Members | Sign In
ScannerDanner Forum > Post your repair questions here
avatar

AFR Wideband O2 - demystifying this creature

posted Jun 21, 2016 05:16:53 by AlexMessina
I'm posting this image linked at very bottom link after watching ScannerDanner's vid on using pre and post O2 sensor response comparison, and response delay of rearO2 as a way of checking Cat Converter function. It's a great vid - https://www.youtube.com/watch?v=vyVnhCIMDnw

The twist here is that I'm using a system that has a AFR Wideband Front O2 and Narrowband rear O2 (or mid cat.
Car is Mazda6 2007 Gen1 2.3lt L3 Engine non-turbo, Auto.


I’m experimenting with wideband AFRs. Damn hard. Adds a layer of complication.
Car runs perfectly; but has a slight but persistent neg LTFT – around -5 to -8%. Never zero, never pos.

Yellow trace is a 5-wire AFR signal wire voltage trace (this has micro current changes to maintain stoich .. it also shows small V fluctuations: V reduces for Rich, Increase for Lean – opposite to narrowband O2)

The Pink trace is Post cat narrowband – it’s actually mid-cat on this car.

You’ll see that when I create a lean condition by pulling a vac hose the AFR does not change: maintains max lean state.

But mid-cat sensor reacts, going lean (normal low V)

When I add hydrocarbon spray via vac hose and pinch off vac leak, you see the AFR go rich (V drop), as does mid-cat (V rise).

Couple of points of interest:
1. AFR wideband behaviour is odd .. no lean response but immediate rich --- (scan data does however show lean/rich uA response)
2. There is a time delay, as you expect, between pre-cat and post-cat recovery from rich to lean/normal condition. But the delay is only 1.4sec … versus ScannerDanner expected delays of 4 seconds+ ... I don’t conclude that short delay is a Cat issue. Rather, I think it is a result of post-cat O2 actually being positioned mid-cat … ie shorter delay time? .

Can anyone explain the AFR behaviour lean response behaviour?
What do you think of response time delay of rear O2?

All comments/thoughts welcome:-) See this link for image (hope link works): https://www.scribd.com/doc/316302622/Waveform-O2-AFR-Narrowband-re-Post-Delay

thanks
alex
page   1
6 replies
avatar
AndyMacFadyen said Jun 21, 2016 07:17:11
The car and presumably the cat is 9+ years old personally would not expect it to perfom like a new cat.
The size of the cat relative to engine size will also be a factor and also the emision requirements the car is built . Another thing to consider is some engine types have tigher emision control pre-cat than others and may be fitted with a smaller cat when they come from the factory.

My own daily runner is 12 years old and built to EU3 standard and doing the same test looking at the O2 sensor output with a scan tool I see 3 frames delay (approx 1.3 seconds) before the rear O2 goes lean. At normal steady idle front O2 is switching at 3Hz and rear O2 sensor voltage is 0.81v, at an MOT tail pipe test in December it showed zero percentage C0 and the CO2 and HC were at the low end of the spec for the vehicle.
[Last edited Jun 21, 2016 11:23:32]
"Rust never sleeps"
avatar
Dylan said Jun 21, 2016 20:58:10
I’m experimenting with wideband AFRs. Damn hard. Adds a layer of complication.
Car runs perfectly; but has a slight but persistent neg LTFT – around -5 to -8%. Never zero, never pos.


What does your STFT read? LTFT is pretty acceptable.

You’ll see that when I create a lean condition by pulling a vac hose the AFR does not change


Hmmm...I would expect a reaction. I can see what you mean looking at your scope capture. How does it react on a snap throttle?

There is a time delay, as you expect, between pre-cat and post-cat recovery from rich to lean/normal condition. But the delay is only 1.4sec … versus ScannerDanner expected delays of 4 seconds+ ...


Just like Andy explained I wouldn't worry about the difference in delay.

Link works as you probably figured 😉
Belgium, Europe
avatar
AndyMacFadyen said Jun 22, 2016 12:30:40
The whole subject of testing wideband sensors is unknown territory for a lot of us outside dealerships and unlike normal Zirconia Lambda sensor there seems to be more variation in the characteristic of the output. I have learned what I can from Paul's videos on the subject but I really haven't had enough exposure to this type of system. On the Schrodingers Box YouTube channel Matt has a couple of vids on the subject that are interesting. The snap throttle test seems to be a very useable method.

Getting back to testing cat reserve oxygen storage capcity testing there are different ways of doing it. Being lazy I tend to prefer graphing the PIDs via OBD2 during a road test. I use prolonged full throttle in a high gear (I have a long steep hill near my location) followed by a coasting.
"Rust never sleeps"
avatar
Dylan said Jun 22, 2016 17:00:13
On the Schrodingers Box YouTube channel Matt has a couple of vids on the subject that are interesting.


You're right. I just had a look and it covers the basics of classic narrow band as well as the A/F sensor wideband.
Belgium, Europe
avatar
Dylan said Jun 24, 2016 19:35:49
In case you haven't seen Paul's videos on testing wideband O2's here are the links:

How To Test a Wideband O2 Sensor Part I

How To Test a Wideband O2 Part II


Belgium, Europe
avatar
AlexMessina said Jun 25, 2016 11:46:52
Hi guys, thanks heaps for engaging and I'm glad others are intersted in trying to crack open this black box of wideband sensors.
To follow up on your great points:
Andy, based on your info, I'm not that fussed about the delay or neg trims. They car runs well. Good mileage, no real issue to speak of (a little vibration at idle, but I'll describe that in a minute cos I think I got to the bottom of it). ... And you're right about the age of the Cat, and also the rear O2 for that matter. Age brings some sluggishness. The Front AFR is new, refit after a failure that threw no code (as usual!!)

Dylan, STFT is fine, around 0, indicating at best no prob, at worst that comp has compensated for some slight influence on LTFT (I have tried a MAF clean and new MAF ... no change) ... Agree LTFT is not an issue on the face of it and, if you're like me, I know it's possible to over think these things ... I do like it though when I plug in my scanner on other cars and I see a STFT and LTFT both nicely around 0 +/- 3-5% ... It's so neat.
It's worth noting, that the scanner reports a Rear O2 LT fuel trim that mostly closely mirrors the overall LTFT with slow,long loping fluctuations from lean/rich/lean ... I'm suspecting that the Rear O2 actually plays some significant role in trimming fuel on Mazdas. It may be a bit sluggish and may be contributing to -LTFT, and again that may suggest a tired CAT, but certainly not dead.

The reluctance of the front AFR to go further lean, at least as my Scope sees it, is a puzzle (or a program strategy??) ... Although the capture is not a classic fast snap throttle, I'm pretty sure that I did a couple of snaps and scope trace did not exceeds its max V ... However, I'll give this another go on next opportunity.
The interesting (confusing?) data is that the Scantool shows a clear lean/rich response in uA data, diving negative at the beginning of a Open Throttle, and soaring positive on the Throttle plate closure. Go figure.


Andy, I've seen Shrodingers stuff ... It's informative and i appreciate his presentation. It related to Toyota's as I recall, with a 3v bias and 3.3v signal wire? His Waveform is certainly typical of Toyotas, but as you say, the widebands seem to have a wide range of engineering protocols and program strategies and Toyota's don't necessarily behave like Mazda's etc.
What I hate is that manufacturers are so tight with information.
In fact, Toyotas seem to have the most OEM info available on widebands and frankly I can't find any OEM data on Mazdas ... only what aftermarket expert techs have pieced together.

I think it's our job to crack the black boxes that manufacturers build.
I look forward to any other thoughts or experiments you guys put together, and I'll do a bit more myself as soon as I get an opp, and will post.

One thing seems clear, however. Despite the many posts suggesting the only way to test a WB AFR is by scantool, it is obvious that you can effectively test and get a waveform if you know the signal wire.

On the rough'ish idle: Mazda has a program strategy that drops RPM by about 50 when at idle, in D (auto) with footbrake applied. It's a fuel/emissions device.
I discovered this after suspecting a slight brake booster leak was causing the idle drop and a slightly but noticeably increased vibration when applying brake at idle (say at stop lights).
However, on clamping off all booster vacuum lines, I discovered the idle drop was a response to a brake-on electrical input to PCM from the brakelight switch! This triggers the idle drop stategy. Recently tried a brand new Mazda3 ... Same behaviour. One for the archive!

cheers
A



Login below to reply: