Emissions Related Help Needed

[Rich_tea]

Hi, my 420R spec Duratec engined car failed the MOT on emissions.  CO was high, lambda was low on all tests.  I replaced the lambda sensor which made no difference.  Looking back at previous MOTs over the last 5 years I've owned the car (total miles only about 2K in that time), all at the same MOT station, the 1st Fast Idle has always been similar but 2nd Fast Idle has been ok.  Since build the car has just under 5K miles. 

The car is running RBTBs and an MBE 992 ECU (Caterham supplied).  From internet research, the high CO and low lambda readings suggest it's running rich.

So I've checked the coolant temperature sensor for the ECU and that appears fine (66 KOhms cold, just over 2 KOhms with the Stack Dash reading around 90C).  I put the old lambda sensor back in, with a cable spliced into the signal wire to allow me to read the lambda signal voltage.  Once warmed up at normal idle, it cycles in the 0.2-0.8V range which is apparently normal but above normal idle (tested between 2-3K RPM), the lambda voltage goes to about 0.9V and stays there, which apparently suggests rich running.  Once it's back to normal idle, it starts to cycle again in the 0.2-0.8V range after it settles down.  

The car hasn't been used since about July and was taken straight from winter hibernation to the MOT, but via a long route with some motorway miles as well to make sure it was thoroughly warmed up.  It seems to drive ok.

Any advice would be appreciated.

[john milner]

My advice is to post the actual numbers as they might mean something to someone.

I recently had a CO (0.63/0.73) emissions failure and I thought the lambda reading (1065) was not too bad but it was pointed out to me that the lambda was on the high side. I then had a closer look for leaks and found a big one. Fixed the leak (lambda 1016) and the CO (0.11/0.47) passed.

[ChrisC]

My first thought would be to check the Throttle Bodies balance.  If that's out the ECU can't control the emissions.  Then I would be looking at the ECU live data to check for errors or figures outside normal.  A resistance reading from a sensor is not a great way to test for faults because what's important is how the ECU is seeing a value.  Ie a perfect sensor won't work if there is a broken wire.

Another assumption is your cat is warm, and the lambda startup cycle is complete before the test was carried out. 
 

Finally you did replace the lambda sensor with a genuine Rover or Bosch part, not an eBay cheap special?

[Rich_tea]

Thanks for the replies guys.  The actual numbers were:

Fast Idle Test: CO 3.09 / HC 85 /Lambda 0.921 (limits are 0.30 / 200 / 0.950-1.090)

2nd Fast Idle Test:  CO 4.32 / HC 166 / Lambda 0.882

Natural Idle Test:  CO 2.70 (limit 0.50)

So the main issue is high CO and low Lambda, indicating a rich mixture.  

The car has done 500 miles since last MOT and about 2K since effectively new.  Any info on checking the RBTB balance would be welcome.  I checked continuity from the temp sensor to the ECU but measured its resistance at the sensor itself.

The car was warm (both coolant and oil) for the MOT and our home diagnosis and the replacement lambda sensor was a genuine Bosch.

The fuel (BP Super Unleaded) is probably at least 8 months old.  Is this a possible cause and if so, how does it cause a rich mixture?

[simon_h]

If the fuel is very old it can cause an incomplete burn and mild misfire that will bugger things up. I had a similar problem with my ordinary car that had sat for ages because of the current situation. A tankful of fresh stuff and a good run and all was ok. 

[Rich_tea]

I've not had a chance to investigate the emissions issue further until now so the car has sat in the garage since the MOT failure and our initial attempts to diagnose the problem back in March.  

I have bought the MBE ECU CAN diagnostic interface and have done my first logging attempt using Easimap 6 software.  After letting the car warm up for 10 minutes, Lambda is running very low (rich mixture) at between 0.5 - 0.8, whether at normal idle or fast idle (2500 - 3000 rpm).  However, "Manifold Pressure 1" which I assume is MAP does not change during the logging session from 1.021 Bar.  

So it seems the MAP sensor is duff/broken wire etc, unless anyone has any better ideas?  Thanks.

[aerobod - near CYYC]

The MAP sensor should not exist at all with RBTBs, as there is no manifold, it should be using TPS based mapping only, I'm inclined to look at RBTB balance, TPS alignment and whether the Lambda sensor is faulty, assuming input from other sensors such as temperature are all as expected.

The Lambda value itself is not very meaningful with a narrow band sensor, it is only capable of registering lean or rich by oscillating around the stoichiometric value, you would need a wide band sensor and driver system (such as from Innovate) to get accurate Lambda readings.

[Rich_tea]

Thanks for the input James.  I had a feeling it wasn't as simple as a faulty MAP sensor!  This is the physical setup:

So I can see how the MAP sensor is effectively redundant given it's location means it is probably not going to see any significant pressure change.  Any idea why it's even installed?

Coolant temp sensing appears ok and I've replaced the Lambda sensor with a new Bosch MHK 10006.  Which other sensors do I need to check and can you give me any pointers to info on checking the RBTB balance and TPS alignment?  I haven't yet changed the fuel either which is now over a year old.  I couldn't siphon the last 9 litres so assume I'll need to disconnect the fuel line from the filter and pump it out, if that is a possible cause of the rich running.

[aerobod - near CYYC]

The map sensor could be used for barometric pressure correction (something that is missing on the plenum intake version of the R400D) or perhaps is just redundant. It also likely contains the IAT sensor, as the Ford one on the plenum unit does, so the 4-wire connector is +5V, 0V, IAT voltage and MAP voltage.

The RBTBs should be setup with a flow meter to get the right idle flow for each pair (not sure of the spec, as I have the plenum R400D, but others should have that info if it isn't in the technical section), along with throttle stop location for proper idle. You then want to adjust the TPS by slackening the screws and rotating it to give 0.0 or 0.1 throttle site at idle (avoid red flashing 0.0, but moving between 0.0 and 0.1 is fine). This may be an iterative process, as rotating the TPS can affect the idle speed which may in turn require the throttle stop to be adjusted which will then affect the throttle site. The target idle speed should be available as a panel, you want it to be stable within about 20RPM of the target. Typically you want a minimum throttle opening to give a smooth idle at the target idle speed.

In terms of other values beside coolant temp (assuming it is in the 80 to 100C range and certainly not below 60C), I would confirm Lambda status is "OK" and you are not getting any error increments while running for things like the reset counter or cam and crank signals. The short term fuel trim should generally be low and preferably moving between positive and negative values. The battery voltage should also be stable and preferably above 13V at idle, although idle speed may put it into the high 12V range.

If you can't get a Lambda value that oscillates between lean and rich when at a steady engine speed and the Lambda state is "OK", I would look for intake vacuum or exhaust leaks before further troubleshooting. I would also check the plugs are all even and light tan in colour after a good run with moderately spirited driving.

[John Vine]

Some answers to your questions:

So I can see how the MAP sensor is effectively redundant given it's location means it is probably not going to see any significant pressure change.  Any idea why it's even installed?

It's a combined barometric/temperature sensor.  I'm not sure how its signals are used by the ECU, but CC have told me that it should support airflow calculations.  Mine was simply transferred from the plenum when I had RBs fitted, so clearly it doubles as a MAP sensor.
...can you give me any pointers to info on checking the RBTB balance and TPS alignment?

CC published Workshop Notice 608 in 2015.  If you'd like a copy, message me your email address.  For info, the text says:

For all 2.0L Ford Duratec engined R500 and R400 using roller barrels the following can be used to help setup.

1. Remove air filter (and air box if fitted)
2. Adjust the throttles in closed position to 6mm using round bar as gauge
3. Adjust the TPS to 4.62V when throttles are in closed position. Please note that the voltage reading will fall as the throttles open.
4. Start engine
5. Using an air flow meter ensure that each cylinder on idle flows between 5-6kg/hr using the air bleed screws which are situated on the engine side of the throttles.
6. Idle rpm should be 900-1000rpm.
7. Stop engine
8. Refit air filters and air boxes as appropriate

JV

[Rich_tea]

Thanks James and JV.  Unfortunately when I went to run Easimap again, it isn't seeing the ECU this time.  After I ran the first logging session, I disconnected the ECU connector to test continuity on some feeds but I think when I reconnected it I must have dislodged the female socket in the loom connector for Pin 8 (CANL) (clear plastic shroud removed for investigation).  

Any tips on getting it back into position?   Looks like I need to open the backshell casing and push it back into position from the rear?

[aerobod - near CYYC]

Yes, the shell should have little locking tabs that you can carefully pry to separate it for removal, then push the pin back in from behind until it locks in position. There may be a tie wrap on the shell where the cable enters that has to be cut and replaced to get the two halves apart.

[aerobod - near CYYC]

If the pin won't lock in position, there are tiny tangs that may have to be carefully bent out slightly.

[Rich_tea]

Thanks James, I've sorted the ECU connector. 

Before adjustment, the closed throttle front 2 cylinder throttle openings were around 5.7 mm and rear 2 around 6.8 mm.  I balanced them to bring the fronts to match the rears then lowered the idle stop to bring all 4 to around 6mm.  I adjusted Throttle Angle 1 (TPS) from 4.59V @ closed throttle to 4.62V.  Throttle Site 1 was reading 0.00 (not flashing) before and after TPS adjustment at closed throttle. 

I ran the engine up to temp and did an Easimap logging session.  I don't yet have a flow meter so couldn't measure or balance air flow.  Normal idle when warm was 1000 rpm with small fluctuation.  I did a couple of fast idles approx 2500 - 3000 rpm and a blip to about 4500 rpm then back to normal idle and shut down.  Throttle Angle 1 and Throttle Site 1 seem to change with throttle input, lowest Throttle Angle was about 4.24V and Throttle Site max was about 6.1 at the 4500 rpm blip.

The car has always had a tendency to kangaroo at slow speed light throttle openings but not sure if this could be linked to the RBTB imbalance?

The only apparent anomaly I can see in the log file graph is Lambda, which fluctuated around 0.65 and only spiked momentarily to about 1.1 when I dropped the throttle back to normal idle.  Lambda Status was displaying  "Control Disabled by User" at normal idle and "Control Disabled by Target Map at fast idle & throttle blip in the Easimap live Panel but only showed a flat line "Disabled by User" in the entire log session graph.  Today's Lambda reading was using the new Bosch Lambda sensor but was consistent with the log data I got the older day using the old Lambda sensor and before adjusting RBTB balance & TPS.  The Lambda reading around 0.65 seems off the scale from what I can find on a quick Google.  When I measured the Lambda signal voltage in March with a voltmeter before getting the Easimap interface (details in an earlier post), the voltages I recorded then appeared to be closer to spec than the 0.65 Lambda figure now suggests (noting Lambda value in Easimap is not voltage).  

I still have not changed the fuel which is now over a year old so that's on the list.  I may also put the old Lambda sensor back in and read the voltage again to see if the RBTB/TPS adjustments have made any difference.  There are no obvious air/exhaust leaks so any other advice/suggestions would be most welcome.  The car is SORN and out of MOT so I can't road test it (legally) unless I'm taking it for the MOT. 

If anyone is willing to look at my Easimap log file, that would be much appreciated.  

[aerobod - near CYYC]

I can have a look at the Easimap log file over the weekend.

With the ECU being locked, it is not possible to see where Lambda should be in closed loop mode, but typically should be when the engine is above 60C water temp, throttle site is below 8.0 and revs are below about 4500RPM. It should go into closed loop mode and Lambda Status should then be "OK".

As I mentioned previously, the voltage (and consequent Lambda value) of the Lambda sensor will only oscillate meaningfully between the extremes of rich and lean with a narrow band sensor, any value in between is meaningless, you need a wide band sensor fitted to get meaningful Lambda values. In the case of Lambda showing as 0.65, it could mean that Lambda is anywhere from that value to less than 1.0, but you won't get values above 1.0 unless either the fuel trim can adjust appropriately in closed loop mode, or you close the throttle quickly to remove fuel before the engine has spooled down and burnt any fuel in the inlet tract.

A log of fuel trim may give more info, especially if it is struggling at high negative values to hit target Lambda, but reaches a maximum adjustment value that could be somewhere in the range of -20%, depending on what is set in the ECU. If this is the case overfuelling could be due to a leaky, stuck or otherwise malfunctioning injector (indicated by a the plugs not being even in colour), a misfire, or an air metering issue due to faulty sensor input or an air leak in the inlet or exhaust.

If you are getting any misfires on an R400D/420R that may be upsetting the mixture control (they can be subtle, with no significant roughness or loss of power), then the plug coil wiring sub loom and the connectors on the coil pack could be causing the problem due to intermittent firing. You can normally check for this by measuring the exhaust primaries temperature with a cheap infrared thermometer, looking for temperature variations when the misfire happens and subsequent raw fuel is sent into the primary, cooling the one down where the misfire is occurring.

[Rich_tea]

Thanks for the offer James.  PM sent.

[7 Minder]

I thought I should update this thread as we left it hanging. 

I took over the car from my son Rich (Rich_tea) in April.  With the extra motivation of it now being my problem, I took another look.  Although I'd done most of the earlier diagnosis attempts, I kicked myself when I realized the throttle bodies were loose against the head.  With the air filter housing removed this time, it was plainly obvious.

Anyhow, after tightening it up, it passed the MOT without a problem.   A couple of posters last year had suggested an air leak as possible cause but I hadn't spotted it.  Thanks to everyone who offered advice.  I'll try to pay closer attention next time!