[s-cars] Motronic musings: ECU guru's take a look (long)

[Mihnea Cotet]

Sandy, see answers below...

At 06:51 18/04/2003 -0700, Sandy Sligh wrote:
>With regards to out of wack emissions at idle, I've
>tried to imagine the senario's that would create an HC
>rich exhaust:
>
>First:
>
>Especially at idle, isn't the A/F ratio monitored in a
>closed-loop fashion by the 02 sensor? if the sensor
>sees anything richer than the 14:1 mixture deemed as
>optimal for converter efficiency, shouldnt the ECU cut
>back on the injector timing?  Unless there is a
>minimum injector pulse width that the ECU is butting
>up against in attempts to lean out the mixture.  Could
>this be happening?

Yes it does.


>Now, what about this senario:
>
>There is a vacuum leak after the MAF. The ECU takes
>the MAF input data, calculates the proper injector
>timing to match that value and then the engine runs.

Well, the ECU takes a little more data than the MAF in order to calculate
the injector opening time, the input data are: TPS, MAF, IAT, CLT and a
"lambda factor". O2 is used to correct the basic injector opening time.


>However, the 02 sensor, seeing an oxygen rich exhaust
>due to more air being let in via the vacuum leak gives
>the ECU a reason to up the injector timing in order to
>compensate for the vacuum leak.  But, what happens in
>the ECU when the O2 sensor reading doesn't mesh with
>the data the MAF sensor is providing. The MAF says
>there shouldn't be a lean conditio, the O2 sensor says
>there IS a lean condition. Which component does the
>ECU yield to?

The O2 and this is called the adaptive function of the Motronic. It means
it can adapt itself to a vacuum leak or to a too rich condition within
certain limits. Once the limit is exceeded, it throws a code and starts
running badly.


>I would guess that the MAF wins that fight, or else
>the ECU would continue to correct for the vacuum leak
>and no one would ever know. The ECU renormalizes
>itself, and the car runs LEAN at idle, leading to a
>"lean misfire" condition.

See above/


>Third Senario:
>
>In the case the fuel pressure was higher than it
>should be, for whatever reason (AAN vs RS2): This
>would have the result that for a particular injector
>opening time, more fuel would come out than the ECU
>would expect. This would be sensed by the O2 sensor as
>a RICH condition. Wouldn't the ECU dial back the
>injector timing to compensate. Then again, there is a
>conflict in that the ECU see's no reason for the
>injectors to be outputting more fuel than necessary so
>perhaps it doesn't dial back the timing (instead using
>the new rich version of the injector output as the
>baseline?) creating a RICH condition at idle.

Well, the ECU can adapt itself better to a too rich condition than to a
"too lean", including at WOT. Don't ask me why but it is so.


>Fourth Senario:
>
>There is a misfire in the ignition system caused by
>faulty boots or faulty PSO or whatever. The misfire,
>preventing combustion, creates oxygen rich exhaust.
>The O2 sensor senses the O2 rich exhaust and informs
>the ECU, who in turn sees no out of wack signal from
>any of its other control inputs, thus continues to
>run. The ECU renormalizes the baseline O2 sensor
>reading and the engine continues to run unoptimally
>resulting in high HC emissions.

Nope, it keeps on adapting itself.



>Anyway, for anyone who understands the control systems
>present in the Motronic software, let me know where my
>reasoning is faulty or is correct.


I think I tried to as much as possible, especially given the extreme
examples :-)

HTH,

Mihnea