Your figures look good, and I’ll admit right now I’m no expert, it is possible the lambda sensor is not reading correctly, as the scanner will only show you what the ECU sees and if the Lambda sensor is telling it that it is at 14.7/1, the ECU will try and maintain that. However my gut feeling is your sensor is okay.
Where you present at the MOT test?
Was the car thoroughly warmed up prior to the test?
Has the car been sat around unused a lot before the test?
And was the ECU remapped after the new CAT was fitted or before?
It may be that the car just needs bit of an “Italian tune-up” prior to the test or the mechanics didn’t let it warm up sufficiently.
Your disadvantaged here because the second O2 sensor monitors the cat efficiency and rumour is that it can affect fueling to keep emissions in check, although I’ve never seen any evidence of this.
I don’t have the figures yet, I need to pop back and collect the paperwork at some point. The exhaust does smell rich and when I disconnected the sensor wiring the EML illuminated but it did not change the running of the engine. This made me think the sensor isn’t actually changing anything and so perhaps failing.
I thought AFR was ok but I’m sat in my lounge and it’s still claiming it can see 14.2 so I’m not sure how much I trust that!
The car was definitely warm when it went through, the cat was a replacement for the original centre section one and all mapping / changes were done about 18 months ago with no changes since. It seems emissions performance has dropped significantly since it’s last MOT - it’s a bit of a puzzle!
If the cars definitely running rich, and it sounds like it is, and the O2 sensor is reading 14.7, then it sounds like it could be the O2 sensor. You could induce a vacuum leak to see if it responds, but I suspect if it wasn’t responding you’d get a EML. I’d be interested to know what the externally measured lambda figure is though as a comparison to your O2 sensor.
I remember trying to diagnose a running rich problem on my mk2. I put a meter on the O2 sensor and it just read about 0.8 volts. Turns out the output swings between 0 and 0.8 about ten times a second which was too fast for my DVM to register. When I used an oscilloscope you could easily see this swing. Also if, as Henry says above, it is current based, not voltage, I doubt if you would see anything unless you put the meter in series with the output. Just my thoughts.
Just one note. If you use penetrating oil to ease removal and wish to reuse the sensor clean all traces of oil off before refitting. Otherwise you will have some really funky readings if it gets on the sensor. If fitting a new one only use a very small amount of anti seize grease on the threads and clean the are in and around the exhaust thread.
Hi,
I have the same package as you fitted by Skuzzle, albeit a 2013 model.
Nick does still operate a Q&A page from his Skuzzle FB page.
He also is working in partnership with another garage.
I have the original centre section (with the cat) still as we didn’t think it was worth replacing.
Just an opinion without getting too technical, but could it be that your new cat has failed even though it is new?
The way I see it, it is either the O2 sensor or the Cat.
IF it was me I would have an independent specialist look at it and diagnose the problem and probably the best £50 you could spend.
You could faff about forever and never get the true fault.
Just an opinion and good luck and hope you get it sorted soon. (across the border in Wilts).
Just tried running torque app on a cold start with the following digital gauge parameters
AFR commanded
AFR measured
O2S1Eq
Coolant temp
AFR commanded on startup is 14.7 and stays stable throughout
AFR Measured is 12.7 ish on cold start and fluctuates rapidly until rising to 14.5 -14.9 at 82 deg coolant
O2S1 is 0.9 at startup and moves quickly to 1
I have the 2nd O2 fitted on my car but I believe that is used more for long term trims so should not affect your readings
All done without throttle and car stationary btw.
I’m leaning towards the lambda being defective. The heater circuit is 3 ohms and running the car showed that I don’t appear to be getting any sort of signal from the sensor.
Interesting article relating to this type of sensor cribbed from a certain US forum
Mentions manifold leaks which could be worth checking for. I had one of the manifold nuts work loose and drop in the undertray several months after a racing beat 4 branch was fitted.
So as this particular snag has piqued my interest, I got the old ELM out and had a look at my car, which is only the 1.8.
The car scanner app didn’t give me the option of looking at voltage, but did let me look at current and air fuel ratio.
It seems that 0 mA corresponds to stoichiometric and a very small current tells the ECU to adjust the fuel and the polarity of the current indicates rich or lean.
I’m led to believe that there should be a pretty constant voltage, the value of which varies by manufacturer.
So from my reckoning a wide band sensor will fail at stoichiometric as there would be 0 A, and a 0 voltage would also indicate a failed sensor.
However, I would expect a completely failed sensor to generate a code and the EML. So I’m still not totally convinced it is your O2 sensor, but it certainly could be.
Thanks once again. The Torque Pro app doesn’t appear to let me see current so I’ll aim to replicate your readings tomorrow and see how they compare.
The 3ohm heater resistance seems low compared with various MX5 forums but it’s within spec based on the manual above so I’ve bought a new sensor while there’s 37% off, I can always return it if necessary!
Likewise, I continue to read with interest.
Again just an opinion, but perhaps your O2 sensor has failed and also damaged the catalytic converter? (I hope not).
I am not saying that has happened, but personally I would be very suspicious that “might” of happened. It’s one or the other or both.
Professional advice needed in my opinion.
A garage with good testing kit will give you the definitive answer and hope it’s a simple cheap fix.
I like your optimism but
a, there is some satisfaction in working, learning and researching your way out of a problem like this and it generally saves a lot of money and stops parts being thrown at the problem.
b, in the absence of a fault code being generated the average garage is going to replace parts at your expense until the problem disappears or you run out of patience.
We all get valuable insight from this process and it benefits the forum community if and when the problem is found and assuming the outcome is published.
I don’t disagree if you have the tools and skill base to do the work.
Just from experience you get to a point…
Good luck with the fault finding and fix.
Only if you have not opened the package. Because of their sensitivity O2 sensors tend to be that you’ve bought it once it’s open.
The fact that your A/F ratio is static according to the sensor despite change in revs [and even engine off] and clearly being rich is a very good sign that your sensor is the culprit. The other end of the sensor chain is the MAF. I changed the O2 sensor and then it brought up a code which needed the MAF changing too.
Thanks everyone for the input. I’m an engineer by trade (ships not cars though) so I quite enjoy the diagnosis puzzle. I intend on going as far as I can (with the support on this forum which has been great so far) but as someone said above, I accept that I might just have to pay for some expert advice at some point! Until then…
MK375 provided an extract from the service manual so I ran that test on my driveway (static - I don’t have a very long driveway). So the engine wasn’t under load but I did the 3000rpm to idle test. The criteria is that the sensor current should peak at 250mA or more under deceleration - mine peaks at 172mA so although (off)load might make a difference that still seems quite low.
Mine needed replacing but was seemingly working fine. It gave sensible readings and was not setting a code. I replaced it and then got CEL code because the MAF needed to go as well.
They can be broken even when they appear to be working.
