Right, stay posted cause i got a fk load of info to chuck in this post:
I spent a tonne of time checking out how to work out how to get enough fuel/load cells available for the CE engine to run boosted air values. Currently, the CE uses a couple tables to run: one which ill call fueling, is Load x RPM, filled in with AFRs, and the other, timing, is Load x RPM filled with Degrees of Advance BTDC (which i think doesnt take into account the OEM timing offset at the crack. Not a biggie).
The Load for the Fuel map goes in lots of 10% (10, 20, 30, etc up to 100 (ten columns), and RPM goes o, 500, 1000, 1500, 2000, 2500 etc up to 6500.
The Load for the Timing map also goes in lots of 10, but as 0, 10, 20, up to 100 (twelve columns). You might have counted that as 11 columns, but there is in fact 12. The final column is 260, and is the limp mode. Heavy retardation to make sure *poo* isnt going to go bang if something isnt running right. (RPM for Timing Map goes 0, 500, 750, 1000, 1250, 1500, 2000, 2500, 3000 up to 7000, and then a final row of 11000). Essentially, think of the fueling as 10x14, and the timing as 12x18 for stock CE.
Looking at Evo, DSMs, Subarus, all sorts from all over the world. Eventually i came across an idea/method the Evo guys use, which is sort of a shortcut to increasing the cell resolution. In essence, what i did is a cheaters way to gain resolution as we are restricted to what our maps are. To gain extra load values (e.g, from 0-100, then say, 110, 120, 130 up to 200 or 300), we need to grab that "space" from the ECUs memory. The thing is, the ECU does not like to give up the space if it isnt located direct in line. For example the ECU memory from byte 0-1000 might be for the load, but 1001-1500 is for the RPM, which means you cannot grab 1001 for fueling cause it will impact the RPM, and you cant grab 1501 cause the ECU doesnt like to split its memory/reference locations.
Thats the back end out of the way. What the Evo guys do, is either get ceddy or tephra mods for big maps, or both ceddy/tephra and then also expand their resolution tables out, which also tightens it up. Their tables might (example) go 0, 50, 100, 110, 115, 120, 125, 130, 135, 140 up to whatever the Evo table is, instead of the stock 0, 50, 70, 90, 100, 110, 120, 130, 140, 150, etc (Lancer is 12 columns, CM5A is like 18? Evo might be 25. Big maps is like 30 or something, because the Evo ECU does have the next memory cells available to be taken for more table resolution). The cheaters column "expansion" is done by removing a neighboring column, if it is a duplicate. If the Fueling Map has the exact same AFR fields for the RPM in the 40% load as the 50% load, and and then increases/decreases for the 60%, then you can cut the 50% column and replace it with the 60% column. This will give you a whole new column to play with! The idea is the cut out a couple columns, and it will give you enough space to play with the fields. In that above example, someone might choose to change the 50% column heading to 53%, so it would go 40, 53, 60. This is what i mean by both expanding out the columns, but tightening it as well.
I am not sure on the limit of the numbering, but I've taken that theory and tested it successfully. I have a fuel map that runs from 10%, 30, 70, 80, 100, 120, 140, 160, 180, 200% load (I was able to cut out the 20, 40, 50, 60, 90% loads as they were close enough to the neighboring cells), and a timing map that is 0, 10, 20, 50, 70, 100, 120, 140, 160, 180, 200, 260. You have to keep in mind that both of these table will reference each other, and a severe mismatch in the numbering will cause it to really fk up. Ive read its a very good idea to keep as much of the Load cells in check with each other as possible (e.g, if fuel map goes in lots of 20s (20, 40, 60 etc), the timing should as well, as much as possible go in 20s too) . In addition, Evos and other boosted cars dont run in boost (e.g 100% load and higher) all of the time, just like our CEs are run NA 100% of the time. Removing too much of the NA mapping (the area less than 100% load) can and will cause sluggish driving amongst a bunch of other problems, so the challenge is to keep the table as 12 columns long, but have 6 or so as below 100% load, and the last 6~ above 100% (a good idea would be 20, 40, 60, 70, 80, 100, 105, 110, 115, 120, 125, 130, which puts 130% load as the peak. Note that 100% load is WOT/0 inHg, but 200% is not 14.7psi, its close, but not exact.
Following all this, i tested out the above on a space ECU i got from the wreckers, and the car started, idled, and revved comfortably. I will need an idea of the timing and fueling/AFR values from a turbo car (ideally a CM5A) to fine tune, but i think its a safe bet that it should run easily enough for very low boost applications.