Pre-codes: anything to add or correct?

[matthew1]

Pre-codes are after the ECU has noticed measurements out of parameters and taken preventative measures*, but before they're logged as DTCs, which in turn happens before a Check Engine light is set.

Do you have anything to add to or correct with that definition?


*I see boost cut in half

[electech]

Thanks for the definition clarification. Your posts are really helpful.

In my mind, I don't like the word "preventative". I would suggest something like "temporary corrective action".

The ECU responds to out-of-bounds measurements by applying substitute values. For example, if it doesn't like the airflow sensor (MAF) reading because it doesn't line up with the throttle position, but it isn't out far enough or long enough to set a code, it ignores the data it doesn't like and puts in it's own calculated substitute value. If it was an anomoly, and the data returns to within range, then the system goes back online and no code is set - but depending on the ECU, or if you have diagnostics on an attached scan tool, it will store a "snapshot" of the data at the time of the fault. This "snapshot" is, I assume, what you are referring to as a "pre-code".

Of course, if the sensor keeps being out of bounds, it will set a code, and stop asking the MAF for data until the next cycle. I recall testing the limits of a Chevy ECU by putting a dial-type low-voltage potentiometer (variable resistor) on the ground wire of the MAF. We could get the engine to surge and almost stall by adjusting the dial, all without setting a code. Of course if we pushed it too far, it would set a code and then the dial would become inactive - the ECM just didn't care what the MAF said any more.

The whole reason behind pre-codes is that there is ALWAYS some level of erroneous data. Even the RF interference of cell phones can mess up Hall Effect sensors. People would get really upset if their cell phones caused CEL or ABS lights to come on. (They actually did back in the late '80s, especially with crankshaft position sensors). Having a pre-determined range of error before the computer calls it a fault saves chasing a lot of unnecessary "ghosts".

As to DTCs ocurring before the CEL is set, that depends. Some codes will set the CEL as soon as they are logged, some have to occur a set number of times (or for a set period of time), and some codes will never cause the CEL to light, but will still be picked up on the scan tool. I think your definition is correctly worded, it's just a clarification more than anything.

Hope this helps!
(or at least makes sense)

[matthew1]

Thanks for your explanations, Electech. Interesting stuff. "Pre-code" isn't necessarily the most accurate term for the scenario it describes, but it's the best I could come up with. It's short and easy to remember.