dabow

iTrader: (19)

well the car failed, but the car was set up via esl chips.

know i'm thinking of the apexi fc know. not really got to much cash to play with.

but this is how it failed:

fast idle test, you should be no higher than 0.30% the car was 2.32%
second fast idle test, again 0.30% the car was 3.53%
& natural idle test, should be no more than 0.50% & the car was 3.77%

will a map/different ecu bring these figers down to normal
& are the apexi's any good, any issues with them at all.

Gear Head

iTrader: (2)

Is your lambda sensor ok? try the cheap things first mate!

Jolly Green Monster

iTrader: (4)

esl uses the original ecu's lambda sensor setup to sort emissions and it is not the esl chip causing the problem imho..

as said check the lambda sensor is working.

Simon

dabow

iTrader: (19)

the lamba passed

fast idle test, limits 0.97 - 1.03. it was 1.00

its as if theres to much petrol, getting in there.
i havent used the car much this year. the first set of chips i had in the ecu, but with the 380 injectors when through perfect

with this map & 440cc injectors
its sky high

i'm not blaming esl.
just seems i need to find out why the emissions were so high

joz8968

iTrader: (13)

dabow, you have PM...

Scott.T@PolarPerformance

Have replied to your PM's
After reading your PM's and these posts it's still not clear to me whether the map you are running is set to use the 440's.
The ECU should be quite capable of running the correct AFR for the MOT as it uses lamda feedback. If the map has a 380cc injector constant and you are running 440's it may be an issue.
But I have seen a cars running 440's idle and cruise perfectly well even on a 380's map.
It ran a bit richer when on boost though i.e not under control of the closed loop lambda feedback.

Jolly Green Monster

iTrader: (4)

that doesn't mean the lambda sensor on the car is working fine

Simon

Bob Rawle

if lambda was 1 the sensor is working fine, so, do you have cats on the car ? If not then ......

bob

Alan Jeffery

The signal voltage from the sensor should be cycling if it's working correctly, assuming it's the early three wire zirconia version.
To clear this one up, because not everybody understands it, Lambda is an industry term to determine the best air fuel ratio level for cat efficiency. Lambda One is 14.7 to 1 AFR. The Lambda sensor is designed to control the mixture as far as possible around 14.7 to 1 or Lambda One. The sensor will only control the mixture when it is in closed loop, roughly idle settings up to medium load.
3 % CO doesn't go with a correctly cycling lambda sensor, so I'd check that first with a voltmeter. You should have a 12v and earth for the heater, and between .1 and .9 of a volt cycling from the signal wire, usually the white one on the car end of the multi plug.

Jolly Green Monster

iTrader: (4)

agreed

Alan Jeffery

To take that test to its conclusion BTW, if the voltage is at .9 constant it means the mix is rich for some other reason than a faulty lambda sensor. If it's steady at around .4 of a volt, fit a new sensor as it's duff. A lean mix would give a constant low voltage, .1 or less.

dabow

iTrader: (19)

the lambda does have three wires. red/white/black

i've sent off for another one just in case, to which has 2 white & a black
the car does have a cat

so i need to test between the red & the black & then black & white
should the car be running when testing?

MartynJ

Connect the black wire of the new sensor to the white wire of the old multiplug , the other 2 wires don't matter as they just form a loop..
You need to check between the Black signal wire at the lambda end and ground , not the other 2 wires...
Yes the car needs to be running and you need to allow a couple of minutes warming up time for the lambda to begin to function...
If the lambda voltage is cycling between 0.1 and 0.9 of a volt then you have no mixture problems , this isn't the end of it though...
It is possible to have to high a CO figure despite mixture being correct , this will normally be down to the cat being goosed...
Have a look on your fail certificate and post the O2 figure out of interest , this will tell me if you have any minor exhaust leaks which can have an effect on mixture figures...

ZEN Performance

Lambda 1 on the test doesn't mean the sensor is working, just that the cat is working well. But if the pre cat lambda is say 0.9lambda, you might get it cleaned up to nearer 1, but with excess HC and CO.

You do have a cat don't you?

dabow

iTrader: (19)

yep

i may be onto something else
the map i had, may have been wrong

i have sti v4 pistons, & 440cc injectors
if i say put my old chips in to which were set up for the 380cc injectors.
the car runs smoother.

so all the may need, is a proper map to sort the fueling out
i've been speeking to one of the tuners on here who used to deal with these ep-roms

i'm first going to change the lambda senser first, just in case
but the senser ain't that old really as the sports car ain't

i've had a car fail becouse the lambda was not working. but when changed, it diden't make much difference to the reading. in the end, it was the injectors were to big in the car. the ecu was tell the injectors to open & close. but when open, they were letting more fuel in becouse they were bigger

you tuners will no more about this, i was not said exactly like that. but to the same effect.

if the lambda is working fine. it may just need this either live map i keep hearing about. or an ecu & having that mapped

is it posable to have the reading needed set
say, if the co level is 0.30% have it set to either 0.20/0.25%

Bob Rawle

Voltage won't cycle if the ecu isn't cycling it, the sensor only produces an output, the ecu adding and subtracting fuel causes the cycling. The sensor doesn't control anything it merely acts as an input to the ecu to allow that to do the work, since we want to be clear.

If the car has a cat then you really do need to measure the afr pre-cat as Paul says, bear in mind though that the ecu's have a plus/minus 25% short term trim and a plus/minus 15%(ish) long term trim in addition so things have to be a very long way out for that not to get it right injector change or not.

The lambda heater could be failed which woul cause it to read with an offset when cooler.

bob

The rookie

iTrader: (1)

If the engine out was Lambda one, then the precat CO should only be 0.67%, I don't believe it was at Lambda 1 frankly.

Sorry Paul/Bob, a cat can't change the true Lambda, only if your using a wideband OXYGEN sensor to infer Lambda (so any exhaust plug in 'lambda meter') will it be changed, the spindt calculation used in 4 gas (CO/CO2/HC and O2) analysers at MOT stations isn't affected by sampling before/after a catalyst.

Simon

Alan Jeffery

The point of mentioning the lambda sensor is to illustrate a simple method of figuring out if a zirconia sensor is faulty or not. I think we all know that the sensor won't help you much without ECU input! Splitting hairs won't help the guy fix his car, now will it?
Rookie is righty...

MartynJ

Sorry Paul , lambda 1 has nothing at all to with the cat...
You can have a car idle and run at lamdba 1 even if not fitted with a cat , the cats job is to remove the CO and HC content of the emissions...
A CO figure of below 0.3 and HC of below 200 shows that the cat is effecient , although this is dependant on mixture being correct....

dabow

iTrader: (19)

bottom line. the car could do with a rolling road to find out what has gone off

if all is working fine, senser ect. then its down to the ecu, telling the system what to do.

which brings the point of my thread up.
yes a map should sort this emission issue out

i would like to thank all that have given ideas, & what to check next.
i've got a new lamba senser coming + other senser.

after this live map. the car should fly through its mot
touch wood.

ZEN Performance

The lambda measurement WILL be altered by the cat. While you may assume that lambda <1 (pre cat) means no spare 02 in the exhaust stream, this is not necessarily the case. You can see this effect easily when putting cars on a 4 gas with a wideband lambda monitoring what's going on. Remembering of course that an early ECU will cycle the fuelling up and down either side of lambda 1 ranging from about 0.95 to 1.05, while the post cat lambda will be much closer to 1. The effect will also depend on the ignition timing, the engine (leaking bores, leaking valves, oil contamination).

I've seen a number of cars with dead or slow lambda sensors causing high CO while the lambda as reported by the 4 gas is pretty close to bang on.

If the ECU is okay, and the lambda is really around 1, then the cat is dead, or not hot enough or both.

dabow

iTrader: (19)

the cat is not even a year old. & is still under warrenty
the lamba is not that old either, & the car has not been used much last year
i think i did a total of 3/4k

the ecu is perfect bar the chips i put in.
the chip were custom mapped but i have sti pistons & 440cc injectors in, the car was not rolling roadfed either to get the map

with the other mods on the car too, as some has said. the car could do with a rolling road to determin what happening.

at the same time, being an early classic. i need either an uprated euc to be able to be mapped or this live map what some are talking about

i'll try the car again when i've changed the lamba again.
but dont think it will bring it down as much as i needs too

i've had this car for just over two years. & loads of things i've picked up
but theres loads to lean with these cars

i'll keep you all postard regards the new lambda senser anyways

The rookie

iTrader: (1)

Poppycock - total twaddle......

Yes there is spare oxygen at Lambda one, just as their should be a CO of circa 0.67%

BUT Lambda relates to AFR (and it value relative to stoichoimetry - which is fuel dependant), that doesn't change from before any combustion exists (when air is about 20% oxygen) to after, (usually about 0.25% - but with a misfire on one cylinder for example it would be 5% 02 but AFR and hence Lambda are the same!) to post cat (hopefully no oxygen), and a proper analyser (you'll need a burner for the liquid fuel for pre combustion) will tell you that.

Only if you use a 'Lambda meter ' which just uses an oxygen sensor will its 'inferred' lambda be changed by a passing the same AFR through a catalyst.

Lambda 1 will give (out of engine) 0.67% CO, which is a fail, if you have a partial misfire it could be higher still, but not likley to be 3+%, which implies an Oxygen sensor failure or other realted failure that disables closed loop fuelling.

Don't screw your good rep here Paul.....This is my day job!

Simon

Scott.T@PolarPerformance

Dave,

In response to your PM. No the car cannot be mapped to have a CO of 0.2-0.25.

If you check back through your MOT's you will see that the CO is quite random and can fall anywhere between 0-0.3 from year to year depending on several factors.
The main one being how hot the CAT is during the test.

The car can be mapped to run with the 440's (which I think it already is, according to you rposts & PM's), the fuelling set and the injector constant adjusted along with a whole host of other things.
But when running closed loop the ECU will be running in response to the feedback from the lambda sensor and using it's own method to set the fueling to Lambda 1 (AFR 14.7).
As mentioned above this is the most efficient AFR and it is down to the CAT to remove the CO and HC.

If you car is hunting at idle or having several misfires, this will effect the CO and HC content. So plugs, filter and check for air leaks.

Leaking exhausts can also effect the results. If you have a poor gasket anywhere on the exhaust then this can effect the result. i.e split up-pipe, cracked headers or blowing downpipe gasket etc....

As I also mentioned to you I had a sports cat and it only just scrapped through the MOT, this was wrapped to maintain heat and 'YES' it was from the same place you got yours.
Hence I ditched it as could feel the restriction and went back to the decat and put the centre cat on for MOT's (15 min job)

I would suggest you do any or all of the following :

1. Fit a new Lambda
2. Fit a decat Centre Section aswell, if available
3. Get the MOT centre to take the revs to about 4,000rpm for about 3-4 minutes to ignite the cat. Then start the MOT emissions test.

dabow

iTrader: (19)

the centre section is decated, & always has
its a down pipe sports cat i have.

the lambda is on its way. too
i did run the car before i got the car to the mot station. but they had probs & the car was stood.

all the exhurst gaskets are new not even a year ago, theres no leaks at all on there. the headers was also changed to sti v2 ones. as was an uprated uppipe too.

again i used new gaskets.
the pistons were cleaned as well as the heads, i cleaned every thing.
i was tempted to put a set of 380cc injectors back in & replace the chips to suit as i still have these.

the cat still looks ok, not that black either
but if i get this car/ecu mapped.

i only asked about the 0.2/0.25% as a guide ect, i know they will be higher/lower

but will the emissions be in the limits to pass
trhe things i've fitted since i had the second set of chips put in is, the front mount/ an oil cooler/ an uprated oil pump & put the oil presure gauge to the top of the engine.

the fm will need to be set up i know. could this have corsed this at all
but if i change the lamba again, & its the same.
what then, if a map wont sort the emissions out

do i jusyt scap the car or what?

Scott.T@PolarPerformance

I know the cat is in the downpipe, it's a scoobyworld sportscat.

I was suggesting you get hold of a catted centre which will help it through.
Many get through with just the centre, infact my centre results were better then the sports cat in the downpipe that I got rid of.

As mentioned via PM putting 440's in doesn't make a great deal of difference to closed loop fueling, the ECU seems to cope quite well with it.
So I don't think thats an issue. It will run a little richer on boost, but not by alot.

The FMIC should not effect closed loop running unless it has any air leaks and the ECU is struggling to maintain the AFR.

dabow

iTrader: (19)

before i deside anything else, i'll change the lamba

take it back with a hot cat, then see what difference there is

but if theres nothing changed. then where do i go from there?
apart from a dodgy mot

scoobyDAZZA

centre section cat
get the car/catS as hot as possible and test immediately

The rookie

iTrader: (1)

If its running too rich (and 3.8% implies about 13.25:1) then a red hot and about to melt cat' won't get it to pass....

Simon

dabow

iTrader: (19)

as advised.

i'm testing the lamba today
if its not that, im stumped.