Right, you know how the coolant temp sensor senses when the engine is up to 'normal operating temperature' and stops the ECU running its rich mixture for start-up.. well at exactly what temperature does this occur?

Been doing some VAG-COMing.. would I be right in saying that once the ECT passes this given temperature, you should be able to see a decrease in the 'Injector On Time' values?

At the moment my thermostat is stuck open and the car idles at like 88degs.. will the ECU consider this as 'cold'?

Car is a 6N 1.4 8v running the Bosch Motronic 9.0 injection system.

Thanks folks

90 degrees is generally about right

aargh daym.. generally 90, or exactly 90?

as in.. if my car was never reaching 90 degs (always just short, at 88 degs) due to open thermostat, then would it always run rich?

no, the difference between 88 and 90 is nominal, the bosch ecu's coolant enrichment maps only go in 10 degree c steps iirc...

you can tell if the car is running rich or not using vag-com, run the engine up to temp and with it idling... engine 01 measuring blocks 08, group 1... one of the 'windows' will have the coolant temp that the ecu is seeing and the voltage displayed in the 3rd 'window' is the output of the probe switching from rich/lean for closed loop idle mixture, it should alternate bewteen 0 and 1v ish roughly every 2 seconds

Edited 3 February, 201016 yr by SkodaBoy

incidently, i *think* that the injection period readouts are not used on the bosch ecu's but i would have to check

no, the difference between 88 and 90 is nominal, the bosch ecu's coolant enrichment maps only go in 10 degree c steps iirc...

you can tell if the car is running rich or not using vag-com, run the engine up to temp and with it idling... engine 01 measuring blocks 08, group 1... one of the 'windows' will have the coolant temp that the ecu is seeing and the voltage displayed in the 3rd 'window' is the output of the probe switching from rich/lean for closed loop idle mixture, it should alternate bewteen 0 and 1v ish roughly every 2 seconds

oh ok cool man, that's something else off the list.

my lambda voltage is fluctuating between those voltages like you said.. and I think it is at around that rate of every 2 seconds.

but the values it varies between is 0.1 and 0.8v. it does not go above 0.8v, and will not drop below 0.1v.. does this signify whether the lambda is sensing a rich/lean mixture?

incidently, i *think* that the injection period readouts are not used on the bosch ecu's but i would have to check

In Group 002 (Engine Measuring Blocks) in second field I think, I get 'Injector On Time = 4.50-4.60ms' (it fluctuates between the two.

But then in Group 003 (Engine Measuring Blocks) in second field again, I get 'Injector On Time = 2.10ms'.

Are any of these actually from valid feeds from the engine, and which one is right? surely every ECU controls the injector on time, so its values will be stored and monitored for VAG-COM to pick up? I've probs understood what you said a bit wrong though lol

a lambda cycling between 0.2-0.8v is doing just fine. Have you fixed all the other issues?

In Group 002 (Engine Measuring Blocks) in second field I think, I get 'Injector On Time = 4.50-4.60ms' (it fluctuates between the two.

But then in Group 003 (Engine Measuring Blocks) in second field again, I get 'Injector On Time = 2.10ms'.

Are any of these actually from valid feeds from the engine, and which one is right? surely every ECU controls the injector on time, so its values will be stored and monitored for VAG-COM to pick up? I've probs understood what you said a bit wrong though lol

ok let me explain a bit better, the ecu does control the fueling and spark timing, but not all of them have the functionality to 'monitor' the injection timing, one of those injection timing settings will be the actual opening period of the injector and 4.5 milliseconds at idle sounds about right, the other one will be the absolute timing ie.. how many milliseconds before TDC does it inject the fuel

0.1 to 0.8v is normal but do you sometimes get the fault code 'lamda signal too low'?? that would suggest that the heater element in the tip of the sensor is getting tired

ok sweet.

georgio, i haven't currently done anything else (I'm in uni till the weekend) but i wanted to understand the values i was currently getting, fix the fault (leaking manifold) and then study the values again to see what has changed.

skodaboy, sounds good mate.. will the injectors have the same opening time on fast idle (MOT definition), and just cycle through injections at a quicker rate? I really need to see if these values change, as it's the fast idle I am failing (although natural idle is borderline limit aswell)

do you think the absolute timing values I had (2.10ms at idle) are roughly correct?

as for the fault code, only fault code I got was for 'NO SIGNAL', not LOW SIGNAL. Which I think is because I ran the engine with lambda unplugged for a bit..

00525 - Oxygen Sensor (G39)

03-10 - No Signal - Intermittent

Have cleared this, and has since not re-appeared..

finally, I was thinking.. if the lambda thinks the overall mix is too lean, and it has to keep correcting it by running engine rich, will this be recorded in the fuel trims? (long term or short term?). That way, the lambda voltages will be 'normal' as the sensor thinks the mixture is back to normal (stoichiometric) after adjusting to a richer mixture (but it's not, because it is fooled by the air leak).

i am imagining that my fuel trims will be highish positive percentages, because the lambda was sensing a lean mixture? (haven't been able to check them yet - dont know how?)

so.. when I solve the leak, the fuel trims will have to be adjusted and brought back closer to 0% as the sensor no longer picks up a lean mixture and doesn't need to constantly richen the mixture?

and then also my injector on times will drop at fast idle, hence lowering CO emissions?

i know it's going into needless detail, but i just want to know if i'm understanding the process correctly, before assuming that i've sorted the rich running problem and then going and failing another test.

appreciate the ongoing help chaps, let me know if i'm totally on the wrong lines!

ah, ok, i was thinking about this again yesterday and i think i might have misled you a bit, the lambda signal only switches when the stoichiometric ratio is right, this of course does not take account of the fact you may have an exhaust leak allowing air in, or possibly the engine is burning oil, or a vacuum leak...

can you be a bit more specific with what emmisions values you failed on? CO?? CO2??

Here are the emissions I got (2x MOT fails)..

FIRST MOT - Fast Idle 1

"CO - 0.543% (0.300% limit)

HC - 71ppm (<=200ppm limit) FINE

Lambda - 1.053 (0.970-1.030 limit)

Fast Idle 2

CO - 0.526%

HC - 73 FINE

Lambda - 1.208"

Changed Cat (may have been dud) and Lambda Sensor (may have also been dud) and got rid of loads of exhaust blows..

"SECOND MOT - Fast Idle 1

CO - 0.559% (0.300% limit) < Got WORSE?

HC - 73ppm (<=200ppm limit) Same, Fine

Lambda - 1.053 (1.030 limit) < Improved

Fast Idle 2

CO - 0.625 < Still WORSE?

HC - 96

Lambda - 1.036 (1.030 limit)" < SO CLOSE!

Oh.. that's just thrown all my ideas out of the window lol.

So.. if the stoichiometric ratio wasn't being met, what does the lambda sensor do?

This is bearing in mind that.. what the lambda senses as the stoiciometric ratio isn't actually going to be correct if there is a leak in front of it, is there?

If more air enters the exhaust fumes before the lambda sensor, the sensor will assume the mixture going into the engine is lean (even though it might have been stoichiometric) and will try and increase the fuel going into the mixture on average to keep it nearer what it thinks is stoichiometric?

So, the ECU picks up that it keeps having to add time onto the injector pulse after every cycle, and stores this in some kind of long term memory / fuel trim after some time, as a % ?

Is that how it all works? :(

you've more or less hit the nail on the head with your description... there are long term additive and siubtractive fuel trims yes, but thier scope of adjustment is fairly limited, it wont contiune to blindly add fuel ad hoc once it has reach a set limit, you get a code like 'bank 1 - additive air - system too lean' or something similar, the most common cause is a leaking flexi pipe (if fitted) or a leaking manifold or downpipe gasket...

you can reset the long term learned values but i will have o look up the channel/ block number for it!!

ok a few questions... have you checked the ignition timing isn't too far advanced?? what condition is the air filter in?

you've more or less hit the nail on the head with your description... there are long term additive and siubtractive fuel trims yes, but thier scope of adjustment is fairly limited, it wont contiune to blindly add fuel ad hoc once it has reach a set limit, you get a code like 'bank 1 - additive air - system too lean' or something similar, the most common cause is a leaking flexi pipe (if fitted) or a leaking manifold or downpipe gasket...

you can reset the long term learned values but i will have o look up the channel/ block number for it!!

ok a few questions... have you checked the ignition timing isn't too far advanced?? what condition is the air filter in?

What if the long term fuel trims required weren't past the correction size needed to set off a code (is it 20% that sets off a code?) If it were just under that, there would be loads of correction going on but still not enough to set off a code like..

if you could find out which measuring block group i could find them in that would be outstanding dude.

as for resetting the trims, i wouldn't have to do that would i? i could just make the correction (ie fix the air leak) and then the ECU would start finding out that the LTFT required is different so will start adjusting itself back to normal again won't it? and the percentages will make their way closer to 0% again, over about 5 miles of driving isn't it?

i've looked at the air filter and it doesn't seem clogged or anything, i will probably take it out for the emissions test though. as for the ignition timing, i haven't checked it but even if it were too high, that would be because of a sensor reading, there is no way to manually adjust it is there?

can i check the timing by vagcom, or would i need a timing gun? i thought the timing changed constantly like the fuelling, to maximise leanness before any knock occurs?

the most common cause is a leaking flexi pipe (if fitted) or a leaking manifold or downpipe gasket...

Oh and.. by flexi pipe, do you mean exhaust flexi-joint thing on the downpipe? or a flexipipe coming off inlet manifold? (brake servo or FPR)?

I thought about the leaking flexijoint and downpipe gasket.. but mine is a later 6N and the lambda is in the manifold, in front of the downpipe gask and flexijoint.. so would leaks there actually affect the closedloop cycle at all?

i've checked for leaks around the inlet side of things although can't really be sure without having it smoke tested or something

as you mentioned, my money is on leaking exhaust mani (ie, the broken stud)

Edited 6 February, 201016 yr by CWA

ok, firstly it will re-adapt itself after a peroid of time, normally after a few miles or so, but you can perform a reset to put them back to the factory values, but i can't remember it off the top of my head at the moment (still at work now)

ref; 20%+-, i'm not 100% sure what the actual percentage is, but there is a very sticky grey area between long term and short term fuel trims...

you can adjust the ignition timing, with the engine running, use vag-com, go into measureing blocks, then group 1, and this check the 4th 'window' if it shows 00000000 then you can do it, if not then not all the pre-requisits have been met.. then go basic settings 04 group 1, then the engine should rise to an idle speed of about 1200rpm on it's own... use an advance timing light and turn the dizzy till you get 6 deg BTDC...... i will need to double check tihs is 100% correct procedure!!!

yes the timing is dynamic, ie; it wil dance about all the time to stabilize the idle etc, but in basic settings iti will be locked to 6 deg BTDC with the throttle closed

ok, next, exhaust leaks and closed loop.... contary to popular opinion gases don't always flow out of the exhaust, at certain points in it's operation you get reversion back into the cylinder, but the favorite symptom is drawing air past the lamda probe on over-run via a leaky pipe/gasket....

ok, another thing to note about closed loop is that it does not operate all the time, generally speaking it only operates on idle and light load/throttle and low down in the rev range. at W OT it is running purely on the mapped settings...

one more question, is yours the one with a maf or a mao sensor? it's worth pulling it out and giving it a clean

Edited 6 February, 201016 yr by SkodaBoy

Fix the broken stud on the exhaust manifold!!

I must say, you've excelled with your use of vag-com.

PoloD.

ok, firstly it will re-adapt itself after a peroid of time, normally after a few miles or so, but you can perform a reset to put them back to the factory values, but i can't remember it off the top of my head at the moment (still at work now)

ref; 20%+-, i'm not 100% sure what the actual percentage is, but there is a very sticky grey area between long term and short term fuel trims...

you can adjust the ignition timing, with the engine running, use vag-com, go into measureing blocks, then group 1, and this check the 4th 'window' if it shows 00000000 then you can do it, if not then not all the pre-requisits have been met.. then go basic settings 04 group 1, then the engine should rise to an idle speed of about 1200rpm on it's own... use an advance timing light and turn the dizzy till you get 6 deg BTDC...... i will need to double check tihs is 100% correct procedure!!!

yes the timing is dynamic, ie; it wil dance about all the time to stabilize the idle etc, but in basic settings iti will be locked to 6 deg BTDC with the throttle closed

ok, next, exhaust leaks and closed loop.... contary to popular opinion gases don't always flow out of the exhaust, at certain points in it's operation you get reversion back into the cylinder, but the favorite symptom is drawing air past the lamda probe on over-run via a leaky pipe/gasket....

ok, another thing to note about closed loop is that it does not operate all the time, generally speaking it only operates on idle and light load/throttle and low down in the rev range. at W OT it is running purely on the mapped settings...

one more question, is yours the one with a maf or a mao sensor? it's worth pulling it out and giving it a clean

Ahh awesome. I understand what you mean by back-pressure now.. had heard of it before but wasn't sure what it meant lol

If you're able to find out the measuring block groups for fuel trims that would be mint, I will just see if any of the figures change when I make all the adjustments lol. If something changes then I will assume it's good and take it for a gas analysis

I haven't been able to get home to the car yet (which is soo frustrating lol) so I don't think I'll mess about with the timing just yet, but cheers for the info. My brain is already mashed with all this lambda info I've been trying to learn lmao

I've been told my set-up is MAP based, although Haynes manual only tells me where to locate the Intake Air Temperature. I'm assuming that this is coupled with a MAP sensor, but I'm only going on what I've been told on here. Either way, I will definately pull it out and give it a clean as I've suspected this might be a cause aswell.

Thanks a ton for all the help mate, will let you know how i get on

Fix the broken stud on the exhaust manifold!!

I must say, you've excelled with your use of vag-com.

PoloD.

Haha yeeah I'm just going to go with that mate and hope for the best!

And cheers dude, I love vag-com now - its awesome! Learnt so much from it and all the back-up help on CP for it lol saving up for the full version now

The lambda is working - we know that - and its located just downstream of the exhaust manifold. Now, the exhaust manifold is leaking. This is drawing over pure (clean) air into the exhaust gasses diluting them as they pass over the lambda - in effect, creating a falsely 'too lean' mixture. The lambda correctly directs an increase in fuelling.... sealing the gap in the manifold may well solve your problem!

Have you had an educated mechanic cast his/her eye over the engine? Internet diagnosis is difficult without seeing the offending engine.

Best of luck!

PoloD.

Edited 11 February, 201016 yr by PoloD