Automotive PMIC for Safety Systems
Automotive PMIC for Safety Systems
TPS6593-Q1 Power Management IC (PMIC) with 5 BUCKs and 4 LDOs for Safety-
Relevant Automotive Applications
• One low-dropout (LDO) linear regulator with low-
1 Features noise performance
• Qualified for automotive applications – 1.2 V to 3.3 V output voltage range in 25-mV
• AEC-Q100 qualified with the following results: steps
– Device operates from 3 V to 5.5 V input supply – 300 mA output current capability with short-
– Device temperature grade 1: –40°C to +125°C circuit and over-current protection
ambient operating temperature range • Configurable power sequence control in non-
– Device HBM classification level 2 volatile memory (NVM):
– Device CDM classification level C4A – Configurable power-up and power-down
• Functional Safety-Compliant sequences between power states
– Developed for functional safety applications – Digital output signals can be included in the
– Documentation to aid ISO26262 and IEC61508 power sequences
system design available upon product release – Digital input signals can be used to trigger
– Systematic capability up to ASIL-D and SIL-3 power sequence transitions
– Hardware integrity up to ASIL-B and SIL-2 – Configurable handling of safety-relevant errors
– Input supply voltage monitor • 32-kHz crystal oscillator with option to output a
– Under/overvoltage monitors and over-current buffered 32-kHz clock output
monitors on all output supply rails • Real-time clock (RTC) with alarm and periodic
– Watchdog with selectable trigger / Q&A mode wake-up mechanism
– Two error signal monitors (ESMs) with • One SPI or two I2C control interfaces, with
selectable level / PWM mode second I2C interface dedicated for Q&A watchdog
– Thermal monitoring with high temperature communication
warning and thermal shutdown • Package option:
– Bit-integrity (CRC) error detection on internal – 8-mm × 8-mm 56-pin VQFNP with 0.5-mm pitch
configuration registers and non-volatile memory
(NVM) 2 Applications
• Low-power consumption • Automotive infotainment and digital cluster,
– 2 μA typical shutdown current navigation systems, telematics, body electronics
– 7 μA typical in back up supply only mode and lighting
– 20 μA typical in low power standby mode • Advanced driver assistance system (ADAS)
• Five step-down switched-mode power supply • Industrial control and automation
(BUCK) regulators:
3 Description
– 0.3 V to 3.34 V output voltage range in 5, 10, or
20-mV steps The TPS6593-Q1 device provides four flexible multi-
– One with 4 A, three with 3.5 A, and one with 2 phase configurable BUCK regulators with 3.5 A output
A output current capability current per phase, and one additional BUCK regulator
– Flexible multi-phase capability for four BUCKs: with 2 A output current.
up to 14 A output current from a single rail Table 3-1. Device Information Table
– Short-circuit and over-current protection
PART NUMBER(1) PACKAGE BODY SIZE (NOM)
– Internal soft-start for in-rush current limitation
TPS6593-Q1 VQFNP (56) 8.00 mm × 8.00 mm
– 2.2 MHz / 4.4 MHz switching frequency
– Ability to synchronize to external clock input (1) See the orderable addendum at the end of the data sheet for
• Three low-dropout (LDO) linear regulators with all available packages.
configurable bypass mode
– 0.6 V to 3.3 V output voltage range with 50-mV
steps in linear regulation mode
– 1.7 V to 3.3 V output voltage range in bypass
mode
– 500 mA output current capability with short-
circuit and over-current protection
An IMPORTANT NOTICE at the end of this data sheet addresses availability, warranty, changes, use in safety-critical applications,
intellectual property matters and other important disclaimers. PRODUCTION DATA.
TPS6593-Q1
SLVSE83B – DECEMBER 2020 – REVISED SEPTEMBER 2023 www.ti.com
nINT
I2C and SPI
128-kHz RC Oscillator
OSC32KCAP
nPWRON/ENABLE
Interrupt Handler OSC32KIN
VBACKUP 32-kHz Crystal Oscillator OSC32KOUT
Backup Supply
VCCA Management WAKEn Real-Time Clock SYNCCLKOUT/CLK32KOUT
nSLEEPn (RTC) With Calendar
VOUT_LDOVRTC LDO Bandgap 20-MHz Monitor Oscillator
Clock Controller
RTC VRTC
Fixed & Monitor
20-MHz RC Oscillator
State Machine
VOUT_LDOVINT LDO Bandgap (FFSM) Trigger Mode or
FSD
INT VINT Question and Answer Clock
Pre- (Q&A) Watchdog DPLL with SYNCCLKIN
Dividers and
Configurable nERR_MCU, SSM (GPIO10)
VIN Monitor Mux
State nERR_SoC Level or PWM Mode
OVP Machine
Bandgap Error Signal Monitors
UVLO Single or Multiphase
(PFSM) (MCU, SoC)
BUCK1 PVIN_B1
Power-Good Monitor
Resource Controller
for Buck and LDO 3.5 A SW_B1
Regulators SRAM Over-Current Monitor,
PVIN_B5
Power Good Short Circuit Monitor, FB_B1
Controller & Monitor SW Short Monitor
VIN Monitor
OVP
Bandgap
UVLO
Registers BUCK2 PVIN_B2
CRC
Thermal Monitor Thermal I2C/SPI/ 3.5 A SW_B2
Controller GPIO/ Over-Current Monitor,
Register Map Short Circuit Monitor,
SPMI FB_B2
VOUT_LDO1 LDO1, Bypass SW Short Monitor
Non-Volatile Memory
LBIST (NVM)
PVIN_LDO12 Over-Current Monitor,
Short Circuit Monitor BUCK3 PVIN_B3
2
LDO2, Bypass SPMI I C and SPI 3.5 A SW_B3
Over-Current Monitor,
Over-Current Monitor, BIST and CRC CRC Short Circuit Monitor,
VOUT_LDO2 FB_B3
Short Circuit Monitor SW Short Monitor
Target Control I2C1 I2C2 SPI
PVIN_LDO3 LDO3, Bypass
BUCK4 PVIN_B4
VOUT_LDO3 Over-Current Monitor, 4 A (Single-Phase)
Short Circuit Monitor or 3.5 A SW_B4
nERR_MCU (GPIO7)
SDA_I2C2 (GPIO2)
SCL_I2C2 (GPIO1)
Over-Current Monitor,
nERR_SoC (GPIO3)
SDO_SPI (GPIO2)
(GPIO2 or GPIO11)
PGOOD (GPIO9)
CS_SPI (GPIO1)
SDATA (GPIO6)
FB_B4
(Low Noise) SW Short Monitor
TRIG_WDOG
SDA_I2C1
SCL_I2C1
Over-Current Monitor,
CLK_SPI
SDI_SPI
VOUT_LDO4
Short Circuit Monitor
BUCK5 PVIN_B5
VIO_IN Safety
SDA_I2C1/SDI_SPI
AMUXOUT
Bandgap
nRST_OUT
GPIO10
GPIO11
GPIO1
GPIO2
GPIO3
GPIO4
GPIO5
GPIO6
GPIO7
GPIO8
GPIO9
EN_DRV
VCCA
Functional Diagram
Table of Contents
1 Features............................................................................1 7.18 I/O Pullup and Pulldown Resistance.......................37
2 Applications..................................................................... 1 7.19 I2C Interface............................................................37
3 Description.......................................................................1 7.20 Serial Peripheral Interface (SPI)............................. 39
4 Revision History.............................................................. 3 7.21 Typical Characteristics............................................ 40
5 Description (continued).................................................. 4 8 Detailed Description......................................................43
6 Pin Configuration and Functions...................................5 8.1 Overview................................................................... 43
6.1 Digital Signal Descriptions........................................ 10 8.2 Functional Block Diagram......................................... 44
7 Specifications................................................................ 17 8.3 Feature Description...................................................45
7.1 Absolute Maximum Ratings...................................... 17 8.4 Device Functional Modes........................................118
7.2 ESD Ratings............................................................. 18 8.5 Control Interfaces....................................................151
7.3 Recommended Operating Conditions.......................18 8.6 Configurable Registers........................................... 158
7.4 Thermal Information..................................................18 8.7 Register Maps.........................................................160
7.5 General Purpose Low Drop-Out Regulators 9 Application and Implementation................................ 358
(LDO1, LDO2, LDO3)..................................................20 9.1 Application Information........................................... 358
7.6 Low Noise Low Drop-Out Regulator (LDO4)............ 21 9.2 Typical Application.................................................. 358
7.7 Internal Low Drop-Out Regulators (LDOVRTC, 9.3 Power Supply Recommendations...........................378
LDOVINT)....................................................................22 9.4 Layout..................................................................... 378
7.8 BUCK1, BUCK2, BUCK3, BUCK4 and BUCK5 10 Device and Documentation Support........................381
Regulators................................................................... 23 10.1 Device Support..................................................... 381
7.9 Reference Generator (BandGap)..............................29 10.2 Device Nomenclature............................................381
7.10 Monitoring Functions ..............................................30 10.3 Documentation Support........................................ 382
7.11 Clocks, Oscillators, and PLL................................... 32 10.4 Receiving Notification of Documentation Updates382
7.12 Thermal Monitoring and Shutdown......................... 33 10.5 Support Resources............................................... 382
7.13 System Control Thresholds.....................................34 10.6 Trademarks........................................................... 382
7.14 Current Consumption..............................................34 10.7 Electrostatic Discharge Caution............................382
7.15 Backup Battery Charger..........................................35 10.8 Glossary................................................................382
7.16 Digital Input Signal Parameters.............................. 35 11 Mechanical, Packaging, and Orderable
7.17 Digital Output Signal Parameters ...........................36 Information.................................................................. 382
4 Revision History
NOTE: Page numbers for previous revisions may differ from page numbers in the current version.
Changes from Revision A (July 2022) to Revision B (September 2023) Page
• Changed the device status from Advance Information to Production Data ....................................................... 1
5 Description (continued)
All of the BUCK regulators can be synchronized to an internal 2.2-MHz or 4.4-MHz or an external 1-MHz, 2-MHz,
or 4-MHz clock signal. To improve the EMC performance, an integrated spread-spectrum modulation can be
added to the synchronized BUCK switching clock signal. This clock signal can also be made available to external
devices through a GPIO output pin. The device provides four LDOs: three with 500-mA capability, which can be
configured as load switches; one with 300-mA capability and low-noise performance.
Non-volatile memory (NVM) is used to control the default power sequences and default configurations, such
as output voltage and GPIO configurations. The NVM is pre-programmed to allow start-up without external
programming. Most static configurations, stored in the register map of the device, can be changed from the
default through SPI or I2C interfaces to configure the device to meet many different system needs. The NVM
contains a bit-integrity-error detection feature (CRC) to stop the power-up sequence if an error is detected,
preventing the system from starting in an unknown state.
The TPS6593-Q1 includes a 32-kHz crystal oscillator, which generates an accurate 32-kHz clock for the
integrated RTC module. A backup-battery management provides power to the crystal oscillator and the real-time
clock (RTC) module from a coin cell battery or a super-cap in the event of power loss from the main supply.
The TPS6593-Q1 device includes protection and diagnostic mechanisms such as voltage monitoring on the
input supply, voltage monitoring on all BUCK and LDO regulator outputs, register and interface CRC, current-
limit, short-circuit protection, thermal pre-warning, and over-temperature shutdown. The device also includes a
Q&A or trigger mode watchdog to monitor for MCU software lockup, and two error signal monitor (ESM) inputs
with fault injection options to monitor the error signals from the attached SoC or MCU. The TPS6593-Q1 can
notify the processor of these events through the interrupt handler, allowing the MCU to take action in response.
PVIN_B4
PVIN_B3
GPIO11
SW_B4
SW_B4
SW_B3
SW_B3
VIO_IN
FB_B4
FB_B3
GPIO4
GPIO3
GND
GND
56
55
54
53
52
51
50
49
48
47
46
45
44
43
AMUXOUT 1 42 GPIO10
VOUT_LDOVINT 2 41 GPIO8
VOUT_LDOVRTC 3 40 OSC32KCAP
VCCA 4 39 OSC32KOUT
REFGND1 5 38 OSC32KIN
REFGND2 6 37 FB_B5
VOUT_LDO3 9 34 SW_B5
PVIN_LDO3 10 33 GPIO2
VOUT_LDO2 11 32 GPIO1
PVIN_LDO12 12 31 SCL_I2C1/SCK_SPI
VOUT_LDO1 13 30 SDA_I2C1/SDI_SPI
nINT 14 29 EN_DRV
15
16
17
18
19
20
21
22
23
24
25
26
27
28
GPIO7
nRSTOUT
PVIN_B2
nPWRON/ENABLE
PVIN_B1
SW_B2
SW_B2
GPIO9
FB_B2
FB_B1
GPIO5
GPIO6
SW_B1
SW_B1
Not to scale
Figure 6-1. 56-Pin RWE (VQFNP) Package, 0.5-mm Pitch, With Thermal Pad (Top View)
(1) Default option before NVM settings are loaded into the device.
7 Specifications
7.1 Absolute Maximum Ratings
Over operating free-air temperature range (unless otherwise noted). Voltage level is with reference to the thermal/ground
pad of the device.(1)
POS MIN MAX UNIT
M1.3 Voltage on OV protected supply input pin VCCA(2) –0.3 6 V
Voltage on all buck supply voltage input
M1.4 PVIN_Bx(2) –0.3 6 V
pins
Voltage difference between supply input
M1.4a Between VCCA and each PVIN_Bx –0.5 0.5 V
pins
PVIN_Bx + 0.3
M1.5a SW_Bx pins –0.3 V
Voltage on all buck switch nodes V, up to 6 V
M1.5b SW_Bx pins, 10-ns transient –2 10 V
M1.6 Voltage on all buck voltage sense nodes FB_Bx –0.3 4 V
M1.7 Voltage on all LDO supply voltage input pins PVIN_LDOx(2) –0.3 6 V
PVIN_LDOx +
M1.8 Voltage on all LDO output pins VOUT_LDOx –0.3 V
0.3 V, up to 6 V
M1.9 Voltage on internal LDO output pins VOUT_LDOVINT, VOUT_LDOVRTC –0.3 2 V
VCCA + 0.3 V,
M1.10 Voltage on I/O supply pin VIO_IN with respect to ground pad –0.3 V
up to 6 V
Voltage on logic pins (input or output) in VIO I2C and SPI pins, nRSTOUT, and nINT pins, and all
M1.11 –0.3 6 V
domain GPIO output buffers except GPIO5 & GPIO6
Voltage on logic pins (input or output) in GPIO5 & GPIO6, and all GPIO input buffers except
M1.12 –0.3 6 V
LDOVINT domain GPIO3 & GPIO4
Voltage on logic pins (input) in LDOVRTC
M1.13 GPIO3 & GPIO4 –0.3 6 V
domain
Voltage on logic pins (input or output) in
M1.14 nPWRON/ENABLE & EN_DRV –0.3 6 V
VCCA domain
VCCA + 0.3 V,
M1.15 Voltage on analog mux output pin AMUXOUT –0.3 V
up to 6 V
M1.16 Voltage on back-up power supply input VBACKUP –0.3 6 V
M1.17 Voltage on crystal oscillator pins OSC32KIN, OSC32KOUT, & OSC32KCAP –0.3 2 V
M1.18 Voltage on REFGND pins REFGND1 & REFGND2 –0.3 0.3 V
M2.1a VCCA, PVIN_Bx (voltage below 2.7 V) 60 mV/µs
Voltage rise slew-rate on input supply pins
M2.1b VIO (only when VCCA < 2 V) 60 mV/µs
M2.3a All pins other than power resources 20 mA
Buck1/2/3/4 regulators: PVIN_Bx and SW_Bx per
M2.3b Peak output current 5 A
phase
M2.3c Buck5 regulator: PVIN_B5 and SW_B5 3 A
M2.4a GPIOx pins, source current 3 mA
GPIO1/2/5/6, SDA_I2C1/SDI_SPI, EN_DRV, nINT,
M2.4b 8 mA
and nRSTOUT pins, sink current
Average output current, 100 k hour, TJ =
M2.4c 125℃ GPIO3/4/7/8/9/10/11 pins, sink current 3 mA
M2.4d LDO1/2/3 regulators 350 mA
M2.4e LDO4 regulators 210 mA
M3 Junction temperature, TJ –45 160 °C
M4 Storage temperature, Tstg –65 150 °C
(1) Operation outside the Absolute Maximum Ratings may cause permanent device damage. Absolute Maximum Ratings do not imply
functional operation of the device at these or any other conditions beyond those listed under Recommended Operating Conditions.
If used outside the Recommended Operating Conditions but within the Absolute Maximum Ratings, the device may not be fully
functional, and this may affect device reliability, functionality, performance, and shorten the device lifetime.
(2) The voltage at VCCA and PVIN pins can exceed the 6 V absolute max condition for a short period of time, but must remain less than 8
V. VCCA at 8 V for a 100 ms duration is equivalent to approximately 8 hours of aging for the device at room temperature.
(1) AEC Q100-002 indicates that HBM stressing must be in accordance with the ANSI/ESDA/JEDEC JS-001 specification.
(1) The maximum output voltage of BUCK1 to BUCK5 and LDO1 to LDO4 can be reduced by an NVM setting to adopt the maximum
voltage to the requirements (or maximum ratings) of the load. This reduction of the maximum output voltage protects the processor
from exceeding the maximum ratings of the core voltage. The default value is defined in the nonvolatile memory (NVM) and can be
updated by software through I2C/SPI interface after device start-up.
(2) Additional cooling strategies may be necessary to keep junction temperature at recommended limits.
(3) The input buffer of a fail-safe GPIO pin is isolated from its input signal. Therefore, the input voltage to a fail-safe pin can be as high as
5.5 V.
(1) For more information about traditional and new thermal metrics, see the Semiconductor and ICPackage Thermal Metrics application
report.
7.5 General Purpose Low Drop-Out Regulators (LDO1, LDO2, LDO3) (continued)
Over operating free-air temperature range (unless otherwise noted). Voltage level is in reference to the thermal/ground pad
of the device.
POS PARAMETER TEST CONDITIONS MIN TYP MAX UNIT
VTH_RV_SC(LDO Threshold voltage for Short
1.20 LDOn_EN = 0 140 150 160 mV
n) Circuit
Timing Requirements
Time between enable of the LDOn to within OV/UV
19.1 ton(LDOn) Turn-on time 500 µs
monitor level
VOUT from 0.3 V to 90% of LDOn_VSET.
19.2a 25 mV/µs
LDOn_SLOW_RAMP = 0
tramp(LDOn) Ramp-up slew rate
VOUT from 0.3 V to 90% of LDOn_VSET.
19.2b 3 mV/µs
LDOn_SLOW_RAMP = 1
19.3a tdelay_OC(LDOn) Over-current detection delay Detection signal delay when IOUT > ILIM 35 µs
tdeglitch_OC(LDOn Over-current detection signal
19.3b Digital deglitch time for the over-current detection signal 38 44 µs
) deglitch time
tlatency_OC(LDOn Over-current signal total
19.4 Total delay from Iout > ILIM to interrupt or PFSM trigger 79 µs
) latency time
(1) Input capacitors must be placed as close as possible to the device pins.
(2) When DC voltage is applied to a ceramic capacitor, the effective capacitance is reduced due to DC bias effect. The table above
therefore lists the minimum value as CAPACITANCE. In order to meet the minimum capacitance requirement, the nominal value of
the capacitor may have to be scaled accordingly to take the drop of capacitance into account for a given dc voltage at the outputs of
regulators.
(3) Ceramic capacitors recommended
(4) Load transient voltage must be considered when selecting UV/OV threshold levels for the LDO output
(5) Additional capacitance, including local and POL, beyond the specified value can cause the LDO to become unstable
(1) Input capacitors must be placed as close as possible to the device pins.
(2) When DC voltage is applied to a ceramic capacitor, the effective capacitance is reduced due to DC bias effect. The table above
therefore lists the minimum value as CAPACITANCE. In order to meet the minimum capacitance requirement, the nominal value of
the capacitor may have to be scaled accordingly to take the drop of capacitance into account for a given dc voltage at the outputs of
regulators.
(3) Ceramic capacitors recommended
(4) Additional capacitance, including local and POL, beyond the specified value can cause the LDO to become unstable
(1) When DC voltage is applied to a ceramic capacitor, the effective capacitance is reduced due to DC bias effect. The table above
therefore lists the minimum value as CAPACITANCE. In order to meet the minimum capacitance requirement, the nominal value of
the capacitor may have to be scaled accordingly to take the drop of capacitance into account for a given DC voltage at the outputs of
regulators.
ILIM_FWD_PEAK
4.10 Forward current limit step Size 1 A
_Step
(1) Input capacitors must be placed as close as possible to the device pins.
(2) When DC voltage is applied to a ceramic capacitor, the effective capacitance is reduced due to DC bias effect. The table above
therefore lists the minimum value as CAPACITANCE. In order to meet the minimum capacitance requirement, the nominal value of
the capacitor may have to be scaled accordingly to take the drop of capacitance into account for a given DC voltage at the outputs of
regulators.
(3) The maximum output current can be limited by the forward current limit. The maximum output current is also limited by the junction
temperature and maximum average current over lifetime. The power dissipation inside the die increases the junction temperature and
limits the maximum current depending on the length of the current pulse, efficiency, board and ambient temperature.
(4) Additional cooling strategies may be necessary to keep the device junction temperature at recommended limits with large output
current.
(5) SLEW_RATEx[2:0] register default comes from NVM memory, and can be re-configured by the MCU. Output capacitance, forward and
negative current limits and load current may limit the maximum and minimum slew rates.
(6) The final PFM-to-PWM and PWM-to-PFM switchover current varies slightly and is dependent on the output voltage, input voltage, and
the inductor current level. The BUCK Regulator does not switch over from PWM to PFM for Fsw=2.2MHz and VOUT < 0.5V
(7) Please refer to the applications section of the data sheet regarding the power delivery network (PDN) used for the transient load step
and output ripple test conditions. All ripple specs are defined across POL capacitor in the described PDN.
(8) The 33.3 mV/µs slew-rate setting is not recommended for LBx ≥ 1 µH, as this can trigger OV detection due to larger overshoot at the
buck output.
(1) The default values of BUCKn_OV_THR & BUCKn_UV_THR registers come from the NVM memory, and can be re-configured by
software.
(2) The default values of LDOn_OV_THR & LDOn_UV_THR registers come from the NVM memory, and can be re-configured by software.
(3) The default values of VCCA_OV_THR & VCCA_UV_THR registers come from the NVM memory, and can be re-configured by
software.
(4) Interrupt status signal is input signal for PGOOD deglitch logic.
(1) Customer must use the XTAL_SEL bit to select the corresponding crystal based on its load capacitance.
(2) External capacitors must be used if crystal load capacitance > 6 pF.
TSD_IMM_INT thermal
8.3a TSD_imm 140 150 160 °C
shutdown rising threshold
Input slew rate of VCCA and Measured at VCCA and PVIN_x pins as voltage rises
9.15 VVCCA_PVIN_SR 60 mV/µs
PVIN_x supplies from 0V to VPOR_Rising
Measured at VIO pin as voltage rises from 0V
9.16 VVIO_SR Input slew rate of VIO supply 60 mV/µs
to VPOR_Rising
Input slew rate of VBACKUP
9.17 VVBACKUP_SR Measured at VBACKUP pin 60 mV/µs
supply
Timing Requirements
VCCA_PG_SEL = 0b. Total delay from detection of
26.3a 15 µs
tlatency_VCCAOV VCCA_OVP signal latency VCCA_OVP to the rise of VCCA_OVP_INT
P from detection VCCA_PG_SEL = 1b. Total delay from detection of
26.3b 15 µs
VCCA_OVP to the rise of VCCA_OVP_INT
Measured time between VVCCA falling from 3.3 V to
tlatency_VCCAUVL VCCA_UVLO signal latency
26.4 2.7 V with ≤ 100mv/µs slope, to the detection of 10 µs
O from detection
VCCA_UVLO signal
LDOVINT OVP and UVLO
26.5 tlatency_VINT With 25-mV overdrive 12 µs
signal latency from detection
26.15 tLBISTrun Run time for LBIST 1.8 ms
Device initialization time to
tINIT_NVM_ANAL load default values for NVM
26.16 2 ms
OG registers, and start-up analog
circuits
Device initialization time for
tINIT_REF_CLK_L
26.17 reference bandgaps, system 1 ms
DO
clock, and internal LDOs
(1) End of charge (EOC) voltage measured when VCCA-VBACKUP > 200mV. When VCCA-VBACKUP is ≤ 200mV, the charger remains
fully functional, although the EOC voltage measurement is not based on final voltage, but on charger dropout.
(1) ENABLE signal deglitch is not available when device is activated from the LP_STANDBY state while the deglitching clock is not
available.
50 LP STANDBY, no OVP
Quiescent Current (µA)
TA = 25°C TA = 25°C
Figure 7-1. Quiescent Current vs Input Voltage Figure 7-2. Standby Current with VCCA Monitor
4 1.5
1.3
3
Total Active Phases
VVOUT_Bn (V)
1.1
2 SR = 33.3 V/ms
0.9 SR = 20 V/ms
SR = 10 V/ms
SR = 5 V/ms
1 2.2 MHz, Adding SR = 2.5 V/ms
2.2 MHz, Shedding 0.7 SR = 1.25 V/ms
4.4 MHz, Adding SR = 0.625 V/ms
4.4 MHz, Shedding SR = 0.3125 V/ms
0 0.5
0 1 2 3 4 5 6 0.5 1 1.5 2 2.5 3 3.5 4
IOUT_Bn (A) Time (ms)
VPVIN_Bn = 3.3 V Buck VSET = 1.0 V TA = 25°C VPVIN_Bn = 3.3 V Buck VSET = 0.6 V TA = 25°C
to 1.4 V
Figure 7-3. Buck Phase Adding and Shedding
Figure 7-4. Buck Ramp-up Slew Rate
1.5 1.2
SR = 33.3 V/ms
SR = 20 V/ms
1
SR = 10 V/ms
1.3
SR = 5 V/ms
SR = 2.5 V/ms 0.8
SR = 1.25 V/ms
VVOUT_Bn (V)
VVOUT_Bn (V)
VPVIN_Bn = 3.3 Buck VSET = 1.4 V to TA = 25°C VPVIN_Bn = 3.3 V Buck VSET = 1 V Slew Rate = 5 V/ms
V 0.6 V Figure 7-6. Buck Start-up with no Load, Auto Mode
Figure 7-5. Buck Ramp-down Slew Rate
1 1
0.8 0.8
VVOUT_Bn (V)
VVOUT_Bn (V)
0.6 0.6
2.2MHz, 1-phase 2.2MHz, 1-phase
0.4 2.2MHz, 2-phase 0.4 2.2MHz, 2-phase
2.2MHz, 3-phase 2.2MHz, 3-phase
0.2 2.2MHz, 4-phase 0.2 2.2MHz, 4-phase
4.4MHz, 1-phase 4.4MHz, 1-phase
4.4MHz, 2-phase 4.4MHz, 2-phase
0 4.4MHz, 3-phase 0 4.4MHz, 3-phase
4.4MHz, 4-phase 4.4MHz, 4-phase
-0.2 -0.2
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4
Time (ms) Time (ms)
VPVIN_Bn = 3.3 V Buck VSET = 1 V Slew Rate = 5 V/ms VPVIN_Bn = 3.3 V Buck VSET = 1 V Slew Rate = 5 V/ms
Figure 7-7. Buck Start-up with 1A Load, Auto Mode Figure 7-8. Buck Shutdown with no Load, Auto Mode
1.2 1.6
1
1.4
0.8
VVOUT_Bn (V)
VVOUT_Bn (V)
1.2
0.6
2.2MHz, 1-phase
0.4 2.2MHz, 2-phase 1
2.2MHz, 3-phase
0.2 2.2MHz, 4-phase
4.4MHz, 1-phase
4.4MHz, 2-phase 0.8
0 4.4MHz, 3-phase No Load
4.4MHz, 4-phase with 1A load
-0.2 0.6
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0 5 10 15 20 25 30 35
Time (ms) Time (us)
VPVIN_Bn = 3.3 V Buck VSET = 1 V Slew Rate = 5 V/ms VPVIN_Bn = 3.3 V Buck VSET = 0.6 V Slew Rate = 33.3
Figure 7-9. Buck Shutdown with 1A Load, Auto Mode to 1.4 V V/ms
Figure 7-10. Buck Ramp-up with and without Load
1.5 3.5
No Load 3.3 V to 0.8 V
1.4 with 1A load 3 3.3 V to 1.8 V
5 V to 0.8 V
1.3 5 V to 1.8 V
2.5
5 V to 3.3 V
1.2 Bypass Mode
VOUT(LDOn) (V)
VVOUT_Bn (V)
2
1.1
1.5
1
1
0.9
0.5
0.8
0.7 0
0.6 -0.5
0 10 20 30 40 50 60 70 80 90 0 0.4 0.8 1.2 1.6 2 2.4 2.8 3.2 3.6 4
Time (us) Time (ms)
VPVIN_Bn = 3.3 Buck VSET = 1.4 V to Slew Rate = 33.3 VIN(LDOn) = 3.3 V or 5 V TA = 25°C
V 0.6 V V/ms Figure 7-12. GPLDO Start-up with LDOn_SLOW_RAMP = 0
Figure 7-11. Buck Ramp-down with and without Load
VOUT(LDOn) (V)
2
2
1.5
1.5
1
1
0.5
0 0.5
-0.5 0
0 0.4 0.8 1.2 1.6 2 2.4 2.8 3.2 3.6 4 0 2 4 6 8 10 12 14 16 18 20
Time (ms) Time (ms)
VOUT(LDOn) (V)
2 2
1.5 1.5
1 1
0.5 0.5
0 0
-0.5 -0.5
0 0.4 0.8 1.2 1.6 2 2.4 2.8 3.2 3.6 4 0 0.4 0.8 1.2 1.6 2 2.4 2.8 3.2 3.6 4
Time (ms) Time (ms)
1.5
0.5
0
0 2 4 6 8 10 12 14 16 18 20
Time (ms)
8 Detailed Description
8.1 Overview
The TPS6593-Q1 device is a Power-Management Integrated Circuit (PMIC), available in a 56-pin, 0.5-mm pitch,
8-mm × 8-mm QFN package. The TPS6593-Q1 is designed for powering embedded systems or System on Chip
(SoC) in automotive or industrial applications. The TPS6593-Q1 provides five configurable BUCK regulators, of
which four rails have the ability to combine outputs in multi-phase mode. BUCK4 has the ability to supply up to
4 A output current in single-phase mode, while BUCK1, BUCK2, and BUCK3 have the ability to supply up to
3.5 A output current in single-phase mode. When working in multi-phase mode, each BUCK1, BUCK2, BUCK3,
and BUCK4 can supply up to 3.5 A output current per phase, adding up to 14 A output current in four-phase
configuration. BUCK5 is a single-phase only BUCK regulator, which supports up to 2 A output current. All five
of the BUCK regulators have the capability to sink a current up to 1 A, and support dynamic voltage scaling.
Double-buffered voltage scaling registers enable each BUCK regulator to transition to a different voltage during
operation by SPI or I2C. A digital PLL enables the BUCK regulators to synchronize to an external clock input,
with phase delays between the output rails.
The TPS6593-Q1 device also provides three LDO rails, which can supply up to 500 mA output current per rail
and can be configured in bypass mode and used as a load switch. One additional low-noise LDO rail can supply
up to 300 mA output current. The 500-mA LDOs support 0.6 V to 3.3 V output voltage with 50-mV step. The
300-mA low-noise LDO supports 1.2 V to 3.3 V output voltage with 25-mV step. The output voltages of the LDOs
can be pre-configured through the SPI or I2C interfaces, which are used to configure the power rails and the
power states of the TPS6593-Q1 device.
I2C channel 1 (I2C1) is the main channel with access to the registers, which control the configurable power
sequencer, the states and the outputs of power rails, the device operating states, the RTC registers and the
Error Signal Monitors. I2C channel 2 (I2C2), which is available through the GPIO1 and GPIO2 pins, is dedicated
for accessing the Q&A Watchdog communication registers. If GPIO1 and GPIO2 are not configured as I2C2
pins, I2C1 can access all of the registers, including the Q&A Watchdog registers. Alternatively, depending on the
NVM-configuration of the orderable part number, SPI is the selected interface for the device and can be used to
access all registers.
The TPS6593-Q1 device includes an internal RC-oscillator to sequence all resources during power up and
power down. Two internal LDOs (LDOVINT and LDOVRTC) generate the supply for the entire digital circuitry of
the device as soon as the external input supply is available through the VCCA input. A backup battery supply
input can also be used to power the RTC block and a 32-kHz Crystal Oscillator clock generator in the event of
main supply power loss.
TPS6593-Q1 device has eleven GPIO pins each with multiple functions and configurable features. All of the
GPIO pins, when configured as general-purpose output pins, can be included in the power-up and power-down
sequence and used as enable signals for external resources. In addition, each GPIO can be configured as a
wake-up input or a sleep-mode trigger. The default configuration of the GPIO pins comes from the non-volatile
memory (NVM), and can be re-programmed by system software if the external connection permits.
The TPS6593-Q1 device includes a watchdog with selectable trigger or Q&A modes to monitor MCU software
lockup, and two error signal monitor (ESM) inputs with fault injection options to monitor the lock-step
signal of the attached SoC or MCU. TPS6593-Q1 includes protection and diagnostic mechanisms such as
voltage monitoring on the input supply, voltage monitoring on all BUCK and LDO regulator outputs, CRC
on configuration registers, CRC on non-volatile memory, CRC on communication interfaces, current-limit and
short-circuit protection on all output rails, thermal pre-warning, and over-temperature shutdown. The device also
includes a Q&A or trigger mode watchdog to monitor for MCU software lockup, and two Error Signal Monitor
inputs with selectable level mode or PWM mode, and with fault injection options to monitor the error signals
from the attached SoC or MCU. The TPS6593-Q1 can notify the processor of these events through the interrupt
handler, allowing the MCU to take action in response.
An SPMI interface is included in the TPS6593-Q1 device to distribute power state information to at most five
satellite PMICs on the same network, thus enabling synchronous power state transition across multiple PMICs in
the application system. This feature allows the consolidation of IO control signals from up to six PMICs powering
the system into one primary TPS6593-Q1 PMIC.
VCCA
VOUT_LDOVINT
VOUT_LDOVRTC
VBACKUP
VIO_IN
<PBKG>
<GND_DIG>
<GND_ANA>
nPWRON/ENABLE
Control
Interface
Grounds Windowed
PGOOD
SCL_I2C1/CLK_SPI Power-Good
LDOVINT BSM LDOVRTC Monitor
SDA_I2C1/SDI_SPI I2C CNTRL,
CS_SPI or SPI VINT VRTC SYNCCLKOUT
SDO_SPI Single or
Output Buffer
VCC internal Multi-Phase
nRSTOUT supply
Internal EN PVIN_B1 VCCA
nINT Interrupt BUCK1
VSEL 3.5 A SW_B1
events
RC DPLL RAMP FB_B1
GPIO1
Oscillator (Phase CLK1 (AVS)
synchronization <GND_B1>
SYNCCLKIN
GPIO2 and dither)
EN PVIN_B2 VCCA
Application Processor
GPIO3 BUCK2
DFT VSEL 3.5 A SW_B2
NVM Controller RAMP FB_B2
GPIO4 NVM Memory CLK2 (AVS)
<GND_B2>
Registers
GPIO5
VCCA Pre-Configurable PVIN_B3 VCCA
Interrupt handler
EN
VCCA_UVLO Power Sequencer BUCK3
GPIO6 VSEL 3.5 A SW_B3
Controller
GPIO
RAMP FB_B3
ECO CLK3 (AVS)
GPIO7 <GND_B3>
WAKEn PWM
NSLEEPn DVS
GPIO8 Default NVM Settings
EN BUCK4 PVIN_B4 VCCA
Thermal VSEL 4 A (1N) SW_B4
GPIO9 Monitoring and 3.5A (multiN)
RAMP FB_B4
Shutdown CLK4
(AVS) <GND_B4>
GPIO10 Hot die detection
EN
VSEL
EN
VSEL
EN
VSEL
VRTC REFGND1
Bypass Bypass Bypass LDO4
LDO1 LDO2 LDO3 300 mA
Internal supply 500 mA 500 mA 500 mA Low Noise
Quiet Ground
PVIN_LDO12
VOUT_LDO2
VOUT_LDO3
VOUT_LDO4
VOUT_LDO1
AMUX_OUT Reference
PVIN_LDO3
PVIN_LDO4
and Bias
REFGND2
VCCA
* These red squares are internal pads for down-bonds to the package thermal/ground pad.
A separate voltage comparator monitors whether or not the VCCA voltage is within the expected PGOOD range
when VCCA is expected to be 5-V or 3.3-V. This voltage comparator checks at device power-up whether the
voltage on the VCCA supply pin is above the VCCA_UV threshold. Refer to Section 8.3.3 for additional detail on
the operation of the PGOOD monitor function.
LDOVINT, which is the internal supply to the digital core of the device, may attempt to restart the device when
the input voltage at VCCA pin falls or stays between VCCA_UVLO and VCCA_UV voltage levels; the voltage at
the VCCA pin, however, must be above the VCCA_UV voltage level for the device to power up properly.
Figure 8-1 shows a block diagram of the VCCA input voltage monitoring.
VSYS VCCA
Preregulator
External Protection
VCCA
Safety
Band Gap
VCCA_UVLO +
VCCA_OVP ±
PVIN
High-Side
Current Limit
Loop
FBP Comparator
Feedback Network
FBN
PWM Gate PDN
± Generator Driver
Low-Side
DAC + Current Limit
Error
CLK
Amplifier
IL_TOT_4PH
IL1
IL2
IL4
IL3
PWM1
PWM2
PWM4
PWM3
Figure 8-3. Example of PWM Timings, Inductor Current Waveforms, and Total Output Current in 4-Phase
Configuration. 1
The converter can be forced to multi-phase operation by the BUCKn_FPWM_MP bit in BUCKn_CTRL1 register.
If the regulator operates in forced multi-phase mode , each phase automatically operates in the forced-PWM
mode. If the multi-phase operation is not forced, the number of phases are added and shed automatically to
follow the required output current.
Best efficiency obtained with
Operation
Operation
Operation
Operation
1-Phase
2-Phase
3-Phase
4-Phase
N=1
N=2
Efficiency
N=3
N=4
Load Current
Figure 8-4. Multiphase BUCK Converter Efficiency vs Number of Phases (Converters in PWM Mode) 2
Table 8-2. Primary BUCK Assignment for Supported Multi-phase Configuration (continued)
Supported Multi-Phase BUCK Regulator Configuration Primary BUCK Assignment
2-Phase: BUCK1 + BUCK2 BUCK1
2-Phase: BUCK3 + BUCK4 BUCK3
When the BUCK regulators are configured in 3-phase or 4-phase configurations, there are exceptions to the
above list of registers that the TPS6593-Q1 ignores. The configuration registers are user-configurable for
the voltage monitor function on BUCK3 and BUCK4 in a 4-phase configuration and BUCK3 in a 3-phase
configuration. The UV/OV voltage monitors of these BUCK3 and BUCK4 regulators can be used to monitor
external supply rails, by connecting these external rails to the FB_Bn pins of these BUCK3 and BUCK4
regulators. The following list of registers and register bits for BUCK3 and BUCK4 can be used to enable and set
the target voltage for the external voltage monitoring function under such configuration:
• BUCKn_VMON_EN bit
• BUCKn_VSEL bit
• BUCKn_SLEW_RATE
• BUCKn_VOUT_1 and BUCKn_VOUT_2 registers
• BUCKn_PG_WINDOW register
Customers are responsible for the values set in these registers when using BUCK3 or BUCK4 to monitor an
external supply under the 3-phase or 4-phase configuration. If the voltage monitor function is not used under
such a scenario, the FB_Bn pins must be connected to the reference ground, and the BUCKn_VMON_EN bits
must be set to '0'.
8.3.2.1.5 Spread-Spectrum Mode
The TPS6593-Q1 device supports spread-spectrum modulation of the switching clock signal used by the BUCK
regulators. Three factory-selectable modulation modes are available: the first mode is modulation from external
input clock at the SYNCCLKIN pin; the second mode is modulating the input clock at the SYNCCLKIN pin using
the DPLL; the third mode is modulating the internal 20-MHz RC-Oscillator clock using the DPLL.
The spread-spectrum modulation mode is pre-configured in NVM. Changing this modulation mode during
operation is not supported.
The modulation frequency range is limited by the DPLL bandwidth. The max frequency spread for the input clock
to the DPLL is ±18% to secure parametric compliance of the BUCK output performance.
The internal modulation is inactive by default and can be enabled and configured after power up. Internal
modulation is activated by setting the SS_EN control bit. The internal modulation must be inactive (SS_EN = 0)
when changing the following parameter:
• SS_DEPTH[1:0] – Spread Spectrum modulation depth
When internal modulation is enabled and configured, it can be made inactive by the system MCU during
operation. The device transition to different mission states does not impact internal modulation when it is
enabled and configured.
8.3.2.1.6 Adaptive Voltage Scaling (AVS) and Dynamic Voltage Scaling (DVS) Support
An AVS or a DVS voltage value can be configured by the attached MCU after the BUCK regulator is powered
up to the default output voltage selected in register BUCKn_VSET1, that loads its default value from NVM. The
purpose of the AVS/DVS voltage is to set the BUCK output voltage to enable optimal efficiency and performance
of the attached SoC.
All of BUCK regulators in the TPS6593-Q1 device support AVS and DVS voltage scaling changes. Once
the AVS/DVS voltage value is written into the BUCKn_VSET1 or BUCKn_VSET2 register, and the MCU sets
the BUCKn_VSEL register to select the AVS/DVS voltage, the output of the BUCK regulator remains at the
AVS/DVS voltage level instead of the default voltage from NVM until one of the following events occur:
• Error that causes the device to re-initialize itself through a power cycle after reaching the SAFE RECOVERY
state
• Error that causes the device to execute warm reset
BUCK ENABLE *)
I2C/SPI
BUCKn_VSEL BUCKn_EN
I2C/SPI BUCKn_VSET1
DCDC
MUX Regulator
I2C/SPI BUCKn_VSET2
The digital control block automatically updates the OV and UV threshold of the BUCK output voltage monitor
during the AVS or DVS voltage change. When the output voltage is increased, the OV threshold is updated at
the same time the BUCKn_VSETx is updated to the AVS voltage level, while the UV threshold is updated after a
delay calculated by Equation 1.
When the output voltage is decreased, the UV threshold is updated at the same time the BUCKn_VSETx is
updated to the AVS voltage level, while the OV threshold is updated after a delay calculated by Equation 1.
In order to prevent erroneous voltage monitoring, the digital block also temporarily masks the results of the OV
and UV monitor from the regulator output when the BUCK regulator is enabled and the voltage is rising to the
BUCKn_VSETx level. The duration of the mask starts from the time the BUCK regulator is enabled. The BUCK
OV monitor output is masked for a fixed delay time of tPG_OV_GATE, that is approximately 115 µs – 128 µs. The
UV monitor output is masked for the time duration calculated by Equation 2. The 370-µs additional delay time in
the formula includes the start-up delay of the BUCK regulator, and the fixed delay after the ramp.
Note
Because output capacitance, forward and negative current limits and load current of the BUCK
regulator may affect the slew rate of the BUCK regulator output voltage, the delay time of tPG_UV_GATE
may not be sufficient long for the slower slew rate setting when the target BUCK regulator output
voltage is higher. Please refer to the PMIC user's guide for detail information about the supported
voltage level and slew rate setting combinations of a particular orderable part number.
Figure 8-6 and Figure 8-7 are timing diagrams illustrating the voltage change for AVS and DVS enabled BUCK
regulators and the corresponding OV and UV monitor threshold changes.
BUCKn VOUT
I2C/SPI write
State control (or State control (or State control (or
I2C/SPI write) I2C/SPI write) I2C/SPI write)
BUCKn_EN 0 1 0 1
0 us
Register bits BUCKn_OV_SET
(internal register, not 0x00 0x5F 0x5A
accessible by user)
tPG_OV_UV_DELAY
BUCKn_UV_SET 0x00 0x5F 0x5A
(internal register, not
accessible by user)
BUCKn_VSEL 0
BUCKn_OV_UV_EN 0 1 0 1
Automac control
0 us by digital
BUCKn_UV Monitor Output
BUCKn_UV Ga ng tPG_UV_GATE
tPG_UV_GATE
Figure 8-6. AVS Voltage and OV UV Threshold Level Change Timing Diagram
OPP_OD OPP_OD
(or OPP_TURBO) (or OPP_TURBO)
OPP_NOM
OV limit
OV limit
UV limit
UV limit
BUCKn VOUT
I2C/SPI write
(DVFS control) State control (or State control (or
I2C/SPI write) I2C/SPI write)
BUCKn_VSET2 0x5F
Automac control
by digital
BUCKn_EN 1 0 1
0 us
BUCKn_OV_SET
Register bits (internal register, not 0x5F 0x73
accessible by user)
tPG_OV_UV_DELAY
BUCKn_UV_SET
(internal register, not
0x5F 0x73
accessible by user)
BUCKn_VSEL 0
BUCKn_OV_UV_EN 1 0 1
0 us Automac control
Automac control by digital
by digital
BUCKn_OV Monitor Output
0 us
BUCKn_UV Monitor Output
BUCKn_UV Ga ng
tPG_UV_GATE
Figure 8-7. DVS Voltage and OV UV Threshold Level Change Timing Diagram
20 MHz
RC
Main CLK
Oscillator RESET
Detector
20 MHz
RC ÷ 18 Buck2
Oscillator
Phase and
DPLL
52.8MHz +/- 20% freq control
SYNCCLKIN /N
Detector
Divider
SYNCCLKOUT SYNCCLKOUT
_FREQ_SEL
Divider Clock Select
Spread-spec
SYNCCLKIN ´(;7_CLK_ Logic
Control
)5(4´
SEL_EXT_CLK
Note
Writing a RESERVED value to the LDOn_VSET[7:0] register bits causes a LDOn_OV_INT or
LDOn_UV_INT interrupt.
The LDO regulators do not have slew rate control for voltage ramp; by setting the LDOn_SLOW_RAMP bit to '1',
however, the ramp up speed of the regulator output voltage is < 3 V/ms.
If an LDO is not needed, its associated UV/OV Voltage Monitor can be used to monitor an external voltage rail
by connecting the external rail to the VOUT_LDOn pin. The voltage level of the monitored external rail must be
within the PGOOD monitor range of the LDOn_VSET[7:0] of the LDO. If an external resistor divider is necessary
in this case, the user must take into account the input impedance at the VOUT_LDOn pin (as shown in Figure
8-9), and adjust the resistor values to compensate for the voltage shift.
External Supply
Output
PVIN_LDOn
VOUT_LDOn
LDO
50 kŸ
Pull-Down resistor 512 kŸ
when LDOs are
Disabled
LDOn_UV_THR
+
UV
±
DAC
±
LDOnOV_THR OV
+
8.3.2.2.1 LDOVINT
The LDOVINT voltage regulator is dedicated to supply the digital and analog functions of the TPS6593-Q1
device, which are not required to be always-on and can be turned-off when the device is in low power states.
The LDOVINT voltage regulator is automatically turned on and off as needed if LP_STANDBY_SEL = '1'. The
automatic control optimizes the overall current consumption when the device is in low power LP_STANDBY
state.
The LDOVINT voltage regulator is dedicated for internal use only, and cannot be used to support external
loads. An output filtering capacitor must be connected at the VOUT_LDOVINT pin. Do not connect any other
components or external loads to this VOUT_LDOVINT pin.
8.3.2.2.2 LDOVRTC
The LDOVRTC voltage regulator supplies always-on functions, such as wake-up functions. This power resource
is active as soon as a valid VCCA is present. The LDOVRTC voltage regulator is dedicated for internal use
only, and cannot be used to support external loads. An output filtering capacitor must be connected at the
VOUT_LDOVRTC pin. Do not connect any other components or external loads to this VOUT_LDOVRTC pin.
This voltage regulator is enabled in normal mode or backup mode. The LDOVRTC voltage regulator functions in
normal mode when supplied from the main system power rail and is able to supply the input buffers of GPIO3
and GPIO4, the digital components, the crystal, and the RTC calendar module of the TPS6593-Q1 device. The
LDOVRTC voltage regulator remains on in BACKUP state when VCCA is below the VCCA_UVLO level, and the
backup power source is above the LDOVRTC_UVLO level.
Only the 32 kHz crystal and the RTC counter are activated in the BACKUP state. The RTC calendar function
remains active in the LP_STANDBY state, but the interrupt functions are reduced to maintaining the wake up
functions only. The RTC calendar and interrupt functions are fully activated in the mission states.
The customer has the option to enable the shelf mode by setting the LDORTC_DIS bit to 1 while the device is in
the MISSION state and the I2C bus is in operation and ramp down VCCA to 0 V immediately after the I2C write
has completed. This shelf mode forces the device to skip the BACKUP state and enters the NO SUPPLY state
under VCCA_UVLO condition. This mode is useful to prevent the continual draining of the backup power source
when the 32 KHz crystal and RTC counter functions are no longer needed.
8.3.2.2.3 LDO1, LDO2, and LDO3
The LDO1, LDO2 and LDO3 regulators can deliver up to 500 mA of current, with a configurable output range
of 0.6 V to 3.3 V in 50-mV steps. These 3 LDO regulators also support bypass mode, which allows an input
voltage at the PVIN_LDOn to show up at the VOUT_LDOn pin. This feature allows the LDOs to be configured
as load switches with power sequencing control. Similar to the buck regulators mentioned in Section 8.3.2.1.4,
the UV/OV Voltage Monitor of an un-used LDO regulator can also be used to monitor an external voltage rail by
connecting the external rail to the VOUT_LDOn pin.
The bypass capability to connect the input voltage to the output in bypass mode is supported when the input
voltage is within the 1.7 V to 3.5 V range. This bypass capability also allows the LDO to switch from 3.3 V in
bypass mode to 1.8 V in LDO mode or from 1.8 V in LDO mode to 3.3 V in bypass mode for an SD card I/O
supply.
The LDO1, LDO2 and LDO3 regulator include a Current-Limit to protect the internal Power-FET against
overcurrent. This Current-Limit has a fixed value between 700 mA and 1800 mA.
It is important to wait until the LDO has settled on the target voltage from the previous change when changing
the LDO output voltage setting. The worst-case voltage scaling time for LDO1, LDO2, and LDO3 is 63 µs x (7 +
the number of 50-mV steps to the new target voltage).
Table 8-4 shows the coding used to select the output voltage for LDO1, LDO2, and LDO3.
Table 8-4. Output Voltage Selection for LDO1, LDO2, and LDO3
Output Voltage Output Voltage Output Voltage Output Voltage
LDOx_VSET LDOx_VSET LDOx_VSET LDOx_VSET
[V] [V] [V] [V]
0x00 Reserved 0x10 1.20 0x20 2.00 0x30 2.80
0x01 Reserved 0x11 1.25 0x21 2.05 0x31 2.85
0x02 Reserved 0x12 1.30 0x22 2.10 0x32 2.90
0x03 Reserved 0x13 1.35 0x23 2.15 0x33 2.95
0x04 0.60 0x14 1.40 0x24 2.20 0x34 3.00
0x05 0.65 0x15 1.45 0x25 2.25 0x35 3.05
0x06 0.70 0x16 1.50 0x26 2.30 0x36 3.10
0x07 0.75 0x17 1.55 0x27 2.35 0x37 3.15
0x08 0.80 0x18 1.60 0x28 2.40 0x38 3.20
0x09 0.85 0x19 1.65 0x29 2.45 0x39 3.25
0x0A 0.90 0x1A 1.70 0x2A 2.50 0x3A 3.30
0x0B 0.95 0x1B 1.75 0x2B 2.55 0x3B Reserved
0x0C 1.00 0x1C 1.80 0x2C 2.60 0x3C Reserved
Table 8-4. Output Voltage Selection for LDO1, LDO2, and LDO3 (continued)
Output Voltage Output Voltage Output Voltage Output Voltage
LDOx_VSET LDOx_VSET LDOx_VSET LDOx_VSET
[V] [V] [V] [V]
0x0D 1.05 0x1D 1.85 0x2D 2.65 0x3D Reserved
0x0E 1.10 0x1E 1.90 0x2E 2.70 0x3E Reserved
0x0F 1.15 0x1F 1.95 0x2F 2.75 0x3F Reserved
PGOOD_SEL_TDIE_WARN
VMON
VMON
BUCKn
VMON
BUCKn
VMON
BUCKn
BUCKn PGOOD_BUCKn
PGOOD_WINDOW BUCKn
Monitor
BUCKn_ILIM
BUCKn_VSETn
BUCKn_UV_THR
BUCKn_OV_THR
BUCKn_VMON_EN
PGOOD_SEL_BUCKn
VMON
VMON
BUCKn
VMON
BUCKn
LDOn PGOOD_LDOn
PGOOD_WINDOW BUCKn
Monitor
PGOOD (GPIO9)
LDOn_VSET
LDOn_UV_THR
LDOn_OV_THR
PGOOD_POL
LDOn_VMON_EN
PGOOD_SEL_LDOn
VCCA PGOOD_VCCA
PGOOD_WINDOW
Monitor
VCCA_PG_SET
VCCA_UV_THR
VCCA_OV_THR
VCCA_VMON_EN
PGOOD_SEL_VCCA
NRSTOUT
PGOOD_SEL_NRSTOUT
NRSTOUT_SoC
PGOOD_SEL_NRSTOUT_SOC
Voltage
Powergood window
BUCKn_VSETn or
LDOn_VSET (1)
BUCKn_VSETn or Power-good
LDOn_VSET (2) window
Time
On Request
VIO
BUCKn_VMON_EN
or LDOn_VMON_EN
NRSTOUT
or NRSTOUT_SoC
tlatency tlatency
_PGOOD _PGOOD
PGOOD
(PGOOD_SEL_NRSTOUT =1
or PGOOD_SEL_NRSTOUT_SOC = 1) tlatency tlatency
_PGOOD _PGOOD
PGOOD
(PGOOD_SEL_NRSTOUT = 0
and PGOOD_SEL_NRSTOUT_SOC = 0)
The OV and UV threshold of the voltage monitors of the BUCK regulators and the LDO regulators are updated
automatically by the digital control block when the output voltage setting changes. When the output voltage of
the regulator is increased, the OV threshold is updated at the same time the _VSET of the regulator is changed.
The UV threshold is updated after a delay calculated by the delta voltage change and the slew rate setting.
When the output voltage is decreased, the UV threshold is updated at the same time the _VSET of the regulator
is changed. The OV threshold is updated after a delay calculated by the delta voltage change and the slew rate
setting. The OV and UV threshold of the BUCK and LDO output voltage monitors are calculated based on the
target output voltage set by the corresponding BUCKn_VSET1, BUCKn_VSET2, or LDOn_VSET registers, and
the deviation from the target output voltage set (the voltage window) by the corresponding BUCKn_UV_THR,
BUCKn_OV_THR, LDOn_UV_THR, and the LDOn_OV_THR registers. For the OV and UV threshold of BUCK
and LDO output monitors to update with the correct timing, the following operating procedures must be followed
when updating the _VSET values of the regulators to avoid detection of OV/UV fault:
• BUCK and LDO regulators must be enabled at the same time as or earlier than as their VMON
• New voltage level must not be set before the start-up has finished
• New voltage level must not be set before the previous voltage change (ramp plus settling time) has
completed
The voltage monitors of unused BUCK or LDO regulators can be used for external supply rails monitoring. In
three-phase configuration, the Voltage Monitor of BUCK3 (on FB_B3 pin) becomes a free available resource
for monitoring an external supply voltage. In four-phase configuration, the Voltage Monitor of both BUCK3
(on FB_B3 pin) and BUCK4 (on FB_B4 pin) become free available resources for monitoring two external
supply voltages.. The target output voltage is set by the corresponding BUCKn_VSET1, BUCKn_VSET2,
or LDOn_VSET registers, and the deviation from the target output voltage set (the voltage window) by
the corresponding BUCKn_UV_THR, BUCKn_OV_THR, LDOn_UV_THR, and the LDOn_OV_THR registers.
Following aspects need to be taken into account if Voltage Monitors of unused BUCK or LDO regulators are
used for monitoring external supply rails:
• For voltage monitors of unused LDO regulators: the voltage level applied at the VOUT_LDOx pin, inclusive
expected tolerances, must be below the supply voltage applied at the PVIN_LDOx pin
• For voltage monitors of unused BUCK and LDO regulators: the maximum nominal supply voltage of the
monitored supply rail is 3.3V
• For voltage monitors of unused BUCK regulators and for voltage monitors of BUCK3 and/or BUCK4
regulators if used in a three-phase or four-phase configuration: the configured values for the BUCKn_VSET
and the BUCKn_SLEW_RATE determine the delay-time for the voltage monitoring to become active after
the corresponding BUCKn_VMON_EN bit is set. See equation (2) in Section 8.3.2.1.6. If BUCK3 and/or
BUCK4 regulators are used in a three-phase or four-phase configuration: even though the values for the
BUCK1_VSET and BUCK1_SLEW_RATE bits determine the output voltage and slew-rate of the three-phase
or four-phase output rail, the values for BUCK3_VSET respectively BUCK4_VSET and BUCK3_SLEW_RATE
respectively BUCK4_SLEW_RATE bits determine the power-good level and the voltage monitoring delay
time for the BUCK3 respectively BUCK4 Voltage Monitors.
• For voltage monitors of unused LDO regulators: the delay-time for the voltage monitoring to become active
after the corresponding LDOn_VMON_EN bit is set it 601..606μs.
Any of the GPIO pin can also be configured as part of the power-up sequence to enable external devices such
as external BUCKs when it is configured as a general-purpose output port.
The nINT pin, the EN_DRV pin, the nRSTOUT pin and the GPIO pin assigned as nRSTOUT_SOC have
readback monitoring to detect errors on the signals. The monitoring of the EN_DRV pin checks for mismatch in
both low and high levels. For the nINT pin, the nRSTOUT pin and the GPIO pin assigned as nRSTOUT_SOC,
the readback monitoring only checks for mismatches in the low level, therefore it is allowed to combine these
signals with other external pull-down sources. The readback mismatch is continuously monitored without deglitch
circuitry during operation, and the monitoring is gated for tgate_readback period when the signal state is changed
or when a new function is selected for the GPIO pin with the GPIOn_SEL bits. NINT_READBACK_INT,
EN_DRV_READBACK_INT, NRSTOUT_READBACK_INT, and NRSTOUT_SOC_READBACK_INT are the
interrupt bits that are set in an event of a readback mismatch for these pins, respectively.
Note
All GPIO pin are set to generic input pins with resistive pull-down before NVM memory is loaded
during device power up. Therefore, if any GPIOs has external pull-up resistors connecting to a voltage
domain that is energized before the NVM memory is loaded, the GPIO pin is pulled high before the
configuration for the pin is loaded from the NVM.
Note
For GPIO pins with internal pull down enabled, additional leakage current flows into the GPIO pin if
this pin is pulled-up to a voltage higher than the voltage level of its output power domain. If the internal
pull down must be enabled, please use a resistor divider to divide down the input voltage, or use a
series resistor to connect to the input source and ensure the voltage level at the GPIO pin is below the
voltage level of its output power domain.
8.3.8 Interrupts
The interrupt registers in the device are organized in hierarchical fashion. The interrupts are grouped into the
following categories:
BUCK ERROR These interrupts indicate over-voltage (OV), under-voltage (UV), short-circuit (SC)
and over-current (ILIM) error conditions found on the BUCK regulators .
LDO ERROR These interrupts indicate OV, UV, and SC error conditions found on the LDO
regulators, as well as OV and UV error conditions found on the VCCA supply.
VMON ERROR These interrupts indicate OV and UV error conditions found on the VCCA supply.
SEVERE ERROR These errors indicate severe device error conditions, such as thermal shutdown,
PFSM sequencing and execution error and VCCA over-voltage, that causes the
device to trigger the PFSM to execute immediate shutdown of all digital outputs,
external voltage rails and monitors, and proceed to the Safe Recovery State.
MODERATE ERROR These interrupts provide warnings to the system to indicate multiple restart attempts
from SAFE RECOVERY state exceeding the allowed recovery count, multiple warm-
reset executions exceeding the allowed recovery count, detection of long press
nPWRON button, SPMI communication error, register CRC error, BIST failure, read-
back error on nRSTOUT or nINT pins, or junction temperature reaching orderly
shutdown level. These warning causes the device to trigger the PFSM to execute
orderly shutdown of all digital outputs, external voltage rails and monitors, and
proceed to the SAFE RECOVERY state.3
MISCELLANEOUS These interrupts provide information to the system to indicate detection of WDOG or
WARNING ESM errors, die temperature crossing thermal warning threshold, device passing BIST
test, or external sync clock availability.
START-UP SOURCE These interrupts provide information to the system on the mechanism that caused the
device to start up, which includes FSD, RTC alarm or timer interrupts, the activation of
the ENABLE pin or the nPRWON pin button detection.
GPIO DETECTION These interrupts indicate a High-Level or Rising-Edge detection, or indicate a Low-
Level or Falling-Edge detection at the GPIO1 through GPIO11 pins.
FSM ERROR These interrupts indicate the detection of an error that causes the device mission
INTERRUPT state changes.
All interrupts are logically combined on a single output pin, nINT (active low). The host processor can read
the INT_TOP register to find the interrupt registers to find out the source of the interrupt, and write '1' to
the corresponding interrupt register bit to clear the interrupt. This mechanism ensures when a new interrupt
occurs while the nINT pin is still active, all of the corresponding interrupt register bits retain the interrupt source
information until it is cleared by the host.
Hierarchical Structure of Interrupt Registers shows the hierarchical structure of the interrupt registers according
to the categories described above. The purpose of this register structure is to reduce the number of interrupt
register read cycles the host has to perform in order to identify the source of the interrupt. Summary of Interrupt
Signals summarizes the trigger and the clearing mechanism for all of the interrupt signals. This table also shows
which interrupt sources can be masked by setting the corresponding mask register to '1'. When an interrupt is
masked, the interrupt bit is not updated when the associated event occurs, the nINT line is not affected, and
the event is not recorded. If an interrupt is masked after the event occurred, the interrupt register bit reflects the
event until the bit is cleared. While the event is masked, the interrupt register bit is not over-written when a new
event occurs.
3 The SEVERE ERROR and the MODERATE ERROR are handled in NVM memory but TI requires that the NVM pre-configurable finite
state machine (PFSM) settings always follow this described error handling to meet device specifications.
INT_TOP[7:0]
INT_FSM_ERR[7:0]
READBACK SOC_PWR MCU_PWR ORD_ IMM_
WD_INT ESM_INT COMM_ERR_INT
_ERR_INT _ERR_INT _ERR_INT SHUTDOWN_INT SHUTDOWN_INT
INT_COMM_ERR[7:0]
I2C2_ADR I2C2_CRC COMM_ADR COMM_CRC COMM_FRM
FSM_ERR_INT
INT_SEVERE_ERR[7:0]
_ERR_INT
SEVERE
INT_MODERATE_ERR[7:0]
_ERR_INT
INT_MISC[7:0]
MISC_INT
INT_STARTUP[7:0]
NPWRON
SOFT_REBOOT_INT FSD_INT RTC_INT ENABLE_INT
_START_INT
INT_GPIO[7:0]
GPIO1_8_INT GPIO11_INT GPIO10_INT GPIO9_INT
GPIO_INT
INT_GPIO1_8[7:0]
GPIO8_INT GPIO7_INT GPIO6_INT GPIO5_INT GPIO4_INT GPIO3_INT GPIO2_INT GPIO1_INT
INT_LDO_VMON[7:0]
VCCA_INT LDO3_4_INT LDO1_2_INT
LDO_VMON_INT
INT_VMON[7:0]
VCCA_UV_INT VCCA_OV_INT
INT_LDO3_4[7:0]
LDO4_ILIM_INT LDO4_SC_INT LDO4_UV_INT LDO4_OV_INT LDO3_ILIM_INT LDO3_SC_INT LDO3_UV_INT LDO3_OV_INT
INT_LDO1_2[7:0]
LDO2_ILIM_INT LDO2_SC_INT LDO2_UV_INT LDO2_OV_INT LDO1_ILIM_INT LDO1_SC_INT LDO1_UV_INT LDO1_OV_INT
INT_BUCK[7:0]
BUCK5_INT BUCK3_4_INT BUCK1_2_INT
INT_BUCK5[7:0]
BUCK_INT
INT_BUCK3_4[7:0]
BUCK4_ILIM_INT BUCK4_SC_INT BUCK4_UV_INT BUCK4_OV_INT BUCK3_ILIM_INT BUCK3_SC_INT BUCK3_UV_INT BUCK3_OV_INT
INT_BUCK1_2[7:0]
BUCK2_ILIM_INT BUCK2_SC_INT BUCK2_UV_INT BUCK2_OV_INT BUCK1_ILIM_INT BUCK1_SC_INT BUCK1_UV_INT BUCK1_OV_INT
(1) The results shown in this column are selected to meet functional safety assumptions and device specifications. The actual results
can be configured differently in NVM memory. TI recommends reviewing of the system and device functional safety goal and
documentation before deviating from these recommendations.
(2) Interrupt is generated during clock detector operation and in case clock is not available when clock detector is enabled.
(3) This event does not occur if RECOV_CNT_THR = 0, even though RECOV_CNT continues to accumulate and increase, and eventually
saturates when it reaches the maximum count of 15.
(4) Due to a digital logic errata in the device, a write or read SPI command which coincide with the rising edge of the CS signal may not
cause the COMM_FRM_ERR_INT interrupt.
(5) I2C1, I2C2, or SPI address error only occur in safety applications if the interface CRC feature is enabled, when both
I2C1_SPI_CRC_EN and I2C2_CRC_EN are set to '1'.
8.3.9 RTC
8.3.9.1 General Description
The RTC is driven by the 32-kHz oscillator and it provides the alarm and time-keeping functions.
32-kHz Frequency
Week days Control
counter compensation
INT_TIMER
The user can round to the closest minute, by setting the ROUND_30S register bit in the RTC_CTRL_REG
register. TC values are set to the closest minute value at the next second. The ROUND_30S bit is automatically
cleared when the rounding time is performed.
Example:
• If current time is 10H59M45S, round operation changes time to 11H00M00S
• If current time is 10H59M29S, round operation changes time to 10H59M00S
8.3.9.2.1 TC Registers Read Access
TC register read access can be done in two ways:
• A direct read to the TC registers. In this case, there can be a discrepancy between the final time read and
the real time because the RTC keeps running because some of the registers can toggle in between register
accesses. Software must manage the register change during the reading.
• Read access to shadowed TC registers. These registers are at the same addresses as the normal TC
registers. They are selected by setting the GET_TIME bit in the RTC_CTRL_REG register. When this bit
is set, the content of all TC registers is transferred into shadow registers so they represent a coherent
timestamp, avoiding any possible discrepancy between them. When processing the read accesses to the
TC registers, the value of the shadowed TC registers is returned so it is completely transparent in terms of
register access.
8.3.9.2.2 TC Registers Write Access
TC registers write accesses can be done while RTC is stopped. MCU can stop the RTC by the clearing the
STOP_RTC bit of the control register and checking the RUN bit of the status to be sure that RTC is frozen. MCU
then updates the TC values and restarts the RTC by setting the STOP_RTC bit, which ensures that the final
written values are aligned with the targeted values.
HOURS_REG 3 4 5 6
HOURS_REG 3 4
SECONDS_REG 58 59 0 2 3
Register Compensation
Update Event
Trigger mode In trigger mode, the MCU applies a pulse signal with a minimum pulse width of tWD_pulse on the
pre-assigned GPIO input pin to send the required watchdog trigger. To select this mode, the
MCU must clear bit WD_MODE_SELECT. Watchdog Trigger Mode provides more details.
Q&A In Q&A mode, the MCU sends watchdog answers through the I2C1 bus , I2C2 bus or SPI
(question and bus. (Which of these communication busses is to be used depends on the NVM configuration.
answer) mode Please refer to the user's guide of the orderable part number for further details). To select this
mode, the MCU must set bit WD_MODE_SELECT. Watchdog Question-Answer Mode provides
more details.
The device clears the WD_FAIL_CNT[3:0] each time the watchdog enters the Long Window. The status bits
WD_FAIL_INT and WD_RST_INT are latched until the MCU writes a ‘1’ to these bits.
Overview of Watchdog Fail Counter Value Ranges and Corresponding Device Status gives an overview of the
Watchdog Fail Counter value ranges and the corresponding device status.
Table 8-8. Overview of Watchdog Fail Counter Value Ranges and Corresponding Device Status
Watchdog Fail Counter value
Device Status
WD_FAIL_CNT[3:0]
MCU can set the ENABLE_DRV bit if WD_FIRST_OK=1 and no
WD_FAIL_CNT[3:0] ≤ WD_FAIL_TH[2:0]
other error-flags are set.
WD_FAIL_TH[2:0] < WD_FAIL_CNT[3:0] ≤ (WD_FAIL_TH[2:0] + The device sets error-flag WD_FAIL_INT and pulls the nINT pin low.
WD_RST_TH[2:0]) Furthermore, , the device clears the ENABLE_DRV bit.
If configuration bit WD_RST_EN=1, device generates WD_ERROR
trigger in the state machine and reacts as defined in the PFSM,
WD_FAIL_CNT[3:0] > (WD_FAIL_TH[2:0] + WD_RST_TH[2:0])
sets the error-flag WD_RST_INT, and pulls the nINT pin low. See
Summary of Interrupt Signals for the interrupt handling of WD_RST.
Note
If the MCU software changes the duration of the Long-Window to an interval shorter than the time in
which the watchdog has been in the Long-Window, the time-out function of the Long-Window does no
longer operate.
When the MCU clears bit WD_EN, the watchdog goes out of the Long Window and disables the watchdog.
When the watchdog is deactivated in this way, the MCU can set bit WD_EN back to ‘1’ to enable the watchdog
again, and the MCU can control the ENABLE_DRV bit when no other error-flags are set. The MCU must clear bit
WD_PWRHOLD before setting bit WD_EN back to ‘1’ to start the watchdog in Long Window.
The watchdog locks the following configuration register bits when it goes out of the Long Window and starts the
first watchdog sequence:
• WD_WIN1[6:0]
• WD_WIN2[6:0]
• WD_LONGWIN[7:0]
• WD_MODE_SELECT
• WD_QA_FDBK[1:0], WD_QA_LFSR[1:0] and WD_QUESTION_SEED[3:0]
• WD_RST_EN, WD_EN, WD_FAIL_TH[2:0] and WD_RST_TH[2:0]
8.3.10.3 MCU to Watchdog Synchronization
In order to go out of the Long Window and start the first watchdog sequence, the MCU must do the following
before elapse of the Long Window time interval:
• Clear bits WD_PWRHOLD (more detail in Section 8.3.10.4)
• Apply a pulse signal with a minimum pulse-width tWD_pulse on the pre-assigned GPIO pin in the case the
watchdog is configured for Trigger mode, or
• Write four times to WD_ANSWER[7:0] in the case the watchdog is configured for Q&A mode
When the MCU fails to get the watchdog out of the Long Window before the configured Long
Window time interval (tLONG_WINDOW) elapses, the device goes through a warm reset, and sets the
WD_LONGWIN_TIMEOUT_INT. This bit latched until the MCU writes a ‘1’ to clear it.
8.3.10.4 Watchdog Disable Function
The watchdog in the TPS6593-Q1 device has a Watchdog Disable function to prevent an unwanted MCU
reset in case the MCU is un-programmed or needs to be reprogrammed. In order to activate this Watchdog
Disable function for an un-programmed MCU, DISABLE_WDOG pin must be asserted to a logic-high level for
a time-interval longer than tWD_DIS prior to the moment the device releases the nRSTOUT pin. If the Watchdog
Disable function is activated in this way, the device sets bit WD_PWRHOLD to keep the watchdog in the Long
Window. The watchdog stays in the Long Window until the MCU clears the WD_PWRHOLD bit.
In case the MCU needs to be reprogrammed while the watchdog monitors the correct operation of the MCU,
the MCU can set bit WD_RETURN_LONGWIN to put the watchdog back in the Long Window. When the MCU
set this bit, the watchdog returns to the Long Window after the current Watchdog Sequence completes. In order
to make the watchdog stay in the Long Window as long as needed the MCU can either re-configure the Long
Window (tLONG_WINDOW) time interval, or set the WD_PWRHOLD bit. Once the MCU starts the first watchdog
sequence (as described in Section 8.3.10.3), the MCU must clear bit WD_RETURN_LONGWIN before the end
of the first watchdog sequence in order to continue the watchdog sequence operation.
8.3.10.5 Watchdog Sequence
Once the watchdog is out of the Long Window, each watchdog sequence starts with a Window-1 followed by
a Window-2. The watchdog ends the current sequence and after one 20-MHz system clock cycle starts a next
sequence when one of the events below occurs:
• The configured Window-2 time period elapses
• The watchdog detects a pulse signal with a minimum pulse-width tWD_pulse on the pre-assigned GPIO pin if
the watchdog is used in Trigger mode
• The watchdog detects four times a write access to WD_ANSWER[7:0] in case the watchdog is used in Q&A
mode
The MCU can configure the time periods of the Window-1 (tWINDOW1) and Window-2 (tWINDOW2) with the bits
WD_WIN1[6:0] and WD_WIN2[6:0] respectively, before starting the sequence.
Use Equation 5 and Equation 6 to calculate the minimum and maximum values for the tWINDOW1 time interval.
Use Equation 7 and Equation 8 to calculate the minimum and maximum values for the tWINDOW-2 time interval.
YES
RESTART
from all Reset-Extension
NO
states except me-interval
elapsed?
NO SUPPLY
YES
- Device sets
WD_TRG_EARLY
NO WD_PWRHOLD=0? error-ag Device has
YES NO WINDOW-1
- Device sets Received trigger-
me-interval
WD_BAD_EVENT pulse ?
elapsed?
YES error-ag
YES
NO WD_EN=0?
WINDOW-2
- If FIRST_WD_OK=0, device forces
YES ENABLE_DRV=0, else device does not
YES change ENABLE_DRV bit
- MCU sends trigger-pulse
NO
NORMAL – NO Watchdog
- MCU reset inacve WD_EN=0? NO
- MCU can set Device sets
ENABLE_DRV=1 if no other WD_TIMEOUT error- Device has
WINDOW-2
error-ags are set ag YES
me-interval
NO Received trigger- Device clears
- Interrupt inac ve if no - Device sets pulse ? WD_BAD_EVENT
elapsed?
other error- ag set WD_BAD_EVENT error-ag
error-ag
YES
NO
Device Increments
WD_FAIL_CNT[3:0]
LONG-WINDOW
NO
me-interval
elapsed?
WATCHDOG-RESET
- Device forces ENABLE_DRV=0
- Device pulls MCU & SoC reset
- Device sets WD_FAIL_INT
pins low (trigger to FSM)
error-ag
- Device forces ENABLE_DRV=0
- Interrupt acve
- Device clears WD_FIRST_OK bit
Reset-Extension
YES NO
me-interval
elapsed?
WD_FAIL_CNT[3:0] >
- Device sets WD_RST_INT (WD_FAIL_TH[2:0] +
error-ag YES
WD_RST_TH[2:0])
NO
- Interrupt ac ve &
- Device clears WD_RST_EN=1
WD_BAD_EVENT error-!ag
Figure 8-16, Figure 8-17, Figure 8-18, Figure 8-19, and Figure 8-20 give examples of watchdog is trigger mode
with good and bad events after device start-up. In these figures, the red bended arrows indicate a delay of one
20-MHz system clock cycle.
tt <
tt < tLONG_WINDOWt tt = tWINDOW-1t tt < tWINDOW-2t tt = tWINDOW-1t
tWINDOW-2t
WD_FIRST_OK x 0 1
WD_FAIL_INT x 0
WD_RST_INT x 0
WD_LONGWIN_TIMEOUT
x 0
_INT
WD_TRIG_EARLY x 0
WD_TIMEOUT x 0
ENABLE_DRV x 0 1
RECOV_CNT[2:0] 000
Figure 8-16. Watchdog in Trigger Mode – Normal MCU Start-up with Correct Watchdog-Triggers
nRSTOUT
(Reset to MCU)
Watchdog-Trigger on
GPIO pin
Internally Generated
Trigger Pulse
tt = tLONG_WINDOWt tt = tLONG_WINDOWt
WD_FIRST_OK x 0 0 0
WD_FAIL_INT x 0
WD_RST_INT x 0
WD_LONGWIN_
x 0 1 1 1
TIMEOUT_INT
WD_TRG_EARLY x 0
WD_TIMEOUT x 0
ENABLE_DRV x 0 0 0 0
Figure 8-17. Watchdog in Trigger Mode – MCU Does Not Send Watchdog-Triggers After Start-up
RESET Extension
RESET Extension Time
Time
nRSTOUT
(Reset to MCU)
t > tWD_pulse t > tWD_pulse t > tWD_pulse t > tWD_pulse
Watchdog-Trigger
on GPIO pin
WD_FAIL_CNT[3:0]
WD_FAIL_TH[2:0]=000 xxxx 0000 0000 0001 0010 0000
WD_RST_TH[2:0]=001
WD_FAIL_CNT WD_FAIL_CNT >
> WD_FAIL_TH WD_FAIL_TH + WD_RST_TH
WD_FIRST_OK x 0 1 0
MCU clears watchdog error-flags
WD_FAIL_INT x 0 1
WD_RST_INT x 0 1
WD_LONGWIN x 0
_TIMEOUT_INT
WD_TRG_EARLY x 0 1 1
WD_TIMEOUT x 0
ENABLE_DRV x 0 1 0
Figure 8-18. Watchdog in Trigger Mode – Bad Event (Watchdog-Triggers in Window-1) After Start-up
RESET
RESET Extension Time Extension Time
nRSTOUT
(Reset to MCU)
t > tWD_pulse t > tWD_pulse t < tWD_pulse
Watchdog-Trigger
on GPIO pin
Internally Generated
Trigger Pulse
Watchdog Windows Long Window Window-1 Window-2 Window-1 Window-2 Window-1 Window-2 Long Window
WD_FAIL_CNT[3:0]
xxxx 0000 0000 0001 0010 0000
WD_FAIL_TH[2:0]=000
WD_RST_TH[2:0]=001 WD_FAIL_CNT >
WD_FAIL_CNT > WD_FAIL_TH WD_FAIL_TH + WD_RST_TH
FIRST_WD_OK x 0 1 0
WD_FAIL_INT x 0 1
WD_RST_INT x 0 1
WD_LONGWIN_ x 0
TIMEOUT_INT
WD_TRIG_EARLY x 0
x 0 1
WD_TIMEOUT
MCU sets ENABLE_DRV (only possible when FIRST_WD_OK=1)
ENABLE_DRV x 0 1 0
Figure 8-19. Watchdog in Trigger Mode – Bad Events (Too Short or no Trigger in Window-2) After Start-up
Watchdog-Trigger
on GPIO pin
tWD_pulse tWD_pulse tWD_pulse tWD_pulse tWD_pulse
Internally Generated
Trigger Pulse tt <
tt < tt < tt <
tLONG_WINDOW tt = tWINDOW-1t tt = tWINDOW-1t tt = tWINDOW-2t tt = tWINDOW-1t tt = tWINDOW-1t
tWINDOW-2t tWINDOW-2t tWINDOW-2t
t
Watchdog Windows Long Window Window-1 Window-2 Window-1 Window-2 Window-1 Window-2 Window-1 Window-2 Window-1
WD_FAIL_CNT[3:0] 00
WD_FAIL_TH[2:0]=000 xxxx 0000 0000 0001 0000
00
WD_RST_TH[2:0]=001
WD_FAIL_CNT > WD_FAIL_TH
FIRST_WD_OK x 0 1
MCU clears WD_FAIL_TH error-flag (only
possible when WD_FAIL_CNT =< WD_FAIL_TH)
MCU clears watchdog error-flags
WD_FAIL_INT x 0 1 0
WD_RST_INT x 0
WD_LONGWIN_ x 0
TIMEOUT_INT
WD_TRIG_EARLY x 0
WD_TIMEOUT x 0 1
MCU sets ENABLE_DRV (only possible when MCU sets ENABLE_DRV (only
FIRST_WD_OK=1) possible when FIRST_WD_OK=1)
ENABLE_DRV x 0 1 0 1
RECOV_CNT[2:0] 000
Figure 8-20. Watchdog in Trigger Mode – Good Events (Correct Watchdog-Triggers) After Start-up, Followed by a Bad-Event (No Watchdog-
Trigger in Window-2) and After That Followed by a Good Event.
Answer An answer is a 32-bit word that is split into four answer bytes: Answer-3, Answer-2, Answer-1, and
Answer-0.
The watchdog receives an answer-byte when the MCU writes to the WD_ANSWER[7:0] bits. For
each question, the watchdog requires four correct answer-bytes from the MCU in the correct timing
and order (Answer-3, Answer-2, and Answer-1 in Window 1 in the correct sequence, and Answer-0
in Window 2) to detect a good event.
The watchdog sequence in Q&A mode ends after the MCU writes the fourth answer byte (Answer-0), or after a
time-out event when the Window-2 time-interval elapses.
Window-1 Window-2
t = tWINDOW-1 t = tWINDOW-2
Three correct answer-bytes must be provided in Window-1 and
The fourth answer-byte, Answer-0, must be provided
in the correct order:
in Window-2.
x Answer-3
x Answer-2
After the MCU writes the fourth Answer-0 to
x Answer-1
WD_ANSWER[7:0], the Watchdog generates the
next question within 1 Internal System Clock Cycle,
After the Window-1 time elapses, Window 2 begins.
after which the next Watchdog Sequence
(Q&A [n + 1]) begins
The MCU needs to write the answer-bytes to the WD_ANSWER[7:0] bits.
Question Answer
(1) (2)
MCU reads question MCU provides answer
Watchdog Sequence
Note
The Question-Generator is only re-initialized (starting with question 0000) at device power-up. In
following situations, the MCU software needs to read the current question in order to synchronize with
the Question-Generator:
• After MCU re-boot from a warm-reset
• After MCU software sets bit WD_RETURN_LONGWIN=1 to put the Watchdog back into Long
Window
• After MCU wrote WD_EN=0, then reenable Watchdog again with WD_EN=1
Figure 8-22 shows the logic combination for the WD_QUESTION[3:0] generation.
Figure 8-23 shows how the logic combination of the question-counter with the WD_ANSW_CNT[1:0] status bits
generates the reference answer-bytes.
x1 x2 x3 x4
Bit 0 Bit 1 Bit 2 Bit 3
4-bit SEED Value Loaded when the device goes to the RESET state
(Configurable Through WD_QUESTION_SEED[3:0])
(Default Value 4'b1010)
x1 x2 x3 x4 x2 00
x1 01 WD_QUESTION[0]
SEED 1 0 1 0 x4 10
x3 11
1 1 1 0 1
2 1 1 1 0 QST_CNT[1] 00
QST_CNT[0] 01
3 1 1 1 1 QST_CNT[3] 10
QST_CNT[2] 11
4 0 1 1 1
Question Sequence Order 1 to 15
5 0 0 1 1 x4 00
x3 01 WD_QUESTION[1]
6 0 0 0 1 x2 10
x1 11
7 1 0 0 0
8 0 1 0 0 QST_CNT[3] 00
QST_CNT[2] 01
QST_CNT[1] 10
9 0 0 1 0 QST_CNT[0] 11
10 1 0 0 1
x1 00
11 1 1 0 0 x4 01 WD_QUESTION[2]
x3 10
12 0 1 1 0 x2 11
13 1 0 1 1
QST_CNT[0] 00
14 0 1 0 1 QST_CNT[3] 01
QST_CNT[2] 10
15 1 0 1 0 QST_CNT[1] 11
(1) If the current value for bits (x1, x2, x3, x4) is 4'b0000, the next value for these bits (x1, x2, x3, x4) is 4'b0001, and all further question
generation begins from this value.
WD_QUESTION[0] 00
WD_QUESTION[1] 01 Reference-Answer-X[0]
WD_QUESTION[2] 10
WD_QUESTION[3] 11 X = 3, 2,1, 0
WD_ANSW_CNT[1]
WD_QUESTION[3] 00
WD_QUESTION[2] 01
WD_QUESTION[1] 10
WD_QUESTION[0] 11
WD_QUESTION[0] 00
WD_QUESTION[1] 01 Reference-Answer-X[1]
WD_QUESTION[2] 10
WD_QUESTION[3] 11 X = 3, 2,1, 0
WD_QUESTION[2] 00
WD_QUESTION[1] 01
WD_QUESTION[0] 10
WD_QUESTION[3] 11
WD_QUESTION[1]
WD_ANSW_CNT[1]
WD_QUESTION[0] 00
WD_QUESTION[3] 01 Reference-Answer-X[2]
WD_QUESTION[1] 10
WD_QUESTION[1] 11 X = 3, 2,1, 0
WD_QUESTION[3] 00
WD_QUESTION[2] 01
WD_QUESTION[1] 10
WD_QUESTION[0] 11
WD_QUESTION[1]
WD_ANSW_CNT[1]
WD_QUESTION[2] 00
WD_QUESTION[1] 01
WD_QUESTION[0] Reference-Answer-X[3]
10
WD_QUESTION[3] 11 X = 3, 2,1, 0
WD_QUESTION[0] 00
WD_QUESTION[3] 01
WD_QUESTION[2] 10
WD_QUESTION[1] 11
WD_QUESTION[3]
WD_ANSW_CNT[1]
WD_QUESTION[1] 00
WD_QUESTION[0] 01 Reference-Answer-X[4]
WD_QUESTION[2] 10
WD_QUESTION[3] 11 X = 3, 2,1, 0
WD_ANSW_CNT[0]
WD_QUESTION[3] 00
WD_QUESTION[2] 01
WD_QUESTION[1] Reference-Answer-X[5]
10
WD_QUESTION[0] 11 X = 3, 2,1, 0
WD_ANSW_CNT[0]
WD_QUESTION[0] 00
WD_QUESTION[3] 01
WD_QUESTION[2] 10 Reference-Answer-X[6]
WD_QUESTION[1] 11 X = 3, 2,1, 0
WD_ANSW_CNT[0]
WD_QUESTION[2] 00
WD_QUESTION[1] 01
WD_QUESTION[0] Reference-Answer-X[7]
10 X = 3, 2,1, 0
WD_QUESTION[3] 11
WD_ANSW_CNT[0]
Feedback settings are controllable through the bits WD_QA_FDBK[1:0] Calculated Reference-Answer-X byte
(Default value is 2'b00; the selected signals are in red)
• WD_ANSW_CNT[1:0] = 2b‘10:
1. The watchdog calculates the reference Answer-2.
2. A write access occurs. The MCU writes the Answer-2 byte in WD_ANSWER[7:0].
3. The watchdog compares the reference Answer-2 with the Answer-2 byte in WD_ANSWER[7:0]..
4. The watchdog decrements the WD_ANSW_CNT[1:0] bits to 2b‘01 and sets the WD_ANSW_ERR status
bit to 1 if the Answer-2 byte was incorrect.
• WD_ANSW_CNT[1:0] = 2b‘01:
1. The watchdog calculates the reference Answer-1.
2. A write access occurs. The MCU writes the Answer-1 byte in WD_ANSWER[7:0].
3. The watchdog compares the reference Answer-1 with the Answer-1 byte in WD_ANSWER[7:0]..
4. The watchdog decrements the WD_ANSW_CNT[1:0] bits to 2b‘00 and sets the WD_ANSW_ERR status
bit to 1 if the Answer-1 byte was incorrect.
• WD_ANSW_CNT[1:0] = 2b‘00:
1. The watchdog calculates the reference Answer-0.
2. A write access occurs. The MCU writes the Answer-0 byte in WD_ANSWER[7:0].
3. The watchdog compares the reference Answer-0 with the Answer-0 byte in WD_ANSWER[7:0].
4. The watchdog sets the WD_ANSW_ERR status bit to 1 if the Answer-0 byte was incorrect.
5. The watchdog starts a new watchdog sequence and sets the WD_ANSW_CNT[1:0] to 2‘b11’.
The MCU needs to clear the bit by writing a '1' to the WD_ANSW_ERR bit.
Table 8-9. Set of Questions and Corresponding Answer-Bytes Using the Default Setting of WD_QA_CFG
Register
ANSWER-BYTES (EACH BYTE TO BE WRITTEN INTO WD_ANSWER[7:0])
WD QUESTION
ANSWER-3 ANSWER-2 ANSWER-1 ANSWER-0
WD_ANSW_CNT [1:0] = WD_ANSW_CNT [1:0] = WD_ANSW_CNT [1:0] = WD_ANSW_CNT [1:0] =
WD_QUESTION[3:0]
2’b11 2’b10 2’b01 2’b00
0x0 FF 0F F0 00
0x1 B0 40 BF 4F
0x2 E9 19 E6 16
0x3 A6 56 A9 59
0x4 75 85 7A 8A
0x5 3A CA 35 C5
0x6 63 93 6C 9C
0x7 2C DC 23 D3
0x8 D2 22 DD 2D
0x9 9D 6D 92 62
0xA C4 34 CB 3B
0xB 8B 7B 84 74
0xC 58 A8 57 A7
0xD 17 E7 18 E8
0xE 4E BE 41 B1
0xF 01 F1 0E FE
1. The WD_SEQ_ERR and WD_BAD_EVENT status bits are set if Window-1 time-interval elapses before
watchdog has received Answer-3, Answer-2 and Answer-1.
2. The WD_ANSW_EARLY and WD_BAD_EVENT status bits are set if watchdog receives all four answers
in Window-1.
3. The WD_FAIL_CNT[2:0] counter increments by one at the end of the watchdog-sequence.
4. The question-counter does not change, and hence the watchdog does not generate a new question.
• A bad event occurs when one or more of the answer-bytes are not correct in value but in correct timing. After
such a bad event, following events occur:
1. The WD_ANSW_ERR and WD_BAD_EVENT status bits are set as soon as the watchdog detects an
incorrect answer-byte.
2. The WD_FAIL_CNT[2:0] counter increments by one at the end of the watchdog-sequence.
3. The question-counter does not change, and hence the watchdog does not generate a new question.
• A bad event occurs when one or more of the answer-bytes are not correct in value and not in correct timing.
After such a bad event, following events occur:
1. The WD_ANSW_ERR and WD_BAD_EVENT status bits are set as soon as the watchdog detects an
incorrect answer-byte.
2. The WD_SEQ_ERR and WD_BAD_EVENT status bits are set if Window-1 time-interval elapses before
watchdog has received Answer-3, Answer-2 and Answer-1.
3. The WD_ANSW_EARLY and WD_BAD_EVENT status bits are set if watchdog receives all four answer-
bytes in Window-1.
4. The WD_FAIL_CNT[2:0] counter increments by one at the end of the watchdog-sequence.
5. The question-counter does not change, and hence the watchdog does not generate a new question.
• A time-out event occurs when the device receives less than 4 answer-bytes before Window-2 time-interval
elapses. After a time-out event occurs, following events occur:
1. WD_SEQ_ERR and WD_BAD_EVENT status bits are set if Window-1 time-interval elapses before
watchdog has received Answer-3, Answer-2 and Answer-1.
2. The WD_TIMEOUT and WD_BAD_EVENT status bits are set at the end of the watchdog-sequence.
3. The WD_FAIL_CNT[2:0] counter increments by one at the end of the watchdog-sequence.
4. The question-counter does not change, and hence the watchdog does not generate a new question.
The status bit WD_BAD_EVENT is read-only. The watchdog clears the WD_BAD_EVENT status bit at the end of
the watchdog-sequence.
The status bits WD_SEQ_ERR, WD_ANSW_EARLY, and WD_TIMEOUT are latched until the MCU writes a ‘1’
to these bits. If one or more of these status bits are set, the watchdog can still detect a good event in the next
watchdog-sequence. These status bits are read-only. The watchdog clears the WD_BAD_EVENT status bit at
the end of the watchdog-sequence.
Figure 8-24 shows the flow-chart of the watchdog in Q&A mode.
WD_RETURN_
YES
LONGWIN=1?
NO
NO SUPPLY
ANSWER-3
Device sets WD_RST_EN=1 per default - Device sets WD_ANSW_CNT[2:0]=2'b11
Device sets WD_EN=1 per default. - If FIRST_WD_OK=0, device forces
Wake-up ENABLE_DRV=0, else device does not
NO
request? change ENABLE_DRV bit
NO NO - MCU sends ANSWER-3
- Device sets
YES WD_SEQ_ERR
error-ag
- Device sets Device has
WINDOW-2 YES WINDOW-1 NO received
WD_BAD
,me-interval me-interval ANSWER-3 ?
_EVENT
elapsed? elapsed?
error-ag
RESTART YES
from all Reset-Extension
NO
states except )me-interval YES - Device sets
NO SUPPLY
elapsed? WD_ANSW_ERR error-ag NO
ANSWER-3
- Device sets correct?
YES WD_BAD_EVENT error-ag
YES
YES
- Device sets
WD_ANSW_ERR error-ag NO
ANSWER-2
- Device sets correct?
NO WD_PWRHOLD=0?
WD_BAD_EVENT error-ag
YES
YES
ANSWER-1
- Device sets WD_ANSW_CNT[2:0]=2'b01
- If FIRST_WD_OK=0, device forces
NO WD_EN=0? ENABLE_DRV=0, else device does not
NO change ENABLE_DRV bit
NO - MCU sends ANSWER-1
YES
YES - Device sets
WD_SEQ_ERR
error-"ag
- Device sets Device has
NORMAL – NO Watchdog WINDOW-2 YES WINDOW-1 NO received
- MCU reset inac ve WD_EN=0? NO *me-interval WD_BAD me-interval
_EVENT ANSWER-1?
- MCU can set elapsed? elapsed?
error-#ag
ENABLE_DRV=1 if no other YES
error- ags are set
- Device released nINT pin if YES - Device sets
no other error- ag set WD_ANSW_ERR error-$ag
- -Device sets NO ANSWER-1
WD_BAD_EVENT error-%ag correct?
YES
Device generates 1st
Device has
QUESTION
received 4 YES ANSWER-0
- Device locks all Watchdog
answers ?
con0gura1on register bits, - Device sets WD_ANSW_CNT[2:0]=2'b00
except - If FIRST_WD_OK=0, device forces
WD_RETURN_LONGWIN bit ENABLE_DRV=0, else device does not
NO NO change ENABLE_DRV bit
- MCU sends ANSWER-0
Device sets
WD_TIMEOUT
LONG-WINDOW error-ag Device has
NO WINDOW-2
me-interval - Device sets YES NO received
elapsed?
Device me-interval
WD_BAD_EVENT ANSWER-0?
Increments elapsed?
error-ag
WD_FAIL_CNT YES
YES [3:0]
Table 8-10. Correct and Incorrect WD Q&A Sequence Run Scenarios (continued)
NUMBER OF WD ANSWERS WD STATUS BITS IN WDT_STATUS REGISTER
RESPONSE RESPONSE ACTION COMMENTS
ANSW_ERR ANSW_EARLY SEQ_ERR TIME_OUT
WINDOW 1 WINDOW 2
3 CORRECT
0 answers 0b 0b 0b 1b
answers -New WD cycle starts after the Less than 4 CORRECT ANSW
end of RESPONSE WINDOW 2 in RESPONSE WINDOW 1
2 CORRECT
0 answers -Increment WD failure counter and more than 0 ANSWER
answers
-New WD cycle starts with the 0b 0b 1b 1b in RESPONSE WINDOW 2
1 CORRECT same WD Question (WD_ANSW_CNT[1:0] < 3)
0 answers
answers
-New WD cycle starts after the
4th WD answer
3 CORRECT
1 CORRECT answer -Decrement WD failure counter 0b 0b 0b 0b CORRECT SEQUENCE
answers
-New WD cycle starts with a new
WD question
-New WD cycle starts after the
4th WD answer
3 CORRECT 1 INCORRECT
-Increment WD failure counter 1b 0b 0b 0b WD_ANSW_CNT[1:0] = 3
answers answers
-New WD cycle starts with the
same WD question
-New WD cycle starts after the
end of RESPONSE WINDOW 2
3 INCORRECT
0 answers -Increment WD failure counter 1b 0b 0b 1b WD_ANSW_CNT[1:0] < 3
answers
-New WD cycle starts with the
same WD question
-New WD cycle starts after the
4th WD answer
3 INCORRECT
1 CORRECT answer -Increment WD failure counter 1b 0b 0b 0b WD_ANSW_CNT[1:0] = 3
answers
-New WD cycle starts with the
same WD question
-New WD cycle starts after the
4th WD answer
3 INCORRECT 1 INCORRECT
-Increment WD failure counter 1b 0b 0b 0b WD_ANSW_CNT[1:0] = 3
answers answer
-New WD cycle starts with the
same WD question
-New WD cycle starts after the
4th WD answer
4 CORRECT
Not applicable -Increment WD failure counter 0b 1b 0b 0b
answers
-New WD cycle starts with the
same WD question
3 CORRECT
answers + 1 Not applicable 4 CORRECT or INCORRECT
INCORRECT answer ANSWER in RESPONSE
-New WD cycle starts after the WINDOW 1
2 CORRECT
4th WD answer
answers +
Not applicable -Increment WD failure counter 1b 1b 0b 0b
2 INCORRECT
-New WD cycle starts with the
answers
same WD question
1 CORRECT answer
+ 3 INCORRECT Not applicable
answers
The MCU can configure each ESM as long as its related start bit is cleared to 0 (bit ESM_MCU_START or
ESM_SOC_START). As soon as the MCU sets a start bit, the device sets a write-protection on the configuration
registers of the related ESM except the related start bits ESM_MCU_START and ESM_SOC_START.
YES
ESM_x pin
NO
level
- Device clears ESM_x_START=0 = 0?
- Device resets the ESM_x_DELAY1 YES
and ESM_x_DELAY2 timers
ESM_x-INTERRUPT
Note: the procedures ^Check - If ESM_x_PIN_MASK=0, device sets
ESM_x_START=1" and ^ESM_x Level ESM_x_PIN_INT interrupt bit And pulls nINT
- Device releases
D} WŒ} µŒ ^ Œµv ]v ‰ Œ oo o. pin low
nINT pin if all
Reset-Extension - Device starts ESM_x_DELAY1 timer, or
NO time-interval
If ESM_x_START=0, the device stops interrupt bits are
continues to run this timer if already started
cleared
elapsed? the ^ESM_x Level Mode - Device does not change ENABLE_DRV bit
- ESM resets
WŒ} µŒ ^ - Device does not change the level of the
ESM_x_DELAY1 and
MCU/SoC reset pins
ESM_x_DELAY2
Check ESM_x_START=1 timers
YES
- Device stops ESM_x
Level Mode Procedure
NO ESM_x_START YES ESM_x pin
- Device resets the
=1? level =1 &
ESM_x_DELAY1 and YES
ESM_x_PIN_INT=0
ESM_x_DELAY2 timers
?
NO
ESM_x-CONFIGURE
- Device releases MCU/SoC reset pin ESM_x_DELAY1
NO
- For ESM_MCU: Device forces ENABLE_DRV = 0 time-interval
- For ESM_SoC: Device forces ENABLE_DRV = 0 if ESM_SoC_ENDRV = 1 elapsed?
- MCU clears all interrupt bits
- Device releases nINT pin if no other interrupt bits are set
- ESM_x configuration registers unlocked YES
- MCU either clears ESM_x_EN, or
1) MCU configures ESM_x in Level-Mode (bit ESM_x_MODE)
2) MCU configures ESM_x_DELAY1, ESM_x_DELAY2 and ESM_x_ENDRV Configuration bit
3) MCU sets ESM_x_START YES Device forces
ESM_x_ENDRV =1? ENABLE_DRV= 0
NO
ESM_x_DELAY2 YES
set to 0?
NO ESM_x_EN
=0? NO
ESM_x Level-Mode NO
NO ESM_x_EN YES
=0? Error-Handling
Procedure ESM_x_DELAY2
NO
time-interval
elapsed?
YES
ESM_x-RESET
- ESM_x_RST trigger send to to FSM
- If ESM_x_RST_MASK=0, device sets
ESM_x_RST_INT interrupt bit and
pulls nINT pin low
nRSTOUT Pin
MCU sets
ESM_MCU_START
ESM_MCU_START x 0 1
ESM_MCU Input
Pin
tdegl_ESMx 15 s
tdegl_ESMx 15 s
Deglitched ESM_MCU
Input Signal
tdegl_ESMx 15 s
ESM_MCU_DELAY1 timer reset after
MCU clears ESM_MCU_PIN_INT
ESM_MCU_DELAY1
MCU clears
ESM_MCU_PIN_INT &
Deglitched ESM_MCU Input Signal = high
ESM_MCU_PIN_INT
0 1 0
(ESM_MCU_PIN_MASK=0)
ESM_MCU_FAIL_INT 0
(ESM_MCU_FAIL_MASK=0)
ESM_MCU_RST_INT 0
(ESM_MCU_RST_MASK=0)
ENABLE_DRV x 0 1
0 t
Case Number 1:
MCU initiated a fault-injection, and MCU clears the ESM_MCU_PIN_INT interrupt bit before elapse of ESM_MCU_DELAY1 time-interval
Figure 8-26. Example Waveform for ESM in Level Mode - Case Number 1: ESM_MCU Signal Recovers Before Elapse of Delay-1 time-interval
nRSTOUT Pin
MCU sets
ESM_MCU_START
ESM_MCU_START x 0 1
tESM_DEGLITCH 15 s
tLOW_ERROR < (ESM_MCU_DELAY1 +
ESM_MCU_DELAY2)
tdegl_ESMx 15 s
tdegl_ESMx 15 s
Deglitched ESM_MCU
Input Signal
tdegl_ESMx 15 s
ESM_MCU_DELAY1 and
ESM_MCU_DELAY2 timers reset after
MCU clears ESM_MCU_PIN_INT and
ESM_MCU_FAIL_INT
ESM_MCU_DELAY1 ESM_MCU_DELAY2
MCU clears
ESM_MCU_PIN_INT &
Deglitched ESM_MCU Input Signal = high
ESM_MCU_PIN_INT
0 1 0
(ESM_MCU_PIN_MASK=0)
MCU clears
ESM_MCU_FAIL_INT
ESM_MCU_FAIL_INT
(ESM_MCU_FAIL_MASK=0) 0 1 0
ESM_MCU_RST_INT 0
(ESM_MCU_RST_MASK=0)
ENABLE_DRV x 0 1 0 1
0 t
Figure 8-27. Example Waveform for ESM in Level Mode – Case Number 2: Delay-2 Not Set To 0 and ESM_MCU_ENDRV=1, ESM_MCU Signal
Recovers Before Elapse of Delay-2 Time-Interval
ESM_MCU_START x 0 1 0 1
tdegl_ESMx 15 s
tLOW_ERROR
ESM_MCU Input
Pin
tdegl_ESMx 15 s tdegl_ESMx 15 s
Deglitched ESM_MCU
Input Signal
tdegl_ESMx 15 s
ESM_MCU_DELAY1 timer
ESM_MCU_DELAY1 reset when ESM resets the
MCU MCU clears
ESM_MCU_PIN_INT &
Deglitched ESM_MCU Input Signal = high
ESM_MCU_PIN_INT
(ESM_MCU_PIN_MASK=0) 0 1 0
ESM_MCU_FAIL_INT
(ESM_MCU_FAIL_MASK=0) 0
MCU clears
ESM_MCU_RST_INT
ESM_MCU_RST_INT
0 1
(ESM_MCU_RST_MASK=0)
ENABLE_DRV x 0 1 0 1
0 t
Figure 8-28. Example Waveform for ESM in Level Mode – Case Number 3a: Delay-2 Set To 0 and ESM_MCU_ENDRV=1, ESM_MCU Input Signal
Recovers Too Late and MCU-Reset Occurs
nRSTOUT Pin
MCU sets MCU sets ESM_MCU_START
ESM_MCU_START after all interrupt bits are cleared
ESM_MCU_START x 0 1 0 1
tdegl_ESMx 15 s
tLOW_ERROR
tdegl_ESMx 15 s tdegl_ESMx 15 s
Internally Deglitched
ESM_MCU Input Signal
ESM_MCU_PIN_INT
0 1 0
(ESM_MCU_PIN__MASK=0)
MCU clears
ESM_MCU_FAIL_INT
ESM_MCU_FAIL_INT
0 1 0
(ESM_MCU_FAIL_MASK=0)
MCU clears
ESM_MCU_RST_INT
ESM_MCU_RST_INT
0 1 0
(ESM_MCU_RST_MASK=0)
ENABLE_DRV x 0 1 0 1
MCU sets ENABLE_DRV (only MCU sets ENABLE_DRV (only
possible when possible when
ESM_MCU_START=1) ESM_MCU_START=1)
0 t
Figure 8-29. Example Waveform for ESM in Level Mode – Case Number 3b: Delay-2 Not Set to 0 and ESM_MCU_ENDRV=1, ESM_MCU Input
Signal Recovers Too Late and MCU-Reset Occurs
nRSTOUT Pin
MCU sets MCU sets ESM_MCU_START
ESM_MCU_START after all interrupt bits are cleared
ESM_MCU_START x 0 1 0 1
tdegl_ESMx 15 s
tLOW_ERROR
tdegl_ESMx 15 s tdegl_ESMx 15 s
Internally Deglitched
ESM_MCU Input Signal
ESM_MCU_PIN_INT
0 1 0
(ESM_MCU_PIN__MASK=0)
ENABLE_DRV x 0 1 0 1
MCU sets ENABLE_DRV (only MCU sets ENABLE_DRV (only
possible when possible when
ESM_MCU_START=1) ESM_MCU_START=1)
0 t
Figure 8-30. Example Waveform for ESM in Level Mode – Case Number 3c: Delay-2 Not Set to 0 and ESM_MCU_ENDRV=1, MCU Fails to Clear
ESM_MCU_PIN_INT Before Elapse of the ESM_MCU_DELAY2
ESM_x_HMIN[7:0] < 0.5 × (ESM_x_HMAX[7:0] + ESM_x_HMIN[7:0]) × 0.95 × (1 - MCU/SoC clock tolerance) (21)
ESM_x_HMAX[7:0] > 0.5 × (ESM_x_HMAX[7:0] + ESM_x_HMIN[7:0]) × 1.05 × (1 + MCU/SoC clock tolerance) (22)
ESM_x_LMIN[7:0] < 0.5 × (ESM_x_LMAX[7:0] + ESM_x_LMIN[7:0]) × 0.95 × (1 - MCU/SoC clock tolerance) (23)
ESM_x_LMAX[7:0] > 0.5 × (ESM_x_LMAX[7:0] + ESM_x_LMIN[7:0]) × 1.05 × (1 + MCU/SoC clock tolerance) (24)
NO NO
YES YES
ESM_x PWM Error- ESM_x PWM Error-
Handling Procedure Handling Procedure
Reset-Extension Time
nRSTOUT /
nRSTOUT_SoC Pin
x 0 1
ESM_x_START
tdegl_ESMx 15 s
tHIGH_MAX_TH = tHIGH_MAX_TH =
(ESM_x_HMAX [7:0] + 1) × 15 s (ESM_x_HMAX [7:0] + 1) × 15 s
tHIGH_MIN_TH = tHIGH_MIN_TH =
(ESM_x_HMIN [7:0] + 1) (ESM_x_HMIN [7:0] + 1)
× 15 s × 15 s
tLOW_MIN_TH = tLOW_MIN_TH =
(ESM_x_LMIN[7:0] + 1) × 15 s (ESM_x_LMIN[7:0] + 1) × 15 s
Internal ESM_x
bad event Trigger
Internal ESM_x
good event Trigger
ESM_x_ERR_CNT[4:0] x 00000
ESM_x_PIN_INT
(ESM_x_PIN_MASK=0) 0
nINT Pin
ESM_x_FAIL_INT
0
(ESM_x_FAIL_MASK=0)
ESM_x_RST_INT
0
(ESM_x_RST_MASK=0)
ENABLE_DRV
x 0 1
Case Number 1:
PWM signal has a low level at the moment the MCU sets bit ESM_x_start, and the PMIC receives a PWM Error Signal with Correct Timing afterwards
Figure 8-32. Example Waveform for ESM in PWM Mode – Case Number 1: ESM Starts with Low-Level at Deglitched Input signal, and Receives
Correct PWM Signal Afterwards. (The _x stand for _MCU or _SoC)
Reset-Extension Time
nRSTOUT /
nRSTOUT_SoC Pin
x 0 1
ESM_x_START
tdegl_ESMx 15 s
tHIGH_MIN_TH = tHIGH_MIN_TH =
(ESM_x_HMIN [7:0] + 1) (ESM_x_HMIN [7:0] + 1)
× 15 s × 15 s
tLOW_MAX_TH = tLOW_MAX_TH =
(ESM_x_LMAX[7:0] + 1) × 15 s (ESM_x_LMAX[7:0] + 1) × 15 s
tLOW_MIN_TH = tLOW_MIN_TH =
(ESM_x_LMIN[7:0] + 1) × 15 s (ESM_x_LMIN[7:0] + 1) × 15 s
High-Pulse Timer
Reset and Started High-Pulse Timer High-Pulse Timer High-Pulse Timer
Stopped, Stopped, Stopped,
Deglitched ESM_x Low-Pulse Timer Low-Pulse Timer Low-Pulse Timer
Input Signal Reset and Started Low-Pulse Timer Reset and Started Low-Pulse Timer Reset and Started
Stopped, Stopped,
High-Pulse Timer High-Pulse Timer
no bad event trigger as Reset and Started Reset and Started
long as falling edge on tPWM_LOW tPWM_HIGH tPWM_LOW tPWM_HIGH
ESM_x signal comes
before elapse of
tHIGH_MAX_TH 1 ESM_x good-event 1 ESM_x good-event
Internal ESM_x
bad event Trigger
Internal ESM_x
good event Trigger
ESM_x_ERR_CNT[4:0] x 00000
ESM_x_PIN_INT
0
(ESM_x_PIN_MASK=0)
nINT Pin
ESM_x_FAIL_INT
(ESM_x_FAIL_MASK=0) 0
ESM_x_RST_INT
0
(ESM_x_RST_MASK=0)
ENABLE_DRV x 0 1
Case Number 2:
PWM signal has a high level at the moment the MCU sets bit ESM_x_start, and the PMIC receives a PWM Error Signal with Correct Timing afterwards
Figure 8-33. Example Waveform for ESM in PWM Mode – Case Number 2: ESM Starts with High-Level at Deglitched Input Signal, and Receives
Correct PWM Signal Afterwards (The _x stand for _MCU or _SoC)
ESM_x_START x 0 1
tdegl_ESMx 15 s
tHIGH_MAX_TH = tHIGH_MAX_TH = tHIGH_MAX_TH =
(ESM_x_HMAX [7:0] + 1) × (ESM_x_HMAX [7:0] + 1) × (ESM_x_HMAX [7:0] + 1) ×
15 s 15 s 15 s
Deglitched ESM_x
Input Signal
ESM_x_RST_INT 0
ENABLE_DRV x 0 1 0 1
MCU sets ENABLE_DRV (ESM_MCU, only possible if Note: PMIC clears ENABLE_DRV MCU sets ENABLE_DRV (ESM_MCU,
ESM_MCU_START=1. ESM_SOC possible if ESM_SOC_ENDRV=0 only possible if ESM_MCU_START=1.
only when configuration bit ESM_SOC possible if
or if ESM_SOC_ENDRV=1 and ESM_SOC_START=1)
ESM_x_ENDRV=1 ESM_SOC_ENDRV=0 or if
ESM_SOC_ENDRV=1 and
ESM_SOC_START=1)
Figure 8-34. Example Waveform for ESM in PWM Mode – Case Number 3: ESM Starts with Low-Level at Deglitched Input Signal, but Receives
Too Late a Correct PWM Signal Afterwards (The _x stand for _MCU or _SoC)
Reset-Extension
Reset-Extension Time
Time
nRSTOUT /
nRSTOUT_SoC Pin MCU sets
ESM_x_START after all
MCU sets ESM_x_START interrupt bits are cleared
ESM_x_START x 0 1 0 1
tdegl_ESMx 15 s tdegl_ESMx 15 s
ESM_x_ERR_
CNT[4:0] x 00000 00010 00100 00110 00101 00000
ESM_x_ERR_CNT_ ESM_x_DELAY1 and ESM_x_DELAY2 timers
TH[3:0] = 0011 ESM_x_DELAY1 ESM_x_DELAY2 reset when ESM_x resets the MCU or SoC
MCU clears ESM_x_PIN_INT &
ESM_x_ERR_CNT[4:0] ”
ESM_x_ERR_CNT_TH[3:0]
ESM_x_PIN_INT
0 1 0
(ESM_x_PIN_MASK
=0)
&
nINT Pin
MCU clears
ESM_x_FAIL_INT ESM_x_FAIL_INT
(ESM_x_FAIL_MASK 0 1 0
=0) MCU clears
ESM_x_RST_INT
ESM_x_RST_INT
(ESM_x_RST_MASK 0 1 0
=0)
ENABLE_DRV x 0 1 0 1
Figure 8-35. Example Waveform for ESM in PWM Mode – Case Number 4: ESM Starts with Low-Level at Deglitched Input Signal and Receives a
Correct PWM Signal. Afterwards the ESM Detects Bad Events, and the PWM Signal Recovers Too Late which Leads to an ESM_x Reset Trigger
to the PFSM (The _x stand for _MCU or _SoC)
device proceeds to power up the device and reach the default mission state. More details
regarding the LP_WAKE function can be found in Section 8.4.1.2.4.5.
INIT The device is powered by a valid supply on the system power rail (VCCA ≥ VCCA_UV). If
the device was previously in LP_STANDBY state, it has received an external wake-up signal
at the LP_WKUP1/2 pins, the RTC alarm or timer wake-up signal, or an On Request from
the nPWRON/ENABLE pin. Device digital and monitor circuits are powered up. The PMIC
reads its internal NVM memory in this state and configures default values to registers, IO
configuration and FSM accordingly.
BOOT BIST The device is running the built-in self-test routine that includes
Note
The ABIST on the voltage monitor circuits for the BUCK regulators, the LDO
regulators is performed after the start-up of these regulators. The ABIST on the
VMON1 and VMON2 is performed after these voltage monitors are enabled. See
Voltage Monitors.
An option is available to shorten the device power up time from the NO_SUPPLY state by
setting the NVM bit FAST_BOOT_BIST = '1' to skip the LBIST. Software can also set the
FAST_BIST = '1' to skip LBIST after the device wakes up from the LP STANDBY state. When
the device arrives at this state from the SAFE_RECOVERY state, LBIST is automatically
skipped if it has not previously failed. If LBIST failed, but passed after multiple re-tries before
exceeding the recovery counter limit, the device powers up normally. The following NVM bits
are pre-configured options to enable or disable parts of the CRC tests if further sequence time
reduction is required (please refer to the user's guide of the orderable part number):
• REG_CRC_EN = '0': disables the register map and SRAM CRC check
Note
Note: the BIST tests are executed as parallel processes, and the longest process
determines the total BIST duration
Note
the ABIST on the voltage monitor circuits of the BUCKx regulator, LDOn regulators,
VMON1 and VMON2 is performed individually after these voltage monitors are
enabled. See Voltage Monitors)
RUNTIME BIST A request was received from the MCU to exercise a run-time built-in self-test
(RUNTIME_BIST) on the device. No rails are modified and all external signals, including
all I2C or SPI interface communications, are ignored during the RUNTIME_ BIST. During the
RUNTIME_BIST, the device performs the same self-test routines as listed for the BOOT_BIST
state. Furthermore, during RUNTIME_BIST the device performs the ABIST routine on all
enabled voltage monitor circuits.
If the device passed BIST, it resumes the previous operation. If the device failed BIST, it shuts
down all of the regulator outputs and proceed to the SAFE RECOVERY state. In order to
avoid a register CRC error, all register writes must be avoided after the request for the BIST
operation until the device pulls the nINT pin low to indicate the completion of BIST. The results
of the RUNTIME_BIST are indicated by the BIST_PASS_INT or the BIST_FAIL_INT interrupt
bits.
Note
After completion of the RUNTIME_BIST, the LDOx_UV detections are masked for
a time-interval equal to the configured LDOx ramp-up time (25 mV/μs or 3 mV/μs
according configuration bit LDOx_SLOW_RAMP). The actual LDOx output voltages
are not affected by this masking of the LDOx_UV detections.
SAFE The device meets the qualified error condition for immediate or ordered shutdown request.
RECOVERY If the error is recovered within the recovery time interval or meets the restart condition, the
device increments the recovery counter, and returns to INIT state if the recovery counter value
does not exceed the threshold value. Until a supply power cycle occurs, the device stays in
the SAFE RECOVERY state if one of the following conditions occur:
• the recovery counter exceeds the threshold value
• the die temperature cannot be reduced to less than TWARN level
• VCCA stays above OVP threshold
When multiple system conditions occur simultaneously that demand power state arbitration, the device goes to
the higher priority state according to the following priority order:
1. NO SUPPLY
2. BACKUP
3. SAFE_RECOVERY
4. LP_STANDBY
5. MISSION STATES
Figure 8-36 shows the power transition states of the FSM engine.
NO
SUPPLY
LDOVRTC UVLO Condion or
Shelf Mode enabled
LP_STANDBY_SEL = 1 and
Valid WAKE Request1
no valid WAKE request1
INIT
Thermal Shutdown or
VCCA OVP
Error Condions
SAFE BOOT
(recovery cnt +1)
RECOVERY BIST
Orderly shutdown
Condion
(recovery cnt +1) BOOT BIST success
Severe or Moderate
PFSM Errors
(recovery cnt +1)
Mission States
RUNTIME
BIST
1
A valid WAKE request consist of:
nPWRON/ENABLE on request detecon if the device arrived the LP_STANDBY state through
the long key-press of the nPWRON pin or by disabling the ENABLE pin, or
RTC Alarm, RTC Timer, LP_WKUP1 or LP_WKUP2 detecon if the device arrived the
LP_STANDBY state through wring to a TRIGGER_I2C_0 bit.
Figure 8-36. State Diagram for Device Power States
Below are the NVM pre-configured register bits in the RTC domain:
• GPIO3_CONF and GPIO4_CONF registers, except the GPIOn_DEGLITCH_EN bits
• GPIO3_RISE_MASK, GPIO3_FALL_MASK, GPIO4_RISE_MASK, and GPIO4_FALL_MASK bits
• NPWRON_CONF register except ENALBE_DEGLITCH_EN and NRSTOUT_OD bits
• FSD_MASK, ENABLE_MASK, NPWRON, START_MASK, and NPWRON_LONG_MASK bits
• STARTUP_DEST, FAST_BIST, LP_STANDBY_SEL, XTAL_SEL, and XTAL_EN bits
• PFSM_DELAYn, and RTC_SPARE_n bits
Below are the register bits without NVM pre-configuration in the RTC domain:
• FIRST_STARTUP_DONE bit
• SCRATCH_PAD_n bits
• All of RTC control and configuration registers
8.4.1.2 Pre-Configurable Mission States
When the device arrives at a mission state, all rail sequencing is controlled by the pre-configurable FSM engine
(PFSM) through the configuration memory. The configuration memory allows configurations of the triggers and
the operation states that together form the configurable sub state machine within the scope of mission states.
This sub state machine can be used to control and sequence the different voltage outputs as well as any
GPIO outputs that can be used as enable for external rails. When the device is in a mission state, it has the
capacity to supply the processor and other platform modules depending on the power rail configuration. The
definitions and transition triggers of the mission states are configurable through the NVM configuration. Unlike
the user registers, the PFSM definition stored in the NVM cannot be modified during normal operation. When the
PMIC determines that a transition to another operation state is necessary, it reads the configuration memory to
determine what sequencing is needed for the state transition. Furthermore, the PFSM has four storage registers,
further referred to as PFSM storage registers (R0-3).
Table 8-15 shows how the trigger signals for each state transition can come from a variety of interface or GPIO
inputs, or potential error sources. Figure 8-37 shows how the device processes all of the possible error sources
inside the PFSM engine, a hierarchical mask system is applied to filter out the common errors that can be
handled by interrupt only, and categorize the other error sources as Severe Global Error, Moderate Global Error,
and so forth. The filtered and categorized triggers are sent into the PFSM engine, that then determines the entry
and exit condition for each configured mission state.
INTERRUPT
First level mask to filter out non-error interrupts vs. interrupts which require error handling
is given
VCCA BUCK1 BUCK2 BUCK3 BUCK4 BUCK5 LDO1 LDO2 LDO3 LDO4
Recovery Counter Limit to FSM
OR WD error to FSM
function Severe Global
Error
MCU Rail Group SoC Error Monitor to FSM
Mask Moderate Global
SoC Rail Group Error MCU Error Monitor to FSM
IMMEDIATE
SHUTDOWN trigger
Immediate Shutdown Trigger Mask
input to FSM
ORDERLY
SHUTDOWN trigger Orderly Shutdown Trigger Mask
input to FSM
MCU Power
MCU Power Error Trigger Mask
Error Signal
SoC Power
SoC Power Error Trigger Mask
Error Signal
Figure 8-39 shows an example of how the PFSM engine utilizes instructions to execute the configured device
state and sequence transitions of the mission state-machine. Table 8-12 provides the instruction set and usage
description of each instruction in the following sections. Section 8.4.1.2.2 describes how the instructions are
stored in the NVM memory.
Table 8-12. PFSM Instruction set
Command Opcode Command Command Description
Write the specified data, except the masked bits, to the specified
"0000" REG_WRITE_MASK_PAGE0_IMM
page 0 register address.
"0001" REG_WRITE_IMM Write the specified data to the specified register address.
Write the specified data, except the masked bits, to the specified
"0010" REG_WRITE_MASK_IMM
register address.
Write the target voltage of a specified regulator after a specified
"0011" REG_WRITE_VOUT_IMM
delay.
Write the operation mode of a specified regulator after a specified
"0100" REG_WRITE_VCTRL_IMM
delay.
Write the data from PFSM storage register (R0-3), except the
"0101" REG_WRITE_MASK_SREG
masked bits, to the specified register address.
Write PFSM storage register (R0-3) with data from a specified
"0110" SREG_READ_REG
address.
Execution is paused until the specified type of the condition is met
"0111" WAIT
or timed out.
"1000" DELAY_IMM Delay the execution by a specified time.
Delay the execution by a time value stored in the specified PFSM
"1001" DELAY_SREG
storage register (R0-3).
Set a trigger destination address for a given input signal or
"1010" TRIG_SET
condition.
"1011" TRIG_MASK Sets a trigger mask that determines which triggers are active.
"1100" END Mark the final instruction in a sequential task.
Write the specified data to the BIT_SEL location of the specified
"1101" REG_WRITE_BIT_PAGE0_IMM
page 0 register address.
Write the specified data to the SHIFT location of the specified
"1110" REG_WRITE_WIN_PAGE0_IMM
page 0 register address.
• SREG_READ_REG R2 0x15 — Read the content of address 0x15 and write the data to PFSM storage
register R2
• SREG_READ_REG ADDR=0x077 REG=R3 — Read the content of address 0x77 and write the data to
PFSM storage register R3
8.4.1.2.1.10 SREG_WRITE_IMM Command
Description: Write the specified data to the scratch register (R0-3)
Assembly command: SREG_WRITE_IMM [REG=]<PFSM Storage Register> [DATA=]<Data>
Data can be in any literal integer format (decimal, hex, and so forth).
PFSM Storage Register can be R0, R1, R2, or R3.
'REG=' and 'DATA=' are options. When included, the parameters can be in any order.
Examples:
• SREG_WRITE_IMM R2 0x15 — Write 0x15 to PFSM storage register R2
• SREG_WRITE_IMM ADDR=0x077 REG=R3 — Read 0x77 to PFSM storage register R3
8.4.1.2.1.11 WAIT Command
Description: Wait upon a condition of a given type. Execution is paused until the specified type of the condition is
met or timed out
Assembly command: WAIT [COND=]<Condition> [TYPE=]<Type> [TIMEOUT=]<Timeout>
[DEST=]<Destination>
Alternative assembly command: JUMP [DEST=]<Destination>
'COND=', 'TYPE=', 'TIMEOUT=', and 'DEST=' are options. When included, the parameters can be in any order.
Condition are listed in Table 8-13. Examples: GPIO1, BUCK1_PG, I2C_1
Type = LOW, HIGH, RISE, or FALL
Timeout = timeout value in ns, µs, ms, or s. If no unit is entered, this field must be an integer value between
0-63. Timeout value is be rounded to the nearest achievable time based on the current step size. Current
step size is based on the default NVM setting or a SET_DELAY value from a previous command in the same
sequence. Assembler reports an error if the step size is too large or too small to meet the delay.
Destination = Label to jump to if when timeout occurs. Destination must be after the WAIT statement in memory.
Memory space can be either '0' or '1'. '0' indicates the destination address is in the PFSM memory space. '1'
indicates the destination address is external and represents a FSM state ID.
When using the jump command, the PFSM performs an unconditional jump. The command is be compiled
as "WAIT COND=63 TYPE=LOW TIMEOUT=0 DEST=<Destination>". Condition 63 is a hardcoded 1, so
the condition is never satisfied and hence always times out. Therefore this command always jumps to the
destination.
Examples:
• WAIT GPIO4 RISE 1 s <Destination> 0 — Wait to execute the command at the specified SRAM address
when a rise edge is detected at GPIO4, or after 1 second
• WAIT COND=BUCK1_PG TYPE=HIGH TIMEOUT=500 µs DEST=<mcu2act_seq> — Wait to execute the
commands at <mcu2act_seq> address as soon as BUCK1 output is within power-good range, or after 500 µs
Table 8-13. WAIT Command Conditions
COND_ Condition Name COND_ Condition Name COND_ Condition Name COND_ Condition Name
SEL SEL SEL SEL
0 GPIO1(1) 16 LDO1_PG 32 I2C_0(2) 48 LP_STANDBY_SEL
1 GPIO2(1) 17 LDO2_PG 33 I2C_1(2) 49 N/A
2 GPIO3(1) 18 LDO3_PG 34 I2C_2(2) 50 N/A
(1) Conditions GPIO1..GPIO6 are only processed if the associated GPIOx pins are configured as GPIO
(2) Conditions I2C0..7 refer to register bits TRIGGER_I2C_0..7, and are located in register FSM_I2C_TRIGGER
(3) Conditions SREG0_0...7 refer to bits 9...7 in PFSM storage register R0
'DEST=', 'ID=', 'SEL='. 'TYPE=', 'IMM=', and 'EXT=' are options. When included, the parameters can be in any
order.
Destination is the label where this trigger starts executing.
Trig_ID is the ID of the hardware trigger module to be configured (value range 0-27). They must be defined in
numeric order based on the priority of the trigger.
Trig_Sel is the 'Trigger Name' from the Table 8-15. This 'Trigger Name' is the trigger signal to be associated with
the specified TRIG_ID.
Trig_type = LOW, HIGH, RISE, or FALL.
IMM can be either '0' or '1'. = '0' if the trigger is not activated until the END command of a given task; '1' if the
trigger is activated immediately and can abort a sequence.
REENTRANT can be either '0' or '1'. '1' permits a trigger to return to the current state, that allows a self-
branching trigger to execute the current sequence again.
Memory space can be either '0' or '1'. '0' indicates the destination address is in the PFSM memory space. '1'
indicates the destination address is external and represents a FSM state ID.
Examples:
• TRIG_SET seq1 20 GPIO_1 LOW 0 0 — Set trigger 20 to be GPIO_1=Low, not immediate. When triggered,
start executing at ‘seq1’ label.
• TRIG_SET DEST=seq2 ID=15 SEL=WD_ERROR TYPE=RISE IMM=0 EXT=0 — Set trigger 15 to be rising
WD_ERROR trigger, not immediate. When triggered, start executing at ‘seq2’ label.
8.4.1.2.1.15 TRIG_MASK Command
Description: Sets a trigger mask that determines which triggers are active. Setting a ‘0’ enables the trigger,
setting a ‘1’ disables (masks) the trigger.
Assembly command: TRIG_MASK <Mask value>
Mask Value = 28-bit mask in any literal integer format (decimal, hex, and so forth).
Examples:
• TRIG_MASK 0x5FF82F0 — Set the trigger mask to 0x5FF82F0
8.4.1.2.1.16 END Command
Table 8-14 shows the format of the END commands.
Table 8-14. END Command Format
Bit[3:0]
CMD
4 bits
pfsm_start:
TRIG_SET DEST=sequence_name1 ID=0 SEL=trigger_name TYPE=high/low/rise/fall IMM=0/1 EXT=0/1
TRIG_SET DEST=sequence_name2 ID=1 SEL=trigger_name TYPE=high/low/rise/fall IMM=0/1 EXT=0/1
TRIG_SET DEST=sequence_name3 ID=2 SEL=trigger_name TYPE=high/low/rise/fall IMM=0/1 EXT=0/1
TRIG_SET DEST=sequence_name4 ID=4 SEL=trigger_name TYPE=high/low/rise/fall IMM=0/1 EXT=0/1
TRIG_SET DEST=sequence_name5 ID=4 SEL=trigger_name TYPE=high/low/rise/fall IMM=0/1 EXT=0/1
««
TRIG_MASK 0xFFFFFF0
END
sequence_name1
REG_WRITE_MASK_IMM ADDR=register_addr DATA=data MASK=mask_setting These TRIG_SET instructions are used to
REG_WRITE_MASK_IMM ADDR=register_addr DATA=data MASK=mask_setting define the trigger types which initiates each
REG_WRITE_MASK_IMM ADDR=register_addr DATA=data MASK=mask_setting power state sequence. There are a total of
REG_WRITE_VCTRL_IMM REGULATOR=regulator VCTRL=ctrl_setting MASK=mask_setting DELAY=delay_time 28 TRIG_SET available for each PMIC. TYPE
REG_WRITE_VCTRL_IMM REGULATOR=regulator VCTRL=ctrl_setting MASK=mask_setting DELAY=delay_time parameter defines the type of trigger as:
DELAY_IMM delay_time
REG_WRITE_MASK_IMM ADDR=register_addr DATA=data MASK=mask_setting
REG_WRITE_VCTRL_IMM REGULATOR=regulator VCTRL=ctrl_setting MASK=mask_setting DELAY=delay_time
High: active high (level sensitive)
REG_WRITE_MASK_IMM ADDR=register_addr DATA=data MASK=mask_setting Low: active low (level sensitive)
TRIG_MASK 0xFC00EDF Rise: active high (edge sensitive)
END Fall: active low (edge sensitive)
««
sequence_name4
REG_WRITE_MASK_IMM ADDR=register_addr DATA=data MASK=mask_setting
DELAY_IMM delay_time
REG_WRITE_MASK_IMM ADDR=register_addr DATA=data MASK=mask_setting
REG_WRITE_VCTRL_IMM REGULATOR=regulator VCTRL=ctrl_setting MASK=mask_setting DELAY=delay_time
REG_WRITE_VCTRL_IMM REGULATOR=regulator VCTRL=ctrl_setting MASK=mask_setting DELAY=delay_time
DELAY_IMM delay_time
REG_WRITE_VCTRL_IMM REGULATOR=regulator VCTRL=ctrl_setting MASK=mask_setting DELAY=delay_time
REG_WRITE_MASK_IMM ADDR=register_addr DATA=data MASK=mask_setting
REG_WRITE_VCTRL_IMM REGULATOR=regulator VCTRL=ctrl_setting MASK=mask_setting DELAY=delay_time
REG_WRITE_MASK_IMM ADDR=register_addr DATA=data MASK=mask_setting
TRIG_MASK 0xFEF6EDC
END
As soon as the PMIC state reaches the mission states, it starts reading from the configuration memory until
it hits the first END command. Setting up the triggers (1-28) must be the first section of the configuration
memory, as well as the first set of trigger configurations. The trigger configurations are read and mapped to
an internal lookup table that contains the starting address associated with each trigger in the configuration
memory. If the trigger destination is an FFSM state then the address contains the fixed state value. After the
trigger configurations are read and mapped into the SRAM, these triggers control the execution flow of the state
transitions. The signal source of each trigger is listed under Table 8-15.
When a trigger or multiple triggers are activated, the PFSM execution engine looks up the starting address
associated with the highest priority unmasked trigger, and starts executing commands until it hits an END
command. The last commands before END statement is generally the TRIG_MASK command, which directs the
PFSM to a new set of unmasked trigger configurations, and the trigger with the highest priority in the new set
is serviced next. Trigger priority is determined by the Trigger ID associated with each trigger. The priority of the
trigger decreases as the associated trigger ID increases. As a result, the critical error triggers are usually located
at the lowest trigger IDs.
The TRIG_SET commands specify if a trigger is immediate or non-immediate. Immediate triggers are serviced
immediately, which involves branching from the current sequence of commands to reach a new target
destination. The non-immediate triggers are accumulated and serviced in the order of priority through the
execution of each given sequence until the END command in reached. Therefore, the trigger ID assignment for
each trigger can be arranged to produce the desired PFSM behavior.
The TRIG_MASK command determines which triggers are active at the end of each sequence, and is usually
placed just before the END instruction. The TRIG_MASK takes a 28-bit input to allow any combination of
triggers to be enabled with a single command. Through the definition of the active triggers after each sequence
execution the TRIG_MASK command can be conceptualized as establishing a power state.
The above sequence of waiting for triggers and executing the sequence associated with an activated trigger
is the normal operating condition of the PFSM execution engine when the PMIC is in the MISSION state. The
fixed device power FSM takes over control from the execution engine each time an event occurs that requires a
transition from the MISSION state of the PMIC to a fixed device state.
Table 8-15. PFSM Trigger Selections
Trigger Name Trigger Source
An error event causes one of the triggers defined in the FSM_TRIG_SEL_1/2 register to activate, and
IMMEDIATE_SHUTDOWN
the intended action for the activated trigger is to immediate shutdown the device.
MCU_POWER_ERROR Output failure detection from a regulator which is assigned to the MCU rail group (x_GRP_SEL = '01').
ORDERLY_SHUTDOWN An event which causes MODERATE_ERR_INT = '1'.
nPWRON long-press event when NPOWRON_SEL = '01', or ENABLE = '0' when NPOWERON_SEL =
FORCE_STANDBY
'00'.
SPMI_WD_BIST_DONE Completion of SPMI WatchDog BIST.
ESM_MCU_ERROR An event that causes ESM_MCU_RST_INT.
WD_ERROR An event that causes WD_RST_INT.
SOC_POWER_ERROR Output failure detection from a regulator which is assigned to the SOC rail group (x_GRP_SEL = '10').
ESM_SOC_ERROR An event that causes ESM_SOC_RST_INT
NSLEEP2 and NSLEEP1 = '11'. . More information regarding the NSLEEP1 and NSLEEP2 functions
A
can be found under Section 8.4.1.2.4.3.
WKUP1 A rising or falling edge detection on a GPIO pin that is configured as WKUP1 or LP_WKUP1.
SU_ACTIVE A valid On-Request detection when STARTUP_DEST = '11'.
NSLEEP2 and NSLEEP1 = '10'. . More information regarding the NSLEEP1 and NSLEEP2 functions
B
can be found under Section 8.4.1.2.4.3.
WKUP2 A rising or falling edge detection on a GPIO pin that is configured as WKUP2 or LP_WKUP2.
SU_MCU_ONLY A valid On-Request detection when STARTUP_DEST = '10'.
NSLEEP2 and NSLEEP1 = '01'. More information regarding the NSLEEP1 and NSLEEP2 functions
C
can be found under Section 8.4.1.2.4.3.
NSLEEP2 and NSLEEP1 = '00'. More information regarding the NSLEEP1 and NSLEEP2 functions
D
can be found under Section 8.4.1.2.4.3.
SU_STANDBY A valid On-Request detection when STARTUP_DEST = '00'.
SU_X A valid On-Request detection when STARTUP_DEST = '01'.
PFSM WAIT command condition timed out. More information regarding the WAIT command can be
WAIT_TIMEOUT
found under Section 8.4.1.2.1.11.
GPIO1 Input detection at GPIO1 pin. Only processed as trigger if GPIO1 is configured as GPIO.
GPIO2 Input detection at GPIO2 pin. Only processed as trigger if GPIO2 is configured as GPIO.
GPIO3 Input detection at GPIO3 pin. Only processed as trigger if GPIO3 is configured as GPIO.
GPIO4 Input detection at GPIO4 pin. Only processed as trigger if GPIO4 is configured as GPIO.
GPIO5 Input detection at GPIO5 pin. Only processed as trigger if GPIO5 is configured as GPIO.
GPIO6 Input detection at GPIO6 pin. Only processed as trigger if GPIO6 is configured as GPIO.
GPIO7 Input detection at GPIO7 pin. Only processed as trigger if GPIO7 is configured as GPIO.
GPIO8 Input detection at GPIO8 pin. Only processed as trigger if GPIO8 is configured as GPIO.
GPIO9 Input detection at GPIO9 pin. Only processed as trigger if GPIO9 is configured as GPIO.
GPIO10 Input detection at GPIO10 pin. Only processed as trigger if GPIO10 is configured as GPIO.
GPIO11 Input detection at GPIO11 pin. Only processed as trigger if GPIO11 is configured as GPIO.
I2C_0 Input detection of TRIGGER_I2C_0 bit (in register FSM_I2C_TRIGGERS)
I2C_1 Input detection of TRIGGER_I2C_1 bit (in register FSM_I2C_TRIGGERS)
I2C_2 Input detection of TRIGGER_I2C_2 bit (in register FSM_I2C_TRIGGERS)
I2C_3 Input detection of TRIGGER_I2C_3 bit (in register FSM_I2C_TRIGGERS)
I2C_4 Input detection of TRIGGER_I2C_4 bit (in register FSM_I2C_TRIGGERS)
I2C_5 Input detection of TRIGGER_I2C_5 bit (in register FSM_I2C_TRIGGERS)
I2C_6 Input detection of TRIGGER_I2C_6 bit (in register FSM_I2C_TRIGGERS)
OFF request
WKUP1 0W1 or
NSLEEP2&NSLEEP1
NSLEEP1
11W10
0W1
NSLEEP2 WKUP1 0W 1 or
1:0 MCU ONLY NSLEEP2&NSLEEP1
00W11
WKUP2 0W 1 or
Warm Reset triggered by NSLEEP2 NSLEEP2&NSLEEP1
ESM or WDOG error 1:0 00W10
DEEP SLEEP
/S2R
Each power state (light blue bubbles in Figure 8-39) defines the ON or OFF state and the sequencing timing of
the external regulators and GPIO outputs. This example defines four power states: STANDBY, ACTIVE, MCU
ONLY, and DEEP_SLEEP/S2R three power states: STANDBY, ACTIVE and RETENTION states. The priority
order of these states is as follows:
1. ACTIVE
2. MCU ONLY
3. DEEP SLEEP/S2R
4. STANDBY
The transitions between each power state is determined by the trigger signals source pre-selected from Table
8-15. These triggers are then placed in the order of priority through the trigger ID assignment of each trigger
source. The critical error triggers are placed first, some specified as immediate triggers that can interrupt an
on-going sequence. The non-error triggers, which are used to enable state transitions during normal device
operation, are then placed according to the priority order of the state the device is transitioning to. Table 8-16 list
the trigger signal sources, in the order of priority, used to define the power states and transitions of the example
mission state machine shown in Figure 8-39. This table also helps to determine which triggers must be masked
by the TRIG_MASK command upon arriving a pre-defined power state to produce the desired PFSM behavior.
Table 8-16. List of Trigger Used in Example Mission State Machine
Trigger ID Trigger Masked In Each User Defined Power State
Trigger Signal State Transitions DEEP
STANDBY ACTIVE MCU ONLY
SLEEP / S2R
0 IMMEDIATE_SHUTDOWN (1) From any state to SAFE
RECOVERY
1 MCU_POWER_ERROR (1) From any state to SAFE
RECOVERY
2 ORDERLY_SHUTDOWN (1) From any state to SAFE
RECOVERY
3 TRIGGER_FORCE_STANDBY From any state
to STANDBY or Masked
LP_STANDBY
4 WD_ERROR Perform warm reset of all
power rails and return to Masked Masked Masked
ACTIVE
5 ESM_MCU_ERROR Perform warm reset of all
power rails and return to Masked Masked Masked
ACTIVE
6 ESM_SOC_ERROR Perform warm reset of
power rails in SOC
Masked Masked Masked
domain and return to
ACTIVE
7 WD_ERROR Perform warm reset of all
power rails and return to Masked Masked Masked
MCU ONLY
8 ESM_MCU_ERROR Perform warm reset of all
power rails and return to Masked Masked Masked
MCU ONLY
9 SOC_POWER_ERROR ACTIVE to MCU ONLY Masked Masked Masked
10 TRIGGER _I2C_1 (self-cleared) Start RUNTIME_BIST Masked Masked
11 TRIGGER_I2C_2 (self-cleared) Enable I2C CRC
Masked Masked
Function
12 TRIGGER_SU_ACTIVE STANDBY to ACTIVE Masked Masked
13 TRIGGER_WKUP1 Any State to ACTIVE
14 TRIGGER_A (NSLEEP2&NSLEEP1 MCU ONLY or DEEP
Masked
= '11') SLEEP/S2R to ACTIVE
15 TRIGGER_SU_MCU_ONLY STANDBY to MCU ONLY Masked Masked
16 TRIGGER_WKUP2 STANDBY or DEEP
SLEEP/S2R to MCU Masked
ONLY
17 TRIGGER_B (NSLEEP2&NSLEEP1 ACTIVE or DEEP
= '10') SLEEP/S2R to MCU Masked
ONLY
18 TRIGGER_D or TRIGGER_C ACTIVE or MCU ONLY
Masked Masked
(NSLEEP2 = '0' ) to DEEP SLEEP/S2R
19 TRIGGER_I2C_0 (self-cleared) Any state to STANDBY Masked Masked
20 Always '1' (2) STANDBY to SAFE
Mask Masked Masked Masked
RECOVERY
21 Not Used Mask Masked Masked Masked
22 Not Used Mask Masked Masked Masked
23 Not Used Mask Masked Masked Masked
24 Not Used Mask Masked Masked Masked
Table 8-16. List of Trigger Used in Example Mission State Machine (continued)
Trigger ID Trigger Masked In Each User Defined Power State
Trigger Signal State Transitions DEEP
STANDBY ACTIVE MCU ONLY
SLEEP / S2R
25 Not Used Mask Masked Masked Masked
26 Not Used Mask Masked Masked Masked
27 Not Used Mask Masked Masked Masked
28-bit TRIG_MASK Value in Hex format: 0xFFE4FF8 0xFF18180 0xFF01270 0xFFC9FF0
If one of the events listed in Table 8-17 occurs, then the event powers on the device unless one of the gating
conditions listed in Table 8-18 is present.
Table 8-18. ON Requests Gating Conditions
EVENT MASKABLE COMMENT
VCCA > VCCA_OVP Device stays in SAFE RECOVERY until VCCA
VCCA_OVP (event) No
< VCCA_OVP
VCCA_UVLO (event) No VCCA < VCCA_UVLO
VINT_OVP (event) No LDOVINT > 1.98 V
The NPWRON_SEL NVM register bit determines whether the nPWRON/ENABLE pin is treated as a power on
press button or a level sensitive enable switch. When this pin is configured as the nPWRON button, a short
button press detection is latched internally as a device enable signal until the NPWRON_START_INT is cleared,
or a long press key event is detected. The short button press detection occurs when an falling edge is detected
at the nPWRON pin. When the NPWRON_START_MASK bit is set to '1', the device does no longer react to the
changing state of the pin as the nPWRON press button.
The nPWRON/ENABLE pin is a level sensitive pin when it is configured as an ENABLE pin, and an assertion
enables the device until the pin is released. When the ENABLE_MASK bit is set to '1', the device does no longer
react to the changing state of the pin as the ENABLE switch.
8.4.1.2.4.2 OFF Requests
An OFF request is used to orderly switch off the device. OFF requests initiate transition from any other mission
state to the STANDBY state or the LP_STANDBY state depending on the setting of the LP_STANDBY_SEL bit.
Table 8-19 lists the conditions to generate the OFF requests and the corresponding destination state.
Table 8-19. OFF Requests
LP_STANDBY_SEL BIT
EVENT DEBOUNCE DESTINATION STATE
SETTING
The long press key event occurs when the nPWRON pin stays low for longer than tLPK_TIME while the device is in
a mission state.
When the nPWRON or ENABLE pin is used as the OFF request, the device wakes up from the STANDBY or
the LP_STANDBY state through the ON request initiated by the same pin. The NPWRON_START_MASK or the
ENABLE_MASK must remain low in this case to allow the detection of the ON request initiated by the pin. If
the device needs to enter the LP_STANDBY state through the OFF request, it is important that the state of the
nPWRON or ENABLE pin must remain the same until the state transition is completed. Failure to do so may
result in unsuccessful wake-up from the LP_STANDBY state when the pin re-initiates On request.
Using the I2C_TRIGGER_0 bit as the OFF request enables the device to wake up from the STANDBY or
the LP_STANDBY states through the detection of LP_WKUPn/WKUPn pins, as well as RTC alarm or timer
interrupts. To enable this feature, the device must set the I2C_TRIGGER_0 bit to '1' while the NSLEEPn signals
are masked, and the ON request (initialized by the nPWRON or ENABLE pins) must remain active. Also the
interrupt bits ENABLE_INT, FSD_INT and the GPIOx_INT bits (for the GPIOx pins assigned as WKUP1 or
WKUP2) must be cleared before setting the I2C_TRIGGER_0 bit to '1'.
8.4.1.2.4.3 NSLEEP1 and NSLEEP2 Functions
The SLEEP requests are activated through the assertion of nSLEEP1 or nSLEEP2 pins that are the secondary
functions of the 11 GPIO pins. These pins can be selected through GPIO configuration using the GPIOx_SEL
register bits. If the nSLEEP1 or nSLEEP2 pins are not available, the NSLEEP1B and NSLEEP2B register bits
can be configured in place for their functions. The input of nSLEEP1 pin and the state of the NSLEEP1B register
bit are combined to create the NSLEEP1 signal through an OR function. Similarly for the input of the nSLEEP2
pin and the NSLEEP2B register bit as they are combined to create the NSLEEP2 signal.
A 1 → 0 logic level transition of the NSLEEP signal generates a sleep request, while a 0 → 1 logic level
transition reverses the sleep request in the example PFSM from Figure 8-39. When a NSLEEPn signal
transitions from 1 → 0, it generates a sleep request to go from a higher power state to a lower power state.
When the signal transitions from 0 → 1, it reverses the sleep request and returns the device to the higher power
state.
The NSLEEP1 signal is designed to control the SoC supply rails. The NSLEEP2 signal is designed to control
the MCU supply rails. When NSLEEP1 signal changes from 1 → 0, depending on the state of NSLEEP2, the
TPS6593-Q1 device exits ACTIVE state and enters either the MCU ONLY or the S2R states. When NSLEEP2
signal changes from 1 → 0, the device enters the S2R state regardless the state of NSLEEP1.
When the NSLEEP2 input signal changes from 0 →1, the MCU supply rails are enabled and the device
exits S2R state. Depending on the state of NSLEEP1 signal, the device enters either the MCU ONLY or the
ACTIVE state. In order for the system to function correctly, the MCU rails must be enabled when the NSLEEP1
input signal changes from 0 →1, which enables the SOC supply rails. NSLEEP1 0 →1 transition is ignored if
NSLEEP2 is 0.
The NSLEEPn_MASK bit can be used to mask the sleep request associated with the corresponding NSLEEPn
signal. When the NSLEEPn_MASK = 1, the corresponding NSLEEPn signal is ignored. Table 8-20 shows how
the combination of the NSLEEPn signals and NSLEEPn_MASK bits creates triggers A/B/C/D to the FSM to
control the power state of the device.
The states of the resources during ACTIVE, SLEEP, and DEEP SLEEP/S2R states are defined in the
LDOn_CTRL and BUCKn_CTRL registers. For each resource, a transition to the MCU ONLY or the DEEP
SLEEP/S2R states is controlled by the FSM when the resource is associated to the SLEEP or DEEP
SLEEP/S2R states.
Table 8-20 shows the corresponding state assignment based on the state of the NSLEEPn and their
corresponding mask signals using the example PFSM from Figure 8-39.
Table 8-20. NSLEEPn Transitions and Mission State Assignments
Current State NSLEEP1 NSLEEP2 NSLEEP1 MASK NSLEEP2 MASK Trigger to FSM Next State
DEEP SLEEP/S2R 0 0→1 0 0 TRIGGER B MCU ONLY
DEEP SLEEP/S2R 0→1 0→1 0 0 TRIGGER A ACTIVE
DEEP SLEEP/S2R Don't care 0→1 1 0 TRIGGER A ACTIVE
DEEP SLEEP/S2R 0→1 Don't care 0 1
TRIGGER A ACTIVE
or MCU ONLY
MCU ONLY 0→1 1 0 0 TRIGGER A ACTIVE
MCU ONLY 0 1→0 0 0 DEEP SLEEP
TRIGGER D
or S2R
MCU ONLY Don't care 1→0 1 0 DEEP SLEEP
TRIGGER D
or S2R
ACTIVE 1→0 1 0 0 TRIGGER B MCU ONLY
ACTIVE 1→0 1→0 0 0 DEEP SLEEP
TRIGGER D
or S2R
ACTIVE Don't care 1→0 1 0 DEEP SLEEP
TRIGGER D
or S2R
ACTIVE 1→0 Don't care 0 1 TRIGGER B MCU ONLY
If multiple edge detections of WKUP signals occur simultaneous, the device goes to the state in the following
priority order:
1. ACTIVE
2. MCU ONLY
When a valid edge is detected at a WKUP pin, the nINT pin generates an interrupt to signal the MCU of the
wake-up event, and the GPIOx_INT interrupt bit is set. The wake request remains active until the GPIOx_INT bit
is cleared by the MCU. While the wake request is executing, the device does not react to sleep requests to enter
a lower power state until the corresponding GPIOx_INT interrupt bit is cleared to cancel the wake request. After
the wake request is canceled, the device returns to the state indicated by the NSLEEP1 and NSLEEP2 signals
as shown in Table 8-20.
8.4.1.2.4.5 LP_WKUP Pins for Waking Up from LP STANDBY
The LP_WKUP functions are activated through the edge detection of LP_WKUP pins, configurable as secondary
functions of GPIO3 and GPIO4. They are specially designed to wake the device up from the LP STANDBY state
when a high speed wake-up signal is detected. Similar to the WKUP1 and WKUP2 pins, when GPIO3 or GPIO4
pin is configured as an LP_WKUP1 pin, a rising or falling edge detected at the input of this pin (configurable by
the GPIOn_FALL_MASK and the GPIOn_RISE_MASK bits) wakes up the device to the ACTIVE state. Likewise,
if the pin is configured as an LP_WKUP2 pin, a detected edge wakes up the device to the MCU ONLY state. If
multiple edge detections of LP_WKUP signals occur simultaneously, the device goes to the state in the following
priority order:
1. ACTIVE
2. MCU ONLY
Note
Due to a digital control erratum in the device, the ENABLE_MASK/NPWRON_START_MASK bit must
be set to '1' before the device enters LP_STANDBY state in order for the LP_WKUP2 pin to correctly
wake up the device to the MCU_ONLY state.
The TPS6593-Q1 device supports limited CAN wake-up capability through the LP_WKUP1/2 pins. When an
input signal (without deglitch) with logic level transition from high-to-low or low-to-high with a minimum pulse
width of tWK_PW_MIN is detected on the assigned LP_WKUP1/2 pins, the device wakes up asynchronously and
executes the power up sequence. CAN-transceiver RXD- or INH-outputs can be connected to the LP_WKUP
pin. If RXD-output is used, it is assumed that the transceiver RXD-pin IO is powered by the transceiver itself from
an external supply when TPS6593-Q1 is in the LP_STANDBY state. If INH-signal is used it has to be scaled
down to the recommenced GPIO input voltage level specified in the electrical characteristics table.
In this PFSM example, the device can wake up from the LP_STANDBY state through the detection of LP_WKUP
pins only if it enters the LP_STANDBY state through the TRIGGER_I2C_0 OFF request while the NSLEEPn
signals are masked, and the on request initialized by the nPWRON/ENABLE pin remains active. Once a valid
wake-up signal is detected at the LP_WKUP pin, it is handled as a WAKE request. The nINT pin generates an
interrupt to signal the MCU of the wake-up event, and the corresponding GPIOx_INT interrupt bit is set. The
wake request remains active until the interrupt bit is cleared by the MCU. Table 8-20 shows how the device
returns to the state indicated by the NSLEEP1 and NSLEEP2 signals after the wake request is canceled.
Figure 8-40 illustrates the valid wake-up signal at the LP_WKUP1/2 pins, and the generation and clearing of the
internal wake-up signal.
> tWK_PW_MIN
device stays in the SAFE RECOVERY state until VCCA voltage is below the VCCA_UVLO threshold and the
device is power-cycled.
The power resources assigned to the SoC rail group are typically assigned to the SOC power error handling
sequence. In this PFSM example depicted in Figure 8-39, when a power resource in this group is detected,
the PFSM typically causes the device to execute the shutdown of all the resources assigned to the SoC rail
group, and the device enters the MCU ONLY state. The device immediately resets the attached SoC by toggling
the nRSTOUT_SoC (GPO1 or GPIO11) pin. The reset output to the MCU and the resources assigned to the
MCU rail group remain unchanged. The EN_DRV pin also remains unchanged, and the nINT pin signals that an
SOC_PWR_ERR_INT interrupt event has occurred. To recover from the MCU_ONLY state after a SOC power
error, the MCU software must set NSLEEP1 signal to '0' while NSLEEP2 signal remains '1'. This action signals
TPS6593-Q1 that MCU has acknowledged the SOC power error, and is ready to return to normal operation.
MCU can then set the NSLEEP1 signal back to '1' for the device to return to ACTIVE state and reattempt the
SoC power up. Refer to Section 8.4.1.2.4.3 for information regarding the setting of the NSLEEP1 and NSLEEP2
signals.
The power resources used for peripheral devices in the system, for which no error-handling action is required,
the rail mapping needs to be selected as "no group assigned".
8.4.1.3.2 Catastrophic Error
Catastrophic errors are errors that affect multiple power resources such as errors detected in supply voltage,
LDOVINT supply for control logic, errors on internal clock signals, device temperature passing the thermal
shutdown threshold, error detected on the SPMI bus, or an error detected in the PFSM sequence.
Following errors are grouped as severe errors:
• VCCA > OVP threshold
• Junction temperature > immediate shutdown level
• Error in PFSM Sequence
For these above listed errors, depending on the setting of bits SEVERE_ERR_TRIG[1:0], an immediate
or orderly shutdown condition is created. The PFSM executes the corresponding sequence for the
IMMEDIATE_SHUTDOWN trigger or the ORDERLY_SHUTDOWN trigger and enters the SAFE RECOVERY
state.
Following errors are grouped as moderate errors:
• Junction temperature > orderly shutdown level
• BIST failure
• CRC error in register map
• Recovery counter exceeding the threshold value
• Error on SPMI bus
• Readback error on nRSTOUT pin or nINT pin
For these above listed errors, depending on the setting of bits MODERATE_ERR_TRIG[1:0], an immediate
or orderly shutdown condition is created. The PFSM executes the corresponding sequence for the
IMMEDIATE_SHUTDOWN trigger or the ORDERLY_SHUTDOWN trigger and enters the SAFE RECOVERY
state.
For following errors, the device performs an immediate shutdown and resets all internal logic circuits:
• VCCA < UVLO threshold
• Error on LDOVINT supply
• Errors on internal clock signals
• Unrecoverable CRC error in the SRAM memory of the PFSM
For all of the above listed errors, the device resets the attached MCU and SoC by driving the nRSTOUT and
the GPIO pin used as nRSTOUT_SoC (GPO1 or GPIO11) pins low. All of the power resources assigned to the
attached MCUand SOC are shut down. The nINT pin is driven low to signal an interrupt event has occurred, and
the EN_DRV pin is forced low.
VCCA
VCCA_UVLO
tINIT_REF_CLK_LDO
NVM
Inializa on
tBOOT_BIST
Boot BIST
Compleon
Power Up Rail
Sequencing Reset delay
nRSTOUT
tINIT_REFCLK_LDO is the start-up time for the reference block. tINIT_NVM_ANALOG is the time for the device to load
the default values of the NVM configurable registers from the NVM memory, and the start-up time for the analog
circuits in the device. tINIT_REFCLK_LDO and tINIT_NVM_ANALOG are defined in the electrical characteristics table.
BOOT_BIST is the sum of tLBISTrun (which is defined in the electrical characterization tables) and a less than 60us
time-interval needed for the cyclic redundancy check on the register map and the SRAM.
The Power Sequence time is the total time for the device to complete the power up sequence. Please refer to
Section 8.4.1.5 for more details.
The reset delay time is a configurable wait time for the nRSTOUT and the nRSTOUT_SoC release after the
power up sequence is completed.
Valid On X X
Request
Valid Off X X
Request
BUCK3 t(inst16)
t(inst1)
LDO1 t(inst15)
t(inst2)
BUCK2 t(inst14)
t(inst3)
LDO2 t(inst13)
t(inst4)
REGEN1 t(inst12)
t(inst5)
LDO4 t(inst11)
t(inst6)
LDO3
As the power sequences of the TPS6593-Q1 device are defined according to the processor requirements, the
total time for the completion of the power sequence varies across various system definitions.
8.4.1.6 First Supply Detection
The TPS6593-Q1 device can be configured to automatically start up from a first supply-detection (FSD)
event detection. This feature is enabled by setting the FSD_MASK register bit to '0', and setting the
NPWRON_SEL[1:0] registers bits to '10' or '11' to mask the functionality of the nPWRON/ENABLE pin. When the
device is powered up from the NO SUPPLY state, the FSD detection is validated after the NVM default for this
feature is loaded into the device memory.
When the FSD feature is enabled, the PMIC immediately powers up from the NO SUPPLY state to an operation
state, configured by the STARTUP_DEST[1:0] bits when VCCA > VCCA_UV, while VCCA_UV gating is
performed, and only when VCCA voltage monitoring is enabled (VCCA_VMON_EN = 1). After the device arrives
the corresponding STARTUP_DEST[1:0] operation state, the MCU must setup the NSLEEP1 and NSLEEP2
signals accordingly before clearing the FSD_INT interrupt. Once the interrupt is cleared, the device either stays
in the current state or moves to the destination state according to the state of the NSLEEP1 and NSLEEP2
signals as specified in Table 8-20.
LDOVINT The LDOVINT registers are powered by the internal LDOVINT, and retain their values until the
registers device enters the LP_STANDBY state or the BACKUP state after the device was fully powered
up and in operation. When this occurs, LDOVINT is powered off, all LDOVINT registers (including
the VSET registers which store the output voltage levels for all of the external power rails)
are reset. As the device re-enters the INIT state from a wake up signal or an On-request, the
registers powered by the LDOVINT are re-written with the default values from NVM (Non-Volatile
Memory). All registers in the device, except the LDOVRTC registers (registers in RTC domain),
are powered by LDOVINT.
LDOVRTC The LDOVRTC registers (registers in RTC domain) retain their values until a Power-On-Reset
registers (POR) occurs. POR occurs when the device loses supply power and enters the NO SUPPLY
(registers in state. When this occurs, LDOVRTC is powered off, and all LDOVRTC registers are reset.
RTC
Following are the LDOVRTC registers:
domain)
• All RTC registers
• RTC and Crystal Oscillator bits
• Status registers for the following events: TSD and RTC reset
• Control registers for PWRON/ENABLE, GPIO3, and GPIO4 pins (for wake signal monitor
during LP_STANDBY state)
• Following interrupt registers:
– FSD_INT
– RECOV_CNT_INT
– TSD_ORD_INT
– TSD_IMM_INT
– PFSM_ERR_INT
– VCCA_OVP_INT
– ESM_MCU_RST_INT
– ESM_SOC_RST_INT
– WD_RST_INT
– WD_LONGWIN_TIMEOUT_INT
– NPWRON_LONG_INT
Error Feedback
Effective Effective
Thermal
Effective
Thermal
Signal list Current Current
Internal condition list STATE_1 Voltage Voltage
STATE_1
... ...
OFF STATE 2 Exit Condition 1 STATE 2 Exit Condition 1
STATE 2 Exit Condition 2
Signal list Signal list
STATE_2 Signal list
Internal condition list Internal condition list
power
Internal condition list
ERROR
OFF
STATE_2
STATE 2 Exit Condition 2
Signal list
Internal condition list power OFF
STATE_2
STATE 2 Exit Condition 2
Signal list
Internal condition list
power
state
(SAFE STATE)
STATE_3
Controller for
ERROR
(SAFE STATE)
state ERROR
(SAFE STATE)
state
STATE_3
Controller for
Output Power
STATE_N
Output Power Output Power
Supply Rails STATE_N STATE_N
ENABLE or WAKE
signals from
external hardware
Power-State Control
Signals such as:
ACTIVE, SLEEP, RESET, Scalable Microprocessor and System on Chip
ERROR, TRACKING
In this scheme, each primary and secondary PMIC runs on its own system clock, and maintains its own register
map. Each PMIC monitors its own activities and pulls down the open-drain output of nINT or PGOOD pin when
errors are detected. The microprocessor must read the status bits from each PMIC device through the I2C or
SPI interface to find out the source of the error that is reported.
To synchronize the timing when entering and exiting from the LP_STANDBY state, the VOUT_LDOVINT of
the TPS6593-Q1 device must be connected to the ENABLE input of the secondary PMICs, which are the
target devices in the SPMI interface bus. Figure 8-44 illustrates the pin connections between the primary, the
secondary, and the application processor or the System-on-Chip.
TPS6593-Q1
www.ti.com SLVSE83B – DECEMBER 2020 – REVISED SEPTEMBER 2023
x VIO
Secondary PMIC
I2C_SCL
I2C_SDA
SDATA
SCLK
Power Interrupt
Sequencer nINT Handler
Power Good
ENABLE PGOOD Monitor
x
Secondary PMIC
I2C_SCL
I2C_SDA
SDATA
SCLK
Power
Sequencer nINT
x
nERRORx
Power nSLEEPx
Sequencer
GPIO
nRSTOUTx
WAKEn
SCLK PG
SDATA
I2C2_SCL
Q&A
I2C2_SDA
WDOG
The power sequencer of the multiple PMICs are synchronized at the beginning of each power up and power
down sequence; a variation in the sequence timing, however, is still possible due to the ±5% clock accuracy of
the independent system clocks on the primary and secondary PMICs. The worst-case sequence timing variation
from different PMIC rails is up to ±10% of the target delay time. Figure 8-45 illustrates the creation of this timing
variation between PMICs.
Primary PMIC
System clock
(+/-5% accuracy)
... ...
Primary PMIC
Rail X
Primary PMIC
Rail Y
Secondary PMIC
...
System clock
...
(+/-5% accuracy)
Secondary PMIC
Rail Z
SCLK
SCLK
SDATA P A7 A6 A5 A4 A3 A2 A1 A0 P
SCLK
SDATA P D7 D6 D5 D4 D3 D2 D1 D0 P
SCLK
SDATA P D7 D6 D5 D4 D3 D2 D1 D0 P 0
VIN
PWRON/ENABLE
SPMI
After the input power is detected and verified to be at the correct level, the TPS6593-Q1 initializes itself by
reading the NVM and performs all actions that are needed to prepare for operation. After this initialization, the
TPS6593-Q1 enters the BOOT BIST state, in which the internal logic performs a series of tests to verify that the
TPS6593-Q1 device is OK. As part of this test, the SPMI- BIST is performed. After it is completed successfully,
the TPS6593-Q1 device goes to the standby state and waits for further signals from the system to initiate the
power-up sequence of the processor.
A valid on request initiates the processor power-up sequence. The controller device communicates this event
through the SPMI bus to all of the target devices. After that, the power-up sequence is executed and TPS6593-
Q1 enters the configured mission state.
8.4.2.4.1 SPMI Bus during Boot BIST and RUNTIME BIST
During Boot BIST and RUNTIME BIST, both the Logic BIST (LBIST) on the SPMI logic and the SPMI-BIST
are performed to check correct operation of the SPMI bus. The LBIST is performed first before the SPMI-BIST
during BOOT BIST and RUNTIME BIST. The SPMI-BIST is implemented by reading TID from each target device
on the SPMI bus into the controller device, and ensuring they are unique and match the expected amount of
target devices. This process of checking the TID of each target device ensures that:
• All SPMI target devices are present in the system as expected
• The SPMI logic blocks are working on the SPMI controller device and all of the SPMI target devices
• The pins and wires on the ICs and PCB are in working order
The SPMI-BIST is initiated by the SPMI controller block in the primary PMIC by writing a request to all SPMI
target device(s) (using GTID) to send their TIDs to the SPMI controller block of the primary PMIC. Upon
receiving this command from the SPMI controller device, the SPMI target devices request SPMI bus arbitration
using the SR-bit protocol. Upon winning the bus arbitration the SPMI target devices transmit their TID into the
SPMI target block of the primary PMIC.
The SPMI controller block of the primary PMIC contains a list of all SPMI target device(s) on the SPMI bus and
their TIDs in the register set. The SPMI controller block of the primary PMIC reads the TID from each SPMI
target device and compares the result with the stored TID for the corresponding SPMI target device. The SPMI
controller device has to ensure that every non-zero TID on its list is returned, in order to support use cases in
which there are two or more identical SPMI target devices, with same TID, in the system. In these cases, it is
mandatory that the expected number of the same TIDs is returned. If no identical PMICs are to be used, then a
return of the same TID multiple times is an error due to incorrect assembly of identical PMICs onto the PCB. An
all-zero TID stored in the list of the primary PMIC indicates that there are no SPMI target device(s) present on
the SPMI Bus.
8.4.2.4.2 Periodic Checking of the SPMI
The SPMI controller block in the primary PMIC executes the SPMI-BIST periodically while device is operating.
The time-period after the SPMI-BIST is repeated according factory-configured settings during the device boot
time, and after the device reaches mission states. The factory-configured settings of this SPMI-BIST time period
must be the same for all PMICs on the same SPMI network. The SPMI target devices on the SPMI bus expect a
request for sending its TID from the SPMI controller device within 1.5x the factory-configured period . This factor
1.5x provides enough margin for clock uncertainty between the SPMI controller device and the SPMI target
device.
During mission state operation, the SPMI controller device expects the SPMI target devices to respond to the
TID request within the factory-configured polling time-out period . In other words, from the polling start command
each SPMI target device must respond within this factory-configured time interval.
During boot time or when the device enters Safe Recovery state, to prevent the SPMI controller device from
polling the SPMI target devices too often while one or more of these recovering from a system error such as a
thermal shutdown event, the device sets a longer timeout period that allows the SPMI target devices to respond
to the SPMI controller device before he SPMI controller device reports an error.
If one or more devices on the SPMI bus cause a violating of the polling time-out period either during start-up
or during normal operation, the SPMI controller block in the affected PMIC(s) sets a SPMI error trigger signal
to the PFSM of the affected PMIC(s), causing a complete shutdown of the affected PMIC(s). As a result, the
affected PMIC(s) no longer respond on the SPMI bus, which in turn is detected by the SPMI controller block off
the non-affected PMICs on the SPMI bus. The SPMI controller block in these PMICs sets an SPMI error to the
PFSM in these PMICs, causing a complete shutdown of these PMICs. Therefore, all PMICs are finally shutdown
if one or more devices on the SPMI bus cause a violating of the polling time-out period .
8.4.2.4.3 SPMI Message Priorities
The SPMI Bus uses the protocol priority levels listed in Table 8-21 for each type of communication message.
Table 8-21. SPMI Message Types and Priorities
SPMI protocol priority level Name of priority level in SPMI standard Message types
State transition messages from
Highest A-bit arbitration target device(s) to controller
device
State transition messages from
priority arbitration controller device to target
device(s)
target device TID to controller
SR-bit arbitration
device
Controller device request of TIDs
Lowest secondary arbitration
from target device(s)
t
rs
24 24-bit bus ordering value for SPI: 0
it
b
nt
i
ign
ts
os
M
1 Q D 1 Q D 1Q D 1Q D 1 Q D 1 Q D 1 Q D 1Q D
Flip Flop Flip Flop Flip Flop Flip Flop Flip Flop Flip Flop Flip Flop Flip Flop
i
ign
ts
os
M
1 1 1 1 1 1 1 1
Q D Q D Q D Q D Q D Q D Q D Q D
Flip Flop Flip Flop Flip Flop Flip Flop Flip Flop Flip Flop Flip Flop Flip Flop
SCL
SDA
SDA
SCL
S P
START STOP
Condition Condition
The I2C bus is considered busy after a START condition and free after a STOP condition. The I2C controller
device can generate repeated START conditions during data transmission. A START and a repeated START
condition are equivalent function-wise. Figure 8-52 shows the SDA and SCL signal timing for the I2C-compatible
bus. For timing values, see the Specification section.
tBUF
SDA
SCL
tHD;STA tSU;STA tSU;STO
tHIGH
tHD;DAT
S
tSU;DAT RS P S
START REPEATED STOP START
START
1 1 0 0 0 0 0 R/W
Bit 7 Bit 6 Bit 5 Bit 4 Bit 3 Bit 2 Bit 1 Bit 0
For safety applications, the device supports read and write protocols with embedded CRC data fields. In a write
cycle, the I2C controller device (the MCU) must provide the 8-bit CRC value after sending the write data bits and
receiving the ACK from the target (the TPS6593-Q1 PMIC). The CRC value must be calculated from every bit
included in the write protocol except the ACK bits from the target. See CRC Calculation for I2C and SPI Interface
Protocols. In a read cycle, the I2C target must provide the 8-bit CRC value after sending the read data bits and
the ACK bit, and expect to receive the NACK from the controller at the end of the protocol. The CRC value must
be calculated from every bit included in the read protocol except the ACK and NACK bits. See CRC Calculation
for I2C and SPI Interface Protocols.
Note
If I2C CRC is enabled in the device and an I2C write without R_CRC bits is done, the device does not
process the write request. The device does not set any interrupt bit and does not pull the nINT pin low.
The embedded CRC field can be enabled or deactivated from the protocol by setting the I2C1_SPI_CRC_EN
register bit to '1' - enabled, '0' - deactivated. The default of this bit is configurable through the NVM.
In case the calculated CRC-value does not match the received CRC-check-sum, an I2C-CRC-error is detected,
the COMM_CRC_ERR_INT bit is set, unless it is masked by the COMM_CRC_ERR_MASK bit. The MCU must
clear this bit by writing a ‘1’ to the COMM_CRC_ERR_INT bit.
When the CRC field is enabled, in the case when MCU attempts to write to a read-only register
or a register-address that does not exist, the device sets the COMM_ADR_ERR_INT bit, unless the
COMM_ADR_ERR_MASK bit is set. The MCU must clear this bit by writing a ‘1’ to the COMM_ADR_ERR_INT
bit.
START ACK ACK ACK STOP
SCL
SDA
SCL
SDA
The I2C controller device (the MCU) provides R_CRC[7:0], which is calculated from the I2C_ID, R/W, ADDR, and the WDATA bits (24
bits). See CRC Calculation for I2C and SPI Interface Protocols.
SCL
SDA
When READ function is to be accomplished, a WRITE function must precede the READ function as shown above.
REPEATED STOP
START
START ACK ACK ACK ACK NCK
SCL
SDA
The I2C target device (the TPS6593-Q1) provides T_CRC[7:0], which is calculated from the I2C_ID, R/W, ADDR, I2C_ID, R/W, and the
RDATA bits (32 bits). See CRC Calculation for I2C and SPI Interface Protocols.
CS_SPI
SCLK_SPI
PAGE
SDI_SPI ADDR[7:0] [2:0]
0 Reserved[3:0] WDATA[7:0]
CS_SPI
SCLK_SPI
PAGE
SDI_SPI ADDR[7:0] [2:0]
0 Reserved[3:0] WDATA[7:0] R_CRC[7:0]
CS_SPI
SCLK_SPI
PAGE
SDI_SPI ADDR[7:0] [2:0]
1 Reserved[3:0]
CS_SPI
SCLK_SPI
PAGE
SDI_SPI ADDR[7:0] [2:0] 1 Reserved[3:0]
Note
Due to a digital control erratum in the device, the TPS6593-Q1 pulls the nINT pin low and sets
interrupt bit COMM_FRM_ERR_INT if the pin CS_SPI is low during device power-up and goes high
after completion of the device power-up sequence. After system start-up, the MCU must clear this
COMM_FRM_ERR_INT bit such that the TPS6593-Q1 can release the nINT pin.
Note
When I2C Interfaces are used, each of the above listed register pages has its own 6-bit I2C device
address. In order to address Page 0 to 3, the two LSBs if the pre-configured I2C1_ID are replaced
with 00 for Page 0, 01 for Page 1, 10 for Page 2 and 11 for Page 3. As an example, if I2C1_ID=0x26
(0100110b) , Page 0 to 3 have following addresses:
• Page 0: 0100100
• Page 1: 0100101
• Page 2: 0100110
• Page 3: 0100111
For Page 4 the I2C device address is according register bits I2C2_ID. Therefore, in case both I2C1
and I2C2 Interfaces are used, each TPS6593-Q1 device occupies four I2C device addresses (for Page
0, Page 1, Page 2 and Page 3) on the I2C1 bus and one I2C device address (for Page 4) on the
I2C2 bus. And in case only I2C1 Interfaces is used, each TPS6593-Q1 device occupies five I2C device
addresses (for Page 0, Page 1, Page 2, Page 3 and Page 4) on the I2C1 bus. In case multiple devices
are used on a common I2C bus, care must be taken to avoid overlapping I2C device addresses.
Note
When SPI Interface is used, the above listed register pages are addresses with the PAGE[2:0] bits:
0x0 addresses Page 0, 0x1 addresses Page 1, 0x2 addresses Page 2, 0x3 addresses Page 3
Note
The CRC-16 engine assumes a default value of '0' for all undefined or reserved bits in all control
registers. Therefore, the software MUST NOT write the value of '1' to any of these undefined or
reserved bits. If the value of '1 is written to any undefined or reserved bit of a writable register, the
CRC-16 engine detects a mismatch between the calculated and expected CRC-16 values, and hence
the interrupt bit REG_CRC_ERR_INT is set and the device forces an orderly shutdown sequence to
return to the SAFE RECOVERY state.
Note
If a RESERVED bit in a R/W configuration register gets set to 1h through a I2C/SPI write, the
TPS6593-Q1 detects a CRC error in the register map. Therefore, it is important that system software
involved in the I2C/SPI write-access to the TPS6593-Q1 keeps all RESERVED bits (all bits with the
word RESERVED in the Register Field Description tables in the Register Map section at 0h.
Complex bit access types are encoded to fit into small table cells. Table 8-24 shows the codes that are used for
access types in this section.
Table 8-24. TPS6593-Q1 Access Type Codes
Access Type Code Description
Read Type
R R Read
Write Type
W W Write
W1C W Write
1C 1 to clear
WSelfClrF W Write
Reset or Default Value
-n Value after reset or the default
value
LDO4 1.8V
(300mA max) 3.3V VDDSHVx
VIO_IN
I2C I2C
nRSTOUT PORz
EN 3.3V System
Load Switch
LPDDR4
1.8V
1.1V
9.2.1.2.1 VCCA
The VCCA pin provides power to the LDOVINT regulator and other internal functions. This VCCA pin is always
connected in parallel with the buck input pins (PVIN_Bx pins). The VCCA pin can be connected to an optional
0.47-µF bypass capacitor close to the pin.
Table 9-1. Recommended VCCA Components
EIA SIZE
COMPONENT MANUFACTURER PART NUMBER VALUE SIZE (mm) USED for VALIDATION
CODE
Capacitor Murata GCM155C71A474KE36 0.47 µF, 10 V, X7R 0402 1.0 × 0.5 Yes
Capacitor TDK CGA2B3X7S1A474K050BB 0.47 µF, 10 V, X7R 0402 1.0 × 0.5 —
CIN1 and CIN2 are both 12-pF for this device. CPCB1 and CPCB2 depend on the board but is generally around
1-pF. The crystal oscillator chosen must have a required load capacitance of either 6-pF, 9-pF, or 12.5-pF and
the value of the XTAL_SEL bit in the RTC_CTRL_2 register must be updated based on the oscillator chosen. To
achieve the required load capacitance (CL) for the oscillator, Equation 25 is used. This equation assumes that
the crystal series capacitance is negligible.
CL = (CL1 + CPCB1 + CIN1) × (CL2 + CPCB2 + CIN2) / ((CL1 + CPCB1 + CIN1) + (CL2 + CPCB2 + CIN2)) (25)
Assuming CL1 = CL2, this simplifies to CL1 = 2 × CL - CPCB - CIN. Simplifying this into the standard capacitor
values typically available results in the following general capacitor recommendations. If more precise matching is
desired, complete the exercise without series capacitance neglected and with exact PCB parasitic capacitance.
Too much capacitance results in a lower than expected oscillator frequency, while not enough capacitance has
the opposite impact.
The recommended components using a 9-pF oscillator as an example are in Table 9-4. If an alternate load
capacitance crystal is used, the values of the load capacitors must be adjusted to match based on the above.
Table 9-4. Recommended Crystal Oscillator Components for 9-pF Crystal
EIA size
Component MANUFACTURER PART NUMBER VALUE SIZE (mm) Used for Validation
code
Capacitor Murata GCM155R71C104JA55D 100-nF, 16-V, X7R 0402 1.0 x 0.5 Yes
Capacitor TDK CGA2B1X7R1C104K050BC 100-nF, 16-V, X7R 0402 1.0 x 0.5 -
Crystal NDK NX3215SD-32.768K-STD- 32.768-kHz, ±20 3.2 x 1.5 x 0.9 Yes
MUS-6 ppm, 9-pF
Crystal Abracon ABS07AIG-32.768kHz-9-T 32.768-kHz, ±20 3.2 x 1.5 x 0.9 -
ppm, 9-pF
Capacitor Murata GCM1555C1H6R0CA16 6-pF, 50-V, 0402 1.0 x 0.5 Yes
C0G/NP0
Capacitor TDK CGA2B2C0G1H060D050BA 6-pF, 50-V, 0402 1.0 x 0.5 -
C0G/NP0
Figure 9-3. Example Power Distribution Network (PDN) of Local and POL Capacitors
Table 9-6. Local and POL Capacitors used for Buck Use Case Validation
LPCB per
Configuration COUT L CL / phase RPCB per phase1 CPOL1 (total) CPOL2 (total)
phase2
4.4 MHz VOUT Less than 1.9 V, Low Load Low 470 nH 22 uF x 1 8 mΩ 2.5 nH 1 uF x 1
Step, Single Phase with low COUT
Low 220 nH 47 µF × 2 8 mΩ 2.5 nH 10 µF × 4
4.4 MHz VOUT Less than 1.9 V, Multiphase
High 220 nH 47 µF × 4 8 mΩ 2.5 nH 10 µF × 2
4.4 MHz VOUT Less than 1.9 V, Single Low 220 nH 47 µF × 1 8 mΩ 2.5 nH 10 µF × 4
Phase with high COUT
High 220 nH 47 µF × 4 8 mΩ 2.5 nH 10 µF × 2
4.4 MHz VOUT Less than 1.9 V, Single Low 220 nH 22 µF × 1 8 mΩ 2.5 nH 10 µF × 2
Phase with low COUT High 220 nH 47 µF × 1 8 mΩ 2.5 nH 10 µF × 4
2.2 MHz Full VOUT Range and VIN Greater Low 1000 nH 47 µF × 3 8 mΩ 2.5 nH 10 µF × 4
than 4.5 V, Single Phase Only High 1000 nH 47 µF × 3 8 mΩ 2.5 nH 10 µF × 4 680 µF × 1
2.2 MHz VOUT Less than 1.9 V Multiphase Low 470 nH 47 µF × 3 4.1 mΩ 1.3 nH 10 µF × 4
or Single Phase High 470 nH 47 µF × 3 4.1 mΩ 1.3 nH 10 µF × 4 680 µF × 1
2.2 MHz Full VOUT and Full VIN Range, Low 1000 nH 47 µF × 3 4.1 mΩ 1.3 nH 10 µF × 2
Single Phase Only High 1000 nH 100 µF × 4 4.1 mΩ 1.3 nH 10 µF × 2
DDR VTT Termination, 2.2 MHz Single 10 µF × 1 + 22 µF x
- 470 nH 22 µF × 1 27 mΩ 6 nH
Phase Only 1
1. RPCB is the PCB wiring resistance between local and POL capacitors including both positive and negative
paths. For multi-phase outputs the total resistance is divided by the number of phases.
2. LPCB is the PCB wiring inductance between local and POL capacitors including both positive and negative
paths. For multi-phase outputs the total inductance is divided by the number of phases.
Power input and output wiring parasitic resistance and inductance must be minimized.
Table 9-7. Recommended Buck Converter Output Capacitor Components
MANUFACTURER PART NUMBER VALUE EIA Size Code SIZE (mm) Used for Validation
Murata NFM15HC105D0G(1) 1 µF, 4 V, X7S 0402 1.0 × 0.5 Yes
TDK YFF18AC0J105M(1) 1 µF, 6.3 V 0603 1.6 × 0.8 -
Murata NFM18HC106D0G(1) 10 µF, 4 V, X7S 0603 1.6 × 0.8 Yes
TDK YFF18AC0G475M(1) 4.7 µF, 6.3 V 0603 1.6 × 0.8 -
Murata GCM31CR71A226KE02 22 µF, 10 V, X7R 1206 3.2 × 1.6 Yes
Murata GCM21BD7CGA5L1X7R0J226MT0J226M 22 µF, 6.3 V, X7T 0805 2.0 × 1.25 × 1.25 -
TDK CGA5L1X7R0J226MT 22 µF, 6.3 V, X7R 1206 3.2 × 1.6 -
TDK CGA4J1X7T0J226MT 22 µF, 6.3 V, X7T 0805 2.0 × 1.25 × 1.25 -
Murata GCM32ER70J476ME19 47 µF, 6.3 V, X7R 1210 3.2 × 2.5 Yes
Murata GCM31CD70G476M 47 µF, 4 V, X7T 1206 3.2 × 1.6 -
TDK CGA6P1X7S1A476MT 47 µF, 10 V, X7S 1210 3.2 × 2.5 -
TDK CGA5L1X7T0G476MT 47 µF, 4 V, X7T 1206 3.2 × 1.6 -
Murata GCM32ED70G107MEC4 100 µF, 4 V, X7S 1210 3.2 × 2.5 Yes
TDK CGA6P1X7T0G107MT 100 µF, 4 V, X7T 1210 3.2 × 2.5 -
Kemet T510X687K006ATA023(2) 680 µF, 6.3 V 2917 7.4 × 5.0 Yes
TDK TFM322512ALMAR22MTAA 220 nH, 7.6 A Max, 150 °C 3.2 x 2.5 x 1.2 Yes
TDK TFM201610ALMAR24MTAA 220 nH, 5 A Max, 150 °C 2.0 x 1.6 x 1.2 -
Murata DFE2MCAHR24MJ0 240 nH, 4.2 A Max, 150 °C 2.0 x 1.6 x 1.2 -
(1) Component minimum and maximum tolerance values are specified in the electrical parameters section of each IP.
For I2C pull-up resistor values, please refer to the I2C standard for the chosen use-case (standard mode, Fast
mode, Fast mode+, High-Speed mode with Cb = 100pF or 400pF)
80 80
Efficiency (%)
Efficiency (%)
60 60
40 40
VVOUT_Bn = 1.8 V Fsw = 2.2 MHz 4-Phase VPVIN_Bn = 3.3 V VVOUT_Bn = 1.8 V 4-Phase
Figure 9-4. BUCK Efficiency at 3.3 V or 5 V Input Voltage Figure 9-5. BUCK Efficiency with Fsw = 2.2 MHz or 4.4 MHz
95 95
90 90
85 85
Efficiency (%)
Efficiency (%)
80 Fsw = 2.2 MHz, 1 Phase 80 Fsw = 2.2 MHz, 1 Phase
Fsw = 2.2 MHz, 2 Phase Fsw = 2.2 MHz, 2 Phase
75 Fsw = 2.2 MHz, 3 Phase 75 Fsw = 2.2 MHz, 3 Phase
Fsw = 2.2 MHz, 4 Phase Fsw = 2.2 MHz, 4 Phase
Fsw = 4.4 MHz, 1 Phase Fsw = 4.4 MHz, 1 Phase
70 Fsw = 4.4 MHz, 2 Phase 70 Fsw = 4.4 MHz, 2 Phase
Fsw = 4.4 MHz, 3 Phase Fsw = 4.4 MHz, 3 Phase
Fsw = 4.4 MHz, 4 Phase Fsw = 4.4 MHz, 4 Phase
65 65
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
IOUT_Bn (A) IOUT_Bn (A)
Data valid for all bucks up to IOUT_Bn Data valid for all bucks up to IOUT_Bn
VPVIN_Bn = 3.3 V VVOUT_Bn = 1.8 V Auto Mode VPVIN_Bn = 3.3 V VVOUT_Bn = 1.8 V Auto Mode
Figure 9-6. BUCK Efficiency in Varied Phase Configuration, 3.3 Figure 9-7. BUCK Efficiency in Varied Phase Configuration, 5 V
V Input Input
100 100
80 80
Efficiency (%)
Efficiency (%)
60 60
40 40
VOUT_Bn = 0.8 V, F sw = 4.4 MHz VOUT_Bn = 0.8 V, F sw = 4.4 MHz
VOUT_Bn = 0.8 V, F sw = 2.2 MHz VOUT_Bn = 0.8 V, F sw = 2.2 MHz
VOUT_Bn = 1.2 V, F sw = 4.4 MHz VOUT_Bn = 1.2 V, F sw = 4.4 MHz
20 VOUT_Bn = 1.2 V, F sw = 2.2 MHz 20 VOUT_Bn = 1.2 V, F sw = 2.2 MHz
VOUT_Bn = 1.8 V, F sw = 4.4 MHz VOUT_Bn = 1.8 V, F sw = 4.4 MHz
VOUT_Bn = 1.8 V, F sw = 2.2 MHz VOUT_Bn = 1.8 V, F sw = 2.2 MHz
0 0
0 0.5 1 1.5 2 2.5 3 3.5 4 0 0.5 1 1.5 2 2.5 3 3.5 4
IOUT_Bn (A) IOUT_Bn (A)
Data valid for all bucks up to IOUT_Bn Data valid for all bucks up to IOUT_Bn
VPVIN_Bn = 3.3 V Single-Phase Forced-PWM Mode VPVIN_Bn = 5 V Single-Phase Forced-PWM Mode
Figure 9-8. BUCK Efficiency with different VOUT_Bn, 3.3 V Input Figure 9-9. BUCK Efficiency with different VOUT_Bn, 5 V Input
Efficiency (%)
80 60
75 40
-40oC
70 20 25oC
85oC
125oC
65 0
0 0.5 1 1.5 2 2.5 3 3.5 4 0 0.5 1 1.5 2 2.5 3 3.5 4
IOUT_Bn (A) IOUT_Bn (A)
Data valid for all bucks up to IOUT_Bn Data valid for all bucks up to IOUT_Bn
VPVIN_Bn = 3.3 V VVOUT_Bn = 1 V Single-Phase VPVIN_Bn = 3.3 V VVOUT_Bn = 1 V Single-Phase
Figure 9-10. BUCK Efficiency at different TA, Auto Mode Figure 9-11. BUCK Efficiency at different TA, Forced-PWM Mode
1.01 1.01
1 Phase 1 Phase
1.0075 2 Phase 1.0075 2 Phase
3 Phase 3 Phase
1.005 4 Phase 1.005 4 Phase
VVOUT_Bn (V)
VVOUT_Bn (V)
1.0025 1.0025
1 1
0.9975 0.9975
0.995 0.995
0.9925 0.9925
0.99 0.99
-40 -20 0 20 40 60 80 100 120 140 -40 -20 0 20 40 60 80 100 120 140
Temperature (°C) Temperature (°C)
VVOUT_Bn (V)
1.0025 1.0025
1 1
0.9975 0.9975
0.995 0.995
0.9925 0.9925
0.99 0.99
-40 -20 0 20 40 60 80 100 120 140 -40 -20 0 20 40 60 80 100 120 140
Temperature (°C) Temperature (°C)
VVOUT_Bn (V)
1.002 1.002
1 1
0.998 0.998
0.996 0.996
0.994 0.994
0.992 0.992
0.99 0.99
0 2 4 6 8 10 12 14 0 2 4 6 8 10 12 14
IOUT_Bn (A) IOUT_Bn (A)
Data valid for all bucks up to IOUT_Bn Data valid for all bucks up to IOUT_Bn
VPVIN_Bn = 3.3 V VVOUT_Bn = 1 V Auto Mode VPVIN_Bn = 3.3 V VVOUT_Bn = 1 V Auto Mode
Figure 9-18. Buck Load Regulation, with Fsw = 2.2 MHz Figure 9-19. Buck Load Regulation, with Fsw = 4.4 MHz
1.01 1.01
1 Phase 1 Phase
1.008 2 Phase 1.008 2 Phase
3 Phase 3 Phase
1.006 1.006
4 Phase 4 Phase
1.004 1.004
VVOUT_Bn (V)
VVOUT_Bn (V)
1.002 1.002
1 1
0.998 0.998
0.996 0.996
0.994 0.994
0.992 0.992
0.99 0.99
3 3.25 3.5 3.75 4 4.25 4.5 4.75 5 5.25 5.5 3 3.25 3.5 3.75 4 4.25 4.5 4.75 5 5.25 5.5
VPVIN_Bn (V) VPVIN_Bn (V)
VPVIN_Bn = 3.3 V VVOUT_Bn = 1 V Auto Mode VPVIN_Bn = 3.3 V VVOUT_Bn = 1 V Auto Mode
Figure 9-20. Buck Line Regulation, with Fsw = 2.2 MHz Figure 9-21. Buck Line Regulation, with Fsw = 4.4 MHz
VVOUT_Bn (10mV/div)
VVOUT_Bn (10mV/div)
VVOUT_Bn (10mV/div)
VVOUT_Bn (10mV/div)
VSW_Bn (2V/div)
VSW_Bn (2V/div)
Figure 9-44. Buck Load Step Transient - 3-Phase, 2.2 MHz, Figure 9-45. Buck Load Step Transient - 3-Phase, 4.4 MHz,
Forced-PWM Mode Forced-PWM Mode
Figure 9-52. Buck Load Step Transient - Buck4, 2.2 MHz, Figure 9-53. Buck Load Step Transient - Buck4, 4.4 MHz,
Forced-PWM Mode Forced-PWM Mode
ILOAD (0.4A/div)
ILOAD (0.4A/div)
VOUT(LDOn) (V)
1.802 3.001
1.8 3
1.798 2.999
1.796 2.998
1.794 2.997
1.792 2.996
1.79 2.995
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Load (A) Load (A)
VOUT(LDOn) (V)
0.802 1.805
0.8 1.8
0.798 1.795
0.796 1.79
0.794 1.785
0.792 1.78
1.2 1.5 1.8 2.1 2.4 2.7 3 3.3 3.6 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3 3.1 3.2 3.3
VIN(LDOn) (V) VIN(LDOn) (V)
3.2 3.2
3 3
2.8 2.8
VOUT(LDOn) (V)
VOUT(LDOn) (V)
2.6 2.6
2.4 2.4
2.2 2.2
2 2
1.8 1.8
1.6 1.6
0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5
Time (ms) Time (ms)
VOUT(LDOn) (V)
1.802
1.8
VOUT(LDOn) (V)
VOUT(LDOn) (V)
1.805
3.001
3 1.8
2.999 1.795
2.998
1.79
2.997
1.785
2.996
2.995 1.78
0 0.05 0.1 0.15 0.2 0.25 0.3 2.2 2.3 2.4 2.5 2.6 2.7 2.8 2.9 3 3.1 3.2 3.3
Load (A) VIN(LDOn) (V)
VVOUT_Bn (20mV/div)
ILOAD (0.2A/div)
Time (20µs/div)
ILOAD = 0.1 A → 0.4 A → 0.1 A, TR = TF = 1 μs
VIN(LDO4) = 3.3 V VOUT(LDO4) = = 1 V
Figure 9-68. LDO4 Load Step Transient
capacitors, especially for the capacitor at the VOUT_LDOVINT pin, must have a low impedance of less than
2 mΩ to the ground (Thermal Pad) of the TPS6593-Q1. For the ground connection of this capacitor at the
VOUT_LDOVINT pin, use multiple vias (at least three) directly at the ground landing pad of the capacitor. See
illustration below:
Due to the overall small solution size, the thermal performance of the PCB layout is important. Many system-
dependent parameters, such as thermal coupling, airflow, added heat sinks and convection surfaces, and the
presence of other heat-generating components affect the power dissipation limits of a given component. Proper
PCB layout, focusing on thermal performance, results in lower die temperatures. Wide and thick power traces
come with the ability to sink dissipated heat. The capability to sink dissipated heat can be improved further
on multi-layer PCB designs with vias to different planes. Improved heat-sinking capability results in reduced
junction-to-ambient (RθJA) and junction-to-board (RθJB) thermal resistances and thereby reduces the device
junction temperature, TJ. TI strongly recommends to perform a careful system-level 2D or full 3D dynamic
thermal analysis at the beginning product design process, by using a thermal modeling analysis software.
Overall recommendation for the PCB is to use at least 12 layers with 60 to 90 mil thickness, and with following
weights for the Copper layers:
• 0.5oz for signal layers
• at least 1.5oz for top layer and other plane layers
A more complete list of layout recommendations can be found in the Schematic and layout checklist.
This example shows a top and bottom layout of the key power components and the crystal oscillator based on
the EVM. Most of the digital routing is neglected in this image, see the EVM design files EVM design files for
full details. The highest priority must be on the buck input capacitors, followed by the inductors, and the output
capacitor on the VOUT_LDOVINT pin. Ensure that there are sufficient vias for high current pathways.
10.8 Glossary
TI Glossary This glossary lists and explains terms, acronyms, and definitions.
www.ti.com 30-Jan-2024
PACKAGING INFORMATION
Orderable Device Status Package Type Package Pins Package Eco Plan Lead finish/ MSL Peak Temp Op Temp (°C) Device Marking Samples
(1) Drawing Qty (2) Ball material (3) (4/5)
(6)
TPS65930400RWERQ1 ACTIVE VQFNP RWE 56 2000 RoHS & Green NIPDAU | NIPDAUAG Level-3-260C-168 HR -40 to 125 TPS6593 Samples
0400-Q1
TPS65931211RWERQ1 ACTIVE VQFNP RWE 56 2000 RoHS & Green NIPDAU | NIPDAUAG Level-3-260C-168 HR -40 to 125 TPS6593 Samples
1211-Q1
TPS6593C3A0RWERQ1 ACTIVE VQFNP RWE 56 2000 RoHS & Green NIPDAU | NIPDAUAG Level-3-260C-168 HR -40 to 125 TPS6593 Samples
C3A0-Q1
(1)
The marketing status values are defined as follows:
ACTIVE: Product device recommended for new designs.
LIFEBUY: TI has announced that the device will be discontinued, and a lifetime-buy period is in effect.
NRND: Not recommended for new designs. Device is in production to support existing customers, but TI does not recommend using this part in a new design.
PREVIEW: Device has been announced but is not in production. Samples may or may not be available.
OBSOLETE: TI has discontinued the production of the device.
(2)
RoHS: TI defines "RoHS" to mean semiconductor products that are compliant with the current EU RoHS requirements for all 10 RoHS substances, including the requirement that RoHS substance
do not exceed 0.1% by weight in homogeneous materials. Where designed to be soldered at high temperatures, "RoHS" products are suitable for use in specified lead-free processes. TI may
reference these types of products as "Pb-Free".
RoHS Exempt: TI defines "RoHS Exempt" to mean products that contain lead but are compliant with EU RoHS pursuant to a specific EU RoHS exemption.
Green: TI defines "Green" to mean the content of Chlorine (Cl) and Bromine (Br) based flame retardants meet JS709B low halogen requirements of <=1000ppm threshold. Antimony trioxide based
flame retardants must also meet the <=1000ppm threshold requirement.
(3)
MSL, Peak Temp. - The Moisture Sensitivity Level rating according to the JEDEC industry standard classifications, and peak solder temperature.
(4)
There may be additional marking, which relates to the logo, the lot trace code information, or the environmental category on the device.
(5)
Multiple Device Markings will be inside parentheses. Only one Device Marking contained in parentheses and separated by a "~" will appear on a device. If a line is indented then it is a continuation
of the previous line and the two combined represent the entire Device Marking for that device.
(6)
Lead finish/Ball material - Orderable Devices may have multiple material finish options. Finish options are separated by a vertical ruled line. Lead finish/Ball material values may wrap to two
lines if the finish value exceeds the maximum column width.
Important Information and Disclaimer:The information provided on this page represents TI's knowledge and belief as of the date that it is provided. TI bases its knowledge and belief on information
provided by third parties, and makes no representation or warranty as to the accuracy of such information. Efforts are underway to better integrate information from third parties. TI has taken and
continues to take reasonable steps to provide representative and accurate information but may not have conducted destructive testing or chemical analysis on incoming materials and chemicals.
TI and TI suppliers consider certain information to be proprietary, and thus CAS numbers and other limited information may not be available for release.
Addendum-Page 1
PACKAGE OPTION ADDENDUM
www.ti.com 30-Jan-2024
In no event shall TI's liability arising out of such information exceed the total purchase price of the TI part(s) at issue in this document sold by TI to Customer on an annual basis.
Addendum-Page 2
PACKAGE MATERIALS INFORMATION
www.ti.com 28-Mar-2024
B0 W
Reel
Diameter
Cavity A0
A0 Dimension designed to accommodate the component width
B0 Dimension designed to accommodate the component length
K0 Dimension designed to accommodate the component thickness
W Overall width of the carrier tape
P1 Pitch between successive cavity centers
Sprocket Holes
Q1 Q2 Q1 Q2
Pocket Quadrants
Pack Materials-Page 1
PACKAGE MATERIALS INFORMATION
www.ti.com 28-Mar-2024
Width (mm)
H
W
Pack Materials-Page 2
GENERIC PACKAGE VIEW
RWE 56 VQFNP - 0.9 mm max height
8 x 8, 0.5 mm pitch PLASTIC QUAD FLATPACK - NO LEAD
Images above are just a representation of the package family, actual package may vary.
Refer to the product data sheet for package details.
4224587/A
www.ti.com
PACKAGE OUTLINE
RWE0056C SCALE 2.000
VQFNP - 0.9 mm max height
PLASTIC QUAD FLATPACK - NO LEAD
8.1 A
B
7.9
0.05 (0.1)
0.00
PIN 1 ID
DETAIL A
8.1 TYPICAL
7.9
DETAIL A
SCALE 20.000
( 7.75)
(0.15)
0.15 0.1
DETAIL B
DETAIL B
SCALE 20.000
0.9 TYPICAL
12 MAX C
0.8
SEATING PLANE
(0.2) 0.08 C
SEE DETAIL A (R0.2)
SEE DETAIL B
4X
45 X 0.6 MAX
15 28
14
29 PIN 1 ID
OPTIONAL
SYMM 57
4X 5.5 0.05
6.5
1 42
52X 0.5 0.3
56 56X
43 0.2
SYMM
PIN 1 ID 0.5 0.1 C B A
OPTIONAL 56X 0.05 C
0.3
4224586/B 03/2021
NOTES:
1. All linear dimensions are in millimeters. Any dimensions in parenthesis are for reference only. Dimensioning and tolerancing
per ASME Y14.5M.
2. This drawing is subject to change without notice.
3. The package thermal pad must be soldered to the printed circuit board for thermal and mechanical performance.
www.ti.com
EXAMPLE BOARD LAYOUT
RWE0056C VQFNP - 0.9 mm max height
PLASTIC QUAD FLATPACK - NO LEAD
( 5.5)
SYMM
56X (0.6)
56 43
56X (0.25)
1
42
52X (0.5)
(7.8)
SYMM 57
(1.32) TYP
(R0.05) (2.5)
TYP TYP
( 0.2) TYP
VIA
29
14
15 28
(1.32)
TYP
(2.5)
TYP
(7.8)
NOTES: (continued)
4. This package is designed to be soldered to a thermal pad on the board. For more information, see Texas Instruments literature
number SLUA271 (www.ti.com/lit/slua271).
5. Vias are optional depending on application, refer to device data sheet. If any vias are implemented, refer to their locations shown
on this view. It is recommended that vias under paste be filled, plugged or tented.
www.ti.com
EXAMPLE STENCIL DESIGN
RWE0056C VQFNP - 0.9 mm max height
PLASTIC QUAD FLATPACK - NO LEAD
(7.8)
52X (0.5)
(1.32) TYP
(7.8)
METAL
TYP
16X
( 1.12)
(R0.05) TYP
14 29
15 28
SYMM
4224586/B 03/2021
NOTES: (continued)
6. Laser cutting apertures with trapezoidal walls and rounded corners may offer better paste release. IPC-7525 may have alternate
design recommendations.
www.ti.com
IMPORTANT NOTICE AND DISCLAIMER
TI PROVIDES TECHNICAL AND RELIABILITY DATA (INCLUDING DATA SHEETS), DESIGN RESOURCES (INCLUDING REFERENCE
DESIGNS), APPLICATION OR OTHER DESIGN ADVICE, WEB TOOLS, SAFETY INFORMATION, AND OTHER RESOURCES “AS IS”
AND WITH ALL FAULTS, AND DISCLAIMS ALL WARRANTIES, EXPRESS AND IMPLIED, INCLUDING WITHOUT LIMITATION ANY
IMPLIED WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE OR NON-INFRINGEMENT OF THIRD
PARTY INTELLECTUAL PROPERTY RIGHTS.
These resources are intended for skilled developers designing with TI products. You are solely responsible for (1) selecting the appropriate
TI products for your application, (2) designing, validating and testing your application, and (3) ensuring your application meets applicable
standards, and any other safety, security, regulatory or other requirements.
These resources are subject to change without notice. TI grants you permission to use these resources only for development of an
application that uses the TI products described in the resource. Other reproduction and display of these resources is prohibited. No license
is granted to any other TI intellectual property right or to any third party intellectual property right. TI disclaims responsibility for, and you
will fully indemnify TI and its representatives against, any claims, damages, costs, losses, and liabilities arising out of your use of these
resources.
TI’s products are provided subject to TI’s Terms of Sale or other applicable terms available either on ti.com or provided in conjunction with
such TI products. TI’s provision of these resources does not expand or otherwise alter TI’s applicable warranties or warranty disclaimers for
TI products.
TI objects to and rejects any additional or different terms you may have proposed. IMPORTANT NOTICE
Mailing Address: Texas Instruments, Post Office Box 655303, Dallas, Texas 75265
Copyright © 2024, Texas Instruments Incorporated