Question on scanning narrow band

[87gmc]

Ok i know the narrow bank sensors are not accurate when scanning wot.

How does the narrow band work at part throttle and idle. I'm trying to understand the switching back and fourth. How does the computer actually know its at 14.7 if the narrow bands are not accurate. Can someone please explain.

The reasone I bring this up is i noticed also in the tuning the tool you change various patters as far as switching time and all that

Chris

[Blacky]

The narrow band O2's are just that: narrow. That means they can detect when the AFR is 14.7 accurately, but greater than 14.7 is a "blur" and less than 14.7 is a"blur". The AFR, outside of the narrow band centered around 14.7, can only be accurately resolved by the sensors as either rich or lean.

The software in the PCM is a "switching algorithm". Basically if it sees the narrow band on the low side of the threshold voltage (which can vary depending on operating conditions - see {B4105} - O2 Switch Point.) it will increase fuel until it sees the O2 voltage on the rich side of the threshold. Then it will reduce fuel. This oscillation process assumes that because it is often crossing from rich to lean then "on average" the PCM is delivering 14.7 AFR - assuming all other components and calibrations are functioning and correct.

The narrow bands never actually measure the AFR, they just tell the PCM: rich - lean - rich - lean - rich - lean and so on.
The software algorithms in the PCM would be *very* different if fuel delivery was controlled by a couple of wideband O2 sensors instead.

The other O2 calibrations are used to:

1. Alter how fast the PCM tries to bring the O2 voltage back across the switch point to the other side. Too fast and it will get too much overshoot, too slow and it will stay rich or lean too long.
The STFT multipliers are multiplied by the STFT correction factor - which determines haw much adjustment the PCM makes to the injector base pulse width when trying to bring the O2 voltage back accross the threshold. The further the O2 voltage is away from the threshold the more the multipliers are driving the fuel delivery back across the threshold.

2. Alter how fast the STFT and LTFT are updated. The higher the airflow, (based on CL Mode which is based on air flow - see {B4107 - Closed Loop Mode) the faster the update frequency.

The LTFT Update Filter determines how much weight the current STFT should carry when applied to the LTFT. The higher the airflow, the more effect the STFT will have when used to updated the LTFT.

Regards
Paul

[87gmc]

Paul as always thanks for helping this stupid guy out :D

[mnotx]

Responses like this, and numerous others scattered about on this forum, should be copied over to the Tutorial forum. This is terrific information that will benefit a lot of the laymen like myself. It would be great to have them all where we could find them easily in the future.

[PSWired]

It's my understanding that the rich-lean switching also improves catalyst efficiency. I thought that was the main reason for the switching.

[Ira]

It's my understanding that the rich-lean switching also improves catalyst efficiency. I thought that was the main reason for the switching.

I'm not sure it's actually rich-lean switching. Well at some level it is, but if you put a wide band next to the narrow band you might discover it's only bouncing between 14.3 and 15.0. If you look at a log showing the front o2 sensors and injector duty cycle you'll notice that the duty cycle is essentially fixed while the O2 is bouncing.

Ira

[mistermike]

How useful are the average NBO2 voltages in the grand scheme of things? I noticed after header install, they went from low .4's to high .5's and LTFT's are generally trending positive. It looks like the ECM is overcompensating for the headers. Is there a workable correlation between NBO2 averages and AFR, or is it only useful as a crude indicator?

[Blacky]

How useful are the average NBO2 voltages in the grand scheme of things? I noticed after header install, they went from low .4's to high .5's and LTFT's are generally trending positive. It looks like the ECM is overcompensating for the headers. Is there a workable correlation between NBO2 averages and AFR, or is it only useful as a crude indicator?

It's an indication, but not a useful indication. Test have been done (I can't find the thread now) that show the NBO2 averages (at WOT) will vary up to 20%, for the same measured wide band O2 value.

Paul

[mistermike]

It's an indication, but not a useful indication. Test have been done (I can't find the thread now) that show the NBO2 averages (at WOT) will vary up to 20%, for the same measured wide band O2 value.

Paul
Thanks, Paul
I know these things are pretty useless at WOT. I was more wondering how indicative the avg O2 volts would be at part throttle. In my particular case, when I noticed the climb in avg voltage after the header install, it correlated with a noticeable drop in fuel economy at a steady 70 MPH. If I temporarily commanded a higher AF, the voltage came right down. Without a WB, of course I don't really know where I was at. Once I get the WB installed pre-cat, it might be interesting to plot avg NBv against WB A/F.

[mistermike]

It's an indication, but not a useful indication. Test have been done (I can't find the thread now) that show the NBO2 averages (at WOT) will vary up to 20%, for the same measured wide band O2 value.

Paul
Thanks, Paul
I know these things are pretty useless at WOT. I was more wondering how indicative the avg O2 volts would be at part throttle. In my particular case, when I noticed the climb in avg voltage after the header install, it correlated with a noticeable drop in fuel economy at a steady 70 MPH. If I temporarily commanded a higher AF, the voltage came right down. Without a WB, of course I don't really know where I was at. Once I get the WB installed pre-cat, it might be interesting to plot avg NBv against WB A/F.

[carneb]

Just wondering why the 02 switching voltages are different under different operating conditions {B4105}? I would have thought that the O2s would indicate 14.7 AFR as the same voltage under any condition? Is this not the case?

[ringram]

Sensor location seems to have an effect on what voltage is best.
I guess you can double check at each CL mode (airflow) point against a wideband if the o2 voltage matches stoich.

Seems a lot of hassle to me these days though. Best just to pull them out and run open loop :)