Are plugs, ignition wires, airfilter, O2 sensor and ignition timing ok? If so, the problem might be related to the relocated airflow meter on the CAI. The orientation of the flapper door is changed, and thus gravity might help to simulate a greater airflow than the real airflow.

To correct this, either more unmetered air has to be allowed past the AFM - this can be done by adjusting the bypass screw on the AFM, or the spring tension on the AFM needs adjustment - a rather tricky procedure which involves taking the sealed cover off the spring mechanism. In my case at the last "MOT", more bypass air was the key to proper readings.

IIRC, too high CO2 readings can also be due to carbon build-up. I heard this can be cured by slowly feeding water or water mixed with brake fluid into the intake tract (running engine). Never tried it myself, though. I have also no idea what this treatment might do to the catalytic converter. (3/00)

Jjust MHO but I don't think this garage understand emissions technology.

The O2 sensor doesn't have that much effect on overall CO (CO - not CO2 - CO2 is what you want - even if "2 jags" doesn't) - it adjusts things second by second. But even with a completely non-operative O2 sensor, CO emissions won't exceed about 2.5% worst case (and more likely 1.5%) - so long as the rest of the engine tuning is OK.

A cat is NEVER required to get a pre 92 MX5 through the appropriate emissions test - the limits are 3.5% CO and 1000ppm hydrocarbons. And if your cat is shot, replacing it won't have much effect - they're not magic - can only reduce CO by a small amount, like 1.5% to zero. And if the tuning problem has caused your cat to fail, the new one will go the same way PDQ.

You have to cure your basic tuning problem (basically overfuelling) - it may be a temperature (input air or water) sensor, plugs, plug leads (if they haven't been replaced recently that's what I'd suspect) or something more obtuse.

Good luck - you need a tuner who understands his equipment! BTW - the O2 sensor and the cat are about the two most expensive items in the system (except the ECU - heaven forbid) - and neither are needed for your emissions test. (3/00)

BTW - the O2 sensor and the cat are about the two most expensive items in the system.

Only true if you pay Mazda prices. I have bought both single and 4 wire O2 sensors locally for £30 and £45 respectively. They were both Lucas brand and fit and work like a charm. I understand the Mazda price for a single wire sensor is £120 or the like. (3/00)

You can get your Mazda garage to check the AFM resistance to see if they match the open/close resistance. Don't know if they have accurate resistance figures listed but this would be the easiest elimination of a possible cause to overfueling...But then again we are talking about greedy garages. I have a slight suspicion that they already know that the AFM could be the major the cause but have decided to get more money by "playing" around with stuff that takes maximum effort with minimal effects ;-)

And if the problem is the AFM positioning.....

The message above about adjusting the bypass screw is correct but please don't try removing the siliconed panel and adjusting the internals. No-one touches the internals except....
(a)AFM manufacturer (Loads of money and either speaks Japanese or German only)
(b) Auto electrician who specialises in AFMs and has worked on the MX5 AFM before. (Loadsa money and speaks so much technical jargon that he might as well be speaking Japanese or German)

The panel is siliconed for a reason ;-) If you try to adjust yourself or let a mechanic with no experience dealing with AFM meters have a go, you may damage the AFM and a replacement one from Mazda would be well over 350 GBP!!! I sell mine for 100GBP but this isn't the point. Why replace something that might not need replacing. Your options are....

(1) Adjust bypass screw ( remember to mark position before you start adjusting just in case) If this isn't enough to do the job go to option 2
(2) Turn the AFM angle to as close to OEM position and see if it alters the emissions and if it does work a bit, use option (1) to try and get emissions lower. Pulling hair out time if this doesn't work so go to option 3
(3) Remove the CAI temporarily and position the AFM back to OEM position. If you're lucky, you might get away with it and the AFM will perform as OEM. If you're not lucky and the emissions are still slightly out, use option (1) to get emissions lower.

Another thing, you'll be wasting your money on another AFM if you're keeping the CAI, the same symptoms may appear again later on. Either get rid of the CAI or find a way of adjusting the AFM positioning to OEM state with CAI or you can play with the AFM-ECU signal.

No-one modifies the internals of a good working AFM to alter signals, there's too much of a chance of damaging delicate and sensitive mechanical parts. You can play around with the Airflow via external gadgets such as electronic Airflow controllers which takes the signals from the AFM and adjusts them to fool the ECU into either richening or leaning the fuel mixture. Quite a good mod even if you don't have fuelling probs as you can adjust the Airflow settings at certain rev ranges (need a rolling road to get best results). If the AFM positioning is causing the ECU to overfuel and you want to keep the CAI, buying the Airflow controller may be the way to go BUT.... it all depends how much your current AFM signals are out by. Although it gives a broader adjustment than the bypass screw, like all things in life, the Airflow controller still has it's limits. Then again....you could always add a electronic fuel controller to lean out even further. Starting to get a bit complicated for 3:30 am in the morning ;-) (3/00)

To do this is of course a bit difficult, because not everybody has access to a source of a 5V reference signal.

.. is this not what the ECU provides .. never played with the AFM but the throttle, cam pickups etc are all supplied from a 5V (or close) reference signal. (3/00)

You can get your Mazda garage to check the AFM resistance to see if they match the open/close resistance. Don't know if they have accurate resistance figures listed but this would be the easiest elimination of a possible cause to overfueling.

A small addition: The resistance values will not say much about the proper function of the AFM. They are all over the place if you move the flapper door from closed to fully open, generating a sawtooth like pattern over the range of movement. A better test is to check this: (From Randy Stockers extremely recommendable web-site:

The 1.6 OEM flapper door flow meter potentiometer measures 3.5 volts at idle and graduates linearly to 0 volts at WOT with a 5 volt reference signal.

To do this is of course a bit difficult, because not everybody has access to a source of a 5V reference signal.

I do agree with the warnings about opening the AFM. I would not recommend doing it without having a spare handy. Yet I know of several people who have successfully adjusted the spring tension when they swapped the 1.6 AFM for an early RX-7 unit (check Randys page for why and how). http://members.aol.com/solomiata/ I will try that myself when I try to combine an early RX-7 AFM with the JR CAI this spring. (3/00)

Not knowing _anything_ about this subject and so having an exceptionally wild stab in the dark..

Perhaps it has had it's cat taken out (what year was it again?) and as it is being tested as a 5 (and not as a Roadster) then it is subsequently failing...?

(Once had a 92 1.1 Metro that I had stuck a 16v badge on the back, and as it had the badge the numbskulls that MOT's it thought it was failing everything as they were testing it as a 1.4 16v (you can't get a 1.1 16v Metro to start with) I had to tell them what it was - worrying!) (11/00)