How to identify map axis from header data

[JaboJG]

Hey guys

Can somebody explain how to easily identify map axis from header data on 8 and 16 bit dumps?

[JaboJG]

Anybody help?

[lee_hdi]

People more than likely won't give away trade secrets for free

[JaboJG]

Hmm... didn't realise these were trade secrets :-/

[Morette]

What info do you want?
Bosch often has axis size/axis break points before each map.

16bit is usually in decimal true figures, and the range will tell you what you are looking at.

8bit often has a register byte or identifyer which will tell you what the axis is, and what factor to use.
8 bit needs this as there isn't enough data width to show info like rpms if you only have 255 numbers to use.

for 8 bit, you need to think what they are. RPM and Load are the most common axis types, also engine temperature, air temperature, throttle position etc.

Sometimes similar ECU will share the description byte, sometimes they will change it.


Some ecus will also use remote axis, so one RPM axis is stored somewhere else, and used for all maps, and you might not find them at all.

[JaboJG]

Thanks for that info there Morette.

Have also been wondering about AFR tables. Are these parameters within ECU dumps or do we compile them ourselves from sample data i.e. doing minor changes and recording wideband sensor readings etc.?

[Morette]

I treat AFR tables as relative, if you need more fuel, increase the tables and see what happens.

Maybe some ECUs can hit them perfectly, but with many, the numbers in the tables won't match real life.
They will have been calibrated, so even if you convert them into something like Lambda or AFR, you need to change them, look at what has happened, and change again.

You will find the tables aren't truely Lambda, but often a correction factor centred on 128 (for 8 bit).
The software already calculates AFR, and uses the fueling table to calibrate this.

It isn't as easy as just setting the entire map to 12.5 AFR and leaving it to do the rest!

[JaboJG]

It isn't as easy as just setting the entire map to 12.5 AFR and leaving it to do the rest!

I know, lol.


The software already calculates AFR, and uses the fueling table to calibrate this.

Does WinOLS do this?

[Morette]

No, WinOLS doesn't do much, but it allows you to do anything.

To add another level to make things more complex, the "Load" axis isn't the airflow for most Bosch 8 bits, it is ms of injector opening time to reach Lambda = 1.
It is proportional to the airflow, but not a measure of it.

Siemens would use a different method, their load axis is relative filling, so Volumetric Efficiency.

In a Siemens fuel map, it looks like you would expect, more air = more fuel, Bosch fuel maps are a deviation factor from lambda, so they won't look like a Siemens ramp.

Different ECU families can have different methods. It sn't made easy for a person, but easy for an old processor.

[JaboJG]

To add another level to make things more complex, the "Load" axis isn't the airflow for most Bosch 8 bits, it is ms of injector opening time to reach Lambda = 1.

Does seem like it gets quite complex. I assumed the load axis would always be a kPa value.

Ah well, my MPPS v12 will be here soon and I'm going to dump my Citroen C5's Marelli ECU, open it in WinOLS and stick with it until I can make sense of it.

Thanks for the info so far :)

[Morette]

Marelli may be kPa, they liked to use MAP sensors, so you may find none of the info above applies, and there might not be nice easy to see axis data either.

[ducatisti]

Marelli may be kPa, they liked to use MAP sensors, so you may find none of the info above applies, and there might not be nice easy to see axis data either.

Agreed . Marelli axis data seems to be scattered all around the file and its a case of either reading code or using AFR data plus the dyno to find the adjustment points :(

I cant read code so they dyno data tells me what i need ....eventually....

[JaboJG]

The software already calculates AFR, and uses the fueling table to calibrate this.

What software can show me an AFR table?

[Morette]

By software, I mean the ECU program file works this out already.
In fact, it might not deal with AFR at all, just having a programed injector opening time to match the 0-5v signal from the MAF or similar.

Because it already does this, you will see an 8 bit Bosch fueling table at lambda being full of "128".
Use factor 0.0078125, offset 0, precision at 2-3.

It will look like lambda, and act a little like lambda, but the truth it when it aims for lambda automatically, it won't reach it.
The fueling table then corrects the difference between the lambda it wants, and the lambda it gets.

To you, this means increase the table for more fuel, decrease it for less fuel. Monitor, and readjust.

[3abas27]

By software, I mean the ECU program file works this out already.
In fact, it might not deal with AFR at all, just having a programed injector opening time to match the 0-5v signal from the MAF or similar.

Because it already does this, you will see an 8 bit Bosch fueling table at lambda being full of "128".
Use factor 0.0078125, offset 0, precision at 2-3.

It will look like lambda, and act a little like lambda, but the truth it when it aims for lambda automatically, it won't reach it.
The fueling table then corrects the difference between the lambda it wants, and the lambda it gets.

To you, this means increase the table for more fuel, decrease it for less fuel. Monitor, and readjust.

thanks to morette for information