v6 cf3 emission fail...where to start? - Alfa Romeo Forum
You are currently unregistered, register for more features.    
Reply
 
Thread Tools
(Post Link) post #1 of 43 Old 07-12-12 Thread Starter
Status: -
AO Member
 
Join Date: Aug 2004
Location: United Kingdom
County: West Yorkshire
Posts: 56
v6 cf3 emission fail...where to start?

My v6 top end rebuild is now complete Just the hurdle of getting an MOT.

I replaced both head gaskets (ful gasket kit so manifold etc gaskets also), valve stem oil seals and both the '1st stage' lambda sensors , front and rear (bosch, exactly as original).

MOT test today, passed on all counts except for emissions as per below:

Fast Idle

CO max 0.3 actual 1.28 FAIL
HC max 200 actual 74 PASS
Lambda 0.97-1.03 actual 0.990 PASS

2nd Fast Idle

CO max 0.3 actual 1.593 FAIL
HC max 200 actual 42 PASS
Lambda 0.97-1.03 actual 0.964 FAIL

Natural Idle

CO max 0.5 actual 2.461 FAIL

So any advice on where to look next gratefully received. Being a CF3 model there are 4 lambdas and 4 cats...

My instinct is an air leak between the first and 2nd stage cats (those flexy pipes??) Any past experiences (and numbers) would be a great help.

This rebuild has been long and challenging...I'm not sure I'll do it in a hurry again. But the 10 mile drive to the MOT station was sheer pleasure. The beast has never gone so well in its life, so at the moment this MOT hiccup is only feeling like a minor hurdle...

Thanks as always for any help,

Mike

Last edited by miked; 15-12-12 at 00:23.
miked is offline  
Sponsored Links
Advertisement
 
Status: GTA facelift + engine rebuild. In progress
Club Member
Membro Premio
 
Joss's Avatar
 
Club Member Number: 1014
Join Date: Sep 2005
Location: United Kingdom
County: Dorset
Posts: 19,112
Garage
I'd say it needs to be run in a bit more. Hopefully settle down
Joss is offline  
Status: I am also Keith......
AO Member
 
kam325i's Avatar
 
Join Date: Sep 2012
Location: United Kingdom
County: Shropshire
Posts: 565
High CO usually relates to the lambda sensor or the cat.
kam325i is offline  
(Post Link) post #4 of 43 Old 08-12-12 Thread Starter
Status: -
AO Member
 
Join Date: Aug 2004
Location: United Kingdom
County: West Yorkshire
Posts: 56
...Been doing a bit more research, Bosch lists the 1st sensor as 'control' and the second as 'diagnostic'. This has worried me a tad as it implies that the first sensor does all the real work, so as both '1st' sensors are new why do I have a problem?

Anyone have a fuller understanding of how all the 4 lambdas work together on a cf3 v6?

I'm gonna check security of inlet and exhaust manifolds over the weekend, though I know they were done up good and tight...

...If I find nothing I'll probably get both exhaust flexis changed on monday, and if that does not work then 2 genuine bosch universal lambdas for the 3rd and 4th lambdas (mainly to avoid having to rethread those long lambda leads round the back of the engine!)

Please jump in and advise any other/better ideas as I'm stabbing in the dark a bit with this...

Cheers,

Mike
miked is offline  
(Post Link) post #5 of 43 Old 08-12-12 Thread Starter
Status: -
AO Member
 
Join Date: Aug 2004
Location: United Kingdom
County: West Yorkshire
Posts: 56
...More reading has revealed...

Apparently the 2nd lambda is only there to confirm that the 1st cat is working correctly (can anyone confirm this?)

So seems the problem is right back at the 1st lambda or earlier. Surely an air leak after the MAF would cause a lean mixture though? Mine is too rich ?? (would introducing a deliberate small air leak help ?!!...only half joking)

Could that point to the MAF itself?

I'm gonna make up some lambda test plugs to monitor the lambda voltages with them still in circuit to the ecu...will post results hopefully tomorrow.

If a small bit of air got in at the exhaust manifold before the lambda would that give these readings?

Sorry loads more questions, i'm just trying to dump my mental ramblings 'on paper' in the hope somebody will see the truth in there somewhere!

Cheers Again,

Mike
miked is offline  
Status: Busy busy busy!
Club Member
Membro Premio
 
Pud237's Avatar
 
Club Member Number: 71
Join Date: Apr 2006
Location: United Kingdom
County: Cumbria
Posts: 44,801
OK this should help:-

There are 4 lambdas in total on a V6 CF3. There is one situated in each exhaust manifold, one on the front bank (4,5,6) and one on the rear bank (1,2,3). The exhaust manifolds are a 3-into-1 design which contains a catalytic converter. The lambda sensor is installed in the cone preceeding the cat monolith, so as the exhaust gasses leave the cylinders, they first collect from 3 into 1, pass over the lambda sensor, and then flow through the cat. From there, two seperate downpipes take the exhaust gas seperately from each bank of cylinders to another pair of cats that are situated underneath the floor of the vehicle. The exhaust gas at this point is still seperate for the front bank and the rear bank, it then passes through 2 seperate cats, and then over 2 lambda sensors. Only then do the two streams of exhaust gas merge into one, and then onwards to the backbox.

So, 4 lambda sensors

A Front bank primary lambda (located in front bank ex. manifold, before the front bank primary cat)
B Rear bank primary lambda (located in the rear bank ex. manifold, before the rear bank primary cat)
C Front bank secondary lambda sensor (situated behind the secondary cat for the front bank)
D Rear bank secondary lambda sensor (situated behind the secondary cat for the rear bank).

Any random exhaust gas particle will first pass over one of the two primary lambda sensors, through a cat, through another cat, and then over the matching secondary lambda sensor. For this reason you can usually get away with removing one pair of cats without triggering an engine check light for low cat efficiency. If you remove both pairs, you should expect errors.

In your case you mention you have already replaced both primary lambda sensors - did you use Genuine Bosch replacement sensors with the correct part number, or something else? The reason I ask is that I've seen loads of people with problems due to having the wrong lambda sensors installed, or using universal sensors.

Autolusso Penrith - UK's leading independent Alfa Romeo specialist with branches in Bedfordshire, Cumbria & Dorset

Wizard Exhaust systems available here

Tel: 01768 879 171
Pud237 is offline  
Status: -
AO Silver Member
 
amclean's Avatar
 
Join Date: May 2006
Location: United Kingdom
County: Inverclyde
Posts: 1,038

Member car:

'04 156 GTA

I know it sounds obvious but have you changed the spark plugs?
amclean is offline  
(Post Link) post #8 of 43 Old 09-12-12 Thread Starter
Status: -
AO Member
 
Join Date: Aug 2004
Location: United Kingdom
County: West Yorkshire
Posts: 56
Thanks for the replies, much appreciated.

The replacement primary lambdas were both bosch genuine originals, same part numbers as the removed items.

The spark plugs were replaced with the correct platinum long life plugs, and I checked the gaps were 0.7mm.

(The spark plugs removed were alfa '12NC' plugs which are the 'normal life' plugs in near new condition. I could try refitting these with a wider gap (0.8mm) if anyone thinks that might help?..once read that a wider gap gives a cleaner burn, so assume that the platinum plugs are gapped at 0.7 to ensure they stay in spec for their entire life?)...just a thought

New o rings and filters were fitted on the injectors + a good clean + exercised with pp3 battery and redex through them.

New air filter fitted aswell...

As the HC levels are reasonably low even with the rich mixture am I safe to assume the cats are okay?

As the 02/ lambda levels are at the bottom end of the acceptable range (or below) can I assume the problem is not an air leak?

From what I've read today I don't see that my original plan of changing the flexipipes is likely to do anything, so for now I won't be...

I have a complete spare set of coil packs (used believed good) which I could also fit if anyone thinks that's a good idea?

One thing I've noticed is a slight buzz from the throttle body with the ignition on. Not sure if it did that before...Relevant?

...This next one may be a no brainer so shoot me if I've been stupid...The car was laid up in May before I started the engine work with a half full tank of V power. I've not filled it up since but it drove faultlessly to the test station...Could the old petrol have cocked up the readings?

My latest ramblings...

Cheers again,

Mike
miked is offline  
Status: Busy busy busy!
Club Member
Membro Premio
 
Pud237's Avatar
 
Club Member Number: 71
Join Date: Apr 2006
Location: United Kingdom
County: Cumbria
Posts: 44,801
Throttlebody buzz on startup is fairly typical. I would say coil packs do not need changing, as when one fails you usually get a noticeable missfire. Are you 100% sure the engine is timed up perfectly? Did you use the locking tools and a dial gauge / TDC indicator? Wouldn't have thought its the petrol..
Pud237 is offline  
(Post Link) post #10 of 43 Old 09-12-12 Thread Starter
Status: -
AO Member
 
Join Date: Aug 2004
Location: United Kingdom
County: West Yorkshire
Posts: 56
Hi Pud, Thanks Again,

I'm confident the cams are set correctly. I used the cam locks and tdc gauge and turned the engine over numerous times and rechecked. First time I found the inlet cam on the rear bank was very slightly out so I reset that one and repeated a few more times until I was 100% happy. It was cold when I did the belt, probably no more than 5 deg in the garage. Would that affect the tension I set on the tensioner? The mark and notch were lined up perfectly. Is there a temp range you are meant to do the tensioning at?

While we are talking cams, just to mention that the cams themselves were in very good condition with hardly any noticeable wear. The valves were also in good condition and were lightly ground to give a clean seat.

Cheers Again,

Mike
miked is offline  
(Post Link) post #11 of 43 Old 10-12-12 Thread Starter
Status: -
AO Member
 
Join Date: Aug 2004
Location: United Kingdom
County: West Yorkshire
Posts: 56
Tested and failed again...

...Found an airleak this morning, just after the MAF sensor, the reuseable hose clip had sprung open, so that was sorted. Also removed and cleaned the MAF.

Also rechecked both primary lambda connectors...all okay

Feeling confident put 3/4 bottle of redex and 45 litres of Vpower in tank and then went for a 1 hour blast. 5000-6000 rpm for much of the time so everything up to temperature. (wasn't really planning to give the new headgasket that much grief until it had settled in a bit!)

Went for an emissions retest and it failed again,

CO fast idle 1.8%
CO natural idle >2%

So something is still really wrong.

Haven't been able to do the lambda voltage tests yet, hopefully tomorrow.

So what's left?

Injectors?
Throttle body?
New MAF?
Is there a full ECU reset procedure that can be done?
If I buy an OBD scanner what engine parameters can I monitor and do you think it would help solve the problem?

Finally, does anyone know a decent emissions specialist in West Yorks who you'd trust with your beast?

Thanks as always for any and every helpful comment,

Cheers,
Mike
miked is offline  
Status: Busy busy busy!
Club Member
Membro Premio
 
Pud237's Avatar
 
Club Member Number: 71
Join Date: Apr 2006
Location: United Kingdom
County: Cumbria
Posts: 44,801
That is particularly high, strange as everything seems OK. How does it drive with the MAF unplugged?
Pud237 is offline  
(Post Link) post #13 of 43 Old 10-12-12 Thread Starter
Status: -
AO Member
 
Join Date: Aug 2004
Location: United Kingdom
County: West Yorkshire
Posts: 56
Just tried it!

Well, seems to drive okay, not really noticeably different, although it was just a quick wizz round the block.

Obviously the ECU fault light is now on...

Last edited by miked; 15-12-12 at 00:31.
miked is offline  
Status: Busy busy busy!
Club Member
Membro Premio
 
Pud237's Avatar
 
Club Member Number: 71
Join Date: Apr 2006
Location: United Kingdom
County: Cumbria
Posts: 44,801
Are your MOT guys pretty friendly? Would they do another test on the emissions machine with the sensor unplugged? It would make an interesting comparison.
Pud237 is offline  
Status: faclifting a 156
AO Silver Member
 
biffa's Avatar
 
Join Date: Oct 2006
Location: United Kingdom
County: Lancashire
Posts: 3,199
Is the MAF fitted in the right direction? (airflow wise)
biffa is offline  
(Post Link) post #16 of 43 Old 11-12-12 Thread Starter
Status: -
AO Member
 
Join Date: Aug 2004
Location: United Kingdom
County: West Yorkshire
Posts: 56
Thanks Again Gents,

The MOT station is quite a busy one, so I don't want to burn my favours too early as I'm not sure how long this is going to go on! Good suggestion though. What would you expect with the MAF off?...past experience with my 155 I didn't expect it to rev at all, but it did okay up to ~4000 (bit tricky getting out on the open road for a real blast as SORN at mo...even worse as I live in Bradford and practically every traffic light has an ANPR camera here...some kind of 'trial' scheme, so full road tests only really feasible if I've booked another retest...)

Yes MAF is correct way (arrow up), but good point though...

Finally making up my lambda test adaptors so should be able to post those results tonight...

Thanks Again,

Mike
miked is offline  
Status: -
AO Member
 
aggy8888's Avatar
 
Join Date: Apr 2009
Location: United Kingdom
County: Kent
Posts: 430

Member car:

164

Temperature sensors? Make sure the engine knows the coolant temperature. It could be as simple as the cold start feature staying on!
aggy8888 is offline  
(Post Link) post #18 of 43 Old 11-12-12 Thread Starter
Status: -
AO Member
 
Join Date: Aug 2004
Location: United Kingdom
County: West Yorkshire
Posts: 56
Temp sensor....Thanks for the suggestion. The dashboard temp gauge rises steadily to 90 then barely rises even with significant thrashing...so seems to work fine (think that is normal on the v6). There only seems to be one temp sender and I believe that goes straight to the ecu which then feeds the dashboard temp gauge directly (I know other Alfas are not like this). So what you see on the gauge is what the ecu sees...I think, but please correct me if I'm wrong!

Cheers!
Mike
miked is offline  
(Post Link) post #19 of 43 Old 11-12-12 Thread Starter
Status: -
AO Member
 
Join Date: Aug 2004
Location: United Kingdom
County: West Yorkshire
Posts: 56
lambda readings...INTERESTING!

Right, finally got the lambda readings,

Front Bank primary lambda-
0.08v -0.78v ...flicking constantly

Rear Bank primary lambda-
0.08v fixed value...not changing

I unplugged the rear lambda from the ecu with the engine running and continued reading its output voltage...the value gradually rose to 0.5v. Then I plugged it back in and the value quickly dropped back down to 0.05-0.08 (still not sure about this, see later)

Plugging and unplugging the lambda did not make the engine fault light come on.

So it seems that the lambda is stuck reading a constantly lean value so the ecu is compensating by enriching the mix on the rear bank....

Luckily I kept the old lambda so will try refitting it tomorrow (not the easiest task on the aft bank, thats why I deliberately changed it out duriing the rebuild, damn!)

Does not answer every question. Why would a genuine Bosch exact replacement fail straight away? Maybe I'll learn more when I get it out...

I still haven't eliminated the possibility that there may be a fault / short in the wiring harness holding the lambda voltage down and maybe the sensor is still okay...

Progress anyway, which is good!

Thanks Again to everyone who has helped so far, Much appreciated!

Cheers,
Mike.
miked is offline  
Status: Remembering why I dislike working on old cars....
AO Silver Member
 
ChrisS's Avatar
 
Join Date: Sep 2001
Location: Worcestershire
Posts: 2,711

Member car:

Mercedes C350e

On a much earlier and simpler car, but I had simialr issues on the old 155 V6 for a long time - finally found the fault after much angst. It was the fuse powering the lambda heater - it had a crack in it, so would make & break at random.

I'd strongly suspect you have a heater issue on that lambda circuit - reason I say that is the sensor has a very high impedance when it is cold (10s of M ohm IIRC), which drops (a bit) once it's up to operating temperature. I'd think your DVM has a higher impedance than the ECU input, certainly higher with just the one connected, so this would explain why the reading drops when you plug it up. Can you check you have a solid 12v on the heater circuit? This is best done by stabbing the wires on the probe itself rather than just measuring at the connector with the probe disconnected as you will be testing under load.

HTH
ChrisS is offline  
(Post Link) post #21 of 43 Old 12-12-12 Thread Starter
Status: -
AO Member
 
Join Date: Aug 2004
Location: United Kingdom
County: West Yorkshire
Posts: 56
Thanks Chris,
I did notice the heater voltages were a bit odd last night so I've been investigating more. I also have now made up test adaptors for all 4 lambdas so can test them all at once...

So, the latest,

None of the lambda heaters come on immediately, i'd estimate that it was 3 or 4 minutes before I could see the heater voltage on the 1st stage lambdas. I expected to see a straight 12v or thereabouts, but the voltage seems to be modulated, not stable. On the primary lambdas the voltage seems to pulse 10.6-12.6v. The 2nd stage lambdas come on even later, 8 or 9 minutes at a guess. Again the voltages seem to be modulated, pulsing roughly 6.6-10.6v...This was not what I was expecting! does it make any sense??

Next I remeasured the lambda outputs.

The 1st stage ones were as yesterday
front bank (black connector) 0.08v--0.78v pulsing continually
rear bank (grey connector) 0.08v--although it seemed to drop to almost zero at one point

2nd stage lambdas were interesting.
grey one ( I initially assumed this was the rear bank one) 0.67-0.74 pulsing continually
black one (I initially assumed this was the fwd bank one) 0.20 fixed, not pulsing

Logic would suggest then that the black one was actually the rear bank and grey the front bank so maybe someone in the past has swapped them round at the sensor end...
(I had assumed that grey connectors meant rear bank and black meant front bank, but the numbers don't seem to bear that out)

[later edit...I was wrong here, the grey connectors DO mean rear bank lambdas and the black ones front bank lambdas]

Just out of curiosity I swapped the 2 2nd stage connectors round (grey connector on black harness and vice versa...)

Result was that the grey one started pulsing 0.28-0.33. Black one stuck solid at 0.70. !!! I would have expected these to stay the same as the sensor was not physically moved from one bank to the other, just plugged into a different harness.

So to my eyes its looking more towards a wiring or ECU fault. Quick check reveals no bent pins on the ECU, will be looking further at this afternoon...

As always, any comments or opinions on any of the above gratefully received!

Cheers,
Mike

Last edited by miked; 15-12-12 at 11:44.
miked is offline  
Status: Remembering why I dislike working on old cars....
AO Silver Member
 
ChrisS's Avatar
 
Join Date: Sep 2001
Location: Worcestershire
Posts: 2,711

Member car:

Mercedes C350e

Well, not seeing a solid 12V isn't a total surprise, the supply could well be PWM modulated to give a more accurate temperature control...in theory at least. I have no idea how far down that line this ECU goes, but it's certainly not beyond the realms of possibilty. That the heaters for the downstream probes come in later is a bit odd....but again, without knowing what strategy the ECU is using, any kind of view on that is pure guesswork.

Do you know if the heater circuit is fed direct from the ECU? If it is via a relay, then I very much doubt it should be anything but a solid 12V.

Does suggest a wiring or possibly ECU problem doesn't it.......gonna be a great 'eureka' moment when you find it though.


OK, managed to find some info, looks as if each probe has independant heater control direct from the ECU, so seeing a PWM drive shouldn't be a worry. Bother - wold have been good to find it that easily!

Last edited by ChrisS; 12-12-12 at 16:19.
ChrisS is offline  
(Post Link) post #23 of 43 Old 12-12-12 Thread Starter
Status: -
AO Member
 
Join Date: Aug 2004
Location: United Kingdom
County: West Yorkshire
Posts: 56
Cheers Again Chris,

That's useful info to know, many thanks for that.

So this afternoon I've had the ecu off the inlet manifold and checked the two multiway harnesses back to where they enter the bulkhead looking for snagged, cut, kinked etc. wires

Nothing found.

Cleaned and refitted the ECU pins and connectors hoping for a miracle, but situation was as before.

Pulled all lambda connectors off in turn and measured the voltage at the lambda input to the ECU (no lambda connected i.e. open circuit). Both primary lambdas and one of the secondary lambdas all read 0.45 v, whereas the other secondary one reads zero v (another strange anomaly, but for now I'm not too worried about the secondary lambdas as I'm convinced they don't do anything for emissions, just monitor cat condition)

Unfortunately, now seem to have a persistent engine fault light which has not gone off despite everything being reconnected and a number of restarts...

(apparently since April this year that is an MOT fail in itself !! )

Final move of the day was to dig out the old rear lambda sensor. Fitted it to the front lambda position (as I can get to that!)...Ran the engine up again and this lambda definitely works perfectly...just as responsive as the new front lambda.

Plan for tomorrow then is to (somehow) change out the rear lambda. I have no idea how I am going to do this as it was hard enough to do with half the engine missing...! Grateful for any cunning ideas...I really don't want to take off the inlet gear and even if you do it barely gives you any more room for maneouvre...

And I still haven't ruled out maybe ultimately having to change out the ECU. Presume if you do that there will be issues with key codes etc?

That's it for now folks!

Mike
miked is offline  
Status: Remembering why I dislike working on old cars....
AO Silver Member
 
ChrisS's Avatar
 
Join Date: Sep 2001
Location: Worcestershire
Posts: 2,711

Member car:

Mercedes C350e

Mike,

At this point, worth getting yourself a cheap OBD reader I think. They aren't the answer to everything, but can at least provide some clues. I got one of these a while ago - pretty impressive for the price. http://www.ebay.co.uk/itm/Any-Car-En...item46036d217d Works on the Alfa just fine. If you have a half-decent laptop, you could maybe go for FES if you want something more capable but more spcialised.

Only significant engine job I've done on mine so far was the 'big service' but I did take all the gubbins off the top to do that, and do seem to recall there are a couple of plastic connector boxes jam-packed with connectors for ECU functions/sensors. One between the plenum and firewall, the other on top of the thermostat area - give those a good looking at too. Never know.

CODE issues after an ECU swap? almost certainly! I don't know the procedure for changing an ECU, but there must be one that doesn't involve swapping the entire suite of bits. I'll see if e-learn is any help with this and get back.
ChrisS is offline  
Status: Busy busy busy!
Club Member
Membro Premio
 
Pud237's Avatar
 
Club Member Number: 71
Join Date: Apr 2006
Location: United Kingdom
County: Cumbria
Posts: 44,801
Quote:
Originally Posted by ChrisS View Post
CODE issues after an ECU swap? almost certainly! I don't know the procedure for changing an ECU, but there must be one that doesn't involve swapping the entire suite of bits. I'll see if e-learn is any help with this and get back.
A brand new ECU is plug and play. Its only if you're fitting a used one that you need to also swap the code box and immobiliser chips.
Pud237 is offline  
Reply

Go Back   Alfa Romeo Forum > Supported Alfa Romeo Models > Technical & Vehicle Assistance > Engines (TS, JTS, JTD & V6)

Tags
cf3 , emission , failwhere , start

Thread Tools
Show Printable Version Show Printable Version
Email this Page Email this Page



Posting Rules  
You may post new threads
You may post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off

 
For the best viewing experience please update your browser to Google Chrome