Low Noise, High Frequency MEMS Accelerometers ADXL1001/ADXL1002 Data Sheet FEATURES FUNCTIONAL BLOCK DIAGRAM VDD STANDBY ADXL1001/ADXL1002 TIMING GENERATOR MOD SENSOR AMP OUTPUT AMPLIFIER DEMOD OVERRANGE DETECTION VOUT OR SELF TEST VSS ST 07510-001 Single in plane axis accelerometer with analog output Linear frequency response range from dc to 11 kHz (3 dB point) Resonant frequency of 21 kHz Ultralow noise density 30 g/Hz in 100 g range (ADXL1001) 25 g/Hz in 50 g range (ADXL1002) Overrange sensing plus dc coupling allows fast recovery time Complete electromechanical self-test Sensitivity performance Sensitivity stability over temperature 5% Linearity to 0.1% of full-scale range Cross axis sensitivity 1% (ZX), 1% (YX), Single-supply operation Output voltage ratiometric to supply Low power consumption 1.0 mA Power saving standby operation mode with fast recovery RoHS compliant -40C to +125C temperature range 5 mm x 5 mm x 1.80 mm LFCSP package Figure 1. APPLICATIONS Condition monitoring Predictive maintenance Asset health Test and measurement Health usage monitoring system (HUMS) GENERAL DESCRIPTION The ADXL1001/ADXL1002 deliver ultralow noise density over an extended frequency range with two full-scale range options, and are optimized for industrial condition monitoring. The ADXL1001 (100 g) and the ADXL1002 (50 g) have typical noise densities of 30 g/Hz and 25 g/Hz, respectively. Both accelerometer devices have stable and repeatable sensitivity, which is immune to external shocks up to 10,000 g. Rev. 0 The ADXL1001/ADXL1002 have an integrated full electrostatic self test (ST) function and an overrange (OR) indicator that allow advanced system level features and are useful for embedded applications. With low power and single-supply operation of 3.3 V to 5.25 V, the ADXL1001/ADXL1002 also enable wireless sensing product design. The ADXL1001/ ADXL1002 are available in a 5 mm x 5 mm x 1.80 mm LFCSP package, and are rated for operation over a -40C to +125C temperature range. Document Feedback 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 (c)2017 Analog Devices, Inc. All rights reserved. Technical Support www.analog.com ADXL1001/ADXL1002 Data Sheet TABLE OF CONTENTS Features .............................................................................................. 1 Operating Modes ...........................................................................9 Applications ....................................................................................... 1 Bandwidth ......................................................................................9 Functional Block Diagram .............................................................. 1 Applications information .............................................................. 10 General Description ......................................................................... 1 Application Circuit ..................................................................... 10 Revision History ............................................................................... 2 On Demand Self Test ................................................................. 10 Specifications..................................................................................... 3 Ratiometric Output Voltage ...................................................... 10 Absolute Maximum Ratings ............................................................ 4 Interfacing Analog Output Below 10 kHz .............................. 11 Thermal Resistance ...................................................................... 4 Interfacing Analog Output Beyond 10 kHz............................ 12 Recommended Soldering Profile ............................................... 4 Overrange .................................................................................... 12 ESD Caution .................................................................................. 4 Mechanical Considerations for Mounting .............................. 13 Pin Configuration and Function Descriptions ............................. 5 Layout and Design Recommendations ................................... 13 Typical Performance Characteristics ............................................. 6 Outline Dimensions ....................................................................... 14 Theory of Operation ........................................................................ 9 Ordering Guide .......................................................................... 14 Mechanical Device Operation .................................................... 9 REVISION HISTORY 3/2017--Revision 0: Initial Version Rev. 0 | Page 2 of 14 Data Sheet ADXL1001/ADXL1002 SPECIFICATIONS TA = 25C, VDD = 5.0 V, acceleration = 0 g, unless otherwise noted. Table 1. Parameter1 SENSOR Measurement Range Linearity Cross Axis Sensitivity2 SENSITIVITY (RATIOMETRIC TO VDD) Sensitivity Sensitivity Change Due to Temperature3 ZERO g OFFSET (RATIOMETRIC TO VDD) 0 g Output Voltage 0 g Output Range over Temperature4 NOISE Noise Density 1/f Frequency Corner FREQUENCY RESPONSE Sensor Resonant Frequency 5% Bandwidth5 3 dB Bandwidth5 SELF TEST Output Change (Ratiometric to VDD) Input Level High, VIH Low, VIL Input Current OUTPUT AMPLIFIER Short-Circuit Current Output Impedance Maximum Resistive Load Maximum Capacitive Load6 POWER SUPPLY (VDD) Operating Voltage Range Quiescent Supply Current Standby Current Standby Recovery Time (Standby to Measure Mode) Turn On Time7 OPERATING TEMPERATURE RANGE Test Conditions/ Comments Min ADXL1001 Typ Max Min ADXL1002 Typ Max Unit 100 0.1 50 0.1 g % 1.0 1.0 1.0 1.0 % % DC TA = -40C to +125C 20 5 40 5 mV/g % -40C to +125C VDD/2 5 VDD/2 5 V g 30 0.1 25 0.1 g/Hz Hz 21 4.7 11 21 4.7 11 kHz kHz kHz 545 mV Percentage of full scale ZX cross axis YX cross axis 100 Hz to 10 kHz ST low to ST high 235 275 510 VDD x 0.7 VDD x 0.7 25 25 V V A 3 <0.1 20 100 22 3 <0.1 20 100 22 mA M pF nF VDD x 0.3 No external resistor With external resistor 3.3 Output settled to 1% of final value 5.0 1.0 225 <50 VDD x 0.3 5.25 1.15 285 3.3 +125 -40 <550 -40 1 5.0 1.0 225 <50 5.25 1.15 285 <550 V mA A s s +125 C All minimum and maximum specifications are guaranteed. Typical specifications may not be guaranteed. Cross axis sensitivity is defined as the coupling of excitation along a perpendicular axis onto the measured axis output. Includes package hysteresis from 25C. 4 Difference between maximum and minimum values in temperature range. 5 Specified as frequency range that is within a deviation range relative to dc sensitivity, range is limited by an increase in response due to response gain at the sensor resonant frequency. 6 For capacitive loads larger than 100 pF, an external series resistor must be connected (minimum 8 k). The output capacitance must not exceed 22 nF. 7 Measured time difference from the instant VDD reaches half its value to the instant at which the output settles to 1% of its final value. 2 3 Rev. 0 | Page 3 of 14 ADXL1001/ADXL1002 Data Sheet ABSOLUTE MAXIMUM RATINGS RECOMMENDED SOLDERING PROFILE Table 2. Figure 2 and Table 4 provide details about the recommended soldering profile. Rating 10,000 g 10,000 g 1.2 m -0.3 V to +5.5 V Indefinite RAMP-UP THERMAL RESISTANCE Thermal performance is directly linked to printed circuit board (PCB) design and operating environment. Careful attention to PCB thermal design is required. JA is the natural convection junction to ambient thermal resistance measured in a one cubic foot sealed enclosure. JC is the junction to case thermal resistance. Table 3. Package Characteristics 1 JA 48C/W JC 14.1C/W Device Weight <0.2 g Thermal impedance simulated values are based on a JEDEC 2S2P thermal test board with nine thermal vias. See JEDEC JESD51. TL tL TSMAX TSMIN tS RAMP-DOWN PREHEAT 15431-002 -55C to +150C Stresses at or above those listed under Absolute Maximum Ratings may cause permanent damage to the product. This is a stress rating only; functional operation of the product at these or any other conditions above those indicated in the operational section of this specification is not implied. Operation beyond the maximum operating conditions for extended periods may affect product reliability. Package Type CP-32-261 CRITICAL ZONE TL TO TP tP TP TEMPERATURE Parameter Acceleration (Any Axis, Unpowered) Acceleration (Any Axis, Powered) Drop Test (Concrete Surface) VDD Output Short-Circuit Duration (Any Pin to Common) Temperature Range (Storage) t25C TO PEAK TIME Figure 2. Recommended Soldering Profile Table 4. Recommended Soldering Profile Profile Feature Average Ramp Rate (TL to TP) Preheat Minimum Temperature (TSMIN) Maximum Temperature (TSMAX) Time, TSMIN to TSMAX (tS) TSMAX to TL Ramp-Up Rate Time Maintained Above Liquidous (TL) Liquidous Temperature (TL) Time (tL) Peak Temperature (TP) Time Within 5C of Actual Peak Temperature (tP) Ramp-Down Rate Time 25C to Peak Temperature (t25C) ESD CAUTION Rev. 0 | Page 4 of 14 Condition Sn63/Pb37 Pb-Free 3C/sec 3C/sec maximum maximum 100C 150C 60 sec to 120 sec 150C 200C 60 sec to 180 sec 3C/sec maximum 3C/sec maximum 183C 60 sec to 150 sec 240C + 0C/-5C 10 sec to 30 sec 6C/sec maximum 6 min maximum 217C 60 sec to 150 sec 260C + 0C/-5C 20 sec to 40 sec 6C/sec maximum 8 min maximum Data Sheet ADXL1001/ADXL1002 32 31 30 29 28 27 26 25 NIC NIC VOUT DNC VSS VSS DNC DNC PIN CONFIGURATION AND FUNCTION DESCRIPTIONS 1 2 3 4 5 6 7 8 + - AXIS OF SENSITIVITY ADXL1001/ ADXL1002 TOP VIEW (Not to Scale) 24 23 22 21 20 19 18 17 DNC DNC DNC DNC OR DNC DNC DNC NOTES 1. NIC = NOT INTERNALLY CONNECTED. 2. DNC = NO NOT CONNECT. LEAVE THIS PIN UNCONNECTED. 3. THE EXPOSED PAD ON THE BOTTOM OF THE PACKAGE MUST BE CONNECTED TO GROUND. 4. AXIS OF SENSITIVITY IS IN-PLANE TO THE PACKAGE AND HORIZONTAL AS SHOWN. 15431-003 NIC DNC DNC VDD VSS VSS STANDBY ST 9 10 11 12 13 14 15 16 NIC NIC NIC NIC NIC NIC NIC NIC Figure 3. Pin Configuration Table 5. Pin Function Descriptions Pin No. 1 to 9, 31, 32 10, 11, 17 to 19, 21 to 26, 29 12 13, 14, 27, 28 15 16 20 Mnemonic NIC DNC Description Not Internally Connected. Do Not Connect. Leave unconnected. VDD VSS STANDBY ST OR 30 33 VOUT EPAD 3.3 V to 5.25 V Supply Voltage. Supply Ground. Standby mode Input, Active High. Self Test Input, Active High. Overrange Output. This pin instantaneously indicates when the overrange detection circuit identifies significant overrange activity. This pin is not latched. Analog Output Voltage. Exposed Pad. The exposed pad on the bottom of the package must be connected to ground. Rev. 0 | Page 5 of 14 ADXL1001/ADXL1002 Data Sheet TYPICAL PERFORMANCE CHARACTERISTICS NORMALIZED AMPLITUDE (dB) 15 10000 10 NOISE PSD (g/Hz) 1000 5 0 ADXL1001 ADXL1002 100 10 1k 10k 1 0.01 15431-004 -10 100 100k FREQUENCY (Hz) Figure 4. Frequency Response of ADXL1001/ADXL1002, High Frequency (>5 kHz) Vibration Response; a Laser Vibrometer Controller Referencing the ADXL1002 Package Used for Accuracy Figure 7. ADXL1001/ADXL1002 Noise Power Spectral Density (Noise PSD) Below 10 Hz NOISE PSD (g/Hz) 40 30 20 10 100 DEVICE 1 DEVICE 2 DEVICE 3 60 1k 10k 100k FREQUENCY (Hz) 50 40 30 20 10 100 100k Figure 8. ADXL1002 Noise Power Spectral Density (PSD) 5 3 3 SENSITIVITY (%) 5 -1 10k FREQUENCY (Hz) Figure 5. ADXL1001 Noise Power Spectral Density (PSD) vs. Frequency 1 1k 15431-008 DEVICE 1 DEVICE 2 DEVICE 3 50 15431-005 1 -1 -3 -3 20 50 80 110 TEMPERATURE (C) -5 -40 15431-006 -10 -20 0 20 40 60 80 100 TEMPERATURE (C) Figure 6. ADXL1001 Sensitivity vs. Temperature Figure 9. ADXL1002 Sensitivity vs. Temperature Rev. 0 | Page 6 of 14 120 15431-009 NOISE PSD (g/Hz) 10 100 90 80 70 60 SENSITIVITY (%) 1 FREQUENCY (Hz) 100 90 80 70 -5 -40 0.1 15431-007 -5 ADXL1001/ADXL1002 10 10 5 5 0 -10 -40 15 70 125 TEMPERATURE (C) -5 -10 -40 15431-010 -5 0 70 125 TEMPERATURE (C) Figure 10. ADXL1001 Normalized Offset vs. temperature Figure 13. ADXL1002 Normalized Offset vs. Temperature 280 240 220 200 180 140 3.3 3.5 3.7 3.9 4.1 4.3 4.5 4.7 4.9 5.1 5.3 SUPPLY VOLTAGE (V) Figure 11. ADXL1001/ADXL1002 Standby Current vs. Supply Voltage 1000 950 900 850 800 750 700 650 600 15431-011 160 1050 3.3 3.5 3.7 3.9 4.1 4.3 4.5 4.7 SUPPLY VOLTAGE (V) 4.9 5.1 5.3 15431-014 MEASURE MODE SUPPLY CURRENT (A) 1100 260 STANDBY CURRENT (A) 15 15431-013 NORMALIZED OFFSET (g) NORMALIZED OFFSET (g) Data Sheet Figure 14. ADXL1001/ADXL1002 Measure Mode Supply Current vs. Supply Voltage 40 60 35 PERCENT OF POPULATION 40 30 20 30 25 20 15 10 10 0 0 19.0 19.2 19.4 19.6 19.8 20.0 20.2 20.4 20.6 20.8 21.0 ADXL1001 SENSITIVITY (mV/g) 38.0 38.2 38.4 38.6 38.8 39.0 39.2 39.4 39.6 39.8 40.0 ADXL1002 SENSITIVITY (mV) Figure 12. ADXL1001 Sensitivity Histogram at 25C Figure 15. ADXL1002 Sensitivity Histogram at 25C Rev. 0 | Page 7 of 14 15431-015 5 15431-112 PERCENT OF POPULATION 50 ADXL1001/ADXL1002 Data Sheet 40 30 35 PERCENTAGE OF POPULATION 20 15 10 5 20 15 10 212 216 220 224 228 232 236 240 244 15431-018 0 15431-016 930 945 960 975 990 1005 1020 1035 1050 1065 1080 1095 MEASURE MODE SUPPLY CURRENT (A) 248 STANDBY CURRENT (A) Figure 16. ADXL1001/ADXL1002 Measure Mode Supply Current Histogram at 5 V Figure 19. ADXL1001/ADXL1002 Standby Current Histogram at 5 V 0.100 0.075 0.050 0.025 0 -0.250 -0.050 -0.075 0 20 40 60 INPUT ACCELERATION (g) 80 15431-017 -0.100 100 0.750 0.500 0.250 0 -0.250 -0.500 -0.750 -0.100 0 10 20 30 INPUT ACCELERATION (g) 40 15431-019 SENSITIVITY NONLINEARITY (% of Full Scale) 0.100 SENSITIVITY NONLINEARITY (% of Full Scale) 25 5 0 50 Figure 20. ADXL1002 Sensitivity Nonlinearity vs. Input Acceleration Figure 17. ADXL1001 Sensitivity Nonlinearity vs. Input Acceleration 40 6 VOUT STANDBY 35 PERCENTAGE OF POPULATION 5 4 VOLTAGE (V) 30 3 2 1 0 30 25 20 15 10 5 0 10 20 TIME (s) 30 40 0 15431-020 -1 2.46 2.47 2.48 2.49 2.50 2.51 2.52 2.53 2.54 2.55 2.56 2.57 0g OUTPUT (V) Figure 18. ADXL1001/ADXL1002 Output (Gray Trace) Settling in Standby (Black); Output Voltage Settles to Midrail (2.5 V) in Standby in <30 s; Effectively Unfiltered (No Low-Pass Filter) Output Rev. 0 | Page 8 of 14 Figure 21. ADXL1001/ADXL1002 0 g Output Population 15431-012 PERCENTAGE OF POPULATION 25 Data Sheet ADXL1001/ADXL1002 THEORY OF OPERATION The ADXL1001/ADXL1002 are high frequency, low noise singleaxis microelectromechanical systems (MEMS) accelerometers that provide an analog output that is proportional to mechanical vibration. The ADXL1001/ADXL1002 have high g ranges of 100 g and 50 g and are suitable for vibration measurements in high bandwidth applications such as vibration analysis systems that monitor and diagnose machine or system health. The low noise and high frequency bandwidth allows the measurement of vibration patterns caused by small moving parts, such as internal bearings, and the high g range provides the dynamic range to be used in high vibration environments such as heating, ventilation, and air conditioning (HVAC) and heavy machine equipment. To achieve proper performance, be aware of system noise, mounting, and signal conditioning. System noise is affected by supply voltage noise. The analog output of the ADXL1001/ADXL1002 is a ratiometric output; therefore, supply voltage modulation affects the output. Use a properly decoupled stable supply voltage to power the ADXL1001/ ADXL1002 and to provide a reference voltage for the digitizing system. The output signal is impacted by an overrange stimulus. An overload indicator output feature is provided to indicate a condition that is critical for an intelligent measurement system. For more information about the overrange features, see the Overrange section. Proper mounting is required to ensure full mechanical transfer of vibration to accurately measure the desired vibration rather than vibration of the measurement system, including the sensor. A common technique for high frequency mechanical coupling is to utilize a sensor stud mount system while considering the mechanical interface of fixing the ADXL1001/ADXL1002 in the stud. For lower frequencies (below the full capable bandwidth of the sensor), it is possible to use magnetic or adhesive mounting. Proper mounting technique ensures proper and repeatable results that are not influenced by measurement system mechanical resonances and/or damping at the desired frequency, and represents an efficient and proper mechanical transfer to the system being monitored. Proper application specific signal conditioning is require to achieve optimal results. An understanding of measurement frequency range and managing overload condition is important to achieve accurate results. The electrical output signal of the ADXL1001/ADXL1002 requires some band limiting and proper digitization bandwidth. See the Interfacing Analog Output Below 10 kHz section and the Interfacing Analog Output Beyond 10 kHz section for more information. MECHANICAL DEVICE OPERATION The moving component of the sensor is a polysilicon surfacemicromachined structure built on top of a silicon wafer. Polysilicon springs suspend the structure over the surface of the wafer and provide a resistance against acceleration forces. 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 structure and unbalances the differential capacitor, resulting in a sensor output with an amplitude proportional to acceleration. Phasesensitive demodulation determines the magnitude and polarity of the acceleration. OPERATING MODES The ADXL1001/ADXL1002 have two operating modes: measurement mode and standby mode. Measurement mode provides a continuous analog output for active monitoring. Standby mode is a nonoperational, low power mode. Measurement Mode Measurement mode is the normal operating mode of the ADXL1001/ADXL1002. In this mode, the accelerometer actively measures acceleration along the axis of sensitivity and consumes 1.0 mA (typical) using a 5.0 V supply. Standby Placing the ADXL1001/ADXL1002 in standby mode suspends the measurement with internal reduction of current consumption to 225 A (typical for 5.0 V supply). The transition time from standby to measurement mode is <50 s. The transition from standby to measure mode is shown in Figure 18. BANDWIDTH The ADXL1001/ADXL1002 circuitry supports an output signal bandwidth beyond the resonant frequency of the sensor, measuring acceleration over a bandwidth comparable to the resonant frequency of the sensor. The output response is a combination of the sensor response and the output amplifier response. Therefore, external band limiting or filtering is required; see the Interfacing Analog Output Below 10 kHz section and the Interfacing Analog Output Beyond 10 kHz section for more information. When using the ADXL1001/ADXL1002 beyond 10 kHz, consider the nonlinearity due to the resonance frequency of the sensor, the additional noise due to the wideband output of the amplifier, and the discrete frequency spurious tone due to coupling of the internal 200 kHz clock. Aliased interferers in the desired band cannot be removed, and observed performance degrades. A combination of high speed sampling and appropriate band limiting filtering is required for optimal performance. Rev. 0 | Page 9 of 14 ADXL1001/ADXL1002 Data Sheet APPLICATIONS INFORMATION APPLICATION CIRCUIT For most applications, a single 1 F capacitor adequately decouples the accelerometer from noise on the power supply. A band limiting filter at the output provides suppression of out of band noise and signal. A capacitive load between 100 pF and 22 nF is recommended. The output amplifier can drive resistive loads up to 2 mA of source current, for example greater than 2.5 k for 5 V operation. If the output is to drive a capacitive load greater than or equal to 100 pF, a series resistor of at least 8 k is required to maintain the amplifier stability. When inactive, the ST and STANDBY pins are forced low. The overrange indicator is an output that can be monitored to identify the status of the system. OPTIONAL LOW-PASS FILTER R The self test feature can be exercised by the user with the following steps: 1. 2. 3. 4. Measure the output voltage. Turn on self test by setting the ST pin to VDD. Measure the output again. Subtract the two readings and compare the result to the expected value from Table 1, while factoring in the response curve due to supply voltage, if necessary, from Figure 23. The self test function can be activated at any point during normal operation by setting the ST pin to VDD. Self test takes approximately 300 s from the assertion of the ST pin to a result, and acceleration outputs return approximately 300 s after the release of the ST pin. While performing the self test measurement, do not use the accelerometer output to measure external acceleration. 700 VOUT TYPICAL SELF TEST DELTA (mV) C VSS 32 31 30 29 28 27 26 25 1 24 2 23 3 22 21 ADXL1001/ ADXL1002 4 5 20 OR 600 500 ADXL1002 ADXL1001 400 300 200 100 19 0 18 7 3.3 10 11 12 13 14 1F 15 3.7 3.9 4.1 4.3 4.5 4.7 4.9 5.1 5.3 Figure 23. ADXL1002 Typical Self Test Delta vs. Supply Voltage 16 ST (ACTIVE HIGH) STANDBY (ACTIVE HIGH) 15431-021 9 VDD (3.3V TO 5.25V SUPPLY VOLTAGE) 3.5 SUPPLY VOLTAGE (V) 17 8 15431-022 6 Figure 22. ADXL1001/ADXL1002 Application Circuit ON DEMAND SELF TEST A fully integrated electromechanical self test function is designed into the ADXL1001/ADXL1002. This function electrostatically actuates the accelerometer proof mass, resulting in a displacement of the capacitive sense fingers. This displacement is equivalent to the displacement that occurs as a result of external acceleration input. The proof mass displacement is processed by the same signal processing circuitry as a true acceleration output signal, providing complete coverage of both the electrical and mechanical responses of the sensor system. RATIOMETRIC OUTPUT VOLTAGE The ADXL1001/ADXL1002 are tested and specified at VDD = 5.0 V; however, it can be powered with VDD as low as 3.3 V or as high as 5.25 V. Some performance parameters change as the supply voltage is varied. The ADXL1001/ADXL1002 output is ratiometric to the supply voltage VDD; therefore, the output sensitivity (or scale factor) varies proportionally to the supply voltage. At VDD = 5.0 V, the output sensitivity is typically 40 mV/g and 20 mV/g in the ADXL1002 and ADXL1001, respectively. The zero g bias output is ratiometric also and is nominally midscale relative to the supply voltage (VDD/2). Rev. 0 | Page 10 of 14 ADXL1001/ADXL1002 45 the desired bandwidth and the chosen ADC sampling rate be faster than the amplifier bandwidth. 40 The output amplifier is ratiometric to the supply voltage, and there are two distinct cases regarding digital conversion, as follows: 35 * 30 20 3.3 3.8 4.3 4.8 5.3 SUPPLY VOLTAGE (V) 15431-023 25 Figure 24. ADXL1002 Sensitivity vs. Supply Voltage * INTERFACING ANALOG OUTPUT BELOW 10 kHz The ADXL1001/ADXL1002 sense mechanical motion along a single axis and produces a voltage output. The system performance depends on the output response that is a result of mechanical vibration sensed and signal processing of the electrical output. The sensor must be effectively mechanically coupled. Mechanical coupling can be a complex integration of multiple components, typically unique for each application. Consideration must be made for all mechanical interfaces including the mounting of the MEMS to the PCB (location on the PCB as well as solder chemistry), the size of the PCB (both thickness and active surface area), and the mounting of the PCB to the system being monitored (either in a module or directly mounted). In general, the following guidelines for effective mechanical interface must be used to support up to 10 kHz bandwidth: * * * * Keep the ADXL1001/ADXL1002 near a stable mechanical mounting on the PCB. Provide multiple hard mounting points. Keep the PCB thick and avoid a large surface area PCB that induces higher magnitude and lower frequency resonances. Ensure the mechanical connection is sufficiently stiff to transfer mechanical forces up to the desired frequency. Below 10 kHz, magnetic and adhesive mounting is possible with proper attention. The EVAL-ADXL1001Z and the EVAL-ADXL1002Z evaluation boards can be used as a reference. The ADXL1001/ADXL1002 electrical output supports a bandwidth beyond the resonance of the sensor. The small signal bandwidth of the output amplifier in the ADXL1001/ADXL1002 is 70 kHz. During the digitization process, aliasing, which is the folding of higher frequency noise and signals into the desired band, can occur. To avoid aliasing noise from the amplifier and other internal circuits (for example, coupling of the internal 200 kHz clock), it is recommended that an external filter be implemented at The user has an analog-to-digital (ADC) downstream of the accelerometer that can use the VDD voltage as a reference. In this case, the voltage supply tolerance and voltage temperature coefficient (commonly associated with external regulators) tracks between the sensor and the ADC and, therefore, the supply and reference voltage induced error cancels out. This design approach is recommended. If the ADC cannot reference the same 5 V supply as the sensor for any reason, the sensitivity of the digitized sensor output reflects the regulator tolerance and temperature coefficient. The ADXL1001/ADXL1002 output amplifier is stable while driving capacitive loads up to 100 pF directly without a series resistor. At loads greater than 100 pF, an 8 k series resistor or greater must be used. See Figure 25 for an example of the interface including components when measuring mechanical vibration from 0 kHz to 5 kHz, using the AD4000 ADC. For a 5 kHz pass band, a singlepole RC filter is acceptable; however, in some applications, use a more aggressive filter and lower sample rate. The following components are recommended to form a two-pole RC filter at the output of the ADXL1001/ADXL1002: R1 = 91 k, C1 = 330 pF, R2 = 0 , and C2 = not required. A minimum ADC sample rate of 16 kHz is recommended to avoid aliasing. See Figure 25 for an example of the interface including components when measuring mechanical vibration from 0 kHz to 10 kHz. The following components are recommended to form a two-pole RC filter at the output of the ADXL1001/ADXL1002: R1 = 16 k, C1 = 300 pF, R2 = 32 k, and C2 = 300 pF. A minimum ADC sample rate of 32 kHz is recommended to avoid aliasing. The two-pole RC filter produces an attenuation of approximately 84 dB at 200 kHz, the internal clock frequency. VDD 3.3V TO 5.0V1 AD4000 V DD 1.8V 0.1F (+1F, OPTIONAL) 10F VDD R2 R1 VOUT IN+ C1 ADXL1001/ ADXL1002 C2 REF VDD AD4000 IN- GND VSS 13.3V LIMITED BY ADXL1001/ADXL1002; 5.0V LIMITED BY AD4000 Figure 25. Application Circuit for the ADXL1001/ADXL1002 Rev. 0 | Page 11 of 14 15431-024 SENSITIVITY (mV/g) Data Sheet ADXL1001/ADXL1002 Data Sheet When using the ADXL1001/ADXL1002 beyond 10 kHz, consider the nonlinearity due to the resonance frequency of the sensor, the additional noise due to the wideband output of the amplifier, and the discrete frequency spurious tone due to coupling of the internal 200 kHz clock. If any of these interferers alias in the desired band, it cannot be removed and observed performance degrades. A combination of high speed sampling and appropriate filtering is required for optimal performance. The first consideration is the effect of the sensor resonance frequency at 21 kHz. Approaching and above this frequency, the output response to an input stimulus peaks, as shown in Figure 4. At frequencies near or above the resonance, the output response is outside the linear response range and, therefore, the sensitivity is different than observed at lower frequencies. In these frequency ranges, the relative response (as opposed to absolute value) over time is typically observed. The ADXL1001/ADXL1002 output amplifier small signal bandwidth is 70 kHz. The user must properly interface to the device with proper signal filtering to avoid issues with out of band noise aliasing into the desired band. The amplifier frequency response roll off can be modeled as a single-pole, low-pass filter at 70 kHz. In the absence of additional external low-pass filtering, to avoid aliasing of high frequency noise, choose a sampling rate of at least twice the equivalent noise bandwidth (ENBW) for a single-pole, low-pass filter, as follows: ENBW = For example, to achieve the lowest rms noise and noise density for extended bandwidth applications, it is recommended to use at least a multiple order low-pass filter at the output of the ADXL1001/ADXL1002 and a digitization sample rate of at least 4x the desired bandwidth, assuming sufficient filtering of the 200 kHz internal clock signal. Use an ADC sample rate of 1 MSPS or greater along with digital low-pass filtering to achieve similar performance. OVERRANGE The ADXL1001/ADXL1002 have an output (OR pin) to signal when an overrange event (acceleration larger than twice the full-scale range). Built in overrange detection circuitry provides an alert to indicate a significant overrange event occurred that is greater than approximately 2x the specified g range. When an overrange is detected, the internal clock is disabled to the sensor for 200 s to maximize protection of the sensor element during an overrange event. If a sustained overrange event is encountered, the overrange detection circuitry triggers periodically, approximately every 500 s. 200 6 DEVICE UNDER TEST REFERENCE OR 150 4 100 2 OR (V) Although a mechanical interface is always important to achieve accurate and repeatable results in MEMS applications, it is critical in cases when measuring greater than a few kilohertz. Typically, magnetic and adhesive mounting are not sufficient to maintain proper mechanical transfer of vibration through these frequencies. Mechanical system analysis is required for these applications. Another issue that must be addressed is the coupling of the internal clock signal at 200 kHz onto the output signal. This clock spur must be filtered by analog or digital filtering so as not to affect the analysis of results. 50 0 0 -50 -2 -2 -1 0 1 2 3 TIME (ms) 4 5 6 7 15431-025 The ADXL1001/ADXL1002 are high frequency, single-axis MEMS accelerometer devices that provide an output signal pass band beyond the resonance frequency range of the sensor. Although the output 3 dB frequency response bandwidth is approximately 11 kHz (note that this is a 3 dB response, meaning there is a gain in sensitivity at this frequency), in some cases, it is desirable to observe frequency beyond this range. To accommodate this, the ADXL1001/ADXL1002 output amplifier supports a 70 kHz small signal bandwidth, which is well beyond the resonant frequency of the sensor. That is, sampling rate must be at least 220 kHz. This sample rate addresses reducing broadband noise due to the amplifier from folding back (aliasing) in-band, but does not prevent out of band signals from aliasing in-band. To prevent out of band responses, additional external low-pass filtering is required. ACCELERATION (g) INTERFACING ANALOG OUTPUT BEYOND 10 kHz Figure 26. ADXL1001/ADXL1002 Behavior During a Continuous Overrange x 70 kHz 110 kHz 2 Rev. 0 | Page 12 of 14 Data Sheet ADXL1001/ADXL1002 MECHANICAL CONSIDERATIONS FOR MOUNTING PCB MOUNTING POINTS Figure 27. Incorrectly Placed Accelerometers LAYOUT AND DESIGN RECOMMENDATIONS Figure 28 shows the recommended printed circuit board land pattern. 0.03"/0.755mm 0.02"/0.5mm 0.012"/0.305mm 32 31 30 29 28 27 26 25 1 24 2 23 3 22 4 21 5 20 6 19 7 18 8 17 9 15431-026 Mount the ADXL1001/ADXL1002 on the PCB in a location close to a hard mounting point of the PCB. Mounting the ADXL1001/ADXL1002 at an unsupported PCB location, as shown in Figure 27, may result in large, apparent measurement errors due to undamped PCB vibration. Placing the accelerometer near a hard mounting point ensures that any PCB vibration at the accelerometer is above the mechanical sensor resonant frequency of the accelerometer and, therefore, effectively invisible to the accelerometer. Multiple mounting points, close to the sensor, and a thicker PCB also help to reduce the effect of system resonance on the performance of the sensor. ACCELEROMETERS 0.146"/3.7mm 0.191"/4.855mm 10 11 12 13 14 15 16 0.191"/4.855mm Figure 28. Recommended Printed Wiring Board Land Pattern Rev. 0 | Page 13 of 14 15431-027 0.146"/3.7mm ADXL1001/ADXL1002 Data Sheet OUTLINE DIMENSIONS DETAIL A (JEDEC 95) 5.10 5.00 SQ 4.90 PIN 1 INDICATOR 0.30 0.25 0.20 PIN 1 INDIC ATOR AREA OPTIONS (SEE DETAIL A) 25 32 24 1 0.50 BSC 3.80 3.70 SQ 3.60 EXPOSED PAD 8 17 *1.85 1.80 1.75 0.05 MAX 0.02 NOM COPLANARITY 0.08 0.203 REF PKG-004829 SEATING PLANE 16 9 BOTTOM VIEW 3.50 REF 0.20 MIN FOR PROPER CONNECTION OF THE EXPOSED PAD, REFER TO THE PIN CONFIGURATION AND FUNCTION DESCRIPTIONS SECTION OF THIS DATA SHEET. *COMPLIANT TO JEDEC STANDARDS MO-220-VHHD-4 WITH EXCEPTION TO PACKAGE HEIGHT. 02-02-2017-A TOP VIEW 0.45 0.40 0.35 Figure 29. 32-Lead Lead Frame Chip Scale Package (LFCSP) 5 mm x 5 mm Body and 1.8 mm Package Height (CP-32-26) Dimensions shown in millimeters ORDERING GUIDE Model1 ADXL1001BCPZ ADXL1001BCPZ-RL ADXL1001BCPZ-RL7 ADXL1002BCPZ ADXL1002BCPZ-RL ADXL1002BCPZ-RL7 EVAL-ADXL1002Z EVAL-ADXL1001Z 1 Temperature Range -40C to +125C -40C to +125C -40C to +125C -40C to +125C -40C to +125C -40C to +125C g Range 100 g 100 g 100 g 50 g 50 g 50 g Package Description 32-Lead Lead Frame Chip Scale Package [LFCSP] 32-Lead Lead Frame Chip Scale Package [LFCSP] 32-Lead Lead Frame Chip Scale Package [LFCSP] 32-Lead Lead Frame Chip Scale Package [LFCSP] 32-Lead Lead Frame Chip Scale Package [LFCSP] 32-Lead Lead Frame Chip Scale Package [LFCSP] ADXL1002 Evaluation Board ADXL1001 Evaluation Board Z = RoHS Compliant Part. (c)2017 Analog Devices, Inc. All rights reserved. Trademarks and registered trademarks are the property of their respective owners. D15431-0-3/17(0) Rev. 0 | Page 14 of 14 Package Option CP-32-26 CP-32-26 CP-32-26 CP-32-26 CP-32-26 CP-32-26