13892_10 Datasheet, PDF (109/161 Pages) Freescale Semiconductor, Inc – Power Management and User Interface IC

FUNCTIONAL DEVICE OPERATION
ADC SUBSYSTEM
As follows from the previous description, using the CC requires a number of programming steps. A typical programming
example is given below.
1. SPI Access 1: Initialize
• Reg 9: Write STARTCC = 1, RSTCC = 1, CCCALA = 1, CCDITHER = 1, CCCALDB = 0
• RSTCC will be self clearing
• Register 10 is NOT to be programmed since by default the ONEC[14:0] scaler is set to 1
2. Wait for analog calibration period
3. SPI Access 2: Set scaler
• Reg 10: Write ONEC to desired value for CC use, for instance 2621DEC
4. SPI Access 3: Read analog offset and reset CC
• Reg 9: Write STARTCC = 1, RSTCC = 1, CCCALA = 0, CCDITHER = 1, CCCALDB = 0
• During the write access, on the MISO read line the most recent CCOUT[15:0] is available
• RSTCC will be self clearing
From this point on the ACC is running properly and CCOUT[15:0] reflects the accumulated charge. In order to be sure the
contents of the CCOUT[15:0] are valid, a CCFAULT bit is available. CCFAULT will be set '1' if the CCOUT content is no longer
valid, this means the bit gets set when a fault condition occurs and stays latched till cleared by software. There is no interrupt
associated to this bit. The following fault conditions are covered.
Counter roll over: CCOUT[15:0] = 8000HEX
This occurs when the contents of CCOUT[15:0] go from a negative to a positive value or vice versa. Software may interpret
incorrectly the battery charge by this change in polarity. When CCOUT[15:0] becomes equal to 8000HEX the CCFAULT is set.
The counter stays counting so its contents can still be exploited.
Battery removal: 'BP<UVDET'
When removing and replacing the battery, the contents of the counter are no longer valid. A battery removal is characterized
by the input supply to the IC dropping below the under voltage detect threshold, so BP<UVDET. To avoid false detection due to
short power cuts, the CCFAULT is set only after a long debounce of 1 second.
Battery removal when charging: BATTDETBS = 1
The battery removal detection as described previously, is not applicable when charging, since the charger will continue to
supply the application and the BP will not drop below UVDET. To still detect a battery removal, one can use the battery detect
function as described in the channel description earlier in this chapter. When the sense bit BATTDETBS becomes a 1, the
CCFAULT is set only after a long debounce of 1 second.
Analog Integrated Circuit Device Data
Freescale Semiconductor
13892
109