my girlfriend has just taken my everyday car back for a retest this morning and it has failed again on emissions.

since its first fail i changed the cat and rear exhaust sections from a car i've just stripped that passed the mot with flying colours six months earlier (i have the emissions printout from the test), so i'm pretty sure it isn't the cat.

i bought some new plugs but didn't have time to fit them last night and am now thinking of swapping the lamda sensor over from the spare engine i have in the garage ( i don't have time to swap the whole engine over and don't really want the hassle of having to do this either).

I fitted new exhaust gaskets so am pretty sure it isn't leaking anywhere, certainly no sound or blowing when running your hands around it.

the car idles very lumpy so something isn't right, anyone got any other ideas of what sensors might be faulty? I've only got til saturday morning on the retest so only a couple of hours light in the evenings to try and sort it

Car is a 1.4 pug 106. The car failed today on lamda, reading 1.06, limit is 1.03 i understand, also failed on hydrocarbons i think, but not got the figures to hand.

thanks,

Ned.


[Edited on 22/9/05 by ned]

Ned

I used to run a 106XSi with the 1.6 injection engine. This used to regularly fail it's emissions tests at MOT time.

In the end I used to put a brand new air filter in on the morning of the test. That seemed to do the trick. Mine was a GpN car and was very sensitive to fuel changes and filters. Perhaps running it on lower spec fuel, instead of Optimax or BP Ultimate might help?

HTH

Mike

Ned,

What about the Lambda sensor! sounds like it could be that.

Rich.

Don't know much about this engine/car. Does it run an airflow meter?

Sometimes cars with airflow meters have problems. The hotwires can get dirty/gummed up and it throws the fueling/emmisions out.

If you can, swap out the AFM with another unit and get them to run it up on the emissions machine again to eliminate it.

quote:

---

Originally posted by shortie
Ned,

What about the Lambda sensor! sounds like it could be that.

Rich.

---

Mike,

I only ever run it on standard unleaded, someone had suggested running on superplus would be better as the higher octane would give a cleaner burn?! will add a new filter to my list..

cheers,

Ned.

just got off the phone to the other half and have the figures (failure points):

last weeks test:

1st fast idle (2450rpm):
lamda 1.09, allowed 0.97-1.03
c02 0.38, allowed 0.3

2nd fast idle (2386rpm)
lamda 1.04
hydrocarbons 102 (pass -allowed upto 200)
c02 0.53

This mornings retest with cat changed:

1st fast idle (2446rpm):
c02 0.04
lamda 1.06

2nd fast idle (2806rpm)
c02 6.27
hyrdocarbons 209
lamda 0.83

the second fast idle today is completely screwed and far worse than last weeks.

Ned.

i shall change the lamda sensor, spark plugs and air filter this evening, but not likely to get tomorrow off work so will have to wait til saturday morning for the retest.

[Edited on 22/9/05 by ned]

almost 100% will be the sensor.. airflow meters very rarely will cause it to fail with those kind of figures, also make sure the exhaust is very well sealed.. some rear silencers have a small hole in them to let out water.. i have had to block them up in the past to get over high lamda

lambda sensors get slower with age -- however it wouldn't be my first suspect mainly because they can be a swine to change. Checking a lambda probe without an ossiclloscope using a multimeter is possible --- check the heater element for voltage, earth and ressistance, check the signal to see it is "switching" up and down rapidily around the lambda voltage.

I'd be concerned about the "lumpy" idle Ned. First place to look is for a vacuum leak -- check all the lines and connections, if anything looks soft or crumbly replace it.

If it's a MAF car it worth popping the airflow sensor out and having a look to see if it's gungy. As mentioned above they can get a bit of oily film on them and act up -- no sense replacing it (if it's dud the car won't run at all) get a spray tin of air intake cleaner and some cotton swabs and give it a thorough (but gentle) washing.

Check the airpipe between the MAF and the TB, any airleaks at all in this piping will give you no end of grief and the moulded accordion hoses on many cars can crack at the seams.

Give the TB butterfly and body a really good dousing with the air intake cleaner, hold the butterfly open and clean it thoroughly with a rag and the air intake cleaning solution. It's worth unscrewing the idle air control solenoid and cleaning it off the car. No point being gentle here, scrub away and get it all as clean as possible.

Also, check your PCV valve, it should have a nice loud metallic rattle when you shake it -- they can get gummed up and stick open or closed, will throw your HC's for a loop as the engine is sucking in oil fumes. If gummed up it can be cleaned with the intake spray -- not usually worth the bother as they're quite inexpensive.

If the car has an EGR valve, test it by hooking up a hand vacuum pump to it and applying full vacuum to it while the car idles -- if the car falls flat on it's face and stalls or threatens to stall this is a good sign, if it just gets a little bit lumpier or no change the EGR is bad -- this test is not definitive, a small leak will cause a lumpy idle on its own and can be tough to diagnose. Depending on the style of valve you can fit a block-off plate snipped out of a lager tin where the gasket goes -- this will stop all leaks and, nudge-nudge, is only for "diagnostic" purposes. The valves, if bad are usually quite expensive, and while blocking them off is frowned upon by those in authority, it can help you pick up a bit of power and keeps your intake cleaner.

Finally, if the idle is still lumpy consider changinging the HT leads and check for an intake manifold leak (while idling, wave an unlit acetylene torch (just acetylene, no oxygen) around all the gasket faces, the intake sensor bungs and the TB shaft seals -- basically anywhere the system could develop a small air leak -- if the idle smooths out or increases as you do this you've found your leak. In a pinch a small squirt bottle of water can be used in place of the acetylene -- the idle won't increase but will very briefly smooth out when the water "plugs" the air-leak.

Cheers, Ted

thanks for the pointers, I will have a look tonight. I have a spare engine out of an identical car i stripped so have a source of spare pipes and sensors.

thanks all,

Ned.

i've changed the plugs and got an air filter ready to go in, but would you believe the lamda sensors have different bloody plugs on them! one is a k reg, the other was an l reg, same colour, engine, trim etc etc. nuisance. i don't even havea soldering iron or crimping kit to change the connectors over

thats assuming this solves the problem. the inlet system has different connectors on all the sensors, so i might be better off swapping the hole inlet manifold over, but then i'll need lots of sundry items and hope the wiring is compatible.

why is nothing ever simple?

looks like i'm gonna have to forgo the retest and fork out for another mot after shedding out for a cripming kit or another lamda sensor

[Edited on 22/9/05 by ned]

would that 'emission reducer' additive help at all? Ive used it in the past to get through emission tests. not sure how it works or if it is that effective though.

are the lambda sensors interchangeable? is there a chance they have different plugs for a reason?

atb

steve

quote:

---

Originally posted by steve_gus
are the lambda sensors interchangeable? is there a chance they have different plugs for a reason?

---

its lambda you toilet

thanks joel

Most lambda sensors are universal 1,2,3 or 4 wires -- if it is a universal type it will a white wire, if it has an earth wire it will be black, the heater connections are grey.

White = earth
Black = earth
Grey = heater power
Grey = heater earth


If the wires are any other colour (such as purple) chances are it is a Zirconia or special type and a universal one won't do.

has the car had timing belt replaced and is it timed up correctly this would also cause rough idle and emmissions to be all screwed up had it before on a clio timing was only slightly out relpaced belt retimed and tensioned emmisssions ok after

mot retest in the morning, have done the following now since the first fail:

had the airbox off and cleaned it inside and out
had the throttle body off and cleaned it
fitted new air filter
swapped the airbox air sensor for a spare (looked cleaner, from known good car)
fitted new spark plugs (old ones were different type and one was odd - completely different to the rest)
put £20 worth of optimax in
fitted spare cat from known good car
fitted replacement lambda sensor from known good car

btw the lambda sensor had 4 wires, 1 black, 1 grey and 2 white. i hope i got the connectors the right way round (only the two white ones that i couldn't tell apart, but guessed from the layout of the connectors), the dash warning light isn't coming on so i hope thats a good thing. its got bosch stamped on the side of it.

I also have some stp petrol treatment/engine cleaner stuff that i was going to pop in it in the morning which is supposed to help with emissions aswell. Its a 12 mile drive to the place i'm mot'ing it.

thanks to everyone who's offered suggestions so far. if it fails tomorrow i will just give it to a garage to sort it out as i've run out of time, patience and ideas!

Ned.


[Edited on 23/9/05 by ned]

car has failed again

the items changed since the last retest have made no difference at all to the figures.

The lamda is low, the hc and co2 are high. The lambda is constantly changing up and down between 0.87 and 0.95 at 2500rpmish (what they test at) the tester thought it must be something telling the ecu to richen and weaken the fueling.

The only other things I could do is change the map sensor and reset or swap the ecu (Assuming the other sensors i've already replacd are ok) but its 25 miles return trip to the mot place and i don't have time to do it now.

I'll have to book it into somewhere with some diagnostics stuff to get it sorted

thanks again for the help.

Ned.

[Edited on 24/9/05 by ned]

quote:

---

Originally posted by ned
car has failed again

---

Thanks Ted, I do have a brand new cat but was reluctant to fit it seeing as the one i had from a car i broke was only a few months old and i had the previous mot printout.

it appears a dodgy connection in the dash that affected the fuel and temp gauges may also have been misleading me with the engine warning liht as on the way back from teh last mot retest it started coming on.

the car is now at a local 'friendly' garage being dealt with. I'll let you know what it turns out to be, probably so you can say 'told you so'!

though seriously, thanks for your suggestions, appreciate it.

Ned.

Ooh, sorry Ned!

I'd forgotten all about the temp sender...

If it's faulty it can definitely cause you to run rich as the computer thinks the car is still warming up. Mine died a few months back flooding the car and leaving me stranded.

Cheers, Ted

The lambda reading on the mot gas analyser print out is different from the reading you get putting a voltmeter on to the lambda sensor.
The lambda test is essentially to see how well the ecu is controlling the emmisions --- in the early days it was possible to get a dirty car through by putting a hole in the exhaust to dillute the exhau gasses.

I would check carefully the inlet and exhaust system for leaks.

Ned. My MOT man is prone to forgetting to test emissions if bought a pint. Quite lucky really when you're running big carbs!

Want his contact details?

Chris

can you believe it!

left the car with my friendly mechanic last week and again this week. checked and changed most sensors (between jobs, no rush, so it doesn't cost me the earth), checked exhaust for air leaks, swapped ecu's, checked fuel pressure regulator, throttle pot and loads of other things. Just spoke to him and they've discovered timing is 1 tooth out. i can't believe it can be something this simple, i didn't think the engine would run if it was a tooth out, i would have thought valve clearances would be minimal on these engines. ho hum.

i think andy lancaster even suggested this and i said it couldn't possibly be that! how annoying!

everyone i spoke to said change the cat and the lambda sensor! to be fair they are probably the normal things to go wrong. just my bleedin luck.