Originally Posted By: badtlc
For about the 100th time. Knock can also be controlled through A/F ratio. Mazda runs rich (as per their engineers) purposely to lower octane requirements. They inject fuel twice for each ignition. They intentionally run rich to CONTROL KNOCK and prevent detonation. The extra fuel absorbs more heat and lowers combustion chamber temperatures. This allows them to run high compression on 87 octane gas. They also have regular timing control.
Well Mr. "For the 100th time" I guess you missed this line of my post:
Quote:
However, you add load, and the ECM will enrich the mixture to compensate for said load and when you go WOT, it will richen the mixture further. This lowers cylinder temps, increases power output
and wards off detonation. All standard practices here.
Quote:
Mazda isn't the only one that does this. I think most of the euro companies do the same thing.
Adding fuel to ward off detonation is built into the base load table of pretty much any ECM. There is an ignition timing table to go along with it. The ignition timing table is altered by data gleaned from the knock sensor. The A/F target table is manipulated during closed loop by feedback from the O2 sensors. Whatever A/F the engineers shot for at a given load/engine temp/air temp, the ECM will try to obtain, and will modify the amount of fuel added to meet those numbers.
In addition to this, ignition timing will be added/removed based on knock-sensor feedback relative to its base-table values. You then have these multiple tables running in parallel, the base tables, which are referenced for conditions not met/experienced in the learned/feedback table and the modified ECM-created table(s) that reflect information learned through feedback/CL operation from which it pulls values for parameters that it populated in that table during CL/feedback operation. These tables are always in flux as they represent whatever combination of parameters for a given throttle angle, load %, ACT, ECT and timing position met the targets of a proper AF related to the desired value programmed in the base table and no knock relative to feedback from the knock sensor relative to the values in the timing table.
When you run higher octane fuel, the A/F stays the same. The ECM doesn't know you put 93 octane in the tank and that it could subsequently run leaner without inducing knock. The opposite of which appears to be the argument you are making here, is it not? (this is based on this quote by you
Originally Posted By: badtlc
I honestly think the reason for low fuel dilution is using exclusively premium gasoline. This engine is designed to enrich the A/F ratio to cool the combustion chamber and prevent knocking. I think the premium allows the ECU to lean out the ratio.
What a conventional feedback based system WILL do is add more ignition timing based on knock-sensor feedback. O2 sensor feedback remains the same regardless of the octane of gasoline used.
My impression is that your premise here is that the engine uses the knock sensor to manipulate A/F in conjunction with timing as some sort of ultra-intelligent feedback mechanism and that is "knows" you have higher octane fuel in the tank based on that feedback and is subsequently running leaner than "designed" because of this. And this in turn is what is lowering your fuel dilution numbers, correct? This would of course mean that the base table for A/F is relative and range based rather than static in regards target ratios and that instead of the ECM attempting to achieve a target A/F for a given operating condition in CL operation, it instead attempts to control knock within a given range of acceptable A/F values by manipulating fuel in addition to ignition timing. This would mean there is no single "target". That "as lean as possible" with "as much timing as possible" is the ECM's goal and that it shoots for both of those things constantly with nothing more than a ceiling and basement in place for its ranges.
If so, this is very easy to verify. Your car would HAVE to have wide-band oxygen sensors for this system to work. If that is the case, then a decent scan tool should be able to access the oxygen sensor information. I've done this in my sister's 330i, which has wideband O2's in case you were wondering, and I've graphed its feedback with my Auto Enginuity software.
You simply need to log the A/F for the same operating conditions on 93 and 87 octane gasoline. If the system is of the traditional feedback type, the O2 sensor graphs for both types of fuel will pretty much mirror each other. If, on the other hand, the system operates as you think it does, it should be significantly leaner on the 93 octane fuel than it is on the 87.
Do you have a scan tool?