Help me understand AFR

[snakebit]

I have heard/read various things about AFR and they seem to conflict. Anyone have the patience to try to straighten me out on this very important subject?

What I've heard:
1. The ideal stoichiometric AFR for air/gasoline is 14.7/1.
2. AFR for a FFSC coldside setup should be 11/1-12/1
2a. An AFR as high as 13/1 will melt pistons in a coldside SCed Miata.

My best guess is that, if pulling the data from the Miata ECU, it will show a very rich number like 11/1 under boost because the PC Pro is telling it to add lots of fuel to make up for all the air the SC is pumping in. If this is so, then the AFR as read by a tailpipe sniffer should still be close to 14.7/1?

Is the simplest/safest way to tune just to read the AFR from a sniffer and dial it in to as close to 14.7/1 under all conditions?

Am I anywhere in the ball park?:confused:

[Kyp J]

I have heard/read various things about AFR and they seem to conflict. Anyone have the patience to try to straighten me out on this very important subject?

What I've heard:
1. The ideal stoichiometric AFR for air/gasoline is 14.7/1.
2. AFR for a FFSC coldside setup should be 11/1-12/1
2a. An AFR as high as 13/1 will melt pistons in a coldside SCed Miata.

My best guess is that, if pulling the data from the Miata ECU, it will show a very rich number like 11/1 under boost because the PC Pro is telling it to add lots of fuel to make up for all the air the SC is pumping in. If this is so, then the AFR as read by a tailpipe sniffer should still be close to 14.7/1?

Is the simplest/safest way to tune just to read the AFR from a sniffer and dial it in to as close to 14.7/1 under all conditions?

Am I anywhere in the ball park?:confused:

In answer to the last line, no. There are better people to answer this than me, but basically the A/F ratio is sampled best just before the Cat by the O2 sensor. The ECM uses this signal to adjust the pulse rate to the injectors or it uses a pre determined map with adjustments with some other parameters (short and long term fuel trim...I just know the words, don't know details) if conditions aren't right for O2 sensor control. Wide open throttle, wrong engine temps and probably more.

Even the ECM uses non Stoich ratios under some conditions as noted above. When under boost, richer mixtures are necessary for reasons I won't elaborate on, I just take it on faith that the Forced induction people know this from past mistakes.

[socal pat]

If the ECU is seeing 11:1 then a sniffer at the tail pipe will also see 11:1 not 14.7:1. Lift the throttle, take the PC-Pro out of the equasion and both sensors will see ~14:1.

[FormerDatsun510Man]

The factory ECU doesn't see the actual AFR because it doesn't have a wideband O2 sensor input. What the factory ECU sees is a voltage of 0-1 V from a narrowband type O2 sensor that is located pre CAT. This type of sensor is really only accurate for telling the ECU when it is 14.7:1 +/- a small amount (hence the term "narrowband"). The purpose of this type of oxygen sensor is for emissions and fuel economy. When the AFR is close to stoich, 14.7:1, the engine runs the cleanest and most fuel efficient (on an aside, actually there are leaner mixtures possible with even better fuel efficiency, but certain emissions suffer sometimes from this). The narrowband O2 sensor reads .45 V for stoich and below this for lean and above this for rich. However, it is not very accurate for how rich or lean the mixture is because its purpose is to have a lot of accuracy right near stoich values. The reason for this is so that it can very accurately tell the ECU if it is running stoich. The reason this is important is so that the engine runs clean and efficient and passes the emissions requirements. Why doesn't the factory ECU have a wideband O2 sensor? Because this is an extra cost that is not necessary on a normally aspirated engine. On many factory turbo engines you will have things like wideband O2 sensors and MAP sensors, which really are necessary (as we have all found out with our boosted Miatas) for safe and efficient boosted operation.

Also, the factory ECU has two basic modes of operation, closed loop and open loop. Closed loop is when the factory ECU is continually checking the voltage value from that narrowband O2 sensor. During this time the factory ECU's goal is to maintain stoich, 14.7:1, AFR at all times. With a scantool hooked up you would see the voltage of the O2 sensor dithering between about .3 to .7 V. This is how the factory ECU maintains the AFR around stoich. What is happening is the factory ECU tries to target about .45 V but never quite gets it there so it will dither as a result. When the voltage is dithering from .3 to .7 V the AFR really is not changing that much... only a few tenths of a point. This is the value of a narrowband O2 sensor.... there is a lot of resolution right around the stoich AFR. The second mode of operation for the factory ECU is open loop. This is basically "power mode" of operation. It is reached under a certain set of conditions that vary by model year of the Miata, but basically is reached when the throttle is depressed past a certain position. The factory ECU knows this from the TPS (throttle position sensor) value. During this mode, the factory ECU ignores the narrowband O2 sensor and simply uses the map that it has burned on the chip for the fuel.

So, how does the Powercard Pro fit in all of this? It extends the pulses to the 4 fuel injectors POST factory ECU. I.e., it does this on the outputs of the factory ECU wires going to the 4 fuel injectors. The factory ECU thus doesn't "know" anything is going on. Furthermore, it does this only in OPEN LOOP mode of the factory ECU. How? Well, the neat thing is the Powercard Pro sends also sends a TPS value of 100% (full throttle) to the factory ECU when the Powercard Pro senses positive boost. This forces the factory ECU into open loop mode, which means it will ignore the narrowband O2 sensor values. During this time, if you adjusted your PC Pro correctly, the AFR reading with a tailpipe sniffer would hopefully be close to 12:1. However, as mentioned earlier here, the factory ECU doesn't really "know" anything about AFR. The narrowband O2 sensor during this time would read something around .9-.95 V, but this would be ignored by the factory ECU.

[FormerDatsun510Man]

Regarding tailpipe sniffers, what it would read depends on the conditions. As mentioned above, during closed loop mode, it would read very close to 14.7:1. Closed loop mode occurs during idling, cruising at a steady speed and most "normal" driving (well I mean "normal" in the general populace terms... :)). This is regardless of the settings on the PC Pro because the PC Pro will only add fuel during open loop mode, which is basically brisk acceleration and greater :). Once in open loop, the tailpipe sniffer will read an AFR that depends on the settings you have on your PC Pro. If adjusted correctly it will be close to about 12:1 AFR. Actually, with a Catalytic convertor in the mix the AFR will read about .5 leaner than it really is. In other words, a reading of 12.5:1 at the tailpipe sniffer is really closer to about 12:1. That is why some of us have welded bungs in the exhaust pre CAT to put the wideband O2 sensor into during testing so as to get a more accurate AFR reading during tuning on the dyno.

Bill

[Race Grandpa]

The only thing that I know something about is the A/F ratio, which I learned the hard way. Tom told me that the maximum A/F ratio as measured on the dyno should be 12:1. I lost an engine as the A/F ratio was set by the tuner on a Mustang dyno from 13:1 to as high as 14:1. When I get it set again I am going to tune it to be between 11 to 11.5. I understand that the Mustand dyno is more accurate than a Dynojet.

[Serpico]

Here's some more info on A/F ratios.

http://www.innovatemotorsports.com/resources/rich.php

[FormerDatsun510Man]

Race Grandpa, the low 11:1 is a bit rich for these applications... If you really want to err on the side of safety a/f wise, 11.5-12:1 range is plenty rich. Even in this range you will give up a small amount of power, maybe 5-10rwhp at the most, but this IMO at least is still not overly rich. With an 11:1 mixture you might start to have problems with fouling plugs, excess carbon buildup, CAT life etc. If you are having any signs of detonation with a 12:1 a/f ratio, this would indicate you either need to lower the compression ratio (not so easy) or retard the ignition timing (pretty easy), if there are no problems with engine or FI system.

I see Serpico has a link to a more technical explanation of this :).

Bill

[jwalton]

To the best of my understanding, the reason you don't want to burning at 14.7:1 even though that's the best ratio for a complete burn, is temperature. Your EGT's get very hot the more you lean out and that is not a good smog thing or a good thing for the exhaust valves etc. Running a little rich allows the extra fuel to absorb some of the compression and combustion heat. If you go too far, the cat gets used up because the unburnt (or continuing to burn) fuel goes in and messes with it. Also, running just a little rich helps alleviate pinging issues due to the same heat absorption issues. Timing will will take care of the rest of it.

[Tom @ Fast Forward]

Our dyno tests have shown that :

01-05 like 13: at start to ~3500 tapering to 12:1 at redline for best power.

99-00 like 11.5-12:1 for best power.

94-97 like 11-12 for best power. 94-97 don't seem to start losing power until well under 11.

[jikemenkins]

Our dyno tests have shown that :

01-05 like 13: at start to ~3500 tapering to 12:1 at redline for best power.

99-00 like 11.5-12:1 for best power.

94-97 like 11-12 for best power. 94-97 don't seem to start losing power until well under 11.

Interesting.

[tann3r]

Tom are the AFRs you listed from a tailpipe sniffer, ie recorded post cat? What AFRs would you tune for pre cat?

thanks

[snakebit]

Thanks, everyone, I am now much better equipped to set my AFR properly at the dyno. Also interested to know if Tom's recommendations are pre or post cat.
SC should get fired up this week!:taz: