Low Noise, Wide Bandwidth, MEMS Accelerometer ADXL1003 Data Sheet FEATURES FUNCTIONAL BLOCK DIAGRAM VDD STANDBY ADXL1003 TIMING GENERATOR MOD SENSOR AMP OUTPUT AMPLIFIER DEMOD OVERRANGE DETECTION XOUT OR SELF TEST VSS ST 16596-001 Single, in plane axis accelerometer with analog output Full-scale range: 200 g Linear frequency response range: dc to 15 kHz typical (3 dB point) Resonant frequency: 28 kHz typical Ultralow noise density: 45 g/Hz Overrange sensing plus dc coupling allows fast recovery time Complete electromechanical self test Sensitivity performance Sensitivity stability over temperature within 5% Linearity to 0.2% of full-scale range Cross axis sensitivity: 1%/0.8% (z-axis acceleration effect on x-axis, y-axis acceleration effect on x-axis) Single-supply operation Output voltage ratiometric to supply Low power consumption: 1.0 mA typical Power saving standby operation mode with fast recovery RoHS compliant -40C to +125C operating temperature range 32-lead, 5 mm x 5 mm x 1.8 mm LFCSP package Figure 1. APPLICATIONS Condition monitoring Predictive maintenance Asset health Test and measurement Health usage monitoring systems (HUMSs) Acoustic emissions GENERAL DESCRIPTION The ADXL1003 delivers ultralow noise density over an extended frequency range and is optimized for bearing fault detection and diagnostics. The ADXL1003 has typical noise density of 45 g/Hz across the linear frequency range. Microelectronicmechanical systems (MEMS) accelerometers have stable and repeatable sensitivity, and are immune to external shocks of up to 10,000 g. Rev. 0 The integrated signal conditioning electronics enable such features as full electrostatic self test (ST) and an overrange (OR) indicator, useful for embedded applications. With low power and single-supply operation of 3.0 V to 5.25 V, the ADXL1003 also enables wireless sensing product design. The ADXL1003 is available in a 5 mm x 5 mm x 1.8 mm LFCSP package, and operates over the -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)2018 Analog Devices, Inc. All rights reserved. Technical Support www.analog.com ADXL1003 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 .................................................................................... 13 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 8/2018--Revision 0: Initial Version Rev. 0 | Page 2 of 14 Data Sheet ADXL1003 SPECIFICATIONS TA = 25C, VDD = 5.0 V, acceleration = 0 g, unless otherwise noted. All minimum and maximum specifications are guaranteed. Typical specifications may not be guaranteed. Table 1. Parameter SENSOR Measurement Range Linearity1 Cross Axis Sensitivity2 SENSITIVITY (RATIOMETRIC TO VDD) Sensitivity3 Sensitivity Change Due to Temperature4 ZERO g OFFSET (RATIOMETRIC TO VDD) 0 g Output Voltage 0 g Output Range over Temperature5 NOISE Noise Density VDD = 5.0 V VDD = 3.0 V 1/f Frequency Corner FREQUENCY RESPONSE Sensor Resonant Frequency 5% Bandwidth6 3 dB Bandwidth7 SELF TEST Output Change (Ratiometric to VDD) Input Voltage Level High, VIH Low, VIL Input Current OUTPUT AMPLIFIER Short-Circuit Current Output Impedance Maximum Capacitive Load8 POWER SUPPLY (VDD) Operating Voltage Range3 Quiescent Supply Current Standby Current Standby Recovery Time (Standby to Measure Mode) Turn On Time9 OPERATING TEMPERATURE RANGE Test Conditions/Comments Min Typ Max g % % % 200 Percentage of full-scale Z-axis acceleration effect on x-axis Y-axis acceleration effect on x-axis DC TA = -40C to +125C 0.2 0.8 1.0 9.2 10 5 Unit 10.8 mV/g % VDD/2 2 V g 45 80 0.1 g/Hz g/Hz Hz 24 28 6.2 15 kHz kHz kHz 57 85 mV -40C to +125C 100 Hz to 14 kHz ST low to ST high VDD x 0.7 25 V V A 3 <0.1 100 22 mA pF nF VDD x 0.3 No series resistor With series resistor 3.0 Output settled to 1% of final value -40 1 5.0 1.0 225 <50 <550 5.25 1.15 285 +125 V mA A s s C Linearity is tested using sine vibration at 13 kHz. Cross axis sensitivity is defined as the coupling of excitation along a perpendicular axis onto the measured axis output. 3 Parameter limits are based on characterization data or are guaranteed by design. 4 Includes package hysteresis from 25C. 5 Difference between the maximum and the minimum values in temperature range. 6 Specified as a frequency range that is within a deviation range relative to dc sensitivity. The range is limited by an increase in response due to response gain at the sensor resonant frequency. 7 Specified as a frequency range that is within a deviation range relative to dc sensitivity. The range is limited by an increase in response due to response gain at the sensor resonant frequency. 8 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. 9 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 Rev. 0 | Page 3 of 14 ADXL1003 Data Sheet ABSOLUTE MAXIMUM RATINGS Table 4. Recommended Soldering Profile Table 2. Parameter Acceleration Any Axis, Powered or Unpowered Drop Test (Concrete Surface) VDD Output Short-Circuit Duration (Any Pin to Common Ground) Temperature Range (Storage) Rating Profile Feature Average Ramp Rate (TL to TP) 10,000 g 1.2 m -0.3 V to +5.5 V Indefinite Preheat Minimum Temperature (TSMIN) Maximum Temperature (TSMAX) Time (TSMIN to TSMAX)(tS) -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. TSMAX to TL Ramp-Up Rate THERMAL RESISTANCE Peak Temperature (TP) Time Within 5C of Actual Peak Temperature (tP) Ramp-Down Rate Thermal performance is directly linked to printed circuit board (PCB) design and operating environment. Careful attention to PCB thermal design is required. Time 25C to Peak Temperature (t25C) 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. ESD CAUTION Table 3. Package Characteristics Package Type CP-32-261 1 JA 48C/W Time Maintained Above Liquidous (TL) Liquidous Temperature (TL) Time (tL) 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. RECOMMENDED SOLDERING PROFILE Figure 2 and Table 4 provide details about the recommended soldering profile. CRITICAL ZONE TL TO TP tP TP TL tL TSMAX TSMIN tS RAMP-DOWN PREHEAT t25C TO PEAK TIME 16596-002 TEMPERATURE RAMP-UP Figure 2. Recommended Soldering Profile 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 240 + 0/-5C 10 sec to 30 sec 6C/sec maximum 6 min maximum 217C 60 sec to 150 sec 260 + 0/-5C 20 sec to 40 sec 6C/sec maximum 8 min maximum Data Sheet ADXL1003 32 31 30 29 28 27 26 25 NIC NIC XOUT DNC VSS VSS DNC DNC PIN CONFIGURATION AND FUNCTION DESCRIPTIONS 1 2 3 4 5 6 7 8 + - ADXL1003 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 = DO NOT CONNECT. LEAVE THIS PIN UNCONNECTED. 3. EXPOSED PAD. THE EXPOSED PAD ON THE BOTTOM OF THE PACKAGE MUST BE CONNECTED TO GROUND AND IS REQUIRED FOR BOTH ELECTRICAL AND MECHANICAL PERFORMANCE. 4. AXIS OF SENSITIVITY IS IN PLANE TO THE PACKAGE AND HORIZONTAL AS SHOWN. 16596-003 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 this pin unconnected. VDD VSS STANDBY ST OR 30 XOUT EPAD 3.0 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 and is required for both electrical and mechanical performance. Rev. 0 | Page 5 of 14 ADXL1003 Data Sheet TYPICAL PERFORMANCE CHARACTERISTICS 0.8 12 0.7 PERCENT OF POPULATION 8 6 4 2 0.6 0.5 0.4 0.3 0.2 0.1 10.7 SENSITIVITY (mV/g) Figure 4. Frequency Response, High Frequency (>5 kHz) Vibration Response; a Laser Vibrometer Controller Referencing the ADXL1003 Package Used for Accuracy 16596-007 10.6 10.5 10.4 10.3 10.2 9.9 10.1 FREQUENCY (Hz) 10.0 100k 9.8 10k 9.5 1k 16596-004 100 9.7 0 0 9.6 NORMALIZED AMPLITUDE (OUTPUT (g)/REFERENCE(g)) 10 Figure 7. Sensitivity Distribution at 25C 10000 100000 1000 NOISE PSD (g/Hz) 10000 1000 100 100 10 1 0.1 1 10 FREQUENCY (Hz) 100 16596-005 1 0.01 1k 100k Figure 8. Noise PSD Above 100 Hz Figure 5. Noise Power Spectral Density (PSD) Below 10 Hz vs. Frequency 5 1 -1 -30 -10 10 30 50 70 90 110 TEMPERATURE (C) 130 150 0.6 0.4 0.2 0 -0.2 -0.4 -0.6 -0.8 -1.0 16596-006 -3 0.8 10 30 50 70 90 110 130 150 170 INPUT ACCELERATION (g) Figure 9. Sensitivity Nonlinearity vs. Input Acceleration Figure 6. Sensitivity vs. Temperature Rev. 0 | Page 6 of 14 190 16596-009 SENSITIVITY ERROR (% of Full-Scale) 1.0 3 -5 -50 10k FREQUENCY (Hz) 16596-008 10 SENSITIVITY CHANGE (% FULL SCALE) POWER SPECTRAL DENSITY (g/Hz) DUT1 DUT2 Data Sheet ADXL1003 40 10 PERCENT OF POPULATION (%) 8 NORMALIZED OFFSET (g) 6 4 2 0 -2 -4 35 30 25 20 15 10 -6 -20 0 20 40 60 80 100 120 TEMPERATURE (C) 0 16596-010 -10 -40 2.45 2.46 2.47 2.48 2.49 2.5 2.51 2.52 2.53 2.54 2.55 0 g OUTPUT DISTRIBUTION (V) Figure 13. 0 g Offset Histogram at 25C Figure 10. Normalized Offset vs. Temperature 280 1000 950 900 850 800 750 700 650 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 5.2 5.4 SUPPLY VOLTAGE (V) 240 220 200 180 160 140 3.0 35 40 PERCENT OF POPULATION (%) 45 30 25 20 15 10 920 940 960 980 1000 1020 1040 1060 MEASURE MODE CURRENT (A) 3.8 4.0 4.2 4.4 4.6 4.8 5.0 5.2 5.4 35 30 25 20 15 10 5 0 16596-012 5 900 3.6 Figure 14. Standby Current vs. Supply Voltage 40 880 3.4 SUPPLY VOLTAGE (V) Figure 11. Measure Mode Supply Current vs. Supply Voltage 0 3.2 200 205 210 215 220 225 230 235 240 245 STANDBY CURRENT (A) Figure 12. Measure Mode Current Histogram at 25C Figure 15. Standby Current Histogram at 25C Rev. 0 | Page 7 of 14 16596-015 600 260 16596-014 STANDBY MODE SUPPLY CURRENT (A) 1050 16596-011 MEASURE MODE SUPPLY CURRENT (A) 1100 PERCENT OF POPULATION (%) 16596-013 5 -8 ADXL1003 Data Sheet 500 3 INPUT ACCELERATION (g) XOUT 2 1 STANDBY 0 400 8 300 6 200 4 100 2 0 0 -1 5 10 15 20 25 30 35 40 TIME (s) -100 13.0 16596-016 0 13.5 14.0 14.5 15.0 15.5 16.0 16.5 17.0 17.5 -2 18.0 TIME (ms) Figure 16. XOUT Output Recovery from Standby Mode to Measure Mode Figure 18. Response to Overload Condition, XOUT Delta is Difference from Midscale Voltage 204 INTERNAL CLOCK FREQUENCY (kHz) 200.5 203 202 201 200 199 198 197 196 195 200.0 199.5 199.0 198.5 198.0 197.5 197.0 196.5 196.0 0 50 TEMPERATURE (C) 100 150 16596-017 194 -50 10 195.5 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 5.2 SUPPLY VOLTAGE (V) Figure 17. Internal Clock Frequency vs. Temperature at 5.0 V Supply Voltage (VDD) Rev. 0 | Page 8 of 14 Figure 19. Internal Clock Frequency vs. Supply Voltage at 25C 16596-019 VOLTAGE (V) 4 INTERNAL CLOCK FREQUENCY, VDD = 5.0V (kHz) REFERENCE XOUT DELTA OVERRANGE OUTPUT (V) 5 -2 12 600 16596-018 6 Data Sheet ADXL1003 THEORY OF OPERATION The ADXL1003 is a low noise, single-axis, MEMS accelerometer, with a 28 kHz resonant frequency that provides an analog output proportional to mechanical vibration. The ADXL1003 has a high g range of 200 g, suitable for vibration measurements in high bandwidth applications. Such applications include vibration analysis systems for monitoring and diagnosing machines or system health. The low noise and high frequency bandwidth allows the measurement of vibration patterns caused by small moving components, such as internal bearings. The high g range provides the dynamic range necessary for 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 ADXL1003 is a ratiometric output. Therefore, supply voltage modulation affects the output. Use a properly decoupled, stable supply voltage to power the ADXL1003 and to provide a reference voltage for the digitizing system. The output signal is impacted by an overrange stimulus. An overload indicator output feature indicates a condition that is critical for an intelligent measurement system. For more information about the overrange features, see the Overrange section. Proper mounting ensures 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 use a sensor stud mount system while considering the mechanical interface of fixing the ADXL1003 in the stud. For lower frequencies (below the full capable bandwidth of the sensor), it may be 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 required to achieve optimal results. Understanding the measurement frequency range and managing overload conditions is important to achieve accurate results. The electrical output signal of the ADXL1003 requires some band limiting and a 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. Differential capacitors that consist of independent fixed plates and plates attached to the moving mass measure the deflection of the structure. Acceleration deflects the structure and unbalances the differential capacitor, resulting in a sensor output with amplitude proportional to acceleration. Phase sensitive demodulation determines the magnitude and polarity of the acceleration. OPERATING MODES The ADXL1003 has two operating modes: measure mode and standby mode. Measure mode provides a continuous analog output for active monitoring. Standby mode is a nonoperational, low power mode. Measure Mode Measure mode is the normal operating mode of the ADXL1003. 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 Mode Placing the ADXL1003 in standby mode suspends the measurement and reduces the internal current consumption to 225 A (typical for the 5.0 V supply). The transition time from standby to measure mode is <50 s. Figure 16 shows the transition from standby to measure mode. BANDWIDTH The ADXL1003 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 ADXL1003 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 ADXL1003 Data Sheet APPLICATIONS INFORMATION 4. 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 a load 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. 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 21. 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. 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. 120 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. SELF TEST DELTA (mV) 100 OPTIONAL LOW-PASS FILTER R VOUT C VSS 80 60 40 20 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 5.0 16596-021 0 5.2 SUPPLY VOLTAGE (V) Figure 21. Typical Self Test Delta vs. Supply Voltage ADXL1003 RATIOMETRIC OUTPUT VOLTAGE OR ST (ACTIVE HIGH) STANDBY (ACTIVE HIGH) Figure 20. Application Circuit ON DEMAND SELF TEST A fully integrated electromechanical self test function is designed into the ADXL1003. 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. The ADXL1003 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 10 mV/g for the ADXL1003. The zero g bias output is ratiometric also and is nominally midscale relative to the supply voltage (VDD/2). The self test feature can be exercised by the user with the following steps: 1. 2. 3. 12 11 10 9 8 7 6 5 Measure the output voltage. Turn on self test by setting the ST pin to VDD. Measure the output again. 4 3.0 3.2 3.4 3.6 3.8 4.0 4.2 4.4 4.6 4.8 SUPPLY VOLTAGE (V) Figure 22. Sensitivity vs. Supply Voltage Rev. 0 | Page 10 of 14 5.0 5.2 16596-022 1F SENSITIVITY (mV/g) VDD (3.0V TO 5.25V SUPPLY VOLTAGE) 16596-020 The ADXL1003 was tested and specified at VDD = 5.0 V. However, the ADXL1003 can be powered with VDD as low as 3.0 V or as high as 5.25 V. Some performance parameters change as the supply voltage is varied. Data Sheet ADXL1003 INTERFACING ANALOG OUTPUT BELOW 10 kHz The ADXL1003 senses mechanical motion along a single axis and produces a voltage output. The system performance depends on the output response resulting from sense mechanical vibration 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 (the location on the PCB as well as the 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 ADXL1003 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-ADXL1003Z evaluation boards can be used as a reference. The ADXL1003 electrical output supports a bandwidth beyond the resonance of the sensor. The small signal bandwidth of the output amplifier in the ADXL1003 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 desired bandwidth and the chosen analog-to-digital converter (ADC) sampling rate be faster than the amplifier bandwidth. The output amplifier is ratiometric to the supply voltage, and there are two distinct cases regarding digital conversion, as follows: The user has an 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. 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 ADXL1003 output amplifier is stable while driving capacitive loads of up to 100 pF directly without a series resistor. At loads greater than 100 pF, an 8 k or greater series resistor must be used. See Figure 23 for an example of the interface, including components when measuring mechanical vibration from 0 kHz to 5 kHz. For a 5 kHz pass band, a single-pole resistor capacitor (RC) filter is acceptable. However, in some applications, use of a more aggressive filter and lower sample ADC sample rate is possible. The following components are recommended to form a 5 kHz low-pass RC filter at the output of the ADXL1003 when interfacing to an ADC, such as the ADAQ7980: R1 = 5 k, C1 = 5 nF, R2 = 0 , and C2 are not required. A minimum ADC sample rate of 20 kHz is recommended to avoid aliasing. When using sampling rates less than the resonance frequency (typically 28 kHz), be aware and account for the effective gain at the output of the sensor due to the resonance to ensure out of band signals are properly attenuated and do not alias into the band. See Figure 23 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 ADXL1003: R1 = 5 k, C1 = 2 nF, R2 = 5 k, and C2 = 2 nF. A minimum ADC sample rate of 200 kHz is recommended to avoid aliasing. Rev. 0 | Page 11 of 14 ADXL1003 Data Sheet 3.0V TO 5.0V* ADR4530 ADR4533 ADR4540 ADR4550 V+ REF REF_OUT LDO_OUT VDD 0.1F (+1F,OPTIONAL) LDO PD_REF 10F R1 ADXL1003 R2 IN+ C1 VIO 20 IN- ADC C2 SDI SCK SDO 1.8nF AMP_OUT PD_LDO 2.2F CNV V- *3.0V PD_AMP ADCN GND LIMITED BY ADXL1003; 5.0V LIMITED BY ADAQ7980/ADAQ79888 16596-023 ADAQ7980/ADAQ7988 Figure 23. Application Circuit for the ADXL1003 INTERFACING ANALOG OUTPUT BEYOND 10 kHz The ADXL1003 is a high frequency, single-axis MEMS accelerometer that provides an output signal pass band beyond the resonance frequency range of the sensor. Although the output 3 dB frequency response bandwidth is approximately 15 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 frequency, the ADXL1003 output amplifier supports a 70 kHz small signal bandwidth, which is well beyond the resonant frequency of the sensor. Although a mechanical interface is always important to achieve accurate and repeatable results in MEMS applications, it is critical 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. When using the ADXL1003 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, the aliasing 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 28 kHz. Approaching and above this frequency, the output response to an input stimulus peaks, as shown in Figure 4. When frequencies are near or above the resonance, the output response is outside the linear response range and 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 ADXL1003 output amplifier small signal bandwidth is 70 kHz. The user must 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 2x the equivalent noise bandwidth (ENBW) for a single-pole, low-pass filter, as follows: ENBW = (/2) x 70 kHz 110 kHz The sample rate must be at least 220 kHz. This sample rate reduces 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. Another artifact 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. 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 ADXL1003 and a digitization sample rate of at least 4x the desired bandwidth, assuming there is 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. Rev. 0 | Page 12 of 14 Data Sheet ADXL1003 The ADXL1003 has an output (OR pin) to signal when an overrange event (when acceleration is greater than 2x the full-scale range) occurs. Built in overrange detection circuitry provides an alert to indicate a significant overrange event occurred that is larger 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 (see Figure 18). 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 effectively invisible to the accelerometer. Multiple mounting points, close to the sensor, and a thicker PCB help reduce the effect of system resonance on the performance of the sensor. ACCELEROMETERS PCB MOUNTING POINTS MECHANICAL CONSIDERATIONS FOR MOUNTING Figure 24. Incorrectly Placed Accelerometers Mount the ADXL1003 on the PCB in a location close to a hard mounting point of the PCB. Mounting the ADXL1003 at an unsupported PCB location, as shown in Figure 24 may result LAYOUT AND DESIGN RECOMMENDATIONS Figure 25 shows the recommended PCB land pattern. 0.03/0.755mm 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 10 11 12 0.146/3.7mm 0.191/4.855mm 13 14 15 16 0.146/3.7mm 0.191/4.855mm Figure 25. PCB Land Pattern Rev. 0 | Page 13 of 14 16596-025 0.02/0.5mm 16596-024 OVERRANGE ADXL1003 Data Sheet OUTLINE DIMENSIONS DETAIL A (JEDEC 95) PIN 1 INDICATOR 0.30 0.25 0.20 25 32 24 1 0.50 BSC 3.80 3.70 SQ 3.60 EXPOSED PAD 8 17 TOP VIEW *1.85 1.80 1.75 0.45 0.40 0.35 0.05 MAX 0.02 NOM COPLANARITY 0.08 0.203 REF PKG-004829 SEATING PLANE PIN 1 INDIC ATOR AREA OPTIONS (SEE DETAIL A) 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 5.10 5.00 SQ 4.90 Figure 26. 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 ADXL1003BCPZ ADXL1003BCPZ-RL ADXL1003BCPZ-RL7 EVAL-ADXL1003Z 1 Temperature Range -40C to +125C -40C to +125C -40C to +125C g Range 200 g 200 g 200 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] ADXL1003 Evaluation Board Z = RoHS Compliant Part. (c)2018 Analog Devices, Inc. All rights reserved. Trademarks and registered trademarks are the property of their respective owners. D16596-0-8/18(0) Rev. 0 | Page 14 of 14 Package Option CP-32-26 CP-32-26 CP-32-26