wideband (another crappy low light video with bad sound...)

[exdeath]

Testing the wideband. Dude I got it from thought it wasn't working and swore the sensor was bad. He had depinned the 6 pin header from the O2 harness to fish it through a firewall and reinstalled them wrong, and broke a few barbs that keep the pins secure in the process.

Half of $1.20 and 5 minutes later:

http://www.youtube.com/watch?v=v1I9ZjdbdcQ

Thought I'd share, very educational for anyone curious how EFI works. Here you can see the O2 sensors in closed loop doing their job oscillating and keeping it idling centered perfectly on 14.8:1, the accelerator enrichment shot when revving it (going to 11.x:1 in a forced induction application), etc. Cold start you would see it idling around 12:1. What you see on that gauge is what that computer in your car spends 99% of it's time doing.

Brand new this is a $200 (I think?) AEM wideband with a Bosch sensor, cable, and gauge with integrated UEGO controller. I'm very impressed with the accuracy and super fast response for such a low cost unit.

PS: goes to lean infinity when the throttle is snapped shut after revving it, for anybody who still doesn't believe injectors are shut off completely when the engine is "coasting" :awe: This settles this long standing ATG debate once and for all.

 

[alkalinetaupehat]

What's with those blips to 15-19? It looks like your Terminator is running lean at idle.

 

[exdeath]

It's me playing with the throttle, or just normal overrun after the throttle closes or when the system bounces around as it recovers from a transient condition before stabilizing again. Perfectly normal. The information from O2 sensors is only useful under stable steady state conditions like a WOT pull or idle, they bounce around a lot if the system is chaotic and unstable such as jiggling the accelerator pedal, etc, when the exhaust system and air/fuel mixture is all over the place and unpredictable.

The first time it shoots to "---" you can hear a very slight rev where I placed my foot on the throttle and moved it a little before I actually revved it.

What is important is that stable idle is between 14.5-15.5 and WOT is ~11.5. Those times it spikes lean after lifting off the throttle are due to DFCO (decel fuel cut off) and there is no fuel present in the system at all at that moment thus full lean and no danger. Then it starts injecting again to stabilize idle as the RPMS come down, hence causing the oscillations before returning to stable between 14.5 to 15.5. And there is no load on the engine at idle, so it's common to idle around 15 for emissions. Any time you see it go way past 15, it's always with closed throttle and decreasing RPM immediately following a throttle opening and sudden closing, no matter how slight (max vacuum and no fuel present).

Watch it again and listen, every time it goes past 15, you'll hear the trail of a rev dying down, even the tiniest ones. EFI is extremely sensitive, it's amazing.

 

[Howard]

PS: goes to lean infinity when the throttle is snapped shut after revving it, for anybody who still doesn't believe injectors are shut off completely when the engine is "coasting" :awe: This settles this long standing ATG debate once and for all.

Depends on what RPM and sometimes what gear you're in, too.

 

[exdeath]

Also keep in mind that while WE are observing on a wideband, the ECU is still getting IT'S information from, and controlling fueling with, the factory narrow band OBDII bank 0/1 sensors, hence the wild swings and oscillations any time it resumes closed loop. A narrow band only knows two states and will report the same thing for a long time, then suddenly switch, causing the computer to overshoot it's target and suddenly reverse direction. So really, in addition to monitoring the absolute physical state of the exhaust stream, we are using a wideband to observe the way the narrowband O2s operate. If a factory ECU used widebands, it could sit at 14.8x perfectly without the slightest hint of oscillation.

 

[Greenman]

Also keep in mind that while WE are observing on a wideband, the ECU is still getting IT'S information from, and controlling fueling with, the factory narrow band OBDII bank 0/1 sensors, hence the wild swings and oscillations any time it resumes closed loop. A narrow band only knows two states and will report the same thing for a long time, then suddenly switch, causing the computer to overshoot it's target and suddenly reverse direction. So really, in addition to monitoring the absolute physical state of the exhaust stream, we are using a wideband to observe the way the narrowband O2s operate. If a factory ECU used widebands, it could sit at 14.8x perfectly without the slightest hint of oscillation.

That's not how it was explained to me. I was told the switching sensors were faster than the wide band, and a great deal more reliable. The problem with switching sensors is they can't read below 13.8:1 AFR.

I don't know any of this as fact, it was explained to me by a fellow who designs tuners, I assumed he knew what he was talking about.

 

[exdeath]

That's not how it was explained to me. I was told the switching sensors were faster than the wide band, and a great deal more reliable.

They are, though that doesn't really have anything to do with what I was saying 😀 What I described is not a lack of speed of a narrowband but their non-linearity and lack of accuracy at the stoichiometric "zero" point. Their output is vertical around that point. That's what causes the computer to always over shoot and change directions repeatedly, causing the oscillations between 14.5 to 15.5, instead of just sitting neatly at a fixed value.

I can see, with desire for better mileage and emissions, and tricky and dangerous things like lean burn with direct injection, production vehicle ECUs moving over to OEM wideband feedback that reports back the actual ratio in real time, instead of just switching rich/lean. In fact I think it's already happening. If I recall there is a newer third type of sensor that is linear and accurate like a wideband, but suitable for use in long term OBDII closed loop use in production vehicles.

 

[Greenman]

They are, though that doesn't really have anything to do with what I was saying 😀 What I described is not a lack of speed of a narrowband but their non-linearity and lack of accuracy at the stoichiometric "zero" point. Their output is vertical around that point. That's what causes the computer to always over shoot and change directions repeatedly, causing the oscillations between 14.5 to 15.5, instead of just sitting neatly at a fixed value.

I can see, with desire for better mileage and emissions, and tricky and dangerous things like lean burn with direct injection, production vehicle ECUs moving over to OEM wideband feedback that reports back the actual ratio in real time, instead of just switching rich/lean. In fact I think it's already happening. If I recall there is a newer third type of sensor that is linear and accurate like a wideband, but suitable for use in long term OBDII closed loop use in production vehicles.

Sounds like you know more about it than I do. But I'd be nothing short of astounded if the oscillation was as great as you claim.

 

[exdeath]

It is. Normal O2 sensors cycle between rich/lean once per second, from .8v to .2v. Around .5v is 14.8:1. This is exactly what's going on as the gauge in the video sweeps from 14.x to 15.x every second. How wide the actual AFR sweeps from lean to rich as it idles trying to keep it centered will vary and is a function of tuning, fuel injector size (precision at lower opening times), type of idle control motor, etc.