PIC24FJ16GA MICROCHIP [Microchip Technology], PIC24FJ16GA Datasheet - Page 179

• Manufacturer Part Number: PIC24FJ16GA
• Description: 28/44-Pin General Purpose, 16-Bit Flash Microcontrollers
• Manufacturer: MICROCHIP [Microchip Technology]

REGISTER 18-10:
18.2
The real-time crystal input can be calibrated using the
periodic auto-adjust feature. When properly calibrated,
the RTCC can provide an error of less than 3 seconds
per month. This is accomplished by finding the number
of error clock pulses and storing the value into the
lower half of the RCFGCAL register. The 8-bit signed
value loaded into the lower half of RCFGCAL is multi-
plied by four and will be either added or subtracted from
the RTCC timer, once every minute. Refer to the steps
below for RTCC calibration:
1.
2.
EQUATION 18-1:
© 2008 Microchip Technology Inc.
(Ideal Frequency† – Measured Frequency) * 60 = Clocks
per Minute
bit 15
bit 7
Legend:
R = Readable bit
-n = Value at POR
bit 15
bit 14-12
bit 11-8
bit 7
bit 6-4
bit 3-0
Using another timer resource on the device, the
user must find the error of the 32.768 kHz
crystal.
Once the error is known, it must be converted to
the number of error clock pulses per minute.
U-0
U-0
—
—
† Ideal frequency = 32,768 Hz
Calibration
Unimplemented: Read as ‘0’
MINTEN2:MINTEN0: Binary Coded Decimal Value of Minute’s Tens Digit; Contains a value from 0 to 5
MINONE3:MINONE0: Binary Coded Decimal Value of Minute’s Ones Digit; Contains a value from 0 to 9
Unimplemented: Read as ‘0’
SECTEN2:SECTEN0: Binary Coded Decimal Value of Second’s Tens Digit; Contains a value from 0 to 5
SECONE3:SECONE0: Binary Coded Decimal Value of Second’s Ones Digit; Contains a value from 0 to 9
SECTEN2
MINTEN2
R/W-x
R/W-x
ALMINSEC: ALARM MINUTES AND SECONDS VALUE REGISTER
W = Writable bit
‘1’ = Bit is set
SECTEN1
MINTEN1
R/W-x
R/W-x
SECTEN0
MINTEN0
R/W-x
R/W-x
PIC24FJ64GA004 FAMILY
Preliminary
U = Unimplemented bit, read as ‘0’
‘0’ = Bit is cleared
SECONE3
MINONE3
R/W-x
R/W-x
3.
4.
Writes to the lower half of the RCFGCAL register
should only occur when the timer is turned off, or
immediately after the rising edge of the seconds pulse.
Note:
a) If the oscillator is faster then ideal (negative
result form step 2), the RCFGCAL register value
needs to be negative. This causes the specified
number of clock pulses to be subtracted from
the timer counter once every minute.
b) If the oscillator is slower then ideal (positive
result from step 2) the RCFGCAL register value
needs to be positive. This causes the specified
number of clock pulses to be subtracted from
the timer counter once every minute.
Divide the number of error clocks per minute by
4 to get the correct CAL value and load the
RCFGCAL register with the correct value.
(Each 1-bit increment in CAL adds or subtracts
4 pulses).
It is up to the user to include in the error
value the initial error of the crystal, drift
due to temperature and drift due to crystal
aging.
SECONE2
MINONE2
R/W-x
R/W-x
x = Bit is unknown
SECONE1
MINONE1
R/W-x
R/W-x
DS39881C-page 177
SECONE0
MINONE0
R/W-x
R/W-x
bit 8
bit 0