Lean combustion?

[chrism]

Hoping to rectify my mistake -> mine's definitely a CrankAltFile "2" :-



And my O2 sensor (now back to Walker) isn't dropping out...



...so I'm absolutely confident that my m100 1499 SE data can be used as a CrankAltFile reference if necessary

[KenS]

Chas54 - That last run is quite a mixture! I am very confused with the CrankAltFire, this plot shows it up at 2 for the first part of the run and at 0 the second half of the run. If the AI description is correct, it should only be active with revs below 800rpm, that is not true in this plot. I guess we cannot make any good progress looking at CrankAltFire until we know its active and inactive state and confirmation of it operation. Looking across several old LEC plots it seems half are running with it at 2 and half with it at 0 ! If this is a fault, a lot of people have it, probably including me. I thought that if it was set active it stopped the ECU from adjusting the timing, but then why do the plots still show that the timing IS being adjusted, shouldn't it be fixed at 16 deg ? Maybe Jaoa is correct, you really cannot trust AI.

bat diss.jpg

Probably not worth taking any action until we understand things a bit better, but while i think of it - is your coil base plate properly earthed - there are insulating bushes to keep it off the chassis and a special large solder tag to make the earth connection.

Chrism - as above, I think I might have to measure the 'bypass' signal on the ECU-Coil pack to see if that makes any sense. I agree that is not clear now if this is a red herring for Chas54 problem (but worrying if a lot of people have an unrealised issue here - including me). I had no reason to doubt mine, but now I just don't know!
One of my plots:

my crankaltfire.jpg

[peteoddcar]

Chris called me today.
I’ve tried to read the thread and I’m getting a tad confused, I am sure you are not just confused but extremely frustrated.
So just because I kind of missed the circumstances of when how this started happening it’s all been ok until the MOT ?
Y/N
How long has the Closed Loop system been in place ? Is the car originally open loop?
Cat or no cat ?
What O2 sensor are you running? What Chip ?

Chris
Do you recall if Andrew O’s car was doing the same thing - I think it was - I drove with him and it sounds a similar failure, pretty sure our WA thread mentioned #44
This turned out to be a wrong / bad O2 sensor. New SJ sensor fixed it.

[peteoddcar]

So I went back and some of my questions have been answered.

To clarify the feed to the O2 sensor heater is an output from the alternator warning lamp terminal.
The alternator lamp extinguishes when the alternator pin reverts from negative ( not charging or not rotating ) to system voltage when the alternator starts to charge, which takes away the potential difference on the neg side of the bulb and therefore puts it out.

[Chas54]

Thanks for taking the time to consider my problem.

The ECU light came up on my way to the MOT test but it did go off when
I stopped the engine and then restarted it. Passed MOT fine.

The car has done very few miles in the past 18months as I moved from Hampshire to Lincolnshire. It ran well without faults when I brought the car up . I remember because there was rain and fog 240miles. The ECU light has come on again a few times doing very local trips.
I can feel when its about to happen as there is a slight hesitation just before.
It seems to correspond with a loss in the O2 sensor feedback voltage and its not clear to me if this is cause or effect.

The car is a pre cat , open loop system.
The wiring for the O2 sensor was already built in and included in the original wiring loom

At some time in its past the ECM chip was changed as the one I removed has hand written date of 09/05 ID9930 . This was running open loop with a CO pot.

I changed the ECM complete with a new unit from a Lotus supplier , with a 1499 chip. This was Sept 2019.
I had a problem meeting the MOT standards .
Around this time (or just before ) I had a vacuum leak and after a lot of searching, including rejointing the plenum and 2ndry throttle body , I found a couple of issues.

IAC valve head was marked badly , I swapped for new.
Exhaust gas recirc valve head was marked through passing gas continuously, it was sticking open due to a mastic like product having been used to glue the black plastic cap onto the actuator pintle valve . It had blocked the spindle which vents the vacuum and allows the EGR valve to close. So it would open the EGR for a long period , and hold it open, even after the engine was stopped . This actually gave me a period of poor running every time I restarted the car. First thing , cold out the garage it would run great. Park up and return to it , it would run badly due to the poor manifold vacuum ,
I did replace the EGR valve but in my current stage of exploration ive blanked it off at the flange. The vacuum pipes to the pintle also are capped off. .


I have a new Walker 350-33022 O2 sensor.
This replaced a Bosch 15701

Both have a low resistance and produce the same results on elanscan .

Thanks for the alternator light function explanation.

At the moment I have to look at the earthing of the ignition pack plate but i'm running out of ideas .

Thanks for taking the time to consider my plight , any suggestions appreciated

Chas54

[Tuga2112]

from what your describing about the light coming on and off. i think your alternator regulator is either randomly malfunctioning. OR the plug is loose and sometimes doesnt get enough contant to allow enough the currnet flow to the o2 sensor.

when i installed my refurbished alternator i did the mistake of not connecting the plug 100% properly seated. its an easier mistake to make then one would think.

unfortunately.. the easiest way to access that plug is to remove the subframe.. if you have triple jointer arms. you may be able to reach it between the subframe and driveshaft from underneath to try re-seat it.

[Chas54]

Tuga
My light is the ECU warning light, not the alternator warning light.

Chas54

[RonR]

Re: Crank Alt Fire - Just looked it up and it's a 1-Bit flag, meaning it can only ever be 0 or 1.
Maybe there's a bug in Elanscan or the definition data which means it's capturing 2 bits instead of 1???

Edit: Although the BIT is only ever 0 or 1, it's the second bit in the 'Mode Word 3' byte meaning it's represented in Elanscan by a 0 or a 2, the bit multipliers being: 1,2,4,8,16,32,64,128

[chrism]

So does this mean that 'accepted' values are usually 0 meaning OFF and 1 meaning ON but our EScan labels this as 0->off and 2->on? - Just a label error?

OR that some versions of EScan are capturing the ON data twice?

Is it worth my doing another scan using the original EScan to see if my "2" is duplicated? (If it's a "1" then it's mistakenly got into beta_Escan and if it's a "2"then it was always there?)

A bit of a thread hijack til Chas54 can verify his wiring...

[peteoddcar]

Chas
My car is very early pre production ( first 100 car build) open loop non cat.
This is one theory I have and it’s based upon a fault that my car developed in 2024.
I would driving and extremely randomly but often on a rough roads engine would stumble recover, possibly stumble and cut out momentarily- check lamp would light and limp mode. Key off and on might be ok for days or come back instantly. Rev counter never dropped indicating that the EMS and ignition was still functioning.
There was never any fault code set.
After a lot of checking this and that new fuel filter made no difference I surmised it was possibly the fuel pump cutting out momentarily.
On investigation I found the imobiliser internal relay was probably at fault I took the immobiliser out of circuit and it’s been fine since a few thousand miles done.
I’d suggest connecting a fuel pressure gauge.
This would correlate with a weak mixture which on a closed loop car would set a code 44

Just a hypothesis.

[Chas54]

My immobiliser was removed around 2019.But my problem does feel like a slight loss of fuel on occasions.
I changed the petrol last week to try and rule out water in the petrol.
Another thought I had was petrol pump suction line leak allowing some air to enter.

Once I've put the ignition pack back in I'll have a look for some petrol pressure testing gear. I think I have some somewhere.

Thanks for suggestions

Chas54

[RonR]

So does this mean that 'accepted' values are usually 0 meaning OFF and 1 meaning ON but our EScan labels this as 0->off and 2->on? - Just a label error?

Not quite Chris, it's a bit / byte / binary thing:
The 8 bits of the 'Mode Word 3' byte are given values according to their position: 1 2 4 8 16 32 64 128, so that the bits can co-exist in a Byte (Max value = 255)
For example if bits 1, 4 and 7 are on ==> 1001 0010 the byte value would be 1 + 8 + 64 = 73.
Our 'Crank Alt Fire' flag is in position 2, so although the bit can only be a 0 or 1, Elanscan masks the other bits in the byte and reports on the byte value, which ends up being 0100 0000 ==> 2.

[chrism]

Oo-'er missus, sounds a little bit more complex then...

That new fuel pressure lead sounds interesting. The Chas54 scan does show 'oxygen sensor lean' despite a:f and closed loop targets(?)

Chas, were you lucky enough to have your sensor wiring factory installed or did a prev owner 'suss' it? I'm asking cos you don't seem to have the w/arch shroud/m/plug that I certainly have.

Your 9930 chip hasn't necessarily been changed. Mine's a 9930 SE and I'm pretty sure it's original. (Ron's 1499 upgrade adds another board that allows the old chip to remain in the ecu.)

[peteoddcar]

Chas
I reckon you have exhausted the bad O2 sensor / wiring as a cause.
‘Over rough road’ got me thinking again.
A further thought on your water in fuel hypothesis ( I like that word! ) I did have similar stumbles way back when I first got the car in 2009. I don’t recall if it set the check light or not.
The stumbles were intermittent- I did a fuel filter change( a fun job probably the first time it had been done I reckon) I did a long run to the Lake District and it stumbled once or twice early in the journey and I just carried on in a kill or cure mentality, but with a couple subsequent tanks of fresh fuel it never did it again. Until 20024.
The fuel filler drain was blocked,so guess that’s how the water ( if that’s what it was got into the tank) I have always made sure it’s unblocked since check when I wash the car.

If you took the pump out and fuel level is very low I reckon it could be possible to see if there are globules of water in the tank. But I’ve never tried that method.

[Chas54]

I have to admit that the discussion on bits and bytes has gone over my head entirely.

Ive now reinstalled the ignition pack .Earth is on the plate and isolated from the longeron.. The other end it think is on the offside end of the Plenum or manifold.. I dont know why it wasnt earthed through the longeron when it was designed, there must be a good reason though.

There is a cable shrouding earth which at the coil pack is attached to the black wire in the upper one of the 3 pin plugs. This being the the cable that comes directley from the magnetic sensor on the CAS , so is ignition related . At the CAS end , the sheathing isnt attached to anything and i wondered if it should go to an earth. The ignition coil drawing doesnt show it as connected to anything and i guess its just to stop any induced interference from the nearby HT leads. Or at least to dissipate it.

I was considering checking the fuel pressure but its a pig of a job without the plenum off. Ive put it off for now as i found some figures written down from when i did test it a few years ago and at the time changed the regulator valve as the original had a disintegrating diaphragm .

I am going to check out the fuel pump and suction line though.

I am not sure that ive exhausted all the electrical problems with the O2 sensor system. My trace on idling from cold shows the sensor produces a voltage very soon after starting but it then drops until the engine is warmed up.. I was thinking this was because the operating temperature is around 350c. This doesnt explain why other peoples O2 sensors dont do this.
Also a run from a couple of day ago from a warm start showed the sensor took much the same amount of time to work and then dropped off for around the same amount of time as when doing a cold start.
As i had a voltmeter tucked into the +ve supply to the O2 heater , i could see the running volts around 13.5v so ive dissmissed any issue with battery or alternator or the heater supply.

I still havent tried substituting the feed back cable entirely. Ive put this off due to the very short cable end at the ECU plug . I was really hopping to cut the wire and solder to extension cables onto the two ends thereby providing an easy route back to the original wire if no fault was found and also an easy way to attach a new cable for a test . But theres not much room so i was waiting for my grandson to visit as he will fit better and learn how to solder.

The wiring to the O2 sensor looks to me to be factory fitted , There was no 02 sensor fitted though. The O2 sensor feedback cable at the ECU end is incorporated into the wiring loom.
I may have to take the wheel arch off again and strip the outer insulating tape on the loom to see if ive missed the underwing connector somehow.

Thanks again to all for advice and suggestions

Chas54

[chrism]

Chas54 some of the SE's had fully-completed O2 sensor wiring with the old promid - others, like mine, had just some of the wiring and needed completing...

Yours seems to have been one of the lucky ones since it has full OEM wiring

ps might not be loose w/arch pins since you say it's a regular occurrence?

[Chas54]

Ive had the petrol pump out of the tank and inspected it for leaks , blowing back down the pump discharge pipework shows there is no leak there. The suction side isnt so easy to test as theres a sock filter i couldnt see how to remove but as the assembly is bellow the fuel level , i cant see a suction leak making any difference.. I dont think theres any water in the fuel either. I syphoned half a jam jar out the lowest part of the well and there is no sign of water. I added some water to the jar just to check i would see it and its really obvious if it has any water in it at all.

I took the advice and cleared the filler flap drain which was partially blocked.

Ive just run the engine to idle and then stopped it for 15 minutes or so , then restarted it to see if i could see the warming up drop off of the O2 sensor signal voltage evem on a warmish start.

Ive attached the elanscan from this as it shows pn the cold start the O2 sensor restarted at about 71c , where the 2nd warm start it picked up at 78c . The overall trend patterns look very similar but with a shorter warming up period.

pet tank and ign pack inspect.zip

Any comments welcome

Chas54

[KenS]

Yes, does still look the same effect going on. It shows again a consistant fault, as the coolant reaches 35 degC, the O2 sensor instantly goes to indicate 'lean'. There must be a change in the ECU control system at 35 degC that considers an input or changes a fuelling map(?) which then makes it 'lean'. It seems odd that it remains in closed loop, since when fully warm, this 'lean' condition should (and does) cause it to drop out of Closed loop after a few seconds (the problem we see when driving). We have considered Crank Alt Fire as a possible errant input but crossed that off the list since we have reference plots with it high and low and not causing this same problem (now separate investigation into Crank Alt Fire!). Assuming the O2 sensor voltages are valid, is there any way in finding out what decision the ECU makes at 35 degC coolant temperature? Comments?

11-9-26.jpg

[chrism]

Yes I can consolidate all of KenS's points from my own pouring over of the latest scan. Some of these oddities are what's making me suspect target values instead of actuals. eg if the ecu sees closed loop does it automatically flag up a:f of 14.7:1 - and if it drops out does it stay in 14.7:1?

Wouldn't it be a lot easier to put a known working promid 1499 into Chas' ecu? This'll pull away a large chunk of is-it-isn't-it ecu or target value stuff? Then we'll know it can only be the O2 sensor wiring, since we can now be pretty sure the O2 sensor isn't duff...

[Chas54]

That's seems like an idea, I'll put the old chip back in . I might have to search about for the. CO pot adjusting instructions .

Might give me a chance to re appraise the O2 sensor cable splicing. Also going to take the inner wing off for another look for the illusive hidden plug. I've got to check the timing after having the CAS off anyway.

But I've got to leave it a couple of days , weekend outing.

Chas54