I've recently purchased excellent OBDII scan software. Now that I have this tool, I wanted to check out how my recently replaced O2 sensors were working. Well, needless to say, I saw some weird behavior.

I have a 97 Explorer EB V8 with 113K miles.

The behavior can be summarized as follows.

1. In general, the Bank 1 and Bank 2 sensors (both pre-Cat and post-Cat) behave quite differently.
2. The pre-Cat sensors on both banks have very different response curves (both period and amplitude).
3. Under varying load conditions, the Bank 2 post-Cat sensor stays high (indicating a high O2 condition), and only under heavy load will it change. But as soon as load returns to more normal conditions, the voltage again goes high.

In addition, after a recent run, I got the following error code.

Fuel and Air Metering
P133
O2 Sensor Circuit Slow Response (Bank 1 Sensor 1)

I wish I could post a couple of plots to help better quantify these differences. I would be most grateful to anyone who would be willing to look at some plots and provide their thoughts.

Thanks,
BB

 

...Welcome to Ford Truck Enthusiasts! If you have a minute check out the Guidelines, it will help you to navigate through and understand FTE.
Stick around someone will be by shortly to answer your question.
We are pleased you have chosen the best source for Fords!

Enjoy FTE
…..See you on the boards.

 

A couple of thoughts to ponder.
1. Pre-cat sensors should exhibit a fairly fast response in interval, and should vary fairly evenly above and below a centerline when engine load is not changing. This type of response indicates they are switching fast enough for the computer to see the results and adjust the fuel trim accordingly. Under a heavy load/ acceleration, the sensors may exhibit longer periods of rich readings which is to be expected as the computer richens the mix for short durations.

2. Post cat sensors should exhibit a fairly steady state reading, assuming constant engine load and the cats are performing properly. If a richer mixture is supplied to the engine under a changing load, the post cat sensor reading may dip for a period of time, but should return to a semi-steady state once the load returns to a more steady state.

3. Both sides of the O2 sensors should report similar readings. If not, then some possibilities to look at
Pre-cat
a. O2 Sensor no reading (defective sensor or wiring) / not switching fast enough to detect PCM commanded fuel trim changes. (your code 133)
B. Fuel trim changes are not being properly applied due to (low/high fuel pressure, injector malfunction(s), incorrect or inaccurate air flow measurements/leaks, ignition related problems (spark misfire))
Post-cat
c. O2 sensor no reading( defective sensor/ wiring) or rapid switching (following pre cat sensor/ or steady state rich); defective or insufficient cat reactance.

In essence, the pre-cat O2 sensors are the most important ones to look at to try to determine the source of a problem.
In your case, if the code matches the cycle rate you see on the display, most likely the O2 sensor is slow to respond to fuel trim changes and should be replaced.
Dialtone

 

Hey Dialtone,

Thanks for the detailed response. If you don't mind helping out a newbie, I have a couple of comments and questions.

1. If I understand your comments about the O2 sensor output correctly, a high voltage indicates a rich condition, and a low voltage indicates a lean condition (opposite to what I wrote below). Is this correct??

2. In my case, both pre-cat sensors are showing what appears to be a high enough switching speed, but are definitely not the same in shape. At times, B1 is switching much faster than B2 and at other times, it's the opposite. For the most part though I'd say that for any period of time, the two responses are dramatically different in switching speed and amplitude.

3. My post-Cat sensor on Bank 1 is never steady state. It follows the pre-Cat sensor with a nearly identical waveform. This makes no sense to me. It would seem to me that the my brand new cat is not doing anything. Would you agree?

4. My B2 post-Cat sensor is steady state, but always at a high voltage level (0.7 volts). This makes more intuitive sense because I would expect the O2 to be used up in the catalytic action and thus, a high voltage would be present. The dips in the response, however, appear to be completely uncorrelated with the engine loading. The Calculated Load can vary dramatically, with little variation in the B2 post-Cat sensor. It definitely does not dip during periods of high load. At one point, I ran the truck very hard, at 100% load, and the post-cat sensor increased slightly from steady state (0.7 volts) up to 0.9 volts, and then when the load was reduced, it returned to it's steady state.

5. I'll definitely take a look at the fuel rail pressure and whether or not I've got a fouled plug indicating some misfires. The tool I'm using to scan the OBDII data is called AutoEnginuity. With this tool, I have yet to see any misfire conditions, but perhaps it's being masked.

Two last questions:
A. My injectors are factory original. What's the best way to detect whether or not I've got a leaky injector ?
B. This is a very general question. Why is the O2 sensor output a switching response? I would think that the O2 content under steady load would be constant.

I know it's difficult to be able to comment on these readings without actually seeing them. I do appreciate your feedback.

Thanks again for your help.
BB

 

1. Sorry for the incorrect wording. Here is an excerpt from an article on how O2 readings correlate to the voltage outputs:

"The difference in oxygen levels between the exhaust and outside air within the sensor causes voltage to flow through the ceramic bulb. The greater the difference, the higher the voltage reading.
An oxygen sensor will typically generate up to about 0.9 volts when the fuel mixture is rich and there is little unburned oxygen in the exhaust. When the mixture is lean, the sensor’s output voltage will drop down to about 0.1 volts. When the air/fuel mixture is balanced or at the equilibrium point of about 14.7 to 1, the sensor will read around 4.5 volts."
2. It would be really hard to say what is normal. Each bank can be independantly trimmed, so the readings could vary slightly based on individual cylinder differences in airflow, compression, spark, and fuel delivery. The best you can hope for is that the trim changes average out for all cylinders on the particular bank. It is intresting to note that if one studies the pinouts on the EEC unit, there are provisions for up to 10 O2 sensor inputs. If all were equipped, a V8 engine could have 1 O2 sensor for each cylinder and 2 leftovers for post cats on each bank.
3&4. I would agree, the reading should not follow the pre-cat sensor. If in doubt, ask a friend to loan you their vehicle and take a look at their readings for a camparison of pre-post cat readings so you can get a baseline comparison. I used this technique for years in electronics when unsure of the exact functions in a circuit. I called it "Stare and Compare"

For your last 2 questions:
a. Short of removing the injectors and subjecting them to a controlled flow test, I do not know of any reliable way diagnosing a defective injector. The soleniod portion (electromagnetic) can be tested using a volt/ ohm meter to verify coil continuity, but this does not help for a stuck closed/ partially clogged tip / potential leaker condition.
One possible indication of a leaker would show up as a leakdown of fuel rail pressure when engine is shut off. The rails should hold pressure for a considerable amount of time after engine is shut off. BUT the fuel pressure regulator could also be leaking past and causing a similar observation. The only way I know to determine which is the cause is to remove the fuel return line and see if fuel is leaking past the regulator, or (in the case a a bad diaphram) into the vacuum regulator line.
b. I suspect 2 reasons. Fuel trim is not exact. Since 1 O2 sensor is uesd to monitor 3-4 cylinders, it is not possible to achieve perfect balance on all cyclinders due to slight differences in each one. The computer uses a strategy of deviating the injector pulse width timing above and below a calculated baseline and then monitors the O2 readings. This produces an average that should be at or near the calculated trim setting. If not, then the trim is recalculated, put into effect, and verified. Should the calculations deviate too far from pre-programmed min-max values, a code is set to indicate this.
Second, the exhaust gas stream is not a constant flow, but a set of compression waves as each cylinder is exhausted. Assuming the O2 sensor can switch fast enough (and is sampled fast enough) to detect these waves, it may also explain some of the peaks and valleys in the readings.
Dialtone

 

Narrow band O2 sensors have a very sharp input/output curve. They don't like to stay in the middle range -- they tend to read 0.9 volts or 0.1 volts but not stay in the middle.

There is a substantial lag between changes in fuel trim and effect on O2 sensor output. Think about the distance involved, the inertia effect of residual exhaust gas, etc. The computer and O2 sensor form a feedback system with sufficient phase shift that it is difficult to get it to not oscillate. Combine this with the response curve of the sensor and it would be hard to build an engine control system that managed to stay exactly at stoichiometric.

Instead, the switching and oscillation is considered a feature. The faster she oscillates, the better. The periods of lean running induce enough oxygen to balance out against the periods of rich running. When the engine is lean, it forms NOx. When the engine is rich, it forms HC. Modern 3-way catalysts can react the HC + NOx by averaging it over time and reduce/oxidize it all to N2 + CO2 + H2O.

A post cat sensor waveform that follows the pre-cat sensor waveform indicates a non-functional cat, a cat that isn't warmed up, or really excessive swing in the fuel trims (possibly caused by slow response in the pre-cat sensor).

 

Hey Dialtone and fefarms,

Thanks for the great replies. This helps out considerably.

I do have one last theoretical question. If you were measuring the O2 content of an exhaust that had a constant, steady level of 02, and the exhaust flowed at a constant rate (unlike a real exhaust that has the exhaust pulses from each cylinder), would the O2 voltage output still have a switching response ?

Basically, is the switching response inherent in a O2 sensing system architecture, or is it a result of the non uniform exhaust output?

Thanks again for the detailed responses.

BB

 

If the amount of O2 is constant, the sensor will stay at a constant voltage.

If the mixture is rich, it will stay at 0.9 volts, even in the face of considerable variation in the actual mixture ratio.

If the mixture is lean, it will stay at 0.1 volts, even in the face of considerable variation.

If the mixture is close to stoichiometric and we hypothesize "a constant steady level" of O2, then the sensor will read some consant steady intermediate voltage. The point is, it is unstable in this regime -- just like balancing a broomstick on the palm of your hand. You can hold it there, maybe forever, but any disturbance will cause it to want to rapidly snap to a more stable state.

The sensor has a steady, albeit sharp-edged, "oxygen content" to "voltage" response curve.

It is the architecture of the SYSTEM (including the sensor, PCM, fuel injectors, and engine) that leads to the switching behavior. The oxygen sensor itself is too slow to react to each individual exhaust pulse -- although the slight time-axis variablity is one of the destabilizing factors in the stoichiometric region (think about the broomstick on the palm analogy -- imagine standing in the rain holding the broomstick on your palm. The broomstick takes longer to fall than the period between rain drops. But it gets irritating having the drops splat on your face and you don't do as well holding it vertical). OK, maybe that last one was a little too strained...

 

One more comment:

A slow oxygen sensor has the same effect on the fuel injection system as several ounces of ethanol have on you.

Imagine balancing the broomstick when you are drunk. Your reaction times are slowed and the broomstick tends to fall farther between corrections...