 Map 1
The exciting question is how a characteristic map is memorized. It must be clear that data is a series of numbers arranged in sequence. If you didn't create it yourself or at least know its structure, you can't do anything with
it. Even if you could completely change the content, you wouldn't know what you were doing.
Let's try entering data into a characteristic map. We are currently doing this manually, which is, of course, completely impractical. To keep things simple, let's assume 8 bits for memorizing our ignition timing. So we could
enter a maximum of 255 different values. We say that anything earlier than 60° before TDC is not possible, so that is our zero point. From there, we continue counting beyond OT.
5° after OT would then be 65. So that should be possible. So now we enter the ignition timing determined for 12 different loads at 600 rpm, then the for 1000 rpm, for 1500 rpm, and so on, always 12 values. There must be
12 values, whether the motor could be operated at this value or not. Then just put in a 255, it won't be queried later anyway.
If, during operation, the control unit receives a speed of 2900 rpm and a load of 2.7 V from the corresponding sensors, the processor skips the first 5 * 12 values and then goes through the following column of numbers
until it reaches the value corresponding to the voltage value of 2.7 V. Let's assume it's the sixth.
Incidentally, jumping is innate to the processor. It often only needs one beat to do so. That's unimaginably fast. Even one of the oldest processors we know of, that of the Commodore 64 from the 80s, operated at a clock
frequency of 1 MHz. Two thousand cycles are possible during the aforementioned 2 ms for combustion. Today's processors, including those in motor vehicles, are at least a thousand times faster.
Of course, these enormous possibilities are not only used to adjust the ignition timing according to load and speed. For example, the temperature of an engine is an important influencing factor. You may already suspect
how the existing map will be expanded to include their influence. You simply add another 12 different temperatures to the measurements and enter the 144 values twelve times in a row.
Now the processor must first make large jumps to the correct temperature, then smaller ones to the correct speed, and then individual steps to the correct load. Can you now imagine that, provided there is sufficient
memory space, even more parameters are possible for aligning the ignition?
It is important to note that the necessary steps, e.g., for advancing the ignition when the engine speed increases, are not all equal. This means that sometimes you get better results if you advance the ignition from 3500
rpm to 4000 rpm by only 3° instead of 5°. This was not possible with the old centrifugal force adjustment, but is no longer a problem today.
There are even engines that require, for example, a slight retardation of the ignition timing above 5000 rpm. This has also only been possible since the introduction of fully electronic ignition. Incidentally, not
every operating point is necessarily addressed during tests. Firstly, you have the values from previous engines and only check critical areas, and secondly, the software can independently generate intermediate values
(interpolate).
| 
