3.6 air/fuel ratios

[rbogh901]

How can one interpret the oxygen sensor reading on a 3.6 to determine if you are running rich or lean?

My understanding is the following:

A rich mixture has less oxygen so the voltage signal is high in the range of .6 up to 1.0 volts.

A lean mixture has more oxygen so the voltage signal is low in the range of .4 down to .1 volts.


With low exhaust oxygen content, we will have high oxygen sensor output which will show as something above .45volts which indicates a rich running condition.

With high exhaust oxygen content, we will have low oxygen sensor output which will show as something below .45 volts which indicates a lead running condition.

From the oxygen content, and the resulting signals, the ECM can determine if the air/fuel ratio is rich or lean and it adjusts the fuel mixture accordingly...

[Steve@Rennsport]

Hi,

Oxysensor voltages are usually expressed as Lambda values and there are charts to show what each step means, relative to AFR values. I have one here that I'm planning to upload to our website soon.

Watching an OEM (narrow-band) O2 sensor work and deriving information from it can be difficult sometimes since Lambda numbers change rapidly when the O2 sensor is healthy. This is why we use wide-band O2 sensors with AFR meters which do not "cycle" and display results you can watch.

Some of these instruments can show either AFR values or Lambda voltages.

[rbogh901]

This is informative. My limited understanding is that lambda values have a normal range, since they switch back and forth so much, but it's not clear how you can use them to know if you are lean or rich. If charts are available it would, indeed, be useful. I'm surprised I haven't found them yet.

Searches have not really answered these questions. If someone has links, here is your chance to buck up.

[Raceboy]

Narrow-band lambda is completely useless as a tuning tool for any engine operating range other than tuning idle and/or very light load where lambda values should be at or around stoich (lambda 1.0 = 14.7 on gasoline).

And most 911 engines like a bit richer idle (creates better vacuum), like 14.0 (lambda ~0.95), especially when dealing with hotter cams and/or ITB's.

At WOT the sensor reads just "rich" and does not give any information how rich. Of course, if your lambda sensor reads "lean" and WOT, it means either trouble with the fuel delivery or too rich (rich misfires read lean on lambda) and still, narrowband lambda is useless.

Its only purpose (and what OEM meant it to be) is to correct mixture at idle and low loads reducing pollution.

[Steve@Rennsport]

Originally Posted by rbogh901

This is informative. My limited understanding is that lambda values have a normal range, since they switch back and forth so much, but it's not clear how you can use them to know if you are lean or rich. If charts are available it would, indeed, be useful. I'm surprised I haven't found them yet.

A healthy O2 sensor's readings change/cycle so fast that its almost, if not virtually impossible to note the Lambda numbers to see precisely what you have.

This is exactly why we use an addition wide-band sensor & AFR meter to actually see what the values are in order to make changes and monitor things.

You can find what you are looking for here:http://www.tatech.fi/cms/uploads/Lambda_vs_AFR.pdf

[rbogh901]

Originally Posted by Steve@Rennsport

A healthy O2 sensor's readings change/cycle so fast that its almost, if not virtually impossible to note the Lambda numbers to see precisely what you have.

You can find what you are looking for here:http://www.tatech.fi/cms/uploads/Lambda_vs_AFR.pdf

Thanks for the chart.
I think the last time I checked my lambda ranged from .92 (rich) to .96 (less rich) but those numbers are just the ones stored by the OBD and read with the Scantool software. Do they have any useful value?

[Steve@Rennsport]

Originally Posted by rbogh901

Thanks for the chart.
I think the last time I checked my lambda ranged from .92 (rich) to .96 (less rich) but those numbers are just the ones stored by the OBD and read with the Scantool software. Do they have any useful value?

Not like the data you get using a wide band unit.

Get yourself an Innovate AFR meter so you have good, accurate info.

[Walt Fricke]

Listen to Steve

The stock sensors and system produce a curve which basically has three slopes when it has voltage on the vertical axis, and AFR (or Lambda) on the other. It starts out a nice shallow angle, with modest changes in AFR for voltage changes. And at the other end, it is also at a nice shallow angle. However, in the area of interest between say 12 and 15 or so, it is very steep - there is just a very small voltage change between 12 and 15. So you don't really know what is going on, other than are you pretty lean or are you pretty rich.

At least for the CIS system using Lambda (maybe the 3.2s also?), this didn't matter to Porsche. If the sensor reads lean, the computer adds more fuel. On the 3.0s this is via the frequency valve in the WUR circuit. When it reads rich as a result, it feeds less fuel. The result is a kind of dynamic average for best part throttle fuel economy. This back and forth is by design. Works just fine to produce the fuel efficiency Porsche wanted. But not for tuning race motors, or home brew EFI systems.

As a result, we fool ourselves if we think we can use a stock sensor and a voltmeter (or an array of colored LEDs) for tuning of any sensitivity, if at all.

[TimT]

The bottom graph in the attached pic shows the output curve of a narrow band sensor (note the narrow band) Interpreting useful information from a curve like this is difficult, the slope of the curve is nearly vertical for a bit....good for keeping the catalyst burning.... not so good for tuning...




Wide band sensors typically put out a more easy to interpret nearly linear graph.



Most quality WB systems have a NB out

[porterdog]

Originally Posted by Walt Fricke

The result is a kind of dynamic average for best part throttle fuel economy.

Respectfully, I think this is incorrect. As the then-lead aftermarket Mustang powertrain calibrator for Roush explained it to me, you can "think of the catalyst as an oxygen bucket." Catalyzing HC requires the cat to give up oxygen, and catalyzing NOx requires the cat to accept oxygen. Obviously, if the bucket were full or empty, one of these two functions would be impossible. "Dithering about stoich," which is what the ECU tries to do during cruise, is all about keeping the cat neither full nor empty, for emissions control purposes.

$0.02

[Raceboy]

Yeah, that's what I stated already above:

Originally Posted by Raceboy

Its only purpose (and what OEM meant it to be) is to correct mixture at idle and low loads reducing pollution.

[Walt Fricke]

$0.02 undervalues the post. At least 0.03! Perhaps the fact that the dithering also keeps the A/F average ratio where it is wanted for fuel efficiency is a happy convergence of two requirements? Your explanation of the pollution effect is really clear (even if borrowed) and I may even be able to remember it.

The point here is the dithering points to the fact that the narrow band sensor isn't useful as a data source for tuning.

Just for grins, does anyone know what kind of O2 sensor the later, water cooled, 911s have? Wide or narrow? And what general system it plays a role in? Does it, too, dither things to make the cat work properly for a reasonable number of miles?

Aftermarket EFI systems often can use data from a wide band O2 sensor to adjust the fuel map. Of course, with these systems I suspect few care about emissions, unless occasionally for when they have to get tested by the state if street driven.

[Raceboy]

Newer watercooled 911's use Bosch LSU wideband lambda sensors and on part loads still run to reduce emissions, but on higher loads rund rich to reduce the possibility of engine failures (=warranty).

When I tune standalones, I always make the part throttle load and idle as good as possible with fuel economy and pollutions in mind but there is always a compromise (best economy doesn't necessarily mean cleanest exhaust gases).

We have annual MOT that includes measuring emissions so street cars have to pass the test.

[psalt]

Perhaps the fact that the dithering also keeps the A/F average ratio where it is wanted for fuel efficiency is a happy convergence of two requirements? Your explanation of the pollution effect is really clear (even if borrowed) and I may even be able to remember it.

Walt,

Best fuel efficiency is around 16:1 AFR. Stoich 14.7:1 is all about converter life and the best compromise of reducing the legally ordained "evil" pollutants (and is mandated by law). GM certainly knows this, it used to put cruise tables above 14.7:1 in the ECU maps and was sued by the Federal government. Just like Cadillac used a separate (leaner warm up) ECU program to meet the IM240 test with the climate control system OFF and instructed the customers to never turn it off when operating the car. On a non cat aftermarket EFI conversion, you can get much better mpg by tuning AFR above 14.7 under low load cruise with ignition advance out near 50 BTDC. Many aftermarket ECUs allow you to shift the narrow band "dither" target higher than 14.7:1. I have done EFI conversions on 50's and 60's land cruisers and mpg can be improved +50% over 4 BBL or SU carburators. In other words, we have chosen to burn more fuel in everyday cruising because of the political decisions made in the name of the environment. Just like we are choosing to ship massive amounts of oil by rail, which burns even more fuel, to satisfy a few eco nazis that failed math class. Or the ethanol mandate. Politics and taxes drive these decisions, not engineering.

[rbogh901]

Originally Posted by psalt

Best fuel efficiency is around 16:1 AFR. Stoich 14.7:1 is all about converter life and the best compromise of reducing the legally ordained "evil" pollutants (and is mandated by law). GM certainly knows this, it used to put cruise tables above 14.7:1 in the ECU maps and was sued by the Federal government. Just like Cadillac used a separate (leaner warm up) ECU program to meet the IM240 test with the climate control system OFF and instructed the customers to never turn it off when operating the car. On a non cat aftermarket EFI conversion, you can get much better mpg by tuning AFR above 14.7 under low load cruise with ignition advance out near 50 BTDC. Many aftermarket ECUs allow you to shift the narrow band "dither" target higher than 14.7:1. I have done EFI conversions on 50's and 60's land cruisers and mpg can be improved +50% over 4 BBL or SU carburators. In other words, we have chosen to burn more fuel in everyday cruising because of the political decisions made in the name of the environment. Just like we are choosing to ship massive amounts of oil by rail, which burns even more fuel, to satisfy a few eco nazis that failed math class. Or the ethanol mandate. Politics and taxes drive these decisions, not engineering.

Very interesting read here. Thanks guys.
Don't get me started on the ethanol debacle.