if a test light is used, the ECM simply shuts down the
heater circuit, since a test light has a resistance of close to 20 Ohms and
only draws about 300 mA of current, as opposed to the 8.00 Amps needed by
the AFR heater.
• Using a low amperage clamp-on meter, obtain a scope
waveform from the AFR’s heater circuit. The waveform should look similar to
a pulsating ignition coil, with a series of current humps indicating a duty
cycle (pulsing) controlled heater element. The highest part of the waveform
should be within 6 to 8 Amps.
• A lack of a current hump on the scope’s display points to
the heater circuit not working. Disconnect the AFR sensor and measure the
heater continuity. It should be close to 1.5 Ohm since the heater is almost
like a straight through wire, which is why it is duty-cycle-controlled. If
the heater circuit continuity reading shows an open circuit, replace the AFR
sensor. Otherwise, perform a resistance check of the heater circuit wires
between the sensor and the ECM/ Power feed circuit.
• A low current reading on the heater circuit indicates a
high resistance fault in the heater circuit.
• A higher than 8.00 amps current draw from the heater
circuit indicates a shorted heater inside the AFR sensor. Replace the AFR
sensor and re-check.