This is awfully confusing. Is there anything that would cause the values or algorithims for the Airflow Hz comp table to flip? At the moment I'm using a table similar to what Honki is using(higher values in the two left columns). So far I'm getting a pretty stable AFR's between 70F to 110F (1.03 to.78)MAF temperatures. However the first time I used this table it seemed that lowering the values in the columns on the right caused an extremely lean condition.

 

MRFred: I have temperature in the X axis because lr299gst rewrote the x axis scaling for me as you see below. I've done this assuming that air pressure is always 1 bar. I know that is not true but baro change is insignificant compared to temperature change for me. Ideally we would have two separate tables for each but this way made most sense for me and, yes, I agree most people shouldn't just rescale their x axis this way without this understanding.

I follow what you're saying about the table values basically being an IPW multplier. I just can't understand why Mitsu would have a table that ENRICHES the mixture when the density is already LOW. Like I said... that will happen even without an adjustment table. Why in the world would the OE scaling add fuel at high temps and pull fuel at low temps? That would lend to it getting even richer in the heat and leaner in the cold. (I understand this table was only intended for idle and cruise, but still... why would you use a table to accentuate an already undesirable phenomenon?)

 

I think the thing to keep in mind with the OE values is that there can be several reasons to enrich as the temperature goes up. The stock enrichment based on this table is pretty small anyhow.

 

honki24, I could be wrong, but this table doesn't compensate for density but it adjusts the fuel delivery based on density to deal with different conditions. Sounds confusing, but I'll explain what I mean and somebody else that knows the ECU can correct my thinking.

The ECU has additional tables to directly compensate for air density. This is seen in jscbanks' thread

When I read that, I read it to mean that there is a table strictly to compensate for changes in density due to air temperature (AIRTEMP, I assume this is the table that was mentioned earlier in this thread by you). There is a second table that compensates for barometric pressure (BARO). Finally, there is a table that adjusts the fuel delivery based on density conditions (TEMPBAROAIRFLOWLKUP).

Now I may be wrong, but it seems like the table you are adjusting is the last table I mentioned. From my understanding, that table is just to adjust the fueling under certain (low load) conditions.

It's not uncommon to lean out the AFR in high density air. This is done because if the ambient air is more dense, it requires less boost to reach the same mass flow rate/LOAD (remember, the factory ECU controls the boost pressure based on LOAD and not actual boost pressure), which means it's cooler and less likely to detonate. Yes, overall, IPW must increase to maintain a given AFR because of the higher air density, but this table seems to be used to change the AFR after the ECU has already compensated for the density.

 

You mentioned dense air requires less boost to reach the same airflow or load. Could this mean anyone who raises boost past a certain boost level (let's say to 25psi) would have to adjust this table?

 

I have no idea, I was just putting out a possible idea as to why the factory tuning may potentially lean out the AFR under higher density conditions.

The fact that it's set to unity above 200Hz though to me seems to say that this table is not for WOT tuning and is more likely only meant for driveability tuning.

 

The airtemp correction in jcsbanks formula is not the table posted earlier. That is an incorrectly defined table. The airtemp correction in jcsbanks formula is the same one used to calculate airtemp-compensated load and baro+airtemp compensated load. Its in this thread:

https://www.evolutionm.net/forums/ec...ion-table.html

I had previously recommended against adjusting this table as it will also affect the airtemp compensated loads which will complicate the results.

 

I see. Adjusting this table does make a difference on WOT. I have very close AFR's at the moment. I'm ranging from between 11.5(117F) to 12.2(55F). I just have to do a little more tweaking. It's a whole lot better then when there was no correction to this table. I had a range of 10.2 to 12.7 in the same temp ranges.

At one point I even tried to change the y axis to account for anything above 200hz, but it was way more difficult to tune than I thought. I've also learned if there are any ranges outside of the x scales (hits .53/180F on the MAF when it not on the table) the ECU does not interpolate the values. It went really lean past that point.

 

I'm coming around to the opinion that this airtemp table might be worth changing from its present gas law configuration. Given that the load calc and the IPW calc both use it, if we get the AFR consistent between different temperatures by using it then we are actually making it hit the "correct" load.

 

That would assume that there is nothing about air temp that affects other aspects of the air/fuel mixing quality, fuel density vs air density, and the combustion process. Not knocking your suggestion, but these are things that I wonder about with regard to IAT effects on AFR.

 

I think the ideal gas law model it comes with is good in theory, but it is not compensating for reality because of the other factors you mention. The table is being used to correct the load and IPW calculations in a way we understand, I think we should use it to make the AFR behave as we want it to. If we make the AFR less variable with temperature (but still leave in a sensible enrichment), we will also reduce the timing variation with temperature. If we see increasing knock sums in high temperatures we can adjust the timing vs temp table to compensate.

In my equation, the AIRTEMP*BARO compensations are part of the density calculation. The MAFSCALING*MAFSMTHG*TEMPBAROAIRFLOWLKUP are for sorting out the non-linearities of the MAF sensor.

 

Okay, I did a bit of logging on the way to work today. I was skeptical that the values in this map are direct IPW multipliers. Here's what I did.:

29F outside this morning
Did several pulls
These 2nd to 3rd pulls show what I've found:

My initial table:


Here is the log:

You'll see the AFR here is in the High 12s. (look at the little box by the cursor)

----------------------------------------------------------------------------------------------

Here is the inverted table:


Here is the log:

You'll see the AFR here is in the 11s. (look at the little box by the cursor)

In both of these cases temperature and baro are comparable, the only difference is the values in the cells. These values are not IPW multipliers. They do the inverse. Higher numbers decrease fuel and lower numbers increase fuel.

In both of these cases I should have been running "just to the left" of my 33F column. I would estimate that temp and baro would have put me 15% to the left of my 33F column and 85% to the right of my 106F column. So basically you could say that both of these pulls are running pretty much off of the cell at the intersect of 200Hz and 33F. As you see, when this value is higher my AFR is leaner. When it is lower my AFR is richer.

I calculate from these findings roughly:

An increase of .07 on this map made me run 1.4 AFR points leaner.

 

I will double check the disasm later. It could well be that it is a divider, these routines load up registers and some are x*y, some are x*y/z etc.

 

Has anyone considered that this table could also be used to compensate for altitude changes? Higher elevation less dense air and vice versa.

 

The table that I linked to and that table you posted about appear to be the same?