MOT Emissions failure argh!

  1. My model of MX-5 is: __2006 NC 2.0 litre
  2. I’m based near: __Bristol
  3. I’m looking for technical help or recommendations on: __Emissions and Lambda Sensors!


I’ve got a 2 litre NC with decat manifold and the second (post cat) sensor mapped out (the Skuzzle fast road pack). It’s also got a brand new 400 cell centre section cat and the car is failing badly on emissions. The MOT certificate says (I don’t have the readings yet):

  • Exhaust carbon monoxide content after 2nd fast idle exceeds default limits ( (b))
  • Exhaust Lambda reading after 2nd fast idle outside specified limits ( ©)

The mechanic is convinced that the map has slipped and the car is just running rich the whole time. He thinks the answer is a remap and it’ll all be fine, the issue is I would need to go to Blink / BBR as I can’t find anyone who can work on a flashed ECU near Bristol. However I’m not so sure…

Google says that a failing lambda sensor can have a similar effect and my OBD2 reader sampling at 0.1 seconds showed the O2 bank 1 sensor 1 steady at 0v on a warm engine between idle and 5000 rpm - I expected to see a fluctuation between about 0.4v and 1v based on what the internet tells me!

I tracked every version of the sensor incase it was logged differently and they’re all the same; I also tracked Mass Air Flow and that varies with throttle so I know the reader is outputting correctly.
Any thoughts on where I go next? This is the first time if used a OBD2 diagnostic tool and I think I’m reading it right but I don’t really want to spend money on a new sensor if I’m reading this wrong or not looking in the right place.

This is just my initial reaction, I am prepared to be wrong.

Map slippage? that’s a new one on me

The emissions test is usually done at idle speed. If it fails idle speed then they do the extended test. So your car is running too rich at idle. The most likely cause of this is the MAF sensor being dirty, the AF sensor [bank 1 sensor 1] being faulty/dirty or the air filter being clogged.

Clean the MAF sensor, takes about 5 minutes. Change the air filter [or clean the skuzzle intake] and at this stage I would put some fuel system cleaner through the engine and give it an italian tune up if it’s been doing short journeys. I might even disconnect the battery so the idle has to re learn in case it has picked up bad habits

Using the torque app I would also look at the fuel trims and historical or pending codes.

You also state that you have a new 400 cel cat fitted. Is this since the remap? was there a less restrictive 200 cel cat on there before?

Have you thought about possible exhaust leak in front of the lambda sensor, where the exhaust gases might be drawing in air making the sensor think it is too lean?

Am I reading this correctly? Mani decat, post mani cat mapped out, so no catalytic converters? THEN, a 400 cell cat introduced beyond lambda sensors. Is it any wonder? Might be wrong but thought it worth mentioning.

Your ecu can’t do anything regardless of map with no o2 signal. You need to work on confirming that you have a working primary o2 sensor before anything else. It could be that the connector has somehow failed or the cable has been damaged by poor routing near the hot manifold

So decat manifold, lambda, centre section cat then backbox. It’s the second lambda that is not needed in this configurstion that has been mapped out.

It passed the MOT with flying colours last year and I’ve changed nothing since. It’s part of what makes me think the sensor is on its way out and not controlling the air / fuel mix. The ECU is definitely going closed loop when warm.

1 Like

Fair comment, wasn’t sure from original post. Thanks for clarifying :+1:

Lambda’s do let go, most definitely be worth swapping it out to see. If not done before, you will need a socket specially for the job, I had to buy one for my wife’s cara few years ago, it’s one of those tools that you buy and then probably never use again :unamused:

Break the old lamda sensor off at the porcelain and use an ordinary socket to get the old, virtually welded in one out and the special tool for the new one.
If it will come out with it, by all means use the tool for the old one but I found it flexed too much.
I found alternate heating with the engine and spraying with lubricant was necessary

Thanks everyone. I’m going to see if I can get a multimeter on the pins to see if the sensor is dead or the wiring is damaged. After that it looks like I’ll have to put my hand in my pocket!

1 Like

It’s a wideband O2 sensor so you should be able to see the airfuel ratio which should be around 14.7/1. If you’re getting anything above that then the O2 sensor is reading rich. It’s also useful to look at your fuel trims, this will be given as a percentage. They should in an ideal world be at 0%, a negative % would indicate a rich condition and the ECU is pulling the fuel. A positive % would indicate a lean condition and the ECU is adding fuel. Also does it smell rich, is the exhaust sooty, have you had a look at the spark plugs? Sooty spark plugs indicate rich running. If the O2 is indicating 14.7/1, but there are indications of running rich, then it’s likely to be the O2 sensor. If the sensor was dead you’d get an engine management light, but if the wiring or the sensor is damaged or contaminated it can read 14.7/1 regardless.


If you want to put a multimeter on the connector while plugged in to check voltage signal directly you can sometimes use a couple of inch piece of single strand copper which is small enough to go in the terminals and exit to outside the connector. Any old multistrand cable can be used

Quick update, i’ve put a multimeter across the ground and live wires on the pre-cat O2 sensor and it’s reading 0.005v the whole time, at all revs and with a rich smelling exhaust. I thought I’d see roughly 1v for a rich condition.

Does that sound like a failing sensor?

Surely you would see close to 1v if the sensor was active and working. If it’s feeding close to 0v back to the ecu then it is sensing a lean condition and trying to compensate. Sounds like a new sensor reqd

since you are not currently using the second sensor, can’t you just physically swap the two sensors?

They are different

No, it a wideband sensor, it doesn’t act the same as a narrowband sensor. It works on current. With a scanner you should be able to see the air/fuel mixture ratio. Looking at this combined with fuel trims and the fact you know it’s running rich should tell you whether the sensor is serviceable or not. They’re reasonably expensive, but if money’s not an object you could just change it and see if that fixes the fault. I’m assuming you haven’t got a engine management light on.

1 Like

There is also a heater circuit in there, just a heads up if you’re putting a multimeter across it.

Thanks Henry! Air/Fuel was between 14.6 - 14.8 and trim (from memory) was less than 5% so it seems odd that that voltage across the sensor (I checked which ones were the heater) was very low at 0.005v. The multimeter backed this up too. I’m assuming the sensor is claiming the engine is lean and so more fuel is being added and so it’s running rich.

I’m trying to use the OBD reader to try and be sure that the sensor is worth the investment. I appreciate there’s always a risk but if rather pay once and get it right of I can.

Although there are conflicting claims about it, I think Henry is right that the Mk.3 front O2 sensor is some kind of wideband type, not the more familiar narrowband which produces a 0-1V output. If so, I don’t think you can measure any meaningful output by looking for voltage generated by the sensor itself.

1 Like

That’s interesting. Perhaps I’ll run another log tomorrow of air /fuel and trim to see what happens.