40 + bhp for 50p?

As James says, the sensors don't work as back-ups for each other, but in conjunction to provide a more accurate result.

Obviously it varies from make to make, but my understanding is that most systems use a series of primary sensors to determine the engine's position within the map, which is a function of load and engine speed, and then a set of secondary sensors that make small corrections to improve the efficiency further.

The primary sensors are the crank speed and, and then some method of determining the load, which tends to be either a mass air flow sensor, or a manifold pressure sensor.

The secondary sensors then provide a correction function, which can either be a fixed linear adjustment to the maps output (such as the air intake temperature), or it can be an additional correction map. (such as throttle position). Air intake temperature is usually required for systems that use MAP for their loading, as they need to know the density of air to determine how much is entering, while a MAF based system would only need it for extra fine tuning of the timing. Lambda is one of these secondary sensors, it just provides a correction factor to get the engine to aim for its target air fuel mix if it sees a problem. Engines with a wideband lambda can correct under load, but like James says, engines with narrowbands can only do it at cruise when you're aiming for stoichiometric to keep the cat happy.

Obviously quite a few of them do more than one job, and some sensors aren't anything to do with the initial map, but may effect things, like a boost sensor on a MAF based engine which may control the actuator, but not have any real say over the fueling or timing.

In the case of that Honda in the link, the air intake sensor tricked it into thinking it was taking in nice cold air, which made the timing advance, suiting the higher octane fuel. But it also tried to richen the mixture, as it was applying a correction to the fuel map thinking the air's more dense, so needing more fuel. But then then at cruise the lambda sensor was kicking in and correcting things back to stoichiometric. When they put their foot down things were getting richer though, because the narrowband lambda couldn't measure the intentionally richer mixtures under load, and the changes often happen too fast for the lambda to keep up with. (Plus fuel evaporation from the inlet ports walls screws things up under changing loadings.)

(Sorry, I may have gone a bit ott with that reply, someone may find it useful/interesting anyway.)


Tricking sensors with resistors can be an effective way of correcting the management when you've moved away from factory in other areas, but as with any other mod, it only works if you know what you're doing.