Mems 2J - Lambda inactive

Hi all,

Apologies first of all, I'm an interloper with an S1 Elise 111s and this is a lengthy first post! The 111s shares the same engine as the MGF VVC so any thoughts are welcome on the issue below.

I have an issue with the cars emissions; it idles and drives perfectly but the engine's running rich with tell tale smell and black exhaust deposits but no smoke.

Pscan has highlighted the lambda sensor is inactive and remains that way. The lambda voltage is mostly static at 0.49v with an uplift in voltage only when on the over-run from c. 2,500rpm where it will briefly climb to c. 0.6v before falling to 0.49v again.

Measuring the output from the lambda using a digi multimeter shows an output of c. 0.8v which does vary a little with engine speed, I think this is telling me the lambda knows the engine is rich but the ECU is not responding to the lambda.

Prior to me owning the pscan the lambda, temp sensor, air temp sensor were replaced recently with genuine MG ones to try and address the issue so I'm reasonably confident they should be OK.

The harness has good continuity all the way back from the various sensors & lambda connector to the ECU (+ve and -ve) yet the ECU appears to not respond to the lambda. I've tried a second hand Mems 2J to rule out the ECU but no change to the lambda response.

A couple of questions:

1) Would anyone be able to share with me some baseline 'live' Pscan data for the VVC? As a newbie to this type of diagnostics it would be helpful to verify I'm interpreting the readings from all of the sensors correctly and importantly to understand what I should be seeing for the lambda, MAP etc.

2) Does anyone know if the Mems 2J defaults to a static voltage (0.49v) in the event of the lambda voltage being outside of a normal range (I'm guessing c. 0.5 to 1.0v)? I'm beginning to wonder if this is a symptom as opposed to the cause which may be caused by an issue elsewhere?

Thanks for any thoughts.

Stuart

Hi and welcome to the forum Stuart

Don't worry, I think of the S1 as a racier more shapely cousin to the F.

I had identical problems on mf mk1 F, right down to replacing the lambda probe and trying a different ECU.

The probe I got had a later plug so I made up a patch lead.
Nothing worked so I got another probe but with the correct plug and it breezed through the MOT.

IIRC the probe voltage alternates between 0.1 and 0.9v.
With no lambda reading the ECU will keep the fueling on open loop.

Keep us informed on when you make progress.

James

I haven't tried it myself (on a Lambda sensor) but if had this problem then I think I would be tempted to look at the lambda signal using something like winscope .

It uses a line in card on a PC soundcard as an oscilloscope and it's free.

From memory a line in is about 1V pk-pk so all you would need is a suitable 3.5mm jack plug, two bits of wire and it would probably be wise to put a 1M ohm resistor in line.

A microphone input is about 10 times more sensitive so using a 10 to 1 resistor divider to reduce the voltage to one tenth would do it.

Before I had a proper scope I was doing this to look at diagnostic protocols and it worked quite will.

A sound card input is AC coupled so it's no good for signals that don't change, but you can use a voltmeter for that.

My understanding is that for a normal narrow band lambda sensor you should see a square wave as the sensor constantly turns on and off. On a mic or audio input this would turn into a sawtooth but at least you can see that something is going on.

Obviously do your own research because if you destroy the line in/mic in on your laptop I'm not going to be responsible for that.

I seem to remember we did a thread on j2 ref values when we were hosted on the old rover.org forum. Phiilip, do you still have access to the original threads? I am pretty sure the contributors were myself, cairnsys and rob bell.
There is a thread on here now called reference values, mems3 That should be very similar to the J2 with regard to fuelling. Every time I have had high emissions it has been the lambda sensor at fault, but do check for air leaks around inlet manifold and exhaust pipe.
Nigel

Everything that I managed to grab from the old Rovertech forum is archived here https://www.the-t-bar.com/forum/pscan-uk-diagnostic-tool-forum/96416-archive-of-lost-support-topics

Thanks to everyone for your thoughts, much appreciated. I'll have a trawl through the old postings to see if the reference data is there as it will be great to get a full set of readings.

What a pain this has proved to be but I've got there in the end. Pre Pscan I tried new lambda, new sensors with no luck (good old stab in the dark method) but no improvement.

Post Pscan I narrowed the fault to the fact that the lambda remained inactive and there was no fuelling compensation, more testing of wires, cleaning of contacts, removal of the misfire system (unique to the Elise 111s) but no progress at all. Tried a new (used) Mems 2J, no progress there either. #getting p*ssed off!

In short after eliminating all causes I've come back to what I assumed was an unlikely culprit....I replaced the new lambda I fitted a few months ago (genuine MG) with a new NGK one and some new plugs for good measure and....

After under 30 seconds the lambda goes active, the lambda voltage varies and I have fuelling compensation too! All looks well. It really does prove that old point don't assume it's good even if it's new!

Just need to fit everything together and give it a thorough road test before getting the emissions checked.

Thanks all.