Jeep Enthusiast Forums banner

Oxygen Sensors (more info than you probably want)

123K views 48 replies 21 participants last post by  LouC  
#1 · (Edited)
I have been collecting O2 sensor info over the years. Half of the following is in my own words so feel free to correct anything you see wrong.

The O2 sensor is there primarily because of emissions! A second reason is to fine tune fuel management. And thirdly they indicate the state of the catalyst.

It was developed to ensure that the correct amount of oxygen is available for a reaction to take place in the CAT which needs O2 to convert other nasty gases to form less nasty gases. (Example carbon monoxide + oxygen = CO2.

The sensors upstream (before the Cat) are important for delivering the correct amount of oxygen and for fine tuning the air-to fuel ratio. They help determine the duration of time that the injectors stay open. The oxygen sensor measures the oxygen content from the exhaust stream and from this it can figure out if the fuel mixture is rich or lean

The downstream O2 sensors are for measuring the catalyst shelf-life. This will tell if the Cat is spent and needs replacement. This sensor has no effect on the engine performance if it the vehicle is OBDI (up to 1995).

A rich mixture consumes most or all of the oxygen so the voltage generated by the O2 sensor will be high (0.6 to 1.0 volts). A leans mixture will have more available O2 left over. In this case the sensor will produce a voltage from 0.1to 0.4v. The perfect average voltage is around 0.45 v for complete combustion.

The sensor will only generate an accurate signal when it is at a minimum of 400 C, so it is ignored until the engine control goes into Closed Loop operation (an engine at operating temperature). The O2 sensor also has its own built in heater (2amp load) to get it to temp quickly and so that it is hotter then the exhaust stream. This cuts down the emissions as well.

Both O2 sensor measures how much O2 is present in the exhaust stream. The computer does a simple calculation (upstream O2 value - downstream O2 value). Obviously, the upstream valve is always a higher valve. Since a chemical reaction takes place, there is less O2 exiting the Cat. So the math as I see it:

If (upstream O2 value - downstream O2 value) = a negative number, than one of the sensors is bad.
If (upstream O2 value - downstream O2 value) = 0, then the catalyst is spent and needs to be replaced.
If (upstream O2 value - downstream O2 value) = a positive number, then the catalyst is converting oxygen and other constituents to form less toxic gases.

Testing the Heater
You can easily test the heater circuit with the two white wires. With the engine cold and off unplug the O2 sensor connector. Check for resistance on the sensor side. An Open circuit or mega-ohms means a bad heater circuit. Throw the O2 sensor in the trash. . I measured 4.5 ohms at 45 degrees F outside. This will always store a code. The heater circuit should draw about 2 amps when hot.

Testing the measuring circuit
The output of the measuring circuit is more difficult. With a hot engine one can back-probe the connector. The voltage output is a Sine-wave. So if you connect a voltmeter across the correct pins (gray & black wires), then the meter will constantly bounce back & forth between 0.1 & 0.9 volts. A good sensor will bounce between a much narrower range like (0.25 to 0.75). The important thing is that the average of the swing equals 0.45 volts. I have not tried to measure this but I think it will be too difficult to measure on a voltmeter unless it has an averaging function. (only the really cool & expensive ones have this)
While measuring this circuit, you can pull off a vacuum hose and check to see if the voltage drops (a lean mixture)

A bad upstream sensor can cause a decrease in mileage, higher CO and/or HC emission levels

Bank 1 sensor 1= the side of the engine which has cylinder number 1 on it and this is upstream of the Cat. This is the driver's side on Jeeps.
Bank 1 sensor 2 = driver's side downstream sensor or after the Cat.

Bank 2 sensor 1 = passenger side upstream sensor
Bank 2 sensor 2 = passenger side downstream sensor

Harness Connections
Heating circuit
Orange/with black stripe goes to white wire on sensor (heater 12 volts DC)
Black/with light blue stripe goes to the other white wire on sensor (heater ground)

Sensing circuit
Black with orange stripe goes to the gray wire on sensor (0.1 to 0.9 volts AC signal from sensor)
Black with tan stripe goes to black wire on sensor (sensor ground)

Many things can cause the sensor to read incorrectly: the built-in heater circuit not working, supplemental gas additives, vacuum leak, and ERG not working can all cause the O2 sensor to appear to be bad.

Unfortunately, the best way to determine if the sensors are bad is to have a proper scan performed. That way the outputs are plotted and it can be determined if a sensor is lazy or not functioning properly

Many people have reported problems with aftermarket O2 sensors! I would go to your local Dodge or Jeep dealer and give them your VIN number. That way you hand you a sensor that will work correctly with your Jeep; or email Nick at Kolak@aol.com and tell him that you got his name from this forum for a discount.
 
  • Like
Reactions: tlspiers
#3 ·
Thanks for putting that together…Very informative. What are your thoughts on when to replace them? I have seen a few posts where some people change these sensors as part of their preventive maintenance program in the hopes of maximizing fuel economy. Others wait until they get an engine diagnostic codes letting you know its time to change them.
 
#5 ·
Additional Info Relative to WJs

In order to be more effective in removing harmful emissions, WJs, and I guess later models, use the downstream O2 sensor for additional fine tuning the A/F. Downstream sensors are still used to determine catalytic converter efficiency also. Attached is a description from the WJ FSM (I only attached info for California emissions system, the federal emissions system works the same except for fewer cats and O2 sensors.)
 

Attachments

#6 ·
I did not realize that. Thanks for additional info:thumbsup:
Since the title of the abstract is "Downstream Sensors (California vehicles)" I am guessing that this applies only for California vehicles.
 
#7 ·
All OBDII Jeeps, that is 1996 and later, utilize the downstream sensor(s) not only to monitor converter health, but also to provide for further fuel trim to ensure converter efficiency. The downstream sensors should thus be replaced at the same intervals as the upstream sensors because a sluggish downstream sensor can mistakenly countermand the input from the upstream sensor.
 
#8 ·
tjtrimble said:
Thanks for putting that together…Very informative. What are your thoughts on when to replace them? I have seen a few posts where some people change these sensors as part of their preventive maintenance program in the hopes of maximizing fuel economy. Others wait until they get an engine diagnostic codes letting you know its time to change them.
I'd like to see some information on this too. Is it worth going a head and replacing mine this summer just to have it done? I'm sure they are probably original after just about 162k. Or should I wait until I know it's shot?

Also, I've heard bad ones can 'soft code' and not give check engine lights. Is it true that an OBDII reader can read this and tell you its bad as well? Reason I ask is that a friend of mine has a small crack in his exhaust manifold but gets really bad gas mileage. I have trouble believing that it's just the manifold and not a bad O2 sensor.
 
#9 ·
Older ones seem to crap out around 60,000 miles. Newer ones are good for at least 100,000 miles. I would replace it if a code is thrown or if a sudden drop in mileage is observed (calculated by doing the math and not using the overhead display/VIC)
 
#10 ·
greasefingers said:
Older ones seem to crap out around 60,000 miles. Newer ones are good for at least 100,000 miles. I would replace it if a code is thrown or if a sudden drop in mileage is observed (calculated by doing the math and not using the overhead display/VIC)
What about it being bad and not throwing a code. I seem to remember it happening before. Would a reader at like Autozone pick it up?

As for my ZJ, I haven't had any mileage drops and either way, the VIC/overhead is always with in a tenth or two of my calculated mileage. So i'll probably just leave them until I get ambitious and get all my other mods in.
 
#11 ·
An O2 sensor can drift off of calibration (like any sensor). So it would not show a code if it was going bad (drifting or off calibration). It will throw a code if it is out of some specified range. Fouling can also insulate it (thermally and chemically) That is why a history or plot of the voltages vs time is so important in determining the behavior of the sensor.

A reader at Autozone is useless (unless you have a 1998 Jeep which the ignition key trick does not work) Readers just read the stored codes. The Jeep dealer can actually read more codes than aftermarket readers can do. But a scanner that can plot is the most valuable tool for determining sensors’ characteristics.
 
#12 ·
I have a scanner ( a black box) that connects to my Laptop, for my Toyota 4runner. The 4runner uses a different protocol than Chrysler vehicles, so I can not use it for the Jeep. But I have plotted the four O2 sensors when I bought the new 4runner. Now or anytime in the future I can collect the raw data again. Then I can co-plot the data sets and see if the O2 sensors are bad or drifting from their original brand new state.
 
#15 ·
Sensors are fun (OK maybe the Geek is coming out in me). I have removed some of them from my Jeep and measured certain values. The best example is the CTS (coolant temp sensor). I removed this from the thermostat housing (on 4.0L engines) and submerged it in a pot of water (when my wife was not at home) Then I increased the stovetop burner and paused for equilibrium and recorded the resistance and temperature. I repeated the procedure at different temps.

Now using Excel I co-plotted my data and the data provided from the FSM (factory service manual). BTW the FSM was the best investment I have ever made on my Jeep! Once the curves are plotted an educated evaluation can be determined about the sensors characteristics.

Below is an example of collecting data. As see below my original 13 year old CTS (coolant temp sensor) is as good as it was back when it was new [1994]
Image
 
#16 ·
Thats pretty cool. Do you have any information on a reader like that? I might have to look into getting on in the future. What about adapters. Would one for a 96 ZJ work for a 99 XJ too? I don't know anything about that, but if it did, I could probably get my buddy to split the cost. We were talking about getting a code reader together but this might be way better. Does this have the same functionality as a code reader though in addition to it data gathering ability?
 
#17 ·
I bought this one.
http://www.myscantool.com/index.html
I wanted a scanner that connected to a Laptop so that I could have a history of information. A reader only reads in real-time and then the info is lost unless you can save the data to a computer.
 
#18 ·
greasefingers said:
...
Since the title of the abstract is "Downstream Sensors (California vehicles)" I am guessing that this applies only for California vehicles.
I wouldn't say it applies only to California vehicles. Model years 1999 - 2000 did have two different exhaust systems available, California (3 cats and 4 O2 sensors, required by more strict Cali emission standards) and Federal (1 cat and 2 O2 sensors, most of the rest of the country). I have no idea of the numbers of WJs with the two different systems. By 2001, essentially all WJs came with the California emissions systems.

Either way, the downstream sensor or sensors performed the same functions.
 
#20 ·
I do not replace a sensor until it throws a CEL at least a couple times. I will always reset the code and see if it comes back. If it keeps coming back relatively quickly, then it probably means that the sensor is dead. I never replace O2 sensors prematurely. You never know if the next one is a new manufacturing defect waiting to happen. If the O2 sensor could leave you stranded, then I would change my tune.
 
#21 ·
Just to add to the discussion on monitoring catalyst efficiency, I have read elsewhere that the computer doesn't simply subtract the upstream voltage from the downstream voltage, it looks at the rate of change.

An (upstream) O2 sensor reading ocillates very rapidly between rich and lean states, this is considered normal operation. It hovers around, but never remains on the .45V "center" mark. The exhaust gasses exiting the converter will also ocillate around a center reading, but at a much slower rate. This is because throughout the catalytic process, one part of the catalyst stores spare oxygen from the lean portion of the cycle to be reacted with the unburnt HCs during the rich portion of the cycle.

So, the catalyst's efficiency is determined by the PCM by it's ability to "buffer" the rich/lean ocillation as noted by the difference in the rate of change between the upstream catalyst and the downstream catalyst.

For any O2 sensor, if the computer sees that the signal is not changing, whether rich, lean, or center, it assumes the sensor is bad and sets a trouble code.
 
#29 ·
Here is what I have, the Actron CP9190: http://www.actron.com/product_detail.php?pid=16317

I got it new on ebay for about $240 shipped. Its amazing. Hooks up to laptop and can be updated for free so it will never be outdated. You can print out the results and also reads Live Data as the vehicle is running. It performs tests as well.

This one can do OBDI and OBDII. It does '84-'95 Ford, '84-'95 GM and 89-'95 Chrysler vehicles. And all '96 and up vehicles (including CAN).

The Actron CP9180 is similar but only does OBDII. Cheaper too.
http://www.actron.com/product_detail.php?pid=16318
 
#23 ·
Splitting the cost is a good idea as it is only used a few times a year.

To plot you have to export the data to Excel. Or you can Screen dump real time plots to the printer.
 
#24 ·
Alright. I've got excel laying around. Only problem is we dont have laptops (yet), getting those for college during the summer...T61's baby :) But I can just jack my moms new Toshiba one. But yeah, I figure spliting the cost is the best way, just not sure if we want Chrysler only or all makes. *shrug*
 
#25 ·
It's a really cool tool to have. I am very much a sensor geek, so I really appreciate it.

Here are some things to keep in mind……………………

The codes are stored in the RAM (random access memory) which always has 12 volts supplied to it. To clear the memory the battery must be disconnected to erase the RAM.

Hint: there is also a fuse in the power distribution center, which can be removed instead of the battery.

Inputs (like the coil, MAP, or coolant temperature sensor) do not speak the same language as the CPU so an analog to digital converter is first needed.

Anything connected to a computer is based on the digital on/off language. Zero (0) = off and one (1) = On. So a 12 volt (analog) signal from an oxygen sensor must be converted into a bunch of zeros and ones. It is either on or off and nothing in-between. . Any numeric (1,2,3,etc) can be expressed by combinations of zeros and ones.

Then it compares the Inputs with a set of instructions (or a curve) preprogrammed into the ROM (read only memory). Then it decides what to control with it (like turn on an actuator which can be a fuel injector) But before it does this the reverse process must happen. The digital signal from the CPU output must be converted back to analog (D to A converter) because the fuel injector is a 12 volt (on or off) device.

Don't give the PCM/ECU too much credit. There is no sticker with "Intel Inside" on it. There is nothing wrong with simple math (addition & subtraction) if you can get the necessary info out of it. Especially since a waveform like a sinewave goes from a positive slope to a negative one several times a second. Vehicle computers do not perform any complicated operations. Instead it performs thousands of very simple operations, very fast. A clock keeps everything in some type of logical order.

So an RPM (speed) signal or an oxygen sensor is delivering a waveform to the computer. The computer is only looking for the "on" signals and not a voltage. If you connect a volt meter to the same output, it misses practically all the data and instead yields only an average valve for volts at that given point in time. The Napkin diagram below is how a waveform type analog signal is converted to a digital one.
Image
 
#26 ·
That's kinda interesting. For what little I apparently know about computer operation, I got a decent amount of what you said. :D Regardless, this should be a good investment even if I just toy with it most of the time and only need it to diagnose once in a while.