AD ADXL346ACCZ

3-Axis, ±2 g/±4 g/±8 g/±16 g
Ultralow Power Digital Accelerometer
ADXL346
FEATURES
GENERAL DESCRIPTION
Ultralow power: as low as 23 μA in measurement mode and
0.2 μA in standby mode at VS = 2.6 V (typical)
Power consumption scales automatically with bandwidth
User-selectable resolution
Fixed 10-bit resolution
Full resolution, where resolution increases with g range,
up to 13-bit resolution at ±16 g (maintaining 4 mg/LSB
scale factor in all g ranges)
Patent pending, embedded memory management system
with FIFO technology minimizes host processor load
Single-tap/double-tap detection
Activity/inactivity monitoring
Free-fall detection
Concurrent four- and six-position orientation detection
Supply and I/O voltage range: 1.7 V to 2.75 V
SPI (3- and 4-wire) and I2C digital interfaces
Flexible interrupt modes mappable to either interrupt pin
Measurement ranges selectable via serial command
Bandwidth selectable via serial command
Wide temperature range (−40°C to +85°C)
10,000 g shock survival
Pb free/RoHS compliant
Small and thin: 3 mm × 3 mm × 0.95 mm LGA package
The ADXL346 is a small, thin, ultralow power, 3-axis accelerometer
with high resolution (13-bit) measurement at up to ±16 g. Digital
output data is formatted as 16-bit twos complement and is accessible through either an SPI (3- or 4-wire) or I2C® digital interface.
The ADXL346 is well suited for mobile device applications. It
measures the static acceleration of gravity in tilt-sensing applications, as well as dynamic acceleration resulting from motion
or shock. Its high resolution (4 mg/LSB) enables measurement
of inclination changes of less than 1.0°.
Several special sensing functions are provided. Activity and
inactivity sensing detect the presence or lack of motion by
comparing the acceleration on any axis with user-set thresholds.
Tap sensing detects single and double taps in any direction. Freefall sensing detects if the device is falling. Orientation detection
is capable of concurrent four- and six-position sensing and a
user-selectable interrupt on orientation change for 2D or 3D
applications. These functions can be mapped individually to
either of two interrupt output pins. An integrated, patent pending
memory management system with 32-level first in, first out (FIFO)
buffer can be used to store data to minimize host processor activity
and lower overall system power consumption.
Low power modes enable intelligent motion-based power
management with threshold sensing and active acceleration
measurement at extremely low power dissipation.
APPLICATIONS
Handsets
Medical instrumentation
Gaming and pointing devices
Industrial instrumentation
Personal navigation devices
Hard disk drive (HDD) protection
The ADXL346 is supplied in a small, thin, 3 mm × 3 mm ×
0.95 mm, 16-lead, plastic package.
FUNCTIONAL BLOCK DIAGRAM
VS
ADXL346
VDD I/O
POWER
MANAGEMENT
SENSE
ELECTRONICS
ADC
3-AXIS
SENSOR
DIGITAL
FILTER
32-LEVEL
FIFO
CONTROL
AND
INTERRUPT
LOGIC
INT1
INT2
SDA/SDI/SDIO
SERIAL I/O
SDO/ALT
ADDRESS
CS
GND
08167-001
SCL/SCLK
Figure 1.
Rev. 0
Information furnished by Analog Devices is believed to be accurate and reliable. However, no
responsibility is assumed by Analog Devices for its use, nor for any infringements of patents or other
rights of third parties that may result from its use. Specifications subject to change without notice. No
license is granted by implication or otherwise under any patent or patent rights of Analog Devices.
Trademarks and registered trademarks are the property of their respective owners.
One Technology Way, P.O. Box 9106, Norwood, MA 02062-9106, U.S.A.
Tel: 781.329.4700
www.analog.com
Fax: 781.461.3113
©2010 Analog Devices, Inc. All rights reserved.
ADXL346
TABLE OF CONTENTS
Features .............................................................................................. 1 Register Definitions ................................................................... 23 Applications ....................................................................................... 1 Applications Information .............................................................. 29 General Description ......................................................................... 1 Power Supply Decoupling ......................................................... 29 Functional Block Diagram .............................................................. 1 Mechanical Considerations for Mounting .............................. 29 Revision History ............................................................................... 2 Tap Detection .............................................................................. 29 Specifications..................................................................................... 3 Improved Tap Detection............................................................ 30 Absolute Maximum Ratings............................................................ 5 Tap Sign ....................................................................................... 30 Thermal Resistance ...................................................................... 5 Threshold .................................................................................... 31 Package Information .................................................................... 5 Link Mode ................................................................................... 31 ESD Caution .................................................................................. 5 Sleep Mode vs. Low Power Mode............................................. 31 Pin Configuration and Function Descriptions ............................. 6 Offset Calibration ....................................................................... 31 Typical Performance Characteristics ............................................. 7 Using Self-Test ............................................................................ 32 Theory of Operation ...................................................................... 12 Orientation Sensing ................................................................... 32 Power Sequencing ...................................................................... 12 Data Formatting of Upper Data Rates ..................................... 34 Power Savings ............................................................................. 13 Noise Performance ..................................................................... 35 Serial Communications ................................................................. 14 Operation at Voltages Other Than 2.6 V ................................ 35 SPI ................................................................................................. 14 Offset Performance at Lowest Data Rates ............................... 36 I2C ................................................................................................. 17 Axes of Acceleration Sensitivity ............................................... 37 Interrupts ..................................................................................... 19 Layout and Design Recommendations ................................... 38 FIFO ............................................................................................. 20 Outline Dimensions ....................................................................... 39 Self-Test ........................................................................................ 21 Ordering Guide .......................................................................... 39 Register Map.................................................................................... 22 REVISION HISTORY
5/10—Revision 0: Initial Version
Rev. 0 | Page 2 of 40
ADXL346
SPECIFICATIONS
TA = 25°C, VS = 2.6 V, VDD I/O = 1.8 V, acceleration = 0 g, CS = 10 μF tantalum, CI/O = 0.1 μF, ODR = 800 Hz, unless otherwise noted.
Table 1. Specifications
Parameter
SENSOR INPUT
Measurement Range
Nonlinearity
Inter-Axis Alignment Error
Cross-Axis Sensitivity 3
OUTPUT RESOLUTION
All g Ranges
±2 g Range
±4 g Range
±8 g Range
±16 g Range
SENSITIVITY
Sensitivity at XOUT, YOUT, ZOUT
Sensitivity Deviation from Ideal
Scale Factor at XOUT, YOUT, ZOUT
Sensitivity Change Due to Temperature
0 g OFFSET
0 g Output for XOUT, YOUT, ZOUT
0 g Output Deviation from Ideal
0 g Offset vs. Temperature for X-, Y-Axes
0 g Offset vs. Temperature for Z-Axis
NOISE
X-, Y-Axes
Z-Axis
OUTPUT DATA RATE AND BANDWIDTH
Output Data Rate (ODR) 4, 5, 6
SELF-TEST 7
Output Change in X-Axis
Output Change in Y-Axis
Output Change in Z-Axis
POWER SUPPLY
Operating Voltage Range (VS)
Interface Voltage Range (VDD I/O)
Measurement Mode Supply Current
Standby Mode Supply Current
Turn-On and Wake-Up Time 8
Test Conditions
Each axis
User selectable
Percentage of full scale
Each axis
10-bit resolution
Full resolution
Full resolution
Full resolution
Full resolution
Each axis
All g ranges, full resolution
±2 g, 10-bit resolution
±4 g, 10-bit resolution
±8 g, 10-bit resolution
±16 g, 10-bit resolution
All g ranges
All g ranges, full resolution
±2 g, 10-bit resolution
±4 g, 10-bit resolution
±8 g, 10-bit resolution
±16 g, 10-bit resolution
Min 1
230
230
115
57
29
3.5
3.5
7.1
14.1
28.6
Typ 2
Max1
Unit
±2, ±4, ±8, ±16
±0.5
±0.1
±1
g
%
Degrees
%
10
10
11
12
13
Bits
Bits
Bits
Bits
Bits
256
256
128
64
32
±1.0
3.9
3.9
7.8
15.6
31.2
±0.02
282
282
141
71
35
0
±35
±0.7
±1.3
+150
4.3
4.3
8.7
17.5
34.5
LSB/g
LSB/g
LSB/g
LSB/g
LSB/g
%
mg/LSB
mg/LSB
mg/LSB
mg/LSB
mg/LSB
%/°C
Each axis
−150
ODR = 100 Hz for ±2 g, 10-bit
resolution or all g ranges, full
resolution
ODR = 100 Hz for ±2 g, 10-bit
resolution or all g ranges, full
resolution
User selectable
ODR = 3200 Hz
Rev. 0 | Page 3 of 40
1.1
LSB rms
1.5
LSB rms
0.10
3200
Hz
0.27
−1.55
0.40
1.55
−0.27
1.95
g
g
g
2.75
VS
V
V
μA
μA
μA
ms
1.7
1.7
ODR ≥ 100 Hz
ODR < 10 Hz
mg
mg
mg/°C
mg/°C
2.6
1.8
140
30
0.2
1.4
ADXL346
Parameter
TEMPERATURE
Operating Temperature Range
WEIGHT
Device Weight
Test Conditions
Min 1
Typ 2
−40
18
1
Max1
Unit
+85
°C
mg
All minimum and maximum specifications are guaranteed. Typical specifications are not guaranteed.
The typical specifications shown are for at least 68% of the population of parts and are based on the worst case of mean ±1 σ except for 0 g output and sensitivity,
which represents the target value. For 0 g offset and sensitivity, the deviation from the ideal describes the worst case of mean ±1 σ.
3
Cross-axis sensitivity is defined as coupling between any two axes.
4
Bandwidth is the −3 dB frequency and is half the output data rate bandwidth = ODR/2.
5
The output format for the 3200 Hz and 1600 Hz ODRs is different from the output format for the remaining ODRs. This difference is described in the Data Formatting of
Upper Data Rates section.
6
Output data rates below 6.25 Hz exhibit additional offset shift with increased temperature, depending on selected output data rate. Refer to the Offset Performance at
Lowest Data Rates section for details.
7
Self-test change is defined as the output (g) when the SELF_TEST bit = 1 (in the DATA_FORMAT register, Address 0x31) minus the output (g) when the SELF_TEST bit = 0.
Due to device filtering, the output reaches its final value after 4 × τ when enabling or disabling self-test, where τ = 1/(data rate). The part must be in normal power
operation (LOW_POWER bit = 0 in the BW_RATE register, Address 0x2C) for self-test to operate correctly.
8
Turn-on and wake-up times are determined by the user-defined bandwidth. At a 100 Hz data rate, the turn-on and wake-up times are each approximately 11.1 ms. For
other data rates, the turn-on and wake-up times are each approximately τ + 1.1 in milliseconds, where τ = 1/(data rate).
2
Rev. 0 | Page 4 of 40
ADXL346
ABSOLUTE MAXIMUM RATINGS
PACKAGE INFORMATION
Parameter
Acceleration
Any Axis, Unpowered
Any Axis, Powered
VS
VDD I/O
Digital Pins
All Other Pins
Output Short-Circuit Duration
(Any Pin to Ground)
Temperature Range
Powered
Storage
Rating
10,000 g
10,000 g
−0.3 V to +3.0 V
−0.3 V to +3.0 V
−0.3 V to VDD I/O + 0.3 V or
3.0 V, whichever is less
−0.3 V to +3.0 V
Indefinite
The information in Figure 2 and Table 4 provide details about
the package branding for ADXL346. For a complete listing of
product availability, see the Ordering Guide section.
Y2Z
vvvv
Figure 2. Product Information on Package (Top View)
−40°C to +105°C
−40°C to +105°C
Table 4. Package Branding Information
Stresses above those listed under Absolute Maximum Ratings
may cause permanent damage to the device. This is a stress
rating only; functional operation of the device at these or any
other conditions above those indicated in the operational
section of this specification is not implied. Exposure to absolute
maximum rating conditions for extended periods may affect
device reliability.
Branding Key
Y2Z
vvvv
ESD CAUTION
THERMAL RESISTANCE
Table 3. Package Characteristics
Package Type
16-Terminal LGA
θJA
150°C/W
08167-047
Table 2.
θJC
85°C/W
Device Weight
18 mg
Rev. 0 | Page 5 of 40
Field Description
Part identifier for ADXL346
Factory lot code
ADXL346
VDD I/O
GND
RESERVED
VS
PIN CONFIGURATION AND FUNCTION DESCRIPTIONS
16
15
14
1
2
NC
ADXL346
13
GND
12
GND
+X
5
NC = NO INTERNAL
CONNECTION
11
INT1
10
NC
+Z
9
6
7
8
INT2
TOP VIEW
(Not to Scale)
08167-002
NC
+Y
CS
4
SDO/
ALT ADDRESS
3
SDA/SDI/SDIO
NC
SCL/SCLK
Figure 3. Pin Configuration (Top View)
Table 5. Pin Function Descriptions
Pin No.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
Mnemonic
VDD I/O
NC
NC
SCL/SCLK
NC
SDA/SDI/SDIO
SDO/ALT ADDRESS
CS
INT2
NC
INT1
GND
GND
VS
RESERVED
GND
Description
Digital Interface Supply Voltage.
Not Internally Connected.
Not Internally Connected.
Serial Communications Clock.
Not Internally Connected.
Serial Data (I2C)/Serial Data Input (SPI 4-Wire)/Serial Data Input and Output (SPI 3-Wire).
Serial Data Output (SPI 4-Wire)/Alternate I2C Address Select (I2C).
Chip Select.
Interrupt 2 Output.
Not Internally Connected.
Interrupt 1 Output.
Must be connected to ground.
Must be connected to ground.
Supply Voltage.
Reserved. This pin must be connected to VS.
Must be connected to ground.
Rev. 0 | Page 6 of 40
ADXL346
30
30
25
25
PERCENT OF POPULATION (%)
20
15
10
10
–100
–50
0
50
ZERO g OFFSET (mg)
100
30
30
25
25
20
15
10
–100
–50
0
50
ZERO g OFFSET (mg)
100
150
15
10
0
–150
150
–100
–50
0
50
ZERO g OFFSET (mg)
100
150
Figure 8. Y-Axis Zero g Offset at 25°C, VS = 1.8 V
30
25
25
PERCENT OF POPULATION (%)
30
20
15
10
5
20
15
10
5
08167-006
PERCENT OF POPULATION (%)
100
20
Figure 5. Y-Axis Zero g Offset at 25°C, VS = 2.6 V
0
–150
–50
0
50
ZERO g OFFSET (mg)
5
08167-005
5
0
–150
–100
Figure 7. X-Axis Zero g Offset at 25°C, VS = 1.8 V
PERCENT OF POPULATION (%)
PERCENT OF POPULATION (%)
Figure 4. X-Axis Zero g Offset at 25°C, VS = 2.6 V
08167-104
0
–150
150
08167-105
0
–150
15
5
08167-004
5
20
–100
–50
0
50
ZERO g OFFSET (mg)
100
0
–150
150
Figure 6. Z-Axis Zero g Offset at 25°C, VS = 2.6 V
08167-106
PERCENT OF POPULATION (%)
TYPICAL PERFORMANCE CHARACTERISTICS
–100
–50
0
50
ZERO g OFFSET (mg)
100
Figure 9. Z-Axis Zero g Offset at 25°C, VS = 1.8 V
Rev. 0 | Page 7 of 40
150
ADXL346
250
60
150
ZERO g OFFSET (mg)
40
30
20
100
50
0
–50
–100
–150
08167-010
10
0
–3
–2
–1
0
08167-013
PERCENT OF POPULATION (%)
200
50
–200
–250
–40
1
–20
0
ZERO g OFFSET TEMPERATURE COEFFICIENT (mg/°C)
20
40
60
80
100
TEMPERATURE (°C)
Figure 10. X-Axis Zero g Offset Temperature Coefficient, VS = 2.6 V
Figure 13. X-Axis Zero g Offset vs. Temperature—
Eight Parts Soldered to PCB, VS = 2.6 V
60
250
150
ZERO g OFFSET (mg)
40
30
20
100
50
0
–50
–100
–150
08167-011
10
0
–3
–2
–1
0
08167-014
PERCENT OF POPULATION (%)
200
50
–200
–250
–40
1
–20
0
ZERO g OFFSET TEMPERATURE COEFFICIENT (mg/°C)
20
40
60
80
100
TEMPERATURE (°C)
Figure 11. Y-Axis Zero g Offset Temperature Coefficient, VS = 2.6 V
Figure 14. Y-Axis Zero g Offset vs. Temperature—
Eight Parts Soldered to PCB, VS = 2.6 V
60
250
150
ZERO g OFFSET (mg)
40
30
20
100
50
0
–50
–100
–150
08167-012
10
0
–3
–2
–1
0
08167-015
PERCENT OF POPULATION (%)
200
50
–200
–250
–40
1
ZERO g OFFSET TEMPERATURE COEFFICIENT (mg/°C)
–20
0
20
40
60
80
TEMPERATURE (°C)
Figure 12. Z-Axis Zero g Offset Temperature Coefficient, VS = 2.6 V
Figure 15. Z-Axis Zero g Offset vs. Temperature—
Eight Parts Soldered to PCB, VS = 2.6 V
Rev. 0 | Page 8 of 40
100
60
60
50
50
PERCENT OF POPULATION (%)
40
30
20
230
240
250
260
SENSITIVITY (LSB/g)
270
0
280
230
Figure 16. X-Axis Sensitivity at 25°C, VS = 2.6 V, Full Resolution
240
250
260
SENSITIVITY (LSB/g)
270
280
Figure 19. X-Axis Sensitivity at 25°C, VS = 1.8 V, Full Resolution
60
60
50
50
PERCENT OF POPULATION (%)
40
30
20
30
20
10
08167-017
10
40
0
230
240
250
260
SENSITIVITY (LSB/g)
270
0
280
230
Figure 17. Y-Axis Sensitivity at 25°C, VS = 2.6 V, Full Resolution
50
50
PERCENT OF POPULATION (%)
60
40
30
20
10
240
250
260
SENSITIVITY (LSB/g)
270
250
260
SENSITIVITY (LSB/g)
270
280
40
30
20
10
08167-018
230
240
Figure 20. Y-Axis Sensitivity at 25°C, VS = 1.8 V, Full Resolution
60
0
08167-117
PERCENT OF POPULATION (%)
20
08167-116
0
PERCENT OF POPULATION (%)
30
10
08167-016
10
40
08167-118
PERCENT OF POPULATION (%)
ADXL346
0
280
230
Figure 18. Z-Axis Sensitivity at 25°C, VS = 2.6 V, Full Resolution
240
250
260
SENSITIVITY (LSB/g)
270
280
Figure 21. Z-Axis Sensitivity at 25°C, VS = 1.8 V, Full Resolution
Rev. 0 | Page 9 of 40
280
90
275
80
270
70
265
SENSITIVITY (LSB/g)
100
60
50
40
30
20
260
255
250
245
10
0
–0.10
–0.05
0
0.05
SENSITIVITY TEMPERATURE COEFFICIENT (%/°C)
235
230
–40
0.10
–20
0
20
40
TEMPERATURE (°C)
60
80
100
Figure 25. X-Axis Sensitivity vs. Temperature—
Eight Parts Soldered to PCB, VS = 2.6 V, Full Resolution
280
90
275
80
270
70
265
SENSITIVITY (LSB/g)
100
60
50
40
30
20
260
255
250
245
10
0
–0.10
–0.05
0
0.05
SENSITIVITY TEMPERATURE COEFFICIENT (%/°C)
08167-026
240
08167-023
235
230
–40
0.10
Figure 23. Y-Axis Sensitivity Temperature Coefficient, VS = 2.6 V
–20
0
20
40
TEMPERATURE (°C)
60
80
100
Figure 26. Y-Axis Sensitivity vs. Temperature—
Eight Parts Soldered to PCB, VS = 2.6 V, Full Resolution
280
90
275
80
270
70
265
SENSITIVITY (LSB/g)
100
60
50
40
30
20
260
255
250
245
08167-024
240
10
0
–0.10
–0.05
0
0.05
SENSITIVITY TEMPERATURE COEFFICIENT (%/°C)
08167-127
PERCENT OF POPULATION (%)
Figure 22. X-Axis Sensitivity Temperature Coefficient, VS = 2.6 V
PERCENT OF POPULATION (%)
08167-025
240
08167-022
PERCENT OF POPULATION (%)
ADXL346
235
230
–40
0.10
Figure 24. Z-Axis Sensitivity Temperature Coefficient, VS = 2.6 V
–20
0
20
40
TEMPERATURE (°C)
60
80
Figure 27. Z-Axis Sensitivity vs. Temperature—
Eight Parts Soldered to PCB, VS = 2.6 V, Full Resolution
Rev. 0 | Page 10 of 40
100
40
40
35
35
PERCENT OF POPULATION (%)
30
25
20
15
10
0.5
0.6
0.7
0.8
SELF-TEST SHIFT (g)
0.9
10
0
90
100
110
120
130
140
150
OUTPUT CURRENT (µA)
160
170
180
Figure 31. Supply Current at 25°C, 100 Hz Output Data Rate, VS = 2.6 V
40
160
35
140
30
SUPPLY CURRENT (µA)
25
20
15
10
08167-008
5
0
–1.0
–0.9
–0.8
–0.7
SELF-TEST SHIFT (g)
–0.6
120
100
80
60
40
20
08167-020
PERCENT OF POPULATION (%)
15
1.0
Figure 28. X-Axis Self-Test Response at 25°C, VS = 2.6 V
0
–0.5
3.13 6.25 12.50 25
50 100 200 400
OUTPUT DATA RATE (Hz)
Figure 29. Y-Axis Self-Test Response at 25°C, VS = 2.6 V
800 1600 3200
Figure 32. Supply Current vs. Output Data Rate at 25°C—10 Parts, VS = 2.6 V
40
150
SUPPLY CURRENT CONSUMPTION (µA)
35
30
25
20
15
10
5
08167-009
PERCENT OF POPULATION (%)
20
08167-019
0
25
5
08167-007
5
30
0
1.0
1.1
1.2
1.3
SELF-TEST SHIFT (g)
1.4
140
130
120
110
100
08167-021
PERCENT OF POPULATION (%)
ADXL346
90
1.5
1.6
Figure 30. Z-Axis Self-Test Response at 25°C, VS = 2.6 V
1.8
2.0
2.2
2.4
SUPPLY VOLTAGE, VS (V)
2.6
Figure 33. Supply Current vs. Supply Voltage at 25°C
Rev. 0 | Page 11 of 40
2.8
ADXL346
THEORY OF OPERATION
The ADXL346 is a complete 3-axis acceleration measurement
system with a selectable measurement range of ±2 g, ±4 g, ±8 g,
or ±16 g. It measures both dynamic acceleration resulting from
motion or shock and static acceleration, such as gravity, which
allows the device to be used as a tilt sensor.
The sensor is a polysilicon surface-micromachined structure
built on top of a silicon wafer. Polysilicon springs suspend the
structure over the surface of the wafer and provide a resistance
against forces due to applied acceleration.
Deflection of the structure is measured using differential capacitors
that consist of independent fixed plates and plates attached to the
moving mass. Acceleration deflects the proof mass and unbalances
the differential capacitor, resulting in a sensor output with an
amplitude proportional to acceleration. Phase-sensitive demodulation is used to determine the magnitude and polarity
of the acceleration.
POWER SEQUENCING
Power can be applied to VS or VDD I/O in any sequence without
damaging the ADXL346. All possible power-on modes are
summarized in Table 6. The interface voltage level is set with
the interface supply voltage, VDD I/O, which must be present to
ensure that the ADXL346 does not create a conflict on the
communication bus. For single-supply operation, VDD I/O can be
the same as the main supply, VS. In a dual-supply application,
however, VDD I/O can differ from VS to accommodate the desired
interface voltage, as long as VS is greater than or equal to VDD I/O.
After VS is applied, the device enters standby mode, where power
consumption is minimized and the device waits for VDD I/O to be
applied and for the command to enter measurement mode to be
received. (This command can be initiated by setting the measure
bit (Bit D3) in the POWER_CTL register (Address 0x2D).) In
addition, any register can be written to or read from to configure
the part while the device is in standby mode. It is recommended
to configure the device in standby mode and then to enable
measurement mode. Clearing the measure bit returns the
device to the standby mode.
Table 6. Power Sequencing
Condition
Power Off
VS
Off
VDD I/O
Off
Bus Disabled
On
Off
Bus Enabled
Off
On
Standby or
Measurement Mode
On
On
Description
The device is completely off, but there is a potential for a communication
bus conflict.
The device is on in standby mode, but communication is unavailable and will
create a conflict on the communication bus. The duration of this state should
be minimized during power-up to prevent a conflict.
No functions are available, but the device will not create a conflict on the
communication bus.
At power-up, the device is in standby mode, awaiting a command to enter
measurement mode, and all sensor functions are off. After the device is
instructed to enter measurement mode, all sensor functions are available.
Rev. 0 | Page 12 of 40
ADXL346
POWER SAVINGS
Table 8. Typical Current Consumption vs. Data Rate, Low
Power Mode (TA = 25°C, VS = 2.6 V, VDD I/O = 1.8 V)
Power Modes
The ADXL346 automatically modulates its power consumption
in proportion to its output data rate, as outlined in Table 7. If
additional power savings is desired, a lower power mode is
available. In this mode, the internal sampling rate is reduced,
allowing for power savings in the 12.5 Hz to 400 Hz data rate
range at the expense of slightly greater noise. To enter low
power mode, set the LOW_POWER bit (Bit D4) in the BW_RATE
register (Address 0x2C). The current consumption in low power
mode is shown in Table 8 for cases where there is an advantage
to using low power mode. Use of low power mode for a data
rate not shown in Table 8 does not provide any advantage over
the same data rate in normal power mode. Therefore, it is
recommended that only data rates listed in Table 8 be used in
low power mode. The current consumption values shown in
Table 7 and Table 8 are for a VS of 2.6 V.
Table 7. Typical Current Consumption vs. Data Rate
(TA = 25°C, VS = 2.6 V, VDD I/O = 1.8 V)
Output Data
Rate (Hz)
3200
1600
800
400
200
100
50
25
12.5
6.25
3.13
1.56
0.78
0.39
0.20
0.10
Bandwidth (Hz)
1600
800
400
200
100
50
25
12.5
6.25
3.13
1.56
0.78
0.39
0.20
0.10
0.05
Rate Code
1111
1110
1101
1100
1011
1010
1001
1000
0111
0110
0101
0100
0011
0010
0001
0000
IDD (μA)
140
90
140
140
140
140
90
55
40
31
27
23
23
23
23
23
Output Data
Rate (Hz)
400
200
100
50
25
12.5
Bandwidth (Hz)
200
100
50
25
12.5
6.25
Rate Code
1100
1011
1010
1001
1000
0111
IDD (μA)
90
55
40
31
27
23
Autosleep Mode
Additional power can be saved if the ADXL346 automatically
switches to sleep mode during periods of inactivity. To enable
this feature, set the THRESH_INACT register (Address 0x25)
and the TIME_INACT register (Address 0x26) each to a value
that signifies inactivity (the appropriate value depends on the
application), and then set the AUTO_SLEEP bit (Bit D4) and the
link bit (Bit D5) in the POWER_CTL register (Address 0x2D).
Current consumption at the sub-8 Hz data rates used in this
mode is typically 23 μA for a VS of 2.6 V.
Standby Mode
For even lower power operation, standby mode can be used.
In standby mode, current consumption is reduced to 0.2 μA
(typical). In this mode, no measurements are made. Standby mode
is entered by clearing the measure bit (Bit D3) in the
POWER_CTL register (Address 0x2D). Placing the device into
standby mode preserves the contents of FIFO.
Rev. 0 | Page 13 of 40
ADXL346
SERIAL COMMUNICATIONS
I2C and SPI digital communications are available. In both cases,
the ADXL346 operates as a slave. I2C mode is enabled if the CS pin
is tied high to VDD I/O. The CS pin should always be tied high to
VDD I/O or be driven by an external controller because there is no
default mode if the CS pin is left unconnected. Therefore, not
taking these precautions may result in an inability to communicate
with the part. In SPI mode, the CS pin is controlled by the bus
master. In both SPI and I2C modes of operation, data transmitted
from the ADXL346 to the master device should be ignored during
writes to the ADXL346.
SPI
For SPI, either 3- or 4-wire configuration is possible, as shown in
the connection diagrams in Figure 34 and Figure 35. Clearing the
SPI bit (Bit D6) in the DATA_FORMAT register (Address 0x31)
selects 4-wire mode, whereas setting the SPI bit selects 3-wire
mode. The maximum SPI clock speed is 5 MHz with 100 pF
maximum loading, and the timing scheme follows clock polarity
(CPOL) = 1 and clock phase (CPHA) = 1. If power is applied to
the ADXL346 before the clock polarity and phase of the host
processor are configured, the CS pin should be brought high
before changing the clock polarity and phase. When using 3-wire
SPI, it is recommended that the SDO pin either be pulled up to
VDD I/O or be pulled down to GND via a 10 kΩ resistor.
CS
SDIO
PROCESSOR
D OUT
D IN/OUT
SDO
SCLK
D OUT
08167-027
ADXL346
Figure 34. 3-Wire SPI Connection Diagram
PROCESSOR
CS
D OUT
SDI
D OUT
SDO
SCLK
D IN
D OUT
To read or write multiple bytes in a single transmission, the
multiple-byte bit, located after the R/W bit in the first byte transfer
(MB in Figure 36 to Figure 38), must be set. After the register
addressing and the first byte of data, each subsequent set of
clock pulses (eight clock pulses) causes the ADXL346 to point
to the next register for a read or write. This shifting continues
until the clock pulses cease and CS is deasserted. To perform reads
or writes on different, nonsequential registers, CS must be
deasserted between transmissions and the new register must be
addressed separately.
The timing diagram for 3-wire SPI reads or writes is shown in
Figure 38. The 4-wire equivalents for SPI writes and reads are
shown in Figure 36 and Figure 37, respectively. For correct
operation of the part, the logic thresholds and timing parameters
in Table 9 and Table 10 must be met at all times.
Use of the 3200 Hz and 1600 Hz output data rates is recommended only with SPI communication rates greater than or
equal to 2 MHz. The 800 Hz output data rate is recommended
only for communication speeds greater than or equal to 400 kHz,
and the remaining data rates scale proportionally. For example,
the minimum recommended communication speed for a 200 Hz
output data rate is 100 kHz. Operation at an output data rate
above the recommended maximum may result in undesirable
effects on the acceleration data, including missing samples or
additional noise.
08167-028
ADXL346
CS is the serial port enable line and is controlled by the SPI master.
This line must go low at the start of a transmission and high at
the end of a transmission, as shown in Figure 36. SCLK is the
serial port clock and is supplied by the SPI master. SCLK should
idle high during a period of no transmission. SDI and SDO are
the serial data input and output, respectively. Data is updated
on the falling edge of SCLK and should be sampled on the
rising edge of SCLK.
Figure 35. 4-Wire SPI Connection Diagram
Rev. 0 | Page 14 of 40
ADXL346
CS
tM
tSCLK
tDELAY
tS
tQUIET
tCS,DIS
SCLK
tSETUP
tHOLD
MB
A5
tSDO
X
SDO
A0
D7
ADDRESS BITS
X
D0
tDIS
DATA BITS
X
X
X
08167-129
W
SDI
X
tR, tF
Figure 36. SPI 4-Wire Write
CS
tSCLK
tDELAY
tM
tS
tCS,DIS
tQUIET
SCLK
tHOLD
R
SDI
MB
tSDO
X
SDO
A0
A5
X
X
tDIS
ADDRESS BITS
X
X
D7
X
D0
08167-130
tSETUP
tR, tF
DATA BITS
Figure 37. SPI 4-Wire Read
CS
tDELAY
tM
tSCLK
tS
tQUIET
tCS,DIS
SCLK
tSETUP
SDIO
tHOLD
R/W
MB
tSDO
tR, tF
A5
A0
ADDRESS BITS
D7
D0
DATA BITS
08167-131
SDO
NOTES
1. tSDO IS ONLY PRESENT DURING READS.
Figure 38. SPI 3-Wire Read/Write
Rev. 0 | Page 15 of 40
ADXL346
Table 9. SPI Digital Input/Output
Parameter
Digital Input
Low Level Input Voltage (VIL)
High Level Input Voltage (VIH)
Low Level Input Current (IIL)
High Level Input Current (IIH)
Digital Output
Low Level Output Voltage (VOL)
High Level Output Voltage (VOH)
Low Level Output Current (IOL)
High Level Output Current (IOH)
Pin Capacitance
1
Test Conditions
Min
Limit 1
Max
0.3 × VDD I/O
0.7 × VDD I/O
VIN = VDD I/O
VIN = 0 V
0.1
−0.1
IOL = 10 mA
IOH = −4 mA
VOL = VOL, max
VOH = VOH, min
fIN = 1 MHz, VIN = 2.6 V
0.2 × VDD I/O
0.8 × VDD I/O
10
−4
8
Unit
V
V
μA
μA
V
V
mA
mA
pF
Limits are based on characterization results; not production tested.
Table 10. SPI Timing (TA = 25°C, VS = 2.6 V, VDD I/O = 1.8 V) 1
Limit 2, 3
Max
5
Parameter
fSCLK
tSCLK
tDELAY
tQUIET
tDIS
tCS,DIS
Min
tS
tM
tSETUP
tHOLD
tSDO
tR 4
tF4
0.3 × tSCLK
0.3 × tSCLK
5
5
200
5
5
10
150
40
20
20
Unit
MHz
ns
ns
ns
ns
ns
Description
SPI clock frequency
1/(SPI clock frequency) mark-space ratio for the SCLK input is 40/60 to 60/40
CS falling edge to SCLK falling edge
SCLK rising edge to CS rising edge
CS rising edge to SDO disabled
CS deassertion between SPI communications
ns
ns
ns
ns
ns
ns
ns
SCLK low pulse width (space)
SCLK high pulse width (mark)
SDI valid before SCLK rising edge
SDI valid after SCLK rising edge
SCLK falling edge to SDO/SDIO output transition
SDO/SDIO output low to output high transition
SDO/SDIO output high to output low transition
The CS, SCLK, SDI, and SDO pins are not internally pulled up or down; they must be driven for proper operation.
Limits are based on characterization results; not production tested.
3
The timing values are measured corresponding to the input thresholds (VIL and VIH) given in Table 9.
4
Output rise and fall times are measured with a capacitive load of 150 pF.
1
2
Rev. 0 | Page 16 of 40
ADXL346
I2C
Due to communication speed limitations, the maximum output
data rate when using 400 kHz I2C is 800 Hz and scales linearly with
a change in the I2C communication speed. For example, using I2C
at 100 kHz would limit the maximum ODR to 200 Hz. Operation
at an output data rate above the recommended maximum may
result in an undesirable effect on the acceleration data, including
missing samples or additional noise.
With CS tied high to VDD I/O, the ADXL346 is in I2C mode,
requiring a simple 2-wire connection as shown in Figure 39.
The ADXL346 conforms to the UM10204 I2C-Bus Specification
and User Manual, Rev. 03—19 June 2007, available from NXP
Semiconductor. It supports standard (100 kHz) and fast (400 kHz)
data transfer modes if the bus parameters given in Table 11 and
Table 12 are met. Single- or multiple-byte reads/writes are supported, as shown in Figure 40. With the ALT ADDRESS pin (Pin 7)
high, the 7-bit I2C address for the device is 0x1D, followed by the
R/W bit. This translates to 0x3A for a write and 0x3B for a read.
An alternate I2C address of 0x53 (followed by the R/W bit) can
be chosen by grounding the ALT ADDRESS pin. This translates
to 0xA6 for a write and 0xA7 for a read.
VDD I/O
ADXL346
RP
RP
PROCESSOR
CS
D IN/OUT
SDA
D OUT
SCL
There are no internal pull-up or pull-down resistors for any
unused pins; therefore, there is no known state or default state
for the CS or ALT ADDRESS pin if left floating or unconnected.
It is required that the CS pin be connected to VDD I/O and that
the ALT ADDRESS pin be connected to either VDD I/O or GND
when using I2C.
08167-032
ALT ADDRESS
Figure 39. I2C Connection Diagram (Address 0x53)
If other devices are connected to the same I2C bus, the nominal
operating voltage level of these other devices cannot exceed VDD I/O
by more than 0.3 V. External pull-up resistors, RP, are necessary for
proper I2C operation. Refer to the UM10204 I2C-Bus Specification
and User Manual, Rev. 03—19 June 2007, when selecting pull-up
resistor values to ensure proper operation.
Table 11. I2C Digital Input/Output
Parameter
Digital Input
Low Level Input Voltage (VIL)
High Level Input Voltage (VIH)
Low Level Input Current (IIL)
High Level Input Current (IIH)
Digital Output
Low Level Output Voltage (VOL)
Test Conditions
Min
Unit
0.3 × VDD I/O
V
V
μA
μA
0.7 × VDD I/O
VIN = VDD I/O
VIN = 0 V
0.1
−0.1
VDD I/O < 2 V, IOL = 3 mA
VDD I/O ≥ 2 V, IOL = 3 mA
VOL = VOL, max
fIN = 1 MHz, VIN = 2.6 V
Low Level Output Current (IOL)
Pin Capacitance
0.2 × VDD I/O
400
V
mV
mA
pF
3
8
Limits are based on characterization results; not production tested.
SINGLE-BYTE WRITE
MASTER START
SLAVE ADDRESS + WRITE
SLAVE
DATA
REGISTER ADDRESS
ACK
ACK
STOP
ACK
MULTIPLE-BYTE WRITE
MASTER START
SLAVE ADDRESS + WRITE
SLAVE
DATA
REGISTER ADDRESS
ACK
ACK
DATA
STOP
ACK
ACK
SINGLE-BYTE READ
MASTER START
SLAVE ADDRESS + WRITE
SLAVE
START1
REGISTER ADDRESS
ACK
SLAVE ADDRESS + READ
ACK
NACK
ACK
DATA
ACK
DATA
STOP
MULTIPLE-BYTE READ
MASTER START
SLAVE
1THIS
SLAVE ADDRESS + WRITE
START1
REGISTER ADDRESS
ACK
ACK
SLAVE ADDRESS + READ
ACK
NACK
STOP
DATA
START IS EITHER A RESTART OR A STOP FOLLOWED BY A START.
NOTES
1. THE SHADED AREAS REPRESENT WHEN THE DEVICE IS LISTENING.
Figure 40. I2C Device Addressing
Rev. 0 | Page 17 of 40
08167-033
1
Limit 1
Max
ADXL346
Table 12. I2C Timing (TA = 25°C, VS = 2.6 V, VDD I/O = 1.8 V)
Parameter
fSCL
t1
t2
t3
t4
t5
t6 3, 4, 5, 6
t7
t8
t9
t10
Limit 1, 2
Max
400
Min
2.5
0.6
1.3
0.6
100
0
0.6
0.6
1.3
Unit
kHz
μs
μs
μs
μs
ns
μs
μs
μs
μs
ns
ns
ns
ns
ns
pF
0.9
300
0
t11
250
300
20 + 0.1 CB 7
CB
7
400
Description
SCL clock frequency
SCL cycle time
tHIGH, SCL high time
tLOW, SCL low time
tHD, STA, start/repeated start condition hold time
tSU, DAT, data setup time
tHD, DAT, data hold time
tSU, STA, setup time for repeated start
tSU, STO, stop condition setup time
tBUF, bus-free time between a stop condition and a start condition
tR, rise time of both SCL and SDA when receiving
tR, rise time of both SCL and SDA when receiving or transmitting
tF, fall time of SDA when receiving
tF, fall time of both SCL and SDA when transmitting
tF, fall time of both SCL and SDA when transmitting or receiving
Capacitive load for each bus line
1
Limits are based on characterization results, with fSCL = 400 kHz and a 3 mA sink current; not production tested.
All values referred to the VIH and the VIL levels given in Table 11.
3
t6 is the data hold time that is measured from the falling edge of SCL. It applies to data in transmission and acknowledge.
4
A transmitting device must internally provide an output hold time of at least 300 ns for the SDA signal (with respect to VIH,min of the SCL signal) to bridge the undefined region of
the falling edge of SCL.
5
The maximum t6 value must be met only if the device does not stretch the low period (t3) of the SCL signal.
6
The maximum value for t6 is a function of the clock low time (t3), the clock rise time (t10), and the minimum data setup time (t5(min)). This value is calculated as t6(max) = t3 − t10 − t5(min).
7
CB is the total capacitance of one bus line in picofarads.
2
SDA
t3
t9
t10
t4
t11
SCL
t6
t2
t5
t7
REPEATED
START
CONDITION
Figure 41. I2C Timing Diagram
Rev. 0 | Page 18 of 40
t1
t8
STOP
CONDITION
08167-034
t4
START
CONDITION
ADXL346
INTERRUPTS
DOUBLE_TAP Bit
The ADXL346 provides two output pins for driving interrupts:
INT1 and INT2. Both interrupt pins are push-pull, low impedance
pins with the output specifications listed in Table 13. The default
configuration of the interrupt pins is active high. This can be
changed to active low by setting the INT_INVERT bit (Bit D5)
in the DATA_FORMAT (Address 0x31) register. All functions
can be used simultaneously, with the only limiting feature being
that some functions may need to share interrupt pins.
The DOUBLE_TAP bit is set when two acceleration events
that are greater than the value in the THRESH_TAP register
(Address 0x1D) occur for less time than is specified in the DUR
register (Address 0x21). The second tap starts after the time
specified by the latent register (Address 0x22) but within the
time specified in the window register (Address 0x23). See the Tap
Detection section for more details.
Interrupts are enabled by setting the appropriate bit in the
INT_ENABLE register (Address 0x2E) and are mapped to either
the INT1 or INT2 pin based on the contents of the INT_MAP
register (Address 0x2F). When initially configuring the interrupt
pins, it is recommended that the functions and interrupt mapping
be done before enabling the interrupts. When changing the configuration of an interrupt, it is recommended that the interrupt be
disabled first, by clearing the bit corresponding to that function in
the INT_ENABLE register, and then the function be reconfigured
before enabling the interrupt again. Configuration of the functions
while the interrupts are disabled helps to prevent the accidental
generation of an interrupt before it is desired.
The activity bit is set when acceleration greater than the value stored
in the THRESH_ACT register (Address 0x24) is experienced on
any participating axis, as set by the ACT_INACT_CTL register
(Address 0x27).
The interrupt functions are latched and cleared by either reading
the DATAX, DATAY, and DATAZ registers (Address 0x32 to
Address 0x37) until the interrupt condition is no longer valid
for the data-related interrupts or by reading the INT_SOURCE
register (Address 0x30) for the remaining interrupts. This section
describes the interrupts that can be set in the INT_ENABLE
register and monitored in the INT_SOURCE register.
DATA_READY Bit
The DATA_READY bit is set when new data is available and is
cleared when no new data is available.
Activity Bit
Inactivity Bit
The inactivity bit is set when acceleration of less than the
value stored in the THRESH_INACT register (Address 0x25) is
experienced for more time than is specified in the TIME_INACT
register (Address 0x26) on all participating axes, as set by the
ACT_INACT_CTL register (Address 0x27). The maximum value
for TIME_INACT is 255 sec.
FREE_FALL Bit
The FREE_FALL bit is set when acceleration of less than the
value stored in the THRESH_FF register (Address 0x28) is
experienced for more time than is specified in the TIME_FF
register (Address 0x29) on all axes (logical AND). The FREE_FALL
interrupt differs from the inactivity interrupt as follows: all axes
always participate and are logically AND’ed, the timer period is
much smaller (1.28 sec maximum), and the mode of operation is
always dc-coupled.
Watermark Bit
The watermark bit is set when the number of samples in FIFO
equals the value stored in the samples bits (Register FIFO_CTL,
Address 0x38). The watermark bit is cleared automatically when
FIFO is read, and the content returns to a value below the value
stored in the samples bits.
SINGLE_TAP Bit
The SINGLE_TAP bit is set when a single acceleration event
that is greater than the value in the THRESH_TAP register
(Address 0x1D) occurs for less time than is specified in
the DUR register (Address 0x21).
Table 13. Interrupt Pin Digital Output
Parameter
Digital Output
Low Level Output Voltage (VOL)
High Level Output Voltage (VOH)
Low Level Output Current (IOL)
High Level Output Current (IOH)
Pin Capacitance
Rise/Fall Time
Rise Time (tR) 2
Fall Time (tF) 3
Test Conditions
IOL = 300 μA
IOH = −150 μA
VOL = VOL, max
VOH = VOH, min
fIN = 1 MHz, VIN = 2.6 V
Min
Limit 1
Max
0.2 × VDD I/O
−150
8
V
V
μA
μA
pF
210
150
ns
ns
0.8 × VDD I/O
300
CLOAD = 150 pF
CLOAD = 150 pF
1
Limits are based on characterization results; not production tested.
Rise time is measured as the transition time from VOL, max to VOH, min of the interrupt pin.
3
Fall time is measured as the transition time from VOH, min to VOL, max of the interrupt pin.
2
Rev. 0 | Page 19 of 40
Unit
ADXL346
Overrun Bit
Stream Mode
The overrun bit is set when new data replaces unread data. The
precise operation of the overrun function depends on the FIFO
mode. In bypass mode, the overrun bit is set when new data
replaces unread data in the DATAX, DATAY, and DATAZ registers
(Address 0x32 to Address 0x37). In all other modes, the overrun
bit is set when FIFO is filled. The overrun bit is automatically
cleared when the contents of FIFO are read.
In stream mode, data from measurements of the x-, y-, and zaxes are stored in FIFO. When the number of samples in FIFO
equals the level specified in the samples bits of the FIFO_CTL
register (Address 0x38), the watermark interrupt is set. FIFO
continues accumulating samples and holds the latest 32 samples
from measurements of the x-, y-, and z-axes, discarding older
data as new data arrives. The watermark interrupt continues
occurring until the number of samples in FIFO is less than the
value stored in the samples bits of the FIFO_CTL register.
Orientation Bit
The orientation bit is set when the orientation of the accelerometer
changes from a valid orientation to a different valid orientation.
An interrupt is not generated, however, if the orientation of the
accelerometer changes from a valid orientation to an invalid
orientation, or from a valid orientation to an invalid orientation
and then back to the same valid orientation. An invalid orientation
is defined as an orientation within the dead zone, or the region of
hysteresis. This region helps to prevent rapid orientation change
due to noise when the accelerometer orientation is close to the
boundary between two valid orientations.
The orientations that are valid for the interrupt depend on which
mode, 2D or 3D, is linked to the orientation interrupt. The mode is
selected with the INT_3D bit (Bit D3) in the ORIENT_CONF
register (Address 0x3B). See the Register 0x3B—ORIENT_CONF
(Read/Write) section for more details on how to enable the
orientation interrupt.
FIFO
Trigger Mode
In trigger mode, FIFO accumulates samples, holding the latest
32 samples from measurements of the x-, y-, and z-axes. After
a trigger event occurs and an interrupt is sent to the INT1 or
INT2 pin (determined by the trigger bit in the FIFO_CTL register),
FIFO keeps the last n samples (where n is the value specified by
the samples bits in the FIFO_CTL register) and then operates in
FIFO mode, collecting new samples only when FIFO is not full.
A delay of at least 5 μs should be present between the trigger event
occurring and the start of reading data from the FIFO to allow
the FIFO to discard and retain the necessary samples. Additional
trigger events cannot be recognized until the trigger mode is
reset. To reset the trigger mode, set the device to bypass mode
and then set the device back to trigger mode. Note that the FIFO
data should be read first because placing the device into bypass
mode clears FIFO.
Retrieving Data from FIFO
The ADXL346 contains patent pending technology for an
embedded memory management system with 32-level FIFO
that can be used to minimize host processor burden. This buffer
has four modes: bypass, FIFO, stream, and trigger (see Table 22).
Each mode is selected by the settings of the FIFO_MODE bits
(Bits[D7:D6]) in the FIFO_CTL register (Address 0x38).
Bypass Mode
In bypass mode, FIFO is not operational and, therefore,
remains empty.
FIFO Mode
In FIFO mode, data from measurements of the x-, y-, and z-axes
are stored in FIFO. When the number of samples in FIFO
equals the level specified in the samples bits of the FIFO_CTL
register (Address 0x38), the watermark interrupt is set. FIFO
continues accumulating samples until it is full (32 samples from
measurements of the x-, y-, and z-axes) and then stops collecting
data. After FIFO stops collecting data, the device continues to
operate; therefore, features such as tap detection can be used
after FIFO is full. The watermark interrupt continues to occur
until the number of samples in FIFO is less than the value
stored in the samples bits of the FIFO_CTL register.
The FIFO data is read through the DATAX, DATAY, and DATAZ
registers (Address 0x32 to Address 0x37). When the FIFO is in
FIFO, stream, or trigger mode, reads to the DATAX, DATAY,
and DATAZ registers read data stored in the FIFO. Each time
data is read from the FIFO, the oldest x-, y-, and z-axes data are
placed into the DATAX, DATAY, and DATAZ registers.
If a single-byte read operation is performed, the remaining bytes of
data for the current FIFO sample are lost. Therefore, all axes of
interest should be read in a burst (or multiple-byte) read operation.
To ensure that the FIFO has completely popped (that is, that new
data has completely moved into the DATAX, DATAY, and DATAZ
registers), there must be at least 5 μs between the end of reading
the data registers and the start of a new read of the FIFO or a
read of the FIFO_STATUS register (Address 0x39). The end of
reading a data register is signified by the transition of data from
Register 0x37 to Register 0x38 or by the CS pin going high.
For SPI operation at 1.6 MHz or less, the register addressing
portion of the transmission is a sufficient delay to ensure that
the FIFO has completely popped. For SPI operation greater than
1.6 MHz, it is necessary to deassert the CS pin to ensure a total
delay of 5 μs; otherwise, the delay will not be sufficient. The total
delay necessary for 5 MHz operation is at most 3.4 μs. This is
not a concern when using I2C mode because the communication
rate is low enough to ensure a sufficient delay between FIFO reads.
Rev. 0 | Page 20 of 40
ADXL346
SELF-TEST
The ADXL346 incorporates a self-test feature that effectively
tests its mechanical and electronic systems simultaneously.
When the self-test function is enabled (via the SELF_TEST bit
(Bit D7 in the DATA_FORMAT register, Address 0x31), an
electrostatic force is exerted on the mechanical sensor. This
electrostatic force moves the mechanical sensing element in the
same manner as acceleration would, and it is additive to the
acceleration experienced by the device. This added electrostatic
force results in an output change in the x-, y-, and z-axes. Because
the electrostatic force is proportional to VS2, the output change
varies with VS. This effect is shown in Figure 42.
The scale factors listed in Table 14 can be used to adjust the
expected self-test output limits for different supply voltages, VS.
The self-test feature of the ADXL346 also exhibits a bimodal
behavior. However, the limits listed in Table 1 and Table 15 to
Table 18 are valid for both potential self-test values due to bimodality. Use of the self-test feature at data rates less than 100 Hz
or at 1600 Hz may yield values outside these limits. Therefore,
the part must be in normal power operation (LOW_POWER
bit = 0 in the BW_RATE register, Address 0x2C) and be placed
into a data rate of 100 Hz through 800 Hz or 3200 Hz for the
self-test function to operate correctly.
3
X-AXIS SELF-TEST HIGH LIMIT
Y-AXIS SELF-TEST HIGH LIMIT
Z-AXIS SELF-TEST HIGH LIMIT
X-AXIS SELF-TEST LOW LIMIT
Y-AXIS SELF-TEST LOW LIMIT
Z-AXIS SELF-TEST LOW LIMIT
Supply Voltage, VS
1.70 V
1.80 V
2.00 V
2.60 V
2.75 V
1
X-, Y-Axes
0.43
0.48
0.59
1.00
1.13
Z-Axis
0.38
0.47
0.58
1.00
1.11
Table 15. Self-Test Output in LSB for ±2 g, 10-Bit or Full
Resolution (TA = 25°C, VS = 2.6 V, VDD I/O = 1.8 V)
Axis
X
Y
Z
Min
70
−400
100
Max
400
−70
500
Unit
LSB
LSB
LSB
Table 16. Self-Test Output in LSB for ±4 g, 10-Bit Resolution
(TA = 25°C, VS = 2.6 V, VDD I/O = 1.8 V)
Axis
X
Y
Z
Min
35
−200
50
Max
200
−35
250
Unit
LSB
LSB
LSB
Table 17. Self-Test Output in LSB for ±8 g, 10-Bit Resolution
(TA = 25°C, VS = 2.6 V, VDD I/O = 1.8 V)
Axis
X
Y
Z
Min
17
−100
25
Max
100
−17
125
Unit
LSB
LSB
LSB
Table 18. Self-Test Output in LSB for ±16 g, 10-Bit Resolution
(TA = 25°C, VS = 2.6 V, VDD I/O = 1.8 V)
0
–1
–2
08167-136
SELF-TEST SHIFT LIMITS (g)
2
Table 14. Self-Test Output Scale Factors for Different Supply
Voltages, VS
–3
1.6
1.8
2.0
2.2
2.4
SUPPLY VOLTAGE, VS (V)
2.6
Axis
X
Y
Z
2.8
Figure 42. Self-Test Output Change Limits vs. Supply Voltage
Rev. 0 | Page 21 of 40
Min
8
−50
12
Max
50
−8
63
Unit
LSB
LSB
LSB
ADXL346
REGISTER MAP
Table 19. Register Map
Address
Hex
Dec
0x00
0
0x01 to 0x1C
1 to 28
0x1D
29
0x1E
30
0x1F
31
0x20
32
0x21
33
0x22
34
0x23
35
0x24
36
0x25
37
0x26
38
0x27
39
0x28
40
0x29
41
0x2A
42
0x2B
43
0x2C
44
0x2D
45
0x2E
46
0x2F
47
0x30
48
0x31
49
0x32
50
0x33
51
0x34
52
0x35
53
0x36
54
0x37
55
0x38
56
0x39
57
0x3A
58
0x3B
59
0x3C
60
Name
DEVID
Reserved
THRESH_TAP
OFSX
OFSY
OFSZ
DUR
Latent
Window
THRESH_ACT
THRESH_INACT
TIME_INACT
ACT_INACT_CTL
THRESH_FF
TIME_FF
TAP_AXES
ACT_TAP_STATUS
BW_RATE
POWER_CTL
INT_ENABLE
INT_MAP
INT_SOURCE
DATA_FORMAT
DATAX0
DATAX1
DATAY0
DATAY1
DATAZ0
DATAZ1
FIFO_CTL
FIFO_STATUS
TAP_SIGN
ORIENT_CONF
Orient
Type
R
Reset Value
11100110
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R/W
R
R/W
R/W
R/W
R/W
R
R/W
R
R
R
R
R
R
R/W
R
R
R/W
R
00000000
00000000
00000000
00000000
00000000
00000000
00000000
00000000
00000000
00000000
00000000
00000000
00000000
00000000
00000000
00001010
00000000
00000000
00000000
00000010
00000000
00000000
00000000
00000000
00000000
00000000
00000000
00000000
00000000
00000000
00100101
00000000
Rev. 0 | Page 22 of 40
Description
Device ID.
Reserved. Do not access.
Tap threshold.
X-axis offset.
Y-axis offset.
Z-axis offset.
Tap duration.
Tap latency.
Tap window.
Activity threshold.
Inactivity threshold.
Inactivity time.
Axis enable control for activity and inactivity detection.
Free-fall threshold.
Free-fall time.
Axis control for single tap/double tap.
Source of single tap/double tap.
Data rate and power mode control.
Power-saving features control.
Interrupt enable control.
Interrupt mapping control.
Source of interrupts.
Data format control.
X-Axis Data 0.
X-Axis Data 1.
Y-Axis Data 0.
Y-Axis Data 1.
Z-Axis Data 0.
Z-Axis Data 1.
FIFO control.
FIFO status.
Sign and source for single tap/double tap.
Orientation configuration.
Orientation status.
ADXL346
Register 0x24—THRESH_ACT (Read/Write)
REGISTER DEFINITIONS
Register 0x00—DEVID (Read Only)
The DEVID register holds a fixed device ID code of 0xE6
(346 octal).
The THRESH_ACT register is eight bits and holds the threshold
value for detecting activity. The data format is unsigned, so the
magnitude of the activity event is compared with the value in
the THRESH_ACT register. The scale factor is 62.5 mg/LSB.
A value of 0 may result in undesirable behavior if the activity
interrupt is enabled.
Register 0x1D—THRESH_TAP (Read/Write)
Register 0x25—THRESH_INACT (Read/Write)
The THRESH_TAP register is eight bits and holds the threshold
value for tap interrupts. The data format is unsigned, so the
magnitude of the tap event is compared with the value in
THRESH_TAP for normal tap detection. For information on
improved tap detection, refer to the Improved Tap Detection
section. The scale factor is 62.5 mg/LSB (that is, 0xFF = +16 g).
A value of 0 may result in undesirable behavior if single-tap/
double-tap interrupts are enabled.
The THRESH_INACT register is eight bits and holds the threshold
value for detecting inactivity. The data format is unsigned, so
the magnitude of the inactivity event is compared with the value
in the THRESH_INACT register. The scale factor is 62.5 mg/LSB.
A value of 0 may result in undesirable behavior if the inactivity
interrupt is enabled.
D7
1
D6
1
D5
1
D4
0
D3
0
D2
1
D1
1
D0
0
Register 0x1E, Register 0x1F, Register 0x20—OFSX,
OFSY, OFSZ (Read/Write)
The OFSX, OFSY, and OFSZ registers are each eight bits and
offer user-set offset adjustments in twos complement format
with a scale factor of 15.6 mg/LSB (that is, 0x7F = 2 g). The
values stored in the offset registers are automatically added to
the acceleration data, and the resulting value is stored in the
output data registers. For additional information regarding
offset calibration and the use of the offset registers, refer to the
Offset Calibration section.
Register 0x21—DUR (Read/Write)
The DUR register is eight bits and contains an unsigned time
value representing the maximum time that an event must be
above the THRESH_TAP threshold to qualify as a tap event. For
information on improved tap detection, refer to the Improved Tap
Detection section. The scale factor is 625 μs/LSB. A value of 0
disables the single-tap/double-tap functions.
Register 0x22—Latent (Read/Write)
The latent register is eight bits and contains an unsigned time
value representing the wait time from the detection of a tap
event to the start of the time window (defined by the window
register) during which a possible second tap event can be detected.
For information on improved tap detection, refer to the Improved
Tap Detection section. The scale factor is 1.25 ms/LSB. A value of 0
disables the double-tap function.
Register 0x23—Window (Read/Write)
The window register is eight bits and contains an unsigned time
value representing the amount of time after the expiration of the
latency time (determined by the latent register) during which a
second valid tap can begin. For information on improved tap
detection, refer to the Improved Tap Detection section. The scale
factor is 1.25 ms/LSB. A value of 0 disables the double-tap
function.
Register 0x26—TIME_INACT (Read/Write)
The TIME_INACT register is eight bits and contains an unsigned
time value representing the amount of time that acceleration
must be less than the value in the THRESH_INACT register for
inactivity to be declared. The scale factor is 1 sec/LSB. Unlike
the other interrupt functions, which use unfiltered data (see the
Threshold section), the inactivity function uses filtered output
data. At least one output sample must be generated for the
inactivity interrupt to be triggered. This results in the function
appearing unresponsive if the TIME_INACT register is set to a
value less than the time constant of the output data rate. A value
of 0 results in an interrupt when the output data is less than the
value in the THRESH_INACT register.
Register 0x27—ACT_INACT_CTL (Read/Write)
D7
ACT ac/dc
D3
INACT ac/dc
D6
ACT_X enable
D2
INACT_X enable
D5
ACT_Y enable
D1
INACT_Y enable
D4
ACT_Z enable
D0
INACT_Z enable
ACT AC/DC and INACT AC/DC Bits
A setting of 0 selects dc-coupled operation, and a setting of 1
enables ac-coupled operation. In dc-coupled operation, the
current acceleration magnitude is compared directly with
THRESH_ACT and THRESH_INACT to determine whether
activity or inactivity is detected.
In ac-coupled operation for activity detection, the acceleration
value at the start of activity detection is taken as a reference
value. New samples of acceleration are then compared to this
reference value, and if the magnitude of the difference exceeds
the THRESH_ACT value, the device triggers an activity interrupt.
Similarly, in ac-coupled operation for inactivity detection, a
reference value is used for comparison and is updated whenever
the device exceeds the inactivity threshold. After the reference
value is selected, the device compares the magnitude of the
difference between the reference value and the current acceleration
with THRESH_INACT. If the difference is less than the value in
THRESH_INACT for the time in TIME_INACT, the device is
considered inactive and the inactivity interrupt is triggered.
Rev. 0 | Page 23 of 40
ADXL346
Register 0x2B—ACT_TAP_STATUS (Read Only)
ACT_x Enable Bits and INACT_x Enable Bits
A setting of 1 enables x-, y-, or z-axis participation in detecting
activity or inactivity. A setting of 0 excludes the selected axis from
participation. If all axes are excluded, the function is disabled.
For activity detection, all participating axes are logically OR’ed,
causing the activity function to trigger when any of the participating axes exceeds the threshold. For inactivity detection, all
participating axes are logically AND’ed, causing the inactivity
function to trigger only if all participating axes are below the
threshold for the specified period of time.
Register 0x28—THRESH_FF (Read/Write)
The THRESH_FF register is eight bits and holds the threshold
value, in unsigned format, for free-fall detection. The acceleration
on all axes is compared with the value in THRESH_FF to determine if a free-fall event occurred. The scale factor is 62.5 mg/LSB.
Note that a value of 0 mg may result in undesirable behavior if
the free-fall interrupt is enabled. Values between 300 mg and
600 mg (0x05 to 0x09) are recommended.
Register 0x29—TIME_FF (Read/Write)
The TIME_FF register is eight bits and stores an unsigned time
value representing the minimum time that the value of all axes
must be less than THRESH_FF to generate a free-fall interrupt.
The scale factor is 5 ms/LSB. A value of 0 may result in undesirable
behavior if the free-fall interrupt is enabled. Values between 100 ms
and 350 ms (0x14 to 0x46) are recommended.
Register 0x2A—TAP_AXES (Read/Write)
D7
0
D6
0
D5
0
D4
Improved
tap
D3
Suppress
D2
TAP_X
enable
D1
TAP_Y
enable
D0
TAP_Z
enable
Improved Tap Bit
The improved tap bit is used to enable improved tap detection.
This mode of operation improves tap detection by performing
an ac-coupled differential comparison of the output acceleration
data. The improved tap detection is performed on the same output
data available in the DATAX, DATAY, and DATAZ registers. Due
to the dependency on the output data rate and the ac-coupled
differential measurement, the threshold and timing values for
single taps and double taps must be adjusted for improved tap
detection. For further explanation of improved tap detection, see
the Improved Tap Detection section. Improved tap is enabled
by setting the improved tap bit to a value of 1 and is disabled
by clearing the bit to a value of 0.
Suppress Bit
Setting the suppress bit suppresses double-tap detection if
acceleration greater than the value in THRESH_TAP is present
between taps. See the Tap Detection section for more details.
D7
0
D6
ACT_X
source
D5
ACT_Y
source
D4
ACT_Z
source
D3
Asleep
D2
TAP_X
source
D1
TAP_Y
source
D0
TAP_Z
source
ACT_x Source and TAP_x Source Bits
These bits indicate the first axis involved in a tap or activity
event. A setting of 1 corresponds to involvement in the event,
and a setting of 0 corresponds to no involvement. When new
data is available, these bits are not cleared but are overwritten by
the new data. The ACT_TAP_STATUS register should be read
before clearing the interrupt. Disabling an axis from participation
clears the corresponding source bit when the next activity or
single-tap/double-tap event occurs.
Asleep Bit
A setting of 1 in the asleep bit indicates that the part is asleep,
and a setting of 0 indicates that the part is not asleep. This bit
toggles only if the device is configured for autosleep. See the
Register 0x2D—POWER_CTL (Read/Write) section for more
information on autosleep mode.
Register 0x2C—BW_RATE (Read/Write)
D7
0
D6
0
D5
0
D4
LOW_POWER
D3
D2
D1
Rate
D0
LOW_POWER Bit
A setting of 0 in the LOW_POWER bit selects normal operation,
and a setting of 1 selects reduced power operation, which is
associated with somewhat higher noise (see the Power Modes
section for details).
Rate Bits
These bits select the device bandwidth and output data rate (see
Table 7 and Table 8 for details). The default value is 0x0A, which
translates to a 100 Hz output data rate. An output data rate should
be selected that is appropriate for the communication protocol and
frequency selected. Selecting too high of an output data rate with a
low communication speed results in samples being discarded.
Register 0x2D—POWER_CTL (Read/Write)
D7
0
D6
0
D5
Link
D4
AUTO_SLEEP
D3
Measure
D2
Sleep
D1 D0
Wakeup
Link Bit
A setting of 1 in the link bit with both the activity and inactivity
functions enabled delays the start of the activity function until
inactivity is detected. After activity is detected, inactivity detection
begins, preventing the detection of activity. This bit serially links
the activity and inactivity functions. When this bit is set to 0,
the inactivity and activity functions are concurrent. Additional
information can be found in the Link Mode section.
TAP_x Enable Bits
A setting of 1 in the TAP_X enable, TAP_Y enable, or TAP_Z
enable bit enables x-, y-, or z-axis participation in tap detection.
A setting of 0 excludes the selected axis from participation in
tap detection.
Rev. 0 | Page 24 of 40
ADXL346
When clearing the link bit, it is recommended that the part be
placed into standby mode and then set back to measurement
mode with a subsequent write. This is done to ensure that the
device is properly biased if sleep mode is manually disabled;
otherwise, the first few samples of data after the link bit is cleared
may have additional noise, especially if the device was asleep
when the bit was cleared.
AUTO_SLEEP Bit
If the link bit is set, a setting of 1 in the AUTO_SLEEP bit enables
the autosleep functionality. In this mode, the ADXL346 automatically switches to sleep mode if the inactivity function is
enabled and inactivity is detected (that is, when acceleration is
below the THRESH_INACT value for at least the time indicated
by TIME_INACT). If activity is also enabled, the ADXL346
automatically wakes up from sleep after detecting activity and
returns to operation at the output data rate set in the BW_RATE
register. A setting of 0 in the AUTO_SLEEP bit disables automatic
switching to sleep mode. See the description of the sleep bit in this
section for more information on sleep mode.
If the link bit is not set, the AUTO_SLEEP feature is disabled,
and setting the AUTO_SLEEP bit does not have any impact on
device operation. Refer to the Link Bit section or the Link Mode
section for more information about using the link feature.
When clearing the AUTO_SLEEP bit, it is recommended that the
part be placed into standby mode and then set back to measurement mode with a subsequent write. This is done to ensure that
the device is properly biased if sleep mode is manually disabled;
otherwise, the first few samples of data after the AUTO_SLEEP
bit is cleared may have additional noise, especially if the device
was asleep when the bit was cleared.
Measure Bit
A setting of 0 in the measure bit places the part into standby mode,
and a setting of 1 places the part into measurement mode. The
ADXL346 powers up in standby mode with minimum power
consumption.
Sleep Bit
A setting of 0 in the sleep bit puts the part into the normal mode
of operation, and a setting of 1 places the part into sleep mode.
Sleep mode suppresses DATA_READY, stops transmission of data
to FIFO, and switches the sampling rate to one specified by the
wakeup bits. In sleep mode, only the activity function can be used.
While the DATA_READY interrupt is suppressed, the output
data registers are still updated at the sampling rate set by the
wakeup bits.
When clearing the sleep bit, it is recommended that the part be
placed into standby mode and then set back to measurement
mode with a subsequent write. This is done to ensure that the
device is properly biased if sleep mode is manually disabled;
otherwise, the first few samples of data after the sleep bit is
cleared may have additional noise, especially if the device was
asleep when the bit was cleared.
Wakeup Bits
These bits control the frequency of readings in sleep mode as
described in Table 20.
Table 20. Frequency of Readings in Sleep Mode
D1
0
0
1
1
Setting
D0
0
1
0
1
Frequency (Hz)
8
4
2
1
Register 0x2E—INT_ENABLE (Read/Write)
D7
DATA_READY
D3
Inactivity
D6
SINGLE_TAP
D2
FREE_FALL
D5
DOUBLE_TAP
D1
Watermark
D4
Activity
D0
Overrun/
orientation
Setting bits in this register to a value of 1 enables their respective
functions to generate interrupts, whereas a value of 0 prevents
the functions from generating interrupts. The DATA_READY,
watermark, and overrun/orientation bits enable only the interrupt
output; the functions are always enabled. It is recommended that
interrupts be configured before enabling their outputs.
Register 0x2F—INT_MAP (Read/Write)
D7
DATA_READY
D3
Inactivity
D6
SINGLE_TAP
D2
FREE_FALL
D5
DOUBLE_TAP
D1
Watermark
D4
Activity
D0
Overrun/
orientation
Bits set to 0 in this register send their respective interrupts to the
INT1 pin, whereas bits set to 1 send their respective interrupts to
the INT2 pin. All selected interrupts for a given pin are OR’ed.
Register 0x30—INT_SOURCE (Read Only)
D7
DATA_READY
D3
Inactivity
D6
SINGLE_TAP
D2
FREE_FALL
D5
DOUBLE_TAP
D1
Watermark
D4
Activity
D0
Overrun/
orientation
Bits set to 1 in this register indicate that their respective functions
have triggered an event, whereas bits set to 0 indicate that the
corresponding events have not occurred. The DATA_READY,
watermark, and overrun/orientation bits are always set if the
corresponding events occur, regardless of the INT_ENABLE
register settings, and are cleared by reading data from the
DATAX, DATAY, and DATAZ registers. The DATA_READY and
watermark bits may require multiple reads, as indicated in the
FIFO mode descriptions in the FIFO section. Other bits, and the
corresponding interrupts, including orientation if enabled, are
cleared by reading the INT_SOURCE register.
Rev. 0 | Page 25 of 40
ADXL346
Register 0x31—DATA_FORMAT (Read/Write)
D7
SELF_TEST
D6
SPI
D5
INT_INVERT
D4
0
D3
FULL_RES
D2
Justify
D1 D0
Range
The DATA_FORMAT register controls the presentation of data
to Register 0x32 through Register 0x37. All data, except that for
the ±16 g range, must be clipped to avoid rollover.
and DATAx1 as the most significant byte, where x represents X,
Y, or Z. The DATA_FORMAT register (Address 0x31) controls
the format of the data. It is recommended that a multiple-byte
read of all registers be performed to prevent a change in data
between reads of sequential registers.
Register 0x38—FIFO_CTL (Read/Write)
SELF_TEST Bit
D7
D6
FIFO_MODE
A setting of 1 in the SELF_TEST bit applies a self-test force to
the sensor, causing a shift in the output data. A value of 0 disables
the self-test force.
FIFO_MODE Bits
SPI Bit
D5
Trigger
D4
D3
D2
D1
Samples
D0
These bits set the FIFO mode, as described in Table 22.
Table 22. FIFO Modes
A value of 1 in the SPI bit sets the device to 3-wire SPI mode,
and a value of 0 sets the device to 4-wire SPI mode.
INT_INVERT Bit
A value of 0 in the INT_INVERT bit sets the interrupts to active
high, and a value of 1 sets the interrupts to active low.
FULL_RES Bit
When this bit is set to a value of 1, the device is in full resolution
mode, where the output resolution increases with the g range
set by the range bits to maintain a 4 mg/LSB scale factor. When
the FULL_RES bit is set to 0, the device is in 10-bit mode, and
the range bits determine the maximum g range and scale factor.
Setting
D7
D6
0
0
0
1
Mode
Bypass
FIFO
1
0
Stream
1
1
Trigger
Justify Bit
A setting of 1 in the justify bit selects left-justified (MSB) mode,
and a setting of 0 selects right-justified mode with sign extension.
Function
FIFO is bypassed.
FIFO collects up to 32 values and then
stops collecting data, collecting new
data only when FIFO is not full.
FIFO holds the last 32 data values.
When FIFO is full, the oldest data is
overwritten with newer data.
When triggered by the trigger bit,
FIFO holds the last data samples
before the trigger event and then
continues to collect data until FIFO is
full. New data is collected only when
FIFO is not full.
Trigger Bit
Range Bits
A value of 0 in the trigger bit links the trigger event of trigger mode
to INT1, and a value of 1 links the trigger event to INT2.
These bits set the g range as described in Table 21.
Samples Bits
Table 21. g Range Setting
The function of these bits depends on the FIFO mode selected (see
Table 23). Entering a value of 0 in the samples bits immediately sets
the watermark bit in the INT_SOURCE register (Address 0x30),
regardless of which FIFO mode is selected. Undesirable operation
may occur if a value of 0 is used for the samples bits when trigger
mode is used.
D1
0
0
1
1
Setting
D0
0
1
0
1
g Range
±2 g
±4 g
±8 g
±16 g
Table 23. Samples Bits Functions
Register 0x32 to Register 0x37—DATAX0, DATAX1,
DATAY0, DATAY1, DATAZ0, DATAZ1 (Read Only)
These six bytes (Register 0x32 to Register 0x37) are eight bits
each and hold the output data for each axis. Register 0x32 and
Register 0x33 hold the output data for the x-axis, Register 0x34 and
Register 0x35 hold the output data for the y-axis, and Register 0x36
and Register 0x37 hold the output data for the z-axis. The output
data is twos complement, with DATAx0 as the least significant byte
FIFO Mode
Bypass
FIFO
Stream
Trigger
Rev. 0 | Page 26 of 40
Samples Bits Function
None.
Specifies how many FIFO entries are needed to
trigger a watermark interrupt.
Specifies how many FIFO entries are needed to
trigger a watermark interrupt.
Specifies how many FIFO samples are retained in
the FIFO buffer before a trigger event.
ADXL346
Register 0x39—FIFO_STATUS (Read Only)
D7
FIFO_TRIG
D6
0
D5
D4
D3
D2
Entries
D1
D0
FIFO_TRIG Bit
A 1 in the FIFO_TRIG bit corresponds to a trigger event occurring,
and a 0 means that a FIFO trigger event has not occurred.
Entries Bits
These bits report how many data values are stored in FIFO.
Access to collect the data from FIFO is provided through the
DATAX, DATAY, and DATAZ registers. FIFO reads must be
done in burst or multiple-byte mode because each FIFO level is
cleared after any read (single- or multiple-byte) of FIFO. FIFO
stores a maximum of 32 entries, which equates to a maximum
of 33 entries available at any given time because an additional
entry is available at the output filter of the device.
Register 0x3A—TAP_SIGN (Read Only)
D7
0
D6
XSIGN
D5
YSIGN
D4
ZSIGN
D3
0
D2
XTAP
D1
YTAP
D0
ZTAP
xSIGN Bits
These bits indicate the sign of the first axis involved in a tap
event. A setting of 1 corresponds to acceleration in the negative
direction, and a setting of 0 corresponds to acceleration in the
positive direction. These bits update only when a new singletap/double-tap event is detected, and only the axes enabled in the
TAP_AXES register (Address 0x2A) are updated. The TAP_SIGN
register should be read before clearing the interrupt. See the Tap
Sign section for more details.
xTAP Bits
These bits indicate the first axis involved in a tap event. A
setting of 1 corresponds to involvement in the event, and a
setting of 0 corresponds to no involvement. When new data is
available, these bits are not cleared but are overwritten by the
new data. The TAP_SIGN register should be read before clearing
the interrupt. Disabling an axis from participation clears the
corresponding source bit when the next single-tap/double-tap
event occurs.
Register 0x3B—ORIENT_CONF (Read/Write)
D7
INT_
ORIENT
D6
D5
Dead zone
D4
D3
INT_
3D
D2
D1
D0
Divisor
An orientation interrupt is generated whenever the orientation
status for the mode selected by the INT_3D bit changes in the
orient register (Address 0x3C). The orientation interrupt is
cleared by reading the INT_SOURCE register. Clearing the
INT_ORIENT bit, or the orientation bit in the INT_ENABLE
register (Address 0x2E), disables and clears the interrupt.
Writing to the BW_RATE register (Address 0x2C) or placing
the part into standby mode resets the orientation feature, clearing
the orientation filter and the interrupt. However, resetting the
orientation feature also resets the orientation status in the orient
register (Address 0x3C) and, therefore, causes an interrupt to be
generated when the next output sample is available if the present
orientation is not the default orientation. A value of 0 for the
INT_ORIENT bit disables generation of the orientation interrupt
and permits the use of the overrun function.
Dead Zone Bits
These bits determine the region between two adjacent orientations,
where the orientation is considered invalid and is not updated. A
value of 0 may result in undesirable behavior when the orientation
is close to the bisector between two adjacent regions. The dead zone
angle is determined by these bits, as described in Table 24. See the
Orientation Sensing section for more details.
Table 24. Dead Zone and Divisor Codes
Decimal
0
1
2
3
4
5
6
7
Binary
000
001
010
011
100
101
110
111
Dead Zone Angle
(Degrees)
5.1
10.2
15.2
20.4
25.5
30.8
36.1
41.4
Divisor
Bandwidth (Hz)
ODR/9
ODR/22
ODR/50
ODR/100
ODR/200
ODR/400
ODR/800
ODR/1600
INT_3D Bit
If the orientation interrupt is enabled, the INT_3D bit determines
whether 2D or 3D orientation detection generates an interrupt.
A value of 0 generates an interrupt only if the 2D orientation
changes from a valid 2D orientation to a different valid 2D
orientation. A value of 1 generates an interrupt only if the 3D
orientation changes from a valid 3D orientation to a different
valid 3D orientation.
INT_ORIENT Bit
Divisor Bits
Setting the INT_ORIENT bit enables the orientation interrupt.
A value of 1 overrides the overrun function of the device and
replaces overrun in the INT_MAP (Address 0x2F), INT_ENABLE
(Address 0x2E), and INT_SOURCE (Address 0x30) registers with
the orientation function. After setting the INT_ORIENT bit, the
orientation bits in the INT_MAP and INT_ENABLE registers must
be configured to map the orientation interrupt to INT1 or INT2
and to enable generation of the interrupt to the pin.
These bits set the bandwidth of the filter used to low-pass filter the
measured acceleration for stable orientation sensing. The divisor
bandwidth is determined by these bits, as detailed in Table 24,
where ODR is the output data rate set in the BW_RATE register
(Address 0x2C). See the Orientation Sensing section for more
details.
Rev. 0 | Page 27 of 40
ADXL346
Register 0x3C—Orient (Read Only)
D7
0
D6
V2
D5
D4
2D_ORIENT
D3
V3
D2
D1
D0
3D_ORIENT
Vx Bits
These bits show the validity of the 2D (V2) and 3D (V3) orientations. A value of 1 corresponds to the orientation being valid. A
value of 0 means that the orientation is invalid because the current
orientation is in the dead zone.
xD_ORIENT Bits
These bits represent the current 2D (2D_ORIENT) and 3D
(3D_ORIENT) orientations of the accelerometer. If the orientation interrupt is enabled, this register is read to determine the
orientation of the device when the interrupt occurs. Because this
register updates with each new sample of acceleration data, it
should be read at the time of the orientation interrupt to ensure
that the orientation change that caused the interrupt has been
identified. Orientation values are shown in Table 25 and Table 26.
See the Orientation Sensing section for more details.
Writing to the BW_RATE register (Address 0x2C) or placing
the part into standby mode resets the orientation feature, clearing
the orientation filter and the orientation status. An orientation
interrupt (if enabled) results from these actions if the orientation
during the next output sample is different from the default
value (+X for 2D orientation detection and undefined for 3D
orientation).
Table 25. 2D Orientation Codes
Decimal
0
1
2
3
Binary
00
01
10
11
Orientation
Portrait positive
Portrait negative
Landscape positive
Landscape negative
Dominant Axis
+X
−X
+Y
−Y
Table 26. 3D Orientation Codes
Decimal
3
4
2
5
1
6
Rev. 0 | Page 28 of 40
Binary
011
100
010
101
001
110
Orientation
Front
Back
Left
Right
Top
Bottom
Dominant Axis
+X
−X
+Y
−Y
+Z
−Z
ADXL346
APPLICATIONS INFORMATION
POWER SUPPLY DECOUPLING
TAP DETECTION
A 1 μF tantalum capacitor (CS) at VS and a 0.1 μF ceramic capacitor
(CI/O) at VDD I/O placed close to the ADXL346 supply pins is
recommended to adequately decouple the accelerometer from
noise on the power supply. If additional decoupling is necessary,
a resistor or ferrite bead, no larger than 100 Ω, in series with VS
may be helpful. Additionally, increasing the bypass capacitance
on VS to a 10 μF tantalum capacitor in parallel with a 0.1 μF
ceramic capacitor may also improve noise.
The tap interrupt function is capable of detecting either single
or double taps. The following parameters are shown in Figure 45
for a valid single-tap event and a valid double-tap event:
Care should be taken to ensure that the connection from the
ADXL346 ground to the power supply ground has low impedance
because noise transmitted through ground has an effect similar
to noise transmitted through VS. It is recommended that VS and
VDD I/O be separate supplies to minimize digital clock noise on
the VS supply. If this is not possible, additional filtering of the
supplies as previously mentioned may be necessary.
VS
•
The tap detection threshold is defined by the THRESH_TAP
register (Address 0x1D).
The maximum tap duration time is defined by the DUR
register (Address 0x21).
The tap latency time is defined by the latent register
(Address 0x22) and is the waiting period from the end of
the first tap until the start of the time window when a
second tap can be detected, which is determined by the
value in the window register (Address 0x23).
The interval after the latency time (set by the latent register) is
defined by the window register. Although a second tap must
begin after the latency time has expired, it need not finish
before the end of the time defined by the window register.
•
•
•
VDD I/O
CS
CI/O
FIRST TAP
GND
CS
3-WIRE OR
4-WIRE SPI
OR I2C
INTERFACE
08167-035
SDA/SDI/SDIO
INT1 SDO/ALT ADDRESS
SCL/SCLK
INT2
XHI BW
ADXL346
INTERRUPT
CONTROL
SECOND TAP
VDD I/O
TIME LIMIT FOR
TAPS (DUR)
Figure 43. Applications Diagram
The ADXL346 should be mounted on the PCB in a location
close to a hard mounting point of the PCB to the case. Mounting
the ADXL346 at an unsupported PCB location, as shown in
Figure 44, may result in large, apparent measurement errors due
to undampened PCB vibration. Locating the accelerometer near
a hard mounting point ensures that any PCB vibration at the
accelerometer is above the accelerometer’s mechanical sensor
resonant frequency and, therefore, effectively invisible to the
accelerometer. Multiple mounting points, close to the sensor,
and/or a thicker PCB also help to reduce the effect of system
resonance on the performance of the sensor.
ACCELEROMETERS
MOUNTING POINTS
08167-036
PCB
LATENCY
TIME
(LATENT)
INTERRUPTS
MECHANICAL CONSIDERATIONS FOR MOUNTING
THRESHOLD
(THRESH_TAP)
TIME WINDOW FOR
SECOND TAP (WINDOW)
SINGLE-TAP
INTERRUPT
DOUBLE-TAP
INTERRUPT
08167-037
VS
Figure 45. Tap Interrupt Function with Valid Single and Double Taps
If only the single-tap function is in use, the single-tap interrupt
is triggered when the acceleration goes below the threshold, as
long as DUR has not been exceeded. If both single and doubletap functions are in use, the single-tap interrupt is triggered
when the double-tap event has been either validated or
invalidated.
Several events can occur to invalidate the second tap of a
double-tap event. First, if the suppress bit in the TAP_AXES
register (Address 0x2A) is set, any acceleration spike above the
threshold during the latency time (set by the latent register)
invalidates the double-tap detection, as shown in Figure 46.
Figure 44. Incorrectly Placed Accelerometers
Rev. 0 | Page 29 of 40
ADXL346
LATENCY
TIME (LATENT)
TIME WINDOW FOR SECOND
TAP (WINDOW)
Figure 46. Double-Tap Event Invalid Due to High g Event
When the Suppress Bit Is Set
A double-tap event can also be invalidated if acceleration above
the threshold is detected at the start of the time window for the
second tap (set by the window register (Address 0x23)). This results
in an invalid double tap at the start of this window, as shown in
Figure 47. Additionally, a double-tap event can be invalidated if
an acceleration exceeds the time limit for taps (set by the DUR
register (Address 0x21)), resulting in an invalid double tap at
the end of the DUR time limit for the second tap event, also
shown in Figure 47.
INVALIDATES DOUBLE TAP
AT START OF WINDOW
XHI BW
LATENCY
TIME
(LATENT)
TIME WINDOW FOR
SECOND TAP (WINDOW)
IMPROVED TAP DETECTION
Improved tap detection is enabled by setting the improved tap
bit of the TAP_AXES register (Address 0x2A). When improved
tap detection is enabled, the filtered output data corresponding to
the output data rate set in the BW_RATE register (Address 0x2C)
is processed to determine if a tap event occurred. In addition, an
ac-coupled differential measurement is used. This results in the
timing values and threshold values for improved tap detection
being different from those used for normal tap detection.
TAP SIGN
08167-039
XHI BW
TIME LIMIT
FOR TAPS
(DUR)
INVALIDATES
DOUBLE TAP AT
END OF DUR
After a tap interrupt has been received, the first axis to exceed
the THRESH_TAP level is reported in the ACT_TAP_STATUS
register (Address 0x2B). This register is never cleared but is
overwritten with new data.
When improved tap detection is used, new values must be
determined based on test results. In general, no timing values
(in the DUR, latent, or window registers) should be set that are
less than the time step resolution set by the output data rate.
The threshold value for improved tap detection can typically be
set much lower than the threshold for normal tap detection.
The value used depends on the value in the BW_RATE register
and should be determined through system testing. Refer to the
Threshold section for more details.
TIME LIMIT
FOR TAPS
(DUR)
TIME LIMIT
FOR TAPS
(DUR)
DUR, latent, window, and THRESH_TAP registers is required.
In general, a good starting point is to set the DUR register to a
value greater than 0x10 (10 ms), the latent register to a value greater
than 0x10 (20 ms), the window register to a value greater than
0x40 (80 ms), and the THRESH_TAP register to a value greater
than 0x30 (3 g). Setting a very low value in the latent, window, or
THRESH_TAP register may result in an unpredictable response
due to the accelerometer picking up echoes of the tap inputs.
Figure 47. Tap Interrupt Function with Invalid Double Taps
Single taps, double taps, or both can be detected by setting the
respective bits in the INT_ENABLE register (Address 0x2E).
Control over participation of each of the three axes in single-tap/
double-tap detection is exerted by setting the appropriate bits in
the TAP_AXES register (Address 0x2A). For the double-tap
function to operate, both the latent and window registers must
be set to a nonzero value.
A negative sign is produced by experiencing a negative acceleration, which corresponds to tapping on the positive face of the
device for the desired axis. The positive face of the device is the
face such that movement in that direction is positive acceleration.
For example, tapping on the face that corresponds to the +X
direction, labeled as front in Figure 48, results in a negative sign
for the x-axis. Tapping on the face labeled as left in Figure 48
results in a negative sign for the y-axis, and tapping on the face
labeled top results in a negative sign for the z-axis. Conversely,
tapping on the back, right, or bottom sides results in positive
signs for the corresponding axes.
Every mechanical system has somewhat different single-tap/
double-tap responses based on the mechanical characteristics of
the system. Therefore, some experimentation with values for the
+z
TOP
(+Z)
+y
LEFT
(+Y)
+x
FRONT
(+X)
08167-046
TIME LIMIT
FOR TAPS
(DUR)
08167-038
XHI BW
INVALIDATES DOUBLE TAP IF
SUPRESS BIT IS SET
Figure 48. 3D Orientation with Coordinate System
Rev. 0 | Page 30 of 40
ADXL346
THRESHOLD
The lower output data rates are achieved by decimating a
common sampling frequency inside the device. The activity,
free-fall, and single-tap/double-tap detection functions without
improved tap enabled are performed using undecimated data.
As the bandwidth of the output data varies with the data rate
and is lower than the bandwidth of the undecimated data, the
high frequency and high g data that is used to determine activity,
free-fall, and single-tap/double-tap events may not be present if
the output of the accelerometer is examined. This may result in
functions triggering when acceleration data does not appear to
meet the conditions set by the user for the corresponding function.
LINK MODE
The function of the link bit is to reduce the number of activity
interrupts that the processor must service by setting the device
to look for activity only after inactivity. For proper operation of
this feature, the processor must still respond to the activity and
inactivity interrupts by reading the INT_SOURCE register
(Address 0x30) and, therefore, clearing the interrupts. If an activity
interrupt is not cleared, the part cannot go into autosleep mode.
The asleep bit in the ACT_TAP_STATUS register (Address 0x2B)
indicates whether the part is asleep.
SLEEP MODE VS. LOW POWER MODE
In applications where a low data rate and low power consumption
are desired (at the expense of noise performance), it is recommended that low power mode be used. The use of low power
mode preserves the functionality of the DATA_READY interrupt
and the FIFO for postprocessing of the acceleration data. Sleep
mode, while offering a low data rate and power consumption, is
not intended for data acquisition.
However, when sleep mode is used in conjunction with the
AUTO_SLEEP mode and the link mode, the part can automatically
switch to a low power, low sampling rate mode when inactivity
is detected. To prevent the generation of redundant inactivity
interrupts, the inactivity interrupt is automatically disabled
and activity is enabled. When the ADXL346 is in sleep mode, the
host processor can also be placed into sleep mode or low power
mode to save significant system power. Once activity is detected,
the accelerometer automatically switches back to the original
data rate of the application and provides an activity interrupt
that can be used to wake up the host processor. Similar to when
inactivity occurs, detection of activity events is disabled and
inactivity is enabled.
OFFSET CALIBRATION
Accelerometers are mechanical structures containing elements
that are free to move. These moving parts can be very sensitive
to mechanical stresses, much more so than solid-state electronics.
The 0 g bias or offset is an important accelerometer metric because
it defines the baseline for measuring acceleration. Additional
stresses can be applied during assembly of a system containing
an accelerometer. These stresses can come from, but are not
limited to, component soldering, board stress during mounting,
and application of any compounds on or over the component. If
calibration is deemed necessary, it is recommended that calibration
be performed after system assembly to compensate for these effects.
A simple method of calibration is to measure the offset while
assuming that the sensitivity of the ADXL346 is as specified in
Table 1. The offset can then be automatically accounted for by
using the built-in offset registers (Register 0x1E, Register 0x1F, and
Register 0x20). This results in the data acquired from the DATAX,
DATAY, and DATAZ registers (Address 0x32 to Address 0x37)
already compensating for any offset.
In a no-turn or single-point calibration scheme, the part is oriented
such that one axis, typically the z-axis, is in the 1 g field of gravity
and the remaining axes, typically the x- and y-axes, are in a 0 g
field. The output is then measured by taking the average of a
series of samples. The number of samples averaged is a choice of
the system designer, but a recommended starting point is 0.1 sec
worth of data for data rates of 100 Hz or greater. This corresponds
to 10 samples at the 100 Hz data rate. For data rates of less than
100 Hz, it is recommended that at least 10 samples be averaged
together. These values are stored as X0g, Y0g, and Z+1g for the 0 g
measurements on the x- and y-axes and the 1 g measurement
on the z-axis, respectively.
The values measured for X0g and Y0g correspond to the offset of
the x- and y-axes, and compensation is done by subtracting those
values from the output of the accelerometer to obtain the actual
acceleration:
XACTUAL = XMEAS − X0g
YACTUAL = YMEAS − Y0g
Because the z-axis measurement is done in a 1 g field, a no-turn or
single-point calibration scheme assumes an ideal sensitivity, SZ,
for the z-axis. This is subtracted from Z+1g to attain the z-axis
offset, which is then subtracted from future measured values to
obtain the actual value:
Z0g = Z1g − SZ
ZACTUAL = ZMEAS − Z0g
The ADXL346 can automatically compensate the output for offset
by using the offset registers (Register 0x1E, Register 0x1F, and
Register 0x20). These registers contain an 8-bit, twos complement
value that is automatically added to all measured acceleration
values, and the result is then placed into the DATAX, DATAY,
and DATAZ registers. Because the value placed in an offset register
is additive, a negative value is placed into the register to eliminate a
positive offset and vice versa for a negative offset. The register
has a scale factor of 15.6 mg/LSB and is independent of the
selected g range.
As an example, assume that the ADXL346 is placed into fullresolution mode with a sensitivity of typically 256 LSB/g. The
part is oriented such that the z-axis is in the field of gravity and
the outputs of the x-, y-, and z-axes are measured as +10 LSB,
−13 LSB, and +9 LSB, respectively. Using the previous equations,
X0g is +10 LSB, Y0g is −13 LSB, and Z0g is +9 LSB. Each LSB of
Rev. 0 | Page 31 of 40
ADXL346
output in full-resolution is 3.9 mg or one-quarter of an LSB of
the offset register.
Because the offset register is additive, the 0 g values are negated
and rounded to the nearest LSB of the offset register:
XOFFSET = −Round(10/4) = −3 LSB
YOFFSET = −Round(−13/4) = 3 LSB
ZOFFSET = −Round(9/4) = −2 LSB
These values are programmed into the OFSX, OFSY, and OFXZ
registers, respectively, as 0xFD, 0x03, and 0xFE. As with all
registers in the ADXL346, the offset registers do not retain the
value written into them when power is removed from the part.
Power-cycling the ADXL346 returns the offset registers to their
default value of 0x00.
Because the no-turn or single-point calibration method assumes an
ideal sensitivity in the z-axis, any error in the sensitivity results in
offset error. For instance, if the actual sensitivity was 250 LSB/g
in the previous example, the offset would be 15 LSB, not 9 LSB.
To help minimize this error, an additional measurement point
can be used with the z-axis in a 0 g field, and the 0 g measurement
can be used in the ZACTUAL equation.
USING SELF-TEST
The self-test change is defined as the difference between the
acceleration output of an axis with self-test enabled and the
acceleration output of the same axis with self-test disabled (see
Endnote 7 of Table 1). This definition assumes that the sensor
does not move between these two measurements, because if the
sensor moves, a non–self-test related shift corrupts the test.
Proper configuration of the ADXL346 is also necessary for an
accurate self-test measurement. The part should be set with a data
rate greater than or equal to 100 Hz. This is done by ensuring that
a value greater than or equal to 0x0A is written into the rate bits
(Bit D3 through Bit D0) in the BW_RATE register (Address 0x2C).
The part also must be placed into normal power operation by
ensuring that the LOW_POWER bit (Bit D4) in the BW_RATE
register is cleared (LOW_POWER bit = 0) for accurate self-test
measurements. It is recommended that the part be set to fullresolution, 16 g mode to ensure that there is sufficient dynamic
range for the entire self-test shift. This is done by setting the
FULL_RES bit (Bit D3) and writing a value of 0x03 to the range
bits (Bit D1 and Bit D0) of the DATA_FORMAT register
(Address 0x31). This results in a high dynamic range for
measurement and a 3.9 mg/LSB scale factor.
After the part is configured for accurate self-test measurement,
several samples of acceleration data for the x-, y-, and z-axes
should be retrieved from the sensor and averaged together. The
number of samples averaged is a choice of the system designer,
but a recommended starting point is 0.1 sec worth of data for
data rates of 100 Hz or greater. This corresponds to 10 samples
at the 100 Hz data rate. For data rates of less than 100 Hz, it is
recommended that at least 10 samples be averaged together. The
averaged values should be stored and labeled appropriately as the
self-test disabled data, that is, XST_OFF, YST_OFF, and ZST_OFF.
Next, self-test should be enabled by setting Bit D7 of the
DATA_FORMAT register (Address 0x31). The output needs some
time (about four samples) to settle once self-test is enabled. After
allowing the output to settle, several samples of acceleration data
for the x-, y-, and z-axes should be taken again and averaged. It
is recommended that the same number of samples be taken for
this average as was previously taken. These averaged values should
again be stored and labeled appropriately as the value with selftest enabled, that is, XST_ON, YST_ON, and ZST_ON. Self-test can then
be disabled by clearing Bit D7 of the DATA_FORMAT register
(Address 0x31).
With the stored values for self-test enabled and disabled, the
self-test change is as follows:
XST = XST_ON − XST_OFF
YST = YST_ON − YST_OFF
ZST = ZST_ON − ZST_OFF
Because the measured output for each axis is expressed in LSBs,
XST, YST, and ZST are also expressed in LSBs. These values can be
converted to g’s of acceleration by multiplying each value by the
3.9 mg/LSB scale factor, if configured for full-resolution mode.
Additionally, Table 15 through Table 18 correspond to the self-test
range converted to LSBs and can be compared with the measured
self-test change when operating at a VS of 2.6 V. For other voltages,
the minimum and maximum self-test output values should be
adjusted based on (multiplied by) the scale factors shown in
Table 14. If the part was placed into ±2 g, 10-bit or full-resolution
mode, the values listed in Table 15 should be used. Although
the fixed 10-bit mode or a range other than 16 g can be used, a
different set of values, as indicated in Table 16 through Table 18,
would need to be used. Using a range below 8 g may result in
insufficient dynamic range and should be considered when
selecting the range of operation for measuring self-test.
If the self-test change is within the valid range, the test is considered
successful. Generally, a part is considered to pass if the minimum
magnitude of change is achieved. However, a part that changes
by more than the maximum magnitude is not necessarily a failure.
ORIENTATION SENSING
The orientation function of the ADXL346 reports both 2D
and 3D orientation concurrently through the orient register
(Address 0x3C). The V2 and V3 bits (Bit D6 and Bit D3 in the
orient register) report the validity of the 2D and 3D orientation
codes. If V2 or V3 are set, their respective code is a valid
orientation. If V2 or V3 are cleared, the orientation of the
accelerometer is unknown, such as when the orientation is
within the dead zone between valid regions.
Rev. 0 | Page 32 of 40
ADXL346
PORTRAIT
For 2D orientation sensing, the relation of the x- and y-axes to
gravity is used to determine the accelerometer orientation (see
Figure 49 and Table 25). Portrait positive corresponds to the x-axis
being most closely aligned to the gravity vector and directed
upwards, opposite the gravity vector. Portrait negative is the
opposite of portrait positive, with the x-axis pointing downwards
along the gravity vector. Landscape positive corresponds to the
y-axis being most closely aligned with the gravity vector and
directed upwards, away from the gravity vector. Landscape
negative is the orientation opposite landscape positive. The
dead zone regions are shown in the orientations for portrait
positive (+X) and portrait negative (−X) of Figure 49. These
regions also exist for landscape positive (+Y) and landscape
negative (−Y), as shown in Figure 49.
The states shown in Table 26 correspond to which side of the
accelerometer is directed upwards, opposite the gravity vector.
As shown in Figure 48, the accelerometer is oriented in the top
state. If the device is flipped over such that the top of the device
is facing down, toward gravity, the orientation is reported as the
bottom state. If the device is adjusted such that the positive x-axis
or positive y-axis direction is pointing upwards, away from the
gravity vector, the accelerometer reports the orientation as front
or left, respectively.
The algorithm to detect orientation change is performed after
filtering the output acceleration data to eliminate the effects of
high frequency motion. This is performed by using a low-pass
filter with a bandwidth set by the divisor bits (ORIENT_CONF
register, Address 0x3B). The orientation filter uses the same
output data available in the output data registers (Address 0x32
to Address 0x37); therefore, the orient register (Address 0x3C)
is updated at the same rate as the data rate that is set in the
BW_RATE register (Address 0x2C). Because the output data
is used, the bandwidth of the orientation filter depends on the
value set in the BW_RATE register, and the divisor bandwidth
values in Table 24 are referenced to the selected output data rate.
To eliminate most human motion, such as walking or shaking, the
value in the divisor bits (Bits[D2:D0]) of the ORIENT_CONF
register (Address 0x3B) should be selected to effectively limit the
orientation bandwidth to 1 Hz or 2 Hz. For example, with an
output data rate of 100 Hz, a divisor selection of 3 (ODR/100)
results in a 1 Hz bandwidth for orientation detection. For best
results, it is recommended that an output data rate of ≥25 Hz in
normal power mode and ≥200 Hz in low power operation be used.
NEGATIVE (01)
DEADZONES
+X
+Y
+Y
+g
+g
+X
LANDSCAPE
POSITIVE (10)
+Y
NEGATIVE (11)
+X
+g
+X
08167-040
+g
+Y
Figure 49. 2D Orientation with Corresponding Codes
The width of the dead zone region between two orientation
positions is determined by setting the value of the dead zone bits
(Bits[D6:D4]) in the ORIENT_CONF register (Address 0x3B).
The dead zone region size can be specified as per the values
shown in Table 24. The dead zone angle represents the total
angle where the orientation is considered invalid. Therefore, a
dead zone of 15.4° corresponds to 7.7° in either direction away
from the bisector of two bordering regions. An example with a
dead zone region of 15.4° is shown in Figure 50. It should be
noted that the values shown in Table 24 correspond to the
typical dead zone angle when the gravity vector is completely
contained in only two axes (xy, xz, or yz) and should be used
only as a starting point. If the device is oriented such that the
projection of gravity onto all three axes is nonzero, the effective
sensitivity is reduced, causing an increase in the dead zone angle.
Therefore, evaluation needs to be performed for specific application uses to determine the optimal setting for the dead zone.
PORTRAIT
POSITIVE
52.7°
DEADZONE
45°
37.3°
+X
LANDSCAPE
POSITIVE
+Y
+g
08167-041
In 3D orientation, the z-axis is also included. If the accelerometer is
placed in a Cartesian coordinate system, as shown in Figure 48 of
the Tap Sign section, the top of the device corresponds to the
positive z-axis direction, the front of the device corresponds to
the positive x-axis direction, and the right side of the device
corresponds to the positive y-axis direction.
POSITIVE (00)
Figure 50. Orientation Showing a 15.4° Dead Zone Region
By setting the INT_ORIENT bit (Bit D7) of the ORIENT_CONF
register (Address 0x3B), an interrupt can be generated when the
device is placed into a new valid orientation. Only one mode of
orientation detection, 2D or 3D, can generate an interrupt at a
time. The orientation detection mode is selected by setting or
clearing the INT_3D bit (Bit D3) of the ORIENT_CONF register
(Address 0x3B). For more details, refer to the Register 0x3B—
ORIENT_CONF (Read/Write) section.
Rev. 0 | Page 33 of 40
ADXL346
the LSB of the output data-word is Bit D6 of the DATAx0 register.
In full-resolution operation when data is left justified, the location
of the LSB changes according to the selected output range. For a
range of ±2 g, the LSB is Bit D6 of the DATAx0 register; for ±4 g,
Bit D5 of the DATAx0 register; for ±8 g, Bit D4 of the DATAx0
register; and for ±16 g, Bit D3 of the DATAx0 register. This is
shown in Figure 52.
Writing to the BW_RATE register or placing the part into standby
mode resets the orientation feature, clearing the orientation filter
and the orientation status. These actions cause an orientation
interrupt (if enabled), however, if the orientation during the
next output sample is different from the default value (+X for
2D orientation detection and undefined for 3D orientation).
DATA FORMATTING OF UPPER DATA RATES
The use of 3200 Hz and 1600 Hz output data rates for fixed 10-bit
operation in the ±4 g, ±8 g, and ±16 g output ranges provides an
LSB that is valid and that changes according to the applied acceleration. Therefore, in these modes of operation, Bit D0 is not
always 0 when output data is right justified, and Bit D6 is not
always 0 when output data is left justified. Operation at any data
rate of 800 Hz or lower also provides a valid LSB in all ranges and
modes that change according to the applied acceleration.
Formatting of output data at the 3200 Hz and 1600 Hz output
data rates changes depending on the mode of operation (fullresolution or fixed 10-bit) and the selected output range.
When using the 3200 Hz or 1600 Hz output data rates in
full-resolution or ±2 g, 10-bit operation, the LSB of the output
data-word is always 0. When data is right justified, this corresponds
to Bit D0 of the DATAx0 register, as shown in Figure 51. When
data is left justified and the part is operating in ±2 g, 10-bit mode,
DATAx1 REGISTER
DATAx0 REGISTER
D7
D6
D5
D4
D3
D2
D1
D0
D7
D6
D5
D4
D3
D2
D1
D0
D7
D6
D5
D4
D3
D2
D1
D0
D7
D6
D5
D4
D3
D2
D1
0
OUTPUT DATA-WORD FOR
±16g, FULL-RESOLUTION MODE.
OUTPUT DATA-WORD FOR THE ±2g,
FULL-RESOLUTION AND ALL 10-BIT MODES.
08167-145
THE ±4g AND ±8g FULL-RESOLUTION MODES HAVE THE SAME LSB LOCATION AS THE ±2g
AND ±16g FULL-RESOLUTION MODES, BUT THE MSB LOCATION CHANGES TO BIT D2 AND
BIT D3 OF THE DATAx1 REGISTER FOR ±4g AND ±8g, RESPECTIVELY.
Figure 51. Data Formatting When Output Data Is Right Justified
DATAx1 REGISTER
DATAx0 REGISTER
D7
D6
D5
D4
D3
D2
D1
D0
D7
D6
D5
D4
D3
D2
D1
D0
D7
D6
D5
D4
D3
D2
D1
D0
D7
D6
D5
D4
D3
D2
D1
0
MSB FOR ALL MODES
OF OPERATION WHEN
LEFT JUSTIFIED.
LSB FOR ±2g, FULL-RESOLUTION
AND ALL 10-BIT MODES.
LSB FOR ±4g, FULL-RESOLUTION MODE.
LSB FOR ±8g, FULL-RESOLUTION MODE.
FOR 3200Hz AND 1600Hz OUTPUT DATA RATES, THE LSB IN THESE MODES IS ALWAYS 0.
ADDITIONALLY, ANY BITS TO THE RIGHT OF THE LSB ARE ALWAYS 0 WHEN THE OUTPUT
DATA IS LEFT JUSTIFIED.
Figure 52. Data Formatting When Output Data Is Left Justified
Rev. 0 | Page 34 of 40
08167-146
LSB FOR ±16g, FULL-RESOLUTION MODE.
ADXL346
10k
NOISE PERFORMANCE
X-AXIS
Y-AXIS
Z-AXIS
OUTPUT NOISE (LSB rms)
5
X-AXIS, NORMAL POWER
Y-AXIS, NORMAL POWER
Z-AXIS, NORMAL POWER
X-AXIS, LOW POWER
Y-AXIS, LOW POWER
Z-AXIS, LOW POWER
4
3
2
08167-147
1
0
3.13 6.25 12.50 25
50 100 200 400
OUTPUT DATA RATE (Hz)
1
10
100
AVERAGING PERIOD, (s)
1k
10k
X-AXIS
140
Y-AXIS
Z-AXIS
130
120
110
100
08167-149
PERCENTAGE OF NORMALIZED NOISE (%)
150
Figure 54 shows the typical Allan deviation for the ADXL346.
The 1/f corner of the device, as shown in this figure, is very low,
allowing absolute resolution of approximately 100 μg (assuming
that there is sufficient integration time). The figure also shows
that the noise density is 420 μg/√Hz for the x- and y-axes and
530 μg/√Hz for the z-axis.
6
0.1
Figure 54. Allan Deviation
The trend of noise performance for both normal power and low
power modes of operation of the ADXL346 is shown in Figure 53.
7
100
10
0.01
For low power operation (LOW_POWER bit = 1 in the BW_RATE
register, Address 0x2C), the noise of the ADXL346 is constant
for all valid data rates shown in Table 8. This value is typically
less than 2.83 LSB rms for the x- and y-axes and typically less
than 4.25 LSB rms for the z-axis.
Figure 55 shows the typical noise performance trend of the
ADXL346 over supply voltage. The performance is normalized
to the tested and specified supply voltage, VS = 2.6 V. The x-axis
offers the best noise performance over supply voltage, increasing by
typically less than 25% from nominal at a supply voltage of 1.8 V.
The performance of the y- and z-axes is comparable, with both
axes increasing by typically less than 35% when operating with a
supply voltage of 1.8 V. It should be noted, as shown in Figure 53,
that the noise on the z-axis is typically higher than that on the
y-axis; therefore, although the noise on the z- and y-axes change
roughly the same in percentage over supply voltage, the magnitude
of change on the z-axis is greater than the magnitude of change
on the y-axis.
1k
08167-148
ALLAN DEVIATION (µg)
The specification of noise shown in Table 1 corresponds to the
typical noise performance of the ADXL346 in normal power operation with an output data rate of 100 Hz (LOW_POWER bit = 0,
rate = 0x0A in the BW_RATE register, Address 0x2C). For normal
power operation at data rates below 100 Hz, the noise of the
ADXL346 is equivalent to the noise at 100 Hz ODR in LSBs. For
data rates greater than 100 Hz, the noise increases approximately by
a factor of √2 per doubling of the data rate. For example, at 400 Hz
ODR, the noise on the x- and y-axes is typically less than 2 LSB
rms, and the noise on the z-axis is typically less than 3 LSB rms.
90
1.6
1.8
2.0
2.2
2.4
2.6
2.8
SUPPLY VOLTAGE, VS (V)
Figure 55. Normalized Noise vs. Supply Voltage
OPERATION AT VOLTAGES OTHER THAN 2.6 V
The ADXL346 is tested and specified at a supply voltage of
VS = 2.6 V; however, it can be powered with a VS as high as 2.75 V
or as low as 1.7 V. Some performance parameters change as the
supply voltage changes, including the offset, sensitivity, noise,
self-test, and supply current.
Due to minuscule changes in the electrostatic forces as supply
voltage is varied, the offset and sensitivity change slightly. When
operating at a supply voltage of VS = 1.8 V, the offset of the x- and
y-axes is typically 25 mg higher than at VS = 2.6 V operation. The
z-axis is typically 20 mg lower when operating at a supply voltage
of 1.8 V than when operating at VS = 2.6 V. Sensitivity on the
x- and y-axes typically shifts from a nominal 256 LSB/g (fullresolution or ±2 g, 10-bit operation) at VS = 2.6 V operation to
250 LSB/g when operating with a supply voltage of 1.8 V. The z-axis
sensitivity is unaffected by a change in supply voltage and is the
same at VS = 1.8 V operation as it is at VS = 2.6 V operation. Simple
linear interpolation can be used to determine typical shifts in
offset and sensitivity at other supply voltages.
800 1600 3200
Figure 53. Noise vs. Output Data Rate for Normal and Low Power Modes,
Full Resolution (256 LSB/g)
Rev. 0 | Page 35 of 40
ADXL346
140
Changes in noise performance, self-test response, and supply
current are discussed elsewhere throughout the data sheet. For
more information about noise performance, review the Noise
Performance section. The Self-Test section discusses both the
operation of self-test over voltage (a square relationship with the
supply voltage) and the conversion of the self-test response in
g’s to LSBs. Finally, Figure 33 shows the impact of supply voltage
on typical current consumption at a 100 Hz output data rate,
with all other output data rates following the same trend.
NORMALIZED OUTPUT (LSB)
120
OFFSET PERFORMANCE AT LOWEST DATA RATES
80
0.10Hz
0.20Hz
0.39Hz
0.78Hz
1.56Hz
3.13Hz
6.25Hz
60
40
08167-056
20
0
25
35
45
55
65
75
85
TEMPERATURE (°C)
Figure 56. Typical X-Axis Output vs. Temperature at Lower Data Rates,
Normalized to 100 Hz Output Data Rate, VS = 2.6 V
140
120
NORMALIZED OUTPUT (LSB)
The ADXL346 offers several output data rates and bandwidths
designed for a wide range of applications. However, at the lowest
data rates, described as those data rates below 6.25 Hz, the offset
performance over temperature can vary significantly from the
remaining data rates. Figure 56, Figure 57, and Figure 58 show
the typical offset performance of the ADXL346 over temperature
for data rates of 6.25 Hz and lower. All plots are normalized to
the offset at 100 Hz output data rate; therefore, a nonzero value
corresponds to additional offset shift due to the temperature for
that data rate.
100
When using the lowest data rates, it is recommended that the
operating temperature range of the device be limited to provide
minimal offset shift across the operating temperature range.
Due to variability between parts, it is also recommended that
calibration over temperature be performed if any data rates
below 6.25 Hz are in use.
100
80
0.10Hz
0.20Hz
0.39Hz
0.78Hz
1.56Hz
3.13Hz
6.25Hz
60
40
08167-057
20
0
25
35
45
55
65
75
85
TEMPERATURE (°C)
Figure 57. Typical Y-Axis Output vs. Temperature at Lower Data Rates,
Normalized to 100 Hz Output Data Rate, VS = 2.6 V
140
NORMALIZED OUTPUT (LSB)
120
100
80
0.10Hz
0.20Hz
0.39Hz
0.78Hz
1.56Hz
3.13Hz
6.25Hz
60
40
20
08167-058
0
–20
25
35
45
55
65
75
85
TEMPERATURE (°C)
Figure 58. Typical Z-Axis Output vs. Temperature at Lower Data Rates,
Normalized to 100 Hz Output Data Rate, VS = 2.6 V
Rev. 0 | Page 36 of 40
ADXL346
AXES OF ACCELERATION SENSITIVITY
AZ
AX
08167-042
AY
Figure 59. Axes of Acceleration Sensitivity (Corresponding Output Increases When Accelerated Along the Sensitive Axis)
XOUT = +1g
YOUT = 0g
ZOUT = 0g
TOP
TOP
XOUT = 0g
YOUT = +1g
ZOUT = 0g
XOUT = –1g
YOUT = 0g
ZOUT = 0g
XOUT = 0g
YOUT = 0g
ZOUT = +1g
Figure 60. Output Response vs. Orientation to Gravity
Rev. 0 | Page 37 of 40
XOUT = 0g
YOUT = 0g
ZOUT = –1g
08167-043
TOP
XOUT = 0g
YOUT = –1g
ZOUT = 0g
TOP
GRAVITY
ADXL346
LAYOUT AND DESIGN RECOMMENDATIONS
Figure 61 shows the recommended printed wiring board land pattern. Figure 62 and Table 27 provide details about the recommended
soldering profile.
0.8000
0.3000
3.3500
0.5000
08167-044
3.3500
Figure 61. Recommended Printed Wiring Board Land Pattern
(Dimensions shown in millimeters)
CRITICAL ZONE
TL TO TP
tP
TP
tL
TSMAX
TSMIN
tS
RAMP-DOWN
PREHEAT
08167-045
TEMPERATURE
RAMP-UP
TL
t25°C TO PEAK
TIME
Figure 62. Recommended Soldering Profile
Table 27. Recommended Soldering Profile 1, 2
Profile Feature
Average Ramp Rate from Liquid Temperature (TL) to Peak Temperature (TP)
Preheat
Minimum Temperature (TSMIN)
Maximum Temperature (TSMAX)
Time from TSMIN to TSMAX (tS)
TSMAX to TL Ramp-Up Rate
Liquid Temperature (TL)
Time Maintained Above TL (tL)
Peak Temperature (TP)
Time of Actual TP − 5°C (tP)
Ramp-Down Rate
Time 25°C to Peak Temperature
1
2
Sn63/Pb37
3°C/sec max
Condition
Pb-Free
3°C/sec max
100°C
150°C
60 sec to 120 sec
3°C/sec max
183°C
60 sec to 150 sec
240 + 0/−5°C
10 sec to 30 sec
6°C/sec max
6 minutes max
Based on JEDEC Standard J-STD-020D.1.
For best results, the soldering profile should be in accordance with the recommendations of the manufacturer of the solder paste used.
Rev. 0 | Page 38 of 40
150°C
200°C
60 sec to 180 sec
3°C/sec max
217°C
60 sec to 150 sec
260 + 0/−5°C
20 sec to 40 sec
6°C/sec max
8 minutes max
ADXL346
OUTLINE DIMENSIONS
3.10
3.00 SQ
2.90
PIN 1
CORNER
0.350
0.10
0.50
BSC
13 14
16
1
0.250
0.50
9
8
6
5
BOTTOM VIEW
TOP VIEW
0.275
1.00
0.95
0.85
END VIEW
SEATING
PLANE
01-13-2010-B
0.79
0.74
0.69
Figure 63. 16-Terminal Land Grid Array [LGA]
(CC-16-3)
Solder Terminations Finish Is Au over Ni
Dimensions shown in millimeters
ORDERING GUIDE
Model 1
ADXL346ACCZ
ADXL346ACCZ-RL
ADXL346ACCZ-RL7
EVAL-ADXL346Z
EVAL-ADXL346Z-M
EVAL-ADXL346Z-S
1
Measurement
Range (g)
±2, ±4, ±8, ±16
±2, ±4, ±8, ±16
±2, ±4, ±8, ±16
Specified
Voltage (V)
2.6
2.6
2.6
Temperature
Range
−40°C to +85°C
−40°C to +85°C
−40°C to +85°C
Package Description
16-Terminal Land Grid Array [LGA]
16-Terminal Land Grid Array [LGA]
16-Terminal Land Grid Array [LGA]
Evaluation Board
Analog Devices Inertial Sensor Evaluation
System, Includes ADXL346 Satellite
ADXL346 Satellite, Standalone
Z = RoHS Compliant Part.
Rev. 0 | Page 39 of 40
Package
Option
CC-16-3
CC-16-3
CC-16-3
Branding
Code
Y2Z
Y2Z
Y2Z
ADXL346
NOTES
I2C refers to a communications protocol originally developed by Philips Semiconductors (now NXP Semiconductors).
Analog Devices offers specific products designated for automotive applications; please consult your local Analog Devices sales representative for details. Standard products sold by
Analog Devices are not designed, intended, or approved for use in life support, implantable medical devices, transportation, nuclear, safety, or other equipment where malfunction
of the product can reasonably be expected to result in personal injury, death, severe property damage, or severe environmental harm. Buyer uses or sells standard products for use
in the above critical applications at Buyer's own risk and Buyer agrees to defend, indemnify, and hold harmless Analog Devices from any and all damages, claims, suits, or expenses
resulting from such unintended use.
©2010 Analog Devices, Inc. All rights reserved. Trademarks and
registered trademarks are the property of their respective owners.
D08167-0-5/10(0)
Rev. 0 | Page 40 of 40