O2 Sensors & AFRs

[Dan D.]

A couple of weeks ago Engine Masters did a segment on Air Fuel Ratio meters. They evaluated 5 different brands. They welded several bungs into the head pipes of their test engine so they could all be tested together. The results of their testing:

1. All 5 tested within a 1/2 point of each other, so all 5 were deemed to be acceptable. No way of knowing what the actual/correct stoichiometric ratio was. This was just comparison testing.

2. They moved the sensors 2 feet back on the head pipe and the change from the original location was negotiable, so they determined that the sensors do not have to be right at the manifold. This requirement on new cars is to reduce light-off time, although they did not suggest this.

3. But placing the sensor in the tailpipe made a huge change. The readings were much leaner, especially at idle, suggesting there was additional room air in the pipe that was being read by the sensor. But I know some members here on the forum use the tailpipe method when tuning engines. I do not see any mention of tail pipe on any of the internet sites - they all say to weld in bungs.

4. There was never any mention of the sensors being heated or not, but I would think that they all would be. I think most are heated now. Again to reduce light-off time.

5. I recorded this show. My FI unit is running rich (not the CSV) and I am thinking about buying an AFR and was going to use the tail pipe method, just to not to have to remove the head pipes, which could be difficult because of the clamps compressing the joints. But it looks like I need to weld in the bungs.

6. So I was just wondering if anyone has any experience/opinions on this - or saw the show.

Dan

[Duke W.]

Re: O2 Sensors & AFRs

Most chassis dyno operators offer wide band A/F ratio measurement via a probe stuck up the tail pipe. I don't have any data on Rochester FI engines, but typical carbureted engines on WOT runs from off idle to redline will be overrich (<12.5:1) at low revs and lean out as revs increase. If it doesn't lean out beyond 13.5:1, then I say leave well enough alone.

In a recent dyno test of a '64 Special 300 HP configuration the OE unmodified AFB did an amazingly good job of maintaining a correct WOT air/fuel ratio for such a crude device compared to modern fuel management systems.

For cruise A/F ratio measurement on an inertia dyno, like a Dynojet, start out at 1500 and slowly work the throttle to accelerate to 3500 to 4000 maintaining manifold vacuum in the range of about 16-12".

The nice thing about Rochester FI is than you can easily adjust the cruise and WOT mixtures by adjusting the rich and lean stops on the fuel meter.

Another way to adjust the FI cruise mixture is to test on a level stretch of road at say 35 to 65 MPH or greater and lean the lean stop in 1/8 turn increments until there is some lean surge or hesitation upon light acceleration, then go back one increment. A good FI system with nozzles that all flow within less than a 5 percent range should be okay with a cruise A/F ratio of 15.5:1.

With a little practice so you know how to make adjustments fast, one hour on a chassis dyno will likely be less expensive and take less time to optimize the fuel flow map than a permanent O2 sensor, and no need to make permanent modifications.

BTW, on engines with full time exhaust air injection like a lot of engines from the sixties and seventies the measured A/F ratios will not be accurate unless the air injection is diverted or disabled. Modern engines with air pumps, often electrically driven, only operate when the engine is warming up prior to entering closed loop operation.

Modern cars with O2 sensors both ahead and aft of the catalyst essentially monitor catalyst performance, and if the aft sensor doesn't detect enough oxygen consumption by the catalyst, you get a CEL and codes.

Duke

[Jim L.]

Re: O2 Sensors &amp; AFRs

Dan,
I use an Innovate tailpipe attachment to mount the O2 sensor for my AFR measuring equipment. I also have a test engine with O2 bungs in the exhaust pipes.

As manufactured, the Innovate tailpipe attachment does not penetrate the exhaust stream far enough to avoid exhaust dilution from ambient air. This makes any AFR readings at idle inaccurate. However, at highway cruise speeds and during WOT runs, exhaust gas velocity is high enough that no noticeable dilution takes place and accurate readings are possible.

This idle mixture dilution isn't really a big deal since you don't set your idle AFR by gauge anyway. You adjust the mixture screw for the highest and best idle, letting the AFR be what it will be.

However, it bugs me that idle AFR readings aren't reliable so I added an extension to the Innovate thingy and the result looks like this:

4064.jpg

Even with this extension, there must be a little bit of exhaust dilution taking place. An FI unit which needs ~12.5:1 on my test engine to idle happily will register ~13.9:1 when it's on my car with the tailpipe sniffer doing the sampling.

In any event, if you do buy AFR measuring equipment, I suggest the following:

To get an accurate cruise mixture reading, be traveling as fast as the law will permit and, if at all possible, be on a very, very slight upgrade. What you are achieving is opening the throttle plate far enough to shut down all idle boost so you are running on only the venturi signal BUT you don't want the grade to be so steep that the enrichment diaphragm transitions the ratio lever to the power stop. The original target cruise AFR for Rochester FI was specified at 15.5:1. To allow for some measurement error, my target is 15.3:1

You can get decent WOT readings by starting ~40 MPH in 4th gear and accelerating for as long as you have the nerve to keep going. Years back, Jerry Bramlett did a lot of experimenting with air/fuel ratios and discovered that FI engines make the best power with an AFR of 13.1:1. Again allowing for measurement error, my target is around 12.9:1

Good luck,

Jim

[George J.]

Re: O2 Sensors &amp; AFRs

Jim's advise is spot on. I have had an Innovate LM-2 for about ten years. Back when I got it, I couldn't really get what I thought were good readings with my FI sidepipe '65. I put it away until this year, when I needed to recalibrate my unit after a rebuild. I added the extension Jim recommended and it works great, now. Not only that, but I used the OBDII reader feature to diagnose my 1998 M3, which needed a cam sensor replaced. Between not needing to go to a dyno, and not having to pay for diagnostic and repair of the M3, the unit has more than paid for itself.
Highly recommended.

George

[Dan D.]

Re: O2 Sensors &amp; AFRs

Dan,
I use an Innovate tailpipe attachment to mount the O2 sensor for my AFR measuring equipment. I also have a test engine with O2 bungs in the exhaust pipes.

As manufactured, the Innovate tailpipe attachment does not penetrate the exhaust stream far enough to avoid exhaust dilution from ambient air. This makes any AFR readings at idle inaccurate. However, at highway cruise speeds and during WOT runs, exhaust gas velocity is high enough that no noticeable dilution takes place and accurate readings are possible.

This idle mixture dilution isn't really a big deal since you don't set your idle AFR by gauge anyway. You adjust the mixture screw for the highest and best idle, letting the AFR be what it will be.

However, it bugs me that idle AFR readings aren't reliable so I added an extension to the Innovate thingy and the result looks like this:

[ATTACH=CONFIG]118996[/ATTACH]

Even with this extension, there must be a little bit of exhaust dilution taking place. An FI unit which needs ~12.5:1 on my test engine to idle happily will register ~13.9:1 when it's on my car with the tailpipe sniffer doing the sampling.

In any event, if you do buy AFR measuring equipment, I suggest the following:

To get an accurate cruise mixture reading, be traveling as fast as the law will permit and, if at all possible, be on a very, very slight upgrade. What you are achieving is opening the throttle plate far enough to shut down all idle boost so you are running on only the venturi signal BUT you don't want the grade to be so steep that the enrichment diaphragm transitions the ratio lever to the power stop. The original target cruise AFR for Rochester FI was specified at 15.5:1. To allow for some measurement error, my target is 15.3:1

You can get decent WOT readings by starting ~40 MPH in 4th gear and accelerating for as long as you have the nerve to keep going. Years back, Jerry Bramlett did a lot of experimenting with air/fuel ratios and discovered that FI engines make the best power with an AFR of 13.1:1. Again allowing for measurement error, my target is around 12.9:1

Good luck,

Jim

Hi Jim,

Thank you very much (and Duke too) for your excellent write-ups. I just have a couple of questions if you don't mind:

Is your sensor heated? I would think yes - I don't see how the tailpipe method would work if it wasn't heated.

Do you see any difference(s) in readings between the right and left banks?

I have the crossover pipe on my car. Do you think I could use this with a bung? I would think yes, but I don't know how much air flow there is in this pipe - its getting equal exhaust pressure from both banks. Lot easier to remove and add a bung to the crossover than it would be to add it to the head pipes. Welding it in is not a problem. Removing the head pipe would be the challenge for me. (I'm 82).

The photo of your tailpipe extension mod is a little confusing to me. Could explain it little bit, perhaps a different photo?

Thanks again Jim for your education and advise.

Dan D.

[Jim L.]

Re: O2 Sensors &amp; AFRs

Hi Jim,

Thank you very much (and Duke too) for your excellent write-ups. I just have a couple of questions if you don't mind:

Is your sensor heated? I would think yes - I don't see how the tailpipe method would work if it wasn't heated.

Do you see any difference(s) in readings between the right and left banks?

I have the crossover pipe on my car. Do you think I could use this with a bung? I would think yes, but I don't know how much air flow there is in this pipe - its getting equal exhaust pressure from both banks. Lot easier to remove and add a bung to the crossover than it would be to add it to the head pipes. Welding it in is not a problem. Removing the head pipe would be the challenge for me. (I'm 82).

The photo of your tailpipe extension mod is a little confusing to me. Could explain it little bit, perhaps a different photo?

Thanks again Jim for your education and advise.

Dan D.

Dan,

AFAIK all sensors are heated out of necessity. The sensor temperature has to be well regulated by the controller in order for the sensor to provide accurate O2 measurements. And it's possible to get the sensor too hot. When you buy Innovate equipment, it comes with a warning to NOT place the sensor too close to the exhaust manifold because the exhaust heat can increase sensor temp beyond what the controller can regulate.

As far as L/R bank differences, I have only one sensor for my test car and it's in the right bank. Looking at spark plug color, I don't see any L/R differences, and I shouldn't since FI nozzles were factory matched to within +/- 2.5% of nominal.

Here is what the basic Innovate tailpipe sniffer thingy looks like:

72-06.jpg

To it I added ~15" of soft copper tubing I got at a hardware store. A small worm clamp joins the copper tube to the Innovate thingy. Being soft copper, the tube can be bent/formed to match the exhaust pipe curvature of many different cars.

The one thing about the Innovate sniffer is that the thumb screw is totally inadequate. Exhaust system vibration will cause it to loosen every single time.

I use a different method of retaining the sniffer:

7291-re.jpg

Regarding putting a bung in the crossover pipe, I've never tried it but my intuition is that it wouldn't work well. That O2 sensor depends on exhaust flow by it. Given roughly equal pressure at the L/R exhaust pipes, I seriously doubt you'd get any significant exhaust flow by the sensor.

Does this help or are things still as clear as mud?

Jim