TomoHawk

Legitimacy of AFR gauges

    Recommended Posts

    I have an L28E with L-Jet EFI.

    I was thinking of building an AFR (air-fuel ratio) gauge using some common electronics stuff.   It lights up LEDs from a signal generated by a narrow-band oxygen sensor in the exhaust pipe.

    But, I question the legitimacy and accuracy of the gauge.     How does oxygen in the exhaust tell you how much air and fuel is going into the cylinders?  There is no sensor that will do that, and it would go into the intake manifold, not the exhaust pipe.  The best you can do is to use a modern engine, with a programmable controller to measure the air (via the AFM) and calculate the amount of fuel inject from the fuel pressure and the injection timing, then calculate the AFR. Measuring the oxygen in the exhaust does not tell you what is going into the cylinders.  The best you can do with that method is to guess at what the AFR is, with presumptions of engine perfection (you would need an engine that is perfectly built, in perfect operating condition, producing a 100% perfect combustion of what did go into the cylinder.)

    So using one of the so-called AFR meters is  a bunch of hooey, and the value it gives you cannot be trusted.  Using a narrow-band oxygen sensor, the signal will jump around, from a relatively lean (say 15:1) to a relatively rich (about 12:1) mixture- IF you believe the theory of these sensors. It's logical to say that the amount of oxygen detected by the sensor, is proportional to an amount f air, but it cannot tell you the AFR.  It only tells you there is oxygen/air in the exhaust.   

    I have read a lot about wide-band sensors, but I'd say it is the same as any other oxygen sensor. 

    The AFR display I will be building, if you are electronically-skilled, is based on the LM3814 LED driver chip, and lights LEDs, linearly, in proportion to the signal.  And the best you can do is to adjust the mixture so the display moves around enough to get an "average." then do some on-road or dynomometer testing to get your desired mixture.  Or, you could use an exhaut gas analyzer to adjust the mixture based on theCO content, whic is similar to the oxygen content adjustments.

    O2 Sensor Display Circuit.jpg

    Capture.PNG

    Share this post


    Link to post
    Share on other sites

    Tjereis no "calibration" needed.  You install the oxygen sensor in the exhaust pipe and hook it up to the gauge.

    Share this post


    Link to post
    Share on other sites

    I thought about building my own system and just using a look-up table, a real paper one, to mark up a voltmeter.  

    Not sure where "legitimacy" comes in to the conversation.  It's just a number from a sensor, some basic scientific principles, and some math.

    Share this post


    Link to post
    Share on other sites

    The problem is that an oxygen sensor in the exhaust stream can't tell you what the fuel mixture in the intake manifold is, precisely.  You make a LOT of presumptions, and people accept what the gauge says as an exact, scientific answer.

     You can buy pre-built displays cheaply from electronic kit companies, but it's just as easy to build one yourself, if you have the skill. Because the  signal can change quickly, you would be better off color-coding the voltmeter, for rich, about-right, and lean.

    Share this post


    Link to post
    Share on other sites
    6 minutes ago, TomoHawk said:

     people accept what the gauge says as an exact, scientific answer.

     you would be better off color-coding the voltmeter, for rich, about-right, and lean.

    I don't.

    That's what I was going to do.

    Then realized it didn't matter for my plans.  I only have to pass emissions every two years and my engine is unmodified.

    • Like 1

    Share this post


    Link to post
    Share on other sites
    23 hours ago, TomoHawk said:

    Using a narrow-band oxygen sensor, the signal will jump around, from a relatively lean (say 15:1) to a relatively rich (about 12:1) mixture- IF you believe the theory of these sensors. It's logical to say that the amount of oxygen detected by the sensor, is proportional to an amount f air, but it cannot tell you the AFR.  It only tells you there is oxygen/air in the exhaust.   

    I'm no expert on O2 sensors, but I don't believe the output will be jumping around unless the mixture is jumping around. When used in the control loop in a car, the only reason the O2 sensor output swings back and forth from rich to lean is because the controller is adjusting the mixture and making the output swing from rich to lean.

    If you hang that 3914 based meter on your car and you're running lean, it'll always show lean. It won't be flipping around.

     

    Share this post


    Link to post
    Share on other sites

    My A/F meter flips around when it's first started and the engine is cold. Don't know if it's because the sensor hasn't warmed up or if it's due to the mix being irregular because the intake runners haven't warmed up. Or maybe both. Once the engine has run about 15 seconds the meter is steady and the numbers on the gauge respond instantly with adjustment of the choke lever. Kind of handy (with the SU's) to avoid running too rich or lean at startup. Need to adjust the lever several times to keep AFR where I want it until it warms up.

    Noticed it also responds to stuff I don't think of as fuel. The AT modulator was bad, leaking ATF into the balance tube occasionally until I replaced it last month. I was taking a drive and saw the gauge suddenly drop to 10. Looked in the rearview and saw a cloud of blue smoke. Guess the engine had just sucked in a gob.

     

    Share this post


    Link to post
    Share on other sites
    10 hours ago, Stanley said:

    My A/F meter flips around when it's first started and the engine is cold. Don't know if it's because the sensor hasn't warmed up or if it's due to the mix being irregular because the intake runners haven't warmed up. Or maybe both. Once the engine has run about 15 seconds the meter is steady and the numbers on the gauge respond instantly with adjustment of the choke lever. Kind of handy (with the SU's) to avoid running too rich or lean at startup. Need to adjust the lever several times to keep AFR where I want it until it warms up.

    Noticed it also responds to stuff I don't think of as fuel. The AT modulator was bad, leaking ATF into the balance tube occasionally until I replaced it last month. I was taking a drive and saw the gauge suddenly drop to 10. Looked in the rearview and saw a cloud of blue smoke. Guess the engine had just sucked in a gob.

     

    Yes the heater needs to heat up the sensor so the initial readings might be off. Containment’s will ruin a sensor . Coolant from a HG failure will do it for sure, but oil will ruin the sensor too. 

    Share this post


    Link to post
    Share on other sites
    10 hours ago, Stanley said:

    My A/F meter flips around when it's first started and the engine is cold. Don't know if it's because the sensor hasn't warmed up or if it's due to the mix being irregular because the intake runners haven't warmed up. Or maybe both.

    Yeah, I'm not surprised the first couple seconds are unreliable. After that though, it sounds like it tracks the mixture very nicely. That's cool. I wish I had something like that installed.

    But back to the point I was trying to make originally... People talk about the mixture oscillating quickly between rich and lean. I propose that the only reason that's occurring is the control system is MAKING that occur. And that's because that's the only way they can measure the mixture. They are using a narrow band sensor and shooting for "an average" mixture at stoichiometric.

    If you stick an O2 sensor in the Z's exhaust stream and hold the engine condition steady, then you will measure a steady voltage out of the sensor. It won't be flipping around. It will probably be pegged rich, or pegged lean, but it'll be steady.   LOL 

    Share this post


    Link to post
    Share on other sites

    That's the advantage of a simple display like this:  to see if there is a pattern in the exhaust for the different throttle settings.  It should be an overall steady value until you press the throttle pedal enough to operate the TPS, then you ought to see the display change to a slightly lean indication for mid-throttle, then richer for an open throttle, then back to rich at idle.  If your AFM and ECU are operating correctly, the display should be fairly steady otherwise.

    Share this post


    Link to post
    Share on other sites

    I think what I'd like to see is scientific PROOF that you can quantify the FR in the intake manifold be measuring the oxygen in the exhaust.  Just because lots of people use it (either the narrow-or wide-ban sesor and AFR gauge) doesn't mean it works and is accurate and recise. 

    Share this post


    Link to post
    Share on other sites

    6 hours ago, TomoHawk said:

    I think what I'd like to see is scientific PROOF that you can quantify the FR in the intake manifold be measuring the oxygen in the exhaust.  Just because lots of people use it (either the narrow-or wide-ban sesor and AFR gauge) doesn't mean it works and is accurate and recise. 

    All modern closed loop systems are built around O2 sensors. The fueling is modulated by the ECU based on the O2 sensors. If it didn't work and was unreliable than it wouldn't be the default engine model. Everything from Ford to Ferrari to Bugatti do it this way...unless I am mistaken.

    Share this post


    Link to post
    Share on other sites
    11 hours ago, TomoHawk said:

     Just because lots of people use it (either the narrow-or wide-ban sesor and AFR gauge) doesn't mean it works and is accurate and recise. 

    Lots of people? How about every modern car manufacturer. Its what passes emission standards and gives you modern gas mileage values. Because it works.

    • Like 1

    Share this post


    Link to post
    Share on other sites

    The point is that people claim the sensor/gauge combination give an accurate and precise  measurement, which it can't.  It's based in faulty presumptions, and the engine is controlled by high-speed electronics to get an "average" that just happens to be acceptable.

    You might as well measure the AFR from the temperature of the tyres. Or, you might as well make the claim of a "clean-burning fuel,"  when the sole purpose of the internal-combustion engine is to burn fuel to create carbon dioxide and water, and nitrogen compounds.

    Until a sensor that directly measures the air-to-fuel relationship in the intake manifold, you cannot claim to measure the air-to-fuel ratio by the waste gases of combustion.

    Edited by TomoHawk

    Share this post


    Link to post
    Share on other sites
    18 minutes ago, TomoHawk said:

    you cannot claim to measure the air-to-fuel ratio by the waste gases of combustion

    Try explaining that to the state of California.

    Share this post


    Link to post
    Share on other sites

    Every sensor anyone has built, or will ever build, to read any physical property, does not read 'perfectly' with no errors in accuracy or precision. They almost all, regardless of what they measure or how they do it, as based on a scientific response models that connect, for example, how resistance changes occur with varying temperature. It ain't perfect. They are designed to deliver a specified accuracy and repeatability. We as consumers do not see this data when we buy a gauge.

     They are however, calibrated to internationally accepted standards (which also have errors, but much smaller ones) that everyone who builds this type of sensor is required to calibrate their sensors to, so that they can be responsible and accountable engineering businesses. Some comply, some don't. Companies that buy these sensors tend to buy from trusted sources. You can't as GM or Nissan, afford to buy whatever sensors you like from the cheapest vendor when they lay at the heart of how the engine runs.

    Nothing you build at home will even remotely approach the precision, accuracy and repeatability or response time of what commecial O2 sensors and their related electronics now provide. It is a wonderful experience to try to do better, you will learn a great deal while you do it. 

    Edited by zKars
    • Like 1

    Share this post


    Link to post
    Share on other sites

    I agree with that, for the most part.  It's just that people buy a narrow-band, or wide band gauge and the corresponding OXYGEN sensor, and set it up, thinking they will get Air-fuel mixture readings from the exhaust gases.

    Share this post


    Link to post
    Share on other sites

    Well, since you are infinitely knowledgeable in gas analysis, explain, exactly, how you convert an oxygen signal from a narrow-band sensor to an air-fuel ratio.

    Share this post


    Link to post
    Share on other sites

    In theory, what you do exactly is measure the voltage generated by the sensor, look it up in a table to find the experimentally determined and theoreticaly backed o2 concentration that corresponds to and then look that up in another table to find what the AFR would have been in the chamber to result in that residual o2 concentration in the exhaust.
    That o2 concentration to AFR table is also experimentally determined and theoretically backed.
    The experiments show that, consistently enough, for a given AFR in the chamber, o2 concentration in the exhaust will be constant regardless of other variables.

    For a narrow band sensor, its responses are only trustworthy in a narrow range either side of lambda=1 or stoichiometric AFR. Outside that it can only tell you lean or rich.
    A wideband sensor is accurate in a wider band of AF ratios, around 10:1 to 20:1.

    An odd first post I know.

    • Like 3

    Share this post


    Link to post
    Share on other sites

    3 hours ago, jonbill said:

    In theory, what you do exactly is measure the voltage generated by the sensor, look it up in a table to find the experimentally determined and theoreticaly backed o2 concentration that corresponds to and then look that up in another table to find what the AFR would have been in the chamber to result in that residual o2 concentration in the exhaust.
    That o2 concentration to AFR table is also experimentally determined and theoretically backed.
    The experiments show that, consistently enough, for a given AFR in the chamber, o2 concentration in the exhaust will be constant regardless of other variables.

    For a narrow band sensor, its responses are only trustworthy in a narrow range either side of lambda=1 or stoichiometric AFR. Outside that it can only tell you lean or rich.
    A wideband sensor is accurate in a wider band of AF ratios, around 10:1 to 20:1.

    An odd first post I know.

    But a good first post!

    • Like 1

    Share this post


    Link to post
    Share on other sites
    5 hours ago, jonbill said:

    In theory...

    That theory is well known, but presumes you have an engine that is working 100% perfectly and you have a 100% perfect combustion.

    Edited by TomoHawk

    Share this post


    Link to post
    Share on other sites

    Create an account or sign in to comment

    You need to be a member in order to leave a comment

    Create an account

    Sign up for a new account in our community. It's easy!

    Register a new account

    Sign in

    Already have an account? Sign in here.

    Sign In Now