The P89LPC920/921/922 is a single-chip microcontroller designed for applications demanding high-integration, low cost
solutions over a wide range of performance requirements. The P89LPC920/921/922 is based on a high performance processor
architecture that executes instructions in two to four clocks, six times the rate of standard 80C51 devices. Many system level
functions have been incorporated into the P89LPC920/921/922 in order to reduce component count, board space, and system
cost.
Pin configuration
20-Pin TSSOP, DIP Package
P0.1/CIN2B/KBI1KBI0/CMP2/P0.0
20
P0.2/CIN2A/KBI2
19
P0.3/CIN1B/KBI3
18
P0.4/CIN1A/KBI4
17
P0.5/CMPREF/KBI5
16
VDD
15
P0.6/CMP1/KBI6
14
P0.7/T1/KBI7
13
P1.0/TXD
12
P1.1/RXD
11
P1.7
P1.6
/P1.5
RST
VSS
XTAL1/P3.1
CLKOUT/XTAL2/P3.0
INT1
/P1.4
SDA/INT0
/P1.3
SCL/T0/P1.2
1
2
3
4
5
6
7
8
9
10
Logic symbol
KBI0
KBI1
KBI2
KBI3
KBI4
KBI5
KBI6
KBI7
CLKOUT
CMP2
CIN2B
CIN2A
CIN1B
CIN1A
CMPREF
CMP1
T1
XTAL2
XTAL1
PORT0PORT3
VDDV
SS
P89
LPC920
/921/
922
PORT1
TxD
RxD
T0
INT0
INT1
RST
SCL
SDA
2003 Nov 6 9
Philips Semiconductors
User’s Manual - Preliminary -
General Description
Block diagram
2KB/ 4KB/ 8KB
Code Flash
Port 3
Configurable I/Os
Port 1
Configurable I/Os
P89LPC920/921/922
High Performance
LPC920/921/922 CPU
256 Byte
Data RAM
Internal Bus
UART
Real-Time Clock/
System Timer
Crystal or
Resonator
Port 0
Configurable I/Os
Keypad
Interrupt
Watchdog Timer
and Oscillator
Programmable
Oscillator Divider
Configurable
Oscillator
CPU
Clock
On-Chip
RC
Oscillator
I2C
Timer0
Timer1
Analog
Comparators
Power Monitor
(Power-On Reset,
Brownout Reset)
2003 Nov 6 10
Philips Semiconductors
User’s Manual - Preliminary -
General Description
Pin Descriptions
MnemonicPin no. Type Name and function
P0.0 - P0.7 1, 20, 19,
18, 17,
16, 14,
13
1I/OP0.0Port 0 bit 0.
20I/OP0.1Port 0 bit 1.
19I/OP0.2Port 0 bit 2.
18I/OP0.3Port 0 bit 3.
17I/OP0.4Port 0 bit 4.
16I/OP0.5Port 0 bit 5.
14I/OP0.6Port 0 bit 6.
13I/OP0.7Port 0 bit 7.
I/OPort 0: Port 0 is an 8-bit I/O port with a user-configurable output type. During reset
Port 0 latches are configured in the input only mode with the internal pullup
disabled. The operation of port 0 pins as inputs and outputs depends upon the port
configuration selected. Each port pin is configured independently. Refer to the
section on I/O port configuration and the DC Electrical Characteristics for details.
The Keypad Interrupt feature operates with port 0 pins.
All pins have Schmitt triggered inputs.
Port 0 also provides various special functions as described below.
OCMP2Comparator 2 output.
IKBI0Keyboard Input 0.
ICIN2B Comparator 2 positive input B.
IKBI1Keyboard Input 1.
ICIN2A Comparator 2 positive input A.
IKBI2Keyboard Input 2.
ICIN1B Comparator 1 positive input B.
IKBI3Keyboard Input 3.
ICIN1A Comparator 1 positive input A.
IKBI4Keyboard Input 4.
I/OT1Timer/counter 1 external count input or overflow output.
IKBI7Keyboard Input 7.
P89LPC920/921/922
2003 Nov 6 11
Philips Semiconductors
User’s Manual - Preliminary -
General Description
MnemonicPin no. Type Name and function
P1.0 - P1.712, 11,
10, 9, 8,
4, 3, 2
12I/OP1.0Port 1 bit 0.
11I/OP1.1Port 1 bit 1.
10I/OP1.2Port 1 bit 2. (Open-drain when used as outputs)
9I/OP1.3Port 1 bit 3. (Open-drain when used as outputs)
8I/OP1.4Port 1 bit 4.
4IP1.5Port 1 bit 5. (Input only)
3I/OP1.6Port 1 bit 6.
2I/OP1.7Port 1 bit 7.
P3.0 - P3.17, 6I/OPort 3: Port 3 is an 2-bit I/O port with a user-configurable output type. During reset
7I/OP3.0Port 3 bit 0.
I/O
Port 1: Port 1 is an 8-bit I/O port with a user-configurable output type, except for
(for
three pins as noted below. During reset Port 1 latches are configured in the input
P1.0-
only mode with the internal pullup disabled. The operation of the configurable port 1
P1.4,
pins as inputs and outputs depends upon the port configuration selected. Each of
P1.6-
the configurable port pins are programmed independently. Refer to the section on
P1.7),
I/O port configuration and the DC Electrical Characteristics for details. P1.2 - P1.3
I (for
are open drain when used as outputs. P1.5 is input only.
P1.5)
All pins have Schmitt triggered inputs.
Port 1 also provides various special functions as described below.
OTxDTransmitter output for the serial port.
IRxDReceiver input for the serial port.
I/OT0Timer/counter 0 external count input or overflow output. (Open-drain
when used as outputs)
2
I/OSCLI
IINT0
I/OSDAI2C serial data input/output.
IINT1
IRST
Port 3 latches are configured in the input only mode with the internal pullup
disabled. The operation of port 3 pins as inputs and outputs depends upon the port
configuration selected. Each port pin is configured independently. Refer to the
section on I/O port configuration and the DC Electrical Characteristics for details.
All pins have Schmitt triggered inputs.
Port 3 also provides various special functions as described below:
OXTAL2 Output from the oscillator amplifier (when a crystal oscillator option is
OCLKOUTCPU clock divided by 2 when enabled via SFR bit (ENCLK - TRIM.6).
C serial clock input/output.
External interrupt 0 input.
External interrupt 1 input.
External Reset input (if selected via FLASH configuration). A low on
this pin resets the microcontroller, causing I/O ports and peripherals to
take on their default states, and the processor begins execution at
address 0.
selected via the FLASH configuration).
It can be used if the CPU clock is the internal RC oscillator, watchdog
oscillator or external clock input, except when XTAL1/XTAL2 are used
to generate clock source for the real time clock/system timer.
P89LPC920/921/922
2003 Nov 6 12
Philips Semiconductors
User’s Manual - Preliminary -
General Description
MnemonicPin no. Type Name and function
6I/OP3.1Port 3 bit 1.
IXTAL1 Input to the oscillator circuit and internal clock generator circuits (when
selected via the FLASH configuration). It can be a port pin if internal
RC oscillator or watchdog oscillator is used as the CPU clock source,
AND if XTAL1/XTAL2 are not used to generate the clock for the real
time clock/system timer.
V
SS
V
DD
5IGround: 0V reference.
15IPower Supply: This is the power supply voltage for normal operation as well as
Idle and Power Down modes.
P89LPC920/921/922
2003 Nov 6 13
Philips Semiconductors
User’s Manual - Preliminary -
General Description
P89LPC920/921/922
Special Function Registers
Note: Special Function Registers (SFRs) accesses are restricted in the following ways:
1. User must NOT attempt to access any SFR locations not defined.
2. Accesses to any defined SFR locations must be strictly for the functions for the SFRs.
3. SFR bits labeled ’-’, ’0’ or ’1’ can ONLY be written and read as follows:
- ’-’ Unless otherwise specified, MUST be written with ’0’, but can return any value when read (even if it was written with ’0’).
It is a reserved bit and may be used in future derivatives.
- ’0’ MUST be written with ’0’, and will return a ’0’ when read.
- ’1’ MUST be written with ’1’, and will return a ’1’ when read
NameDescription
ACC*AccumulatorE0H00H 00000000
AUXR1#Auxiliary Function RegisterA2HCLKLPEBRRENT1ENT0SRST0-DPS00H
WDCON# Watchdog Control RegisterA7HPRE2PRE1PRE0--WDRUN WDTOF WDCLKNotes 3,5
WDL#Watchdog LoadC1HFFH 11111111
WFEED1# Watchdog Feed 1C2H
WFEED2# Watchdog Feed 2C3H
2003 Nov 6 16
Philips Semiconductors
User’s Manual - Preliminary -
General Description
P89LPC920/921/922
Notes:
* SFRs are bit addressable.
# SFRs are modified from or added to the 80C51 SFRs.
- Reserved bits, must be written with 0’s.
§ BRGR1 and BRGR0 must only be written if BRGEN in BRGCON SFR is ’0’. If any of them is written if BRGEN = 1, result is unpredictable.
Unimplemented bits in SFRs (labeled ’-’ ) are X (unknown) at all times. Unless otherwise specified, ones should not be written to these bits since
they may be used for other purposes in future derivatives. The reset values shown for these bits are ’0’s although they are unknown when
read.
1. All ports are in input only (high impendance) state after power-up.
2. The RSTSRC register reflects the cause of the LPC920/921/922 reset. Upon a power-up reset, all reset source flags are cleared except POF
and BOF - the power-on reset value is xx110000.
3. After reset, the value is 111001x1, i.e., PRE2-PRE0 are all 1, WDRUN=1 and WDCLK=1. WDTOF bit is 1 after watchdog reset and is 0 after
power-on reset. Other resets will not affect WDTOF.
4. On power-on reset, the TRIM SFR is initialized with a factory preprogrammed value. Other resets will not cause initialization of the TRIM
register.
5. The only reset source that affects these SFRs is power-on reset.
2003 Nov 6 17
Philips Semiconductors
User’s Manual - Preliminary -
General Description
Memory organization
The P89LPC920/921/922 memory map is shown in Figure 1-1.
FF00h
FFEFh
1FFFh
1E00h
1C00h
1BFFh
1800h
17FFh
1400h
13FFh
1000h
0FFFh
0C00h
0BFFh
0800h
07FFh
0400h
03FFh
0000h
IAP entry-points
ISP CODE (512B)*
Sector 7
Sector 6
Sector 5
Sector 4
Sector 3
Sector 2
Sector 1
Sector 0
Read- protected
IAP calls only
IAP routines
entry points for:
- 51 ASM. code
C code
-
ISP
serial loader
-
UART (auto-baud)
-I2C, SPI etc. *
Flexible choices:
- as supplied (UART)
- Phil ips libraries
-user-defined
Note: ISP code is located at the end of Sector
1 on the LPC920, at the end of Sector 4 on the
LPC92 1, AND AT THE END OF Sector 7 on
the LPC922.
P89LPC920/921/922
FFEFh
FF1Fh
FF00h
1FFFh
1E00h
Special Function
Regi sters
(directly addressable)
entry points
I DATA (incl. DATA)
128 Bytes On-Chip
Data Memory (stack
and indirect addr. )
DATA
128 Bytes On-Chip
Data Memory (stack,
direct and indir. addr.)
4 Reg. Banks R0-R7
Data Memory
(DATA, IDATA)
*
Figure 1-1: P89LPC920/921/922 memory map
The various P89LPC920/921/922 memory spaces are as follows:
DATA128 bytes of internal data memory space (00h..7Fh) accessed via direct or indirect addressing, using instructions
other than MOVX and MOVC. All or part of the Stack may be in this area.
IDATAIndirect Data. 256 bytes of internal data memory space (00h:FFh) accessed via indirect addressing using
instructions other than MOVX and MOVC. All or part of the Stack may be in this area. This area includes the DATA
area and the 128 bytes immediately above it.
SFRSpecial Function Registers. Selected CPU registers and peripheral control and status registers, accessible only via
direct addressing.
CODE64 KB of Code memory space, accessed as part of program execution and via the MOVC instruction. The
P89LPC920/921/922 has 2 KB/ 4 KB/ 8 KB of on-chip Code memory.).
2003 Nov 6 18
Philips Semiconductors
User’s Manual - Preliminary -
General Description
P89LPC920/921/922
DATA RAM ARRANGEMENT
The 256 bytes of on-chip RAM is organized as follows:
TypeData RAMSize (Bytes)
DATAMemory that can be addressed directly and indirectly128
IDATAMemory that can be addressed indirectly (includes DATA)256
Table 1-1: On-chip data memory usage.
2003 Nov 6 19
Philips Semiconductors
User’s Manual - Preliminary -
General Description
P89LPC920/921/922
2003 Nov 6 20
Philips Semiconductors
User’s Manual - Preliminary -
CLOCKS
P89LPC920/921/922
2. CLOCKS
Enhanced CPU
The P89LPC920/921/922 uses an enhanced 80C51 CPU which runs at 6 times the speed of standard 80C51 devices. A machine
cycle consists of two CPU clock cycles, and most instructions execute in one or two machine cycles.
Clocks
Clock definitions
The P89LPC920/921/922 device has several internal clocks as defined below:
• OSCCLK - Input to the DIVM clock divider. OSCCLK is selected from one of four clock sources (see Figure 2-3) and can also
be optionally divided to a slower frequency (see section "CPU Clock (CCLK) modification: DIVM register").
Note: f
• CCLK - CPU clock; output of the DIVM clock divider. There are two CCLK cycles per machine cycle, and most instructions are
executed in one to two machine cycles (two or four CCLK cycles).
• RCCLK - The internal 7.373 MHz RC oscillator output.
• PCLK - Clock for the various peripheral devices and is CCLK/2.
is defined as the OSCCLK frequency.
OSC
Oscillator clock (OSCCLK)
The P89LPC920/921/922 provides several user-selectable oscillator options. This allows optimization for a range of needs from
high precision to lowest possible cost. These options are configured when the FLASH is programmed and include an on-chip
watchdog oscillator, an on-chip RC oscillator, an oscillator using an external crystal, or an external clock source. The crystal
oscillator can be optimized for low, medium, or high frequency crystals covering a range from 20 kHz to 12 MHz.
Low speed oscillator option
This option supports an external crystal in the range of 20 kHz to 100 kHz. Ceramic resonators are also supported in this
configuration.
Medium speed oscillator option
This option supports an external crystal in the range of 100 kHz to 4 MHz. Ceramic resonators are also supported in this
configuration.
High speed oscillator option
This option supports an external crystal in the range of 4 MHz to 12 MHz. Ceramic resonators are also supported in this
configuration.
Clock output
The P89LPC920/921/922 supports a user-selectable clock output function on the XTAL2 / CLKOUT pin when the crystal
oscillator is not being used. This condition occurs if a different clock source has been selected (on-chip RC oscillator,watchdog
oscillator, external clock input on X1) and if the Real-time Clock is not using the crystal oscillator as its clock source. This allows
external devices to synchronize to the P89LPC920/921/922. This output is enabled by the ENCLK bit in the TRIM register
The frequency of this clock output is 1/2 that of the CCLK. If the clock output is not needed in Idle mode, it may be turned off prior
to entering Idle, saving additional power. Note: on reset, the TRIM SFR is initialized with a factory preprogrammed value.
2003 Nov 6 21
Philips Semiconductors
User’s Manual - Preliminary -
CLOCKS
P89LPC920/921/922
Therefore when setting or clearing the ENCLK bit, the user should retain the contents of bits 5:0 of the TRIM register. This can
be done by reading the contents of the TRIM register (into the ACC for example), modifying bit 6, and writing this result back into
the TRIM register. Alternatively,the “ANL direct” or “ORL direct” instructions can be used to clear or set bit 6 of the TRIM register.
Quartz crystal or
ceramic resonator
The oscillator must be configured in
one of the following modes:
- Low Frequency Crystal
- Medium Frequency Crystal
- High Frequency Crystal
*
* A series resistor may be required to limit
crystal drive levels. This is especially
important for low frequency crystals (see
text).
Figure 2-1: Using the crystal oscillator
On-chip RC oscillator option
P89LPC920/921/922
XTAL1
XTAL2
The P89LPC920/921/922 has a 6-bit TRIM register that can be used to tune the frequency of the RC oscillator. During reset, the
TRIM value is initialized to a factory pre-programmed value to adjust the oscillator frequency to 7.373 MHz, ±1%. (Note: the initial
value is better than 1%; please refer to the datasheet for behavior over temperature). End user applications can write to the TRIM
register to adjust the on-chip RC oscillator to other frequencies. Increasing the TRIM value will decrease the oscillator frequency.
TRIM
Address: 96h
Not bit addressable
Reset Source(s): Power-up only
Reset Value: On power-up reset, ENCLK = 0, and TRIM.5-0 are loaded with the factory programmed value.
BITSYMBOLFUNCTION
TRIM.7-Reserved.
TRIM.6ENCLKWhen ENCLK =1, CCLK/ 2 is output on the XTAL2 pin (P3.0) provided that the crystal
TRIM.5-0Trim value.
Note: on reset, the TRIM SFR is initialized with a factory preprogrammed value. When setting or clearing the ENCLK bit,
the user should retain the contents of bits 5:0 of the TRIM register. This can be done by reading the contents of the TRIM
register (into the ACC for example), modifying bit 6, and writing this result back into the TRIM register. Alternatively, the
“ANL direct” or “ORL direct” instructions can be used to clear or set bit 6 of the TRIM register.
76543210
-ENCLKTRIM.5TRIM.4TRIM.3TRIM.2TRIM.1TRIM.0
oscillator is not being used. When ENCLK=0, no clock output is enabled.
Figure 2-2: On-chip RC oscillator TRIM register
Watchdog oscillator option
The watchdog has a separate oscillator which has a frequency of 400 kHz. This oscillator can be used to save power when a
high clock frequency is not needed.
2003 Nov 6 22
Philips Semiconductors
User’s Manual - Preliminary -
CLOCKS
P89LPC920/921/922
External clock input option
In this configuration, the processor clock is derived from an external source driving the XTAL1 / P3.1 pin. The rate may be from
0 Hz up to 12 MHz. The XTAL2 / P3.0 pin may be used as a standard port pin or a clock output.
RTCS1:0
XTAL1
XTAL2
RC Oscillator
(7.3728MHz)
Watchdog
Oscillator
High freq.
Med freq.
Low freq.
FOSC2:0
OSC
CLK
DIVM
CPU
Clock
CCLK
RTC
CPU
/2
WDT
PCLK
(400KHz)
Baud Rate
Generator
UART
Figure 2-3: Block diagram of oscillator control
Timer 0 & 1
Peripheral Clock
I2C
Oscillator Clock (OSCCLK) wakeup delay
The P89LPC920/921/922 has an internal wakeup timer that delays the clock until it stabilizes depending to the clock source used.
If the clock source is any of the three crystal selections, the delay is 992 OSCCLK cycles plus 60-100µs. If the clock source is
either the internal RC oscillator or the Watchdog oscillator, the delay is 224 OSCCLK cycles plus 60-100µs.
CPU Clock (CCLK) modification: DIVM register
The OSCCLK frequency can be divided down, by an integer, up to 510 times by configuring a dividing register, DIVM, to provide
CCLK. This produces the CCLK frequency using the following formula:
CCLK frequency = f
Where: f
Since N ranges from 0 to 255, the CCLK frequency can be in the range of f
This feature makes it possible to temporarily run the CPU at a lower rate, reducing power consumption. By dividing the clock, the
CPU can retain the ability to respond to events other than those that can cause interrupts (i.e. events that allow exiting the Idle
mode) by executing its normal program at a lower rate. This can often result in lower power consumption than in Idle mode. This
is the frequency of OSCCLK
OSC
N is the value of DIVM.
OSC
/ (2N)
OSC
to f
/510. (for N =0, CCLK = f
OSC
OSC
) .
2003 Nov 6 23
Philips Semiconductors
User’s Manual - Preliminary -
CLOCKS
P89LPC920/921/922
can allow bypassing the oscillator start-up time in cases where Power down mode would otherwise be used. The value of DIVM
may be changed by the program at any time without interrupting code execution.
Low power select
The P89LPC920/921/922 is designed to run at 12 MHz (CCLK) maximum. However, if CCLK is 8 MHz or slower, the CLKLP
SFR bit (AUXR1.7) can be set to a ‘1’ to lower the power consumption further. On any reset, CLKLP is ‘0’ allowing highest
performance. This bit can then be set in software if CCLK is running at 8 MHz or slower.
2003 Nov 6 24
Philips Semiconductors
User’s Manual - Preliminary -
INTERRUPTS
P89LPC920/921/922
3. INTERRUPTS
The P89LPC920/921/922 uses a four priority level interrupt structure. This allows great flexibility in controlling the handling of the
P89LPC920/921/922’s 12 interrupt sources.
Each interrupt source can be individually enabled or disabled by setting or clearing a bit in the interrupt enable registers IEN0 or
IEN1. The IEN0 register also contains a global enable bit, EA, which enables all interrupts.
Each interrupt source can be individually programmed to one of four priority levels by setting or clearing bits in the interrupt priority
registers IP0, IP0H, IP1, and IP1H. An interrupt service routine in progress can be interrupted by a higher priority interrupt, but
not by another interrupt of the same or lower priority. The highest priority interrupt service cannot be interrupted by any other
interrupt source. If two requests of different priority levels are received simultaneously, the request of higher priority level is
serviced.
If requests of the same priority level are pending at the start of an instruction cycle, an internal polling sequence determines which
request is serviced. This is called the arbitration ranking. Note that the arbitration ranking is only used for pending requests of
the same priority level.
Table summarizes the interrupt sources, flag bits, vector addresses, enable bits, priority bits, arbitration ranking, and whether
each interrupt may wake up the CPU from a Power down mode.
Interrupt priority structure
There are four SFRs associated with the four interrupt levels: IP0, IP0H, IP1, IP1H. Every interrupt has two bits in IPx and IPxH
(x = 0,1) and can therefore be assigned to one of four levels, as shown in Table .
1. SSTAT.5 = 0 selects combined Serial Port (UART) Tx and Rx interrupt; SSTAT.5 = 1 selects Serial Port Rx interrupt only
(Tx interrupt will be different, see Note 3 below).
2. This interrupt is used as Serial Port (UART) Tx interrupt if and only if SSTAT.5 = 1, and is disabled otherwise.
3. If SSTAT.0 = 1, the following Serial Port additional flag bits can cause this interrupt: FE, BR, OE
2
1,3
1,3
Interrupt
flag bit(s)
TI & RI
RI
WDOVF/
RTCF
TI006BhEST (IEN1.6)P1H.6, IP1.610No
Vector
address
0023h
0053h
Interrupt
enable bit(s)
ES/ESR
(IEN0.4)
EWDRT
(IEN0.6)
Interrupt
priority
IP0H.4, IP0.411No
IP0H.6, IP0.63Yes
P89LPC920/921/922
Arbitration
ranking
Power down
wakeup
The P89LPC920/921/922 has two external interrupt inputs in addition to the Keypad Interrupt function. The two interrupt inputs
are identical to those present on the standard 80C51 microcontrollers.
These external interrupts can be programmed to be level-triggered or edge-triggered by clearing or setting bit IT1 or IT0 in
Register TCON. If ITn = 0, external interrupt n is triggered by a low level detected at the INTn
is edge triggered. In this mode if consecutive samples of the INTn pin show a high level in one cycle and a low level in the next
cycle, interrupt request flag IEn in TCON is set, causing an interrupt request.
Since the external interrupt pins are sampled once each machine cycle, an input high or low level should be held for at least one
machine cycle to ensure proper sampling. If the external interrupt is edge-triggered, the external source has to hold the request
pin high for at least one machine cycle, and then hold it low for at least one machine cycle. This is to ensure that the transition is
detected and that interrupt request flag IEn is set. IEn is automatically cleared by the CPU when the service routine is called.
If the external interrupt is level-triggered, the external source must hold the request active until the requested interrupt is
generated. If the external interrupt is still asserted when the interrupt service routine is completed, another interrupt will be
generated. It is not necessary to clear the interrupt flag IEn when the interrupt is level sensitive, it simply tracks the input pin level.
If an external interrupt is enabled when the P89LPC920/921/922 is put into Power down or Idle mode, the interrupt occurance
will cause the processor to wake up and resume operation. Refer to the section on Power Reduction Modes for details.
pin. If ITn = 1, external interrupt n
External Interrupt pin glitch suppression
Most of the P89LPC920/921/922 pins have glitch suppression circuits to reject short glitches (please refer to the P89LPC920/
921/922 datasheet, AC Electrical Characteristics for glitch filter specifications) . However, pins SDA/INT0
do not have the glitch suppression circuits. Therefore, INT1
has glitch suppression while INT0 does not.
/P1.3 and SCL/T0/P1.2
2003 Nov 6 26
Philips Semiconductors
User’s Manual - Preliminary -
INTERRUPTS
RTCF
ERTC
(RTCCON.1)
WDOVF
IE0
EX0
IE1
EX1
BOPD
EBO
KBIF
EKBI
EWDRT
CMF2
CMF1
EC
EA (IE0.7)
TF0
ET0
TF1
ET1
TI & RI/RI
ES/ESR
EST
EI2C
P89LPC920/921/922
Wakeup (if in
Power down)
TI
SI
Interrupt to CPU
Figure 3-1: Interrupt sources, interrupt enables, and power down wake-up sources
2003 Nov 6 27
Philips Semiconductors
User’s Manual - Preliminary -
INTERRUPTS
P89LPC920/921/922
2003 Nov 6 28
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
I/O PORTS
4. I/O PORTS
The P89LPC920/921/922 has 3 I/O ports: Port 0, Port 1, and Port 3. Ports 0 and 1 are 8-bit ports and Port 3 is a 2-bit port. The
exact number of I/O pins available depends upon the clock and reset options chosen (see Table 4-1)
Table 4-1: .Number of I/O pins available
Clock sourceReset optionNumber of I/O pins
On-chip oscillator or watchdog
oscillator
External clock input
No external reset (except during power-up)18
External RST
No external reset (except during power-up)17
External RST
pin supported17
pin supported16
Low/medium/high speed oscillator
(external crystal or resonator)
No external reset (except during power-up)16
External RST
pin supported15
Port configurations
All but three I/O port pins on the P89LPC920/921/922 may be configured by software to one of four types on a pin-by-pin basis,
as shown in Table 4-3. These are: quasi-bidirectional (standard 80C51 port outputs), push-pull, open drain, and input-only. Two
configuration registers for each port select the output type for each port pin.
P1.5 (RST) can only be an input and cannot be configured.
P1.2 (SCL/T0) and P1.3 (SDA/INT0) may only be configured to be either input-only or open drain.
Table 4-2: Port output configuration settings
PxM1.yPxM2.yPort output mode
00Quasi-bidirectional
01Push-Pull
10Input Only (High Impedance)
11Open Drain
Quasi-bidirectional output configuration
Quasi-bidirectional outputs can be used both as an input and output without the need to reconfigure the port. This is possible
because when the port outputs a logic high, it is weakly driven, allowing an external device to pull the pin low. When the pin is
driven low, it is driven strongly and able to sink a large current. There are three pull-up transistors in the quasi-bidirectional output
that serve different purposes.
One of these pull-ups, called the “very weak” pull-up, is turned on whenever the port latch for the pin contains a logic 1. This very
weak pull-up sources a very small current that will pull the pin high if it is left floating.
A second pull-up, called the “weak” pull-up, is turned on when the port latch for the pin contains a logic 1 and the pin itself is also
at a logic 1 level. This pull-up provides the primary source current for a quasi-bidirectional pin that is outputting a 1. If this pin is
pulled low by an external device, the weak pull-up turns off, and only the very weak pull-up remains on. In order to pull the pin
low under these conditions, the external device has to sink enough current to overpower the weak pull-up and pull the port pin
below its input threshold voltage.
2003 Nov 6 29
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
I/O PORTS
The third pull-up is referred to as the “strong” pull-up. This pull-up is used to speed up low-to-high transitions on a quasibidirectional port pin when the port latch changes from a logic 0 to a logic 1. When this occurs, the strong pull-up turns on for two
CPU clocks quickly pulling the port pin high .
The quasi-bidirectional port configuration is shown in Figure 4-1.
Although the P89LPC920/921/922 is a 3 V device most of the pins are 5 V-tolerant. If 5 V is applied to a pin configured in quasibidirectional mode, there will be a current flowing from the pin to VDD causing extra power consumption. Therefore, applying 5 V
to pins configured in quasi-bidirectional mode is discouraged.
A quasi-bidirectional port pin has a Schmitt-triggered input that also has a glitch suppression circuit.
(Please refer to the P89LPC920/921/922 datasheet, AC Electrical Characteristics for glitch filter specifications)
port latch data
2 CPU
clock delay
V
DD
strong
V
DD
very
weak
V
DD
weak
port
pin
input data
Figure 4-1: Quasi-bidirectional output
glitch rejection
Open drain output configuration
The open drain output configuration turns off all pull-ups and only drives the pulldown transistor of the port pin when the port latch
contains a logic 0. To be used as a logic output, a port configured in this manner must have an external pull-up, typically a resistor
tied to V
The open drain port configuration is shown in Figure 4-2.
An open drain port pin has a Schmitt-triggered input that also has a glitch suppression circuit.
Please refer to the P89LPC920/921/922 datasheet, AC Electrical Characteristics for glitch filter specifications).
. The pulldown for this mode is the same as for the quasi-bidirectional mode.
DD
port latch data
input data
glitch rejection
port
pin
Figure 4-2: Open drain output
2003 Nov 6 30
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
I/O PORTS
Input-only configuration
The input port configuration is shown in Figure 4-3. It is a Schmitt-triggered input that also has a glitch suppression circuit.
(Please refer to the P89LPC920/921/922 datasheet, AC Electrical Characteristics for glitch filter specifications)
input data
glitch rejection
Figure 4-3: Input -only
port
pin
Push-pull output configuration
The push-pull output configuration has the same pulldown structure as both the open drain and the quasi-bidirectional output
modes, but provides a continuous strong pull-up when the port latch contains a logic 1. The push-pull mode may be used when
more source current is needed from a port output.
The push-pull port configuration is shown in Figure 4-4.
A push-pull port pin has a Schmitt-triggered input that also has a glitch suppression circuit.
(Please refer to the P89LPC920/921/922 datasheet, AC Electrical Characteristics for glitch filter specifications)
V
DD
strong
port latch data
port
pin
input data
Figure 4-4: Push-pull output
glitch rejection
Port 0 analog functions
The P89LPC920/921/922 incorporates two Analog Comparators. In order to give the best analog performance and minimize
power consumption, pins that are being used for analog functions must have both the digital outputs and digital inputs disabled.
Digital outputs are disabled by putting the port pins into the input-only mode as described in the Port Configurations section (see
Table 4-3).
Digital inputs on Port 0 may be disabled through the use of the PT0AD register. Bits 1 through 5 in this register correspond to
pins P0.1 through P0.5 of Port 0, respectively. Setting the corresponding bit in PT0AD disables that pin’s digital input. Port bits
that have their digital inputs disabled will be read as 0 by any instruction that accesses the port.
On any reset, PT0AD bits 1 through 5 default to ‘0’s to enable the digital functions.
2003 Nov 6 31
Philips Semiconductors
I/O PORTS
Table 4-3: Port output configuration
Port pin
P0.0P0M1.0P0M2.0KBI0,CMP2
P0.1P0M1.1P0M2.1KBI1,CIN2B
P0.2P0M1.2P0M2.2KBI2,CIN2A
P0.3P0M1.3P0M2.3KBI3,CIN1B
P0.4P0M1.4P0M2.4KBI4,CIN1A
P0.5P0M1.5P0M2.5KBI5,CMPREF
P0.6P0M1.6P0M2.6KBI6,CMP1
P0.7P0M1.7P0M2.7KBI7,T1
P1.0P1M1.0P1M2.0TxD
P1.1P1M1.1P1M2.1RxD
P1.2P1M1.2P1M2.2T0,SCLinput-only or open-drain
P1.3P1M1.3P1M2.3INT0
P1.4P1M1.4P1M2.4INT1
P1.5not configurableRST
P1.6P1M1.6P1M2.6
P1.7P1M1.7P1M2.7
P3.0P3M1.0P3M2.0XTAL2,CLKOUT
P3.1P3M1.1P3M2.1XTAL1
Configuration SFR bits
PxM1.yPxM2.y
Alternate usageNotes
Refer to section "Port 0 analog functions" for usage as
analog inputs (CIN2B, CIN2A, CIN1B, CIN1A and
CMPREF)
,SDAinput-only or open-drain
Input only. Usage as general purpose input or RST is
determined by User Configuration Bit RPD (UCFG1.6).
Always a reset input during a power-on sequence.
User’s Manual - Preliminary -
P89LPC920/921/922
Additional port features
After power-up, all pins are in Input-Only mode. Please note that this is different from the LPC76x series of devices.
• After power-up, all I/O pins except P1.5, may be configured by software.
• Pin P1.5 is input only. Pins P1.2 and P1.3 are configurable for either input-only or open drain.
Every output on the P89LPC920/921/922 has been designed to sink typical LED drive current. However, there is a maximum
total output current for all ports which must not be exceeded. Please refer to the P89LPC920/921/922 Datasheet for detailed
specifications.
All ports pins that can function as an output have slew rate controlled outputs to limit noise generated by quickly switching output
signals. The slew rate is factory-set to approximately 10 ns rise and fall times.
2003 Nov 6 32
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
POWER MONITORING FUNCTIONS
5. POWER MONITORING FUNCTIONS
The P89LPC920/921/922 incorporates power monitoring functions designed to prevent incorrect operation during initial poweron and power loss or reduction during operation. This is accomplished with two hardware functions: Power-on Detect and
Brownout Detect.
Brownout Detection
The Brownout Detect function determines if the power supply voltage drops below a certain level. The default operation for a
Brownout Detection is to cause a processor reset. However, it may alternatively be configured to generate an interrupt by setting
the BOI (PCON.4) bit and the EBO (IEN0.5) bit.
Enabling and disabling of Brownout Detection is done via the BOPD (PCON.5) bit, bit field PMOD1-0 (PCON.1-0) and user
configuration bit BOE (UCFG1.5). If BOE is in an unprogrammed state, brownout is disabled regardless of PMOD1-0 and BOPD.
If BOE is in a programmed state, PMOD1-0 and BOPD will be used to determine whether Brownout Detect will be disabled or
enabled. PMOD1-0 is used to select the power reduction mode. If PMOD1-0 = ‘11’, the circuitry for the Brownout Detection is
disabled for lowest power consumption. BOPD defaults to ‘0’, indicating brownout detection is enabled on power-on if BOE is
programmed.
If Brownout Detection is enabled, the operating voltage range for V
VDD falls below the Brownout trip voltage, V
If Brownout Detection is disabled, the operating voltage range for VDD is 2.4 V-3.6 V. If the P89LPC920/921/922 device is to
operate with a power supply that can be below 2.7 V, BOE should be left in the unprogrammed state so that the device can
operate at 2.4 V, otherwise continuous brownout reset may prevent the device from operating.
If Brownout Detect is enabled (BOE programmed, PMOD1-0 ≠ ‘11’, BOPD = 0), BOF (RSTSRC.5) will be set when a brownout
is detected, regardless of whether a reset or an interrupt is enabled, . BOF will stay set until it is cleared in software by writing ‘0’
to the bit. Note that if BOE is unprogrammed, BOF is meaningless. If BOE is programmed, and a initial power-on occurs, BOF
will be set in addition to the power-on flag (POF - RSTSRC.4).
For correct activation of Brownout Detect, certain V
specifications.
(see D.C. Electrical Characteristics), and is negated when VDD rises above VBO.
BO
rise and fall times must be observed. Please see the datasheet for
DD
is 2.7 V-3.6 V, and the brownout condition occurs when
DD
2003 Nov 6 33
Philips Semiconductors
POWER MONITORING FUNCTIONS
Table 5-1: Brownout options
BOE
(UCFG1.5)
0 (erased)XX XXXX
1 (programmed)
PMOD1-0
(PCON.1-0)
11
(total power
down)
≠ 11
(any mode
other than
total power
down)
BOPD
(PCON.5)
XXXX
1
(brownout
detect
powered
down)
0
(brownout
detect
active)
BOI
(PCON.4)
XXX
0
(brownout
detect
generates
reset)
1
(brownout
detect
generates
an
interrupt)
EBO
(IEN0.5)
XX
1
(enable
brownout
interrupt)
0XBoth brownout reset and interrupt disabled. V
X0
EA
(IEN0.7)
1
(global
interrupt
enable)
User’s Manual - Preliminary -
P89LPC920/921/922
Description
Brownout disabled. V
Brownout disabled. V
However, BOPD is default to ‘0’ upon power-up.
Brownout reset enabled. V
3.6 V. Upon a brownout reset, BOF (RSTSRC.5) will be
set to indicate the reset source. BOF can be cleared by
writing ‘0’ to the bit.
Brownout interrupt enabled. V
3.6 V. Upon a brownout interrupt, BOF (RSTSRC.5) will
be set. BOF can be cleared by writing ‘0’ to the bit.
operating range is 2.4 V-3.6 V. However, BOF
(RSTSRC.5) will be set when V
Detection trip point. BOF can be cleared by writing ‘0’ to
the bit.
operating range is 2.4 V-3.6 V.
DD
operating range is 2.4 V-3.6 V.
DD
operating range is 2.7 V-
DD
operating range is 2.7 V-
DD
DD
falls to the Brownout
DD
Power-on Detection
The Power-On Detect has a function similar to the Brownout Detect, but is designed to work as power initially comes up, before
the power supply voltage reaches a level where the Brownout Detect can function. The POF flag (RSTSRC.4) is set to indicate
an initial power-on condition. The POF flag will remain set until cleared by software by writing ‘0’ to the bit. Note that if BOE
(UCFG1.5) is programmed, BOF (RSTSRC.5) will be set when POF is set. If BOE is unprogrammed, BOF is meaningless.
Power reduction modes
The P89LPC920/921/922 supports three different power reduction modes as determined by SFR bits PCON.1-0 (see Table 5-2):
2003 Nov 6 34
Philips Semiconductors
POWER MONITORING FUNCTIONS
User’s Manual - Preliminary -
P89LPC920/921/922
PMOD1
(PCON.1)
PMOD0
(PCON.0)
00Normal mode (default) - no power reduction.
01
Idle mode. The Idle mode leaves peripherals running in order to allow them to activate the processor
when an interrupt is generated. Any enabled interrupt source or reset may terminate Idle mode.
Power down mode:
The Power down mode stops the oscillator in order to minimize power consumption.
The P89LPC920/921/922 exits Power down mode via any reset, or certain interrupts - external pins
/INT1, brownout Interrupt, keyboard, Real-time Clock/System Timer), watchdog, and comparator
INT0
trips. Waking up by reset is only enabled if the corresponding reset is enabled, and waking up by interrupt
is only enabled if the corresponding interrupt is enabled and the EA SFR bit (IEN0.7) is set.
In Power down mode the internal RC oscillator is disabled unless both the RC oscillator has been
selected as the system clock AND the RTC is enabled.
In Power down mode, the power supply voltage may be reduced to the RAM keep-alive voltage V
This retains the RAM contents at the point where Power down mode was entered. SFR contents are not
guaranteed after V
has been lowered to V
DD
via Reset in this situation. VDD must be raised to within the operating range before the Power down mode
10
is exited.
When the processor wakes up from Power down mode, it will start the oscillator immediately and begin
execution when the oscillator is stable. Oscillator stability is determined by counting 1024 CPU clocks
after start-up when one of the crystal oscillator configurations is used, or 256 clocks after start-up for the
internal RC or external clock input configurations.
Some chip functions continue to operate and draw power during Power down mode, increasing the total
power used during Power down. These include:
• Brownout Detect
• Watchdog Timer if WDCLK (WDCON.0) is ‘1’.
• Comparators (Note: Comparators can be powered down separately with PCONA.5 set to ‘1’ and
comparators disabled);
• Real-time Clock/System Timer (and the crystal oscillator circuitry if this block is using it, unless
RTCPD, i.e., PCONA.7 is ‘1’).
Description
, therefore it is recommended to wake up the processor
RAM
RAM
.
Total power down mode: This is the same as Power down mode except that the Brownout Detection
circuitry and the voltage comparators are also disabled to conserve additional power. Note that a
brownout reset or interrupt will not occur. Voltage comparator interrupts and Brownout interrupt cannot
be used as a wakeup source.The internal RC oscillator is disabled unless both the RC oscillator has
been selected as the system clock AND the RTC is enabled.
The following are the wakeup options supported:
• Watchdog Timer if WDCLK (WDCON.0) is ‘1’. Could generate Interrupt or Reset, either one can wake
11
up the device
• External interrupts INTO/INT1
• Keyboard Interrupt
• Real-time Clock/System Timer (and the crystal oscillator circuitry if this block is using it, unless
RTCPD, i.e., PCONA.7 is ‘1’).
• Note: Using the internal RC-oscillator to clock the RTC during Power down may result in relatively high
power consumption. Lower power consumption can be achieved by using an external low frequency
clock when the Real-time Clock is running during Power down.
Table 5-2: Power reduction modes.
2003 Nov 6 35
Philips Semiconductors
POWER MONITORING FUNCTIONS
User’s Manual - Preliminary -
P89LPC920/921/922
PCON
Address: 87h
Not bit addressable
Reset Source(s): Any reset
Reset Value:00000000B
BITSYMBOLFUNCTION
PCON.7SMOD1Double Baud Rate bit for the serial port (UART) when Timer 1 is used as the baud rate
PCON.6SMOD0Framing Error Location:
PCON.5BOPDBrownout Detect Power down. When 1, Brownout Detect is powered down and therefore
PCON.4BOIBrownout Detect Interrupt Enable. When 1, Brownout Detection will generate a interrupt .
PCON.3GF1General Purpose Flag 1. May be read or written by user software, but has no effect on
PCON.2GF0General Purpose Flag 0. May be read or written by user software, but has no effect on
PCON.1-0PMOD1-PMOD0Power Reduction Mode (see section "Power reduction modes").
76543210
SMOD1 SMOD0BOPDBOIGF1GF0PMOD1 PMOD0
source. When 1, the Timer 1 overflow rate is supplied to the UART. When 0, the Timer 1
overflow rate is divided by two before being supplied to the UART. (See Figure 9-2)
-When 0, bit 7 of SCON is accessed as SM0 for the UART.
-When 1, bit 7 of SCON is accessed as the framing error status (FE) for the UART.
This bit also determines the location of the UART receiver interrupt RI (see description
on RI in Figure 9-3).
disabled. When 0, Brownout Detect is enabled. (Note: BOPD must be ‘0’ before any
programming or erasing commands can be issued. Otherwise these commands will be
aborted.)
When 0, Brownout Detection will cause a reset.
operation.
operation.
Figure 5-1: Power Control register (PCON)
2003 Nov 6 36
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
POWER MONITORING FUNCTIONS
PCONA
Address: B5H
Not bit addressable
Reset Source(s): Any reset
Reset Value: 00000000B
BITSYMBOLFUNCTION
PCONA.7RTCPDReal-time Clock Power down: When ‘1’, the internal clock to the Real-time Clock is
PCONA.6-Not used. Reserved for future use.
PCONA.5VCPDAnalog Voltage Comparators Power down: When ‘1’, the voltage comparators are
PCONA.4-Not used. Reserved for future use.
PCONA.3I2PDI
PCONA.2-Not used. Reserved for future use.
PCONA.1SPDSerial Port (UART) Power down: When ‘1’, the internal clock to the UART is disabled. Note
PCONA.0-Not used. Reserved for future use.
76543210
RTCPD-VCPD-I2PD-SPD-
disabled.
powered down. User must disable the voltage comparators prior to setting this bit.
2
C Power down: When ‘1’, the internal clock to the I2C is disabled. Note that in either
Power down mode or Total Power down mode, the I
of this bit.
that in either Power down mode or Total Power down mode, the UART clock will be
disabled regardless of this bit.
NOTE: Brownout Detect Power down is located in PCON.5.
2
C clock will be disabled regardless
Figure 5-2: Power Control register A (PCONA)
2003 Nov 6 37
Philips Semiconductors
POWER MONITORING FUNCTIONS
User’s Manual - Preliminary -
P89LPC920/921/922
2003 Nov 6 38
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
RESET
6. RESET
The P1.5/RST pin can function as either an active low reset input or as a digital input, P1.5. The RPE (Reset Pin Enable) bit in
UCFG1, when set to 1, enables the external reset input function on P1.5. When cleared, P1.5 may be used as an input pin.
NOTE: During a power-on sequence, The RPE selection is overriden and this pin will always functions as a reset input. An
external circuit connected to this pin should not hold this pin low during a Power-on sequence as this will keep the device in reset.
After power-on this input will function either as an external reset input or as a digital input as defined by the RPE bit. Only a poweron reset will temporarily override the selection defined by RPE bit. Other sources of reset will not override the RPE bit.
NOTE: During a power cycle, V
must fall below V
DD
reapplied, in order to ensure a power-on reset.
Reset can be triggered from the following sources (see Figure 6-1):
• External reset pin (during power-on or if user configured via UCFG1);
• Power-on Detect;
• Brownout Detect;
• Watchdog Timer;
• Software reset;
• UART break detect reset.
For every reset source, there is a flag in the Reset Register, RSTSRC. The user can read this register to determine the most
recent reset source. These flag bits can be cleared in software by writing a ‘0’ to the corresponding bit. More than one flag bit
may be set:
• During a power-on reset, both POF and BOF are set but the other flag bits are cleared.
• For any other reset, any previously set flag bits that have not been cleared will remain set.
(see "DC electrical characteristics" in the datasheet) before pwoer is
POR
RPE (UCFG1.6)
RST
Pin
WDTE (UCFG1.7)
Watchdog Timer Reset
Software Reset SRST (AUXR1.3)
Power-on Detect
UART Break Detect
EBRR (AUXR1.6)
Brownout Detect Reset
BOPD (PCON.5)
Chip Reset
Figure 6-1: Block diagram of Reset
2003 Nov 6 39
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
RESET
RSTSRC
Address: DFH
Not bit addressable
Reset Sources: Power-on only
Reset Value: xx110000B (This is the power-on reset value. Other reset sources will set corresponding bits.)
BITSYMBOLFUNCTION
RSTSRC.7-6-Reserved for future use. Should not be set to 1 by user programs.
RSTSRC.5BOFBrownout Detect Flag. When Brownout Detect is activated, this bit is set. It will remain set
RSTSRC.4POFPower-on Detect Flag. When Power-on Detect is activated, the POF flag is set to indic ate
RSTSRC.3R_BKBreak detect reset. If a break detect occurs and EBRR (AUXR1.6) is set to ‘1’, a system
RSTSRC.2R_WDWatchdog Timer reset flag. Cleared by software by writing a ‘0’ to the bit or a Power-on
RSTSRC.1R_SFSoftware reset Flag. Cleared by software by writing a ‘0’ to the bit or a Power-on reset.
RSTSRC.0R_EXExternal reset Flag. When this bit is ‘1’, it indicates external pin reset. Cleared by software
76543210
--BOFPOFR_BKR_WDR_SFR_EX
until cleared by software by writing a ‘0’ to the bit. (Note: On a Power-on reset, both POF
and this bit will be set while the other flag bits are cleared.)
an initial power-up condition. The POF flag will remain set until cleared by software by
writing a ‘0’ to the bit.. (Note: On a Power-on reset, both BOF and this bit will be set while
the other flag bits are cleared.)
reset will occur. This bit is set to indicate that the system reset is caused by a break detect.
Cleared by software by writing a ‘0’ to the bit or on a Power-on reset.
reset.(NOTE: UCFG1.7 must be = 1).
by writing a ‘0’ to the bit or a Power-on reset. If RST
reset is over, R_EX will be set.
is still asserted after the Power-on
Figure 6-2: Reset Sources register
Reset vector
Following reset, the P89LPC920//921/922 will fetch instructions from either address 0000h or the Boot address. The Boot
address is formed by using the Boot Vector as the high byte of the address and the low byte of the address =00h. The Boot
address will be used if a UART break reset occurs or the non-volatile Boot Status bit (BOOTSTAT.0) = 1, or the device has been
forced into ISP mode.Otherwise, instructions will be fetched from address 0000H.
2003 Nov 6 40
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
TIMERS 0 AND 1
7. TIMERS 0 AND 1
The P89LPC920/921/922 has two general-purpose counter/timers which are upward compatible with the 80C51 Timer 0 and
Timer 1. Both can be configured to operate either as timers or event counters (see Figure 7-1). An option to automatically toggle
the Tx pin upon timer overflow has been added.
In the “Timer” function, the timer is incremented every PCLK.
In the “Counter” function, the register is incremented in response to a 1-to-0 transition on its corresponding external input pin (T0
or T1). The external input is sampled once during every machine cycle. When the pin is high during one cycle and low in the next
cycle, the count is incremented. The new count value appears in the register during the cycle following the one in which the
transition was detected. Since it takes 2 machine cycles (4 CPU clocks) to recognize a 1-to-0 transition, the maximum count rate
is 1/4 of the CPU clock frequency. There are no restrictions on the duty cycle of the external input signal, but to ensure that a
given level is sampled at least once before it changes, it should be held for at least one full machine cycle.
The “Timer” or “Counter” function is selected by control bits TnC/T
Function Register TMOD. Timer 0 and Timer 1 have five operating modes (modes 0, 1, 2, 3 and 6), which are selected by bitpairs (TnM1, TnM0) in TMOD and TnM2 in TAMOD. Modes 0, 1, 2 and 6 are the same for both Timers/Counters. Mode 3 is
different. The operating modes are described later in this section.
(x = 0 and 1 for Timers 0 and 1 respectively) in the Special
TMOD
Address: 89h
Not bit addressable
Reset Source(s): Any source
Reset Value: 00000000B
BITSYMBOLFUNCTION
TMOD.7T1GATEGating control for Timer 1. When set, Timer/Counter is enabled only while the INT1
TMOD.6T1C/T
TMOD.5, 4T1M1,T1M0Mode Select for Timer 1.These bits are used with the T1M2 bit in the TAMOD register to
TMOD.3T0GATEGating control for Timer 0. When set, Timer/Counter is enabled only while the INT0
TMOD.2T0C/T
TMOD.1, 0T0M1,T0M0Mode Select for Timer 0. These bits are used with the T0M2 bit in the TAMOD register to
Figure 7-1: Timer/Counter Mode Control register (TMOD)
76543210
T1GATET1C/T
high and the TR1 control pin is set. When cleared, Timer 1 is enabled when the TR1
control bit is set.
Timer or Counter Selector for Timer 1. Cleared for Timer operation (input from CCLK). Set
for Counter operation (input from T1 input pin).
determine the Timer 1 mode (see Figure 7-2).
high and the TR0 control pin is set. When cleared, Timer 0 is enabled when the TR0
control bit is set.
Timer or Counter Selector for Timer 0. Cleared for Timer operation (input from CCLK). Set
for Counter operation (input from T0 input pin).
determine the Timer 0 mode (see Figure 7-2).
T1M1T1M0T0GATET0C/TT0M1T0M0
pin is
pin is
2003 Nov 6 41
Philips Semiconductors
TIMERS 0 AND 1
User’s Manual - Preliminary -
P89LPC920/921/922
TAMOD
Address: 8Fh
Not bit addressable
Reset Source(s): Any reset
Reset Value: xxx0xxx0B
BITSYMBOLFUNCTION
TAMOD.7-5-Reserved for future use. Should not be set to 1 by user programs.
TAMOD.4T1M2Mode Select bit 2 for Timer 1. It is used with T1M1 and T1M0 in the TMOD register to
TAMOD.3-1-Reserved for future use. Should not be set to 1 by user programs.
TAMOD.0T0M2Mode Select bit 2 for Timer 0. It is used with T0M1 and T0M0 in the TMOD register to
0 0 116-bit Timer/Counter “THn” and “TLn” are cascaded; there is no prescaler.(Mode 1)
0 1 08-bit auto-reload Timer/Counter. THn holds a value which is loaded into TLn when it
0 1 1Timer 0 is a dual 8-bit Timer/Counter in this mode. TL0 is an 8-bit Timer/Counter controlled
1 0 0Reserved. User must not configure to this mode.
1 0 1Reserved. User must not configure to this mode.
1 1 0PWM mode (see section "Mode 6").
1 1 1Reserved. User must not configure to this mode.
76543210
---T1M2---T0M2
determine Timer 1 mode.
determine Timer 0 mode.
Timer Mode
overflows. (Mode 2)
by the standard Timer 0 control bits. TH0 is an 8-bit timer only, controlled by the Timer 1
control bits (see text). Timer 1 in this mode is stopped. (Mode 3)
Figure 7-2: Timer/Counter Auxiliary Mode Control register (TAMOD)
Mode 0
Putting either Timer into Mode 0 makes it look like an 8048 Timer, which is an 8-bit Counter with a divide-by-32 prescaler. Figure
7-4 shows Mode 0 operation.
In this mode, the Timer register is configured as a 13-bit register. As the count rolls over from all 1s to all 0s, it sets the Timer
interrupt flag TFn. The count input is enabled to the Timer when TRn = 1 and either TnGATE = 0 or INTn
= 1 allows the Timer to be controlled by external input INTn, to facilitate pulse width measurements). TRn is a control bit in the
Special Function Register TCON (Figure 7-3). The TnGATE bit is in the TMOD register.
The 13-bit register consists of all 8 bits of THn and the lower 5 bits of TLn. The upper 3 bits of TLn are indeterminate and should
be ignored. Setting the run flag (TRn) does not clear the registers.
Mode 0 operation is the same for Timer 0 and Timer 1. See Figure 7-4. There are two different GATE bits, one for Timer 1
(TMOD.7) and one for Timer 0 (TMOD.3).
= 1. (Setting TnGATE
Mode 1
Mode 1 is the same as Mode 0, except that all 16 bits of the timer register (THn and TLn) are used. See Figure 7-5.
2003 Nov 6 42
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
TIMERS 0 AND 1
Mode 2
Mode 2 configures the Timer register as an 8-bit Counter (TLn) with automatic reload, as shown in Figure 7-6. Overflow from TLn
not only sets TFn, but also reloads TLn with the contents of THn, which must be preset by software. The reload leaves THn
unchanged. Mode 2 operation is the same for Timer 0 and Timer 1.
Mode 3
When Timer 1 is in Mode 3 it is stopped. The effect is the same as setting TR1 = 0.
Timer 0 in Mode 3 establishes TL0 and TH0 as two separate 8-bit counters. The logic for Mode 3 on Timer 0 is shown in Figure
7-7. TL0 uses the Timer 0 control bits: T0C/T, T0GATE, TR0, INT0, and TF0. TH0 is locked into a timer function (counting
machine cycles) and takes over the use of TR1 and TF1 from Timer 1. Thus, TH0 now controls the “Timer 1” interrupt.
Mode 3 is provided for applications that require an extra 8-bit timer. With Timer 0 in Mode 3, an P89LPC920/921/922 device can
look like it has three Timer/Counters.
Note: When Timer 0 is in Mode 3, Timer 1 can be turned on and off by switching it into and out of its own Mode 3. It can still be
used by the serial port as a baud rate generator, or in any application not requiring an interrupt.
Mode 6
In this mode, the corresponding timer can be changed to a PWM with a full period of 256 timer clocks (see Figure 7-8). Its
structure is similar to mode 2, except that:
• TFn (n = 0 and 1 for Timers 0 and 1 respectively) is set and cleared in hardware;
• The low period of the TFn is in THn, and should be between 1 and 254, and;
• The high period of the TFn is always 256-THn.
• Loading THn with 00h will force the Tx pin high, loading THn with FFh will force the Tx pin low.
Note that interrupt can still be enabled on the low to high transition of TFn, and that TFn can still be cleared in software like in
any other modes.
2003 Nov 6 43
Philips Semiconductors
TIMERS 0 AND 1
User’s Manual - Preliminary -
P89LPC920/921/922
TCON
Address: 88h
Bit addressable
Reset Source(s): Any reset
Reset Value: 00000000B
BITSYMBOLFUNCTION
TCON.7TF1Timer 1 overflow flag. Set by hardware on Timer/Counter overflow. Cleared by hardware
TCON.6TR1Timer 1 Run control bit. Set/cleared by software to turn Timer/Counter 1 on/off.
TCON.5TF0Timer 0 overflow flag. Set by hardware on Timer/Counter overflow. Cleared by hardware
TCON.4TR0Timer 0 Run control bit. Set/cleared by software to turn Timer/Counter 0 on/off.
TCON.3IE1Interrupt 1 Edge flag. Set by hardware when external interrupt 1 edge is detected. Cleared
TCON.2IT1Interrupt 1 Type control bit. Set/cleared by software to specify falling edge/low level
TCON.1IE0Interrupt 0 Edge flag. Set by hardware when external interrupt 0 edge is detected. Cleared
TCON.0IT0Interrupt 0 Type control bit. Set/cleared by software to specify falling edge/low level
76543210
TF1TR1TF0TR0IE1IT1IE0IT0
when the interrupt is processed, or by software (except in mode 6, see above, when it is
cleared in hardware).
when the processor vectors to the interrupt routine, or by software. (except in mode 6, see
above, when it is cleared in hardware)
by hardware when the interrupt is processed, or by software.
triggered external interrupts.
by hardware when the interrupt is processed, or by software.
triggered external interrupts.
PCLK
Tn Pin
INTn
TRn
Gate
Pin
Figure 7-3: Timer/Counter Control register (TCON)
C/T
= 0
C/T = 1
Figure 7-4: Timer/Counter 0 or 1 in Mode 0 (13-bit counter)
Control
TLn
(5-bits)
(8-bits)
Overflow
THn
Toggle
TFn
ENTn
Interrupt
Tn Pin
2003 Nov 6 44
Philips Semiconductors
TIMERS 0 AND 1
User’s Manual - Preliminary -
P89LPC920/921/922
PCLK
Tn Pin
INTn
Tn Pin
INTn
TRn
Gate
Pin
PCLK
TRn
Gate
Pin
= 0
C/T
= 1
C/T
Figure 7-5: Timer/Counter 0 or 1 in Mode 1 (16-bit counter)
= 0
C/T
C/T = 1
Control
Control
TLn
(8-bits)
TLn
(8-bits)
THn
(8-bits)
(8-bits)
Reload
Overflow
THn
Toggle
Overflow
Toggle
TFn
ENTn
TFn
ENTn
Interrupt
Tn Pin
Interrupt
Tn Pin
T0 Pin
INT0
PCLK
TR0
Gate
Pin
Figure 7-6: Timer/Counter 0 or 1 in Mode 2 (8-bit auto-reload)
Figure 7-8: Timer/Counter 0 or 1 in Mode 6 (PWM auto-reload)
Timer overflow toggle output
Timers 0 and 1 can be configured to automatically toggle a port output whenever a timer overflow occurs. The same device pins
that are used for the T0 and T1 count inputs and PWM outputs are also used for the timer toggle outputs. This function is enabled
by control bits ENT0 and ENT1 in the AUXR1 register, and apply to Timer 0 and Timer 1 respectively. The port outputs will be a
logic 1 prior to the first timer overflow when this mode is turned on.In order for this mode to function, the C/T
selecting PCLK as the clock source for the timer.
= 0
Control
TLn
(8-bits)
Reload THn on falling transition
and (256-THn) on rising transition
THn
(8-bits)
Overflow
Toggle
TFn
ENTn
bit must be cleared
Interrupt
Tn Pin
2003 Nov 6 46
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
REAL-TIME CLOCK/SYSTEM TIMER
8. REAL-TIME CLOCK/SYSTEM TIMER
The P89LPC920/921/922 has a simple Real-time Clock/System Timer that allows a user to continue running an accurate timer
while the rest of the device is powered down. The Real-time Clock can be an interrupt or a wake-up source (see Figure 8-1). The
Real-time Clock is a 23-bit down counter. The clock source for this counter can be either the CPU clock (CCLK) or the XTAL1-2
oscillator, provided that the XTAL1-2 oscillator is not being used as the CPU clock. If the XTAL1-2 oscillator is used as the CPU
clock, then the RTC will use CCLK as its clock source regardless of the state of the RTCS1:0 in the RTCCON register. There are
three SFRs used for the RTC:
• RTCCON - Real-time Clock control.
• RTCH - Real-time Clock counter reload high (bits 22-15).
The Real-time Clock/System Timer can be enabled by setting the RTCEN (RTCCON.0) bit. The Real-time Clock is a 23-bit down
counter (initialized to all 0’s when RTCEN = 0) that is comprised of a 7-bit prescaler and a 16-bit loadable down counter. When
RTCEN is written with ‘1’, the counter is first loaded with (RTCH,RTCL,’1111111’) and will count down. When it reaches all 0’s,
the counter will be reloaded again with (RTCH,RTCL,’1111111’) and a flag - RTCF (RTCCON.7) - will be set.
Any write to RTCH and RTCL in-between the Real-time Clock reloading will not cause reloading of the counter. When the current
count terminates, the contents of RTCH and RTCL will be loaded into the counter and the new count will begin. An immediate
reload of the counter can be forced by clearing the RTCEN bit to ‘0’ and then setting it back to ‘1’ .
RTCS1-0 (RTCCON.6-5) are used to select the clock source for the RTC if either the Internal RC oscillator or the internal WD
oscillator is used as the CPU clock. If the internal crystal oscillator or the external clock input on XTAL1 is used as the CPU clock,
then the RTC will use CCLK as its clock source.
Changing RTCS1-0
RTCS1-0 cannot be changed if the RTC is currently enabled (RTCCON.0 =1). Setting RTCEN and updating RTCS1-0 may be
done in a single write to RTCCON. However, if RTCEN = 1, this bit must first be cleared before updating RTCS1-0.
2003 Nov 6 47
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
REAL-TIME CLOCK/SYSTEM TIMER
Real-time Clock interrupt/wake up
If ERTC (RTCCON.1), EWDRT (IEN1.0.6) and EA (IEN0.7) are set to ‘1’, RTCF can be used as an interrupt source. This interrupt
vector is shared with the watchdog timer. It can also be a source to wake up the device.
Reset sources affecting the Real-time Clock
Only power-on reset will reset the Real-time Clock and its associated SFRs to their default state
RTCCON.7RTCFReal-time Clock Flag. This bit is set to ‘1’ when the 23-bit Real-time Clock reaches a count
of ‘0’. It can be cleared in software.
RTCCON.6-5RTCS1-0Real-time Clock source select (see Table 8-1).
RTCCON.4-2-Reserved for future use. Should not be set to 1 by user programs.
RTCCON.1ERTCReal-time Clock interrupt enable. The Real-time Clock shares the same interrupt as the
watchdog timer. Note that if the user configuration bit WDTE (UCFG1.7) is ‘0’, the
watchdog timer can be enabled to generate an interrupt. Users can read the RTCF
(RTCCON.7) bit to determine whether the Real-time Clock caused the interrupt.
RTCCON.0RTCENReal-time Clock enable. The Real-time Clock will be enabled if this bit is ‘1’. Note that this
bit will not Power down the Real-time Clock. The RTCPD bit (PCONA.7) if set, will Power
down and disable this block regardless of RTCEN.
76543210
RTCFRTCS1RTCS0---ERTCRTCEN
Figure 8-2: RTCCON Register
2003 Nov 6 49
Philips Semiconductors
REAL-TIME CLOCK/SYSTEM TIMER
User’s Manual - Preliminary -
P89LPC920/921/922
2003 Nov 6 50
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
UART
9. UART
The P89LPC920/921/922 has an enhanced UART that is compatible with the conventional 80C51 UART except that Timer 2
overflow cannot be used as a baud rate source. The P89LPC920/921/922 does include an independent Baud Rate Generator.
The baud rate can be selected from the oscillator (divided by a constant), Timer 1 overflow, or the independent Baud Rate
Generator. In addition to the baud rate generation, enhancements over the standard 80C51 UART include Framing Error
detection, break detect, automatic address recognition, selectable double buffering and several interrupt options.
The UART can be operated in 4 modes:
Mode 0
Serial data enters and exits through RxD. TxD outputs the shift clock. 8 bits are transmitted or received, LSB first. The baud rate
is fixed at 1/16 of the CPU clock frequency.
Mode 1
10 bits are transmitted (through TxD) or received (through RxD): a start bit (logical 0), 8 data bits (LSB first), and a stop bit (logical
1). When data is received, the stop bit is stored in RB8 in Special Function Register SCON. The baud rate is variable and is
determined by the Timer 1 overflow rate or the Baud Rate Generator (see “Baud Rate Generator and selection” section).
Mode 2
11 bits are transmitted (through TxD) or received (through RxD): start bit (logical 0), 8 data bits (LSB first), a programmable 9th
data bit, and a stop bit (logical 1). When data is transmitted, the 9th data bit (TB8 in SCON) can be assigned the value of 0 or 1.
Or, for example, the parity bit (P, in the PSW) could be moved into TB8. When data is received, the 9th data bit goes into RB8
in Special Function Register SCON and the stop bit is not saved. The baud rate is programmable to either 1/16 or 1/32 of the
CCLK frequency, as determined by the SMOD1 bit in PCON.
Mode 3
11 bits are transmitted (through TxD) or received (through RxD): a start bit (logical 0), 8 data bits (LSB first), a programmable 9th
data bit, and a stop bit (logical 1). Mode 3 is the same as Mode 2 in all respects except baud rate. The baud rate in Mode 3 is
variable and is determined by the Timer 1 overflow rate or the Baud Rate Generator (see “Baud Rate Generator and selection”
section).
In all four modes, transmission is initiated by any instruction that uses SBUF as a destination register. Reception is initiated in
Mode 0 by the condition RI = 0 and REN = 1. Reception is initiated in the other modes by the incoming start bit if REN = 1.
SFR space
The UART SFRs are at the following locations:
Table 9-1: SFR Locations for UARTs
RegisterDescriptionSFR Location
PCONPower Control87H
SCONSerial Port (UART) Control98H
SBUFSerial Port (UART) Data Buffer99H
SADDRSerial Port (UART) AddressA9H
SADENSerial Port (UART) Address EnableB9H
2003 Nov 6 51
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
UART
RegisterDescriptionSFR Location
SSTATSerial Port (UART) StatusBAH
BRGR1Baud Rate Generator Rate High ByteBFH
BRGR0Baud Rate Generator Rate Low ByteBEH
BRGCON Baud Rate Generator ControlBDH
Baud Rate Generator and selection
The P89LPC920/921/922 enhanced UART has an independent Baud Rate Generator. The baud rate is determined by a value
programmed into the BRGR1 and BRGR0 SFRs. The UART can use either Timer 1 or the baud rate generator output as
determined by BRGCON.2-1 (see Figure 9-2). Note that Timer T1 is further divided by 2 if the SMOD1 bit (PCON.7) is set. The
independent Baud Rate Generator uses CCLK.
Updating the BRGR1 and BRGR0 SFRs
The baud rate SFRs, BRGR1 and BRGR0 must only be loaded when the Baud Rate Generator is disabled (the BRGEN bit in
the BRGCON register is ‘0’). This avoids the loading of an interim value to the baud rate generator. (CAUTION: If either BRGR0
or BRGR1 is written when BRGEN = 1, the result is unpredictable.)
Table 9-2: Baud rate generation for UART
SCON.7
(SM0)
SCON.6
(SM1)
PCON.7
(SMOD1)
BRGCON.1
(SBRGS)
Receive/transmit baud rate for UART
00XXCCLK/16
00CCLK/(256-TH1)64
01
10
11
10CCLK/(256-TH1)32
X1CCLK/((BRGR1,BRGR0)+16)
0XCCLK/32
1XCCLK/16
00CCLK/(256-TH1)64
10CCLK/(256-TH1)32
X1CCLK/((BRGR1,BRGR0)+16)
2003 Nov 6 52
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
UART
BRGCON
Address: BDh
Not bit addressable
Reset Source(s): Any reset
Reset Value: xxxxxx00B
BITSYMBOLFUNCTION
BRGCON.7-2-Reserved for future use. Should not be set to 1 by user programs.
BRGCON.1SBRGSSelect Baud Rate Generator as the source for baud rates to UART in modes 1 & 3 (see
BRGCON.0BRGENBaud Rate Generator Enable. Enables the baud rate generator. BRGR1 and BRGR0 can
76543210
------SBRGS BRGEN
Table 9-2 for details)
only be written when BRGEN =0.
Figure 9-1: BRGCON register
Timer 1 Overflow
(PCLK-based)
Baud Rate Generator
(CCLK-based)
SMOD1 = 1
÷2
SMOD1 = 0
SBRGS = 0
Baud Rate Modes 1 and 3
SBRGS = 1
Figure 9-2: Baud rate generation for UART (Modes 1, 3)
Framing Error
A Framing error occurs when the stop bit is sensed as a logic ‘0’. A Framing error is reported in the status register (SSTAT). In
addition, if SMOD0 (PCON.6) is 1, framing errors can be made available in SCON.7. If SMOD0 is 0, SCON.7 is SM0. It is
recommended that SM0 and SM1 (SCON.7-6) are programmed when SMOD0 is ‘0’.
Break Detect
A break detect is reported in the status register (SSTAT). A break is detected when any 11 consecutive bits are sensed low.
Since a break condition also satisfies the requirements for a framing error, a break condition will also result in reporting a framing
error. Once a break condition has been detected, the UART will go into an idle state and remain in this idle state until a stop bit
has been received. The break detect can be used to reset the device and force the device into ISP mode by setting the EBRR
bit (AUXR1.6)
2003 Nov 6 53
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
UART
.
SCON
Address: 98h
Bit addressable
Reset Source(s): Any reset
Reset Value: 00000000B
BITSYMBOLFUNCTION
SCON.7SM0/FEThe use of this bit is determined by SMOD0 in the PCON register. If SMOD0 = 0, this bit
is read and written as SM0, which with SM1, defines the serial port mode. If SMOD0 = 1,
this bit is read and written as FE (Framing Error). FE is set by the receiver when an invalid
stop bit is detected. Once set, this bit cannot be cleared by valid frames but is cleared by
software. (Note: UART mode bits SM0 and SM1 should be programmed when SMOD0 is
‘0’ - default mode on any reset.)
SCON. 6SM1With SM0, defines the serial port mode (see table below).
SM0, SM1
0 00: shift registerCCLK/16 (default mode on any reset)
0 11: 8-bit UARTVariable (see Table 9-2)
1 02: 9-bit UARTCCLK/32 or CCLK/16
1 13: 9-bit UART Variable (see Table 9-2)
SCON.5SM2Enables the multiprocessor communication feature in Modes 2 and 3. In Mode 2 or 3, if
SCON.4RENEnables serial reception. Set by software to enable reception. Clear by software to disable
SCON.3TB8The 9th data bit that will be transmitted in Modes 2 and 3. Set or clear by software as
SCON.2RB8The 9th data bit that was received in Modes 2 and 3. In Mode 1 (SM2 must be 0), RB8 is
SCON.1TITransmit interrupt flag. Set by hardware at the end of the 8th bit time in Mode 0, or at the
SCON.0RIReceive interrupt flag. Set by hardware at the end of the 8th bit time in Mode 0, or
UART ModeUART 0 Baud Rate
SM2 is set to 1, then Rl will not be activated if the received 9th data bit (RB8) is 0. In Mode
0, SM2 should be 0. In Mode 1, SM2 must be 0.
reception.
desired.
the stop bit that was received. In Mode 0, RB8 is undefined.
the stop bit (see description of INTLO bit in SSTAT register) in the other modes. Must be
cleared by software.
approximately halfway through the stop bit time in Mode 1. For Mode 2 or Mode 3, if
SMOD0, it is set near the middle of the 9th data bit (bit 8). If SMOD0 = 1, it is set near the
middle of the stop bit (see SM2 - SCON.5 - for exceptions). Must be cleared by software.
76543210
SM0/FESM1SM2RENTB8RB8TIRI
Figure 9-3: Serial Port Control register (SCON)
2003 Nov 6 54
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
UART
SSTAT
Address: BAh
Not bit addressable
Reset Source(s): Any reset
Reset Value: 00000000B
BITSYMBOLFUNCTION
SSTAT.7DBMODDouble buffering mode. When set = 1 enables double buffering. Must be ‘0’ for UART
SSTAT.6INTLOTransmit interrupt position. When cleared = 0, the Tx interrupt is issued at the beginning
SSTAT.5CIDISCombined Interrupt Disable. When set = 1, Rx and Tx interrupts are separate. When
SSTAT.4DBISELDouble buffering transmit interrupt select. Used only if double buffering is enabled.This bit
SSTAT.3FEFraming error flag is set when the receiver fails to see a valid STOP bit at the end of the
SSTAT.2BRBreak Detect flag. A break is detected when any 11 consecutive bits are sensed low.
SSTAT.1OEOverrun Error flag is set if a new character is received in the receiver buffer while it is still
SSTAT.0STINTStatus Interrupt Enable. When set =1, FE, BR, or OE can causean interrupt. The interrupt
76543210
DBMOD INTLOCIDISDBISELFEBROESTINT
mode 0. In order to be compatible with existing 80C51 devices, this bit is reset to ‘0’ to
disable double buffering.
of the stop bit. When set =1, the Tx interrupt is issued at end of the stop bit. Must be ‘0’
for mode 0. Note that in the case of single buffering, if the Tx interrupt occurs at the end
of a STOP bit, a gap may exist before the next start bit.
cleared = 0, the UART uses a combined Tx/Rx interrupt (like a conventional 80C51
UART). This bit is reset to ‘0’ to select combined interrupts.
controls the number of interrupts that can occur when double buffering is enabled. When
set, one transmit interrupt is generated after each character written to SBUF, and there is
also one more transmit interrupt generated at the beginning (INTLO = 0) or the end
(INTLO = 1) of the STOP bit of the last character sent (i.e., no more data in buffer). This
last interrupt can be used to indicate that all transmit operations are over. When cleared
= 0, only one transmit interrupt is generated per character written to SBUF. Must be ‘0’
when double buffering is disabled.
Note that except for the first character written (when buffer is empty), the location of the
transmit interrupt is determined by INTLO. When the first character is written, the transmit
interrupt is generated immediately after SBUF is written.
frame. Cleared by software.
Cleared by software.
full (before the software has read the previous character from the buffer), i.e., when bit 8
of a new byte is received while RI in SCON is still set. Cleared by software.
used (vector address 0023h) is shared with RI (CIDIS = 1) or the combined TI/RI (CIDIS
= 0). When cleared = 0, FE, BR, OE cannot cause an interrupt. (Note: FE, BR, or OE is
often accompanied by a RI, which will generate an interrupt regardless of the state of
STINT). Note that BR can cause a break detect reset if EBRR (AUXR1.6) is set to ‘1’.
Figure 9-4: Serial Port Status register (SSTAT)
More about UART Mode 0
In Mode 0, a write to SBUF will initiate a transmission. At the end of the transmission, TI (SCON.1) is set, which must be cleared
in software. Double buffering must be disabled in this mode.
Reception is initiated by clearing RI (SCON.0). Synchronous serial transfer occurs and RI will be set again at the end of the
transfer. When RI is cleared, the reception of the next character will begin. Refer to Figure 9-5 for timing.
Figure 9-5: Serial Port Mode 0 (double buffering must be disabled)
More about UART Mode 1
Reception is initiated by detecting a 1-to-0 transition on RxD. RxD is sampled at a rate 16 times the programmed baud rate. When
a transition is detected, the divide-by-16 counter is immediately reset. Each bit time is thus divided into 16 counter states. At the
7th, 8th, and 9th counter states, the bit detector samples the value of RxD. The value accepted is the value that was seen in at
least 2 of the 3 samples. This is done for noise rejection. If the value accepted during the first bit time is not 0, the receive circuits
are reset and the receiver goes back to looking for another 1-to-0 transition. This provides rejection of false start bits. If the start
bit proves valid, it is shifted into the input shift register, and reception of the rest of the frame will proceed.
The signal to load SBUF and RB8, and to set RI, will be generated if, and only if, the following conditions are met at the time the
final shift pulse is generated: RI = 0 and either SM2=0 or the received stop bit =1. If either of these two conditions is not met, the
received frame is lost. If both conditions are met, the stop bit goes into RB8, the 8 data bits go into SBUF, and RI is activated.
TX Clock
Write to SBUF
Shift
TxD
TI
RX Clock
RxD
Shift
RI
÷ 16 Reset
Start BitStop Bit
Start Bit
D0D1D5D2D6D3D4D7
D0D1D5D2D6D3D4D7
Figure 9-6: Serial Port Mode 1 (only single transmit buffering case is shown)
2003 Nov 6 56
INTLO = 0
Transmit
INTLO = 1
Stop Bit
Receive
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
UART
More about UART Modes 2 and 3
Reception is the same as in Mode 1.
The signal to load SBUF and RB8, and to set RI, will be generated if, and only if, the following conditions are met at the time the
final shift pulse is generated. (a) RI = 0, and (b) Either SM2 = 0, or the received 9th data bit = 1. If either of these conditions is
not met, the received frame is lost, and RI is not set. If both conditions are met, the received 9th data bit goes into RB8, and the
first 8 data bits go into SBUF.
TX Clock
Write to SBUF
Shift
TxD
TI
RX Clock
Start BitStop Bit
D0D1D5D2D6D3D4D7
TB8
INTLO = 0
Transmit
INTLO = 1
RxD
Shift
RI
÷ 16 Reset
Start Bit
D0D1D5D2D6D3D4D7
Figure 9-7: Serial Port Mode 2 or 3 (only single transmit buffering case is shown)
Framing Error and RI in Modes 2 and 3 with SM2 = 1
If SM2 = 1 in modes 2 and 3, RI and FE behaves as in the following table.
Table 9-3: FE and RI when SM2 = 1 in Modes 2 and 3
Mode
20
31
PCON.6
(SMOD0)
RB8RIFE
0No RI when RB8 = 0Occurs during STOP bit
1
Similar to Figure 9-7, with SMOD0 = 0, RI
occurs during RB8, one bit before FE
0No RI when RB8 = 0Will NOT occur
1
Similar to Figure 9-7, with SMOD0 = 1, RI
occurs during STOP bit
Stop Bit
RB8
SMOD0 = 0
SMOD0 = 1
Occurs during STOP bit
Occurs during STOP bit
Receive
Break Detect
. A break is detected when 11 consecutive bits are sensed low and is reported in the status register (SSTAT). For Mode 1, this
consists of the start bit, 8 data bits, and two stop bit times. For Modes 2 & 3, this consists of the start bit, 9 data bits, and one stop
2003 Nov 6 57
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
UART
bit. The break detect bit is cleared in software or by a reset. The break detect can be used to reset the device and force the device
into ISP mode. This occurs if the UART is enabled and the the EBRR bit (AUXR1.6) is set and a break occurs.
Double buffering
The UART has a transmit double buffer that allows buffering of the next character to be wriiten to SBUF while the first character
is being transmitted. Double buffering allows transmission of a string of characters with only one stop bit between any two
characters, provided the next character is written between the start bit and the stop bit of the previous character.
Double buffering can be disabled. If disabled (DBMOD, i.e. SSTAT.7 = 0), the UART is compatible with the conventional 80C51
UART. If enabled, the UART allows writing to SnBUF while the previous data is being shifted out.
Double buffering in different modes
Double buffering is only allowed in Modes 1, 2 and 3. When operated in Mode 0, double buffering must be disabled (DBMOD = 0).
Transmit interrupts with double buffering enabled (Modes 1, 2 and 3)
Unlike the conventional UART, when double buffering is enabled, the Tx interrupt is generated when the double buffer is ready
to receive new data. The following occurs during a transmission (assuming eight data bits):
1. The double buffer is empty initially.
2. The CPU writes to SBUF.
3. The SBUF data is loaded to the shift register and a Tx interrupt is generated immediately.
4. If there is more data, go to 6, else continue on 5.
5. If there is no more data, then:
- If DBISEL is ‘0’, no more interrupts will occur.
- If DBISEL is ‘1’ and INTLO is ‘0’, a Tx interrupt will occur at the beginning of the STOP bit of the data currently in the shifter
(which is also the last data).
- If DBISEL is ‘1’ and INTLO is ‘1’, a Tx interrupt will occur at the end of the STOP bit of the data currently in the shifter (which
is also the last data).
6. If there is more data, the CPU writes to SBUF again. Then:
- If INTLO is ‘0’, the new data will be loaded and a Tx interrupt will occur at the beginning of the STOP bit of the data currently in the shifter.
- If INTLO is ‘1’, the new data will be loaded and a Tx interrupt will occur at the end of the STOP bit of the data currently in
the shifter.
Go to 3.
Note that if DBISEL is ‘1’ and the CPU is writing to SBUF when the STOP bit of the last data is shifted out, there can be an
uncertainty of whether a Tx interrupt is generated already with the UART not knowing whether there is any more data following.
2003 Nov 6 58
Philips Semiconductors
UART
TxD
Write to
SBUF
Tx Interrupt
TxD
Write to
SBUF
Tx Interrupt
Double Buffering (DBMOD/SSTAT.7 = 1), Early Interrupt (INTLO/SSTAT.6 = 0) is Shown, No End-
User’s Manual - Preliminary -
P89LPC920/921/922
Single Buffering (DBMOD/SSTAT.7 = 0), Early Interrupt (INTLO/SSTAT.6 = 0) is Shown
ing Tx Interrupt (DBISEL/SnSTAT.4 = 0)
TxD
Write to
SBUF
Tx Interrupt
Double Buffering (DBMOD/SSTAT.7 = 1), Early Interrupt (INTLO/SSTAT.6 = 0) is Shown, With
Ending Tx Interrupt (DBISEL/SSTAT.4 = 1)
Figure 9-8: Transmission with and without double buffering
The 9th bit (bit 8) in double buffering (Modes 1, 2 and 3)
If double buffering is disabled (DBMOD, i.e. SSTAT.7 = 0), TB8 can be written before or after SBUF is written, provided TB8 is
updated before that TB8 is shifted out. TB8 must not be changed again until after TB8 shifting has been completed, as indicated
by the Tx interrupt.
If double buffering is enabled, TB8 MUST be updated before SBUF is written, as TB8 will be double-buffered together with SBUF
data. The operation described in the section "Transmit interrupts with double buffering enabled (Modes 1, 2 and 3)" becomes as
follows:
1. The double buffer is empty initially.
2. The CPU writes to TB8.
3. The CPU writes to SBUF.
4. The SBUF/TB8 data is loaded to the shift register and a Tx interrupt is generated immediately.
5. If there is more data, go to 7, else continue on 6.
6. If there is no more data, then:
- If DBISEL is ‘0’, no more interrupt will occur.
- If DBISEL is ‘1’ and INTLO is ‘0’, a Tx interrupt will occur at the beginning of the STOP bit of the data currently in the shifter
(which is also the last data).
- If DBISEL is ‘1’ and INTLO is ‘1’, a Tx interrupt will occur at the end of the STOP bit of the data currently in the shifter (which
is also the last data).
7. If there is more data, the CPU writes to TB8 again.
2003 Nov 6 59
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
UART
8. The CPU writes to SBUF again. Then:
- If INTLO is ‘0’, the new data will be loaded and a Tx interrupt will occur at the beginning of the STOP bit of the data currently in the shifter.
- If INTLO is ‘1’, the new data will be loaded and a Tx interrupt will occur at the end of the STOP bit of the data currently in
the shifter.
Go to 4.
Note that if DBISEL is ‘1’ and the CPU is writing to SBUF when the STOP bit of the last data is shifted out, there can be an
uncertainty of whether a Tx interrupt is generated already with the UART not knowing whether there is any more data
following.
Multiprocessor communications
UART modes 2 and 3 have a special provision for multiprocessor communications. In these modes, 9 data bits are received or
transmitted. When data is received, the 9th bit is stored in RB8. The UART can be programmed such that when the stop bit is
received, the serial port interrupt will be activated only if RB8 = 1. This feature is enabled by setting bit SM2 in SCON. One way
to use this feature in multiprocessor systems is as follows:
When the master processor wants to transmit a block of data to one of several slaves, it first sends out an address byte which
identifies the target slave. An address byte differs from a data byte in that the 9th bit is 1 in an address byte and 0 in a data byte.
With SM2 = 1, no slave will be interrupted by a data byte. An address byte, however, will interrupt all slaves, so that each slave
can examine the received byte and see if it is being addressed. The addressed slave will clear its SM2 bit and prepare to receive
the data bytes that follow. The slaves that weren’t being addressed leave their SM2 bits set and go on about their business,
ignoring the subsequent data bytes.
Note that SM2 has no effect in Mode 0, and must be ‘0’ in Mode 1.
Automatic address recognition
Automatic address recognition is a feature which allows the UART to recognize certain addresses in the serial bit stream by using
hardware to make the comparisons. This feature saves a great deal of software overhead by eliminating the need for the software
to examine every serial address which passes by the serial port. This feature is enabled by setting the SM2 bit in SCON. In the
9 bit UART modes (mode 2 and mode 3), the Receive Interrupt flag (RI) will be automatically set when the received byte contains
either the “Given” address or the “Broadcast” address. The 9 bit mode requires that the 9th information bit is a 1 to indicate that
the received information is an address and not data.
Using the Automatic Address Recognition feature allows a master to selectively communicate with one or more slaves by
invoking the Given slave address or addresses. All of the slaves may be contacted by using the Broadcast address. Two special
Function Registers are used to define the slave’s address, SADDR, and the address mask, SADEN. SADEN is used to define
which bits in the SADDR are to be used and which bits are “don’t care”. The SADEN mask can be logically ANDed with the
SADDR to create the “Given” address which the master will use for addressing each of the slaves. Use of the Given address
allows multiple slaves to be recognized while excluding others. The following examples will help to show the versatility of this
scheme:
Slave 0 SADDR = 1100 0000
Slave 1 SADDR = 1100 0000
In the above example SADDR is the same and the SADEN data is used to differentiate between the two slaves. Slave 0 requires
a 0 in bit 0 and it ignores bit 1. Slave 1 requires a 0 in bit 1 and bit 0 is ignored. A unique address for Slave 0 would be 1100 0010
since slave 1 requires a 0 in bit 1. A unique address for slave 1 would be 1100 0001 since a 1 in bit 0 will exclude slave 0. Both
slaves can be selected at the same time by an address which has bit 0 = 0 (for slave 0) and bit 1 = 0 (for slave 1). Thus, both
could be addressed with 1100 0000.
In a more complex system the following could be used to select slaves 1 and 2 while excluding slave 0:
SADEN = 1111 1101
Given= 1100 00X0
SADEN = 1111 1110
Given= 1100 000X
2003 Nov 6 60
Philips Semiconductors
UART
User’s Manual - Preliminary -
P89LPC920/921/922
Slave 0 SADDR = 1100 0000
Slave 1 SADDR = 1110 0000
Slave 2 SADDR = 1110 0000
In the above example the differentiation among the 3 slaves is in the lower 3 address bits. Slave 0 requires that bit 0 = 0 and it
can be uniquely addressed by 1110 0110. Slave 1 requires that bit 1 = 0 and it can be uniquely addressed by 1110 and 0101.
Slave 2 requires that bit 2 = 0 and its unique address is 1110 0011. To select Slaves 0 and 1 and exclude Slave 2 use address
1110 0100, since it is necessary to make bit 2 = 1 to exclude slave 2. The Broadcast Address for each slave is created by taking
the logical OR of SADDR and SADEN. Zeros in this result are treated as don’t-cares. In most cases, interpreting the don’t-cares
as ones, the broadcast address will be FF hexadecimal. Upon reset SADDR and SADEN are loaded with 0s. This produces a
given address of all “don’t cares” as well as a Broadcast address of all “don’t cares”. This effectively disables the Automatic
Addressing mode and allows the microcontroller to use standard UART drivers which do not make use of this feature.
SADEN = 1111 1001
Given= 1100 0XX0
SADEN = 1111 1010
Given= 1110 0X0X
SADEN = 1111 1100
Given= 1110 00XX
2003 Nov 6 61
Philips Semiconductors
UART
User’s Manual - Preliminary -
P89LPC920/921/922
2003 Nov 6 62
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
I2C INTERFACE
10. I2C INTERFACE
The I2C-bus uses two wires, serial clock (SCL) and serial data (SDA) to transfer information between devices connected to the
bus, and has the following features:
• Bidirectional data transfer between masters and slaves
• Multimaster bus (no central master)
• Arbitration between simultaneously transmitting masters without corruption of serial data on the bus
• Serial clock synchronization allows devices with different bit rates to communicate via one serial bus
• Serial clock synchronization can be used as a handshake mechanism to suspend and resume serial transfer
•The I
A typical I
are possible on the I
• Data transfer from a master transmitter to a slave receiver. The first byte transmitted by the master is the slave address. Next
• Data transfer from a slave transmitter to a master receiver. The first byte (the slave address) is transmitted by the master. The
The P89LPC920/921/922 device provides a byte-oriented I
Master Receiver Mode, Slave Transmitter Mode and Slave Receiver Mode.
2
C-bus may be used for test and diagnostic purposes
2
C-bus configuration is shown in Figure 1. Depending on the state of the direction bit (R/W), two types of data transfers
follows a number of data bytes. The slave returns an acknowledge bit after each received byte.
slave then returns an acknowledge bit. Next follows the data bytes transmitted by the slave to the master. The master returns
an acknowledge bit after all received bytes other than the last byte. At the end of the last received byte, a “not acknowledge”
is returned. The master device generates all of the serial clock pulses and the START and STOP conditions. A transfer is ended
with a STOP condition or with a repeated START condition. Since a repeated START condition is also the beginning of the
next serial transfer, the I
2
C-bus:
2
C-bus will not be released.
2
C interface. It has four operation modes: Master Transmitter Mode,
R
P
I2C-bus
P1.3/SDA P1.2/SCL
P89LPC920/921/922
The P89LPC920/921/922 CPU interfaces with theI
Register), I2DAT (I
Cycle Register High Byte), and I2SCLL (SCL Duty Cycle Register Low Byte).
2
C Data Register), I2STAT (I2C Status Register), I2ADR (I2C Slave Address Register), I2SCLH (SCL Duty
Other Device with I
Interface
Figure 1: I
2
2
C-bus through six Special Function Registers (SFRs): I2CON (I2C Control
R
P
2
C
C-bus configuration
Other Device with I
Interface
SDA
SCL
2
C
2003 Nov 6 63
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
I2C INTERFACE
I2C Data register
I2DAT register contains the data to be transmitted or the data received. The CPU can read and write to this 8-bit register while
it is not in the process of shifting a byte. Thus this register should only be accessed when the SI bit is set. Data in I2DAT remains
stable as long as the SI bit is set. Data in I2DAT is always shifted from right to left: the first bit to be transmitted is the MSB (bit
7), and after a byte has been received, the first bit of received data is located at the MSB of I2DAT.
I2ADR register is readable and writable, and is only used when the I2C interface is set to slave mode. In master mode, this
register has no effect. The LSB of I2ADR is general call bit. When this bit is set, the general call address (00h) is recognized.
I2ADR7, 1I2ADR.6, 07 bit own slave address. When in master mode, the contents of this register has no effect.
I2ADR7.0GCGeneral call bit. When set, the general call address (00H) is recognized, otherwise it is
ignored.
2
Figure 3: I
C Slave Address register
I2C Control register
The CPU can read and write this register. There are two bits are affected by hardware: the SI bit and the STO bit. The SI bit is
set by hardware and the STO bit is cleared by hardware.
CRSEL determines the SCL source when the I2C is in master mode. In slave mode this bit is ignored and the bus will
automatically synchronize with any clock frequency up to 400 kHz from the master I
interface uses the Timer1 overflow rate divided by 2 for the I
2
C clock rate. Timer 1 should be programmed by the user in 8 bit
C data rate range is 11.72 Kbit/sec - 3000 Kbit/sec.
2
C interface uses the internal clock generator based on the value of I2SCLL and I2CSCLH register. The
duty cycle does not need to be 50%.
2
C device. When CRSEL = 1, the I2C
2003 Nov 6 64
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
I2C INTERFACE
The STA bit is START flag. Setting this bit causes the I2C interface to enter master mode and attempt transmitting a START
condition or transmitting a repeated START condition when it is already in master mode.
The STO bit is STOP flag. Setting this bit causes the I2C interface to transmit a STOP condition in master mode, or recovering
from an error condition in slave mode.
2
If the STA and STO are both set, then a STOP condition is transmitted to the I
START condition afterwards. If it is in slave mode, an internal STOP condition will be generated, but it is not transmitted to the
bus.
I2CON
Address: D8h
Bit addressable
76543210
-I2ENSTASTOSIAA-CRSEL
Reset Source(s): Any reset
Reset Value: x00000x0B
BITSYMBOLFUNCTION
I2CON.7-Reserved for future use. Should not be set to 1 by user programs.
2
I2CON.6I2ENI
C Interface Enable. When set, enables the I2C interface. When clear, the I2C function is
disabled.
2
I2CON.5STAStart Flag. STA = 1: I
C enters master mode, checks the bus and generates a START
condition if the bus is free. If the bus is not free, it waits for a STOP condition (which will
free the bus) and generates a START condition after a delay of a half clock period of the
internal clock generator. When the I
2
is transmitted or received, it transmits a repeated START condition. STA may be set at
any time, it may also be set when the I
STA = 0: no START condition or repeated START condition will be generated.
I2CON.4STOSTOP Flag. STO = 1: In master mode, a STOP condition is transmitted to the I
When the bus detects the STOP condition, it will clear STO bit automatically.
In slave mode, setting this bit can recover from an error condition. In this case, no STOP
condition is transmitted to the bus. The hardware behaves as if a STOP condition has
been received and it switches to “not addressed” Slave Receiver Mode. The STO flag is
cleared by hardware automatically.
2
I2CON.3SII
C Interrupt Flag. This bit is set when one of the 25 possible I2C states is entered.
When EA bit and EI2C (IEN1.0) bit are both set, an interrupt is requested when SI is set..
Must be cleared by software by writing 0 to this bit.
I2CON.2AAThe Assert Acknowledge Flag. When set to 1, an acknowledge (low level to SDA) will be
returned during the acknowledge clock pulse on the SCL line on the following situations:
(1)The “own slave address” has been received. (2)The general call address has been
received while the general call bit(GC) in I2ADR is set. (3) A data byte has been received
while the I
while the I
2
C interface is in the Master Receiver Mode. (4)A data byte has been received
2
C interface is in the addressed Slave Receiver Mode
When cleared to 0, an not acknowledge (high level to SDA) will be returned during the
acknowledge clock pulse on the SCL line on the following situations: (1) A data byte has
been received while the I
2
C interface is in the Master Receiver Mode. (2) A data byte has
been received while the I2C interface is in the addressed Slave Receiver Mode.
I2CON.1-Reserved for future use. Should not be set to 1 by user programs.
I2CON.0CRSELSCL clock selection. When set = 1, Timer1 overflow generates SCL, when cleared = 0,
the internal SCL generator is used base on values of I2SCLH and I2SCLL.
C-bus if it is in master mode, and transmits a
C interface is already in master mode and some data
2
C interface is in an addressed slave mode.
2
C-bus.
2
Figure 4: I
2003 Nov 6 65
C Control register
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
I2C INTERFACE
I2C Status register
This is a read-only register. It contains the status code of I2C interface. The least three bits are always 0. There are 26 possible
status codes. When the code is F8H, there is no relevant information available and SI bit is not set. All other 25 status codes
correspond to defined I
I2STAT
Address: D9h
Not bit addressable
Reset Source(s): Any reset
Reset Value: 11111000B
BITSYMBOLFUNCTION
I2STAT7, 3STA.4, 0I
I2STAT2, 0-These three bits are not used and always set to 0.
2
C states. When any of these states entered, the SI bit will be set. Refer to Table 2 to Table 5 for details.
76543210
STA.4 STA.3 STA.2STA.1 STA.0000
2
C the status code.
2
Figure 5: I
C Status register
I2C SCL Duty Cycle registers I2SCLH and I2SCLL
When the internal SCL generator is selected for the I2C interface by setting CRSEL = 0 in the I2CON register, the user must set
values for registers I2SCLL and I2SCLH to select the data rate. I2SCLH defines the number of PCLK cycles for SCL = high,
I2SCLL defines the number of PCLK cycles for SCL = low. The frequency is determined by the following formula:
Bit Frequency = f
Where f
is the frequency of PCLK.
PCLK
The values for I2SCLL and I2SCLH do not have to be the same; the user can give different duty cycle’s for SCL by setting these
two registers. However, the value of the register must ensure that the data rate is in the I
the values of I2SCLL and I2SCLH have some restrictions and values for both registers greater than 3 PCLKs are recommended.
2
Table 1: I
C clock rates selection
I2SCLL
+
CRSEL7.373 MHz3.6865 MHz1.8433 MHz12 MHz6 MHz
I2SCLH
60-307154--
70-263132--
80-230115-375
90-205102-333
10036918492-300
15024612361400200
2501477437240120
3001236131200100
50074371812060
60061311510050
/ (2*(I2SCLH + I2SCLL))
PCLK
Bit data rate (Kbit/sec) at f
OSC
2
C data rate range of 0 - 400 kHz. Thus
2003 Nov 6 66
Philips Semiconductors
I2C INTERFACE
2
Table 1: I
C clock rates selection
User’s Manual - Preliminary -
P89LPC920/921/922
Bit data rate (Kbit/sec) at f
I2SCLL
+
CRSEL7.373 MHz3.6865 MHz1.8433 MHz12 MHz6 MHz
I2SCLH
1000371896030
1500251264020
200018953015
-1
3.6 - 922 Kbps
timer1 in mode2
1.8 - 461 Kbps
timer1 in mode 2
0.9 - 230 Kbps
timer1 in mode 2
OSC
5.86 - 1500 Kbps
timer1 in mode 2
2.93 - 750 Kbps
timer1 in mode 2
I2C operation mode
Master Transmitter Mode
In this mode data is transmitted from master to slave. Before the Master Transmitter Mode can be entered, I2CON must be
initialized as follows:
76543210
I2CON (D8h)
-I2ENSTASTOSIAA - CRSEL
-1000x -bit rate
2
Figure 10-6: I
CRSEL defines the bit rate. I2EN must be set to 1 to enable the I2C function. If the AA bit is 0, it will not acknowledge its own
slave address or the general call address in the event of another device becoming master of the bus and it can not enter slave
mode. STA, STO, and SI bits must be cleared to 0.
The first byte transmitted contains the slave address of the receiving device (7 bits) and the data direction bit. In this case, the
data direction bit (R/W) will be logic 0 indicating a write. Data is transmitted 8 bits at a time. After each byte is transmitted, an
acknowledge bit is received. START and STOP conditions are output to indicate the beginning and the end of a serial transfer.
2
C will enter Master Transmitter Mode by setting the STA bit. The I2C logic will send the START condition as soon as the
The I
bus is free. After the START condition is transmitted, the SI bit is set, and the status code in I2STAT should be 08h. This status
code must be used to vector to an interrupt service routine where the user should load the slave address to I2DAT (Data Register)
and data direction bit (SLA+W). The SI bit must be cleared before the data transfer can continue.
When the slave address and R/W bit have been transmitted and an acknowledgment bit has been received, the SI bit is set again,
and the possible status codes are 18h, 20h, or 38h for the master mode or 68h, 78h, or 0B0h if the slave mode was enabled
(setting AA = Logic 1). The appropriate action to be taken for each of these status codes is shown in Table 2.
C Control register
2003 Nov 6 67
Philips Semiconductors
I2C INTERFACE
S Slave Address R/W A DATA A DATA A/A P
User’s Manual - Preliminary -
P89LPC920/921/922
“0” - Write
“1” - Read
From Master to Slave
From Slave to Master
Figure 7: Format in the Master Transmitter Mode
Data Transferred
(n Bytes + Acknowledge
A = Acknowledge (SDA low)
A = Not Acknowledge (SDA high)
S = START condition
P = STOP Condition
Master Receiver Mode
In the Master Receiver Mode, data is received from a slave transmitter. The transfer started in the same manner as in the Master
Transmitter Mode. When the START condition has been transmitted, the interrupt service routine must load the slave address
and the data direction bit to I
When the slave address and data direction bit have been transmitted and an acknowledge bit has been received, the SI bit is
set, and the Status Register will show the status code. For master mode, the possible status codes are 40H, 48H, or 38H. For
slave mode, the possible status codes are 68H, 78H, or B0H. Refer to Table 3 for details.
S Slave Address R A DATA A DATA A P
From Master to Slave
From Slave to Master
2
C Data Register (I2DAT). The SI bit must be cleared before the data transfer can continue.
“0” - Write
“1” - Read
Data Transferred
(n Bytes + Acknowledge
A = Acknowledge (SDA low)
A = Not Acknowledge (SDA high)
S = START condition
Figure 8: Format of Master Receiver Mode
2
After a repeated START condition, I
C may switch to the Master Transmitter Mode.
S RA SLAAARS WAA P DATADATA SLA DATA
Data Transferred
(n Bytes +
Acknowledge
A = Acknowledge (SDA low)
= Not Acknowledge (SDA high)
A
From Master to Slave
From Slave to Master
Figure 9: A Master Receiver switches to Master Transmitter after sending Repeated Start
2003 Nov 6 68
S = START condition
P = STOP Condition
SLA = Slave Address
RS = Repeat START condition
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
I2C INTERFACE
Slave Receiver Mode
In the Slave Receiver Mode, data bytes are received from a master transmitter. To initialize the Slave Receiver Mode, the user
should write the slave address to the Slave Address Register (I2ADR) and the I2C Control Register (I2CON) should be configured
as follows:
76543210
I2CON (D8h)
CRSEL is not used for slave mode. I2EN must be set = 1 to enable I
slave address or the general call address. STA, STO and SI are cleared to 0.
After I2ADR and I2CON are initialized, the interface waits until it is addressed by its own address or general address followed by
the data direction bit which is 0(W). If the direction bit is 1(R), it will enter Slave Transmitter Mode. After the address and the
direction bit have been received, the SI bit is set and a valid status code can be read from the Status Register(I2STAT). Refer to
Table 4 for the status codes and actions.
S Slave Address W A DATA A DATA A /A P/RS
From Master to Slave
From Slave to Master
-I2ENSTASTOSIAA- CRSEL
-10001 --
Figure 10-10: I2C Control register
2
C function. AA bit must be set = 1 to acknowledge its own
“0” - Write
“1” - Read
Data Transferred
(n Bytes + Acknowledge
A = Acknowledge (SDA low)
A = Not Acknowledge (SDA high)
S = START condition
P = STOP Condition
RS = Repeated START condition
Figure 11: Format of Slave Receiver Mode
Slave Transmitter Mode
The first byte is received and handled as in the Slave Receiver Mode. However, in this mode, the direction bit will indicate that
the transfer direction is reversed. Serial data is transmitted via P1.3/SDA while the serial clock is input through P1.2/SCL. START
and STOP conditions are recognized as the beginning and end of a serial transfer. In a given application, I
master and as a slave. In the slave mode, the I2C hardware looks for its own slave address and the general call address. If one
of these addresses is detected, an interrupt is requested. When the microcontrollers wishes to become the bus master, the
hardware waits until the bus is free before the master mode is entered so that a possible slave action is not interrupted. If bus
arbitration is lost in the master mode, I
same serial transfer.
2003 Nov 6 69
2
C switches to the slave mode immediately and can detect its own slave address in the
2
C may operate as a
Philips Semiconductors
I2C INTERFACE
S Slave Address R A DATA A DATA A P
User’s Manual - Preliminary -
P89LPC920/921/922
“0” - Write
“1” - Read
Data Transferred
(n Bytes + Acknowledge
A = Acknowledge (SDA low)
From Master to Slave
From Slave to Master
A = Not Acknowledge (SDA high)
S = START condition
P = STOP Condition
Figure 12: Format of Slave Transmitter Mode
2003 Nov 6 70
Philips Semiconductors
I2C INTERFACE
User’s Manual - Preliminary -
P89LPC920/921/922
8
P1.3/SDA
P1.2/SCL
P1.3
Input
Filter
Output
Stage
Input
Filter
Output
Stage
P1.2
Timer 1
Overflow
I2CON
I2SCLH
I2SCLL
Address Register
Comparator
Shift Register
8
Bit Counter /
Arbitration &
Sync Logic
Serial Clock
Generator
Control Register & SCL Duty
Cycle Registers
Timing
&
Control
Logic
I2ADR
ACK
I2DAT
CCLK
Interrupt
Internal Bus
Status Bus
I2STAT
Figure 13: I2C-bus serial interface block diagram
2003 Nov 6 71
8
Status
Decoder
Status Register
8
Philips Semiconductors
I2C INTERFACE
Table 2: Master Transmitter Mode
User’s Manual - Preliminary -
P89LPC920/921/922
Status
code
(I2STAT)
08H
10H
18h
20h
28h
Status of the
2
I
C-bus hardware
A START condition
has been
transmitted
A repeat START
condition has been
transmitted
SLA+W has been
transmitted; ACK
has been received
SLA+W has been
transmitted;NOTACK has been
received
Data byte in I2DAT
has been
transmitted; ACK
has been received
Application software response
Next action taken by I2C
to/from I2DATto I2CON
STASTOSIAA
SLA+W will be transmitted; ACK bit will be
Load SLA+Wx00x
Load SLA+W or
Load SLA+R
Load data byte or000x
no I2DAT action or100xRepeated START will be transmitted;
no I2DAT action or
no I2DAT action
Load data byte or000x
no I2DAT action or100xRepeated START will be transmitted;
no I2DAT action or010x
no I2DAT action 110x
Load data byte or000x
no I2DAT action or100xRepeated START will be transmitted;
no I2DAT action or010x
no I2DAT action 110x
x00x
010x
110x
received
As above;SLA+W will be transmitted;
2
C switches to Master Receiver Mode
I
Data byte will be transmitted; ACK bit will be
received
STOP condition will be transmitted;
STO flag will be reset
STOP condition followed by a START
condition will be transmitted; STO flag will be
reset.
Data byte will be transmitted;ACK bit will be
received
STOP condition will be transmitted; STO flag
will be reset
STOP condition followed by a START
condition will be transmitted; STO flag will be
reset
Data byte will be transmitted;
ACK bit will be received
STOP condition will be transmitted; STO flag
will be reset
STOP condition followed by a START
condition will be transmitted; STO flag will be
reset
hardware
2003 Nov 6 72
Philips Semiconductors
I2C INTERFACE
Table 2: Master Transmitter Mode(Continued)
User’s Manual - Preliminary -
P89LPC920/921/922
Status
code
(I2STAT)
Status of the
2
C-bus hardware
I
Data byte in I2DAT
has been
transmitted,NOT
30h
ACK hasbeen
received
Arbitration lost in
38H
SLA+R/W or data
bytes
Table 3: Master Receiver Mode
Status
code
(I2STAT)
Status of the
2
C-bus hardware
I
Application software response
Next action taken by I2C
to/from I2DATto I2CON
hardware
STASTOSIAA
Load data byte or000x
Data byte will be transmitted;
ACK bit will be received
no I2DAT action or100xRepeated START will be transmitted;
no I2DAT action or
010x
STOP condition will be transmitted; STO flag
will be reset
STOP condition followed by a START
no I2DAT action110x
condition will be transmitted. STO flag will be
reset.
No I2DAT action
or
000x
No I2DAT action 100x
2
I
C-bus will be released; not addressed
slave will be entered
A START condition will be transmitted when
the bus becomes free.
Application software response
Next action taken by I
to/from I2DATto I2CON
hardware
2
C
08H
10H
38H
40h
A START condition
has been
transmitted
A repeat
STARTcondition
has been
transmitted
Arbitration lost in
NOT ACK bit
SLA+R has been
transmitted;ACK
has been received
STASTOSIAA
SLA+R will be transmitted;
Load SLA+Rx00x
Load SLA+R or
X00xAs above
Load SLA+Wx00x
ACK bit will be received
SLA+W will be transmitted;
2
I
Transmitter Mode
no I2DAT action or
000xI
2
slave mode
no I2DAT action
100x
A START condition will be
transmitted when the bus becomes
free
no I2DAT action or0000
no I2DAT action or0001
Data byte will be received;
NOT ACK bit will be returned
Data byte will be received; ACK bit
will be returned
C will be switches to Master
C will be released; it will enter a
2003 Nov 6 73
Philips Semiconductors
I2C INTERFACE
Table 3: Master Receiver Mode(Continued)
User’s Manual - Preliminary -
P89LPC920/921/922
Status
code
(I2STAT)
48h
50h
58h
Status of the
2
C-bus hardware
I
SLA+R has been
transmitted; NOT
ACK has been
received
Data byte has been
received; ACK has
been returned
Data byte has been
received; NACK has
been returned
Application software response
Next action taken by I2C
to/from I2DATto I2CON
STASTOSIAA
No I2DAT action or
no I2DAT action or010x
no I2DAT action or110x
Read data byte
read data byte0001
Read data byte or100xRepeated START will be transmitted;
read data byte or010x
read data byte110x
100x
0000
Repeated START will be transmitted;
STOP condition will be transmitted;
STO flag will be reset
STOP condition followed by a
START condition will be transmitted;
STO flag will be reset
Data byte will be received; NOT ACK
bit will be returned
Data byte will be received; ACK bit
will be returned
STOP condition will be transmitted;
STO flag will be reset
STOP condition followed by a
START condition will be transmitted;
STO flag will be reset
hardware
Table 4: Slave Receiver Mode
Status
code
(I2STAT)
60H
68H
70H
Status of the
2
C-bus hardware
I
Own SLA+W has
been received; ACK
has been received
Arbitration lost in
SLA+R/Was
master;Own SLA+W
has been received,
ACK returned
General call
address(00H) has
beenreceived, ACK
has been returned
Application software response
to/from I2DATto I2CON
STASTOSIAA
no I2DAT action or
no I2DAT action x001
No I2DAT action or
no I2DAT action x001
No I2DAT action or
no I2DAT action x001
x000
x000
x000
Next action taken by I2C
hardware
Data byte will be received and NOT ACK will
be returned
Data byte will be received and ACK will be
returned
Data byte will be received and NOT ACK will
be returned
Data byte will be received and ACK will be
returned
Data byte will be received and NOT ACK will
be returned
Data byte will be received and ACK will be
returned
2003 Nov 6 74
Philips Semiconductors
I2C INTERFACE
Table 4: Slave Receiver Mode(Continued)
User’s Manual - Preliminary -
P89LPC920/921/922
Status
code
(I2STAT)
78H
80H
88H
90H
Status of the
2
C-bus hardware
I
Arbitration lost in
SLA+R/W as
master; General call
addresshas been
received, ACK bit
has been returned
Previously
addressed with own
SLA address; Data
has been received;
ACK has been
returned
Previously
addressed with own
SLA address; Data
has been received;
NACK has been
returned
Previously
addressed with
General call;
Datahas been
received; ACK has
been returned
Application software response
to/from I2DATto I2CON
STASTOSIAA
no I2DAT action or
no I2DAT action x001
Read data byte or
read data bytex001
Read data byte or
read data byte
or
read data byte
or
read data byte1001
Read data byte orx000
read data bytex001
x000
x000
0000
0001
1000
Next action taken by I2C
hardware
Data byte will be received and NOT ACK will
be returned
Data byte will be received and ACK will be
returned
Data byte will be received and NOT ACK will
be returned
Data byte will be received; ACK bit will be
returned
Switched to not addressed SLA mode; no
recognition of own SLA or general address
Switched to not addressed SLA mode; Own
SLA will be recognized; general call address
will be recognized if I2ADR.0=1
Switched to not addressed SLA mode; no
recognition of own SLA or General call
address. A START condition will be
transmitted when the bus becomes free
Switched to not addressed SLA mode; Own
slave address will be recognized; General
call address will be recognized if I2ADR.0=1.
A START condition will be transmitted when
the bus becomes free.
Data byte will be received and NOT ACK will
be returned
Data byte will be received and ACK will be
returned
2003 Nov 6 75
Philips Semiconductors
I2C INTERFACE
Table 4: Slave Receiver Mode(Continued)
User’s Manual - Preliminary -
P89LPC920/921/922
Status
code
(I2STAT)
98H
A0H
Status of the
2
C-bus hardware
I
Previously
addressed with
General call; Data
has been received;
NACK has been
returned
A STOP condition or
repeated START
condition has been
received while still
addressed as SLA/
REC or SLA/TRX
Application software response
to/from I2DATto I2CON
STASTOSIAA
Read data byte
0000
Switched to not addressed SLA mode; no
recognition of own SLA or General call
address
Switched to not addressed SLA mode; Own
read data byte
0001
slave address will be recognized; General
call address will be recognized if I2ADR.0=1.
Switched to not addressed SLA mode; no
recognition of own SLA or General call
read data byte
1000
address. A START condition will be
transmitted when the bus becomes free.
Switched to not addressed SLA mode; Own
slave address will be recognized; General
read data byte
1001
call address will be recognized if I2ADR.0=1.
A START condition will be transmitted when
the bus becomes free.
Switched to not addressed SLA mode; no
No I2DAT action0000
recognition of own SLA or General call
address
no I2DAT action
0001
Switched to not addressed SLA mode; Own
slave address will be recognized; General
call address will be recognized if I2ADR.0=1.
Switched to not addressed SLA mode; no
no I2DAT action
1000
recognition of own SLA or General call
address. A START condition will be
transmitted when the bus becomes free.
Switched to not addressed SLA mode; Own
slave address will be recognized; General
no I2DAT action
1001
call address will be recognized if I2ADR.0=1.
A START condition will be transmitted when
the bus becomes free.
Next action taken by I2C
hardware
2003 Nov 6 76
Philips Semiconductors
I2C INTERFACE
Table 5: Slave Transmitter Mode
User’s Manual - Preliminary -
P89LPC920/921/922
Status
code
(I2STAT)
A8h
B0h
B8H
C0H
Status of the
I2C-bus hardware
Own SLA+R has
been received; ACK
has been returned
Arbitration lost in
SLA+R/W as
master; Own SLA+R
has been received,
ACK has been
returned
Data byte in I2DAT
has been
transmitted; ACK
has been received
Data byte in I2DAT
has been
transmitted; NACK
has been received
Application software response
to/from I2DATto I2CON
STASTOSIAA
Load data byte orx000
load data bytex001
Load data byte orx000
load data bytex001
Load data byte or
load data byte
No I2DAT action or
no I2DAT action or0001
no I2DAT action or1000
x000
x001
0000
Last data byte will be transmitted and ACK
bit will be received
Data byte will be transmitted; ACK will be
received
Last data byte will be transmitted and ACK
bit will be received
Data byte will be transmitted;
ACK bit will be received
Last data byte will be transmitted and ACK
bit will be received
Data byte will be transmitted; ACK will be
received
Switched to not addressed SLA mode; no
recognition of own SLA or General call
address.
Switched to not addressed SLA mode; Own
slave address will be recognized; General
call address will be recognized if I2ADR.0=1.
Switched to not addressed SLA mode; no
recognition of own SLA or General call
address. A START condition will be
transmitted when the bus becomes free.
Next action taken by I
hardware
2
C
no I2DAT action 1001
2003 Nov 6 77
Switched to not addressed SLA mode; Own
slave address will be recognized; General
call address will be recognized if I2ADR.0=1.
A START condition will be transmitted when
the bus becomes free.
Philips Semiconductors
I2C INTERFACE
Table 5: Slave Transmitter Mode(Continued)
User’s Manual - Preliminary -
P89LPC920/921/922
Status
code
(I2STAT)
C8H
For more information about the I2C interface, please refer to the I2C specification.
Status of the
2
C-bus hardware
I
Last data byte in
I2DAT has been
transmitted(AA=0);
ACK has been
received
Application software response
to/from I2DATto I2CON
STASTOSIAA
No I2DAT action or
no I2DAT action or0001
no I2DAT action or1000
no I2DAT action 1001
0000
Next action taken by I2C
hardware
Switched to not addressed SLA mode; no
recognition of own SLA or General call
address.
Switched to not addressed SLA mode; Own
slave address will be recognized; General
call address will be recognized if I2ADR.0=1.
Switched to not addressed SLA mode; no
recognition of own SLA or General call
address. A START condition will be
transmitted when the bus becomes free.
Switched to not addressed SLA mode; Own
slave address will be recognized; General
call address will be recognized if I2ADR.0=1.
A START condition will be transmitted when
the bus becomes free.
2003 Nov 6 78
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
ANALOG COMPARATORS
11. ANALOG COMPARATORS
Two analog comparators are provided on the P89LPC920/921/922. Input and output options allow use of the comparators in a
number of different configurations. Comparator operation is such that the output is a logical one (which may be read in a register
and/or routed to a pin) when the positive input (one of two selectable pins) is greater than the negative input (selectable from a
pin or an internal reference voltage). Otherwise the output is a zero. Each comparator may be configured to cause an interrupt
when the output value changes.
Comparator configuration
Each comparator has a control register, CMP1 for comparator 1 and CMP2 for comparator 2. The control registers are identical
and are shown in Figure 11-1.
The overall connections to both comparators are shown in Figure 11-2. There are eight possible configurations for each
comparator, as determined by the control bits in the corresponding CMPn register: CPn, CNn, and OEn. These configurations
are shown in Figure 11-3.
When each comparator is first enabled, the comparator output and interrupt flag are not guaranteed to be stable for 10
microseconds. The corresponding comparator interrupt should not be enabled during that time, and the comparator interrupt flag
must be cleared before the interrupt is enabled in order to prevent an immediate interrupt service.
CMPn
Address: ACh (CMP1), ADh (CMP2)
Not bit addressable
Reset Source(s): Any reset
Reset Value: xx000000B
BITSYMBOLFUNCTION
CMPn.7, 6-Reserved for future use. Should not be set to 1 by user programs.
CMPn.5CEnComparator enable. When set, the corresponding comparator function is enabled.
Comparator output is stable 10 microseconds after CEn is set.
CMPn.4CPnComparator positive input select. When 0, CINnA is selected as the positive comparator
input. When 1, CINnB is selected as the positive comparator input.
CMPn.3CNnComparator negative input select. When 0, the comparator reference pin CMPREF is
selected as the negative comparator input. When 1, the internal comparator reference,
Vref, is selected as the negative comparator input.
CMPn.2OEnOutput enable. When 1, the comparator output is connected to the CMPn pin if the
comparator is enabled (CEn = 1). This output is asynchronous to the CPU clock.
CMPn.1COnComparator output, synchronized to the CPU clock to allow reading by software.
CMPn.0CMFnComparator interrupt flag. This bit is set by hardware whenever the comparator output
COn changes state. This bit will cause a hardware interrupt if enabled. Cleared by
software.
Figure 11-1: Comparator control registers (CMP1 and CMP2)
76543210
--CEnCPnCNnOEnCOnCMFn
2003 Nov 6 79
Philips Semiconductors
ANALOG COMPARATORS
User’s Manual - Preliminary -
P89LPC920/921/922
(P0.4) CIN1A
(P0.3) CIN1B
(P0.5) CMPREF
(P0.2) CIN2A
(P0.1) CIN2B
Vref
CP1
Comparator 1
OE1
+
CO1
-
Change Detect
CN1
CP2
Change Detect
Comparator 2
CMF1
CMF2
+
CO2
-
OE2
CN2
Figure 11-2: Comparator input and output connections
CMP1 (P0.6)
Interrupt
EC
CMP2 (P0.0)
Internal reference voltage
An internal reference voltage, Vref, may supply a default reference when a single comparator input pin is used. Please refer to
the Datasheet for specifications.
Comparator interrupt
Each comparator has an interrupt flag CMFn contained in its configuration register. This flag is set whenever the comparator
output changes state. The flag may be polled by software or may be used to generate an interrupt. The two comparators use one
common interrupt vector. The interrupt will be generated when the interrupt enable bit EC in the IEN1 register is set and the
interrupt system is enabled via the EA bit in the IEN0 register. If both comparators enable interrupts, after entering the interrupt
service routine, the user will need to read the flags to determine which comparator caused the interrupt.
When a comparator is disabled the comparator’s output, COx, goes high. If the comparator output was low and then is disabled,
the resulting transition of the comparator output from a low to high state will set the the comparator flag, CMFx. This will cause
an interrupt if the comparator interrupt is enabled. The user should therefore disable the comparator interrupt prior to disabling
the comparator. Additionally, the user should clear the comparator flag, CMFx, after disabling the comparator.
Comparators and power reduction modes
Either or both comparators may remain enabled when Power down or Idle mode is activated, but both comparators are disabled
automatically in Total Power down mode.
If a comparator interrupt is enabled (except in Total Power down mode), a change of the comparator output state will generate
an interrupt and wake up the processor. If the comparator output to a pin is enabled, the pin should be configured in the pushpull mode in order to obtain fast switching times while in Power down mode. The reason is that with the oscillator stopped, the
temporary strong pull-up that normally occurs during switching on a quasi-bidirectional port pin does not take place.
2003 Nov 6 80
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
ANALOG COMPARATORS
Comparators consume power in Power down and Idle modes, as well as in the normal operating mode. This should be taken into
consideration when system power consumption is an issue. To minimize power consumption, the user can Power down the
comparators by disabling the comparators and setting PCONA.5 to ‘1’, or simply putting the device in Total Power down mode.
CPn, CNn, OEn = 0 0 0
CINnA
CMPREF
CPn, CNn, OEn = 0 1 0
CINnA
Vref (1.23V)
CPn, CNn, OEn = 1 0 0
CINnB
CMPREF
CPn, CNn, OEn = 1 1 0
CINnB
Vref (1.23V)
CPn, CNn, OEn = 0 0 1
+
COn
-
+
COn
-
+
COn
-
+
COn
-
CINnA
CMPREF
CPn, CNn, OEn = 0 1 1
CINnA
Vref (1.23V)
CPn, CNn, OEn = 1 0 1
CINnB
CMPREF
CPn, CNn, OEn = 1 1 1
CINnB
Vref (1.23V)
+
COn
CMPn
-
+
COn
CMPn
-
+
COn
CMPn
-
+
COn
CMPn
-
Figure 11-3: Comparator configurations
Comparator configuration example
The code shown below is an example of initializing one comparator. Comparator 1 is configured to use the CIN1A and CMPREF
inputs, outputs the comparator result to the CMP1 pin, and generates an interrupt when the comparator output changes.
CMPINIT:
MOVPT0AD,#030h; Disable digital INPUTS on pins that are used for analog functions: CIN1A, CMPREF.
ANLP0M2,#0CFh; Disable digital OUTPUTS on pins that are used
ORLP0M1,#030h; for analog functions: CIN1A, CMPREF.
MOVCMP1,#024h; Turn on comparator 1 and set up for:
; - Positive input on CIN1A.
; - Negative input from CMPREF pin.
; - Output to CMP1 pin enabled.
CALLdelay10us; The comparator has to start up for at least 10 microseconds before use.
ANLCMP1,#0FEh; Clear comparator 1 interrupt flag.
SETBEC; Enable the comparator interrupt. The priority is left at the current value.
SETBEA; Enable the interrupt system (if needed).
2003 Nov 6 81
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
ANALOG COMPARATORS
RET; Return to caller.
The interrupt routine used for the comparator must clear the interrupt flag (CMF1 in this case) before returning.
2003 Nov 6 82
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
KEYPAD INTERRUPT (KBI)
12. KEYPAD INTERRUPT (KBI)
The Keypad Interrupt function is intended primarily to allow a single interrupt to be generated when Port 0 is equal to or not equal
to a certain pattern. This function can be used for bus address recognition or keypad recognition. The user can configure the port
via SFRs for different tasks.
There are three SFRs used for this function. The Keypad Interrupt Mask Register (KBMASK) is used to define which input pins
connected to Port 0 are enabled to trigger the interrupt. The Keypad Pattern Register (KBPATN) is used to define a pattern that
is compared to the value of Port 0. The Keypad Interrupt Flag (KBIF) in the Keypad Interrupt Control Register (KBCON) is set
when the condition is matched while the Keypad Interrupt function is active. An interrupt will be generated if it has been enabled
by setting the EKBI bit in IEN1 register and EA = 1. The PATN_SEL bit in the Keypad Interrupt Control Register (KBCON) is used
to define equal or not-equal for the comparison.
In order to use the Keypad Interrupt as an original KBI function like in the 87LPC76x series, the user needs to set KBPATN =
0FFH and PATN_SEL = 0 (not equal), then any key connected to Port0 which is enabled by KBMASK register is will cause the
hardware to set KBIF = 1 and generate an interrupt if it has been enabled. The interrupt may be used to wake up the CPU from
Idle or Power down modes. This feature is particularly useful in handheld, battery powered systems that need to carefully manage
power consumption yet also need to be convenient to use.
In order to set the flag and and cause an interrupt, the pattern on Port 0 must be held longer than 6 CCLKs.
KBPATN
Address: 93h
Not bit addressable
Reset Source(s): Any reset
Reset Value: 11111111B
BITSYMBOLFUNCTION
KBPATN.7-0-Pattern bit 7 - bit 0
KBCON
Address: 94h
Not bit addressable
Reset Source(s): Any reset
Reset Value: xxxxxx00B
BITSYMBOLFUNCTION
KBCON.7-2-Reserved
KBCON.1PATN_SELPattern Matching Polarity selection. When set, Port 0 has to be equal to the user-defined
KBCON.0KBIFKeypad Interrupt Flag. Set when Port 0 matches user defined conditions specified in
The watchdog timer subsystem protects the system from incorrect code execution by causing a system reset when it underflows
as a result of a failure of software to feed the timer prior to the timer reaching its terminal count. The watchdog timer can only be
reset by a power-on reset.
Watchdog Function
The user has the ability using the WDCON and UCFG1 registers to control the run /stop condition of the WDT, the clock source
for the WDT, the prescaler value, and whether the WDT is enabled to reset the device on underflow. In addition, there is a safety
mechanism which forces the WDT to be enabled by values programmed into UCFG1 either through IAP or a commercial
programmer.
The WDTE bit (UCFG1.7), if set, enables the WDT to reset the device on underflow. Following reset, the WDT will be running
regardless of the state of the WDTE bit.
The WDRUN bit (WDCON.2) can be set to start the WDT and cleared to stop the WDT. Following reset this bit will be set and
the WDT will be running. All writes to WDCON need to be followed by a feed sequence (see section "Feed Sequence" on page
86). Additional bits in WDCON allow the user to select the clocksource for the WDT and the prescaler.
When the timer is not enabled to reset the device on underflow, the WDT can be used in "timer mode" and be enabled to produce
an interrupt (IEN0.6) if desired.
The Watchdog Safety Enable bit, WDSE (UCFG1.4) along with WDTE, is designed to force certain operating conditions at powerup. Refer to the Table 13-1 for details
Table 13-1: Watchdog timer configuration.
WDTE
(UCFG1.7)
0x
WDSE
(UCFG1.4)
FUNCTION
The watchdog reset is disabled. The timer can be used as an internal timer and
can be used to generate an interrupt. WDSE has no effect.
10
11
Figure 13-3 shows the watchdog timer in watchdog mode. It consists of a programmable 13-bit prescaler, and an 8-bit down
counter. The down counter is clocked (decremented) by a tap taken from the prescaler. The clock source for the prescaler is
either PCLK or the watchdog oscillator selected by the WDCLK bit in the WDCON register. (Note that switching of the clock
sources will not take effect immediately - see section "Watchdog Clock Source" on page 88).
The watchdog asserts the watchdog reset when the watchdog count underflows and the watchdog reset is enabled. When the
watchdog reset is enabled, writing to WDL or WDCON must be followed by a feed sequence for the new values to take effect.
If a watchdog reset occurs, the internal reset is active for at least one watchdog clock cycle (PCLK or the watchdog oscillator
clock). If CCLK is still running, code execution will begin immediately after the reset cycle. If the processor was in Power down
mode, the watchdog reset will start the oscillator and code execution will resume after the oscillator is stable.
The watchdog reset is enabled. The user can set WDCLK to choose the clock
source.
The watchdog reset is enabled, along with additional safety features:
1. WDCLK is forced to 1 (using watchdog oscillator)
2. WDCON and WDL register can only be written once
3. WDRUN is forced to 1
2003 Nov 6 85
Philips Semiconductors
WATCHDOG TIMER
Watchdog
Oscillator
PCLK
WDCLK after a
watchdog feed
sequence
PRE2
PRE1
PRE0
Feed Sequence
DECODE
÷32
÷32÷64÷128÷256÷512÷1024÷2048÷4096
000
001
010
011
100
101
110
111
÷2÷2÷2÷2÷2÷2÷2
Figure 13-1: Watchdog Prescaler
User’s Manual - Preliminary -
P89LPC920/921/922
TO
WATCHDOG
DOWN
COUNTER
(after one
prescaler
count delay
The watchdog timer control register and the 8-bit down counter (Figure 13-3) are not directly loaded by the user. The user writes
to the WDCON and the WDL SFRs. At the end of a feed sequence, the values in the WDCON and WDL SFRs are loaded to the
control register and the 8-bit down counter. Before the feed sequence, any new values written to these two SFRs will not take
effect. To avoid a watchdog reset, the watchdog timer needs to be fed (via a special sequence of software action called the feed
sequence) prior to reaching an underflow.
To feed the watchdog, two write instructions must be sequentially executed successfully. Between the two write instructions, SFR
reads are allowed, but writes are not allowed. The instructions should move A5H to the WFEED1 register and then 5AH to the
WFEED2 register. An incorrect feed sequence will cause an immediate watchdog reset. The program sequence to feed the
watchdog timer is as follows:
CLREA; disable interrupt
MOVWFEED1,#0A5h ; do watchdog feed part 1
MOVWFEED2,#05Ah ; do watchdog feed part 2
SETBEA; enable interrupt
This sequence assumes that the P89LPC901/902/903 interrupt system is enabled and there is a possibility of an interrupt request
occuring during the feed sequence. If an interrupt was allowed to be serviced and the service routine contained any SFR writes,
it would trigger a watchdog reset. If it is known that no interrupt could occur during the feed sequence, the instructions to disable
and re-enable interrupts may be removed.
In watchdog mode (WDTE = 1), writing the WDCON register must be IMMEDIATELY followed by a feed sequence to load the
WDL to the 8-bit down counter, and the WDCON to the shadow register. If writing to the WDCON register is not immediately
followed by the feed sequence, a watchdog reset will occur.
For example: setting WDRUN = 1:
MOVACC,WDCON; get WDCON
SETBACC.2; set WD_RUN=1
MOVWDL,#0FFh; New count to be loaded to 8-bit down counter
CLREA; disable interrupt
MOVWDCON,ACC; write back to WDCON (after the watchdog is enabled, a feed must occur
; immediately)
2003 Nov 6 86
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
WATCHDOG TIMER
MOVWFEED1,#0A5h ; do watchdog feed part 1
MOVWFEED2,#05Ah ; do watchdog feed part 2
SETBEA; enable interrupt
In timer mode (WDTE = 0), WDCON is loaded to the control register every CCLK cycle (no feed sequence is required to load the
control register), but a feed sequence is required to load from the WDL SFR to the 8-bit down counter before a time-out occurs.
WDCON
Address: A7h
Not bit addressable
Reset Source(s): See reset value below
Reset Value:111xx1?1B(Note: WDCON.7,6,5,2,0 - set to ’1’ any reset; WDCON.1 - cleared to ’0’ on Power-on
BITSYMBOLFUNCTION
WDCON.7-5PRE2-PRE0Clock Prescaler Tap Select. Refer to Table 13-2 for details.
WDCON.4-3-Reserved for future use. Should not be set to 1 by user program.
WDCON.2WDRUNWatchdog Run Control. The watchdog timer is started when WDRUN = 1 and stopped
WDCON.1WDTOFWatchdog Timer Time-Out Flag. This bit is set when the 8-bit down counter underflows.
WDCON.0WDCLKWatchdog input clock select. When set, the watchdog oscillator is selected. When cleared,
76543210
PRE2PRE1PRE0--WDRUN WDTOF WDCLK
reset, set to ’1’ on watchdog reset, not affected by any other reset)
when WDRUN = 0. This bit is forced to 1 (watchdog running) and cannot be cleared to
zero if both WDTE and WDSE are set to 1.
In watchdog mode, a feed sequence will clear this bit. It can also be cleared by writing ’0’
to this bit in software.
PCLK is selected. (If the CPU is powered down, the watchdog is disabled if WDCLK = 0,
see section "Power down operation"). (Note: If both WDTE and WDSE are set to 1, this
bit is forced to 1.) Refer to section "Watchdog Clock Source" on page 88 for details.
Figure 13-2: Watchdog Timer Control Register
The number of watchdog clocks before timing out is calculated by the following equations:
(5+PRE)
tclks = (2
where:
• PRE is the value of prescaler (PRE2-PRE0) which can be the range 0-7, and;
• WDL is the value of watchdog load register which can be the range of 0-255.
The minimum number of tclks is:
tclks = (2
The maximum number of tclks is:
tclks = (2
The following table show sample P89LPC920/921/922 timeout values
The watchdog timer system has an on-chip 400KHz oscillator. The watchdog timer can be clocked from either the watchdog oscillator or
from PCLK (refer to Figure 13-1) by configuring the WDCLK bit in the Watchdog Control Register WDCON. When the watchdog feature is
enabled, the timer must be fed regularly by software in order to prevent it from resetting the CPU.
After changing WDCLK (WDCON.0), switching of the clock source will not immediately take effect. As shown in Figure 13-3, the selection
is loaded after a watchdog feed sequence. In addition, due to clock synchronization logic, it can take two old clock cycles before the old clock
source is deselected, and then an additional two new clock cycles before the new clock source is selected.
Since the prescaler starts counting immediately after a feed, switching clocks can cause some inaccuracy in the prescaler count. The
inaccuracy could be as much as 2 old clock source counts plus 2 new clock cycles.
Note: When switching clocks, it is important that the old clock source is left enabled for 2 clock cycles after the feed completes. Otherwise,
the watchdog may become disabled when the old clock source is disabled. For example, suppose PCLK (WCLK=0) is the current clock
source. After WCLK is set to ’1’, the program should wait at least two PCLK cycles (4 CCLKs) after the feed completes before going into
Power down mode. Otherwise, the watchdog could become disabled when CCLK turns off. The watchdog oscillator will never become
selected as the clock source unless CCLK is turned on again first.
2003 Nov 6 88
Philips Semiconductors
WATCHDOG TIMER
MOV WFEED1, #0A5H
MOV WFEED2, #05AH
Watchdog
Oscillator
PCLK
÷32
PRESCALER
control register
WDL (C1H)
8-Bit Down
Counter
User’s Manual - Preliminary -
P89LPC920/921/922
Watchdog reset can also be caused
by an invalid feed sequence, or by
writing to WDCON not immediately
followed by a feed sequence
RESET
SHADOW
REGISTER FOR
WDCON
PRE2PRE1PRE0WDRUN WDTOFWDCLK
WDCON(A7H)
Figure 13-3: Watchdog Timer in Watchdog Mode (WDTE = 1)
Watchdog Timer in Timer Mode
Figure 13-4 shows the Watchdog Timer in Timer Mode. In this mode, any changes to WDCON are written to the shadow register
after one watchdog clock cycle. A watchdog underflow will set the WDTOF bit. If IEN0.6 is set, the watchdog underflow is enabled
to cause an interrupt. WDTOF is cleared by writing a '0' to this bit in software. When an underflow occurs, the contents of WDL
is reloaded into the down counter and the watchdog timer immediately begins to count down again.
A feed is necessary to cause WDL to be loaded into the down counter before an underflow occurs. Incorrect feeds are ignored
in this mode.
2003 Nov 6 89
Philips Semiconductors
WATCHDOG TIMER
MOV WFEED1, #0A5H
MOV WFEED2, #05AH
Watchdog
Oscillator
CLK
÷32
PRESCALER
control register
WDL (C1H)
8-Bit Down
Counter
User’s Manual - Preliminary -
P89LPC920/921/922
Interrupt
SHADOW
REGISTER FOR
WDCON
PRE2PRE1PRE0WDRUN WDTOFWDCLK
WDCON(A7H)
Figure 13-4: Watchdog Timer in Timer Mode (WDTE = 0)
Power down operation
The WDT oscillator will continue to run in power down, consuming approximately 50uA, as long as the WDT oscillator is selected
as the clock source for the WDT. Selecting PCLK as the WDT source will result in the WDT oscillator going into power down
with the rest of the device (see section "Watchdog Clock Source" on page 88 ). Power down mode will also prevent PCLK from
running and therefore the watchdog is effectively disabled.
Periodic wakeup from Power down without an external oscillator
Without using an external oscillator source, the power consumption required in order to have a periodic wakeup is determined
by the power consumption of the internal oscillator source used to produce the wakeup.The Real-time clock running from the
internal RC oscillator can be used. The power consumption of this oscillator is approximately 300uA. Instead, if the WDT is used
to generate interrupts the current is reduced to approximately 50uA. Whenever the WDT underflows, the device will wake up.
2003 Nov 6 90
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
ADDITIONAL FEATURES
14. ADDITIONAL FEATURES
The AUXR1 register contains several special purpose control bits that relate to several chip features. AUXR1 is described in
Figure 14-1.
AUXR1
Address: A2h
Not bit addressable
Reset Source(s): Any reset
Reset Value: 000000x0B
BITSYMBOLFUNCTION
AUXR1.7CLKLPClock Low Power Select. When set, reduces power consumption in the clock circuits. Can
AUXR1.6EBRRUART Break Detect Reset Enable. If ‘1’, UART Break Detect will cause a chip reset and
AUXR1.5ENT1When set, the P0.7 pin is toggled whenever Timer1 overflows. The output frequency is
AUXR1.4ENT0When set the P1.2 pin is toggled whenever Timer0 overflows. The output frequency is
AUXR1.3SRSTSoftware Reset. When set by software, resets the P89LPC920/921/922 as if a hardware
AUXR1.20This bit contains a hard-wired 0. Allows toggling of the DPS bit by incrementing AUXR1,
AUXR1.1-Not used. Allowable to set to a “1” .
AUXR1.0DPSData Pointer Select. Chooses one of two Data Pointers.
76543210
CLKLPEBRRENT1ENT0SRST0-DPS
be used when the clock frequency is 8 MHz or less. After reset this bit is cleared to support
up to 12 MHz operation.
force the device into ISP mode.
therefore one half of the Timer1 overflow rate. Refer to the Timer/Counters section for
details.
therefore one half of the Timer0 overflow rate. Refer to the Timer/Counters section for
details.
reset occurred.
without interfering with other bits in the register.
Figure 14-1: AUXR1 register
Software reset
The SRST bit in AUXR1 gives software the opportunity to reset the processor completely, as if an external reset or watchdog
reset had occurred. If a value is written to AUXR1 that contains a 1 at bit position 3, all SFRs will be initialized and execution will
resume at program address 0000. Care should be taken when writing to AUXR1 to avoid accidental software resets.
Dual Data Pointers
The dual Data Pointers (DPTR) adds to the ways in which the processor can specify the address used with certain instructions.
The DPS bit in the AUXR1 register selects one of the two Data Pointers. The DPTR that is not currently selected is not accessible
to software unless the DPS bit is toggled.
Specific instructions affected by the Data Pointer selection are:
• INCDPTRIncrements the Data Pointer by 1.
• JMP @A+DPTRJump indirect relative to DPTR value.
• MOV DPTR, #data16 Load the Data Pointer with a 16-bit constant.
2003 Nov 6 91
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
ADDITIONAL FEATURES
• MOVCA, @A+DPTRMove code byte relative to DPTR to the accumulator.
• MOVX A, @DPTRMove data byte the accumulator to data memory relative to DPTR.
• MOVX @DPTR, AMove data byte from data memory relative to DPTR to the accumulator.
Also, any instruction that reads or manipulates the DPH and DPL registers (the upper and lower bytes of the current DPTR) will
be affected by the setting of DPS. The MOVX instructions have limited application for the P89LPC920/921/922 since the part
does not have an external data bus. However, they may be used to access Flash configuration information (see Flash
Configuration section) or auxiliary data (XDATA) memory.
Bit 2 of AUXR1 is permanently wired as a logic 0. This is so that the DPS bit may be toggled (thereby switching Data Pointers)
simply by incrementing the AUXR1 register, without the possibility of inadvertently altering other bits in the register.
2003 Nov 6 92
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
FLASH MEMORY
15. FLASH MEMORY
General description
The P89LPC920/921/922 Flash memory provides in-circuit electrical erasure and programming. The Flash can be read and
written as bytes. The Sector and Page Erase functions can erase any Flash sector (1 KB) or page (64 bytes). The Chip Erase
operation will erase the entire program memory. Five Flash programming methods are available. On-chip erase and write timing
generation contribute to a user-friendly programming interface. The P89LPC920/921/922 Flash reliably stores memory contents
even after 100,000 erase and program cycles. The cell is designed to optimize the erase and programming mechanisms. The
P89LPC920/921/922 uses V
Features
• Parallel programming with industry-standard commercial programmers
• In-Circuit serial Programming (ICP) with industry-standard commercial programmers.
• IAP-Lite allows individual and multiple bytes of code memory to be used for data storage and programmed under control of the
end application.
• Internal fixed boot ROM, containing low-level In-Application Programming (IAP) routines that can be called from the end
application (in addition to IAP-Lite).
• Default serial loader providing In-System Programming (ISP) via the serial port, located in upper end of user program memory.
• Boot vector allows user provided Flash loader code to reside anywhere in the Flash memory space, providing flexibility to the
user.
• Programming and erase over the full operating voltage range
• Read/Programming/Erase using ISP/IAP/IAP-Lite
• Any flash program operation in 2 ms (4ms for erase/program)
• Programmable security for the code in the Flash for each sector.
• > 100,000 typical erase/program cycles for each byte.
• 10-year minimum data retention.
as the supply voltage to perform the Program/Erase algorithms.
DD
Flash programming and erase
The P89LPC920/921/922 program memory consists 1 KB sectors. Each sector can be further divided into 64-byte pages. In
addition to sector erase and page erase, a 64-byte page register is included which allows from 1 to 64 bytes of a given page to
be programmed at the same time, substantially reducing overall programming time. Five mthods of programming this device are
available.
• Parallel programming with industry-standard commercial programmers.
• In-Circuit serial Programming (ICP) with industry-standard commercial programmers.
• IAP-Lite allows individual and multiple bytes of code memory to be used for data storage and programmed under control of
the end application.
• Internal fixed boot ROM, containing low-level In-Application Programming (IAP) routines athat can be called from the end
application (in addition to IAP-Lite).
• A factory-provided default serial loader, located in upper end of user program memory, providing In-System Programming
(ISP) via the serial port.
Using Flash as data storage: IAP-Lite
The Flash code memory array of this device supports IAP-Lite in addition to standard IAP functions. Any byte in a non-secured
sector of the code memory array may be read using the MOVC instruction and thus is suitable for use as non-volatile data storage. IAP-Lite provides an erase-program function that makes it easy for one or more bytes within a page to be erased and pro-
2003 Nov 6 93
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
FLASH MEMORY
grammed in a single operation without the need to erase or program any other bytes in the page. IAP-Lite is performed in the
application under the control of the microcontroller’s firmware using four SFRs and an internal 64-byte "page register" to facilitate erasing and programing within unsecured sectors. These SFRs are:
• FMCON (Flash Control Register). When read, this is the status register. When written, this is a command register. Note that
the status bits are cleared to ’0’s when the command is written.
• FMDATA (Flash Data Register). Accepts data to be loaded into the page register.
• FMADRL, FMADRH (Flash memory address low, Flash memory address high). Used to specify the byte address within the
page register or specify the page within user code memory.
The page register consists of 64 bytes and an update flag for each byte. When a LOAD command is issued to FMCON the page
register contents and all of the update flags will be cleared. When FMDATA is written, the value written to FMDATA will be
stored in the page register at the location specified by the lower 6 bits of FMADRL. In addition, the update flag for that location
will be set. FMADRL will auto-increment to the next location. Auto-increment after writing to the last byte in the page register will
"wrap -around" to the first byte in the page register, but will not affect FMADRL[7:6].. Bytes loaded into the page register do not
have to be continous. Any byte location can be loaded into the page register by changing the contents of FMADRL prior to writing to FMDATA. However, each location in the page register can only be written once following each LOAD command. Attempts
to write to a page register location more than once should be avoided.
FMADRH and FMADRL[7:6] are used to select a page of code memory for the erase-program function. When the erase-program command is written to FMCON, the locations within the code memory page that correspond to updated locations in the
page register, will have their contents erased and programmed with the contents of their corresponding locations in the page
register.Only the bytes that were loaded into the page register will be erased and programmed in the user code array. Other
bytes within the user code memory will not be affected.
Writing the erase-program command (68H) to FMCON will start the erase-program process and place the CPU in a programidle state.The CPU will remain in this idle state until the erase-program cycle is either completed or terminated by an interrupt.
When the program-idle state is exited FMCON will contain status information for the cycle.
If an interrupt occurs during an erase/programming cycle, the erase/programming cycle will be aborted and the OI flag (Operation Interrupted) in FMCON will be set. If the application permits interrupts during erasing-programming the user code should
check the OI flag (FMCON.0) after each erase-programming operation to see if the operation was aborted. If the operation was
aborted, the user’s code will need to repeat the process starting with loading the page register.
The erase-program cycle takes 4ms (2ms for erase, 2ms for programming) to complete, regardless of the number of bytes that
were loaded into the page register.
Erasing-programming of a single byte (or multiple bytes) in code memory is accomplished using the following steps:
• Write the LOAD command (00H) to FMCON. The LOAD command will clear all locations in the page register and their
corresponding update flags.
• Write the address within the page register to FMADRL. Since the loading the page register uses FMADRL[5:0], and since the
erase-program command uses FMADRH and FMADRL[7:6], the user can write the byte location within the page register
(FMADRL[5:0]) and the code memory page address (FMADRH and FMADRL[7:6]) at this time.
• Write the data to be programmed to FMDATA. This will increment FMADRL pointing to the next byte in the page register.
• Write the address of the next byte to be programmed to FMADRL, if desired. (Not needed for contiguous bytes since FMADRL
is auto-incremented). All bytes to be programmed must be within the same page.
• Write the data for the next byte to be programmed to FMDATA.
• Repeat writing of FMADRL and/or FMDATA until all desired bytes have been loaded into the page register.
• Write the page address in user code memory to FMADRH and FMADRL[7:6], if not previously included when writing the page
register address to FMADRL[5:0].
• Write the erase-program command (68H) to FMCON,starting the erase-program cycle.
• Read FMCON to check status. If aborted, repeat starting with the LOAD command.
An assembly language routine to load the page register and perform an erase/program operation is shown in Figure 15-2. A
2003 Nov 6 94
Philips Semiconductors
FLASH MEMORY
similar C-language routine is shown in Figure 15-3.
User’s Manual - Preliminary -
P89LPC920/921/922
FMCON
Address: E4h
Not bit addressable
Reset Source(s): Any reset
Reset Value:
BITSYMBOLFUNCTION
FMCON.7-4-Reserved.
FMCON.3HVAHigh voltage abort. Set if either an interrupt or a brown-out is detected during a program
FMCON.2HVEHigh voltage error. Set when an error occurs in the high voltage generator.
FMCON.1SVSecurity violation. Set when an attempt is made to program, erase, or CRC a secured
FMCON.0OIOperation interrupted. Set when cycle aborted due to an interrupt or reset.
76543210
----HVAHVESVOI
or erase cycle. Also set if the brown-out detector is disabled at the start of a program or
erase cycle.
sector or page.
Figure 15-1: Flash Memory Control Registe
2003 Nov 6 95
Philips Semiconductors
FLASH MEMORY
r
;* Inputs: *
;* R3 = number of bytes to program (byte) *
;* R4 = page address MSB(byte) *
;* R5 = page address LSB(byte) *
;* R7 = pointer to data buffer in RAM(byte) *
;* Outputs: *
;* R7 = status (byte) *
;* C = clear on no error, set on error *
LOAD EQU 00H
EP EQU 68H
PGM_USER:
MOV FMCON,#LOAD ;load command, clears page register
MOV FMADRH,R4 ;get high address
MOV FMADRL,R5 ;get low address
MOV A,R7 ;
MOV R0,A ;get pointer into R0
LOAD_PAGE:
MOV FMDAT,@R0 ;write data to page register
INC R0 ;point to next byte
DJNZ R3,LOAD_PAGE ;do until count is zero
MOV FMCON,#EP ;else erase & program the page
MOV R7,FMCON ;copy status for return
MOV A,R7 ;read status
ANL A,#0FH ;save only four lower bits
JNZ BAD ;
CLR C ;clear error flag if good
RET ;and return
BAD:
SETB C ;set error flag
RET ;and return
Figure 15-2: Assembly language routine to erase/program all or part of a page
User’s Manual - Preliminary -
P89LPC920/921/922
2003 Nov 6 96
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
FLASH MEMORY
#include <REG931.H>
unsigned char idata dbytes[64]; // data buffer
unsigned char Fm_stat; // status result
bit PGM_USER (unsigned char, unsigned char);
bit prog_fail;
for (i=0;i<64;i=i+1)
{
FMDATA = dbytes[i];
}
FMCON = EP; //erase & prog page command
Fm_stat = FMCON; //read the result status
if ((Fm_stat & 0x0F)!=0) prog_fail=1; else prog_fail=0;
return(prog_fail);
}
Figure 15-3: C-language routine to erase/program all or part of a page
In-Circuit Programming (ICP)
In-Circuit Programming is a method intended to allow commercial programmers to program and erase these devices without
removing the microcontroller from the system. The In-Circuit Programming facility consists of a series of internal hardware
resources to facilitate remote programming of the P89LPC920/921/922 through a two-wire serial interface. Philips has made incircuit programming in an embedded application possible with a minimum of additional expense in components and circuit
board area. The ICP function uses five pins (Vdd, Vss, P0.5, P0.4, and RST
interface your application to an external programmer in order to use this feature.
). Only a small connector needs to be available to
ISP and IAP capabilities of the P89LPC920/921/922
An In-Application Programming (IAP) interface is provided to allow the end user’s application to erase and reprogram the user
code memory. In addition, erasing and reprogramming of user-programmable bytes including UCFG1, the Boot Status Bit, and
the Boot Vector is supported. As shipped from the factory, the upper 512 bytes of user code space contains a serial In-System
Programming (ISP) loader allowing for the device to be programmed in circuit through the serial port. This ISP boot loader will,
in turn, call low-level routines through the same common entry point that can be used by the end-user application.
2003 Nov 6 97
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
FLASH MEMORY
Boot ROM
When the microcontroller contains a a 256 byte Boot ROM that is separate from the user’s Flash program memory. This Boot
ROM contains routines which handle all of the low level details needed to erase and program the user Flash memory . A user
program simply calls a common entry point in the Boot ROM with appropriate parameters to accomplish the desired operation.
Boot ROM operations include operations such as erase sector, erase page, program page, CRC, program security bit, etc. The
Boot ROM occupies the program memory space at the top of the address space from FF00 to FFFF hex, thereby not conflicting
with the user program memory space.This function is in addition to the IAP-Lite feature.
Power-On reset code execution
The P89LPC920/921/922 contains two special Flash elements: the BOOT VECTOR and the Boot Status Bit. Following reset,
the P89LPC920/921/922 examines the contents of the Boot Status Bit. If the Boot Status Bit is set to zero, power-up execution
starts at location 0000H, which is the normal start address of the user’s application code. When the Boot Status Bit is set to a va
one, the contents of the Boot Vector is used as the high byte of the execution address and the low byte is set to 00H.
The factory default settings for these devcies are show in Table 15-1, below.The factory pre-programmed boot loader can be
erased by the user. Users who wish to use this loader should take cautions to avoid erasing the last 1KB sector on the device. Instead, the page erase function can be used to erase the eight 64-byte pages located in this sector. A custom
boot loader can be written with the Boot Vector set to the custom boot loader, if desired.
Table 15-1: Boot Loader Address and Default Boot Vector
PRODUCT
P89LPC922 8K x 81FFF 15H DDH 0CH
P89LPC921 4K x 80FFF 15H DDH 0BH
P89LPC920 2K x 807FF 15H DDH 1AH1Kx864x80600H-07FFH07H
FLASH
SIZE
END
ADDRESS
SIGNATURE BYTES
MFG ID1 ID2
or 15H DDH 05H
or 15H DDH 05H
SECTOR
SIZE
1Kx864x81E00H-1FFFH1FH
1Kx864x80E00H-0FFFH0FH
PAGE
SIZE
PRE-PROGRAMMED
SERIAL LOADER
DEFAULT
BOOT
VECTOR
Hardware activation of the Boot Loader
The boot loader can also be executed by forcing the device into ISP mode during a power-on sequence (see Figure 15-4).This
is accomplished by powering up the device with the reset pin initially held low and holding the pin low for a fixed time after VDD
rises to its normal operating value. This is followed by three, and only three, properly timed low-going pulses. Fewer or more
than three pulses will result in the device not entering ISP mode. Timing specifications may be found in the datasheet for this
device.
This has the same effect as having a non-zero status bit. This allows an application to be built that will normally execute the user
code but can be manually forced into ISP operation. If the factory default setting for the Boot Vector is changed, it will no longer
point to the factory pre-programmed ISP boot loader code. If this happens, the only way it is possible to change the contents of
the Boot Vector is through the parallel or ICP programming method, provided that the end user application does not contain a
customized loader that provides for erasing and reprogramming of the Boot Vector and Boot Status Bit. After programming the
Flash, the status byte should be programmed to zero in order to allow execution of the user’s application code beginning at
address 0000H.
2003 Nov 6 98
Philips Semiconductors
User’s Manual - Preliminary -
P89LPC920/921/922
FLASH MEMORY
VDD
t
t
VR
RST
In-System Programming (ISP)
In-System Programming is performed without removing the microcontroller from the system. The In-System Programming facility consists of a series of internal hardware resources coupled with internal firmware to facilitate remote programming of the
P89LPC920/921/922 through the serial port. This firmware is provided by Philips and embedded within each P89LPC920/921/
922 device. The Philips In-System Programming facility has made in-circuit programming in an embedded application possible
with a minimum of additional expense in components and circuit board area. The ISP function uses five pins (Vdd, Vss, TxD,
RxD, and RST). Only a small connector needs to be available to interface your application to an external circuit in order to use
this feature.
RH
t
RL
Figure 15-4: Forcing ISP Mode
Using the In-System Programming
The ISP feature allows for a wide range of baud rates to be used in your application, independent of the oscillator frequency. It
is also adaptable to a wide range of oscillator frequencies. This is accomplished by measuring the bit-time of a single bit in a
received character. This information is then used to program the baud rate in terms of timer counts based on the oscillator frequency. The ISP feature requires that an initial character (an uppercase U) be sent to the P89LPC920/921/922 to establish the
baud rate. The ISP firmware provides auto-echo of received characters. Once baud rate initialization has been performed, the
ISP firmware will only accept Intel Hex-type records. Intel Hex records consist of ASCII characters used to represent hexadecimal values and are summarized below:
:NNAAAARRDD..DDCC<crlf>
In the Intel Hex record, the “NN” represents the number of data bytes in the record. The P89LPC920/921/922 will accept up to
64 (40H) data bytes. The “AAAA” string represents the address of the first byte in the record. If there are zero bytes in the
record, this field is often set to 0000. The “RR” string indicates the record type. A record type of “00” is a data record. A record
type of “01” indicates the end-of-file mark. In this application, additional record types will be added to indicate either commands
or data for the ISP facility. The maximum number of data bytes in a record is limited to 64 (decimal). ISP commands are summarized in Table 15-2 . As a record is received by the P89LPC920/921/922, the information in the record is stored internally and
a checksum calculation is performed. The operation indicated by the record type is not performed until the entire record has
been received. Should an error occur in the checksum, the P89LPC920/921/922 will send an “X” out the serial port indicating a
checksum error. If the checksum calculation is found to match the checksum in the record, then the command will be executed.
In most cases, successful reception of the record will be indicated by transmitting a “.” character out the serial port
2003 Nov 6 99
Philips Semiconductors
FLASH MEMORY
Table 15-2: In-System Programming (ISP) hex record formats
Record typeCommand/data function
Program User Code Memory Page
:nnaaaa00dd..ddcc
Where:
nn = number of bytes to program
00
01
aaaa = page address
dd..dd= data bytes
cc = checksum
Example:
:100000000102030405006070809cc
Read Version Id
:00xxxx01cc
Where:
xxxx = required field but value is a “don’t care”
cc = checksum
Example:
:00000001cc
User’s Manual - Preliminary -
P89LPC920/921/922
2003 Nov 6 100
Loading...
+ hidden pages
You need points to download manuals.
1 point = 1 manual.
You can buy points or you can get point for every manual you upload.