Air/Fuel Mixture Meter

Some issue to consider. Most O2 sensors are non-linear, which means that the voltage they put out is not proportionate to the amount of O2 in the exhaust stream. As a result, almost all cars using standard-type O2 sensors use them as rich/lean switches rather than actually trying to measure the amount of exhaust O2. Another issue is that the measuring device need to have very high input impedence, otherwise there is a very real chance of damaging the sensor permanently.

You mentioned that it's a Honda? If it has the V-TEC engine then your O2 sensor is of a type referred to as a Wide Band O2 sensor. This kind of sensor is designed to measure O2 directly and within a fairly wide range of fuel/air mixtures. This kind of sensor requires special control circuitry because it's heated, among other reasons. It is a fairly advanced project to try designing for a wideband sensor, it would be much easier to just buy either a kit or an assembled unit.

Back to the regular sensor. Outside of the stochiometric ratio of 14.7:1 where the O2 sensor voltage is 450mV the voltage is not really meaningful.

JazzMan

learn

the

program

about

It's not a wideband sensor. It's the regular type on a non-vtec engine (D15B7).

Are there any good articles on Honda sensors or regular sensors in general?

Why doesn't the regular O2 sensor measure linearly? Also, would an A/F meter like the one I'm planning for a regular O2 sensors be of any use? I think it would definitely be an interesting project to work on even if it's not as accurate.

Also, what do you mean by "having a high impedance input" as far as the meter goes?

Thanks!

John

in

yellow...

be

to

general?

meter

it

Hmm.. high input impedance... I think I know what you mean. If the measuring device has low impedance, the current might flow towards the sensor instead of the metering device which might damage the sensor, is this what you mean?

Well, from my basic understanding of electronics, I was thinking I could grab a very small sample from the sensor signal and amplify it with a transistor at my meter side. So basically, I could place a high resistor value in parallel with the signal which would drive the transistor and by amplified for the meter to read and display. But then again, the transistor will require at least 0.7V in order to open up (I think saturate is the proper term here) which is pretty much what the O2 signal can provide. So how do I go about making this work without inversing the flow of current and damaging the O2 sensor?

Thanks!

John

expected

Yeah, it's making sense now. I think you said pretty much what I said the post after. So that's definitely a good thing :-)

can

high

Could you give a little more detail as to how a high impedance buffer works and what components/ideas are behind it?

Based on your description, I think that's exactly what I need. So basically I tap into the O2 sensor signal somewhere between the O2 and the ECU, and use that as the input to the buffer? As far as the display goes, how would I convert that tiny voltage/current(say 0.50mV) to display as 500 on a digital display? And what display would work for this meter/gauge gadget I'm working on?

Thanks!

John

John,

high input impedance means that as far as the sensor is concerned it sees a high resistance from whatever it is connected to. If you connect that O2 sensor to a low impedance circuit you reduce the output voltage, this can affect the operation of the cars ECU as that now sees a lower than expected voltage and so compensates by altering the fuel/air mixture.

There are several high input impedance circuits out there on the net you can use, you can buy high impedance single rail supply op-amps which would be ideal for this circuit. You would use one to buffer the circuit (and so prevent interfering with the cars ECU operation) and then another (normal op-amp) as a non-inverting amplifier to boost the signal before feeding it into the digital meter. Although you could also do this by butchering a high imput impedance DVM!

Chris

John

Try this link it gives a pretty good overview about op-amps and their uses:

The specific sections you want are voltage follower, difference amplifier and non-inverting amplifier. The voltage follower acts as a buffer and provides no signal amplification (called gain) or you can use a difference amplifier circuit. Which would give you a variable output depending on the difference in voltage between the two inputs (this can be an advantage in an electrically noisy environment such as a car as it means that any common noise signal in both the ground and signal wire gets ignored and your voltage reaching your display will be more stable).

The final part is to use a non inverting amplifier to boost (or multiply by the gain factor) the input signal: e.g for a gain of 6 your 0-2v sensor voltage would vary between 0 and 12V provided the op-amp can push the output to the power supply limits (not all of them can). You're probably better off staying with a 2V range and using a digital panel meter calibrated for a max reading of 2V then allmost all your resistors in the amplifiers will be the same value. It's better to keep the signal going to the display at a reasonable level rather than reducing to your 500mv as noise will tend to drown out the signal at that level (think of an old audio amplifier - the noise level is pretty much the same and at low volume levels it almost drowns out the music).

For the 3 Red, Green, Yellow LED's you can use three op-amp comparators all reading the same signal and just referenced to different voltages. So that op-amp 1 turns on at 0V op-amp2 at 0.2V etc or you can get a bar graph chip (LM3916 from National Semiconductor) and connect LED's only to 3 outputs, calibrating the chip so the LED's are at the switching points.

For any further circuits just do a google for op-amp schematics or have a look at a few manufacturers website for application notes, you'll then learn about decoupling capacitors and voltage regulators which you'll need as well.

As for specific devices to use everybody has their favourites; National, Maxim etc. and if I told you everything to do there'd be nothing for you to work out for yourself would there ;-) Half the fun with electronics is learning from your mistakes and learning to keep the magic smoke inside the components!

Chris

--FWIW there was a neat widget for sale in gag shops a few years back called the "Fart detector". It consisted of a battery powered heated grid; I think the technical term for this is a hot pipe anemometer. Anyway, when confronted by something other than air the cooling rate of the device would change; this condition caused a recording to play. Accompanied by siren whoops it would shout: "Warning! Fart detected!" I extended the probe wires and stuck one inside the combustion chamber of a potato gun, trying to sense when the mixture was right for firing. Results not great, but it made for an interesting project, heh. --The trick is not so much to detect a mixture, but to detect the *right* mixture. Still workin' on that one...

"John" schreef in bericht news: snipped-for-privacy@comcast.com...

Maybe this page is of interest to you.

Jeroen

learn

Excellent! Thank you!

That's what I was looking for--nice explanation with a schematic.

John

Here's a kit that I'm contemplating buying:

Comes highly recommended.

JazzMan

Here's a better question... Whats the point? A standard narrow band oxygen sensor is next to worthless for giving you mixture readings. They have part/part variation. They vary with age and temperature (which is constantly fluctuating). They are highly non-linear, and they exhibit hysteresis. In otherwords, they read "lean", "rich", and maybe if you are lucky stoichiametric. But under normal running conditions the ECU will keep the AFR bouncing between rich and lean. The only exceptions to this are when the car's cold (at which point the sensor isn't working) or you are at wide open throttle, in which case presumably you are running rich but you can't know how rich. Anyone who tells you differently is trying to blow smoke up your backside. As for your "device", why not just use a cheap digital multimeter? That should do exactly what you want, and it's probably cheaper, more versatile and better packaged.

dan