The JN5148-001 is an ultra low power, high performance wireless
microcontroller targeted at ZigBee PRO networking applications. The device
features an enhanced 32-bit RISC processor offering high coding efficiency
through variable width instructions, a multi-stage instruction pipeline and low
power operation with programmable clock speeds. It also includes a 2.4GHz
IEEE802.15.4 compliant transceiver, 128kB of ROM, 128kB of RAM, and a
rich mix of analogue and digital peripherals. The large memory footprint
allows the device to run both a network stack, e.g. ZigBee PRO, and an
embedded application or in a co-processor mode. The operating current is
below 18mA, allowing operation direct from a coin cell.
Enhanced peripherals include low power pulse counters running in sleep
mode designed for pulse counting in AMR applications and a unique Time
of Flight ranging engine, allowing accurate location services to be
implemented on wireless sensor networks. It also includes a 4-wire I
audio interface, to interface directly to mainstream audio CODECs, as well
as conventional MCU peripherals.
Block Diagram
RAM
Time of Flight
Engine
2.4GHz
Radio
XTAL
Watchdog
Timer
Power
Management
O-QPSK
Modem
IEEE802.15.4
MAC
Acceler ator
128-bi t AES
Encryption
Acceler ator
Benefits
• Single chip integrates
transceiver and
microcontroller for wireless
sensor networks
• Large memory footprint to
run ZigBee PRO together
with an application
• Very low current solution for
long battery life
• Highly featured 32-bit RISC
CPU for high performance
and low power
• System BOM is low in
component count and cost
• Extensive user peripherals
128kB
32-bit
RISC CPU
32-byte
OTP eFuse
ROM
128kB
SPI
2-Wir e Serial
Timer s
UARTs
4-Wire Audio
Sleep Counters
12-bit ADC,
Comparators
12-bit DACs,
Temp Se n s or
Applications
• Robust and secure low power
wireless applications
• ZigBee PRO networks
• Smart metering
(e.g. AMR)
• Home and commercial building
automation
• Location Aware services – e.g.
Asset Tracking
• Industrial systems
• Telemetry
• Remote Control
• Toys and gaming peripherals
Features: Transceiver
• 2.4GHz IEEE802.15.4 compliant
• Time of Flight ranging engine
• 128-bit AES security processor
• MAC accelerator with packet
• 500 & 667kbps data rate modes
• Integrated sleep oscillator for low
• On chip power regulation for 2.0V
• Deep sleep current 100nA
2
• Sleep current with active sleep
S
• <$0.50 external component cost
• Rx current 17.5mA
• Tx current 15.0mA
• Receiver sensitivity -95dBm
• Transmit power 2.5dBm
Features: Microcontroller
• Low power 32-bit RISC CPU, 4 to
• Variable instruction width for high
• Multi-stage instruction pipeline
• 128kB ROM and 128kB RAM for
• JTAG debug interface
• 4-input 12-bit ADC, 2 12-bit
• 3 application timer/counters,
• 2 UARTs
• SPI port with 5 selects
• 2-wire serial interface
• 4-wire digital audio interface
• Watchdog timer
• Low power pulse counters
• Up to 21 DIO
Industrial temp (-40°C to +85°C)
8x8mm 56-lead Punched QFN
The JN5148-001 is an IEEE802.15.4 wireless microcontroller that provides a fully integrated solution for applications
using the IEEE802.15.4 standard in the 2.4 - 2.5GHz ISM frequency band
of the functionality required to meet the IEEE802.15.4 and ZigBee PRO specifications and has additional processor
capability to run a wide range of applications including, but not limited to Smart Energy, Automatic Meter Reading,
Remote Control, Home and Building Automation, Toys and Gaming.
Applications that transfer data wirelessly tend to be more complex than wired ones. Wireless protocols make
stringent demands on frequencies, data formats, timing of data transfers, security and other issues. Application
development must consider the requirements of the wireless network in addition to the product functionality and user
interfaces. To minimise this complexity, Jennic provides a series of software libraries and interfaces that control the
transceiver and peripherals of the JN5148. These libraries and interfaces remove the need for the developer to
understand wireless protocols and greatly simplifies the programming complexities of power modes, interrupts and
hardware functionality.
In view of the above, the register details of the JN5148 are not provided in the datasheet.
The device includes a Wireless Transceiver, RISC CPU, on chip memory and an extensive range of peripherals.
Hereafter, the JN5148-001 will be referred to as JN5148.
1.1 Wireless Transceiver
The Wireless Transceiver comprises a 2.45GHz radio, a modem, a baseband controller and a security coprocessor.
In addition, the radio also provides an output to control transmit-receive switching of external devices such as power
amplifiers allowing applications that require increased transmit power to be realised very easily. Appendix B.4,
describes a complete reference design including Printed Circuit Board (PCB) design and Bill Of Materials (BOM).
The security coprocessor provides hardware-based 128-bit AES-CCM, CBC
specified by the IEEE802.15.4 standard. It does this in-band on packets during transmission and reception, requiring
minimal intervention from the CPU. It is also available for off-line use under software control for encrypting and
decrypting packets generated by software layers such as ZigBee and user applications. This means that these
algorithms can be off-loaded by the CPU, increasing the processor bandwidth available for user applications.
The transceiver elements (radio, modem and baseband) work together to provide IEEE802.15.4 Medium Access
Control (MAC) under the control of a protocol stack. Applications incorporating IEEE802.15.4 functionality can be
rapidly developed by combining user-developed application software with a protocol stack library.
(1) AES-CBC processing is only available off-line for use under software control.
[1], including ZigBee PRO. It includes all
(1)
, CTR and CCM* processing as
1.2 RISC CPU and Memory
A 32-bit RISC CPU allows software to be run on chip, its processing power being shared between the IEEE802.15.4
MAC protocol, other higher layer protocols and the user application. The JN5148 has a unified memory architecture,
code memory, data memory, peripheral devices and I/O ports are organised within the same linear address space.
The device contains 128kbytes of ROM, 128kbytes of RAM and a 32-byte One Time Programmable (OTP) eFuse
memory.
• Two UARTs with support for hardware or software flow control
• Three programmable Timer/Counters – all three support Pulse Width Modulation (PWM) capability, two have
capture/compare facility
• Two programmable Sleep Timers and a Tick Timer
2
• Two-wire serial interface (compatible with SMbus and I
• Four-wire digital audio interface (compatible with I²S)
• Slave SPI port for Intelligent peripheral mode (shared with digital I/O)
• Twenty-one digital I/O lines (multiplexed with peripherals such as timers and UARTs)
• Four channel, 12-bit, Analogue to Digital converter
• Two 12-bit Digital to Analogue converters
• Two programmable analogue comparators
• Internal temperature sensor and battery monitor
• Time Of Flight ranging engine
• Two low power pulse counters
• Random number generator
• Watchdog Timer and Voltage Brown-out
• Sample FIFO for digital audio interface or ADC/DAC
• JTAG hardware debug port
User applications access the peripherals using the Integrated Peripherals API. This allows applications to use a
tested and easily understood view of the peripherals allowing rapid system development.
5 DIO19 TXD1 JTAG_TDO CMOS DIO19 or UART 1 Transmit
31 DIO 20 RXD1
(default)
Digital Peripheral I/O Pin
IP_DI JTAG_TDI
(optional)
JTAG_TDI
(default)
Signal
Type
32K External Crystal Output
Input or Timer2 PWM Output
Antenna Diversity or Digital
Audio Word Select
Antenna Diversity or Digital
Audio Data Input
or Intelligent Peripheral Clock
Input
Intelligent Peripheral Data Out
CMOS DIO16, UART 1 Receive Data
Input or Intelligent Peripheral
Data In or JTAG data In
(optional)
Input, Intelligent Peripheral
Device Select Input or Digital
Audio Clock or JTAG CLK
Send Output, Intelligent
Peripheral Interrupt Output or
Digital Audio Data Output or
JTAG Mode Select
Data Output or JTAG Data Out
CMOS DIO 20, UART 1 Receive Data
Input or JTAG data In (default)
Description
The PCB schematic and layout rules detailed in Appendix B.4
must be followed. Failure to do so will likely result in the
JN5148 failing to meet the performance specification detailed
herein and worst case may result in device not functioning in
the end application.
The device is powered from the VDD1 and VDD2 pins, each being decoupled with a 100nF ceramic capacitor. VDD1
is the power supply to the analogue circuitry; it should be decoupled to ground. VDD2 is the power supply for the
digital circuitry; and should also be decoupled to ground. A 10uF tantalum capacitor is required. Decoupling pins for
the internal 1.8V regulators are provided which require a 100nF capacitor located as close to the device as practical.
VB_VCO, VB_RF, VB_A and VB_SYNTH should be decoupled with an additional 47pF capacitor, while VB_RAM
and VB_DIG require only 100nF. VB_RF and VB_RF2 should be connected together as close to the device as
practical, and only require one 100nF capacitor and one 47pF capacitor. Refer to
VSSA, VSSS, VSS1, VSS2, VSS3 are the ground pins.
2.2.2 Reset
RESETN is a bi-directional active low reset pin that is connected to a 40kΩ internal pull-up resistor. It may be pulled
low by an external circuit, or can be driven low by the JN5148 if an internal reset is generated. Typically, it will be
used to provide a system reset signal. Refer to section 6.2, External Reset, for more details.
2.2.3 32MHz Oscillator
A crystal is connected between XTALIN and XTALOUT to form the reference oscillator, which drives the system
clock. A capacitor to analogue ground is required on each of these pins. Refer to section
for more details. The 32MHz reference frequency is divided down to 16MHz and this is used as the system clock
throughout the device.
B.4.1 for schematic diagram.
5.1 16MHz System Clock
2.2.4 Radio
The radio is a single ended design, requiring a capacitor and just two inductors to match to 50Ω microstrip line to the
RF_IN pin.
An external resistor (43kΩ) is required between IBIAS and analogue ground to set various bias currents and
references within the radio.
Several of the analogue peripherals require a reference voltage to use as part of their operations. They can use
either an internal reference voltage or an external reference connected to VREF. This voltage is referenced to
analogue ground and the performance of the analogue peripherals is dependant on the quality of this reference.
There are four ADC inputs, two pairs of comparator inputs and two DAC outputs. The analogue I/O pins on the
JN5148 can have signals applied up to 0.3v higher than VDD1. A schematic view of the analogue I/O cell is shown in
Figure 3: Analogue I/O Cell
In reset and deep sleep, the analogue peripherals are all off and the DAC outputs are in a high impedance state.
In sleep, the ADC and DACs are off, with the DAC outputs in high impedance state. The comparators may optionally
be used as a wakeup source.
Unused ADC and comparator inputs should be left unconnected.
VDD1
Analogue
Analogue
Peripheral
I/O Pin
VSSA
Figure 3: Analogue I/O Cell
2.2.6 Digital Input/Output
Digital I/O pins on the JN5148 can have signals applied up to 2V higher than VDD2 (with the exception of pins DIO9
and DIO10 that are 3V tolerant) and are therefore TTL-compatible with VDD2 > 3V. For other DC properties of these
pins see section 22.2.3 I/O Characteristics.
When used in their primary function all Digital Input/Output pins are bi-directional and are connected to weak internal
pull up resistors (40kΩ nominal) that can be disabled. When used in their secondary function (selected when the
appropriate peripheral block is enabled through software library calls) then their direction is fixed by the function. The
pull up resistor is enabled or disabled independently of the function and direction; the default state from reset is
enabled.
In reset, the digital peripherals are all off and the DIO pins are set as high-impedance inputs. During sleep and deep
sleep, the DIO pins retain both their input/output state and output level that was set as sleep commences. If the DIO
pins were enabled as inputs and the interrupts were enabled then these pins may be used to wake up the JN5148
from sleep.
The CPU of the JN5148 is a 32-bit load and store RISC processor.It has been architected for three key
requirements:
• Low power consumption for battery powered applications
• High performance to implement a wireless protocol at the same time as complex applications
• Efficient coding of high-level languages such as C provided with the Jennic Software Developers Kit
It features a linear 32-bit logical address space with unified memory architecture, accessing both code and data in the
same address space. Registers for peripheral units, such as the timers, UARTs and the baseband processor are
also mapped into this space.
The CPU has access to a block of 15 32-bit General-Purpose (GP) registers together with a small number of special
purpose registers which are used to store processor state and control interrupt handling. The contents of any GP
register can be loaded from or stored to memory, while arithmetic and logical operations, shift and rotate operations,
and signed and unsigned comparisons can be performed either between two registers and stored in a third, or
between registers and a constant carried in the instruction. Operations between general or special-purpose registers
execute in one cycle while those that access memory require a further cycle to allow the memory to respond.
The instruction set manipulates 8, 16 and 32-bit data; this means that programs can use objects of these sizes very
efficiently. Manipulation of 32-bit quantities is particularly useful for protocols and high-end applications allowing
algorithms to be implemented in fewer instructions than on smaller word-size processors, and to execute in fewer
clock cycles. In addition, the CPU supports hardware Multiply that can be used to efficiently implement algorithms
needed by Digital Signal Processing applications.
The instruction set is designed for the efficient implementation of high-level languages such as C. Access to fields in
complex data structures is very efficient due to the provision of several addressing modes, together with the ability to
be able to use any of the GP registers to contain the address of objects. Subroutine parameter passing is also made
more efficient by using GP registers rather than pushing objects onto the stack. The recommended programming
method for the JN5148 is by using C, which is supported by a software developer kit comprising a C compiler, linker
and debugger.
The CPU architecture also contains features that make the processor suitable for embedded, real-time applications.
In some applications, it may be necessary to use a real-time operating system to allow multiple tasks to run on the
processor. To provide protection for device-wide resources being altered by one task and affecting another, the
processor can run in either supervisor or user mode, the former allowing access to all processor registers, while the
latter only allows the GP registers to be manipulated. Supervisor mode is entered on reset or interrupt; tasks starting
up would normally run in user mode in a RTOS environment.
Embedded applications require efficient handling of external hardware events. Exception processing (including reset
and interrupt handling) is enhanced by the inclusion of a number of special-purpose registers into which the PC and
status register contents are copied as part of the operation of the exception hardware. This means that the essential
registers for exception handling are stored in one cycle, rather than the slower method of pushing them onto the
processor stack. The PC is also loaded with the vector address for the exception that occurred, allowing the handler
to start executing in the next cycle.
To improve power consumption a number of power-saving modes are implemented in the JN5148, described more
fully in section
software control, the processor can be shut down and on an interrupt it will wake up to service the request.
Additionally, it is possible under software control, to set the speed of the CPU to 4, 8, 16 or 32MHz. This feature can
be used to trade-off processing power against current consumption.
21 - Power Management and Sleep Modes. One of these modes is the CPU doze mode; under
This section describes the different memories found within the JN5148. The device contains ROM, RAM, OTP eFuse
memory, the wireless transceiver and peripherals all within the same linear address space.
0xFFFFFFFF
0xF0020000
RAM
(128kB)
0xF0000000
Unpopulated
0x04000000
0x02000000
0x00020000
0x00000000
RAM Ec ho
Peripherals
ROM
(128kB)
Figure 5: JN5148 Memory Map
4.1 ROM
The ROM is 128k bytes in size, and can be accessed by the processor in a single CPU clock cycle. The ROM
contents include bootloader to allow external Flash memory contents to be bootloaded into RAM at runtime, a default
interrupt vector table, an interrupt manager, IEEE802.15.4 MAC and APIs for interfacing on-chip peripherals. The
operation of the boot loader is described in detail in Application Note
to the application’s soft interrupt vector table contained within RAM. Section
the handling of interrupts. ROM contents are shown in
The JN5148 contains 128kBytes of high speed RAM. It can be used for both code and data storage and is accessed
by the CPU in a single clock cycle. At reset, a boot loader controls the loading of segments of code and data from an
external memory connected to the SPI port, into RAM. Software can control the power supply to the RAM allowing
the contents to be maintained during a sleep period when other parts of the device are un-powered. Typical RAM
contents are shown in Figure 7.
0x04020000
CPU Stack
(Grows D own)
Application
MAC Address
MAC Data
Interrupt Vector Table
0x04000000
Figure 7: Typical RAM Contents
4.3 OTP eFuse Memory
The JN5148 contains a total of 32bytes of eFuse memory; this is a One Time Programmable (OTP) memory that is
used to support on chip 64-bit MAC ID and a 128-bit AES security key. A limited number of bits are available for
customer use for storage of configuration information; configuration of these is made through use of software APIs.
For full details on how to program and use the eFuse memory, please refer to application note
OTP efuse Memory
Alternatively, Jennic can provide an eFuse programming service for customers that wish to use the eFuse but do not
wish to undertake this for themselves. For further details of this service, please contact your local Jennic sales office.
[2]
JN-AN-1062 Using
4.4 External Memory
An external memory with an SPI interface may be used to provide storage for program code and data for the device
when external power is removed. The memory is connected to the SPI interface using select line SPISEL0; this
select line is dedicated to the external memory interface and is not available for use with other external devices. See
Figure 8 for connection details.
At reset, the contents of this memory are copied into RAM by the software boot loader. The Flash memory devices
that are supported as standard through the JN5148 bootloader are given in Table 1. Jennic recommends that where
possible one of these devices should be selected.
Applications wishing to use an alternate Flash memory device should refer to application note [3] JN-AN-1038
Programming Flash devices not supported by the JN51xx ROM-based bootloader. This application note provides
guidance on developing an interface to an alternate device.
4.4.1 External Memory Encryption
The contents of the external serial memory may be encrypted. The AES security processor combined with a user
programmable 128-bit encryption key is used to encrypt the contents of the external memory. The encryption key is
stored in eFuse.
When bootloading program code from external serial memory, the JN5148 automatically accesses the encryption key
to execute the decryption process. User program code does not need to handle any of the decryption process; it is
transparent.
With encryption enabled, the time taken to boot code from external flash is increased.
4.5 Peripherals
All peripherals have their registers mapped into the memory space. Access to these registers requires 3 clock
cycles. Applications have access to the peripherals through the software libraries that present a high-level view of
the peripheral’s functions through a series of dedicated software routines. These routines provide both a tested
method for using the peripherals and allow bug-free application code to be developed more rapidly. For details, see
the Integrated Peripherals API Reference Manual (JN-RM-2001).[6]
4.6 Unused Memory Addresses
Any attempt to access an unpopulated memory area will result in a bus error exception (interrupt) being generated.
Two system clocks are used to provide timing references into the on-chip subsystems of the JN5148. A 16MHz clock,
generated by a crystal-controlled 32MHz oscillator, is used by the transceiver, processor, memory and digital and
analogue peripherals. A 32kHz clock is used by the sleep timer and during the startup phase of the chip.
5.1 16MHz System Clock
The 16MHz system clock is used by the digital and analogue peripherals and the transceiver. A scaled version
(4,8,16 or 32MHz) of this clock is also used by the processor and memories. For most operations it is necessary to
source this clock from the 32MHz oscillator.
Crystal oscillators are generally slow to start. Hence to provide a faster start-up following a sleep cycle a fast RC
oscillator is provided that can be used as the source for the 16MHz system clock. The oscillator starts very quickly
and is typically 24MHz causing the system clock to run at 12MHz. Using a clock of this speed scales down the speed
of the processor and any peripherals in use. For the SPI interface this causes no functional issues as the generated
SPI clock is slightly slower and is used to clock the external SPI slave. Use of the radio is not possible when using the
24MHz RC oscillator. Additionally, timers and UARTs should not be used as the exact frequency will not be known.
The JN5148 device can be configured to wake up from sleep using the fast RC oscillator and automatically switch
over to use the 32MHz xtal as the clock source, when it has started up. This could allow application code to be
downloaded from the flash before the xtal is ready, typically improving start-up time by 550usec. Alternatively, the
switch over can be controlled by software, or the system could always use the 32MHz oscillator as the clock source.
5.1.1 32MHz Oscillator
The JN5148 contains the necessary on chip components to build a 32MHz reference oscillator with the addition of an
external crystal resonator and two tuning capacitors. The schematic of these components are shown in
The two capacitors, C1 and C2, should typically be 15pF and use a COG dielectric. Due to the small size ofthese
capacitors, it is important to keep the traces to the external components as short as possible. The on chip
transconductance amplifier is compensated for temperature variation, and is self-biasing by means of the internal
resistor R1. The electrical specification of the oscillator can be found in section 22.3.13. Please refer to Appendix B
for development support with the crystal oscillator circuit.
Figure 9.
JN5148
XTALIN
Figure 9: 32MHz Crystal Oscillator Connections
R1
XTALOUT
C2 C1
5.1.2 24MHz RC Oscillator
An on-chip 24MHz RC oscillator is provided. No external components are required for this oscillator. The electrical
specification of the oscillator can be found in section
The 32kHz system clock is used for timing the length of a sleep period (see section 21 Power Management and
Sleep Modes) and also to generate the system clock used internally during reset. The clock can be selected from
one of three sources through the application software:
32kHz RC Oscillator
•
32kHz Crystal Oscillator
•
• 32kHz External Clock
Upon a chip reset or power-up the JN5148 defaults to using the internal 32kHz RC Oscillator. If another clock source
is selected then it will remain in use for all 32kHz timing until a chip reset is performed.
5.2.1 32kHz RC Oscillator
The internal 32kHz RC oscillator requires no external components. The internal timing components of the oscillator
have a wide tolerance due to manufacturing process variation and so the oscillator runs nominally at 32kHz ±30%. To
make this useful as a timing source for accurate wakeup from sleep, a frequency calibration factor derived from the
more accurate 16MHz clock may be applied. The calibration factor is derived through software, details can be found
in section 11.3.1. For detailed electrical specifications, see section 22.3.11.
5.2.2 32kHz Crystal Oscillator
In order to obtain more accurate sleep periods, the JN5148 contains the necessary on-chip components to build a
32kHz oscillator with the addition of an external 32.768kHz crystal and two tuning capacitors. The crystal should be
connected between 32KXTALIN and 32KXTALOUT (DIO9 and DIO10), with two equal capacitors to ground, one on
each pin. Due to the small size of the capacitors, it is important to keep the traces to the external components as
short as possible.
The electrical specification of the oscillator can be found in section 22.3.12. The oscillator cell is flexible and can
operate with a range of commonly available 32.768kHz crystals with load capacitances from 6 to 12.5pF. However,
the maximum ESR of the crystal and the supply current are both functions of the actual crystal used, see appendix
B.1 for more details.
JN5148
32KXTALOUT 32KXTALIN
Figure 10: 32kHz crystal oscillator connections
5.2.3 32kHz External Clock
An externally supplied 32kHz reference clock on the 32KIN input (DIO9) may be provided to the JN5148. This would
allow the 32kHz system clock to be sourced from a very stable external oscillator module, allowing more accurate
sleep cycle timings compared to the internal RC oscillator. (See section
tolerant input)
A system reset initialises the device to a pre-defined state and forces the CPU to start program execution from the
reset vector. The reset process that the JN5148 goes through is as follows.
When power is applied, the 32kHz RC oscillator starts up and stabilises, which takes approximately 100μsec. At this
point, the 32MHz crystal oscillator is enabled and power is applied to the processor and peripheral logic. The logic
blocks are held in reset until the 32MHz crystal oscillator stabilises, typically this takes 0.75ms. Then the internal
reset is removed from the CPU and peripheral logic and the CPU starts to run code beginning at the reset vector,
consisting of initialisation code and the resident boot loader. [8] Section 22.3.1 provides detailed electrical data and
timing.
The JN5148 has five sources of reset:
• Internal Power-on Reset
• External Reset
• Software Reset
• Watchdog timer
• Brown-out detect
Note: When the device exits a reset condition, device operating
parameters (voltage, frequency, temperature, etc.) must be met to ensure
operation. If these conditions are not met, then the device must be held in
reset until the operating conditions are met. (See section 22.3)
6.1 Internal Power-on Reset
For the majority of applications the internal power-on reset is capable of generating the required reset signal. When
power is applied to the device, the power-on reset circuit monitors the rise of the VDD supply. When the VDD
reaches the specified threshold, the reset signal is generated and can be observed as a rising edge on the RESETN
pin. This signal is held internally until the power supply and oscillator stabilisation time has elapsed, when the internal
reset signal is then removed and the CPU is allowed to run.
VDD
Internal RESET
RESETN Pin
Figure 11: Internal Power-on Reset
When the supply drops below the power on reset ‘falling’ threshold, it will re-trigger the reset. Use of the external
reset circuit show in
The external resistor and capacitor provide a simple reset operation when connected to the RESETN pin.
470nF
6.2 External Reset
An external reset is generated by a low level on the RESETN pin. Reset pulses longer than the minimum pulse width
will generate a reset during active or sleep modes. Shorter pulses are not guaranteed to generate a reset. The
JN5148 is held in reset while the RESETN pin is low. When the applied signal reaches the Reset Threshold Voltage
) on its positive edge, the internal reset process starts.
(V
RST
Multiple devices may connect to the RESETN pin in an open-collector mode. The JN5148 has an internal pull-up
resistor connect to the RESETN pin. The pin is an input for an external reset, an output during the power-on reset
and may optionally be an output during a software reset. No devices should drive the RESETN pin high.
RESETN pin
Reset
Internal Reset
Figure 13: External Reset
6.3 Software Reset
A system reset can be triggered at any time through software control, causing a full chip reset and invalidating the
RAM contents. For example this can be executed within a user’s application upon detection of a system failure. When
performing the reset, the RESETN pin is driven low for 1µsec; depending on the external components this may or
may not be visible on the pin.
In addition, the RESETN line can be driven low by the JN5148 to provide a reset to other devices in the system (e.g.
external sensors) without resetting itself. When the RESETN line is not driven it will pull back high through either the
internal pull-up resistor or any external circuitry. It is essential to ensure that the RESETN line pulls back high within
100µsec after the JN5148 stops driving the line; otherwise a system reset will occur. Due to this, careful consideration
should be taken of any capacitance on this line. For instance, the RC values recommended in section
to be replaced with a suitable reset IC
An internal brown-out detect module is used to monitor the supply voltage to the JN5148; this can be used whilst the
device is awake or is in CPU doze mode. Dips in the supply voltage below a variable threshold can be detected and
can be used to cause the JN5148 to perform a chip reset. Equally, dips in the supply voltage can be detected and
used to cause an interrupt to the processor, when the voltage either drops below the threshold or rises above it.
The brown-out detect is enabled by default from power-up and can extend the reset during power-up. This will keep
the CPU in reset until the voltage exceeds the brown-out threshold voltage. The threshold voltage is configurable to
2.0V, 2.3V, 2.7V and 3.0V and is controllable by software. From power-up the threshold is set by eFuse settings and
the default chip configuration is for the 2.3V threshold. It is recommended that the threshold is set so that, as a
minimum, the chip is held in reset until the voltage reaches the level required by the external memory device on the
SPI interface.
6.5 Watchdog Timer
A watchdog timer is provided to guard against software lockups. It operates by counting cycles of the 32kHz system
clock. A pre-scaler is provided to allow the expiry period to be set between typically 8ms and 16.4 seconds. Failure
to restart the watchdog timer within the pre-configured timer period will cause a chip reset to be performed. A status
bit is set if the watchdog was triggered so that the software can differentiate watchdog initiated resets from other
resets, and can perform any required recovery once it restarts. If the source of the 32kHz system clock is the 32kHz
RC oscillator then the watchdog expiry periods are subject to the variation in period of the RC oscillator.
After power up, reset, start from deep sleep or start from sleep, the watchdog is always enabled with the largest
timeout period and will commence counting as if it had just been restarted. Under software control the watchdog can
be disabled. If it is enabled, the user must regularly restart the watchdog timer to stop it from expiring and causing a
reset. The watchdog runs continuously, even during doze, however the watchdog does not operate during sleep or
deep sleep, or when the hardware debugger has taken control of the CPU. It will recommence automatically if
enabled once the debugger un-stalls the CPU.
The interrupt system on the JN5148 is a hardware-vectored interrupt system. The JN5148 provides several interrupt
sources, some associated with CPU operations (CPU exceptions) and others which are used by hardware in the
device. When an interrupt occurs, the CPU stops executing the current program and loads its program counter with a
fixed hardware address specific to that interrupt. The interrupt handler or interrupt service routine is stored at this
location and is run on the next CPU cycle. Execution of interrupt service routines is always performed in supervisor
mode. Interrupt sources and their vector locations are listed in
Bus error 0x08 Typically cause by an attempt to access an invalid address or a
disabled peripheral
Tick timer 0x0e Tick timer interrupt asserted
Alignment error 0x14 Load/store address to non-naturally-aligned location
Illegal instruction 0x1a Attempt to execute an unrecognised instruction
Hardware interrupt 0x20 interrupt asserted
System call 0x26
Trap 0x2c caused by the b.trap instruction or the debug unit
Reset 0x38 Caused by software or hardware reset.
Stack Overflow 0x3e Stack overflow
System call initiated by b.sys instruction
Table 2: Interrupt Vectors
Table 2 below:
7.1 System Calls
The b.trap and b.sys instructions allow processor exceptions to be generated by software.
A system call exception will be generated when the b.sys instruction is executed. This exception can, for example, be
used to enable a task to switch the processor into supervisor mode when a real time operating system is in use. (See
section 3 for further details.)
The b.trap instruction is commonly used for trapping errors and for debugging.
7.2 Processor Exceptions
7.2.1 Bus Error
A bus error exception is generated when software attempts to access a memory address that does not exist, or is not
populated with memory or peripheral registers or when writing to ROM.
7.2.2 Alignment
Alignment exceptions are generated when software attempts to access objects that are not aligned to natural word
boundaries. 16-bit objects must be stored on even byte boundaries, while 32-bit objects must be stored on quad byte
boundaries. For instance, attempting to read a 16-bit object from address 0xFFF1 will trigger an alignment exception
as will a read of a 32-bit object from 0xFFF1, 0xFFF2 or 0xFFF3. Examples of legal 32-bit object addresses are
0xFFF0, 0xFFF4, 0xFFF8 etc.
7.2.3 Illegal Instruction
If the CPU reads an unrecognised instruction from memory as part of its instruction fetch, it will cause an illegal
instruction exception.
7.2.4 Stack Overflow
When enabled, a stack overflow exception occurs if the stack pointer reaches a programmable location.
Hardware interrupts generated from the transceiver, analogue or digital peripherals and DIO pins are individually
masked using the Programmable Interrupt Controller (PIC). Management of interrupts is provided in the peripherals
[6]. For details of the interrupts generated from each peripheral see the respective section in this datasheet.
library
Interrupts can be used to wake the JN5148 from sleep. The peripherals, baseband controller, security coprocessor
and PIC are powered down during sleep but the DIO interrupts and optionally the pulse counters, wake-up timers and
analogue comparator interrupts remain powered to bring the JN5148 out of sleep.
Prioritised external interrupt handling (i.e., interrupts from hardware peripherals) is provided to enable an application
to control an events priority to provide for deterministic program execution.
The priority Interrupt controller provides 15 levels of prioritised interrupts. The priority level of all interrupts can be set,
with value 0 being used to indicate that the source can never produce an external interrupt, 1 for the lowest priority
source(s) and 15 for the highest priority source(s). Note that multiple interrupt sources can be assigned the same
priority level if desired.
If while processing an interrupt, a new event occurs at the same or lower priority level, a new external interrupt will
not be triggered. However, if a new higher priority event occurs, the external interrupt will again be asserted,
interrupting the current interrupt service routine.
Once the interrupt service routine is complete, lower priority events can be serviced.
The wireless transceiver comprises a 2.45GHz radio, modem, a baseband processor, a security coprocessor and
PHY controller. These blocks, with protocol software provided as a library, implement an IEEE802.15.4 standardsbased wireless transceiver that transmits and receives data over the air in the unlicensed 2.4GHz band.
8.1 Radio
Figure 14 shows the single ended radio architecture.
Radio
LNA
Lim1
Lim2
Lim3
Lim4
D-Type
Switch
Calibration
Reference
& Bias
ADC
PA
sigma
synth
Figure 14: Radio Architecture
The radio comprises a low-IF receive path and a direct modulation transmit path, which converge at the TX/RX
switch. The switch connects to the external single ended matching network, which consists of two inductors and a
capacitor, this arrangement creates a 50Ω port and removes the need for a balun. A 50Ω single ended antenna can
be connected directly to this port.
The 32MHz crystal oscillator feeds a divider, which provides the frequency synthesiser with a reference frequency.
The synthesiser contains programmable feedback dividers, phase detector, charge pump and internal Voltage
Controlled Oscillator (VCO). The VCO has no external components, and includes calibration circuitry to compensate
for differences in internal component values due to process and temperature variations. The VCO is controlled by a
Phase Locked Loop (PLL) that has an internal loop filter. A programmable charge pump is also used to tune the loop
characteristic.
The receiver chain starts with the low noise amplifier / mixer combination whose outputs are passed to a lowpass
filter, which provides the channel definition. The signal is then passed to a series of amplifier blocks forming a limiting
strip. The signal is converted to a digital signal before being passed to the Modem. The gain control for the RX path
is derived in the automatic gain control (AGC) block within the Modem, which samples the signal level at various
points down the RX chain. To improve the performance and reduce current consumption, automatic calibration is
applied to various blocks in the RX path.
In the transmit direction, the digital stream from the Modem is passed to a digital sigma-delta modulator which
controls the feedback dividers in the synthesiser, (dual point modulation). The VCO frequency now tracks the applied
modulation. The 2.4 GHz signal from the VCO is then passed to the RF Power Amplifier (PA), whose power control
can be selected from one of three settings. The output of the PA drives the antenna via the RX/TX switch
In order to realise the full performance of the radio it is essential that the reference PCB layout and BOM are carefully
followed. See Appendix
The radio is powered from a number of internal 1.8V regulators fed from the analogue supply VDD1, in order to
provide good noise isolation between the digital logic of the JN5148 and the analogue blocks. These regulators are
also controlled by the baseband controller and protocol software to minimise power consumption. Decoupling for
internal regulators is required as described in section
For single ended antennas or connectors, a balun is not required, however a matching network is needed.
The RF matching network requires three external components and the IBIAS pin requires one external component as
shown in schematic in B.4.1. These components are critical and should be placed close to the JN5148 pins and
analogue ground as defined in
Constraints
8.1.2 . Antenna Diversity
Support is provided for antenna diversity. Antenna diversity is a technique that maximises the performance of an
antenna system. It allows the radio to switch between two antennas that have very low correlation between their
received signals. Typically, this is achieved by spacing two antennas around 0.25 wavelengths apart or by using two
orthogonal polarisations. So, if a packet is transmitted and no acknowledgement is received, the radio system can
switch to the other antenna for the retry, with a different probability of success.
The JN5148 provides an output (ADO) on DIO12 that is asserted on odd numbered retries that can be used to control
an antenna switch; this enables antenna diversity to be implemented easily (see
Figure 16 Antenna Diversity ADO Signal for TX with Acknowledgement
DIO13 can be configured to be ADE, the inverse of ADO. In this configuration, an antenna diversity scheme can be
implemented without the need for the inverter shown in
The modem performs all the necessary modulation and spreading functions required for digital transmission and
reception of data at 250kbps in the 2450MHz radio frequency band in compliance with the IEEE802.15.4 standard. It
also provides a high data rate modes at 500 and 667kbps.
Features provided to support network channel selection algorithms include Energy Detection (ED), Link Quality
Indication (LQI) and fully programmable Clear Channel Assessment (CCA).
The Modem provides a digital Receive Signal Strength Indication (RSSI) that facilitates the implementation of the
IEEE 802.15.4 ED function.
The LQI is defined in the IEEE 802.15.4 standard as a characterisation of the strength and/or data quality of a
received packet. The Modem produces a signal quality metric based upon correlation magnitudes, which may be
used in conjunction with the ED value to formulate the LQI.
The CCA capability of the Modem supports all modes of operation defined in the IEEE 802.15.4 standard, namely
Energy above ED threshold, Carrier Sense and Carrier Sense and/or energy above ED threshold.
8.3 Baseband Processor
The baseband processor provides all time-critical functions of the IEEE802.15.4 MAC layer. Dedicated hardware
guarantees air interface timing is precise. The MAC layer hardware/software partitioning, enables software to
implement the sequencing of events required by the protocol and to schedule timed events with millisecond
resolution, and the hardware to implement specific events with microsecond timing resolution. The protocol software
layer performs the higher-layer aspects of the protocol, sending management and data messages between endpoint
and coordinator nodes, using the services provided by the baseband processor.
Tx
Bitstream
Append
Checksum
Supervisor
Serialiser
Status
Encrypt
Port
AES
AES
Codec
Codec
Tx/Rx
Frame
Buffer
Radio
Rx
Bitstream
Protocol Timing Engine
CSMACCA
Verify
Checksum
Backoff
Control
Control
Deserialiser
Inline
Security
Decrypt
Port
Protocol
Timers
Processor
Bus
Figure 17: Baseband Processor
8.3.1 Transmit
A transmission is performed by software writing the data to be transferred into the Tx/Rx Frame Buffer, together with
parameters such as the destination address and the number of retries allowed, and programming one of the protocol
timers to indicate the time at which the frame is to be sent. This time will be determined by the software tracking the
higher-layer aspects of the protocol such as superframe timing and slot boundaries. Once the packet is prepared and
protocol timer set, the supervisor block controls the transmission. When the scheduled time arrives, the supervisor
controls the sequencing of the radio and modem to perform the type of transmission required. It can perform all the
algorithms required by IEEE802.15.4 such as CSMA/CA, GTS without processor intervention including retries and
random backoffs.
When the transmission begins, the header of the frame is constructed from the parameters programmed by the
software and sent with the frame data through the serialiser to the Modem. At the same time, the radio is prepared
for transmission. During the passage of the bitstream to the modem, it passes through a CRC checksum generator
that calculates the checksum on-the-fly, and appends it to the end of the frame.
If using slotted access, it is possible for a transmission to overrun the time in its allocated slot; the Baseband
Processor handles this situation autonomously and notifies the protocol software via interrupt, rather than requiring it
to handle the overrun explicitly.
8.3.2 Reception
During reception, the radio is set to receive on a particular channel. On receipt of data from the modem, the frame is
directed into the Tx/Rx Frame Buffer where both header and frame data can be read by the protocol software. An
interrupt may be provided on receipt of the frame header. As the frame data is being received from the modem it is
passed through a checksum generator; at the end of the reception the checksum result is compared with the
checksum at the end of the message to ensure that the data has been received correctly. An interrupt may be
provided to indicate successful packet reception.
During reception, the modem determines the Link Quality, which is made available at the end of the reception as part
of the requirements of IEEE802.15.4.
8.3.3 Auto Acknowledge
Part of the protocol allows for transmitted frames to be acknowledged by the destination sending an acknowledge
packet within a very short window after the transmitted frame has been received. The JN5148 baseband processor
can automatically construct and send the acknowledgement packet without processor intervention and hence avoid
the protocol software being involved in time-critical processing within the acknowledge sequence. The JN5148
baseband processor can also request an acknowledge for packets being transmitted and handle the reception of
acknowledged packets without processor intervention.
8.3.4 Beacon Generation
In beaconing networks, the baseband processor can automatically generate and send beacon frames; the repetition
rate of the beacons is programmed by the CPU, and the baseband then constructs the beacon contents from data
delivered by the CPU. The baseband processor schedules the beacons and transmits them without CPU
intervention.
8.3.5 Security
The baseband processor supports the transmission and reception of secured frames using the Advanced Encryption
Standard (AES) algorithm transparently to the CPU. This is done by passing incoming and outgoing data through an
in-line security engine that is able to perform encryption and decryption operations on-the-fly, resulting in minimal
processor intervention. The CPU must provide the appropriate encrypt/decrypt keys for the transmission or
reception. On transmission, the key can be programmed at the same time as the rest of the frame data and setup
information.
During reception, the CPU must look up the key and provide it from information held in the header of the incoming
frame. However, the hardware of the security engine can process data much faster than the incoming frame data
rate. This means that it is possible to allow the CPU to receive the interrupt from the header of an incoming packet,
read where the frame originated, look up the key and program it to the security hardware before the end of the frame
has arrived. By providing a small amount of buffering to store incoming data while the lookup process is taking place,
the security engine can catch up processing the frame so that when the frame arrives in the receive frame buffer it is
fully decrypted.
8.4 Security Coprocessor
As well as being used during in-line encryption/decryption operations over a streaming interface and in external
memory encryption, it is also possible to use the AES core as a coprocessor to the CPU of the JN5148. To allow the
hardware to be shared between the two interfaces an arbiter ensures that the streaming interface to the AES core
always has priority, to ensure that in-line processing can take place at any time.