Lambda reading and lambda sensor are often confused. For the MoT the lambda reading is taken from the exhaust and a difference could be indicating a leak in the exhaust itself. Therefore the fault most probably won't be with the sensors, these can be tested by taking the voltage across them and looking at the wave-form and the max/min readings of the voltage.
Have a look:-
Check the basic engine tune; check plugs, leads, cap, rotor, air filter etc. Make sure the engine condition is ok (see vacuum testing tech tips). Ensure that the proper testing procedure is carried out, i.e. ensure the engine is fully warmed up and the sniffer pipe is fully inserted into the exhaust etc. All basic stuff, but you would be surprised at how many people change ECU’s etc without even taking a look at the plug condition. Don’t take the customers word for it check it for yourself. You could save yourself a lot of time. You should ensure that the data that you are using as a comparison is accurate. Ensure that it is correct for the car, inc. year, cc model etc. We’ve all, at one time or another spent far too long chasing our tails as a result of a simple oversight or misread number/spec. Test the car, check the emissions at all the required rpm levels and either print out the results or write them down. As a rough guide the expected levels of emissions with a catalyst cat with closed loop injection system are as follows; CO 0.3% max. HC 75 PPM max. CO2 14.5% min. O2 Probably the most common mistake made is when the lambda reading is incorrect. Lambda refers to the air fuel ratio, it is a reading which is calculated from the other emission readings (lambda (l) comes from the Greek word for balance). When faced with a car that has failed the emission test on lambda reading the most common response is to condemn the lambda sensor. This guesswork will sometimes pay dividends but more often than not you will have made an expensive mistake. The lambda reading on a gas tester is, to repeat, an indication of the air to fuel ratio, too high a lambda reading relates to too much oxygen. Too low a reading relates to too much fuel. Check the other readings before condemning the lambda sensor.
The voltage reading from the lambda sensor is often a very good place to start with your diagnose of an emissions failure fault. The lambda sensor can be used as a pre-cat emission tester when used with either an accurate voltmeter or an oscilloscope. With the engine up-to normal working temperature, check the output of the sensor wire, normally black with Zirconia sensors, you should expect an output of 0.2 – 0.8 volts fluctuating between these reading approximately 2 – 3 times per second. If the output voltage is lower than this, i.e. 0 – 0.2 volts there will be a lean running or excess oxygen problem. If the voltage is higher than this, i.e. 0.8 – 1.2 volts then there will be a rich running or excess fuel fault. A point to note is that conventional Zirconia sensor is not capable of producing a voltage above about 1.3 volts; if the voltage exceeds this then there will probably be an earth fault with the sensor. Some cars notably Rover have different requirements in respect to internal lambda connections, should you measure a voltage above 1.3 volts it might be worth checking for a lambda sensor mismatch fault.
If the fault is a high lambda reading, with a high O2 reading but with normal CO and HC readings, the most common cause is post combustion air leaks. A small exhaust system air leak will leak air into the exhaust before any blows can be detected. Only a small volume of clean air introduced into the exhaust will cause a car to fail the MOT test.