BMW E36 Touring MOT Emissions' Test Failure

BMW E36 Touring MOT Emissions' Test Failure

Author
Discussion

tobytronicstereophonic

Original Poster:

54 posts

69 months

Thursday 16th February 2023
quotequote all
Having finally completed my 1998 BMW's MOT work - rear w/bearing, front wishbone, brake light and exhaust emissions too high - today I went for the retest.

I knew all would be fine apart from the nagging doubt about the catalytic converter's condition as I'd been getting around 15mpg. I was worried that it had been damaged with all that unburt fuel passing through it...

A 1.8% CO on the initial test, I'd replaced an open-circuit coolant temp sensor and changed a stuck fuel pressure regulator. It was obviously an improvement as it was now idling better.

On today's retest, CO was a remarkably low 0.01%, but Lambda was now a fail at 1.6.

As I used to have my own garage business, the MOT people know me well and they tried to resolve it, rather than issuing another fail.
He thought it was going to be the slight exhaust leak on the back box. After sealing the small hole, it was still just about the same reading. He said it needs a Cat.

As I needed the car to go on holiday, I asked if I could try something:

I pulled the cover from the AFM and loosened the small screw that is set by Bosch and anti-tamper sealed with hard wax. This is the adjustment for the relationship between 'wiper' and carbon resistive 'track'. I nudged the adjustment to 'rich' and straight away, lambda was now a 'pass' but CO was 4.8% and predictably accompanied with now sky-high HC (hydrocarbons) was also a 'fail'...

An ever so tiny tweak towards lean brought things to close to a 'pass'. Another tweak - about one-thou' of an inch - instantly all three readings were just about as perfect as could ever be. It was - at least for me - a quite remarkable event, that such a tiny +/- adjustment made all of this 25 year old tech work together again in harmony.

Old cars can be a pain, but when you can get NOS parts for pennies, there are some plusses...

Whilst wrestling with this BMW last week, I had a funny turn. I've rented a workshop and I'm going to have a go at coming out of retirement. Wish me luck. I need it, as it was 22 years ago since I last closed the doors of my Citroen specialist business :-)



















stevemcs

8,993 posts

100 months

Thursday 16th February 2023
quotequote all
If the CO and HC are within and there are no air leaks why don't you try a lambda sensor?

SS427 Camaro

6,504 posts

177 months

Thursday 16th February 2023
quotequote all
tobytronicstereophonic said:
Having finally completed my 1998 BMW's MOT work - rear w/bearing, front wishbone, brake light and exhaust emissions too high - today I went for the retest.

I knew all would be fine apart from the nagging doubt about the catalytic converter's condition as I'd been getting around 15mpg. I was worried that it had been damaged with all that unburt fuel passing through it...

A 1.8% CO on the initial test, I'd replaced an open-circuit coolant temp sensor and changed a stuck fuel pressure regulator. It was obviously an improvement as it was now idling better.

On today's retest, CO was a remarkably low 0.01%, but Lambda was now a fail at 1.6.

As I used to have my own garage business, the MOT people know me well and they tried to resolve it, rather than issuing another fail.
He thought it was going to be the slight exhaust leak on the back box. After sealing the small hole, it was still just about the same reading. He said it needs a Cat.

As I needed the car to go on holiday, I asked if I could try something:

I pulled the cover from the AFM and loosened the small screw that is set by Bosch and anti-tamper sealed with hard wax. This is the adjustment for the relationship between 'wiper' and carbon resistive 'track'. I nudged the adjustment to 'rich' and straight away, lambda was now a 'pass' but CO was 4.8% and predictably accompanied with now sky-high HC (hydrocarbons) was also a 'fail'...

An ever so tiny tweak towards lean brought things to close to a 'pass'. Another tweak - about one-thou' of an inch - instantly all three readings were just about as perfect as could ever be. It was - at least for me - a quite remarkable event, that such a tiny +/- adjustment made all of this 25 year old tech work together again in harmony.

Old cars can be a pain, but when you can get NOS parts for pennies, there are some plusses...

Whilst wrestling with this BMW last week, I had a funny turn. I've rented a workshop and I'm going to have a go at coming out of retirement. Wish me luck. I need it, as it was 22 years ago since I last closed the doors of my Citroen specialist business :-)

Good luck and hope your new biz works out.

tobytronicstereophonic

Original Poster:

54 posts

69 months

Thursday 16th February 2023
quotequote all
It has a new Bosch oxygen sensor on it. The car is now fixed:
When I got back home tonight, with better lighting, I looked at the AFM vane screws and there is evidence it'd been taken apart before. When I looked with a magnifying glass, it now looks to be in the same position when it was made by Bosch. It had been messed with in the past.

If you look on YT there are quite a few videos showing how to 'compensate' for wear on vane type AFMs.
Bosch used a laser to trim the resistors on their AFMs to ensure each was was practically identical.

My point being that unless you have access to a three-way gas analyser, tuning* by ear/modifying stuff like this is a waste of time. Although I was already aware of this, I needed the car to 'pass'. I expected the AFM adjustment was a 'workround' and something else was amiss. But it was the final bit of the puzzle: somebody had altered the AFM vane settings in the past.


  • Unless it's injection timing on an XUD diesel. I used to bet that I could set one to 14 deg btdc +/- Citroen's tolerance on only a short roadtest, at the dealership I worked at. I only lost once.

stevieturbo

17,535 posts

254 months

Friday 17th February 2023
quotequote all
CO at 0.01 is far too low, Lambda far too high.

That shouts massively lean and perhaps also an air leak.

Fix those problems before going nuts thinking of buying a cat.

And take a mixture reading pre-cat to see exactly how that is.

If the MAF adjustment has helped....ok, but why did it need adjusted would be a good question, and now what are mixtures like across the board when driving ?