DALLAS DS87C550-FNL

PRODUCT PREVIEW
DS87C550
EPROM High-Speed
Micro with A/D and PWM
www.dalsemi.com
FEATURES
PIN ASSIGNMENT
§ 87C52-Compatible
-
61
10
§ On-chip Memory
-
1
9
8051 pin- and instruction set-compatible
Three 16-bit timer/counters
256 bytes scratchpad RAM
60
8 kbytes EPROM (OTP & Windowed Packages)
- 1 kbyte extra on-chip SRAM for MOVX access
§ On-chip Analog to Digital Converter
-
Eight channels of analog input, 10-bit resolution
Fast conversion time
Selectable internal or external reference voltage
DALLAS
DS87C550
§ Pulse Width Modulator Outputs
-
Four channels of 8-bit PWM
Channels cascadable to 16-bit PWM
§ 4 Capture + 3 Compare Registers
§ 55 I/O Port Pins
§ New Dual Data Pointer Operation
-
26
44
27
Either data pointer can be incremented or
decremented
68-Pin PLCC
68-Pin WINDOWED CLCC
§ ROMSIZE Feature
-
43
Selects effective on-chip ROM size from 0 to 8k
Allows access to entire external memory map
Dynamically adjustable by software
80
65
§ High-Speed Architecture
-
-
4 clocks/machine cycle (8051 = 12)
Runs DC to 33 MHz clock rates
Single-cycle instruction in 121 ns
New Stretch Cycle feature allows access to
fast/slow memory or peripherals
1
64
§ Unique Power Savings Modes
§ EMI Reduction Mode disables ALE when not
needed
§ High integration controller includes:
-
DALLAS
DS87C550
Power-fail reset
Early-warning power-fail interrupt
Two full-duplex hardware serial ports
Programmable watchdog timer
§ 16 total interrupt sources with 6 external
24
Available in 68-pin PLCC, 80-pin PQFP, and
68-pin windowed CLCC
41
25
40
80-Pin PQFP
1 of 50
061499
DS87C550
DESCRIPTION
The DS87C550 EPROM High-Speed Micro with A/D and PWM is a member of the fastest 100% 8051compatible microcontroller family available. It features a redesigned processor core that removes wasted
clock and memory cycles. As a result, it executes 8051 instructions up to three times faster than the
original architecture for the same crystal speed. The DS87C550 also offers a maximum crystal speed of
33 MHz, resulting in apparent execution speeds of up to 99 MHz.
The DS87C550 uses an industry standard 8051 pin-out and includes standard resources such as three
timer/counters, and 256 bytes of scratchpad RAM. This device also features 8 kbytes of EPROM with an
extra 1 kbyte of data RAM (in addition to the 256 bytes of scratchpad RAM), and 55 I/O ports pins. Both
One-Time-Programmable (OTP) and windowed packages are available.
Besides greater speed, the DS87C550 includes a second full hardware serial port, seven additional
interrupts, a programmable watchdog timer, brownout monitor, and power-fail reset.
The DS87C550 also provides dual data pointers (DPTRs) to speed block data memory moves. The user
can also dynamically adjust the speed of external accesses between two and 12 machine cycles for
flexibility in selecting memory and peripherals.
Power Management Mode (PMM) is useful for portable or battery-powered applications. This feature
allows software to select a lower speed clock as the main time base. While normal operation has a
machine cycle rate of 4 clocks per cycle, the PMM allows the processor to run at 1024 clocks per cycle.
For example, at 12 MHz, standard operation has a machine cycle rate of 3 MHz. In Power Management
Mode, software can select an 11.7 kHz (12 MHz/1024) machine cycle rate. There is a corresponding
reduction in power consumption due to the processor running slower.
The DS87C550 also offers two features that can significantly reduce electromagnetic interference (EMI).
One EMI reduction feature allows software to select a reduced emission mode that disables the ALE
signal when it is unneeded. The other EMI reduction feature controls the current to the address and data
pins interfacing to external devices producing a controlled transition of these signals.
ORDERING INFORMATION
PART NUMBER
DS87C550-QCL
DS87C550-FCL
DS87C550-QNL
DS87C550-FNL
DS87C550-KCL
PACKAGE
68-pin PLCC
80-pin PQFP
68-pin PLCC
80-pin PQFP
68-pin windowed CLCC
2 of 50
MAX. CLOCK SPEED
33 MHz
33 MHz
33 MHz
33 MHz
33 MHz
TEMPERATURE RANGE
0°C to +70°C
0°C to +70°C
-40°C to +85°C
-40°C to +85°C
0°C to 70°C
DS87C550
DS87C550 BLOCK DIAGRAM Figure 1
3 of 50
DS87C550
PIN DESCRIPTION Table 1
PLCC/
CLCC
2
36
37
15
QFP
72
34
35
9
SIGNAL NAME
VCC
GND
35
34
32
31
XTAL1
XTAL2
47
48
PSEN
48
49
ALE
49
50
EA
16-23
10-17
P1.0-P1.7
16
17
18
19
20
21
22
23
10
11
12
13
14
15
16
17
RST
DESCRIPTION
VCC - Digital +5V power input.
GND – Digital ground.
RST - I/O. The RST input pin contains a Schmitt voltage input to recognize
external active high Reset inputs. The pin also employs an internal pulldown
resistor to allow for a combination of wired OR external Reset sources. An RC is
not required for power-up, as the DS87C550 provides this function internally.
This pin also acts as an output when the source of the reset is internal to the device
(i.e., watchdog timer, power-fail, or crystal-fail detect). In this case, the RST pin
will be held high while the processor is in a Reset state, and will return to low as
the processor exits this state. When this output capability is used, the RST pin
should not be connected to an RC network or a logic output driver.
Input - The crystal oscillator pins XTAL1 and XTAL2 provide support for
fundamental mode, parallel resonant, AT cut crystals. XTAL1 acts also as an input
if there is an external clock source in place of a crystal. XTAL2 serves as the
output of the crystal amplifier. Note that this output cannot be used to drive any
additional load when a crystal is attached as this can disturb the oscillator circuit.
- Output. The Program Store Enable output. This signal is commonly
connected to optional external ROM memory as a chip enable. PSEN will provide
an active low pulse during a program byte access, and is driven high when not
accessing external program memory.
ALE - Output. The Address Latch Enable output functions as a clock to latch the
external address LSB from the multiplexed address/data bus on Port 0. This signal
is commonly connected to the latch enable of an external 373 family transparent
latch. ALE is driven high when the DS87C550 is in a Reset condition. ALE can
also be disabled and forced high using the EMI reduction mode ALEOFF.
EA - Input. An active low input pin that when connected to ground will force the
DS87C550 to use an external program memory. The internal RAM is still
accessible as determined by register settings. EA should be connected to VCC to
use internal program memory. The input level on this pin is latched at reset.
Port 1 - I/O. Port 1 functions as both an 8-bit, bi-directional I/O port and an
alternate functional interface for several internal resources. The reset condition of
Port 1 is all bits at logic 1. In this state, a weak pullup holds the port high. This
condition allows the pins to serve as both input and output. Input is possible since
any external circuit whose output drives the port will overcome the weak pullup.
When software writes a 0 to any Port 1 pin, the DS87C550 will activate a strong
pulldown that remains on until either a 1 is written or a reset occurs. Writing a 1
after the port has been at 0 will cause a strong transition driver to turn on, followed
by a weaker sustaining pullup. Once the momentary strong driver turns off, the
port again returns to a weakly held high output (and input) state. The alternate
functions of Port 1 pins are detailed below. Note that when the Capture/Compare
functions of timer 2 are used, the interrupt input pins become capture trigger
inputs.
PSEN
Port
P1.0
P1.1
P1.2
P1.3
P1.4
P1.5
P1.6
P1.7
Alternate Function
INT2/CT0 External Interrupt 2/Capture Trigger 0
INT3/CT1 External Interrupt 3/Capture Trigger 1
INT4/CT2 External Interrupt 4/Capture Trigger 2
INT5/CT3 External Interrupt 5/Capture Trigger 3
T2
External I/O for Timer/Counter 2
T2EX
Timer/Counter 2 Capture/Reload Trigger
RXD1
Serial Port 1 Input
TXD1
Serial Port 1 Output
4 of 50
DS87C550
PLCC/
CLCC
50-57
57
56
55
54
53
52
51
50
39-46
39
40
41
42
43
44
45
46
24-31
QFP
51-58
58
57
56
55
54
53
52
51
38-42
45-47
38
39
40
41
42
45
46
47
18-20
23-27
24
25
26
27
28
29
30
31
7-14
18
19
20
23
24
25
26
27
80
1-2
4-8
7
8
9
10
11
12
13
14
80
1
2
4
5
6
7
8
SIGNAL NAME
DESCRIPTION
P0.0 (AD0)
P0.1 (AD1)
P0.2 (AD2)
P0.3 (AD3)
P0.4 (AD4)
P0.5 (AD5)
P0.6 (AD6)
P0.7 (AD7)
Port 0-I/O - AD0-7. Port 0 is an open-drain 8-bit, bi-directional general-purpose
I/O port. When used in this mode pullup resistors are required to provide a logic 1
output. As an alternate function, Port 0 operates as a multiplexed address/data bus
to access off-chip memory or peripherals. In this mode, the LSB of the memory
address is output on the bus during the time that ALE is high. When ALE falls to
a logic 0, the port transitions to a bi-directional data bus. In this mode, the port
provides active high drivers for logic 1 output. The reset condition of Port 0 is tristate (i.e., the open drain devices are off).
P2.0 (A8)
P2.1 (A9)
P2.2 (A10)
P2.3 (A11)
P2.4 (A12)
P2.5 (A13)
P2.6 (A14)
P2.7 (A15)
P3.0-P3.7
Port 2-I/O Address A15:A8. Port 2 functions as an 8-bit bi-directional I/O port
or alternately as an external address bus (A15-A8). The reset condition of Port 2 is
logic high I/O state. In this state, weak pullups hold the port high allowing the
pins to be used as an input or output as described above for Port 1. As an alternate
function Port 2 can function as MSB of the external address bus. This bus can be
used to read external memory or peripherals.
P4.0-P4.7
Port 3 - I/O. Port 3 functions as an 8-bit bi-directional I/O port or alternately as
an interface for External Interrupts, Serial Port 0, Timer 0 & 1 Inputs, and RD and
WR strobes. When functioning as an I/O port, these pins operate as indicated
above for Port 1. The alternate modes of Port 3 are detailed below.
Port Alternate Mode
P3.0 RXD0
Serial Port 0 Input
P3.1 TXD0
Serial Port 0 Output
P3.2
External Interrupt 0
INT0
P3.3
External Interrupt 1
INT1
P3.4 T0
Timer 0 External Input
P3.5 T1
Timer 1 External Input
P3.6
External Data Memory Write Strobe
WR
RD
External Data Memory Read Strobe
P3.7
Port 4 - I/O. Port 4 functions as an 8-bit bi-directional I/O port or alternately as
an interface to Timer 2’s Capture Compare functions. When functioning as an I/O
port, these pins operate as indicated in the Port 1 description. The alternate modes
of Port 4 are detailed below.
Port 4
P4.0
P4.1
P4.2
P4.3
P4.4
P4.5
P4.6
P4.7
Alternate Mode
CMSR0
Timer 2 compare match set/reset output 0
CMSR1
Timer 2 compare match set/reset output 1
CMSR2
Timer 2 compare match set/reset output 2
CMSR3
Timer 2 compare match set/reset output 3
CMSR4
Timer 2 compare match set/reset output 4
CMSR5
Timer 2 compare match set/reset output 5
CMT0
Timer 2 compare match toggle output 0
CMT1
Timer 2 compare match toggle output 1
5 of 50
DS87C550
PLCC/
CLCC
1,
62-68
QFP
64-71
SIGNAL NAME
P5.0-P5.7
1
68
67
66
65
64
63
62
3-6, 32
33, 38
71
70
69
68
67
66
65
64
28, 29
37
74-77
4
5
32
33
6
38
3
59
75
76
28
29
77
37
74
60
A vref+
58
59
A vref-
61
60
63
61
3, 21
22, 30
33, 36
43, 44
62, 73
78, 79
AVCC
AVSS
NC
P6.0-P6.5
P6.7
DESCRIPTION
Port 5 - I/O. Port 5 functions as an open-drain 8-bit bi-directional I/O port or
alternately as an interface to the A/D converter. When used for general purpose
I/O, these pins operate in a quasi-bi-directional mode. Writing a logic 1 to these
pins (reset condition) will cause them to tri-state. This allows the pins to serve as
inputs since the tri-state condition can be driven by an external device. If a logic 0
is written to a pin, it is pulled down internally and therefore serves as an output
pin containing a logic 0. Because these pins are open-drain, external pullup
resistors are required to create a logic 1 level when they are used as outputs. As
an alternate function Port 5 pins operate as the analog inputs for the A/D
converter as described below.
Port Alternate Mode
P5.0 ADC0
Analog to Digital Converter input channel 0
P5.1 ADC1
Analog to Digital Converter input channel 1
P5.2 ADC2
Analog to Digital Converter input channel 2
P5.3 ADC3
Analog to Digital Converter input channel 3
P5.4 ADC4
Analog to Digital Converter input channel 4
P5.5 ADC5
Analog to Digital Converter input channel 5
P5.6 ADC6
Analog to Digital Converter input channel 6
P5.7 ADC7
Analog to Digital Converter input channel 7
Port 6 - I/O. Port 6 functions as an 6-bit bi-directional I/O port or alternately as
an interface to the PWM and A/D on-board peripherals. As an I/O port, these pins
operate as described in Port 1. The alternate modes of Port 6 are detailed below.
Port Alternate Function
P6.0 PWMO0 PWM channel 0 output
P6.1 PWMO1 PWM channel 1 output
P6.2 PWMO2 PWM channel 2 output
P6.3 PWMO3 PWM channel 3 output
P6.4 PWMC0 PWM0 clock input
P6.5 PWMC1 PWM1 clock input
P6.7 STADC
External A/D conversion start signal (active low)
A/D +Reference - Input. When selected, supplies the positive reference voltage
for the A/D converter. This signal should be isolated from digital VCC to prevent
noise from affecting A/D measurements.
A/D -Reference - Input. When selected, supplies the negative reference voltage
for the A/D converter. This signal should be isolated from digital GND to prevent
noise from affecting A/D measurements.
Analog VCC
Analog Ground
NC-Reserved. These pins should not be connected. They are reserved for use
with future devices in this family.
6 of 50
DS87C550
COMPATIBILITY
The DS87C550 is a fully static, CMOS 8051-compatible microcontroller designed for high performance.
While remaining familiar to 8051 family users, it has many new features. With very few exceptions,
software written for existing 8051-based systems works without modification on the DS87C550. The
exception is critical timing since the High Speed Micro performs its instructions much faster than the
original for any given crystal selection. The DS87C550 runs the standard 8051 family instruction set and
is pin-compatible with existing devices with similar features in PLCC or QFP packages.
The DS87C550 provides three 16-bit timer/counters, two full-duplex serial ports, 256 bytes of direct
RAM plus 1 kbyte of extra MOVX RAM. I/O ports have the same operation as a standard 8051 product.
Timers will default to a 12 clock per cycle operation to keep their timing compatible with original 8051
family systems. However, timers are individually programmable to run at the new 4 clocks per cycle if
desired.
The DS87C550 provides several new hardware features implemented by new Special Function Registers.
A summary of all SFRs is provided in Table 2.
PERFORMANCE OVERVIEW
The DS87C550 features a high-speed, 8051-compatible core. Higher speed comes not just from
increasing the clock frequency, but from a newer, more efficient design.
This updated core does not have the dummy memory cycles that are present in a standard 8051. A
conventional 8051 generates machine cycles using the clock frequency divided by 12. In the DS87C550,
the same machine cycle takes 4 clocks. Thus the fastest instruction, 1 machine cycle, executes three times
faster for the same crystal frequency. Note that these are identical instructions. The majority of
instructions on the DS87C550 will see the full 3 to 1 speed improvement. However, some instructions
will achieve between 1.5 and 2.4 to 1 improvement. Regardless of specific performance improvements,
all instructions are faster than the original 8051.
The numerical average of all opcodes gives approximately a 2.5 to 1 speed improvement. Improvement of
individual programs will depend on the actual mix of instructions used. Speed sensitive applications
would make the most use of instructions that are 3 times faster. However, the sheer number of 3 to 1
improved opcodes makes dramatic speed improvements likely for any arbitrary combination of
instructions. These architecture improvements and the sub-micron CMOS design produce a peak
instruction cycle in 121 ns (8.25 MIPs). The Dual Data Pointer feature also allows the user to eliminate
wasted instructions when moving blocks of memory.
INSTRUCTION SET SUMMARY
All instructions in the DS87C550 perform exactly the same functions as their 8051 counterparts. Their
effect on bits, flags, and other status functions is identical. However, the timing of each instruction is
different. This applies both in absolute and relative number of clocks.
For absolute timing of real-time events, the timing of software loops can be calculated using a table in the
High Speed Micro User’s Guide. However, counter/timers default to run at the old 12 clocks per
increment. In this way, timer-based events occur at the standard intervals with software executing at
higher speed. Timers optionally can run at 4 clocks per increment to take advantage of faster processor
operation.
The relative time of two instructions might be different in the new architecture than it was previously. For
example, in the original architecture, the “MOVX A, @DPTR” instruction and the “MOV direct, direct”
7 of 50
DS87C550
instruction used two machine cycles or 24 oscillator cycles. Therefore, they required the same amount of
time. In the DS87C550, the MOVX instruction takes as little as two machine cycles or eight oscillator
cycles, but the “MOV direct, direct” uses three machine cycles or 12 oscillator cycles. While both are
faster than their original counterparts, they now have different execution times. This is because the
DS87C550 usually uses one instruction cycle for each instruction byte. Examine the timing of each
instruction for familiarity with the changes. Note that a machine cycle now requires just 4 clocks, and
provides one ALE pulse per cycle. Many instructions require only one cycle, but some require five. In the
original architecture, all were one or two cycles except for MUL and DIV. Refer to the High Speed Micro
User’s Guide for details and individual instruction timing.
SPECIAL FUNCTION REGISTERS
Special Function Registers (SFRs) control most special features of the DS87C550. This allows the
DS87C550 to have many new features but use the same instruction set as the 8051. When writing
software to use a new feature, an equate statement defines the SFR to an assembler or compiler. This is
the only change needed to access the new function. The DS87C550 duplicates the SFRs contained in the
standard 80C52. Table 2 shows the register addresses and bit locations. Many are standard 80C52
registers. The High Speed Micro User’s Guide describes all SFRs in full detail.
SPECIAL FUNCTION REGISTER LOCATION: Table 2
REGISTER
BIT7
BIT6
BIT5
BIT4
BIT3
BIT2
BIT1
BIT0
ADDRESS
PORT0
SP
DPL
DPH
DPL1
DPH1
DPS
PCON
TCON
TMOD
TL0
TL1
TH0
TH1
CKCON
PORT1
RCON
SCON0
SBUF0
PMR
PORT2
SADDR0
SADDR1
IE
CMPL0
CMPL1
CMPL2
CPTL0
CPTL1
CPTL2
CPTL3
PORT3
ADCON1
P0.7
P0.6
P0.5
P0.4
P0.3
P0.2
P0.1
P0.0
ID1
SMOD_0
TF1
GATE
ID0
SMOD0
TR1
C/ T
TSL
OFDF
TF0
M1
OFDE
TR0
M0
GF1
IE1
GATE
GF0
IT1
C/ T
STOP
IE0
M1
SEL
IDLE
IT0
M0
WD1
P1.7
SM0/FE_0
WD0
P1.6
SM1_0
T2M
P1.5
SM2_0
T1M
P1.4
REN_0
T0M
P1.3
CKRDY
TB8_0
MD2
P1.2
RGMD
RB8_0
MD1
P1.1
RGSL
TI_0
MD0
P1.0
BGS
RI_0
CD1
P2.7
CD0
P2.6
SWB
P2.5
CTM
P2.4
4X/ 2X
P2.3
ALEOFF
P2.2
DEM1
P2.1
DME0
P2.0
EA
EAD
ES1
ES0
ET1
EX1
ET0
EX0
P3.7
P3.6
EOC
P3.5
CONT/SS
P3.4
ADEX
P3.3
WCQ
P3.2
WCM
P3.1
ADON
P3.0
WCIO
80h
81h
82h
83h
84h
85h
86h
87h
88h
89h
8Ah
8Bh
8Ch
8Dh
8Eh
90h
91h
98h
99h
9Fh
A0h
A1h
A2h
A8h
A9h
AAh
ABh
ACh
ADh
AEh
AFh
B0h
B2h
STRT/BSY
8 of 50
DS87C550
SPECIAL FUNCTION REGISTER LOCATION: Table 2 cont’d
REGISTER
BIT7
BIT6
BIT5
BIT4
BIT3
BIT2
BIT1
BIT0
ADDRESS
ADCON2
ADMSB
ADLSB
WINHI
WINLO
IP
SADEN0
SADEN1
T2CON
T2MOD
PORT4
OUTCF
MUX2
MUX1
MUX0
APS3
APS2
APS1
APS0
-
PAD
PS1
PS0
PT1
PX1
PT0
PX0
TF2
CMT1
ADC7
PIP
EXF2
CMT0
ADC6
HIP
RCLK
CMSR5
ADC5
LIP
TCLK
CMSR4
ADC4
-
EXEN2
CMSR3
ADC3
SPTA1
TR2
CMSR2
RMS2
ADC2
SPRA1
C/ T2
T2OE
CMSR1
RMS1
ADC1
SPTA0
DCEN
CMSR0
RMS0
ADC0
SPRA0
-
CM2F
CM1F
CM0F
IE5/CF3
IE4/CF2
IE3/CF1
IE2/CF0
CY
AC
F0
RS1
RS0
OV
F1
P
PWMADR
ADRS
SCON1
SBUF1
PWM0
PWM1
PWM2
PWM3
ACC
PW01CS
PW23CS
SM0/FE_1
SM1_1
SM2_1
REN_1
TB8_1
RB8_1
PWE1
TI_1
PWE0
RI_1
PW0S2
PW2S2
PW0F
PW2F
PW0S1
PW2S1
PW0DC
PW2DC
PW0S0
PW2S0
PW0OE
PW2OE
PW0EN
PW2EN
PW0T/C
PW2T/C
PW1S2
PW3S2
PW1F
PW3F
PW1S1
PW3S1
PW1DC
PW3DC
PW1S0
PW3S0
PW1OE
PW3OE
PW1EN
PW3EN
PW1T/C
PW3T/C
ET2
TF2S
ECM1
-
EX2/EC0
CT1
EX4/EC2
CT1
EX3/EX1
T2P1
CT2
ECM0
TF2B
CT2
EX5/EC3
-
CT3
ECM2
TF2BS
CT3
CT0
T2P0
CT0
TGFF1
CMTE1
TGFF0
CMTE0
CMS5
CMR5
CMS4
CMR4
CMS3
CMR3
CMS2
CMR2
CMS1
CMR1
CMS0
CMR0
STADC
PT2
SMOD_1
PCM2
POR
PWMC1
PCM1
EPFI
PWMC0
PCM0
PFI
PWMO3
PX5/PC3
WDIF
PWMO2
PX4/PC2
WTRF
PWMO1
PX3/PC1
EWT
PWMO0
B3h
B4h
B5h
B6h
B7h
B8h
B9h
BAh
BEh
BFh
C0h
C2h
C4h
C5h
C7h
C8h
C9h
CAh
CBh
CCh
CDh
CEh
CFh
D0h
D2h
D3h
D4h
D5h
D6h
D8h
D9h
DCh
DDh
DEh
DFh
E0h
E1h
E2h
E3h
E4h
E6h
E7h
E8h
EAh
EBh
ECh
EDh
EEh
EFh
F0h
F1h
F8h
FFh
ROMSIZE
PORT5
STATUS
TA
T2IR
CMPH0
CMPH1
CMPH2
CPTH0
CPTH1
CPTH2
CPTH3
PSW
PW0FG
PW1FG
PW2FG
PW3FG
PW01CON
PW23CON
RL2
RLOADL
RLOADH
EIE
T2SEL
CTCON
TL2
TH2
SETR
RSTR
B
PORT6
EIP
WDCON
9 of 50
PX2/PC0
RWT
DS87C550
MEMORY RESOURCES
As is convention within the 8051 architecture, the DS87C550 uses three memory areas. The total memory
configuration of the DS87C550 is 8 kbytes of EPROM, 1 kbyte of data SRAM and 256 bytes of
scratchpad or direct RAM. The 1 kbyte of data space SRAM is read/write accessible and is memory
mapped. This on-chip SRAM is reached by the MOVX instruction. It is not used for executable
memory. The scratchpad area is 256 bytes of register mapped RAM and is identical to the RAM found on
the 80C52. There is no conflict or overlap among the 256 bytes and the 1k as they use different
addressing modes and separate instructions.
OPERATIONAL CONSIDERATION
The erasure window of the windowed CLCC package should be covered without regard to the
programmed/unprogrammed state of the EPROM. Otherwise, the device may not meet the AC and DC
parameters listed in the datasheet.
PROGRAM MEMORY
On-chip ROM begins at address 0000h and is contiguous through 1FFFh (8k). Exceeding the maximum
address of on-chip ROM will cause the DS87C550 to access off-chip memory. However, the maximum
on-chip decoded address is selectable by software using the ROMSIZE feature. Software can cause the
DS87C550 to behave like a device with less on-chip memory. This is beneficial when overlapping
external memory, such as Flash, is used.
With the ROMSIZE feature the maximum on-chip memory size is dynamically variable. Thus a portion
of on-chip memory can be removed from the memory map to access off-chip memory, then restored to
access on-chip memory. In fact, all of the on-chip memory can be removed from the memory map,
allowing the full 64k memory space to be addressed as off-chip memory. ROM addresses that are larger
than the selected maximum are automatically fetched from outside the part via Ports 0 & 2. A depiction
of the ROM memory map is shown in Figure 2.
The ROMSIZE register is used to select the maximum on-chip decoded address for ROM. Bits RMS2,
RMS1, RMS0 (ROMSIZE2:0) have the following effect.
RMS2
0
0
0
0
1
1
1
1
RMS1
0
0
1
1
0
0
1
1
RMS0
0
1
0
1
0
1
0
1
Maximum on-chip
ROM Address
0k
1k (0h - 03FFh)
2k (0h - 07FFh)
4k (0h - 0FFFh)
8k (0h – 1FFFh) default
invalid - reserved
invalid - reserved
invalid - reserved
The reset default condition is a maximum on-chip ROM address of 8 kbytes. Thus no action is required if
this feature is not used. Therefore when accessing external program memory, the first 8 kbytes would be
inaccessible. To select a smaller effective ROM size, software must alter bits RMS2-RMS0. Altering
these bits requires a Timed Access procedure as explained below. The ROMSIZE register should be
manipulated from a safe area in the program memory map. This is a program memory address that will
not be affected by the change. For example, do not select a maximum ROM size of 4k from an internal
ROM address of 5k. This would cause the current address to switch from internal to external and
potentially cause invalid operation. Similarly, do not instantly switch from external to internal memory.
10 of 50
DS87C550
For example, do not select a maximum ROM address of 8k from an external ROM address of 7k (if
ROMSIZE is set for 4k or less).
Off-chip memory is accessed using the multiplexed address/data bus on P0 and the MSB address on P2.
While serving as a memory bus, these pins are not available as I/O ports. This convention follows the
standard 8051 method of expanding on-chip memory. Off-chip ROM access also occurs if the EA pin is
logic 0. EA overrides all bit settings. The PSEN signal will go active (low) to serve as a chip enable or
output enable when Ports 0 & 2 fetch from external ROM.
ROM MEMORY MAP Figure 2
DATA MEMORY
Unlike many 8051 derivatives, the DS87C550 contains additional on-chip data memory. In addition to
the standard 256 bytes of data RAM accessed by direct instructions, the DS87C550 contains another 1
kbyte of data memory that is accessed using the MOVX instruction. Although physically on-chip,
software treats this area as though it was located off-chip. The 1 kbyte of SRAM is permanently located
from address 0000h to 03FFh (when enabled).
Access to the on-chip data RAM is optional under software control. When enabled by software, the data
SRAM is between 0000h and 03FFh. Any MOVX instruction that uses this area will go to the on-chip
RAM while enabled. MOVX addresses greater than 1k automatically go to external memory through
Ports 0 & 2.
When disabled, the 1k memory area is transparent to the system memory map. Any MOVX directed to
the space between 0000h and FFFFh goes to the expanded bus on Ports 0 & 2. This also is the default
condition. This default allows the DS87C550 to drop into an existing system that uses these addresses for
other hardware and still have full compatibility.
The on-chip data area is software selectable using two bits in the Power Management Register (DME1,
DME0). This selection is dynamically programmable. Thus access to the on-chip area becomes
transparent to reach off-chip devices at the same addresses. These bits have the following operation:
11 of 50
DS87C550
DATA MEMORY ACCESS CONTROL Table 3
DME1
0
0
DME0
0
1
1
1
0
1
DATA MEMORY ADDRESS
0000h - FFFFh
0000h - 03FFh
0400h - FFFFh
Reserved
0000h - 03FFh
0400h – FFFBh
FFFCh
FFFDh - FFFFh
MEMORY FUNCTION
External Data Memory *Default condition
Internal SRAM Data Memory
External Data Memory
Reserved
Internal SRAM Data Memory
Reserved - no external access
Read access to the status of lock bits
Reserved
Notes on the status byte read at FFFCh with DME1, 0 = 1, 1: bits 2-0 reflect the programmed status of the
security lock bits LB2-LB0. They are individually set to a logic 1 to correspond to a security lock bit that
has been programmed. These status bits allow software to verify that the part has been locked before
running if desired. The bits are read only.
STRETCH MEMORY CYCLE
The DS87C550 allows software to adjust the speed of off-chip data memory and/or peripheral access by
adjusting the number of machine cycles it takes to execute a MOVX instruction. The micro is capable of
performing the MOVX in as little as two machine cycles. The on-chip SRAM uses this speed and any
MOVX instruction directed internally always uses two cycles. However, the time for the instruction
execution can be stretched for slower interface to external devices. This allows access to both fast
memory and slow memory or peripherals with no glue logic. Even in high-speed systems, it may not be
necessary or desirable to perform off-chip data memory access at full speed. In addition, there are a
variety of memory mapped peripherals such as LCDs or UARTs that are slow and require more time to
access.
The Stretch MOVX function is controlled by the MD2-MD0 SFR bits in the Clock Control Register
(CKCON.2-0) as described below. They allow the user to select a Stretch value between 0 and 7. A
Stretch of 0 will result in a two-machine cycle MOVX instruction. A Stretch of 7 will result in a MOVX
of 12 machine cycles. Software can dynamically change the stretch value depending on the particular
memory or peripheral being accessed. The default stretch of one allows the use of commonly available
SRAMs without dramatically lengthening the memory access times.
Note that the STRETCH MOVX function is slightly different in the DS87C550 than in earlier members
of the high-speed microcontroller family. In all members of this family (including the DS87C550),
increasing the stretch value from 0 to 1 causes setup and hold times to be increased by 1 crystal clock
each. In older members of the family, there is no further change in setup and hold times regardless of the
number of stretch cycles selected. In the DS87C550 however, when a stretch value of 4 or above is
selected, the timing of the interface changes dramatically to allow for very slow peripherals. First, the
ALE signal is increased by 1 machine cycle. This increases the address setup time into the peripheral by
this amount. Next, the address is held on the bus for one additional machine cycle, increasing the address
hold time by this amount. The Read or Write signal is then increased by a machine cycle. Finally, the data
is held on the bus (for a write cycle) one additional machine cycle, thereby increasing the data hold time
by this amount. For every Stretch value greater than 4, the setup and hold times remain constant, and only
the width of the read or write signal is increased.
On reset, the Stretch value will default to a 1, resulting in a three-cycle MOVX for any external access.
Therefore, the default off-chip RAM access is not at full speed. This is a convenience to existing designs
that may not have fast RAM in place. Internal SRAM access is always at full speed regardless of the
12 of 50
DS87C550
Stretch setting. When maximum speed is desired, software should select a Stretch value of 0. When using
very slow RAM or peripherals, the application software can select a larger Stretch value. Note that this
affects data memory accesses only and that there is no way to slow the accesses to program memory other
than to use a slower crystal (or external clock).
The specific timing of the variable speed Stretch MOVX is provided in the Electrical Specifications
section of this data sheet. Table 4 shows the resulting MOVX instruction timing and the read or write
strobe widths for each Stretch value.
DATA MEMORY CYCLE STRETCH VALUES Table 4
CKCON.2-0
M2 M1 M0
0
0
0
0
0
1
0
1
0
0
1
1
1
0
0
1
0
1
1
1
0
1
1
1
MOVX MACHINE
CYCLES
2 (forced internal)
3 (default external)
4
5
9
10
11
12
RD OR WR STROBE WIDTH
IN MACHINE CYCLES
0.5
1
2
3
4
5
6
7
Dual Data Pointer With Inc/Dec
The DS87C550 contains several new, unique features that are associated with the Data Pointer register. In
the original 8051 architecture, the DPTR was a 16-bit value that was used to address off-chip data RAM
or peripherals. To improve the efficiency of data moves, the DS87C550 contains two Data Pointer
registers (DPTR0 and DPTR1). By loading one DPTR with the source address and the other with the
destination address, block data moves can be made much more efficient. Since DPTR0 is located at the
same address as the single DPTR in the original 8051 architecture, code written for the original
architecture will operate normally on the DS87C550 with no modification necessary.
The second data pointer, DPTR1 is located at the next two register locations (up from DPTR0) and is
selected using the data pointer select bit SEL (DPS.0). If SEL = 0, then DPTR0 is the active data pointer.
Conversely, if SEL = 1, then DPTR1 is the active data pointer. Any instruction that reference the DPTR
(ex. MOVX A, @ DPTR) refers to the active data pointer as determined by the SEL bit. Since the bit
adjacent to SEL in the DPS register is not used, the fastest means of changing the SEL (and thereby
changing the active data pointer) is with an INC instruction. Each INC DPS Instruction will toggle the
active data pointer.
Unlike the standard 8051, the DS87C550 has the ability to decrement as well as increment the data
pointers without additional instructions. When the INC DPTR instruction is executed, the active DPTR is
incremented or decremented according to the ID1, ID0 (DPS.7-6), and SEL (DPS.0) bits as shown. The
inactive DPTR is not affected.
ID1
X
X
0
1
ID0
0
1
X
X
SEL
0
0
1
1
RESULT OF INC DPTR
INCREMENT DPTR0
DECREMENT DPTR0
INCREMENT DPTR1
DECREMENT DPTR1
13 of 50
DS87C550
Another useful feature of the device is its ability to automatically switch the active data pointer after a
DPTR-based instruction is executed. This feature can greatly reduce the software overhead associated
with data memory block moves, which toggle between the source and destination registers. When the
Toggle Select bit (TSL;DPS.5) is set to 1, the SEL bit (DPS.0) is automatically toggled every time one of
the following DPTR related instructions are executed:
§
§
§
§
§
INC DPTR
MOV DPTR, #data16
MOVC A, @A+DPTR
MOVX A, @DPTR
MOVX @DPTR, A
As a brief example, if TSL is set to 1, then both data pointers can be updated with the two instruction
series shown.
INC DPTR
INC DPTR
With TSL set, the first increment instruction increments the active data pointer, and then causes the SEL
bit to toggle making the other DPTR active. The second increment instruction increments the newly
active data pointer and then toggles SEL to make the original data pointer active again.
CLOCK CONTROL and POWER MANAGEMENT
The DS87C550 includes a number of unique features that allow flexibility in selecting system clock
sources and operating frequencies. To support the use of inexpensive crystals while allowing full-speed
operation, a clock multiplier is included in the processor’s clock circuit. Also, along with the Idle and
power-down (Stop) modes of the standard 80C52, the DS87C550 provides a new Power Management
mode. This mode allows the processor to continue instruction execution at a very low speed to
significantly reduce power consumption (below even idle mode). The DS87C550 also features several
enhancements to Stop mode that make this extremely low power mode more useful. Each of these
features is discussed in detail below.
SYSTEM CLOCK CONTROL
As mentioned previously, the DS87C550 contains special clock control circuitry that simultaneously
provides maximum timing flexibility and maximum availability and economy in crystal selection. There
are two basic functions to this circuitry: a frequency multiplier and a clock divider. By including a
frequency multiplier circuit, full-speed operation of the processor may be achieved with a lower
frequency crystal. This allows the user the ability to choose a more cost-effective and easily obtainable
crystal than would be possible otherwise.
The logical operation of the system clock divide control function is shown in Figure 3. The clock signal
from the crystal oscillator (or external clock source) is provided to the frequency multiplier module, to a
divide-by-256 module, and to a 3-to-1 multiplexer. The output of this multiplexer is considered the
system clock. The system clock provides the time base for timers and internal peripherals, and feeds the
CPU State Clock Generation circuitry. This circuitry divides the system clock by 4, and it is the four
phases of this clock that make up the instruction execution clock. The four phases of a single instruction
execution clock are also called a single machine cycle clock. Instructions in the DS87C550 all use the
machine cycle as the fundamental unit of measure and are executed in from one to five of these machine
cycles. It is important to note the distinction between the system clock and the machine cycle clock as
they are often confused, creating errors in timing calculations. In performing timing calculations, it is
14 of 50
DS87C550
important to remember that all timers and internal peripherals operate off of some version of the system
clock while the instruction execution engine always operates off of the machine cycle clock.
When CD1 and CD0 (PMR.7-6) are both cleared to a logic 0, the multiplexer selects the frequency
multiplier output. The frequency multiplier can supply a clock that is 2 times or 4 times the frequency of
the incoming signal. If the times-4 multiplier is selected by setting the 4X/ 2X bit (PMR.3) to 1, for
example, the incoming signal is multiplied by 4. This 4X clock is then passed through the multiplexer,
and then output to the CPU State Clock Generation circuits. These CPU State Clock Generation circuits
always divide the incoming clock by 4 to arrive at the four states (called a machine cycle) necessary for
correct processor operation. In this example, since the clock multiplier multiplies by four and the CPU
State Clock Generation circuit divides by 4, the apparent instruction execution speed is 1 external (or
crystal oscillator) clock per instruction. If the 4X/ 2X bit is set to 0, then the apparent instruction
execution speed is 2 clocks per instruction.
It is important to note that the clock multiplier function does not increase the maximum clock (system
clock) rate of the device. The DS87C550 operates at a maximum system clock rate of 33 MHz. Therefore,
the maximum crystal frequency is 8.25 MHz when a clock multiplier of 4 is used, and is 16.5 MHz when
a clock multiplier of 2 is used. The purpose of the clock multiplier is to simplify crystal selection when
maximum processor operation is desired. Specifically, an 8.25 MHz fundamental mode, AT cut, parallel
resonant crystal is much easier to obtain than the same crystal at 33 MHz. Most crystals in that frequency
range tend to be third overtone type.
As illustrated in Figure 3, the programmable Clock Divide control bits CD1-CD0 (PMR.7-6) provide the
processor with the ability to adapt to different crystal (and external clock) frequencies and also to allow
extreme division of the incoming clock providing lower power operation when desired. The effect of
these bits is shown in Table 5.
CD1:CD0 OPERATION Table 5
CD1
0
0
1
1
CD0
0
1
0
1
Instruction Execution
Frequency multiplier (1 or 2 clocks per machine cycle)
Reserved
Clock divided by 4 (4 clocks per machine cycle) Default
Clock divided by 1024 (1024 clocks per machine cycle)
Besides the ability to use a multiplied clock signal, the normal mode of operation, i.e. the reset default
condition (CD1 = 1, CD0 = 0) passes the incoming crystal or external oscillator clock signal straight
through as the system clock. Because of the CPU State Clock generation circuitry’s normal divide-by-4
function, the default execution speed of the DS87C550’s basic instruction is one-fourth the clock
frequency.
The selection of instruction cycle rate takes effect after a delay of one machine cycle. Note that the clock
divider choice applies to all functions including timers. Since baud rates are altered, it may be difficult to
conduct serial communication while in divide-by-1024 mode. This is simplified by the use of switchback
mode (described later) included on the DS87C550.
15 of 50
DS87C550
CLOCK SWITCHING RESTRICTIONS
To ensure clean “glitch-free” switching of the system clock and to ensure that all clocks are running and
stable before they are used, there are minor restrictions on accessing the clock selection bits CD1:0 and
the 4X/ 2X bit.
One restriction is that any change in the CD1 and CD0 bits from a condition other than a 1 0 state (i.e.,
clock divided by 4 mode) must pass through the divide-by-4 state before proceeding to the desired state.
As a specific example, if the clock divisor bits are set to use the frequency multiplier in 4X mode, no
other clock setting is possible until after the CD1:0 bits are set to divide-by-4 mode. After setting clock
divided-by-4 mode, then clock divided by 1024 can be selected by setting CD1 and CD0 to “11b”. Any
attempt to change these bits to a disallowed state will be ignored by the hardware.
There are also some minor restrictions when changing from one clock multiplier to another. Changing the
clock multiplier can only be performed when the Crystal Multiplier Enable bit CTM (PMR.4) is set to 0.
This bit disables the clock multiplication function. However, the CTM bit can only be changed when CD1
and CD0 are set to divide-by-4 mode (i.e., “10b”) and the ring mode (RNGMD = RCON.2) bit is 0
(discussed later). Changing the clock multiplication factor also requires that the new frequency be stable
prior to effecting the change. The SFR bit CKRDY (RCON.3) indicates the state of the stabilization
timeout. Setting the CTM bit to a 0 from a 1 disables the clock multiplier function, automatically clears
the CKRDY bit, and starts the stabilization timeout.
SYSTEM CLOCK CONTROL Figure 3
During the stabilization period, CKRDY will remain low, and software will be unable to set the CD1:0
bits to select the frequency multiplier. After the stabilization delay, CKRDY will be set to a 1 by
hardware. Note that this bit cannot be set to 1 by software. After hardware sets CKRDY bit, then the
CD1:0 bits can be set to use the clock multiplier function. However, before changing CD1:0, the 4X/ 2X
bit must be set to the desired state. Following this, the CTM bit must be set to 1 to enable the crystal
multiplier. Finally the CD1:0 bits may be set to select the crystal multiplier function. By following this
procedure, the processor is guaranteed to receive a stable, glitch-free clock.
OSCILLATOR-FAIL DETECT
The DS87C550 contains a unique safety mechanism called an on-chip Oscillator-Fail Detect circuit.
When enabled, this circuit causes the processor to be reset if the oscillator frequency falls below TBD
kHz. The processor is held in reset until the oscillator frequency rises above TBD kHz. In operation, this
circuit can provide a backup for the watchdog timer. Normally, the watchdog timer is initialized so that it
will timeout and will cause a processor reset in the event that the processor loses control. This works
perfectly as long as there is a clock from the crystal or external oscillator, but if this clock fails, there is
the potential for the processor to fail in an uncontrolled and possibly undesirable state. With the use of the
oscillator-fail detect circuit, the processor will be forced to a known state (i.e., reset) even if the oscillator
stops.
16 of 50
DS87C550
The oscillator-fail detect circuitry is enabled by software setting the enable bit OFDE (PCON.4) to a 1.
Please note that software must use a “Timed Access” procedure (described later) to write to this bit. There
is an oscillator-fail detect flag, OFDF (PCON.5), that is set to a 1 by the hardware when it detects an
oscillator failure. The processor will be forced into a reset state when this occurs if enabled by OFDE.
The oscillator-fail detect flag can only be cleared to a 0 by a power-up reset or by software. It should be
noted that the oscillator-fail detect circuitry is not disabled by entering Stop mode. There fore, the user
must ensure that this feature is disabled before entering Stop mode.
POWER MANAGEMENT MODE (PMM)
Power Management Mode offers a complete scheme of reduced internal clock speeds that allow the CPU
to run software but to use substantially less power. Normally, during default operation, the DS87C550
uses 4 clocks per machine cycle. Thus the instruction cycle (machine cycle clock) rate is Clock/4. At 33
MHz crystal speed, the instruction cycle speed is 8.25 MHz. In PMM the microcontroller operates, but
from an internally divided version of the clock source. This creates a lower power state without external
components. As shown in Figure 3, the system clock may be selected to use the crystal (or external
oscillator) frequency divided by 256. This produces a machine cycle that consists of the crystal frequency
divided by 1024, which is considered Power Management Mode (PMM). With the processor executing
instructions at this much lower rate, a significant amount of power is saved.
Software is the only mechanism to invoke the PMM. Table 6 illustrates the instruction cycle rate in
PMM for several common crystal frequencies. Since power consumption is a direct function of operating
speed, PMM runs very slowly and provides the lowest power consumption without stopping the CPU.
This is illustrated in Table 7.
MACHINE CYCLE RATE Table 6
Crystal Speed
Full Operation
(4 clocks per machine cycle)
PMM
(1024 clocks per machine cycle)
11.0592 MHz
2.765 MHz
10.8 kHz
16 MHz
4.0 MHz
15.6 kHz
25 MHz
6.25 MHz
24.4 kHz
33 Mhz
8.25 MHz
32.2 kHz
OPERATING CURRENT ESTIMATES IN PMM Table 7
Crystal Speed
Full Operation
(4 clocks per machine cycle)
PMM
(1024 clocks per machine cycle)
11.0592 MHz
13.1 mA
4.8 mA
16 MHz
17.2 mA
5.6 mA
25 MHz
25.7 mA
7.0 mA
33 Mhz
32.8 mA
8.2 mA
Note that PMM provides a lower power condition than Idle mode. This is because in Idle, all clocked
functions such as timers run at a rate of crystal divided by 4. Since wakeup from PMM is as fast as or
17 of 50
DS87C550
faster than wakeup from Idle, and since PMM allows the CPU to continue to execute instructions (even if
doing NOPs), there is little reason to use Idle in new designs.
Switchback
One of the other unique features included on the DS87C550 is Switchback. Simply, Switchback when
enabled will allow serial ports and interrupts to automatically switch back from divide-by-1024 (PMM) to
divide-by-4 (standard speed operation). This feature makes it very convenient to use the Power
Management Mode in real time applications. Of course to return to a divide-by-4 clock rate from divideby-1024 PMM, software can simply select the CD1 & CD0 clock control bits to the 4 clocks per cycle
state. However, the DS87C550 provides hardware alternatives for automatic Switchback to standard
speed operation.
The Switchback feature is enabled by setting the SFR bit SWB (PMR.5) to a 1. Once it is enabled and
when PMM is selected, there are two possible events that can cause an automatic switchback to divideby-4 mode. First, if an interrupt occurs and is set so that it will be acknowledged, this event will cause
the system clock to revert from PMM to divide-by-4 mode. For example, if INT0 is enabled then
Switchback will occur on INT0 . However, if INT0 is not enabled, then activity on INT0 will not cause
switchback to occur.
A Switchback can also occur when an enabled UART detects the start bit indicating the beginning of an
incoming serial character or when the SBUF register is loaded initiating a serial transmission. Note that a
serial character’s start bit does not generate an interrupt. This occurs only on reception of a complete
serial word. The automatic Switchback on detection of a start bit allows hardware to correct baud rates in
time for a proper serial reception or transmission. So with Switchback enabled and a serial port enabled,
the automatic switch to normal speed operation occurs automatically in time to receive or transmit a
complete serial character as if nothing special had happened.
Once Switchback causes the processor to make the transition back to divide-by-4 mode, software must
modify SFR bits CD1 & CD0 to re-enter Power Management Mode. However, if a serial port is in the
process of transmitting or receiving a character, then this change back to PMM will not be allowed as the
hardware prevents a write to CD1 & CD0 during any serial port activity.
Since the reception of a serial start bit or an interrupt priority lockout is normally undetectable by
software in an 8051, the Status register features several new flags that are useful. These are described
below.
Status
Information in the Status register assists decisions about switching into PMM. This register contains
information about the level of active interrupts and the activity on the serial ports.
The DS87C550 supports three levels of interrupt priority. These levels are Power-fail, High, and Low.
Status bits STAT.7-5 indicate the service status of each level. If PIP (Power-fail Interrupt Priority;
STATUS.7) is a 1, then the processor is servicing this level. If either HIP (High Interrupt Priority;
STATUS.6) or LIP (Low Interrupt Priority; STATUS.5) is high, then the corresponding level is in
service.
Software should not rely on a lower priority level interrupt source to remove PMM (Switchback) when a
higher level is in service. Check the current priority service level before entering PMM. If the current
service level locks out a desired Switchback source, then it would be advisable to wait until this condition
clears before entering PMM.
18 of 50
DS87C550
Alternately, software can prevent an undesired exit from PMM by entering a low priority interrupt service
level before entering PMM. This will prevent other low priority interrupts from causing a Switchback.
Status also contains information about the state of the serial ports. Serial Port Zero Receive Activity
(SPRA0; STATUS.0) indicates a serial word is being received on Serial Port 0 when this bit is set to a 1.
Serial Port Zero Transmit Activity (SPTA0; STATUS.1) indicates that the serial port is still shifting out a
serial transmission. STATUS.2 (SPRA1) and STATUS.3 (SPTA1) provide the same information for
Serial Port 1, respectively. While one of these bits is set, hardware prohibits software from entering PMM
(CD1 & CD0 are write-protected) since this would corrupt the corresponding serial transmissions.
IDLE MODE
Setting the LSB of the Power Control register (PCON.0) invokes the Idle mode. Idle will leave internal
clocks, serial ports and timers running. Power consumption drops because memory is not being accessed
and instructions are not being executed. Since clocks are running, the Idle power consumption is a
function of crystal frequency. It should be approximately ½ of the operational power at a given
frequency. The CPU can exit the Idle state with any interrupt or a reset. Idle is available for backward
software compatibility. However, due to improvements over the original architecture, the processor’s
power consumption can be reduced to below Idle levels by invoking Power Management Mode (PMM)
and running NOPs.
STOP MODE
Setting bit 1 of the Power Control register (PCON.1) invokes the Stop mode. Stop mode is the lowest
power state (besides power-off) since it turns off all internal clocking. The I CC of a standard Stop mode is
approximately 1 uA (but is specified in the Electrical Specifications). All processor operation ceases at
the end of the instruction that sets PCON.1. The CPU can exit Stop mode from an external interrupt or a
reset condition. Internally generated interrupts (timer, serial port, etc.) are not useful since they require
clocking activity.
BAND-GAP SELECT
The DS87C550 provides two enhancements to the Stop mode. As described below, the DS87C550
provides a band-gap reference to determine Power-fail Interrupt and Reset thresholds. The default state is
that the band-gap reference is off while in Stop mode. This mode allows the extremely low-power state
mentioned above. A user can optionally choose to have the band-gap enabled during Stop mode. With the
band-gap reference enabled, PFI and Power-fail Reset are functional and are valid means for leaving Stop
mode. This allows software to detect and compensate for a brownout or power supply sag, even when in
Stop mode.
In Stop mode with the band-gap enabled, I CC will be approximately 100 uA compared with 1 uA with the
band-gap off. If a user does not require a Power-fail Reset or Interrupt while in Stop mode, the band-gap
can remain disabled. Only the most power-sensitive applications should turn off the band-gap, as this
results in an uncontrolled power-down condition.
The control of the band-gap reference is located in the Ring Oscillator Control Register (RCON). Setting
BGS (RCON.0) to a 1 will keep the band-gap reference enabled during Stop mode. The default or reset
condition is with the bit at a logic 0. This results in the band-gap being off during Stop mode. Note that
this bit has no control of the reference during full power, PMM, or Idle modes.
19 of 50
DS87C550
RING OSCILLATOR
The second enhancement to Stop mode on the DS87C550 allows an additional power saving option while
also making Stop easier to use. This is the ability to start instantly when exiting Stop mode. It is the
internal ring oscillator that provides this feature. This ring can be a clock source when exiting Stop mode
in response to an interrupt. The benefit of the ring oscillator is as follows.
Entering Stop mode turns off the crystal oscillator and all internal clocks to save power. When exiting
Stop mode, the external crystal may require up to 10 ms to begin oscillating again. The DS87C550 can
eliminate that delay through the use of the internal ring oscillator, resuming operation in less than 100 ns
when exiting Stop mode. If a user selects the ring to provide the start-up clock and the processor remains
running, hardware will automatically switch to the crystal once a power-on reset interval (65536 crystal
clocks) has expired.
The ring oscillator runs at approximately 4 MHz but will not be a precise value. Do not conduct real-time
precision operations (including serial communication) during this ring period. The default state is to exit
Stop mode without using the ring oscillator, so action to enable the ring must be taken before entering
stop mode.
The Ring Select (RGSL) bit in the RCON register (RCON.1) controls this function. When RGSL = 1, the
CPU will use the ring oscillator to exit Stop mode quickly. As mentioned above, the processor will
automatically switch from the ring to the crystal after a delay of 65,536 crystal clocks. For a 3.57 MHz
crystal, this is approximately 18 ms. The processor sets a flag called Ring Mode (RGMD = RCON.2) that
tells software that the ring is being used. The bit will be a logic 1 when the ring is in use.
TIMED ACCESS PROTECTION
Selected SFR bits are critical to operation, making it desirable to protect them against an accidental write
operation. The Timed Access procedure prevents an errant processor from accidentally altering a bit that
would seriously affect processor operation. The Timed Access procedure requires that the write of a
protected bit be preceded by the following instructions:
MOV 0C7h, #0AAh
MOV 0C7h, #55h
By writing an AAh followed by a 55h to the Timed Access register (location C7h), the hardware opens a
three-cycle window that allows software to modify one of the protected bits. If the instruction that seeks
to modify the protected bit is not immediately preceded by these instructions, the write will not take
effect. The protected bits are:
WDCON.6
POR
Power-On Reset Flag
WDCON.3
WDIF
Watchdog Interrupt Flag
WDCON.1
EWT
Watchdog Reset Enable
WDCON.0
RWT
Reset Watchdog Timer
RCON.0
BGS
Band-Gap Select
ROMSIZE.2 RMS2
Program Memory Select
Bit 2
20 of 50
DS87C550
ROMSIZE.1 RMS1
Program Memory Select
Bit 1
ROMSIZE.0 RMS0
Program Memory Select
Bit 0
EMI REDUCTION
One of the major contributors to radiated noise in an 8051-based system is the toggling of ALE. The
DS87C550 allows software to disable ALE when not used by setting the ALEOFF (PMR.2) bit to a 1.
When ALEOFF = 1, ALE will still toggle during an off-chip MOVX. However, ALE will remain inactive
when performing on-chip memory access. The default state is ALEOFF = 0 so ALE normally toggles at a
frequency of XTAL/4.
PERIPHERAL OVERVIEW
The DS87C550 provides several of the most commonly needed peripheral functions in microcomputerbased systems. New functions include a second serial port, power-fail reset, power-fail interrupt flag, and
a programmable watchdog timer. In addition, the DS87C550 contains an analog-to-digital converter and
four channels of pulse width modulation for industrial control and measurement applications. Each of
these peripherals is described below. More details are available in the High-Speed Micro Data Book (or
its most recent addendum).
SERIAL PORTS
The DS87C550 provides a serial port (UART) that is identical to the 80C52. In addition, it includes a
second hardware serial port that is a full duplicate of the standard one. This port optionally uses pins P1.6
(RXD1) and P1.7 (TXD1). It has duplicate control functions included in new SFR locations.
Both ports can operate simultaneously but can be at different baud rates or even in different modes. The
second serial port has similar control registers (SCON1, SBUF1) to the original. The new serial port can
only use Timer 1 for timer generated baud rates.
Control for serial port 0 is provided by the SCON0 register while its I/O buffer is SBUF0. The registers
SCON1 and SBUF1 provide the same functions for the second serial port. A full description of the use
and operation of both serial ports may be found in the “High-Speed Microcontroller Data Book.”
ANALOG TO DIGITAL CONVERTER
The DS87C550 contains a 10-bit successive approximation analog-to-digital converter. This converter
provides eight multiplexed channels of analog input and allows the user to select either an external or
internal precision voltage reference to be used for the conversion process. The A/D converter provides
true 9-bit accuracy with a total error of less than ±2 LSBs.
The A/D Converter may be disabled to conserve power by writing a 0 to the SFR ADON bit
(ADCON1.1). At reset, this is the default condition, and the user must write a 1 to this bit prior to using
the A/D Converter.
A/D CONVERTER INPUT
The A/D Converter of the DS87C550 provides eight channels of analog input on device pins ADC7
through ADC0 (P5.7-P5.0). The signals on these pins are input into an analog multiplexer. The magnitude
(and polarity) of these signals is limited by the reference voltage used by the converter (see DC
21 of 50
DS87C550
Specifications section). This reference voltage may be selected to be either an internal band-gap voltage
(VBG) or an external reference ( Avref+, Avref-). This selection is made by writing a 0 (uses internal
reference VBG) or writing a 1 (uses external reference ( Avref+, Avref-) to the ADRS bit (PWMADR.7).
The default reset condition is for the internal reference to be selected.
Selecting a single analog signal for conversion is achieved by software writing the desired channel
number (0 through 7) into SFR bits MUX2 through MUX0 bits of the A/D Control Register 2
(ADCON2.6-4). The single output of the multiplexer is then provided to a sample and hold circuit that
maintains a steady signal during the conversion process.
A/D CONVERSION PROCESS
The A/D conversion process can be configured for one-shot or continuous mode operation. For one-shot
operation, the SFR bit CONT/SS (ADCON1.5) must be a 0. The conversion process is then initiated by
software writing a 1 to the STRT/BSY SFR bit (ADCON1.7) if the ADEX (ADCON1.4) bit is a 0. If the
ADEX bit is a 1, then the conversion is initiated by an active low signal on the external pin STADC
(P6.7). If continuous mode is selected (CONT/SS = 1), then the first conversion is initiated as described
above, but another conversion will be automatically started at the completion of the previous conversion.
Once initiated, the conversion process requires 16 A/D clock periods (T ACLK) to complete. Because of the
dynamic nature of the converter, the A/D clock period can be no less that 1 us and no more than 6.25 us.
This requirement is expressed as follows:
1.0 us <= TACLK <= 6.25 us
Therefore any single conversion time can range from 16 us minimum to 100 us maximum, depending on
the selected A/D clock frequency.
The A/D clock frequency is a function of the processor’s machine cycle clock and the A/D clock’s
prescaler setting as shown by the following equation:
TACLK = TMCLK * (N+1)
where N is the prescaler setting in APS3:0.
The processor’s machine cycle clock period (T MCLK) is normally the external crystal (or oscillator)
frequency multiplied by 4 (but can be affected by the CD1, CD0, and 4X/ 2X bits). The A/D clock period
must be set by the user to ensure that it falls within the minimum and maximum values specified above.
As an example, assume the processor’s crystal frequency is 33 MHz and that the processor is running in a
standard divide-by-4 mode. This means that the period of the processors machine cycle clock, i.e.,
TMCLK, will be (1/33 MHz)*4 or 121.2 ns. If it is assumed that the application requires the fastest possible
conversion time, then the desired T ACLK is 1.0 us. The necessary prescale value can then be calculated as:
N = (TACLK/TMCLK)-1
Therefore for this example, N = 7.25. Since N must be an integer, the value of N must be 8 (rounded up
to the next integer). This results in a conversion clock T ACLK = 1.091 us.
The prescaler value must be stored in SFR bits APS3 through APS0 (ADCON2.3-0) to achieve the proper
A/D clock. These bits default to 0 on a processor reset, so they must be set as desired by the processor’s
initialization software.
22 of 50
DS87C550
A/D OUTPUT
There are two SFR locations that contain the result of the A/D conversion process. They are ADMSB
(most significant byte) and ADLSB (least significant byte). The ADLSB byte always contains the 8 least
significant bits of the 10-bit result. The ADMSB can be configured in two different ways through the use
of the SFR bit OUTCF (ADCON2.7). If OUTCF is a 0, then ADMSB contains the 8 most significant bits
of the 10-bit conversion (i.e., bits 9-2). If OUTCF is a 1, then ADMSB contains A/D output bits 9-8 (right
justified). The upper 6 bits of the register are set to 0 in this case.
The value stored in the output registers is given by the following equation:
1024 x ((Vin-Avref-)/(Avref+ - Avref-))
This equation shows that the A/D conversion result is a 10-bit binary number that represents what fraction
of the available reference voltage the input signal is. As you can see with a reference voltage of 2.5 volts,
the output has a resolution of 2.44 millivolts. This shows that the reference voltage must be very well
regulated to ensure satisfactory performance. It should be noted that the output of the A/D conversion
process will be “0000000000” for voltages from Avref- to (Avref- +1/2 LSB). In addition, “1111111111”
will be output for voltages from (Avref+ - 3/2 LSB) to Avref+.
The DS87C550 offers a unique feature that allows the result of an A/D conversion to be compared with
two user-defined values stored in the WINHI and WINLO registers. The results of this comparison will
set or clear the WCM (ADCON 1.2) bit, and this bit can be used as a qualifier to the A/D interrupt. This
comparison is built into hardware so that this feature is performed without any burden on the software,
and A/D results that are not of particular interest to the application can be ignored. Special function
registers WINHI and WINLO are loaded by application software with 8-bit numbers that are compared
with the 8 MSBs of the A/D result. These user-defined numbers form a range of values, and the A/D
result is evaluated to be inside or outside of this range. When WCIO (ADCON.1) is 0, then WCM is set if
the A/D result is found to be inside the range. Otherwise WCM is cleared. When WCIO is a 1, then
WCM is set if the A/D result is found to be outside the range. Otherwise WCM is cleared. The state of the
WCM bit is expressed by the following equation:
WCM = WCIO ⊕ (WINHI ≤ADMSB) ⊕ (WINLO ≤ADMSB)
This equation precisely identifies the relationship between the window registers (WINHI and WINLO),
the MSB of the A/D conversion (ADMSB), and the WCIO and WCM bits. However by observation, it is
not particularly intuitive as to how this interaction works in a practical sense. If the user makes the
assumption that the value stored in WINHI is greater that the value stored in WINLO (this is normally but
not necessarily the case), then this equation can be simplified to the following two cases:
For WCIO = 0: WCM = (WINHI > ADMSB) AND (ADMSB ≥ WINLO)
For WCIO = 1: WCM = (WINHI ≤ADMSB) OR (ADMSB < WINLO)
It is clear that these two equations now express the cases where the A/D result is inside the comparison
window (WCIO = 0) and outside the comparison window (WCIO = 1). It is important to note the ≤and
≥ symbols and account for the specific values that are included in the comparison.
There is another SFR bit, WCQ, that further defines the action taken when the WCM is set. If WCQ is 0,
then an A/D Interrupt will occur (if enabled) regardless of the comparison results. When WCQ is set to a
23 of 50
DS87C550
1, then an A/D Interrupt will only occur if WCM is set (i.e., the A/D result comparison was true). This
feature allows software to respond only to conditions that meet the programmed range.
PULSE WIDTH MODULATION
The DS87C550 contains four independent 8-bit pulse width modulator ( PWMs) functions each with
independently selectable clock sources. For more precise modulation operations, two 8-bit PWM
functions (PWM0 & PWM1 and/or PWM2 & PWM3) can be cascaded together to form a 16-bit PWM
function.
The PWM function is divided into three major blocks: a clock prescaler, a clock generator, and a pulse
generator. A single prescaler provides selectable clocks of different frequencies to each of the four clock
generator blocks. Each clock generator is an 8-bit reloadable counter that determines the repetition rate
(frequency) of its associated PWM. Each pulse generator PWM block is an 8-bit timer clocked by the
clock generator’s output. When this timer reaches zero, the output of the PWM is set to 1. When the timer
reaches the user selected PWM match value stored in SFR PWMx, the PWM output is cleared to 0. In
this way, the frequency and duty cycle of the PWM is varied under software control.
PWM PRESCALER
The prescaler block of the PWM function accepts as a clock input the system clock provided to the CPU
(and other peripherals), and divides it by 1, 4, 16, and 64. Each of these clocks is available at the output
of the prescaler, and is provided to all four of the PWM clock generator blocks. The actual clock used by
the clock generator block is dependant on the setting of SFR bits PWxS2:0 (where x is the PWM channel
number 0-3) located in the PW01CS or PW23CS registers. In addition to selecting one of the prescaler’s
CPU clock divided outputs, setting PWxS2 to a 1 allows an external clock to be used as an input to the
clock generators. The external clocks are input on device pins PWMC0 (P6.4 for PWM0 or PWM1) or
PWMC1 (P6.5 for PWM2 or PWM3). Like all other inputs to the 8051, these inputs are synchronized by
sampling them using the internal machine cycle clock. Therefore these inputs must be of sufficient
duration for the clock to sample them properly (i.e., 2 machine cycles). The complete functionality of the
clock selection SFR bits is as follows:
Prescaler Output
PWxS2:0
Machine Cycle_Clock/1
000
Machine Cycle_Clock/4
001
Machine Cycle_Clock/16
010
Machine Cycle_Clock/64
011
PWMCx (external)
1xx
In determining the exact frequency output of the prescaler, it is important to note that the machine cycle
clock provided to the prescaler is also software-selectable. The machine cycle clock can be the crystal (or
oscillator frequency) divided by 1, 2, 4, or 1024 as determined by the CD1:0 and the 4X/ 2X SFR bits (see
Clock Divide Control section for details).
PWM CLOCK GENERATOR
The clock generator blocks of the PWM modules are pre-loaded by software with an 8-bit value, and this
value determines the frequency or repetition rate of the PWM function. A value of 0 causes the selected
24 of 50
DS87C550
output of the prescaler to be passed directly to the pulse generator function (i.e., divide by 1). A value of
FFh passes a clock to the pulse generator function that is the selected prescaler output divided by 256. In
general, the clock generators provide a divide by N+1 selectable repetition rate (i.e., frequency) for their
PWM channel.
Each clock generator has an associated SFR that contains the 8-bit reload value. These registers are called
PW0FG, PW1FG, PW2FG, and PW3FG (see SFR map for addresses). In addition, there is a frequency
generator enable bit (PW0EN, PW1EN, PW2EN, & PW3EN) for each of the clock generator blocks that
must be set to a 1 before these blocks will function. These bits are set to 0 after all resets so software must
set them to 1 to enable the PWM clocks.
The output of the clock generator block is supplied to the input of the pulse generator block.
PWM PULSE GENERATOR
The pulse generator block of the PWM function produces the PWM output signal on device pins
PWMO0 (P6.0), PWMO1(P6.1), PWMO2 (P6.2), AND PWMO3 (P6.3). Each of these output bits has an
enable bit: PW0OE (PW01CON.5), PW1OE (PW01CON.1), PW2OE (PW23CON.5), and PW3OE
(PW23CON.1) that are cleared to 0 on all resets, and must be set to 1 by software before the PWMs will
output a signal.
As described earlier, the pulse generator block is basically a free-running timer with a comparison register
that is loaded with an 8-bit value by software. The value of this register establishes the duty cycle of the
PWM function. The comparison values are stored in SFRs PWM0, PWM1, PWM2, and PWM3 for the
respective PWM channels, and it is these values that determine the pulse duration.
Actually, in accessing these specific SFRs, software has access to both the compare registers and the
timer registers of the pulse generator blocks. When the PWM Timer/Compare Value Select SFR bits
PW0T/C (PW01CON.4), PW1T/C (PW01CON.0), PW2T/C (PW23CON.4), and PW3T/C
(PW23CON.0) are cleared to 0, a read or write to the respective PWMx register accesses the compare
register. When these bits are set to 1, a read or write accesses the timer value. With the use of these bits,
the timers in the pulse generator sections of the PWM functions can be used as general purpose timers if
desired.
When the free-running timer of the pulse generator block rolls over from FFh to 00h, the PWM’s output
is set to a 1. As the timer continues to count up from 0, the output of the PWM is cleared to 0 when the
timer value is equal to the comparison register value. This cycle continues automatically without
processor intervention until software or a reset changes some condition.
The value of 0 in the comparison register is a special case of each PWM function. Rather than allow a set
and a reset of the PWM output bit, special hardware ensures that 0 will be output continuously if 0 is
loaded into the compare register.
There are other SFR bits that affect PWM operation for special modes. Bits PW0DC (PW01CON.6),
PW1DC (PW01CON.2), PW2DC (PW23CON.6), and PW3DC (PW23CON.2) cause the output of the
respective PWM function to be a constant 1. This feature may be useful for driving a fixed DC voltage
into any circuitry attached to the PWM output. Bits PW0F (PW01CON.7), PW1F (PW01CON.3), PW2F
(PW23CON.7), and PW3F (PW23CON.3) are flags that are set by the hardware when the respective
PWM pulse generator timer rolls over from FFh to 0. These flags must be cleared by software to remove
their set condition.
25 of 50
DS87C550
16-BIT MODE
For more precise PWM operations, two 8-bit PWMs may be combined into a single 16-bit PWM
function. By setting SFR bit PWE0 (PWMADR.0) to a 1, a new 16-bit PWM0 function is formed from
the 8-bit PWM functions PWM0 (LSB) and PWM1 (MSB). Similarly, by setting PWE1 (PWMADR.1) to
a 1, a new 16-bit PWM1 function is formed from PWM2 (LSB) and PWM3 (MSB). Since each pair of
PWMs can be independently configured into a 16-bit arrangement, the user has the option of having four
8-bit PWM functions, two 8-bit PWM functions and a 16-bit PWM function, or two 16-bit PWM
functions.
In 16-bit PWM mode, the prescaler operates exactly as it did in 8-bit mode. Its outputs are available to all
four clock generator blocks. However in 16-bit mode, the clock generators for 8-bit PWM1 and PWM3
are not functional. The clock for 16-bit PWM0 function is provided by the clock generator for 8-bit
PWM0 and the clock for 16-bit PWM1 function is provided by the clock generator for 8-bit PWM2. The
SFR bits PW0EN (clock generator enable) and PW0S2:0 (clock select bits) provide the programmable
clock controls for 16-bit PWM channel 0, and bits PW2EN and PW2S2:0 provide the programmable
clock controls for 16-bit PWM channel 1. The clock divisor values for the 16-bit PWM operating
frequency are contained in the PW0FG and PW2FG registers for 16-bit PWM0 and PWM1
(respectively). Note that these registers remain 8-bit values so the clock division remains the same for
16-bit and 8-bit operation.
When in 16-bit mode, the two 8-bit pulse generator timers are concatenated together forming a 16-bit
timer. Therefore the pulse generator section of a 16-bit PWM channel has a repetition rate of the input
clock divided by 65,536. As in 8-bit mode when the counter reaches 0, the output of the 16-bit PWM
channel is set (i.e., logic 1), and when it reaches the pre-loaded match value it is cleared (i.e., logic 0).
GENERAL PURPOSE TIMERS/COUNTERS
The DS87C550 contains three general-purpose timer/ counters. Timers 0 and 1 are standard 8051 16-bit
timer/counters with three modes of operation. Each of these devices can be used as a 13-bit timer/counter,
16-bit timer/counter or 8-bit timer/counter with auto-reload. Timer 0 can also operate as two 8-bit timer
counters. Each timer can also be used as a counter of external pulses on the corresponding T0 or T1 pin.
The mode of operation is controlled by the Timer Mode (TMOD) register. Each timer/counter consists of
a 16-bit register in 2 bytes, which can be found in the SFR map as TL0, TH0, TL1, and TH1. These two
timers are enabled by the Timer Control (TCON) register. A complete description of use and operation of
these timers may be found in the “High-Speed Microcontroller Data Book.”
Timer 2 is a true 16-bit timer/counter with several additional features as compared to timers 1 and 0. With
a 16-bit reload register (RLOADL, RLOADH), it provides up/down auto-reload timer/counters and timer
output clock generation. Timer 2 also supports a capture/compare function. This new feature provides
additional timing control capabilities for real-time applications with less CPU intervention. A more
detailed description of this capture/compare feature is provided below.
TIMER 2
The selection of a timer or counter function is controlled by the C/ T2 (T2CON.1) bit When C/ T2 is set to
1, Timer 2 acts as a counter where it counts 1 to 0 transitions on the T2 pin. When C/ T2 is cleared to a 0,
Timer 2 functions as a timer where it counts the system clock as determined by the T2M bit (CKCON.5)
and the clock divide control bits CD1, CD0 (PMR.7:6) and the 4X/ 2X (PMR.3) bit. A prescaler is used to
further divide the input clock by a programmable ratio. The prescaler value is programmable to divide by
1, 2, 4, and 8 as defined by the T2P1 and T2P0 (T2SEL.1:0) bits. Timer 2 is enabled by setting bit TR2
(T2CON.2) to a 1, and disabled by clearing it to a 0.
26 of 50
DS87C550
When the LSB of timer/counter 2 (TL2) overflows, flag TF2B (T2SEL.4) is set, and flag TF2 (T2CON.7)
is set when the high byte (TH2) overflows. Setting flag TF2 also sets flag TF2B. Even though only one
interrupt is available for Timer 2, either or both of these overflows can be programmed to request an
interrupt. To enable the interrupt, the Timer 2 interrupt enable bit ET2 (EIE.7) must be set to a 1. The 8bit overflow interrupt or the 16-bit overflow interrupt is then individually enabled by setting TF2BS
(T2SEL.6) or TF2S (T2SEL.7). Since there is only one interrupt vector for both possible Timer 2
interrupts, the interrupt service routine must determine which event caused the interrupt by polling the
available flags. For both interrupt flags, software must clear them upon servicing the interrupt. There is
no automatic hardware clearing of these flags.
TIMER 2 CAPTURE FEATURE
One of the new features added to Timer 2 is the capture function. The output of Timer 2 is available to
four independent 16-bit capture register pairs (CPTH3:CPTL3, CPTH2:CPTL2, CPTH1:CPTL1, &
CPTH0:CPTL0). These registers are loaded with the 16-bit value contained in Timer 2 when transitions
occur on the corresponding input pin INT5/CT3, INT4/CT2, INT3/CT1 or INT2/CT0 (P1.3, P1.2, P1.1,
or P1.0) respectively. When the capture function is not being used, these input pins also serve as external
interrupt inputs. The Capture Trigger Control register (CTCON) can be programmed to make the capture
occur on a rising edge, a falling edge, or on either a rising or a falling edge on these input pins. The
functionality of the CTCON register is illustrated below. Note that the edge sensitivity established by the
setting of CTCON bits applies to both the capture function and the external interrupt function of these
input pins. This addition allows maximum flexibility in selecting interrupt polarity. Whether these input
pins are used as external interrupt inputs or as capture commands, the input will set the appropriate flag in
the External Interrupt Flag register (T2IR.3:0) and will create an interrupt if the associated enable in the
Extended Interrupt Enable (EIE.3:0) register is set.
CTCON REGISTER FUNCTIONALITY
CTCON.7
CTCON.6
CTCON.5
CTCON.4
CTCON.3
CTCON.2
CTCON.1
CTCON.0
CT3
CT3
CT2
CT2
CT1
CT1
CT0
CT0
Capture register 3 triggered by a falling edge on INT5/CT3
Capture register 3 triggered by a rising edge on INT5/CT3
Capture register 2 triggered by a falling edge on INT4/CT2
Capture register 2 triggered by a rising edge on INT4/CT2
Capture register 1 triggered by a falling edge on INT3/CT1
Capture register 1 triggered by a rising edge on INT3/CT1
Capture register 0 triggered by a falling edge on INT2/CT0
Capture register 0 triggered by a rising edge on INT2/CT0
TIMER 2 COMPARE FEATURE
Another new feature added to Timer 2 capabilities is the compare function. Prior to enabling this
function, the associated compare register pair (CMPH0:CMPL0, CMPH1:CMPL1, CMPH2:CMPL2) is
loaded by software with a 16-bit number. Each time Timer 2 is incremented, the contents of these
registers are compared with the new value of the timer. When a match occurs, the corresponding interrupt
flag (T2IR.6:4) is set to a 1 on the next machine cycle and an interrupt will occur if the corresponding
enable bit is set in the Extended Interrupt Enable (EIE.6:4) register. When a match with CMPH0:CMPL0
occurs, port pins P4.0 through P4.5 are set to a 1 if the corresponding bits of the Set Enable register
(SETR) are at logic 1. If the match is with CMPH1:CMPL1, port pins P4.0 through P4.5 are reset to 0
when the corresponding bits in the reset/toggle enable register RSTR are at logic 1. A match with
CMPH2:CMPL2 toggles port pins P4.6 and 4.7 if the corresponding bits in the RSTR register are at logic
1. Note that for the toggle function it is not the port pin latch that is actually toggled. Instead, separate
flip-flops output the SFR bits TGFF1 and TGFF0 that actually determine the state of the respective port
27 of 50
DS87C550
pin. A 0 in a bit position in either the SETR or the RSTR register disables the corresponding port pin
function. The functionality of the SETR and RSTR registers is shown below.
SETR REGISTER FUNCTIONALITY
SETR.7
SETR.6
SETR.5
SETR.4
SETR.3
SETR.2
SETR.1
SETR.0
TGFF1
TGFF0
CMS5
CMS4
CMS3
CMS2
CMS1
CMS0
This bit toggles if CMPH2:CMPL2 and Timer 2 match and CMTE1 is 1
This bit toggles if CMPH2:CMPL2 and Timer 2 match and CMTE0 is 1
If 1 then P4.5 is set on a match between CMPH0:CMPL0 and Timer 2
If 1 then P4.4 is set on a match between CMPH0:CMPL0 and Timer 2
If 1 then P4.3 is set on a match between CMPH0:CMPL0 and Timer 2
If 1 then P4.2 is set on a match between CMPH0:CMPL0 and Timer 2
If 1 then P4.1 is set on a match between CMPH0:CMPL0 and Timer 2
If 1 then P4.0 is set on a match between CMPH0:CMPL0 and Timer 2
RSTR REGISTER FUNCTIONALITY
RSTR.7
RSTR.6
RSTR.5
RSTR.4
RSTR.3
RSTR.2
RSTR.1
RSTR.0
CMTE1
CMTE0
CMR5
CMR4
CMR3
CMR2
CMR1
CMR0
If 1 then P4.7 toggles on a match between CMPH2:CMPL2 and Timer 2
If 1 then P4.6 toggles on a match between CMPH2:CMPL2 and Timer 2
If 1 then P4.5 is reset on a match between CMPH1:CMPL1 and Timer 2
If 1 then P4.4 is reset on a match between CMPH1:CMPL1 and Time r 2
If 1 then P4.3 is reset on a match between CMPH1:CMPL1 and Timer 2
If 1 then P4.2 is reset on a match between CMPH1:CMPL1 and Timer 2
If 1 then P4.1 is reset on a match between CMPH1:CMPL1 and Timer 2
If 1 then P4.0 is reset on a match between CMPH1:CMPL1 and Timer 2
WATCHDOG TIMER
The free-running watchdog timer, if enabled, will set a flag and cause a reset if not restarted by software
within the user selectable timeout period.
A typical application is to allow the flag to cause a reset. When the watchdog times out, it sets the
Watchdog Timer Reset Flag (WTRF=WDCON.2) which generates a reset if enabled by the Enable
Watchdog Timer Reset (EWT=WDCON.1) bit. In this way if the code execution goes awry and software
does not reset the watchdog as scheduled, the processor is put in a known good state: reset.
In a typical initialization, software selects the desired timeout period using the WD1:0 and the system
clock control bits. Then, it resets the timer and enables the processor reset function. After enabling the
processor reset function, software must then reset the timer before its timeout period or hardware will
reset the CPU. Both the EWT and the Watchdog Reset control (RWT = WDCON.0) bits are protected by
a Timed Access circuit. This prevents errant software from accidentally clearing the watchdog.
The watchdog timer is controlled by the Clock Control (CKCON) and the Watchdog Control (WDCON)
SFRs. CKCON.7 and CKCON.6 are WD1 and WD0 respectively, and they select the watchdog timeout
period. Of course, the 4X/ 2X (PMR.3) and CD1:0 (PMR.7:6) system clock control bits also affect the
timeout period. Selection of timeout is shown in Table 8.
28 of 50
DS87C550
WATCHDOG TIMEOUT VALUES Table 8
INTERRUPT TIMEOUT (CLOCKS)
4X/ 2X
1
0
x
x
x
CD1:0
00
00
01
10
11
RESET TIME-CLOCKS
WD1:0=00
WD1:0=01
WD1:0=10
WD1:0=11
WD1:0=00
WD1:0=01
WD1:0=10
WD1:0=11
215
216
217
217
225
218
219
220
220
228
221
222
223
223
231
224
225
226
226
234
215+512
216+512
217+512
217+512
225+512
218+512
219+512
220+512
220+512
228+512
221+512
222+512
223+512
223+512
231+512
224+512
225+512
226+512
226+512
234+512
The watchdog timer uses the internal system clock as a time base so its timeout periods are very accurate.
From the table, it can be seen that for a 33 MHz crystal frequency, the watchdog timer is capable of
producing timeout periods from 3.97 ms (2 17 * 1/33 MHz) to over two seconds (2.034 = 226 * 1/33 MHz)
with the default setting of CD1:0 (=10). This wide variation in timeout periods allows very flexible
system implementation.
In a typical initialization, the user selects one of the possible counter values to determine the timeout.
Once the counter chain has completed a full count, hardware will set the interrupt flag
(WDIF=WDCON.3). There is no hardware support for a watchdog interrupt, but this flag may be polled
to determine if the timeout period has been completed. Regardless of whether the software makes use of
this flag, there are then 512 clocks left until the reset flag (WTRF=WDCON.2) is set. Software can
enable (1) or disable (0) the reset using the Enable Watchdog Reset (EWT=WDCON.1) bit. Note that the
watchdog is a free running timer and does not require an enable.
POWER-FAIL RESET
The DS87C550 incorporates an internal precision band-gap voltage reference which, when coupled with a
comparator circuit, provides a full power-on and power-fail reset function. This circuit monitors the
processor’s incoming power supply voltage (V CC) and holds the processor in reset while V CC is out of
tolerance. Once V CC has risen above V RST, the DS87C550 will restart the oscillator for the external
crystal and count 65,536 clock cycles before program execution begins at location 0000h. This power
supply monitor will also invoke the reset state when V CC drops below the threshold condition. This reset
condition will remain while power is below the minimum voltage level. When power exceeds the reset
threshold, a full power-on reset will be performed. In this way, this internal voltage monitoring circuitry
handles both power-up and power down conditions without the need for additional external components.
The processor exits the reset condition automatically once V CC meets VRST. This helps the system
maintain reliable operation by only permitting processor operation when its supply voltage is in a known
good state. Software can determine that a Power-On Reset has occurred by checking the Power-On Reset
flag (POR=WDCON.6). Software should clear the POR bit after reading it.
The Reset pin of the DS87C550 is both an input and an output. When the processor is being held in reset
by the power-fail detection circuitry, the reset pin will be actively pulled high by the processor, and can
therefore be used as an input to other external devices.
POWER-FAIL INTERRUPT
The band-gap voltage reference that sets a precise reset threshold also generates an optional early warning
Power-fail Interrupt (PFI). When enabled by software, the processor will vector to ROM address 0033h if
VCC drops below VPFW. PFI has the highest priority. The PFI enable is in the Watchdog Control SFR
(EPFI=WDCON.5). Setting this bit to a logic 1 will enable the PFI. Application software can also read
the PFI flag at WDCON.4. A PFI condition sets this bit to a 1. The flag is independent of the interrupt
enable and software must manually clear it.
29 of 50
DS87C550
INTERRUPTS
The DS87C550 provides 16 interrupt sources with three priority levels. The Power-fail Interrupt (PFI) has
the highest priority. All interrupts, with the exception of the Power-fail Interrupt, are controlled by a
series combination of individual enable bits and a global interrupt enable EA (IE.7). Setting EA to a 1
allows individual interrupts to be enabled. Clearing EA disables all interrupts regardless of their
individual enable settings.
The three available priority levels are low, high, and highest. The highest priority level is reserved for the
Power-Fail Interrupt only. All other interrupt priority levels have individual priority bits that when set to a
1 establish the particular interrupt as high priority. In addition to the user selectable priorities, each
interrupt also has an inherent or “natural priority”. Given that all interrupt sources maintain the default
low priority, the natural priority determines the priority of simultaneously occurring interrupts. The
available interrupt sources, their flags, their enables their natural priority, and their available priority
selection bits are identified in Table 9.
INTERRUPT SOURCES AND PRIORITIES Table 9
NAME
PFI
INT0
SCON1
A/D
TF0
INT2/CF0
CM0F
INT1
INT3/CF1
CM1F
TF1
INT4/CF2
CM2F
SCON0
INT5/CF3
TF2
DESCRIPTION
VECTOR
NATURAL
PRIORITY
Power Fail Interrupt
External Interrupt 0
TI1 or RI1 from serial
port 1
A/D Converter
Interrupt
Timer 0
External Interrupt 2 or
Capture 0
Compare Match 0
External Interrupt 1
External Interrupt 3 or
Capture 1
Compare Match 1
Timer 1
External Interrupt 4 or
Capture 2
Compare Match 2
TI0 or RI0 from serial
port 0
External Interrupt 5 or
Capture 3
33h
03h
0
1
0Bh
Timer 2
FLAG
BIT
PFI(WDCON.4)
IEO(TCON.1)
ENABLE
BIT
EPFI(WDCON.5)
EX0(IE.0)
PRIORITY
CONTROL BIT
N/A
PX0(IP.0)
2
RI_1(SCON1.0)
TI_1(SCON1.1)
ES1(IE.5)
PS1(IP.5)
13h
3
EOC(ADCON1.6)
EAD(IE.6)
PAD(IP.6)
1Bh
4
TF0(TCON.5)
ET0(IE.1)
PT0(IP.1)
23h
5
IE2/CF0(T2IR.0)
EX2/EC0(EIE.0)
PX2/PC0(EIP.0)
2Bh
3Bh
6
7
CM0F(T2IR.4)
IE1(TCON.3)
ECM0(EIE.4)
EX1(IE.2)
PCM0(EIP.4)
PX1(IP.2)
43h
8
IE3/CF1(T2IR.1)
EX3/EC1(EIE.1)
PX3/PC1(EIP.1)
4Bh
53h
9
10
CM1F(T2IR.5)
TF1(TCON.7)
ECM1(EIE.5)
ET1(IE.3)
PCM1(EIP.5)
PT1(IP.3)
5Bh
11
IE4/CF2(T2IR.2)
EX4/EC2(EIE.2)
PX4/PC2(EIP.2)
63h
12
ECM2(EIE.6)
PCM2(EIP.6)
6Bh
13
CM2F(T2IR.6)
RI_0(SCON0.0)
TI_0(SCON0.1)
ES0(IE.4)
PS0(IP.4)
73h
14
IE5/CF3(T2IR.3)
EX5/EC3(EIE.3)
PX5/PC3(EIP.3)
7Bh
15
TF2(TCON.7)
TF2B(T2SEL.4)
ET2(EIE.7)
PT2(EIP.7)
EPROM PROGRAMMING
The DS87C550 follows 8-kbyte EPROM standards for the 8051 family. It is available in a UV erasable,
ceramic windowed package and in plastic packages for one-time user-programmable versions. The part
has unique signature information so programmers can support its specific EPROM options.
PROGRAMMING PROCEDURE
The DS87C550 should run from a clock speed between 4 and 6 MHz when programmed. The
programming fixture should apply address information for each byte to the address lines and the data
value to the data lines. The control signals must be manipulated as shown in Table 10. The diagram in
Figure 5 shows the expected electrical connection for programming. Note that the programmer must
30 of 50
DS87C550
apply addresses in demultiplexed fashion to Ports 1 and 2 with data on Port 0. Waveforms and timing are
provided in the Electrical Specifications.
Program the DS87C550 as follows:
1. Apply the address value.
2. Apply the data value.
3. Select the programming option from Table 10 using the control signals.
4. Increase the voltage on VPP from 5V to 12.75V if writing to the EPROM.
5. Pulse the PROG signal 5 times for EPROM array and 25 times for encryption table, lock bits, and
other EPROM bits.
6. Repeat as many times as necessary.
EPROM PROGRAMMING MODES Table 10
ALE/
MODE
Program Code Data
Verify Code Data
Program Encryption Array
Address 0-3Fh
Program Lock Bits
LB1
LB2
LB3
Program Option Register
Address FCh
Read Signature or Option
Register 30, 31, 60, FCh
RST
H
PSEN
PROG
L
PL5*
H
H
L
L
H
PL25*
H
L
PL25*
H
L
PL25*
H
L
PL25*
H
L
PL25*
H
L
H
EA/
VPP
12.75
V
H
12.75
V
12.75
V
12.75
V
12.75
V
12.75
V
H
P2.6
L
P2.7
H
P3.3
H
P3.6
H
P3.7
H
L
L
L
H
L
H
H
L
H
H
H
H
H
H
H
H
H
H
L
L
H
L
H
H
L
L
H
H
L
L
L
L
L
L
L
*PLn indicates pulse to a logic low n times
EPROM LOCK BITS Table 11
Level
1
Lock Bits
LB1 LB2 LB3
U
U
U
2
P
U
U
3
P
P
U
4
P
P
P
Protection
No program lock. Encrypted; verify if Encryption table was programmed.
Prevent MOVC instructions in external memory from reading program
bytes in internal memory. EA is sampled and latched on reset. Allow
no further programming of EPROM.
Level 2 plus no verify operation. Also, prevents MOVX instructions in external
memory from reading SRAM (MOVX) in internal memory.
Level 3 plus no external execution.
31 of 50
DS87C550
SECURITY OPTIONS
The DS87C550 employs a standard three-level lock that restricts viewing of the EPROM contents. A 64byte Encryption Array allows the authorized user to verify memory by presenting the data in encrypted
form.
Lock Bits
The security lock consists of 3 lock bits. These bits select a total of four levels. Higher levels provide
increasing security but also limit application flexibility. Table 11 shows the security settings. Note that
the programmer cannot directly read the state of the security lock.
Encryption Array
The Encryption Array allows an authorized user to verify EPROM without allowing the true memory to
be dumped. During a verify operation, each byte is Exclusive NORed (XNOR) with a byte in the
Encryption Array. This results in a true representation of the EPROM while the Encryption is
unprogrammed ( FFh). Once the Encryption Array is programmed in a non- FFh state, the verify value will
be encrypted.
For encryption to be effective, the Encryption Array must be unknown to the party that is trying to verify
memory. The entire EPROM also should be a non- FFh state or the Encryption Array can be discovered.
The Encryption Array is programmed as shown in Table 10. Note that the programmer cannot read the
array. Also note that the verify operation always uses the Encryption Array. The array has no impact
while FFh. Simply programming the array to a non- FFh state will cause the encryption to function.
EPROM ERASURE CHARACTERISTICS
Erasure of the information stored in the DS87C550’s EPROM occurs when the isolated gate structure of
the EPROM stage element is exposed to certain wavelengths of light. While the gate structure is to some
degree sensitive to a wide range of wavelengths, it is mostly wavelengths shorter than approximately
4,000 angstroms that are most effective in erasing the EPROM. Since fluorescent lighting and sunlight
have wavelengths in this range, they can cause erasure if the device is exposed to them over an extended
period of time (weeks for sunlight, years in room-level fluorescent light). For this reason (and others
mentioned previously), it is recommended that an opaque covering be placed over the window of the -K
(windowed PLCC) package type.
For complete EPROM erasure, exposure to ultraviolet light at approximately 2537 angstroms to a dose of
15W-sec/cm2 at minimum is recommended. In practice, exposing the EPROM to an ultraviolet lamp of
12,000uW/cm 2 rating for 20 to 39 minutes at a distance of approximately 1 inch will normally be
sufficient.
32 of 50
DS87C550
OTHER EPROM OPTIONS
The DS87C550 has user-selectable options that must be set before beginning software execution. These
options use EPROM bits rather than SFRs.
The EPROM selectable options may be programmed as shown in Table 10. The Option Register sets or
reads these selections. The bits in the Option Register have the following function:
Bit 7 -4
Bit 3
Bit 2-0
Reserved. Program to a 1.
Watchdog POR default. Set to 1:
Watchdog reset function is disabled on power-up. Set to 0: Watchdog reset function is
enabled automatically.
Reserved. Program to a 1.
SIGNATURE
The Signature bytes identify the product and programming revision to EPROM programmers. This
information is located at programming addresses 30h, 31h, and 60h. This information is as follows:
Address
30h
31h
60h
Value
DAh
55
00
Meaning
Manufacturer
Model
Extension
EPROM PROGRAMMING CONFIGURATION Figure 5
33 of 50
DS87C550
ELECTRICAL SPECIFICATIONS
ABSOLUTE MAXIMUM RATINGS*
Voltage on Any Pin Relative to Ground
Voltage on VCC Relative to Ground
Operating Temperature
Storage Temperature
Soldering Temperature
*
-0.3V to (VCC +0.5V)
-0.3V to 6.0V
-40°C to +85°C
-55°C to +125°C
160°C for 10 seconds
This is a stress rating only and functional operation of the device at these or any other conditions
above those indicated in the operation sections of this specification is not implied. Exposure to
absolute maximum rating conditions for extended periods of time may affect reliability.
DC ELECTRICAL CHARACTERISTICS
PARAMETER
Supply Voltage
Power-fail Warning
Minimum Operating Voltage
Supply Current Active Mode
Supply Current Idle Mode
Supply Current Stop Mode
Supply Current Stop Mode
W Band-gap
Input Low Level
Input High Level
Input High Level XTAL2 and
RST
Output Low Voltage Ports 1,
3, 4, 5, 6 @ IOL = 1.6 mA
Output Low Voltage Ports 0,
2, ALE , PSEN
@ IOL = 3.2 mA
Output High Voltage Ports 1,
2, 3, 4, 6, ALE, PSEN
@ IOH = -50 µA
Output High Voltage Ports 1,
2, 3, 4, 6, @ IOH = -1.5 mA
Output High Voltage Ports 0,
2, 5,
ALE, PSEN in Bus Mode
IOH = -8 mA
Input Low Current Ports 1, 2,
3 @ 0.45V
Transition Current from 1 to 0
Ports 1, 2, 3 @ 2V
Input Leakage Port 0, 5 and
SYMBOL
VCC
VPFW
VRST
ICC
IIDLE
ISTOP
ISPBG
MIN
4.5
4.25
4.0
VIL
VIH
VIH2
-0.3
2.0
3.5
TYP
5.0
4.38
4.13
30
15
1
100
MAX
5.5
4.5
4.25
UNITS
V
V
V
mA
mA
µA
µA
NOTES
1
1
1
2
3
4
4
+0.8
VCC+0.3
VCC+0.3
V
V
V
1
1
1
VOL1
0.15
0.45
V
1
VOL2
0.15
0.45
V
1
VOH1
2.4
V
1, 6
VOH2
2.4
V
1, 7
VOH3
2.4
V
1, 5
IIL
-55
µA
11
ITL
-650
µA
8
+10
µA
10
IL
-10
EA
34 of 50
DS87C550
DC ELECTRICAL CHARACTERISTICS
Input Leakage Port 0, Bus
Mode
RST Pulldown Resistance
IL
-300
+300
µA
RRST
50
170
kO
9
NOTES FOR DS87C550 DC ELECTRICAL CHARACTERISTICS:
All parameters apply to both commercial and industrial temperature operation unless otherwise noted.
1. Voltage referenced to digital ground (GND).
2. Active current measured with 33 MHz clock source on XTAL1, V CC=RST=5.5V, other pins
disconnected.
3. Idle mode current measured with 33 MHz clock source on XTAL1, V CC=5.5V, RST at ground, other
pins disconnected.
4. Stop mode current measured with XTAL1 and RST grounded, V CC=5.5V, all other pins disconnected.
This value is not guaranteed. Users that are sensitive to this specification should contact Dallas
Semiconductor for more information.
5. When addressing external memory.
6. RST=VCC. This condition mimics operation of pins in I/O mode. Port 0 is tristated in reset and when
at a logic high state during I/O mode.
7. During a 0 to 1 transition, a one-shot drives the ports hard for two oscillator clock cycles. This
measurement reflects port in transition mode.
8. Ports 1, 2, 3, 5, 6 source transition current when being pulled down externally. Current reaches its
maximum at approximately 2V.
9. 0.45<Vin<V CC. Not a high impedance input. This port is a weak address holding latch in Bus Mode.
Peak current occurs near the input transition point of the latch, approximately 2V.
10. 0.45<Vin<V CC. RST=VCC. This condition mimics operation of pins in I/O mode.
11. This is the current required from an external circuit to hold a logic low level on an I/O pin while the
corresponding port latch is set to 1. This is only the current required to hold the low level; transitions
from 1 to 0 on an I/O pin will also have to overcome the transition current.
35 of 50
DS87C550
A/D CONVERTER ELECTRICAL
CHARACTERISTICS
PARAMETER
Analog Supply Voltage
Analog Supply Current
Analog Idle Mode Current
Analog Power-Down Mode
Current
Analog Input Voltage
Internal Reference
External Reference
External Analog Reference
Voltage
Internal Reference Voltage
Change with Temperature
Change with AV CC
Analog Input Capacitance
A/D Clock
Sampling Time
Conversion Time
Resolution
Differential non-linearity
Integral non-linearity
Offset Error
Gain Error
Cross-talk between A/D
inputs
(@ VCC = 5V ± 10%, 0-70ºC)
SYMBOL
AVCC
AVSS
AIDD
AIDDI
AIDDPD
MIN
VCC
GND
ADC7ADC0
AVREFAVREF+
VBG
VBGT
VBGV
MAX
VCC
GND
1
TBD
300
UNITS
V
0
AVREF-
2.50
AVREF+
V
AVSS-0.2
AVCC +
0.2
2.5625
V
2.4375
CIN
tACLK
tADS
tADC
TYP
2.50
± 50
± 10
10
1
5 tACLK
16 tACLK
15
6.25
10
±1.0
±2.0
±2.0
±1.0
-60
EDL
EIL
EOS
EG
ECT
NOTES
mA
µA
nA
V
PPM/°
C
mV/V
PF
us
tACLK
tACLK
Bits
LSB
LSB
LSB
%
dB
NOTES FOR A/D CONVERTER ELECTRICAL CHARACTERISTICS
1. The following condition must not be exceeded: GND-0.2V < AV SS < VCC + 0.2V.
2. Due to the dynamic nature of the A/D converter, tACLK has both min and max specifications.
3. A complete conversion cycle requires 16 ACLK periods, including five input sampling periods.
36 of 50
2
3
DS87C550
TYPICAL ICC VERSUS FREQUENCY
AC ELECTRICAL CHARACTERISTICS
PARAMETER
Oscillator Freq. (Ext. Osc.)
(Ext. Crystal)
ALE Pulse Width
Port 0 Address Valid to ALE
Low
Address Hold after ALE Low
ALE Low to Valid
Instruction In
ALE Low to PSEN Low
PSEN Pulse Width
PSEN Low to Valid
Instruction In
Input Instruction Hold after
PSEN
Input Instruction Float after
PSEN
Port 0 Address to Valid
Instruction In
Port 2 Address to Valid
Instruction In
PSEN Low to Address Float
SYMBOL
1/tCLCL
tLHLL
33 MHz
MIN
MAX
0
33
1
33
40
VARIABLE CLOCK
MIN
MAX
0
33
1
33
0.375 tMCS -5
UNITS
MHz
ns
tAVLL
10
0.125 tMCS -5
ns
tLLAX1
10
0.125 tMCS -5
ns
tLLIV
tLLPL
tPLPH
0.625
tMCS -20
56
10
55
tPLIV
0.125 tMCS -5
0.5 tMCS -5
tPXIX
0
tPXIZ
26
tAVIV
71
tAVIV2
81
tPLAZ
0
ns
ns
0.5tMCS
-20
41
ns
0
ns
ns
0.25tMCS
-5
0.75tMCS
-20
0.875
tMCS -25
0
ns
ns
ns
ns
NOTES FOR AC ELECTRICAL CHARACTERISTICS
§ The value for tMCS is a time period related to the machine cycle clock in terms of the processor’s input
clock frequency. Its value is highlighted in the table “STRETCH VALUE TIMING” for all possible
settings of the 4X/ 2X and CD1:0 bits. The default condition is CD1 = 1 and CD0 = 0, where 4X/ 2X
is disregarded.
§ All parameters apply to both commercial and industrial temperature operation unless otherwise noted.
37 of 50
DS87C550
§ All signals characterized with load capacitance of 80 pF except Port 0, ALE, PSEN , RD and WR
with 100 pF.
§ Interfacing to memory devices with float times (turn off times) over 25 ns may cause contention. This
will not damage the parts, but will cause an increase in operating current.
§ Specifications assume a 50% duty cycle for the oscillator. Port 2 and ALE timing will change in
relation to duty cycle variation.
MOVX CHARACTERISTICS
VARIABLE CLOCK
PARAMETER
SYMBOL
Data Access ALE
Pulse Width
tLHLL2
Address Hold after
ALE Low for
MOVX Write
tLLAX2
RD
Pulse Width
WR Pulse Width
RD Low to
Valid Data In
Data Hold
after Read
tRLRH
tWLWH
MIN
MAX
0.375tMCS-5
0.5tMCS-5
1.5tMCS-10
0.125tMCS-5
0.25tMCS-5
1.25tMCS-10
0.5tMCS-5
CST*tMCS-10
0.5tMCS-5
CST*tMCS-10
ns
ns
ns
ns
0.5tMCS-20
CST*tMCS-20
tRLDV
tRHDX
Data Float
after Read
tRHDZ
ALE Low to
Valid Data In
tLLDV
Port 0 Address
to Valid Data In
tAVDV1
Port 2 Address
to Valid Data In
tAVDV2
ALE Low to RD
or WR Low
tLLWL
Port 0 Address to
RD or WR Low
tAVWL1
Port 2 Address to
RD or WR Low
tAVWL2
Data Valid to
WR Transition
tQVWX
UNITS
0
ns
STRETCH
VALUES CST
(MD2:0)
CST=0
1<=CST<=3
4<=CST<=7
CST=0
1<=CST<=3
4<=CST<=7
CST=0
1<=CST<=7
CST=0
1<=CST<=7
CST=0
1<=CST<=7
ns
0.125tMCS-5
0.25tMCS-5
1.25tMCS-10
0.25tMCS-5
0.5tMCS-5
2.5tMCS-10
0.375tMCS-5
0.625tMCS-5
2.625tMCS-10
0.25tMCS-5
0.5tMCS-5
1.5tMCS-15
0.625tMCS-20
(CST+0.25)*tMCS-40
(CST+1.25)*tMCS-40
0.75tMCS-20
(CST+0.375)*tMCS-20
(CST+1.375)*tMCS-20
0.875tMCS-20
(CST+0.5)*tMCS-20
(CST+1.5)*tMCS-20
1.25tMCS+5
0.25tMCS+5
1.25tMCS+10
-5
ns
ns
ns
ns
ns
ns
ns
ns
38 of 50
CST=0
1<=CST<=3
4<=CST<=7
CST=0
1<=CST<=3
4<=CST<=7
CST=0
1<=CST<=3
4<=CST<=7
CST=0
1<=CST<=3
4<=CST<=7
CST=0
1<=CST<=3
4<=CST<=7
CST=0
1<=CST<=3
4<=CST<=7
CST=0
1<=CST<=3
4<=CST<=7
DS87C550
MOVX CHARACTERISTICS CONT’D
Data Hold
after Write
tWHQX
Low to
Address Float
tRLAZ
RD
or WR High
to ALE High
RD
tWHLH
0.25tMCS-5
0.5tMCS-5
1.5tMCS-10
ns
-((0.125 tMCS)-5)
ns
10
0.25tMCS+5
1.25tMCS+10
0
0.25tMCS-5
1.25tMCS-10
CST=0
1<=CST<=3
4<=CST<=7
ns
CST=0
1<=CST<=3
4<=CST<=7
NOTES FOR MOVX CHARACTERISTICS USING STRETCH MEMORY
CYCLES
§ tMCS is a time period related to the Stretch memory cycle selection. The following table shows the
value of tMCS for each Stretch selection.
§ CST is the stretch cycle value as determined by the MD2, MD1, & MD0 bits of the CKCON register.
tMCS TIME PERIODS
System Clock Selection
4X/ 2X , CD1, CD0 = 100
4X/ 2X , CD1, CD0 = 000
4X/ 2X , CD1, CD0 = x10
4X/ 2X , CD1, CD0 = x11
tMCS
0.5 tCLCL
1 tCLCL
4 tCLCL
1024 tCLCL
RD , WR PULSE WIDTH WITH STRETCH CYCLES
RD , WR Pulse Width (in oscillator clocks)
MOVX
MD2
MD1
MD0
Machine
4X/ 2X =1 4X/ 2X =0 4X/ 2X =x
4X/ 2X =x
Cycles
CD1:0=00 CD1:0=00 CD1:0=10 CD1:0=11
2048 tCLCL
0
0
0
2
0.5 tCLCL
1 tCLCL
2 tCLCL
4096 tCLCL
0
0
1
3
tCLCL
2 tCLCL
4 tCLCL
8192 tCLCL
0
1
0
4
2 tCLCL
4 tCLCL
8 tCLCL
12288 tCLCL
0
1
1
5
3 tCLCL
6 tCLCL
12 tCLCL
16384 tCLCL
1
0
0
9
4 tCLCL
8 tCLCL
16 tCLCL
20480 tCLCL
1
0
1
10
5 tCLCL
10 tCLCL
20 tCLCL
24576 tCLCL
1
1
0
11
6 tCLCL
12 tCLCL
24 tCLCL
28672 tCLCL
1
1
1
12
7 tCLCL
14 tCLCL
28 tCLCL
EXTERNAL CLOCK CHARACTERISTICS
PARAMETER
Clock High Time
Clock Low Time
Clock Rise Time
Clock Fall Time
SYMBOL
tCHCX
tCLCX
tCLCL
tCHCL
MIN
10
10
39 of 50
TYP
MAX
5
5
UNITS
ns
ns
ns
ns
NOTES
DS87C550
SERIAL PORT MODE 0 TIMING CHARACTERISTICS
PARAMETER
Serial Port Clock Cycle Time
SM2=0, 12 clocks per cycle
SM2=1, 4 clocks per cycle
Output Data Setup to Clock Rising
SM2=0, 12 clocks per cycle
SM2=1, 4 clocks per cycle
Output Data Hold from Clock Rising
SM2=0, 12 clocks per cycle
SM2=1, 4 clocks per cycle
Input Data Hold from Clock Rising
SM2=0, 12 clocks per cycle
SM2=1, 4 clocks per cycle
Clock Rising Edge to Input Data Valid
SM2=0, 12 clocks per cycle
SM2=1, 4 clocks per cycle
SYMBOL
MIN
TYP
MAX
UNITS
tXLXL
12tCLCL
4tCLCL
ns
ns
tQVXH
12tCLCL
4tCLCL
ns
ns
tXHQX
12tCLCL
4tCLCL
ns
ns
tXHDX
12tCLCL
4tCLCL
ns
ns
tXHDV
12tCLCL
4tCLCL
ns
ns
NOTES
EXPLANATION OF AC SYMBOLS
In an effort to remain compatible with the original 8051 family, this device specifies the same parameters
as such devices, using the same symbols. For completeness, the following is an explanation of the
symbols.
t
A
C
D
H
L
I
P
Q
R
V
W
X
Z
Time
Address
Clock
Input data
Logic level high
Logic level low
Instruction
PSEN
Output data
RD signal
Valid
WR signal
No longer a valid logic level
Tristate
40 of 50
DS87C550
POWER CYCLE TIMING CHARACTERISTICS
PARAMETER
Cycle Start-up Time
Power-on Reset Delay
SYMBOL
tCSU
tPOR
MIN
TYP
1.8
MAX
65536
UNITS NOTES
ms
1
tCLCL
2
NOTES FOR POWER CYCLE TIMING CHARACTERISTICS
1. Start-up time for crystals varies with load capacitance and manufacturer. Time shown is for an
11.0592 MHz crystal manufactured by Fox.
2. Reset delay is a synchronous counter of crystal oscillations after crystal start-up. At 33 MHz, this time
is 1.99 ms.
EPROM PROGRAMMING AND VERIFICATION
PARAMETER
Programming Voltage
Programming Supply Current
Oscillator Frequency
POR Delay
Address Setup to PROG Low
Address Hold after PROG
Data Setup to PROG Low
Data Hold after PROG
Enable High to V PP
VPP Setup to PROG Low
VPP Hold after PROG
PROG Width
Address to Data Valid
Enable Low to Data Valid
Data Float after Enable
PROG High to PROG Low
SYMBOL
VPP
IPP
1/tCLCL
tDELAY
tAVGL
tGHAX
tDVGL
tGHDX
tEHSH
tSHGL
tGHSL
tGLGH
tAVQV
tELQV
tEHQZ
tGHGL
MIN
12.5
4
65536
TYP
MAX
13.0
75
6
UNITS
V
mA
MHz
tCLCL
NOTES
1
2
48tCLCL
48tCLCL
48tCLCL
48tCLCL
48tCLCL
10
10
90
0
10
110
µs
µs
µs
48tCLCL
48tCLCL
48tCLCL
µs
NOTES FOR EPROM PROGRAMMING AND VERIFICATION
1. All voltage referenced to ground.
2. The microcontroller holds itself in reset for this duration when power is applied. No signals should be
manipulated during this interval since the microcontroller is ignoring inputs. At a 4 MHz oscillator
frequency, this period is 16.4 ms.
41 of 50
DS87C550
EXTERNAL PROGRAM MEMORY READ CYCLE
EXTERNAL DATA MEMORY READ CYCLE
42 of 50
DS87C550
EXTERNAL DATA MEMORY WRITE CYCLE
DATA MEMORY WRITE WITH STRETCH=1
43 of 50
DS87C550
DATA MEMORY WRITE WITH STRETCH=2
DATA MEMORY WRITE WITH STRETCH=4
EXTERNAL CLOCK DRIVE
44 of 50
DS87C550
SERIAL PORT MODE 0 TIMING
SERIAL PORT 0 (SYNCHRONOUS MODE)
HIGH SPEED OPERATION SM2=1=>TXD CLOCK=XTAL/4
SERIAL PORT 0 (SYNCHRONOUS MODE)
SM2=0=>TXD CLOCK=XTAL/12
45 of 50
DS87C550
POWER CYCLE TIMING
EPROM PROGRAMMING AND VERIFICATION WAVEFORMS
46 of 50
DS87C550
80-PIN QUAD FLAT PACK (14.0 MM X 20.0 MM)
NOTES:
1.
2.
3.
DIM
A
A1
A2
B
C
D
D1
E
E1
e
L
MIN
MAX
3.40
0.25
2.55
2.87
0.30
0.45
0.13
0.23
23.70
24.10
19.90
20.10
17.70
18.10
13.90
14.10
0.80 BSC
0.65
0.95
47 of 50
DIMENSIONS D1 AND E1 INCLUDE MOLD MISMATCH,
BUT DO NOT INCLUDE MOLD PROTRUSION;
ALLOWABLE PROTRUSION IS 0.25 MM PER SIDE.
DETAILS OF PIN 1 IDENTIFIER ARE OPTIONAL BUT
MUST BE LOCATED WITHIN THE ZONE INDICATED.
ALLOWABLE DAMBAR PROTRUSION IS 0.08 MM
TOTAL IN EXCESS OF THE B DIMENSIONS;
PROTRUSION NOT TO BE LOCATED ON LOWER
RADIUS OR FOOT OF LEAD.
DS87C550
68-PIN WINDOWED CLCC
LTR
A
A1
B
B1
c
CH1
D
D1
D2
E
E1
E2
e1
MIN
MAX
.160
.190
.034
.026
.032
.017
.023
.005
.010
.035
.045
.978
.998
.940
.960
.910
.930
.978
.998
.940
.960
.910
.930
.050 BSC
NOTES:
1.
2.
PIN-1 IDENTIFIER TO BE LOCATED IN ZONE INDICATED.
ALL DIMENSIONS SHOWN ARE IN INCHES.
48 of 50
DS87C550
68-PIN PLCC
NOTE:
1.
2.
PIN-1 IDENTIFIER TO BE LOCATED IN ZONE INDICATED.
DIMENSIONS ARE IN INCH UNITS.
DIM
A
A1
A2
b
b1
c
CH1
D
D1
D2
E
E1
E2
e1
MIN
MAX
.165
.180
.090
.120
.020
.026
.032
.013
.021
.0075
.0125
.042
.048
.985
.995
.950
.958
.882
.938
.985
.995
.950
.958
.882
.938
.050 BSC
49 of 50
DS87C550
HISTORY REVISION
The following represent the key differences between the 06/16/98 and the 06/14/99 version of the
DS87C550 data sheet. Please review this summary carefully.
1. Corrected minor typographical errors on pages 3, 5, 7, 9, 17, 20, 21, 22, 26 and 33.
2. Added standard “Absolute Maximum Ratings” notation.
3. Added AID Specification conditions (V CC = 5V ± 10%, 0-70°C).
4. Updated port 5’s VOL and I L specification.
50 of 50