RomRaider

It only interpolate between the cells ... should be liniarized interpolation, so no higher polynom.
Also you have to rescale you axis ... done interpolation in excel.

values arent correct ... Why should engine now need more fuel in lower loads / rpm ...?
It doesnt get more air there.
Also you need extrapolation not interpolation ... and its only approximation you need to log and tune by yourself for better values after all!

So nearly you done things you could do if you change injectors for rescale tables for beginning ... but you dont you want extrapolate not interpolate.
If ecu could extrapolate you can see by reverse engineering , disambling like via ida pro...

Yes I get your point, which is what I commented about in my previous post.
What makes sense to me is ADDING rows and columns to the current table cells (including axis of course) in a linear fashion, all the way to 7000rpm. But given that I can't paste tables in RR even with the current number of rows/columns, how can I do it with a higher number of rows/columns?

in simple words:
no there is no option in RR to do this and it isnt needen.
There is a table for chaning x and y values of axis, but sometimes need to correct some maps by yourself and not with axis descriptor.
So put in last colum your 7000 or 6800, 6900 this 200rpm isnt so important ... and rescale .
After this do remap (have a look re-map vs re-scale) the map via logging and testing.
So you are done.

Ok so here is how I proceeded:

- to reach 6944rpm with the same amount of rows, I increased the "rpm steps" for some of the values, as such:
* I left the first 5 rows alone, to help precision around idle
* I increased the "rpm step" between 5th and 6th row by 384rpm (median value between 256 and 512) instead of 256
*the next one by 480 instead of 256
* the one after that by 480 as in the original table
* all next "480" original steps by 512 (there were three 480 steps in between 512 ones originally)

- I then divided the new rpm values by the old ones where indicated, to come up with a ratio
- I multiplied the corresponding IPW rows by that ratio, so IPW increases proportionately to RPM
- I left the load values all the same, as they're not RPM dependent.

The one thing I'd like to draw guys your attention at is that the "384rpm" step doesn't match any of the original steps' values (96/256/480/512) and I don't know if the DME would be able to recognize a "rpm step" value that is not present in the original table.

Attached is the new table, which I hope makes sense. Thanks for your feedbacks in advance.

Bump, how does that table look guys? Given how sensitive a matter fueling is, I'd rather not screw this one up
I guess if it's ok I should do the same with full load enrichment and part load ones right?

Your way of thinking is completly wrong to me.

Just change the rpm axis in your fueling table.

That's what I did, didn't I? The values are different from the original ones for that table. And I updated the IPW cells proportionately. The increase in IPW's duration matches the increase in RPM. If I change the axis without changing the IPW values I end up "giving" the same amount of fuel for a higher rpm. I know as rpm increase there is less time for injectors to operate given that the combustion cycle takes less time, but there has to be some sort of compensation I guess?

Attached is a screenshot of the original base fuel table, for comparison purposes.

No, you are thinking the wrong way.

Ipw is depending on load only. Not rpm.

That why you need bigger injector on turbo cars because you would need to open longen than the cycle duration.

I hate to disagree, but I don't think that's the case. If IPW was dependent on load only then why BMW would populate the table with an rpm axis in the first place? There would be only a load axis with corresponding IPW values wouldn't it?
I'm only quoting theoretical stuff here, but as rule of thumb, for a same amount of load, higher rpm means higher fuel requirements up to the point where the duty cycle of the injector can't catch up with the combustion cycle, which makes injectors go static...which explains why IPW decreases after a certain rpm point. It doesn't mean that load is not the primary factor to be taken into account, as the table shows, but if it was the only one, then I don't think there would an RPM axis.
Anyway, my method consists just of updating the IPW values proportionately to keep sticking to variations as intended by BMW (decrease or increase IPW by z% between rpm level x and rpm level y for e.g.). That way if the DME interpolates it will do so between more accurate IPW values given that they are proportional to the new RPM axis rather than the old one. I might very well be wrong, but that does seem to make sense to me...

If we take as an example the last rpm rows as shown here, in the original table, we can see that for a same given load, IPW increases until the last cell, where it decreases.

Why are you editing your fuel table? Is it just because you raised the rev limiter?
Is the car a daily driver?

Yes it is a daily (well, mostly on weekends but I walk to my workplace most week days). I'm currently working on something in between what I originally did and what dx4picco advised. It goes like this:

I understand that my approach was somewhat wrong because there are other "invisible" parameters taken into account to populate that table, with obviously lots of logging by BMW, and that IPW do not necessarily/always increase linearly with rpm (sometimes it does, sometimes it decreases). But still , I couldn't reasonably (IMHO) leave the same IPW's for higher rpm tresholds, to avoid running lean, at least for the rows where IPW clearly increases with RPM.

So I'm upscaling the RPM axis and modifying IPW's, but this time proceeding by analogy with what is in the original table. For e.g. the RPM row which shows 5408 in my new table, I'll paste the values for the 5504rpm row in the original table. And for the last RPM row (6944), I'll paste the original value for 6496 and slightly increase it, just to be on the safe side, since I don't know what pattern IPW would follow (up or down) at that rpm (most likely decrease, but how about combustion chamber - insufficient- cooling? I'd rather be too rich than too lean).

lets try to be short.

In gasoline engine you try to reach lambda 1, meaning stoichiometric ratio between air and fuel, namely 14.5-14.7 afr.
Now lets see how bmw ecu work. It uses a MAF sensor, that measure the air mass going into the engine.
It then divides the air mass by the number of cycle happening, namely rpm. This air mass (kg/h) divided by the number of cycle per time, is the Load (mg per stroke, mg of air per intake stroke).
Now, 600mg of air per intake stroke at 1500 rpm, to be at labmda 1 you need 15ms injector pulse width, which is 18.75% duty cycle on your injector.
600mg of air per intake stroke at 6000 rpm, to be at lambda 1 you need exactly as much fuel per stroke as at 1500rpm, 15ms, equivalent to 75% duty cycle of the injector.

why is the map 2D table and not 1D? because its not the only map ruling the ipw, you have accel table , deccel table, vanos correction table and all interact. Can you run same ipw in each load column? yes I do. Can you change to higher PW for higher rpm? yes you can if you decide to control your WOT enrichment with main table and not the wot one.

No offense but try to school yourself a bit more on the MS platform to avoid mistakes.

Ok so here is the "new" table. Basically what I did is this:

- take a "new" (after rescaling) rpm axis value
- compare it to the closest value(s) in the original table:
* if it's very close to an original value (say 200rpm or less), just copy and paste the IPW values from the original table to the rescaled one
* if it somewhat falls in between two original rpm values, compare IPW in rows respectively associated to both OV and copy and paste the richest IPW values between them to the new table
- For the last row (6944rpm), paste the values from the OV's last row (6496) and richen them slightly.

No offense taken. Answers in blue in your text to make things less tedious.

I really think you are over thinking this since your car isnt going to be hitting the new rev limiter one hundred percent of the time. Horse Power drops off before 6500rpm so there is absolutely no point in running any gear to rev limiter.

If you wanna just smash the gas in neutral every once in a while, to hear your newly set and aggressive rev limiter, using the stock fuel table is not going to be an issue, unless perhaps you plan on putting a brick on your gas pedal and walking away from the car.

But your last few posts about calculating IPW and such has intrigued me, so I did some quick reading and it seems to me, that to do the calculations as correctly as possible without actually having a wideband afr gauge....you would need to figure out the volumetric efficiency for your particular setup. Im sure there is a formula for “reverse” calculating the stock fuel table to get an idea of what BMW feels is the VE of the engine.

It is definitely interesting, and at some point, I myself will have to dive into my fuel tables as well.