Yep, this time it's me. The engine is a 1.2 MPI turbo with lowered compression ratio by a sandwich plate, no cat. Engine runs fine, everything seems to be in order, AFRs on my display, all that. But it failed miserably on the CO test today. At idle it has more than 1.0 out of allowed 0.50, at high idle 0.80 out of allowed 0.30. They were kind enough to let me play with the ignition timing on my ECU and see if I can get it any better, but this brought me nowhere.

Now, apart from the obvious advice that I probably should put the cat in (I don't think it should fail so bad without it, or should it?), I was also thinking if I possibly have setup the vernier pulley wrong. I did everything by the book and the engine runs smooth as I said. But I could have as well messed something up, and it might be a couple of degrees off. So my question is, which way am I likely to be off with the pulley? I will get another chance for one test tomorrow, so I can try to change the pulley settings before that, I just don't really know which way. I want to change it a bit and see if that brings me anywhere better.

EDIT: Am I right in thinking that I should retard a bit, so that exhaust fumes do not start to escape so early, ?

what ecu you running? is the mpi timed at all? sequential? semi sequential? batch fire? alternating?

Ive had the same problems in the past even before the turbo.

Think he had a map it then wrote a chip for the standard ECu.

Ming

craigb244 said:

what ecu you running? is the mpi timed at all? sequential? semi sequential? batch fire? alternating?

Ive had the same problems in the past even before the turbo.

Click to expand...

It is a stock ECU reprogrammed and remapped for Turbo. This part is done properly, you have to trust me here I had three tests today, one with absolute factory settings for ignition, one slightly advanced (well, for the map areas where they test), and one even more advanced, actually here I used ignition values from your map Craig. This leads absolutely nowhere. CO stays at ca. 1.0%, lambda is 1.16 on idle, 1.06 on fast idle. So I assume that's not it, at least not entirely. According to ECU diagnostics and everything I can measure everything is OK.

I just came back from the car, I retarded the cam timings on the vernier pulley by 4 degrees. Visibly nothing changed, which actually brings hope. I also thing I smell less crap now, the guy at the garage before said he does not need to run the machine, he can smell that it is really bad. My plan is to see what the emissions are now, and if it got any better I will just ask them to play with the pulley around until they are satisfied. If it is still bad, I guess I will have to incorporate the cat, but I know from good sources (here) that these engines in principle should be fine without it.

OK my friends, question for both guys in NL and UK. I am googling and googling and to no avail. What I just found out is that the norms I am failing (CO max 0.5% at idle and 0.3% at 2000-3000 RPM, lambda between 0.97 and 1.03), at least in Poland, are required for cars first registered after 1st July 1995. Anybody have any faintest idea for what registration date this is required in NL (but I am also interested in UK for comparison).

All this because my car was first registered in May 1995, so I am thinking about an easy way out of this...

Well, changing cam timings also didn't do much good. So it seems that it is a cat time for me...

can you get methanol where you are? some petrol stations also sell it as e85

my dads sprint bike is run on methanol. when it comes to mot time i pour a gallon into a nearly empty tank with about a gallon of premium unleaded. car passed emissions easy like that. no cat or anything.

but your right... in theory the car should pass without a cat.
as your emissions seem far out i doubt a cat will fix it (my mate who does my mots says they only make a small change.)

craigb244 said:

but your right... in theory the car should pass without a cat.
as your emissions seem far out i doubt a cat will fix it (my mate who does my mots says they only make a small change.)

Click to expand...

Well, the cats are there for a reason, so I trust they should make it better. I read somewhere that in some cases they bring CO down by 80%. Moreover, it is clear to me that on fast idle I have too much O2 (lambda 1.06), and too much of CO, which the cat should together convert to CO2.

Ethanol I think is not an option around here. So if anybody still has a good idea what this could be caused by, I am still listening. Could for e.g. colder plugs be a problem?

EDIT: oh, now that I read about cold spark plugs, this may as well be the problem.

woj said:

EDIT: oh, now that I read about cold spark plugs, this may as well be the problem.

Click to expand...

Well, that did not change things either . Tomorrow I have an appointment with turbo guys in an engineering shop close to where I live and we will see what can be done.

keep us updated with your finds mate

Lambda 1.16 at idle seems very lean, about 17:1 , I'm wondering if you're getting lean misfire causing the poor combustion and high CO.

Does it idle smoothly?

Just to add to the above, if it was me I would put the cam timing back, make sure there's sensible values in the idle region of the ign map, then adjust the fuelling till it's stoich and go from there.

54hp said:

Just to add to the above, if it was me I would put the cam timing back, make sure there's sensible values in the idle region of the ign map, then adjust the fuelling till it's stoich and go from there.

Click to expand...

I did put the timing back, and on my wideband display it says stoich. And this with both lambda sensors, one stock one, one emulated signal from the wb module. Also today on test, 2000-3000 actually had lambda 1.03, which is a (narrow) pass. But the CO is still high. At idle they do not care about the lambda I was told, CO is what matters.

Setting total factory ignition did not do anything either. All this is very suspicious. The car idles fine.

Tomorrow I have an appointment at an engineering shop that builds turbo systems and fabricate manifolds/exhausts. I will talk to them and see what to do next.

Meanwhile I also have some woes with headlamps for the MOT, but this I am not even going to bore you with.

Found an interesting read:

http://www.autoshop101.com/forms/h56.pdf

For me the conclusions are that lean at idle drops CO, which is not the case for me. Also, a cat drastically drops down CO at stoich, by about 80%. This figure I also found elsewhere. There are also mentionings of faulty injectors (probable, but this would show on my AFR readings), and fuel in oil, which is actually something I had a couple of weeks ago after an unfortunate ECU test, but since then the oil had been changed. So still clueless.

OK, talked to a guy at the engineering shop, he looked at my results, listened about what I have done to the car and said there is now way I can drop this CO emission down without a cat. They will fix it for me, but only in two weeks time, so for now I have to keep a low profile

OK folks, cat in (see picture), CO emissions dropped down to the floor MOT tomorrow. But there is one little bit that is confusing / unsolved. Goes like this:

We hooked the car up to the exhaust emissions tester, CO is very fine, but lambda on fast idle outside the range (ca. 1.08). First small remapping, but to no success, closed loop kicks back to 1.08 all the time. Careful look, seriously leaking exhaust in the back, we welded it up, lambda dropped, but still down to 1.04-1.05 (it has to be 1.03). I solved the problem by switching the ECU narrow lambda signal with wideband lambda emulation and changing the stoich point to AFR 14.4. This makes the lambda OK, although rises the CO a bit (but still within limits). It will make it OK for the MOT, but it should not be like this. Any ideas (apart from further leaks, but we eliminated that I think)...

Good luck mate I am stumped with mine thinking new cat and lambda

Well, this is just f....ing unbelievable. The car is very OK for the MOT, no problems, quoting the statement about my emissions "This is just beatiful.". But, the central registration computer system refused to acknowledge my car / its registration / whatever, and issue new MOT papers :bang: :bang: Now I have to wait till they solve whatever the problem is and still without the car till next week

the central registration computer system refused to acknowledge my car / its registration / whatever,

Click to expand...

This happened to my car , the MOT station asked me for my V5 to check it over so i gave them my V5 and my last MOT cert and they managed to sort it out within an hr and i drove away with 12 months MOT

OK, paper work sorted out, though still not clear what the problem was / is... Anyhow :slayer: :slayer: :slayer: :slayer: now I can go and do my car some harm, hopefully not myself :devil:

On this occasion after my experience I would like to put two doubts for discussion and for future turbo generation:

1. I do not believe you can legally (without cheating by using methanol, redex, etc.) push the car through the MOT without a cat. In my case the emissions differences were huge and I just do not see it happening with the cat out...

2. The other thing is the overall claim that turbo by itself provides enough back pressure. Well the engine works fine, that is correct, but now with the cat in I got even smoother idle and cleaner engine sound. My impression is that the cat's back pressure plus turbo back pressure is what my engine really needed.