S2 Running rich.... Why?

[HJ2]

Hi all,


Last friday it was time for the dreaded MOT again. Everything was perfect, but the trend of running richer every year continues.
CO used to be 0.46% @ idle (0.5 is the max here), and a couple of years this was never a problem, even without a CAT it would be a steady 0.46%

In the last few years the CO% rised to 0.7 (!), all @ idle, and revving it would not drop the CO level at all.
Lambda shows also that the car is running rich (no surprise here!)

So my S2 is running rich and I want to find the root cause of this to fix it.
I will continuously update this first post as a reference for others
I allready done some things to get things better, but no result so far. I will not post this again (yet) as i want this topic to be complete for other poor buggers who suffer the same problem.


My car:
S2, No CAT, 2.25" downpipe and exhaust, Switcher

Symptoms:
1. Sooty exhaust
2. Exhaust gas smells rich
3. less mpg
4. High CO
5. Lambda off

Gas Analyser read outs (idle with hot engine):
CO = 0.692
Lambda = 1.129
HC = 112
CO2 = 13.30
O2 = 3.22

Gas Analyser read outs (2600 rpm-ish with hot engine):
CO = 0.790
Lambda = 1.073
HC = 70
CO2 = 13.26
O2 = 2.08

Identified potential causes to check so far:
1. Coolant temperature Sensor
2. Manifold Air Temperature Sensor
3. Thermostat
4. Oxygen sensor
5. Vacuum Leak
6. Exhaust leak
7. EGR Valve (can this cause rich running when it is stuck open?)
8. Partially blocked air filter
9. injector (Stuck open)
10. Wrong Thermostat
11. Base Timing off

Also read here: viewtopic.php?f=38&t=21004&p=279749&hilit=+paradise#p279749
And here: viewtopic.php?f=29&t=24735
Please let me know if i missed anything so far!

[GeoffSmith]

ELANSCAN!

[HJ2]

Gesundheit!

I will post a scan.
Presumably hot and @ idle?

[John_W]

I just solved rich running at idle after changing injectors on my SE by attention to the throttle bypass screw. It was about half a turn too far out. I had re-adjusted it and the CO pot "by ear" in service mode when I fitted the new injectors, but must have got it wrong. It was about half a turn too far out. Elanscan at idle showed the IAC reading and the injectors' pulse width and duty cycle too high compared with previous "normal" logs with the old injectors. I found that a log from cold was useful to see how the parameters changed as the engine warmed up to operating temperature, as it made it maybe more obvious than a hot idle log, although the differences were there when hot.

A couple of slight adjustments (half a turn of the bypass screw and quarter a turn of the CO pot) and all looks back to normal. (Idle's not exactly silky smooth, but it never has been!)

My symptoms were just an uneven idle at rest and "kangarooing" in traffic at idle. If you're reporting increased fuel consumption, your solution might not be as simple, but as Geoff alludes to an ElanScan log might point you in the right direction.

John

[HJ2]

Thanks John!

Unfortunately it’s an S2, so no pot on the intake to turn
Next step: Elanscan

[tb10]

It could be the Coolant Temperature Sensor? When mine went I used a tank of fuel in 100 miles....

[HJ2]

Yes, i write that in my initial post.
I have changed that earlier, without result. I will check everything again so i will measure resistance...

[GeoffSmith]

Run Elanscan and it will show you whether the CTS is working and will also show if the O2 sensor output is correct.

[HJ2]

I just now installed Elanscan + Cable drivers and did a quick check-run (7.7 seconds, stone cold engine) to see if it all works again.

One strange thing I spotted at first glance: I have a PROM error? (Fault 20) Perhaps the result of the Switcher?
Could this have to do anything with it? (don't think so)

And what is "Op. temp.- temp.: 58.0 °C" (this looks a tad high, whatever it is)
Also: Coolant raw 120 ?

Anyway. I'll do a full scan tomorrow from cold till hot + full boost

[John_W]

“Op. temp.- temp“ is just how many degC you’ve still got to go to warm the engine up to its operating temperature.
I’ve never found a use for “coolant raw” and never spent the time to work out what it is.
Prom id error doesn’t sound good - don’t know what that means.

[Simon_P]

“Op. temp.- temp“ is just how many degC you’ve still got to go to warm the engine up to its operating temperature.

now I know - thanks for that.

HJ,
You have a good list add to that barometric pressure compared to a barometer (failing that use the weather forecast). Also check that the MAT is similar to the CTS on a stone cold engine.
Hopefully Elan Scan will indicate something.

If elanscan shows everything is OK then I have a theory for you. What concerns me about your results is the lambda showing really lean. This is indicative of an exhaust leak suckling air into the exhaust. It might just be a loose joint which you should fix anyway, but.... Here's the theory! If it was a cracked or leaking manifold, it might be sucking air in upstream of the O2 sensor which would then read lean and put more fuel in giving you high lambda and high CO ... Just a theory but it looks like you have an exhaust leak and our manifolds are prone to crack - have you scraped your exhaust on a bump recently?

[HJ2]

Not recently, but i am on my third exhaust manifold right now.
So you definately are not talking rubbish!

[GeoffSmith]

I had a PROM error on 2/8/10. That was after a flat battery and it soon cleared.

Coolant temp is the one to look at.

Coolant raw is a strange beastie that has a negative slope with increasing temperature and a step change at about 50°C. I think that [geek] Coolant raw is the measured voltage at CTS which is fed from a fixed voltage via a resistor. The resistance of the CTS with temperature is non-linear and consequently so too is the CTS voltage. This is particularly so below 50°C and the transfer function to temperature is done by a second order polynomial. Above 50°C the gain of the ADC is increased for better sensitivity and the transfer function is linear.[/geek]

[GeoffSmith]

If the O2 sensor output is good (<0.5V to >4.0V) then Simon's theory sounds good to me. You should see something like this:

[HJ2]

The O2 signal looks similar Geof!

[lotusflasherman]

In almost 60 years of playing with internal combustion engines I have never heard of an exhaust leak 'sucking air in'. In my experience exhaust gas pressure is above atmospheric so it only comes out of the leak. Would somebody please educate me on this...

[Simon_P]

Phil,
I can't believe that you of all people and with almost 60 years.....
Exhaust gas isn't at constant pressure it comes in high and consequently low pressure pulses. I believe that this is the basis of cam overlap and exhaust design ie using the low pressure to scavenge the cylinder (already at partial vacuum). I'm no expert and I'm sure that someone more knowledge can explain better or explain if I'm wrong.
My car failed the MOT emissions because it had a loose joint that let air in. The air has no other way of getting in to HJs exhaust look at the AFR for that lambda and then contrast with the rich AFR for that CO.

[HJ2]

Hi Phil,

Let's assume you are correct,
What then could be other causes of my troubles?

[lotusflasherman]

Hi Phil,

Let's assume you are correct,
What then could be other causes of my troubles?

I know I'm correct.. I'll explain later ...

If you have an S2 (Closed Loop) with No Cat and the emissions are starting to wander I'd be looking at the Lamda Sensor output or maybe check EGR system is functioning correctly .. but first I'd use Elanscan.
I presume emissions were initially OK after you removed the Cat..

In the UK they will not test a car with a leaking exhaust as the Emissions Test is not valid so I reckon that's why Simon's car failed.

[HJ2]

Yep,

It's an S2, Closed loop, the O2 signal looks good in Elanscan.
Emmisions were fine for 2 years, then it started to creep every MOT to the current readings

I cleaned the EGR valve 3 years ago, behaps I should unplug and blank it now, to be on the safe side...
I had a cracked exhaust manifold 2 times, and it can be hard to see and detect as a leak sometimes, depending of the location of the crack...

IF there is a leak, then it will be in the manifold or in the flexible section after the downpipe.