Tuning for emmisions test

[luke2152]

I put my kit car through its IVA test (that is the UK test for amateur built vehicles) and it failed on the emissions test. I was allowed to tweek the map and have one more try. I just had a bit of a guess with my changes and improved it but still failed. I'll probably ask a local garage if I can pay for some time on their emissions gear and would love some advice on adjustments especially for the idle for getting it in line with the standards for my retest.
Details of my engine: mazda mx5 1.8 bp engine, MS2, td04 turbo, 3" sports cat that I cut off an unkown exhaust and welded in. Was in unknown used but visually goodcondition (that is to say there was no melting or black deposits on the honeycomb) when I put it in. No idle air control. Cooling fan triggered @ 95 degrees C (205f). Auto tuned the VE table especially the low load parts of it with an innovate wideband but had to remove it for the test because (don't laugh!) it has an edge with radius less then 2.5mm which presents a hazard to pedestrians....
Some details of the test (for those of you from abroad who don't have to deal with this test annually):
It is done in 2 stages. Fast idle which means bringing the engine to 2500-3000rpm under no load and holding it there for a few minutes (not sure exactly how long it was) and natural idle which is as you can imagine the car idling normally and rpm between 500 and 1500rpm. Time is allowed before the test for the engine and cat to thoroughly warm up.
Fast idle limits:
CO 0.3%
HC 200ppm
Lambda 0.97-1.03
Natural idle limits:
CO 0.5%

First try results: msq:

1sttry.msq

Fast idle:
CO 1.05% (fail)
HC 106ppm (pass)
Lambda 0.989 pass
natural idle:
CO 1.66% (fail)

After adjusting map - I was only guessing but I leaned the midrange low load area by 1AFR and retarded timing by 8 degrees which made it difficult to hold a steady rpm but did improve the fast idle results. I didn't get a chance to experiment further.
2nd try results: msq:

2ndtry.msq

Fast idle:
CO 0.2% (pass)
HC 21 (pass)
Lambda 0.994 (pass) (I was a little surprised that the lambda changed little after adjusting the AFR)
natural idle:
CO 2.11% (fail)

Attached is msq
Many thanks,
Luke

[turbo conversion]

Get you a gallon of 106 race fuel to 3 gallons of 93 pump gas and you will pass, it has worked for me many times.

[DaveEFI]

Must admit to not having found autotune much use for idle settings. And don't think it's recommended to use it for this.

There was a UK site which gave the sort of CO you'd expect to see at various AFRs, but it's sadly been down for ages. It was pretty well spot on for my non cat RV8. The factory spec for that is 1.5% CO and 14.1:1 AFR was the figure it gave for that. So that's what I set the idle at using my wideband. The MOT results show it in that ballpark.

I'd sort of guess it should pass the tests at 14.7:1 AFR (lambda1) - if you can get a reliable idle at that.

[TheSilverBuick]

Do they not check NOx? If not it should be cake to pass, just lean it out! Leaning it out drives the NOx through the roof, but will pass a bunch of O2 which should catalyze the CO to CO2 in the converter. Like wise, add the timing back in, it mostly effects NOx. Just measuring CO and HC's is just measuring the efficiency of the combustion process.

Alternatively, smog pumps typically are used to introduce O2 into the exhaust stream to catalyze with the CO.

[luke2152]

Do they not check NOx? If not it should be cake to pass, just lean it out! Leaning it out drives the NOx through the roof, but will pass a bunch of O2 which should catalyze the CO to CO2 in the converter. Like wise, add the timing back in, it mostly effects NOx. Just measuring CO and HC's is just measuring the efficiency of the combustion process.

Alternatively, smog pumps typically are used to introduce O2 into the exhaust stream to catalyze with the CO.

Correct - they don't check NOx. But they check the lambda so it can't be too lean.

Cheers for the replies guys I will see what I can do

[quan3165]

Can you increase your spark plug gap for the test? I imagine it is small since you are boosted. Just a thought.

[Yves]

I read somewhere that mixing in some alcohol to the tune of 10-15 does wonders.

[morrische]

Hi there

first of all it seems (at least to me) that EGO controll is disabled? Can you confirm that it is working? (otherwise rewrite the AFR table is useless)

Second also it seems to me that you passed high idle (wich is already fine) and you just have to work on the normal idle area (low rpm low MAP). try to even this area no big changes) and lean it a little bit out.

If you change the AFR table you should re-tune the VE table. The AFR tables are just for "correction" and not a main size in injectiontime calculations...........

Also Check the injection time in Tunerstudio. try to switch all accel enrichments of untill you have a more or less steady injection time...........Wich helps you to tune the VE map in idle area properly........

More ignition advance give the mixture more time to burn. (Normally during idle 10-15°)

And most important of all..... You need one of this Gas Analyzers...............

Where is lambda during normal Idle?