Lean combustion?

[Chas54]

Ive just had my Elan MOT done and passed Ok but on the way back i could feel a bit of a stutter when reducing speed from say 50mph to 30 mph to pass through rural villages. The engine fault light came on and is a 4.4 when read off the light in service mode, This indicates a lean exhaust at the O2 sensor.
This is a precat 1991 now converted to closed loop .
The elanscan data shows that at the time the lean O2 sensor error occured (850 secs) , the Stochiometric dropped from 14:7 (pretty much as good as it gets) to 14.0 ( which would be richer?)

The IAC is static at the time although the manifold pressure increases a little but this could be a change in throttle position.

Once back up to speed the fault generally clears after a short period as the light goes out on the dashboard .

The petrol is fresh E5 from Tesco .
The fuel reg valve is relatively new
The timing is spot on
TPS is 40 at idle.
I am hopefully pasting an elanscan file here so if anyone has any insights they would be welcome....... no i'm not , cant remember how to paste an ecu file here

charlie

[Chas54]

My elanscan trace may be here as a zip

back from louth vvg louth vvg.zip

file

[Tuga2112]

i would have expected the TPS (throtle position sensor) to read something very low (near zero) at idle. are we talking about the same data ?

[chrism]

Chas54, like Tuga said I'd expect TPS to be more like 0.35V at idle so not sure where your figure of 40 comes from.

Static scans from cold make more sense to me since the sensors can be seen affecting each other more easily than a 'hot run' and also your idle TPS might be available to verify...

I've overlapped your coolant temp with a/f ratio and noticed that your thermostat seems to be opening around 60 - 70C instead of the more usual 82C and there's a largish dip in your a/f ratio at this point. You'll see that later on, around 800s when at 90ishC, all sorts of things start to happen:-



...coolant temp falls and a/f ratio has a 'hissy fit'...

I think I'd be inclined to suspect or replace the t/stat first and post a (say 10min) static warm-up EScan so's we could then look at the temp sensor and O2 sensor contributions. Another thought; I'm wondering whether there's some kind of a vacuum leak when very warm?

I'd certainly be interested to see what others think?

[Chas54]

Sorry about that, it should be 0.4 volts at idle .

I hadn't spotted the erratic coolant temperature temperature but a few weeks ago i thought it was running too cool and put a cardboard blank over part of the radiator grill and it all felt better. So i took the thermostat out and found a split around the centre of the diaphragm .. I replaced it with a new one at 88c i think , ill have to check that. But i haven't taken the cardboard blank off yet (forgot all about it).

I'll have to do a dataset from cold tomorrow , but i've tried and i don't think the fault shows up until the cars moving .

Only other thing now is it's reluctant to idle initially when cold , could be the learned memory wiped when i disconnected to clear fault codes . I think it should reset itself at 30ish.

Thanks for quick responses though.

[Chas54]

I have to apologise as i have posted the wrong elanscan dataset .. I think the data is from before i changed the thermostat.

I will post a correct set as soon as i find it

sorry

[Chas54]

Sorry about earlier foul up.

I have two elanscan data sets
One before the MOT whilst car at idle from cold . All looks good. This is with the new thermostat but my forgotten cardboard radiator blank in place.

The second is a short run after the MOT at at around 970 seconds a code 44 O2 sensor lean fault is recorded . air fuel ratio drops to 13.9 . I haven't spotted anything else happening yet.
I can feel it just before the ECU light comes on.

Any suggestions welcome,

regards
Charlie

8th august 26 pre mot fine.zip

faults aftr mot 2026.zip

[chrism]

Chas54 I'm not ignoring you but:-

1. imo your before MoT idle isn't 'okay' - eg look again at O2 sensor data. (Your t/stat seems good tho'.)
2. I don't mean to be disrespectful but you've included data eg the short run which, to me, doesn't seem to point towards 'code 44 O2 sensor 13.9 a/f ratio' and in my view can't be connected to the warm-up data.

Sorry but I can't 'see' anything...perhaps others LEC-ers can? Is it me or..?

[KenS]

Interesting plots! I agree with Chris that the Pre MOT plot does show something odd with the O2 sensor voltage, the sensor starts up OK around 50sec but dies at 90 sec and recovers again around 450sec. Plotting the 'Closed Loop' flag indicates that it was still in closed loop throughout.

Pre MOT O2-Closedloop.jpg

Then looking at the After MOT plots, again the O2 Sensor data is erratic and drops out, and it drops out of closed loop. (but there are points where the engine is restarted)

afterMOT O2 Closedloop.jpg

Then comparing the Closed/Open loop periods against the Fuel/air ratio, while in closed loop it looks to be lean, when in Open loop near normal or rich.

afterMOT air-fuel.jpg

So question is if its simply an intermittant O2 Sensor or connections to it or is it correctly reporting fueling issues and defaulting periodically to open loop?
My guess its the O2 sensor since when in open loop it can supply more than enough fuel, but not confident enough to say go buy a new O2 Sensor!

It would be interesting to try to get plot while in Service Mode since I think that would remove the O2 sensor from the equation. But dependant on your ElanScan hardware, I don't think its easy to apply the service mode link while the ElanScan interface plugged in.

Any thoughts Chris? Any other members spot anything hiding in the plots?

[chrism]

Erm, I reckon Ken's right that your O2 sensor might be struggling - and in the first instance I'd do another EScan idling from cold for ~10 min for us to see. Your original O2 scan showed an intermittency (black) and I've overlapped my own (healthy) scan in blue:



If your next O2 scan showed 'struggles' I'd substitute your O2 sensor. I've got a spare that you can borrow - paying homage to Ken's reluctance and careful advice about buying a new one. (Perhaps a pal closer to home has one on the shelf?)

I've tested my own O2 sensor in the dim and distant past ( - and I'd be happy to test yours?)

The lhs slate/green lead is the signal lead back to the ecu and enters the rhs O2 sensor as the black lead in the pic.



The lhs brown/yellow is the heater supply. It only gives a supply when the engine is running or ticking over and the red ign light is OFF. It enters the sensor as the bottom white lead in the pic.

The slightly hidden lhs black (under the slate/green) is the heater 'return', which is an earth, and enters the sensor as another white lead on the rhs of the pic.

I'd check the two whites of the (disconnected rhs) m/plug with the resistance side of your m/meter to see if the heater supply circuit is intact. You should get a reading between ~3ohms and ~15ohms or so if the sensor isn't broken. It's a kind of 'continuity check'.

I'm struggling to remember but I seem to think I checked the supply voltage with the sensor disconnected from the multiplug. The engine should be ticking over with the red ign light OFF. I set the m/meter to read 20Vmax and probed the lhs m/plug part between the brown/yellow supply and a convenient earth - the braid at the bonnet hinge - looking for a (12V-ish) voltage to supply the O2 sensor. I think I also checked that the black sensor m/plug earth really was an earth by connecting it to the bonnet braid and looking for continuity when bolted into place - I think the turbo might provide a path to earth.

Other random thoughts:- the brown/yellow O2 heater supply 'feeds' off the red ignition supply light on the dashboard. Has this been flickering at all giving you an interrupted O2 heater feed(?) - cos this might be a cause of your O2 drop-out. I'm thinking "is the alternator rectifier providing enough 'oomph' to supply the heater current?"

I think, too, that heat from the exhaust gases can restart signals from a broken O2 heater circuit supporting/explaining Ken's open/closed intermittency?

Hmm...is all this something to pick the bones out of, I wonder? Or does it add confusion?

[Chas54]

Thanks for replies
I may have restarted the car whilst running , in an attempt to clear the fault
I can confirm I have no issues with the ignition light flicker or fail.

I have two 02 sensors both bought in the last 5 years or so. The one I have in at the moment is one I cleaned by soaking in a baked bean tin of petrol in the sonic bath.

As they both seem to produce a similar result, I suspect a possible wiring fault?

When I converted to closed loop, I rang through the cable return to the ECU pin plug OK. The live side for the heater element and the earth. All were present at the unused plug in the engine bay.

I am going to do a "from cold" idling scan today , and if it looks reasonable go for a short run with a scan as well.

I can see the O2 sensor stops working at a point in my previous scans for some reason.

I have a "dapinky switch" on my elanscan cable so I can select service mode while scanning but have only done this ever at idle. I have a slight fear of switching to service mode as any raised engine speed as I think the ignition will not advance.
I will progress my tests and post results here.
Thanks again for your interest and comments. I was getting a bit stuck about my next steps.
Regards
Chas54

Also , if I can't find anything conclusive I may need to take up the kind offer of a O2 sensor test.

[chrism]

Yep, look forward to the next warm-up EScan - tho' I still think it's worth re-doing those continuity checks.

To explain, my spare O2 sensor came about because I discovered I'd lost closed loop, instinctively ordered a new RAuto part, then discovered that one of the connections behind the wheel arch had decided to 'let go'...

(Like you despite all the forebodings I washed my sensor in carb cleaner then white spirit (thinking it was duff anyway...) only to test it later and find out that it was working perfectly - so I left it in )

[KenS]

I have compared each parameter plot against the O2Voltage for the pre MOT run and have spotted a significant correlation with the IAC values. Not sure what it implies but thought its worth sharing.

IAC stable.jpg

It shows that when the IAC values become stable (in normal range?) the O2 sensor also starts showing normal characteristics (dynamic). Again hard to know which is cause and which is effect. Maybe the O2 Sensor really is showing correct values prior to 450sec and the mixture really is very lean and the IAC is trying but unable to reduce the volume of incoming air up to the point where it all starts working normally at around 450seconds. But what could lead to excessive air but only in the first 450 seconds (IAC, manifold leak?)
So I think any tests around the O2 sensor to ensure it has heater voltage and connections are beneficial, even if it only helps to rule it out, and your plan of a few more cold start scans with one is service mode is good, may be they will provide a good clue as to the initial cause.

I'll have alook at the injector pulse widths and duty cycles and see if its possible to deduce if they are trying to trying to get more fuel in during those lean periods, might be able to rule out a fuel supply issue. Too many possible causes in a closed loop system!

Good luck.

[RonR]

I can't remember who it was but some time ago someone on here had an intermittent connection at the connector under the RHF wheel arch - I think it's behind the wheel arch liner:
RHFHC pin 8 is the feed from O2 sensor to ECU.

Just saw it was Chris - 3 posts ago!

[Chas54]

Bearing in mind all these comments I think I will take the offside front wheel arch inner off tomorrow and have a poke around the connector under there and possibly ring the cable through to the ECU plug.
Once the plugs separated I can slip some fusewire up the side of the pin to check with a meter.
I'll do my test run after.

Thanks again everyone for valuable comments.

Chas54

[Chas54]

idlng then switch to service.zip

I thought id do a quick scan today with my No2 O2 sensor in place .
Prior to start i cleared the faults by battery disconnection and the switched the ignition on then off for ten seconds to allow the IAC to reset.
The engine was then started and ran pretty rough but i guess thats normal as the learned memory would be cleared. The engine actually stopped once before i got it idling better .
Once warmed the idle was stable and running reasonably although the O2 sensor didnt start working until the temp was around 90c. I know the O2 sensor isnt in use until a certain temp is reached .

At around 1000 secs i switched to service mode with my in built switch and the engine speed hunts a little. On switching back to normal and at around 1200secs i reved the engine and released at which point the fault O2 sensor lean was activated and looking at the trace the O2 sensor stops working.

I think tomorrow i will check the O2 sensor connections .

Any comments welcomed

Regards Chas 54

[chrism]

My first thought is that the O2 heater supply isn't working to initiate the sensor and that the exhaust gas temp is finally activating it. I wonder if, meanwhile, your 1499 ecu is trying to compensate with a/f ratio changes?

Let's see what those connections are like.

[chrism]

ps Ken's already noted the correlation between IAC and O2 sensor on your prev. scan - and it's still there on the later scan



pps something strange happening at 400s?

[Chas54]

I think 400secs when the engine stopped and i restarted .

At the moment i'm planning checking the cables to / from the O2 sensor .

[Chas54]

I have been checking out supplies at the O2 sensor and have reverted back to my No.2 sensor as well.
I haven’t checked through to the ECU yet as i had the car on the wrong side of the garage ( passenger door against the wall to allow driver wheel arch removal)
The key on voltage value is 0.396v at the O2 sensor connection plug to the ECU (.32 to .49 acceptable)
The voltage at the heater to earth terminal is 12v and 14.4v running .
The terminal are all clean and good under the wheel arch as well.
The resistance of the heater circuit in my two O2 sensors are 206 ohms and 260 ohms

I have to next test between the pin 7 J2 plug at the ECU and the O2 feedback at the sensor plug.

One or two other points Ive been considering :-
Closed loop operation is started once engine has been running long enough and coolant temperature high enough EMJ1 page 13.

The O2 sensor is out of use at wider throttle openings but it must still be heated as the ignition light connection shared would be live.. Its to suppress engine knock and changes to open loop. At these times the Block Learned Memory must be being used ??

Ive just been up the road in the car after it idle well on the driveway . Almost as soon as i got underway the light came on the ECU , i carried on and then pulled over but it remained on. Heading home the light cleared and then initiated again . Back home to idle on the drive and the light had cleared..

Looking at the elanscan traces i see the O2 sensor stopped working again for some reason. I think i may have to rig my voltmeter so i can see the terminal voltage at the O2 sensor so when the fault is initiated i can glance at the meter and see if ive lost the heater voltage. Surely it should be there all the time ?, its unregulated by the ECU ? The supply point behind the dashboard is awkward to get at

To clear my O2 sensor lean fault , I have to disconnect the battery thereby clearing the block learned memory. ( i didn’t clear it for today’s short run)

How to tell on elanscan if its in open or closed loop operation ? Flag 42 bit 0 indicates closed loop but if i am trying to look at certain points to see if it changes (with the cursor position) i don’t see any change.

We dont see Block Learned memory cell count or cell in use on elanscan..... I'm not sure they would mean much to me but they can indicate what’s happening with fuelling the engine.

Again ive included the scan for any comments

regards Chas54

idle no2 O2 sensor and short run FAULT.zip