A New Spin on MAP and Tuning Fuel

[sstevojr]

Just for Kyle :p This has been accumulating over the last month or two, chopping out the best parts and moving over to these neck of the woods:
I originally started to explore tuning the MAP tables in hopes of cleaning up my AFR’s. The 3x’s MAP tables can be found in Fuel, as they provide the basis for establishing Load. This Load is then applied to all the Fuel tables (with RPM as the other cross reference point) to establish your AFR’s. Many, if not most, people take the general approach of using the fuel maps as +/- to establish desired AFR for WOT. What if you could set an AFR value in your Fuel table, and it directly corresponds to an actual value rather than a +/- value? What if you could get rid of the lurching and bucking at slow speeds? What if you could narrow the apparent disconnect in Load between MAFcalc and MAPcalc? This is especially important for SST equipped vehicles, in that the SST seems to use anticipated Load for…..well everything! I wanted to put this out there, how I went about getting to this point, to let others who might be inclined try and see if they can replicate the same results I’ve achieved. This is something that can’t be done in a generic base-map sort of way, it’s for self-tuning. It’s definitely time consuming (not something that can be done in a 2-3 hour dyno session), but the resulting smooth driveability and ease of future tuning are worth it to me. If you are the type who finds it necessary to smooth EVER table in your ROM, this is for you, lol!
Step One: Establish your MIVEC and Boost Maps
The load as registered by our cars is directly related to the volume of airflow, i.e. Volumetric Efficiency. Boost and MIVEC play the largest roles in this, due to the dramatic affect these can have on cylinder volume. For example, based on my (now old) Boost and MIVEC curves there is a dramatic increase in airflow starting at 2500rpm, and carries to 6000rpm;
http://i902.photobucket.com/albums/ac226/steckers/MIVECBoost.jpg
Step Two: Log RPM/Load/psi
After determining your airflow, you need to log. I use about 100,000 lines of logs over several different days and many different driving styles; it may sound like a lot, but it’s only a couple hours worth of logs. The goal is to hit as many RPM/psi combinations as possible, so you can get a good idea of what Load is actually doing. Take all this data, drop it into excel and create a pivot table; set it up exactly as the MAP tables, and change Load to ‘average’. To make my life simple I group the RPM’s by increments of 250 and psi by 0.5’s;
http://i902.photobucket.com/albums/ac226/steckers/VEtable.jpg

[sstevojr]

Step Three: Scale your MAF and Calibration Table
First, you will need to set the Calibration Fuel Map to 100, otherwise the ECU will spit out a % of the Fuel map. Second you need make sure your MAF is scaled properly. Swiftus has an awesome tool to help scale you’re MAF, check out his thread for details. The only other thing I do after using his tool, is take the two columns of data and drop them into excel. Then I create a scatter plot and tweak the values slightly just so that the graph has the cleanest curve I can make (yes, I’m totally OCD);
http://i902.photobucket.com/albums/ac226/steckers/MAFscatterplot.jpg

Step Four: Create the MAP Table
This is where things get interesting, and where I had the most difficulty. It’s not possible to hit all the cell combinations by DD and some spirited driving. This is where knowing your MIVEC/Boost curve comes in real handy. The stock MAP tables look like this, EVO and RA respectively;
http://i902.photobucket.com/albums/ac226/steckers/MAPstockEVO.jpg
http://i902.photobucket.com/albums/ac226/steckers/MAPstockRA.jpg

[sstevojr]

Many take the first step of smoothing, and get something along the lines of the top graph. But based on the logs, and a continuation of the how the Load was being plotted, I came up with the bottom graph. Neither one is correct (EDIT:'accurate', the top one can be used in a simpler fashion), and if used will have some ‘odd’ side effects (letting off the throttle, idle, WOT shifting, etc);
http://i902.photobucket.com/albums/ac226/steckers/MAPalt1.jpg
I experimented with about a dozen variations, but kept getting unintended side effects. Until I finally figured it out; Load will rise at one rate while in vacuum, and a different rate in positive pressure. The final graph that seems to have solved most my problems looks like this;
http://i902.photobucket.com/albums/ac226/steckers/MAPfinal.jpg
There are a couple of key features that you need to adhere to: 1) below 1000rpm needs to drop off substantially as shown; 2) the angle of the Load increase needs to change when crossing into positive pressure; and 3) keep idle as close to stock as is reasonable. This allows me to have actual AFR within 0.3 of whatever value I drop in my Fuel Map (close enough, lol):
http://i902.photobucket.com/albums/ac226/steckers/Fuelfinal.jpg

[sstevojr]

Warning's
First: if you read the post and didn't understand a significant amount of what I wrote, please please do not attempt any of this! There really are numerous ways that something can go wrong, and w/o knowing exactly what is being done it would be extremely difficult to try and fix it.
And Finally: For the LOVE OF GOD, do not drop in the same values in your ECU as can be seen in the above examples. This should not need clarification.

[sstevojr]

Got to spend a little time on the dyno (thanks Mike), logs helped create a more accurate version, especially for the top end:
http://i902.photobucket.com/albums/ac226/steckers/MAP2.jpg

[sstevojr]

So I've had a slight problem with leaning out on up-shifts in 1st and 2nd, but I think I finally got it resolved. It still leans out a little (no more than 0.5-1.0 afr, which is for a half a second, and reasonable in my book), but there are no longer any side effects. I would occasionally get a real bad leaning out in the 1st to 2nd up-shift, causing the car to fall on it's face for a split second (AFR of 17-17.5 :mad:)
But I think I've figured it out by changing my MAFhz bins cross-over points. About 2 months ago I changed the Open/Closed Loop X-over to 100 load max, but the MAF bins were still left at ~140-160 load (350). I started digging into my logs and noticed my fuel trims were fighting me in one small area. It seems like the fuel trims were trying to correct all the medium throttle driving (80-150 load), despite going to Open Loop at 90-100 load (when the FT no longer had a say in the matter). So when I shifted it was quickly dropping down into this small region of -7 fuel trim (I apparently didn't look closely enough to this region, as I spent such little time in it).
Reset the MAFhz bins to 300 instead of 350, and the fuel trims this morning settled on -1.1 LT w/ a +.7 ST for that small region, success!! :D

[sstevojr]

Update:
I got a 3" CBE installed, giving me a full TBE (I have been on a stock 2.38"CBE up until now). If changing your airflow (i.e. exhaust, cams, intake, etc) you will need to re-create new tables. This CBE caused me max out my WB (10.0 afr) at only 17psi :eek:
That's how much MORE cfm the engine can flow by removing the stock CBE ;)

[sstevojr]

Still working on the top end, but I think I got the bottom end re-sorted. It crazy how small the difference is between a strong, stable idle (bottom) and a car that constantly stalls (top):
http://i902.photobucket.com/albums/ac226/steckers/stallnostall-1.png

Again, top one stalled out when coming to a stop (crossing from positive to negative pressure, the MAP load was too low, causing fuel to use that value and leaning out).

[sstevojr]

I have a track day on Sunday, so I'm trying to get everything sorted out before then, haha.
I've tried digging around but couldn't find any helpful threads about MAFhz and what values equal what rpm/load. So I did a similar thing with logging/pivot table for MAFhz to find out what was actually happening, rather than continually guesstimating:
http://i902.photobucket.com/albums/ac226/steckers/MAFhzPT.png
http://i902.photobucket.com/albums/ac226/steckers/FuelControl.png

That is what I came up with, based on when I want OL/CL to occur, and my HOFT.

[sstevojr]

<DataListItem DataLog="N" Color="" Display="MAPCalcs"
LogReference="MAPCalcs" RequestID="238087CE" Eval="x*100/16384"
Unit="load" MetricEval="" MetricUnit="" ResponseBytes="2" GaugeMin="0"
GaugeMax="255" ChartMin="0" ChartMax="255" ScalingFactor="1" Notes=""
Priority="1" Visible="False" />

<DataListItem DataLog="N" Color="" Display="MAFCalcs"
LogReference="MAFCalcs" RequestID="23808F88" Eval="x*100/16384"
Unit="load" MetricEval="" MetricUnit="" ResponseBytes="4" GaugeMin="0"
GaugeMax="255" ChartMin="0" ChartMax="255" ScalingFactor="1" Notes=""
Priority="1" Visible="False" />

These should work for logging on 2010 X/RA.

[BravoZero]

Post whoring much? :p

Good info dude!

[KyleKitchen]

This is beautiful.
Thats as detailed of a walkthrough as you can get.

[crak]

I have an idea of what's going on, but I am no where near tweaking with numbers yet. You guys are lucky to have friends in town that can help explain how to do what you are doing.

Good research though! I congratulate you.

[sstevojr]

Just for clarity, and to be redundant:
Warning's
First: if you read the post and didn't understand a significant amount of what I wrote, please please do not attempt any of this! There really are numerous ways that something can go wrong, and w/o knowing exactly what is being done it would be extremely difficult to try and fix it.
And Finally: For the LOVE OF GOD, do not drop in the same values in your ECU as can be seen in the above examples. This should not need clarification.

[KyleKitchen]

This is meant as information sharing for those of us that do know how to tune and as a way to get those of you that are just dipping your toes into tuning a little more curious.