I've been playing around in the ralliart maps. I looked up the constants for the math in EcuFlash. I discovered people had labelled these maps high octane ignition map and low octane ignition map.

I am wondering if anyone has discovered how these things work? It appears to me that for the timing, it loads the low octane timing number and the high octane timing number and then performs some interpolation between the two. Since I haven't mapped out all the variables in memory yet I'm not sure what it uses for the interpolation weight value but I suspect it may be the knock number. If this is the case then the difference between these maps is actually the max knock retard.

It's also possible that the ralliart uses a different strategy.

Any comments?

-Michael

 

If you search you will find this has been discussed multiple times.

There is a octane rating on the 7/8/9, which is labeled as learned knock retard on the Evo 10.

It is 100% when everything is fine and no knock is registered. Once a certain amount of knock is registered the octane/LKR percentage drops % and starts interpolating between the high/low octane maps.

 

Ok, I did do some searching and didn't find anything. I guess then that my view of the ralliart knock strategy is consistent with the evo then.

 

One more thing. I read through 8 pages of posts about the open loop/closed loop switches and did not find any mention of this... Both conditions appear to have to be met in order to switch to open loop. Each condition flips it's own bit but at the end of the day the one I suspect is the open loop switch seems to depend on both the TPS and load...

Now on to figuring out how the load is actually calculated...

-Michael

 

yes the OL/CL switch is based on a couple tables, TPS/Load/RPM

 

Its more that OL is the default condition, and there are actually a boat-load of conditions that need to be satisfied for CL to activate. There are a series of several subroutines where all the conditions are checked. If you're interested, I can make my disassem available to you.

 

I did some more playing. Zero'd out the load map and it then switches into closed loop when either condition is met, not both as I had originally thought. Sometimes I get things screwed up when comparing doing a compare like CMPLT R18,R4... UGH!

Does anyone know exactly how the load is calculated on the evo?

 

https://www.evolutionm.net/forums/ec...tor-pulse.html

How deep does the rabbit hole go?

 

Interestingly the ralliart doesn't use a KV type MAF sensor. It seems to have an analog/current type output. I'll have a look at the link. Thanks.

 

DSM cars (Diamond Star Motors) use a specific type of sensor called a "Karman Vortex airflow sensor". The upper inlet of the sensor has an obstruction in the airflow path that generates a constant stream of vortices (think of them as miniature tornadoes) directly behind the obstruction.

The velocity of the incoming airflow determines rate at which these vortices are spun off. By placing an ultrasonic speaker and microphone across the stream of vortices, their rate can be counted. (the vortices spin off in alternating directions and thereby produce a alternating FM shift of the received signal from the speaker)

In addition to measuring incoming airflow velocity, the mass airflow sensor unit also measures air temperature and the local barometric pressure in order to provide the ECU all the data it needs to do a true determination of the MASS of the air that the engine is consuming.

Generally speaking, Mass Airflow measurement is more precise than the speed-density system which has been largely replaced in newer automotive designs. Its (MAF) calibration is not affected when mods that change the engines volumetric efficiency are performed.

The MAF used on DSM vehicles is actually 3 separate sensors that output signals to the EMS allowing the computation of mass airflow into the engine. The main output is a frequency signal that is proportional to the air flow velocity through the sensor. The second is a pressure sensor and the 3rd is a temperature sensor.

Since the physical configuration (x-section) of the MAF is already is known, everything necessary to calculate the volume and therefore the mass flowrate of the inlet air is present.

Flow sum number is proportional to mass flow per engine revolution and is divided by Flow max to get the load percentage.

The ECU counts pulses from the airflow meter, and estimates fractions of pulses between cam timing events. Then converts for air density and scales the output into load units by dividing by Flow max.

 

So does the EVO MAF send the measurement signal as frequency or as an analog signal?

-Michael

 

Frequency. Its KV air velocity.