It is possible to read back filter calibration result. In datasheet it is showed that the upper most 8 bits should be all zeros and the filter calibration is placed in the 8 lower bits. In my application the bit number 9 (DB8) is one. Should this bit be treated as zero?
The calibration result are 8 bits and if I want to save this result and store it in "IF_FILTER_ADJUST" it should be calculated as "FILTER_CAL_READBACK" - 128. This calculation has the effect that the upper most bit (DB7) is inverted and nothing else. Then this result should be stored in "IF_FILTER_ADJUST" that occupies only 6 bits. How should I treat the 8 bit result so it can take place in only 6 bits?
My read back is 0x78 and after the subtraction the result is 0xF8. If I truncate it to 6 bits or if I right shift two times the result is a very big negative number and the receiver will not work good at all after I reprogram the register 5 with this value for "IF_FILTER_ADJUST"!
If I program register 5 with "IF_FILTER_ADJUST" set to all zeros then the receiver will work OK.