Hi,
I saw the application report (A Simple Glass-Breakage Detector Using an MSP430 MCU (Rev. A) ) and I have a question about pre-scaling process in Figure 12.
In figure 8, the summation of p(n) is two words that represented to Signal_overflow_count (high word) and iinteg_total (low word), and the summation of OUTP(n) is two words that represented to the Filtered_signal_overflow_count (high word) and integ_HPF_total (low word).
In figure 12, in order to easy to calculate the ratio of the total integration and the total HPF integration for next process, it seems to do pre-scaling.
However, in the pre-scaling process why the lower word execute right shift and the high word only execute decrement?
I think the ratio of the the total integration and the total HPF integration should be the same as after doing pre-scaling, but it actually is not.
for example:
- the total integration: 0x49BC0
- the total HPF integration: 0x19858
- the ratio: 2.889
After pre-scaling:
- the total integration: 0xF9BC
- the total HPF integration: 0x1985
- the ratio: 7.786
It has a big different ratio between 2.889 (raw) and 7.786 (after pre-scaling).
Is this is a bug or special algorithm?
