Bronco Forum - Full Size Ford Bronco Forum banner

1 - 7 of 7 Posts

·
Registered
Joined
·
295 Posts
Discussion Starter #1
I'm trying to diagnose a check engine light with no codes, and I am looking at the oxygen sensor now. I've got a snap on scanner/troubleshooting computer on it and set to data read mode, and as I'm testing it by driving it I have noticed that the O2 sensor voltage swings wildly from 2mV to 800mV and in between constantly. It does this even when rpm's and load are static.

Is this normal? Or should I get a new sensor?

Thanks

Sy
 

·
aka: kemicalburns
Joined
·
7,857 Posts
if your getting a CEL then there should be codes period. why not try the paper clip method to get your codes. www.fordfuelinjection.com has instructions and may also have the volt info your looking for.

also whats changed recently? any work done on the rig?
 

·
Satyr of the Midwest
Joined
·
17,734 Posts
I'm trying to diagnose a check engine light with no codes, and I am looking at the oxygen sensor now. I've got a snap on scanner/troubleshooting computer on it and set to data read mode, and as I'm testing it by driving it I have noticed that the O2 sensor voltage swings wildly from 2mV to 800mV and in between constantly. It does this even when rpm's and load are static.
That's how zirconia-based oxygen sensors work; they have to 'switch' about a stoichimetric point as a median to be an accurate representation of exhaust stream oxygen content.
 

·
Registered
Joined
·
295 Posts
Discussion Starter #4
if your getting a CEL then there should be codes period. why not try the paper clip method to get your codes. www.fordfuelinjection.com has instructions and may also have the volt info your looking for.

also whats changed recently? any work done on the rig?
I already fixed all errors and issues, and I now have a clear code status (111), so that's what I meant by no codes.

That's how zirconia-based oxygen sensors work; they have to 'switch' about a stoichimetric point as a median to be an accurate representation of exhaust stream oxygen content.
so you're saying that's how it's supposed to function? It doesn't seem logical, the reading is fluctuating every half second by anywhere from 100 to 400 percent. How does the computer get any reliable sense of that the air/fuel mixture is at any time with such a fluctuation?
 

·
Super Moderator
Joined
·
4,971 Posts
As far as I understand it, the only type of vehicle that is going to have a steady reading when you read the output of the O2 sensor is if you put one in a carb'd vehicle (like I have in mine). This is because the carb doesn't continually change it's fuel mixture (at idle), so the signal is steady.

.4 - .6 volts is right around stoich (proper fuel mixture). Less than that is lean and more than that is rich.

A narrowband O2 sensor like you have isn't very accurate. It basically tells if it's lean, stoich, rich but no real basis of HOW lean or rich it is, just THAT it is. A 14.7:1 air/fuel mixture is right around .5 volts. But even 14.5 or 14.3 can quickly jump that reading to around 0.8 - 0.9 volts, or a 14.9 or 15.1 mixture can drop it down to 0.3 - 0.1 volts. It doesn't take much change in air/fuel mixture to change the reading a LOT.

When the computer sees it lean, it richens it up. When it sees it's rich, it leans it out. It's almost like a guess and check, but when it does this multiple times per second, it gets pretty close in the long run.

So what you're seeing on your reading is the computer bouncing the fuel mixtures back and forth rapidly to try to hone in on that perfect mixture. I imagine it'd be pretty hard to read.
 

·
Satyr of the Midwest
Joined
·
17,734 Posts
:stupid

You're a human, not a µC, so trying to read and interpret that signal with your eyes is moot. The frequency of switching, and the duration it remains high or low allow the PCM to get a reliable feedback, albeit there is a great deal of calculation you're not seeing. It's not a 'stupid' sensor like an ECT/ACT or similar, in that a simple transfer function will tell you what the temperature, pressure, or incoming air mass is.

There are more factors here you'd need to research to thoroughly understand what's really going on -- oxygen sensor design, signals and systems, control theory, etc. -- but trust me, it's how the system works.

As an example, this is straight from Ford's own documentation, the LHBH1 strategy book:

OVERVIEW

The goals of the closed loop strategy are:

- add the capability of introducing A/F ratio biasing,

- maximize the feedback limit cycle frequency for all bias values, and

- maintain a simple calibration procedure to describe the closed loop limit
cycle.

The fuel flow is driven in a limit cycle manner about stoichiometry. Using
the EGO (Exhaust Gas Oxygen) sensor, the computer increases or decreases the
injector pulsewidths in a controlled manner. If the EGO reads rich, the
pulsewidths will be decreased (made leaner) at a calculated rate. If the EGO
reads lean, the pulsewidths will be increased (made richer) at a calculated
rate.

When an EGO switch occurs, an instantaneous change (or "jumpback") is made in
the A/F ratio back towards stoichiometry. The jump is made relative to the
A/F ratio (LAMBSE) value at the EGO switch.

The limit cycle can be biased to operate on the average richer or leaner of
stoichiometry.

An example of the closed loop limit cycle is shown on the next page.
LIMIT CYCLE DESCRIPTION

{Text diagram here of the closed-loop limit cycle and relevant HEGO signal}

***** NOTE *****

The direction of the bias is controlled by the sign of the bias value. If
the bias term is negative, a rich bias is indicated. If the sign of the bias
term is positive, a lean bias is indicated.

****************



***********************
***** WARNING *****
***************

It is imperative that an accurate value for the system transport delay be
entered. An incorrect value will result in greatly reduced catalyst
efficiencies due to excessively fast or slow ramp rates, incorrect jumpback
amounts, etc.

**************
***** WARNING *****
***********************
CALIBRATION PHILOSOPHY

Although appearing somewhat complicated, this closed loop algorithm has been
designed to be easy to use.

There are 3 tables which must be calibrated. They are:

1) limit cycle peak to peak amplitude; FN1354(N,MAP); a typical value is
0.034 lambdas (0.034 * 14.64 = 0.49776 A/F ratio) = PTPAMP

2) fuel system transport delay; FN1343(N,MAP); typical values have been 5 -
10 engine revolutions, but note earlier WARNING.

3) BIAS; FN1355(N,MAP) for closed throttle mode:

a) a positive bias value is lean; a negative bias value is rich

b) any calculated absolute value of BIAS/PTPAMP exceeding .2 is clipped
to .2. This is done to avoid extremely long limit cycle periods.
You can get the whole document here or here. Read up on the closed-loop LAMBSE calculation section (middle of Chapter 6, IIRC).
 

·
Registered
Joined
·
295 Posts
Discussion Starter #7
Sig & AB, I never thanked you guys for your responses... Sorry bout that. So thank you. I've since learned quite a bit more about this stuff, so I we ever get the Tuning/FI sub forum going maybe I can actually contribute something of value back.

Regards,

Sy
 
1 - 7 of 7 Posts
Top