• Register Today It's free! This box and some ads will disappear once registered!

Engine Sensors - Part 1 Oxygen Sensors

Glacier991

EF Tranny Guru
Moderator Emeritus
Joined
February 8, 2003
Messages
9,832
Reaction score
33
City, State
Sacramento, CA 95827
Year, Model & Trim Level
1992 XLT
Engine Sensors - Part 1 Oxygen Sensors

Oxygen Sensors…. One of the most misunderstood parts of today’s modern automotive engines. Also the part that is most often associated as being "bad" by the general population when they see a “check engine” light (sometimes labeled “Service engine” or “MIL”). What are hey? How are they made? How do they work? What do they do? What does Closed loop mean? What if I get a "check engine" light relating to an O2 sensor? I hope by the time you get done reading SENSORS – PART ONE, you’ll know.

What are they? Oxygen sensors measure the oxygen content of exhaust gases. Nothing more complicated than that. They report the amount of oxygen as a voltage – they output a generated voltage (the exception being titania sensors, discussed below). A rich mixture has more unused oxygen in it, and the sensor will output up to about .9 volts in that condition. A lean condition will have less unused oxygen in it. The sensor will output much less voltage in that situation, down around .1 volts. They are used to provide feedback to the auto’s computer to tell it when to enrich, and when to lean the fuel mixture. [Which it does by lengthening or shortening the time the fuel injectors are open – pulse width it is called]. Practically speaking, what this means is that the computer is chasing the cat’s tail constantly. Too rich, too lean, and so on and so forth. This back and forth of the mixture produces a sinusoidal wave (well kinda) out of a healthy 02 sensor. The “perfect” air/fuel ratio is considered to be 14.7 parts air to 1 part fuel, sometimes seen as 14.7:1 or called Lambda in the parlance of the Society of Automotive Engineers (SAE). Do you remember the late 70’s Volvo with a “Lambda sensor” Now you know what it was. The first 02 sensor. Back to the wave. It will vary between about .2 and .8 volts crossing over the “Lambda” voltage (which happens to be .445 V) about 2 – 8 times a second for a healthy sensor on a multiport engine... speed varies depending on the type of fuel injection – feedback carburetors being slowest (and almost extinct) throttle body injection the next slowest and multi port the fastest. Here is a picture of a scanner readout of a 4 O2 sensor engine.....reading bank 1 and bank 2... you can see that one bank is rich and one is lean, which one is which?

15286dscn3556.jpg


then, taken immediately after

15286dscn3555.jpg


see the change?

How are they made? There are two designs of sensors. One, which we will not discuss is the Titania sensor, and instead of creating a varying voltage offers a varying resistance. Jeep and some imports use it. The other is a Zirconia sensor, which produces a voltage. This is the one we will discuss. Zirconia Oxygen sensors consist of a bulb, which is part of an assembly which screws into the exhaust manifold so the bulb is in the exhaust stream. The bulb is either made of or contains zirconium ceramic. Inside the bulb are platinum electrode(s). The bulb itself has slits in it so that the inside vents to the outside and exhaust gases pass through it. That is the basic construction. Oh, maybe this is a good place to note something. Fastest way to kill an O2 sensor? Silicone. Spray a silicone spray into the intake of a modern engine and your O2 sensors will die in a heartbeat. Ever wondered what it means when you see products labeled “O2 sensor safe?” It just means there is no silicone.

Here is a sensor for a 92 Explorer:
15286DSCN6306.JPG

15286DSCN6307.JPG

15286DSCN6304-med.JPG


How do they work? Well actually I think this is pretty well answered above. I cannot tell you the chemistry or physics, so I guess continuing this one is kind of pointless, other than to note that heat is the important factor (but I understand NOT the only factor), so they closely track exhaust gas temperature, which also varies based on A/F ratios. (In fact aircraft engines are adjusted for proper A/F ratios by adjusting mixture observing exhaust gas temps in fact since they do not have fixed injector pulse widths or carburetors). Anyway the best thing I can tell you is O2 sensors do work, and do it well. How long? Well in many cases hundred thousand miles or even two hundred thousand miles. As they age, deposits build up and they get slower in their crossover rates, and this can affect fuel efficiency. Expect 75,000 miles easily provided your engine is running ok, and no one is running around with a can of silicone spray nearby and no one has used silicone seal on gaskets in the engine.

What do they do? When they first arrived on the scene, engineers tended to use a single O2 sensor to feedback to the computer to adjust air/fuel (sometimes referred to as A/F) ratio. Later as emissions relied more heavily on the Catalytic Convertor (CAT), we added one AFTER the convertor, the idea being that the after CAT (or post CAT) O2 sensor would not vary as much in output voltage since the CAT should have used up most of the oxygen…. So the output should be “lean” or at least switch much less. Here’s a quiz – what do you think the output voltage on a post-cat O2 sensor will be closer to? .1 V or .8 V? If you do not know, go back and re-read the beginning.

As we moved into more sophisticated systems, we added more sensors, including O2 sensors. Now we sport four O2 sensors in most cases, one on each bank, before the CAT and one after (post) CAT.

Open Loop or Closed Loop? An engineer will tell you that a system which uses output feedback to control its operation is considered a “closed loop” system. One which does not is “open loop”. We know that an O2 sensor runs on heat. An engine in startup does not produce enough heat to make them work. SO…. On startup, the engine computer fuel system runs in “open loop”… and utilizes stored information to provide fuel metering. Once the O2 sensors get warm enough to start operating, the system switches over the them, and into “closed loop”. Starting in the mid 90’s we added heaters to the sensors to speed up this process, since closed loop was friendlier to emissions. Thus the Exhaust Gas Oxygen Sensor, or EGO sensor became “heated” and is now called the Heated Exhaust Gas Oxygen sensor or HEGO sensor. When people see HEGO sometimes they get all confused. It is just the Oxygen Sensor, with heat. Heated sensors usually have 4 wires, non heated 1 or 2.

What does it mean when I get an O2 Sensor Diagnostic Trouble Code (DTC)? The computer watches the performance of the sensors. For example, if pre-CAT sensors stay above or below the .445V it sees this and will set a code, either rich or lean depending. Does that mean the sensors are bad? NO! It means in about 99% of the cases that the sensors are reporting a condition that is either RICH or LEAN, that’s all. Now WHY… is up to you. Lean is often a vacuum leak, and rich is often fuel pressure or intake air related. So if you get a code involving “O2 sensors indicate lean, bank 1” do not assume your sensor is bad and replace it. That would be like having a scout come back and tell you “Sir, the enemy is 5 miles ahead in numbers” and shooting him because you didn’t like hearing his message. (This is a little simplified but you get my idea here how this works, right?)

How do I remove them? Well, truth is they can be real buggers to get out. You may want to try and spray them with penetrating oil, remembering that it will smoke until it burns off once you start up again. They have a wire coming out of the end, so an ordinary socket won;t do, you need a special tool. Here are two. A silver aftermarket one, and the FORD Rotunda tool made by Owatonna Tool Company, or OTC (quality stuff).

15286DSCN6278-med.JPG


here is the OTC one in my hand, sharp eyes may notice the year and number, which until recently was how OTC identified their tools for FORD. When you have room, this one seems to work best, but sometimes a straight shot with the other one is all you can get. The wires go into the slots, in case you wondered.

15286DSCN6279-med.JPG


I am sure I will think of more to add to this, but this seemed like a good start on the sensor series. For those of you for whom this may have saved $75 by NOT replacing a perfectly good O2 sensor, why not use $20 of that and become an Elite explorer and "join" this site? No, ok, then how about buying me a cold one someday? Or better yet... both?
 


Join the Elite Explorers for $20 per year. Gets rid of the ads! New $5 per month "try out" option.

Explorer Forum has probably saved you that much already, and will continue to save you money as you learn how to diagnose fix problems yourself and learn which modifications work without having to experiment on your own. Elite Explorer members see no advertisements, no banner ads, no double underlined links, can add their own profile photo, upload photo attachments in all forums, and Media Gallery, create and save more private Conversations, and more. Join Today. Your support is greatly appreciated.




ExplorerDMB

Moderator/Technician
Moderator Emeritus
Joined
January 26, 2004
Messages
6,162
Reaction score
8
City, State
Mechanicsville, Virginia
Year, Model & Trim Level
2004 Acura TL
Thank you for posting this. Very good information and I agree that O2 Sensors are very misunderstood and misdiagnosed.

-Drew
 




410Fortune

ICE Season
Staff member
Moderator
Joined
August 3, 2000
Messages
26,612
Reaction score
2,288
City, State
NORTH IDAHO, 7B
Year, Model & Trim Level
B2 "Slightly" Modified
Callsign
FOURTEN
Excellent info! Thanks for taking the time to write that up :)

I have had a few issues with HEGO sensors on the Ford drive trains I have worked on.
Many times the problem was with the heated power wire from the PCM to the sensor. I have seen problems with the wiring itself, sometimes melted wires, old Ford splices, or the connectors not being fully seated.
So the wiring to a sensor may also be the source of a code. Easiest way to find out is to monitor the sensor output voltage and also check the wiring between the PCM and the sensor harness for continuity.
 




spindlecone

Explorer Addict
Joined
April 30, 2004
Messages
4,836
Reaction score
0
City, State
San Jose, Calif
Year, Model & Trim Level
2003 sport trac
Great Info
Question about OL,CL, with my scanner, at startup I am in OL for about 5 secs, than the sensors kick in and switch to CL (normal)
At WOT, you go back to OL,Why, what is happining.
 




ExplorerDMB

Moderator/Technician
Moderator Emeritus
Joined
January 26, 2004
Messages
6,162
Reaction score
8
City, State
Mechanicsville, Virginia
Year, Model & Trim Level
2004 Acura TL
What happens is that to get maximum performance it richs the mixture. That is why it goes into OL at Wide Open Throttle. I believe if it was to stay in CL, you wouldn't get the performance you do at WOT.

-Drew
 




Glacier991

EF Tranny Guru
Moderator Emeritus
Joined
February 8, 2003
Messages
9,832
Reaction score
33
City, State
Sacramento, CA 95827
Year, Model & Trim Level
1992 XLT
Drew is right. The computer is programmed to drop out of closed loop at some % of WOT as indicated by the TPS sensor (usually 99%). The tabular value of 12.5:1 A/F is substituted and the computer will try and reach that goal. Once it drops out of OL, it will once again shoot for 14.7:1 A/F (Lambda or Stoichiometric.)

Interestingly, a similar strategy is used in fuel injection to mimic a carburetor. We all know that pumping the throttle on a fuel injected engine does nothing to aid starting, as there is no accelerator pump, BUT.. if you hold the throttle wide open, the computer will sense that and sense a cranking mode, and STOP fuel delivery, mimicking what we used to do to clear a flooded engine on a carburetted engine. I think this is called "Clear Flood Mode".

Just an interesting tidbit (I think).

One other note. because in the WOT process described above we are defeating a device designed to improve auto emissions, whenever an automaker programs something like into their vehicles computer systems, they have to get advance EPA approval. Interesting, eh ? Device defeat exceptions is the name for this process as best I recall.
 




410Fortune

ICE Season
Staff member
Moderator
Joined
August 3, 2000
Messages
26,612
Reaction score
2,288
City, State
NORTH IDAHO, 7B
Year, Model & Trim Level
B2 "Slightly" Modified
Callsign
FOURTEN
Ah-HA! that is why I have always been told to hold the pedal to the floor during hard start situations where the engine may be flooded...always seemed backwards, but worked!
 




spindlecone

Explorer Addict
Joined
April 30, 2004
Messages
4,836
Reaction score
0
City, State
San Jose, Calif
Year, Model & Trim Level
2003 sport trac
410Fortune said:
Ah-HA! that is why I have always been told to hold the pedal to the floor during hard start situations where the engine may be flooded...always seemed backwards, but worked!
Ah the days of carbed flooded engines, vapor lock etc
Thx for the answeres to the Loop questions
 




ExplorerDMB

Moderator/Technician
Moderator Emeritus
Joined
January 26, 2004
Messages
6,162
Reaction score
8
City, State
Mechanicsville, Virginia
Year, Model & Trim Level
2004 Acura TL
I found these graphs:

31_o2_sensor_b1s1.gif


32_o2_sensor_b1s2.gif



O2 sensors should vary from .2 and .8 volts and should look somewhat like the first graph. Bank One Sensor One (the first graph) looks decent. The Bank One Sensor Two (second Graph) is ok - I wish I had a better one to choose from, but it should be a little tighter in variation. This, however, should give a rough estimate of what to expect if scoping/graphing O2 sensors out.

-Drew
 




spindlecone

Explorer Addict
Joined
April 30, 2004
Messages
4,836
Reaction score
0
City, State
San Jose, Calif
Year, Model & Trim Level
2003 sport trac
sort of explaines why inexpensive A/F ratio gauges are unreadable
 




Glacier991

EF Tranny Guru
Moderator Emeritus
Joined
February 8, 2003
Messages
9,832
Reaction score
33
City, State
Sacramento, CA 95827
Year, Model & Trim Level
1992 XLT
While I don't want this thread to morf into something too far afield (which is is not even in any danger of yet - and Thanks Drew for your graphs - I am still trying to count the crossovers in that 20 second graph). I have been asked how difficult is it to replace a sensor. The answer is, it depends.

First off, they are screwed into a manifold that alternately gets red hot and then cools. They tend to require some effort to break loose. [as a result of that question I added to the main thread showing tools]

Secondly, and most importantly, they can be placed in places awkward to get to. Forum member 5.0Dad did a beautiful reply to someone with a 5.0 V-8 having trouble. It is a short thread with excellent photos by 5.0Dad of how to get to certain O2 sensors on the 5.0. Here is a link:

http://www.explorerforum.com/forums/showthread.php?t=73275

[One final note about that above thread you just visited. You see in that thread where there is a complaint voiced that someone went to Autozone and got the code read and it said "slow O2 sensor" - then they bitched they had no clue which one of 4? Well therein lies the rub...he is dead wrong. DO YOUR HOMEWORK. The code definitions are available on the internet, and now with this forum there is a link to them. Contrary to what that member said (and may have been told at Autozone) Code P0153 DOES tell you EXACTLY which sensor is being focused on - and it is Bank 2 sensor 1. The guys at Autozone are not always the sharpest crayons in the box, but their free reads for NUMBER codes is invaluable and I salute them. I am not intending to rag on them. But... fer chrissakes when you GET a code read for you, get the code number and then, DO YOUR HOMEWORK. Don'y rely on the Autozone employee's interpretation - you don't have to, get the FULL story on that code's definition. - Thank you for your cooperation.

End of rant
 








spindlecone

Explorer Addict
Joined
April 30, 2004
Messages
4,836
Reaction score
0
City, State
San Jose, Calif
Year, Model & Trim Level
2003 sport trac
In replacing sensors, a sensor socket is an invaluable tool, if they are accesable as they are on my trac, my sons trac, my youngests eclipse, and the wifes car.
I have changed 4 sensors in the past year on the above cars using Bosch sensors.
The biggest PITA is that they are not plug and play, on the heated sensors you need to make 4 splices in sometimes cramped awkward places.
Each time I have used the crimp on male female, plug type connectors, will make the next time a snap, literally.
 




sparky2263

Well-Known Member
Joined
December 23, 2005
Messages
265
Reaction score
1
City, State
Melrose, Fl
Year, Model & Trim Level
99 Limited AWD 302
jtsmith said:
Why is a wideband O2 sensor different?

Our O2 sensors measure a very narrow range. On its 0-1 VDC scale a 14.7-1 A/F mixture will be around .5 VDC. .1 VDC will be about 13.7-1 and .9 VDC will report as about 15.5 -1. And that's being generous. Wideband sensors can measure a range of 8-1 all the way up to 20-1 (or more). They are a VERY handy tuning tool for performance applications.

Honda actually started using them in 1992. They found they could run A/F mixtures as high as 17-1 under certain cruise conditions. More and more mfgs. are turning to this technology as the limits of A/F ratios will be pushed further and further.

Some of the neater technologies include a prechamber type system where the spark ignites a small area of A/F which in turn will burn a very lean mixture (like 20/22-1). This is said to be capable of providing normal (?) power yet capable of VERY high cruise mileage figures.
 




BrooklynBay

Moderator & long time member.
Staff member
Moderator
Elite Explorer
Joined
November 11, 2005
Messages
54,925
Reaction score
611
City, State
Brooklyn, NY
Year, Model & Trim Level
88 89 93 95 96 Aerostars
I remember reading an interesting article in Gears magazine that gave a comparison of an O2 sensor to a fuel cell. It runs the exhaust gases throught it, and the catalyst (Zirconium I think) changed the gas into an electrical current. This is the reason why they need a specific operating temperature in order to function correctly.
 








Glacier991

EF Tranny Guru
Moderator Emeritus
Joined
February 8, 2003
Messages
9,832
Reaction score
33
City, State
Sacramento, CA 95827
Year, Model & Trim Level
1992 XLT
Excellent links! With the addition of those, this thread sort of becomes the "one stop shop" for everything you might need/want to know about 02 sensors.
 




dogfriend

Human-Animal Hybrid
Joined
July 22, 2002
Messages
7,566
Reaction score
7
City, State
Carmichael, CA
Year, Model & Trim Level
'97 Sport 4x4
Here is one more link that helps to explain why the O2 sensors need to be hot before they start to generate voltage:

http://en.wikipedia.org/wiki/Zirconium_dioxide

Stabilized zirconia is used in oxygen sensors and fuel cell membranes because it has a unique ability to allow oxygen ions to move freely through the crystal structure at high temperatures. This high ionic conductivity (and a low electronic conductivity) makes it one of the most useful electroceramics.

Zirconia is one of few compounds that actually becomes conductive at high temperatures, and more conductive, as its temperature increases. Zirconia starts out with a very high resistance at room temperature, greater than 1 trillion ohm-cm. As the temperature increases it has less than 20,000 ohm-cm at 500 degrees Celsius, to having less than 1,000 ohm-cm of resistance at 1,000 degrees Celsius. It loses nearly all of its resistance around 2,000 degrees Celsius, and becomes a very good conductor.

The O2 sensor has to heat up so that the ZrO2 becomes conductive. When conductive, the ZrO2 becomes a solid state electrolyte, allowing Oxygen ions to travel from the region of high O2 concentration (ambient air outside the exhaust) to the region of low O2 concentration (inside the exhaust pipe). The electrons have to travel through the wire to the PCM and back to keep the electrical charge neutral and the voltage (proportional to O2 concentration) tells the PCM if you are running lean or rich.
 




SIF

Member
Joined
April 26, 2005
Messages
29
Reaction score
0
City, State
UK
Year, Model & Trim Level
4.0l OHV in a TVR!!
So if the sensors are disconnected completely the PCM reverts to an internal tabular value. What are the real world results of running an engine in this state permanently.... as I guess this is how the 'old' injection systems used to work?
 


Join the Elite Explorers for $20 per year. Gets rid of the ads! New $5 per month "try out" option.

Explorer Forum has probably saved you that much already, and will continue to save you money as you learn how to diagnose fix problems yourself and learn which modifications work without having to experiment on your own. Elite Explorer members see no advertisements, no banner ads, no double underlined links, can add their own profile photo, upload photo attachments in all forums, and Media Gallery, create and save more private Conversations, and more. Join Today. Your support is greatly appreciated.




ExplorerDMB

Moderator/Technician
Moderator Emeritus
Joined
January 26, 2004
Messages
6,162
Reaction score
8
City, State
Mechanicsville, Virginia
Year, Model & Trim Level
2004 Acura TL
unplugging the 02 sensors will give you worse gas mileage. 02 sensors are there to optomize fuel economy.

-Drew
 




Top