HIGHLY ACCELERATED LIFE TEST REPORT COM840

HIGHLY ACCELERATED LIFE TEST REPORT COM840 Prepared for: Ampro Computers, Inc. Tom Davenport Director of Quality & Engineering Services 5215 Hellyer Avenue Suite 110 San Jose, California 95138-1007 Phone: (408) 360-4307 TDavenport@ampro.com Page 1 of 37

Highly Accelerated Life Test Report Document Number 1379 Test Dates: 08/11-08/12/2008 COM840 Testing Performed for: Ampro Computers, Inc. Phone: (408) 360-4307 Tom Davenport Email: TDavenport@ampro.com 5215 Hellyer Avenue Suite 110 San Jose, California 95138-1007 Testing Performed by: Phone: (408) 737-7500 Carlos Carcamo Fax: (408) 850-1852 925 Thompson Place Email: carlos@reliantlabs.com Sunnyvale, CA 94085 Reference Documents Document Revision Number Date Issued Service Proposal - SV2395 5/12/2008 HALT Test Report - 1101 8/06/2007 Page 2 of 37

Table of Contents: 1. Objective...4 2. Scope...4 3. Acronyms and Definitions...4 4. Operating and Destruct Limits...5 5. Executive Summary...5 6. HALT Results...7 6.1. Cold Temperature Step Stress...7 6.2. Hot Temperature Step Stress...9 6.3. Rapid Thermal Transitions...11 6.4. Vibration Step Stress...15 6.5. Vibration Measurements...16 6.6. Combined Environment...18 7. Products Tested...20 8. Temperature Test Setup...20 9. Vibration Test Setup...21 10. Test Equipment and Calibration Records...22 10.1. Customer Test Equipment...22 10.2. Reliant Labs Calibration Records...22 11. Functional Test Description...23 11.1. Failure Characterization...23 11.2. Fault Recovery...23 Appendix A - Photographs...24 Appendix B IR Thermal Imaging...32 Appendix C - Vibration Plots...33 Appendix D Screen Captures...37 Page 3 of 37

1. Objective Highly Accelerated Life Testing (H.A.L.T.) is performed to uncover latent defects in product design, component selection, and/or manufacturing that would not otherwise be found through conventional qualification methods. The process subjects the test product to progressively higher stress levels, incorporating thermal dwells, rapid temperature transitions, vibration, and a combination of temperature and vibration to precipitate inherent defects. Moreover, H.A.L.T. stresses the product to failure in order to assess design robustness and margin above its intended operation. 2. Scope To find the weak links in product design, document failure modes, and determine the true operating and destruct limits using repeatable testing techniques such as thermal step stress, rapid thermal transitions, vibration step stress, and combined temperature and vibration environments. By subjecting the product to increasing levels of stress, long term failure modes that would show up under normal operating conditions in months or years can be revealed in just hours or days. An essential component of H.A.L.T. is root cause analysis and the identification and implementation of corrective action to ensure the product integrity, thus increasing the product s reliability and the robustness of design. 3. Acronyms and Definitions H.A.L.T. Highly accelerated life test H.A.S.S. Highly accelerated stress screen UUT Unit under test Operating Limit (OL) - The operating limit is defined as the last operational temperature or vibration set point prior to failure. Destruct Limit (DL) - The destruct limit is the level at which the product stops functioning and remains inoperable at normal operating conditions. Grms Gravity Root Means Squared Page 4 of 37

4. Operating and Destruct Limits Table 1: Defined Operating and Destruct Limits Stress Condition Chamber Setpoint Cold Temperature OL <-60 C (-76 F) Hot Temperature OL +90 C (+194 F) Vibration OL >50 Grms Cold Temperature DL <-60 C (-76 F) Hot Temperature DL >+100 C (+212 F) Vibration DL >50 Grms 5. Executive Summary Failure modes and observations discovered during the HALT process: Cold Thermal Step Stress: The COM840 remained operational during the cold temperature test down to -60 C. Hot Thermal Step Stress: The COM840 was subjected to the hot temperature step stress up to +100 C and remained operational to +90ºC. At the end of the +100ºC temperature step, a Memory (RAM) Error verifying data in RAM was displayed. The unit recovered at room temperature (+23 C). Rapid Thermal Transitions: The COM840 was subjected to five rapid temperature cycles from -60 C to +100 C at a set transition rate of 60 C per minute. During first cycle while transitioning to +100ºC, a Memory (RAM) Error verifying data in RAM was displayed. During the fourth cycle at -60ºC, distortion on the video output was observed (black horizontal lines). The unit was power cycled at +30ºC and the video output recovered; however, the failure mode repeated when the temperature transitioned to +100ºC. The unit recovered after being power cycled at +20ºC at the end of the fourth temperature cycle and remained operational for the remaining of the temperature cycling test. A picture showing the distortion on the video output is documented in Appendix D. Page 5 of 37

Vibration Step Stress: The COM840 remained operational during the vibration step stress up to 50 Grms. Combined Environment: The COM840 was subjected to five combined environment cycles from -60 C to +90 C with a set transition rate of 60 C per minute. The vibration started at 10 Grms and was increased in 10 Grms steps for each additional cycle. Distortion (black horizontal lines) on the video output was observed during the first cycle at -60ºC and 10 Grms. The combined environment test was paused and the unit was power cycled at ambient temperature with no vibration and recovered. After completing the first cycle of combined environment, unit SN_09 was removed from the chamber replaced with SN_30. The combined environment test was re-started. The new unit remained operational throughout all five combined environment cycles. During the combined environment, the chamber was stopped after the second cycle and during the fourth cycle to secure the Heatsink and fan screws. Page 6 of 37

6. HALT Results 6.1. Cold Temperature Step Stress The cold temperature step stress began at +20 C and decreased in 10 C increments. The dwell time at each thermal step was 15 minutes. The dwell time began after the chamber control thermocouples on the UUT stabilized. Airflow from the chamber was directed onto the UUT using 3 aluminum ducting. During the cold temperature test, a cover was placed over the COM840 to reduce the airflow to <100CFM. Pictures illustrating the temperature step stress setup are located in Appendix A. The UUT was monitored for functionality throughout the cold temperature step stress process. Lower Operating Limit (LOL): Lower Destruct Limit: (LDL): <-60 C / -76 F <-60 C / -76 F Table 2: Cold Step Stress Results Chamber Set point ( C) Pass/Fail +20 Pass Unit tested: SN_09. +10 Pass 0 Pass -10 Pass -20 Pass -30 Pass -40 Pass -50 Pass -60 Pass +20 Pass Comments Page 7 of 37

Table 3: Temperatures Recorded During the Cold Temperature Step Stress ( C) Ambient Between CPU & North Bridge Ethernet Set point Control CPU South Bridge Controller U1 +20 C 20.2 28.2 29.3 26.0 37.0 33.5 +10 C 10.4 18.3 19.0 16.1 27.4 24.3 0 C 0.3 8.5 9.6 6.3 18.0 14.8-10 C -9.9-1.8-0.8-3.9 8.3 5.1-20 C -19.8-11.2-10.2-13.4-1.3-4.2-30 C -29.9-21.0-19.7-23.2-11.0-13.8-40 C -39.7-30.6-29.4-32.8-20.2-23.2-50 C -50.1-38.6-37.9-40.9-29.0-30.8-60 C -60.1-47.1-45.4-49.5-38.3-38.8 Note: Thermocouple Locations are documented in Table 13. PCB Temperature 50.0 Cold Temperature Step Stress 40.0 30.0 20.0 10.0 Temperature (C) 0.0-10.0-20.0-30.0-40.0-50.0-60.0-70.0 8:57 9:04 9:11 9:19 9:26 9:33 9:41 9:48 9:55 10:03 10:10 10:17 10:25 10:32 10:39 10:47 10:54 11:01 11:09 11:16 11:23 11:31 Time Control CPU Ambient Between CPU & North Bridge South Bridge Ethernet Controller U1 PCB Temp. Figure 1 Cold Step Stress Chart Page 8 of 37

6.2. Hot Temperature Step Stress The hot temperature step stress began at +30 C and increased in 10 C increments. The dwell time at each thermal step was 15 minutes. The dwell time began after the chamber control thermocouples on the UUT stabilized. Airflow from the chamber was directed onto the UUT using 3 aluminum ducting. During the hot temperature test, a cover was placed over the COM840 to reduce the airflow to <100CFM. Pictures illustrating the temperature step stress setup are located in Appendix A. The UUT was monitored for functionality throughout the hot temperature step stress process. Upper Operating Limit (UOL): Upper Destruct Limit (UDL): +90 C / +194 F >+100 C / +212 F Table 4: Hot Step Stress Results Chamber Set point ( C) Pass/Fail +30 Pass Unit tested: SN_09. +40 Pass +50 Pass +60 Pass +70 Pass +80 Pass +90 Pass +100 Fail +20 Recover Comments A Memory (RAM) error, Error verifying data in RAM, was displayed at the end of the dwell. The error was cleared on the BurnInTest and there were more errors displayed. Page 9 of 37

Table 5: Temperatures Recorded During the Hot Temperature Step Stress ( C) Ambient Between CPU & North Bridge Ethernet Set point Control CPU South Bridge Controller U1 +30 C 30.1 37.4 37.9 35.2 43.3 39.9 +40 C 40.0 47.3 47.8 45.0 53.1 49.6 +50 C 49.9 57.5 58.1 55.2 63.0 59.4 +60 C 60.0 67.9 68.3 65.4 72.8 69.1 +70 C 69.9 78.2 78.6 75.6 82.7 78.9 +80 C 79.9 88.7 89.1 85.9 92.8 88.9 +90 C 89.9 96.5 97.3 94.7 102.1 97.3 +100 C 100.0 106.8 107.4 104.8 111.9 106.8 Note: Thermocouple Locations are documented in Table 13. PCB Temperature 120.0 Hot Temperature Step Stress 110.0 100.0 90.0 80.0 Temperature (C) 70.0 60.0 50.0 40.0 30.0 20.0 10.0 0.0 11:34 11:42 11:49 11:57 12:04 12:12 12:19 12:27 12:34 12:42 12:49 12:57 13:04 13:12 13:19 13:27 13:34 13:42 13:49 13:57 Time Control CPU Ambient Between CPU & North Bridge South Bridge Ethernet Controller U1 PCB Temp. Figure 2 Hot Step Stress Chart Page 10 of 37

6.3. Rapid Thermal Transitions The UUT was exposed to 5 rapid temperature cycles from -60 C to +100 C. The dwell time at each extreme was 15 minutes and the thermal transition rate was set to 60 C per minute. The actual chamber air thermal transition rate was 61 C per minute. The rate of change calculated using the product response thermocouples was 29 C per minute. This transition rate was calculated by averaging the product response thermocouples. The UUT was monitored for functionality throughout the rapid thermal transition process. Upper Set Point: +100 C / +212 F Lower Set Point: -60 C / -76 F Calculated Chamber Air Rate of Change: Calculated Product Response Rate of Change: 61 C / minute 29 C / minute Table 6: Rapid Thermal Transition Results Cycle Chamber Set point ( C) Pass/Fail 0 +20 Pass Unit tested: SN_09. 1 Transition Pass 1-60 Pass 1 Transition Pass 1 +100 Pass 1/2 Transition Fail 2-60 Pass 2 Transition Pass 2 +100 Pass 2/3 Transition Pass 3-60 Pass 3 Transition Pass 3 +100 Pass 3/4 Transition Pass 4-60 Fail 4 Transition No Change Comments A Memory (RAM) error, Error verifying data in RAM, was displayed at the end of the dwell. At ~10 minutes into the dwell, horizontal black lines were observed on the video output. The unit was power cycled but did not recover. A picture showing the video output are documented in Appendix D. Page 11 of 37

Cycle Chamber Set point ( C) Pass/Fail 4 +30 Recover 4 Transition Pass Comments The unit was power cycled and recovered; however, when the BurnInTest was started, the unit rebooted. It was suspected that the unit could not properly boot up due to the internal temperature being too cold. The unit was allowed ~3 minutes to warm up and the BurnInTest was successfully launched. At ~5 minutes into the dwell, the video output failure mode previously described repeated. 4 +100 Fail 4 Transition No Change 4 +20 Recover The unit was power cycled and recovered. 4/5 Transition Pass 5-60 Pass 5 Transition Pass 5 +100 Pass 5 Transition Pass +20 Pass Page 12 of 37

120.0 Rapid Thermal Transitions 100.0 80.0 60.0 Temperature (C) 40.0 20.0 0.0-20.0-40.0-60.0-80.0 14:00 14:09 14:17 14:26 14:35 14:43 14:52 15:01 15:09 15:18 15:27 15:35 15:44 15:53 16:01 16:10 16:19 16:27 16:36 8:18 8:26 8:35 8:44 8:52 Time Control CPU Ambient Between CPU & North Bridge South Bridge Ethernet Controller U1 PCB Temp. Figure 3 Rapid Thermal Transition Chart Page 13 of 37

120.0 Average Rate of Change 80.0 100.0 80.0 60.0 60.0 40.0 Temperature (C) 40.0 20.0 0.0 20.0 0.0-20.0 Rate of Change (ROC) -20.0-40.0-40.0-60.0-60.0-80.0 14:00 14:17 14:33 14:50 15:07 15:23 15:40 15:57 16:13 16:30 8:20 8:36 8:53-80.0 Time Air Temp Product Temp Air ROC Product ROC Figure 4 Air and Product Rate of Change Chart Page 14 of 37

6.4. Vibration Step Stress The UUT was mounted to the vibration table using one 3/8 aluminum plate with aluminum stand-offs to elevate the unit. The plate was bolted to the vibration table at eight locations using 3/8-1 bolts. Pictures illustrating the fixture setup are located in Appendix A. Accelerometers were attached to the UUT to measure the vibration response on the product. The location of each accelerometer is located in Table 10. Pictures illustrating the locations of the accelerometers are located in Appendix A. The vibration step stress began at a set point of 5 Grms and was increased in 5 Grms increments. The dwell time at each set point was 15 minutes. After the 30 Grms level, the vibration was decreased to 5 Grms to determine if a failure occurred that was not detected at the higher vibration level. If a failure was not detected during this tickle vibration, the vibration was increased to the next level. The temperature was set to 25 C. The UUT was monitored for functionality throughout the test. Table 7: Vibration Step Stress Results Chamber Set point (Grms) Pass/Fail 0 Pass Unit tested: SN_09. 5 Pass 10 Pass 15 Pass 20 Pass 25 Pass 30 Pass 5 Pass 35 Pass 5 Pass 40 Pass 5 Pass 45 Pass 5 Pass 50 Pass 5 Pass 0 Pass Comments Page 15 of 37

6.5. Vibration Measurements Vibration levels were measured at three product locations during the vibration step stress. The vibration measurements on the product were taken using a National Instruments spectrum analyzer. The bandwidths used to calculate the Grms levels on the product were 2Hz to 2,500Hz and 2Hz to 10,000Hz using a digital filter at 2,500Hz and 10,000Hz. During each vibration step, a power spectral density plot was taken showing the calculated Grms level on the product and the distribution of energy over the desired frequency bands 2Hz to 2,500Hz and 2Hz to 10,000Hz. The Grms levels measured on the product are shown in the table below and the spectral density plots are located in Appendix B (See Table 10 for accelerometer locations). Table 8: Vibration Levels Measured During Vibration Step Stress (Grms) Chamber Channel 1 Channel 2 Channel 3 Channel 4 Setpoint 2Hz-5k 2Hz-2.5k 2Hz-10k 2Hz-2.5k 2Hz-10k 2Hz-2.5k 2Hz-10k 2Hz-2.5k 2Hz-10k 5 5.72 5.08 5.80 6.22 6.49 2.76 3.20 1.11 1.14 10 11.97 9.64 12.51 12.37 13.23 5.47 6.87 1.45 1.64 15 18.22 13.96 19.80 17.87 19.48 18.05 10.45 2.08 2.58 20 24.54 18.70 27.01 27.03 29.59 10.57 14.92 2.82 3.77 25 29.74 22.31 34.07 33.65 37.00 13.66 19.32 3.66 5.13 30 35.72 27.16 41.44 44.39 48.01 15.43 22.97 4.33 6.41 35 39.44 25.02 46.18 46.16 50.35 17.97 25.55 5.74 8.06 40 46.27 36.02 53.86 50.28 55.05 19.49 28.21 6.07 8.07 45 51.91 40.52 62.59 65.83 70.72 22.84 32.25 6.16 9.38 50 54.12 41.31 63.83 56.80 62.50 23.54 33.07 6.14 9.20 Table 9: Transmissibility Ratio (Energy at Product (response) vs. Input Energy) Transmissibility of Energy from Vibration Table to Product Chamber Setpoint Channel 1 Input Channel 2 Response / Input Channel 3 Response / Input Channel 4 Response / Input 10 9.64 1.28 0.57 0.15 30 27.16 1.63 0.57 0.16 50 41.31 1.37 0.57 0.15 Note: Calculation made using Channel 1 (2 Hz 2.5 khz) frequency band. Page 16 of 37

Table 10: Accelerometer Placement Spectrum Analyzer Channel Assignment Channel Axis Location or Description 1 Z Mounted underneath the vibration table at center. 2 Z Base Board Mounted on the PCB next to J27. 3 Z Base Board Mounted on the PCB between U17 and the battery. 4 Z COM840 Mounted on the Heatsink. Note: The X-axis runs through the chamber doors, the Y-axis runs from the access portal to air plenum, and the Z-axis is vertical to the vibration table. 55.0 Vibration Step Stress 50.0 45.0 Temperature (C) / Vibration (grms) 40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 9:34 9:43 9:51 10:00 10:09 10:18 10:27 10:36 10:44 10:53 11:02 11:11 11:20 11:29 11:37 11:46 11:55 12:04 12:13 12:22 Time Vibe Setpoint Vibe Product Control CPU Ambient Between CPU & North Bridge South Bridge Ethernet Controller U1 PCB Temp. Figure 5 Vibration Step Stress Chart Page 17 of 37

6.6. Combined Environment The UUT was exposed to 5 rapid temperature cycles from -60 C to +90 C combined with vibration. The vibration level was set to 10 Grms for the first temperature cycle and then increased in 10 Grms increments before each additional cycle. The dwell time at each temperature extreme was 15 minutes and the thermal transition rate was set to 60 C per minute. The actual chamber air thermal transition rate was 61 C (This transition rate was calculated during the rapid thermal transition process). The UUT was monitored for functionality throughout the combined environment process. Table 11: Combined Environment Results Cycle Chamber Set point ( C) Chamber Set point (Grms) Pass/Fail 0 +20 0 Unit tested: SN_09. 1 Transition 0/10 1-60 10 Fail 1 Transition 10/0 +20 0 Recover 1 Transition 0/10 Pass 1-60 10 Pass 1 Transition 10 Pass 1 +90 10 Pass 1/2 Transition 10/20 Pass 2-60 20 Pass 2 Transition 20 Pass 2 +90 20 Pass 2 Transition 20/0 Pass 2 +20 0 Pass 2/3 Transition 20/30 Pass 3-60 30 Pass 3 Transition 30 Pass 3 +90 30 Pass 3/4 Transition 30/40 Pass 4-60 40 Pass Comments At ~10 minutes into the dwell, black horizontal lines were displayed on the video output. The unit was recovered after being power cycled. Unit SN_09 was removed from the chamber and replaced with SN_30. Two loose screws were observed from outside the chamber. Four out of five screws on the heatsink became lose. The screws were then secured with Locktite. All four holding screws on the fan on top the heatsink were loose. Page 18 of 37

Cycle Chamber Set point ( C) Chamber Set point (Grms) Pass/Fail 4 Transition 40 Pass 4 +90 40 Pass 4/5 Transition 40/50 Pass 5-60 50 Pass 5 Transition 50 Pass 5 +90 50 Pass 5 Transition 50/0 Pass +20 0 Pass Comments The chamber was stopped to secure the screws with Locktite. When the unit was turned on, the compact flash drive could not be recognized. The compact flash drive was replaced and the unit recovered. 100.0 Combined Temperature and Vibration 90.0 80.0 70.0 60.0 Temperature (C) / Vibration (Grms) 50.0 40.0 30.0 20.0 10.0 0.0-10.0-20.0-30.0-40.0-50.0-60.0-70.0 12:50 13:01 13:13 13:25 14:02 14:14 14:26 14:37 14:49 15:01 15:12 15:24 15:36 15:47 15:59 16:11 16:22 16:34 16:46 16:57 17:09 17:21 Time Temp Setpoint Temp Product Vibe Setpoint Vibe Product Figure 6 Combined Environment Chart Page 19 of 37

7. Products Tested The units subjected to the H.A.L.T process are documented below. Table 12: Product Identification Description Model Number Revision Serial Number COM840 COM-840-R-31 - PRWKBRA3018009 COM840 COM-840-R-31 - PRWKBRA4098030 8. Temperature Test Setup During the thermal testing, a cover was placed over the COM840 to reduce the airflow to <100CFM. Thermocouples were attached to the UUT to monitor the temperature of the product at various locations. The locations of the thermocouple placements are located in the table below. Pictures illustrating the temperature test setup are located in Appendix A. Table 13: Data collection points for temperature monitoring Agilent Data logger Channel Assignment Channel Type Location or Description 1 TC Next to chamber control thermocouple. 2 TC COM840 Monitoring temperature on CPU. 3 TC COM840 Monitoring ambient temperature between CPU and North Bridge. 4 TC COM840 Monitoring temperature on South Bridge. 5 TC Base Board Monitoring temperature on Ethernet Controller U1. 6 TC Base Board Monitoring PCB temperature next to J17. Page 20 of 37

9. Vibration Test Setup The UUT was mounted to the vibration table using one 3/8 aluminum plate with aluminum stand-offs to elevate the unit. The plate was bolted to the vibration table at eight locations using 3/8-1 bolts. Pictures illustrating the fixture setup are located in Appendix A. Accelerometers were attached to the UUT to measure the vibration response on the product at various locations. The location of each accelerometer is located in the table below. Detailed pictures illustrating the vibration fixture and the locations of accelerometers are located in Appendix A. Table 14: Accelerometer Locations Spectrum Analyzer Channel Assignment Channel Axis Location or Description 1 Z Mounted underneath the vibration table at center. 2 Z Base Board Mounted on the PCB next to J27. 3 Z Base Board Mounted on the PCB between U17 and the battery. 4 X COM840 Mounted on the side of Heatsink. Note: The X-axis runs through the chamber doors, the Y-axis runs from the access portal to air plenum, and the Z-axis is vertical to the vibration table. Page 21 of 37

10. Test Equipment and Calibration Records 10.1. Customer Test Equipment Functional test equipment provided by the customer and used to conduct the H.A.L.T. testing is detailed below. The model number and serial number have been documented to aid in the reproducibility and repeatability of test results. Table 15: Customer Test Equipment Description Manufacturer Model Number Serial Number 500 Watt Power Supply Antec BP500U F07080406560 Monitor Dell - - Mouse Belkin - - Keyboard Belkin - - 10.2. Reliant Labs Calibration Records Equipment used to conduct the H.A.L.T. testing is detailed below. All test equipment that requires periodic calibration was in current calibration at time of test. The calibration certifications are traceable to the National Institute of Standards and Technology (NIST). Table 16: Reliant Labs Calibration Records Lab 5 Description Manufacturer Model Number Serial Number Calibration Due HALT Chamber Chart TVC-9 TVC9-01-ON1 02/11/2009 Data Logger Agilent 34970A MY41031262 08/13/2009 Thermocouples Omega C03-T-60 / TT-T-30 NA NA Accelerometer Control Dytran 3030B5 9783 03/07/2009 Accelerometer Ch. 1 Dytran 3030B5 9476 11/29/2008 Accelerometer Ch. 2 PCB 352A21 41738 09/14/2008 Accelerometer Ch. 3 PCB 352A21 57293 09/14/2008 Accelerometer Ch. 4 PCB 352A21 74973 07/21/2009 Accelerometer Ch. 5 Dytran 3035A1G 945 09/14/2008 Accelerometer Ch. 6 Dytran 3035A1G 949 01/25/2009 Spectrum Analyzer/DAQ National Instruments 187570H-01 135EAE1 05/27/2009 Spectrum Analyzer/Signal Cond. Dytran 4120 299 05/20/2009 Video Microscope PTEM Zoom 70 29-95-91 NA Page 22 of 37

11. Functional Test Description During the HALT the UUT was continually monitored for functionality. When a failure occurred the test diagnostics logged the failure data and all attempts to determine the cause of failure were made. Test equipment used to monitor the UUT was set up on a bench outside the H.A.L.T. chamber. Cables, wires, and power cords were fed through the chamber access port. Before running the H.A.L.T., the following modifications were made to the UUT: Functional test cables were modified to reach outside the chamber. BurnInTest Version 5.0 Pro was used to monitor functional test results. Functional test files were stored on the compact flash residing on the UUT. Table 17: Functional test list for COM840 using BurnInTest Version 5.0 Pro 1 CPU Math 7 Network 2 2 CPU SIMD 8 Video Playback 3 Memory (RAM) 9 Serial Port 1 4 2D Graphics 10 Serial Port 2 5 3D Graphics 11 Serial Port 3 6 Network 1 12 Serial Port 4 11.1. Failure Characterization During the HALT process, any abnormalities are considered possible failure modes until they can be determined otherwise. To establish clear failure characterization the list below details acceptable specifications and/or unit functionality. Monitor the functional test software for errors or failures. 11.2. Fault Recovery When a failure occurred the environmental stress was reduced to determine when the UUT recovered. Once the UUT recovered, the environmental stress was increased to determine if the failure mode was repeatable. 1. If the UUT was still functional after repeating the failure mode, the environmental stress was increased until additional failure modes were determined or a hard failure occurred. 2. If the UUT was not functional after repeating the failure mode, an attempt was made to mask the failure and continue testing in pursuant of additional failure modes. Page 23 of 37

Appendix A - Photographs Figure A.1: Thermal Test Setup Page 24 of 37

Figure A.2: Control Thermocouple Location Page 25 of 37

Figure A.3: Vibration Test Setup Page 26 of 37

Figure A.4: Location of Accelerometer 2 Page 27 of 37

Figure A.5: Location of Accelerometer 3 Page 28 of 37

Figure A.6: Location of Accelerometer 4 Page 29 of 37

Figure A.7: Test Equipment Setup Page 30 of 37

Figure A.8: Chamber Lab Setup Page 31 of 37

Appendix B IR Thermal Imaging U1 (~36 C) Base Board PCB (~36.26 C) Fan Battery U11 (~33 C) Figure B.1: Temperature Scan for COM840 Page 32 of 37

Appendix C - Vibration Plots Vibration Plot - 5 Grms 2 Hz 2.5 khz 1.0E+00 Power Spectral Density 1.0E-01 1.0E-02 Grms2/Hz 1.0E-03 1.0E-04 1.0E-05 1.0E-06 0 500 1000 1500 2000 2500 Frequency (Hz) Channel 1 (5.08 Grms) Channel 2 (6.22 Grms) Channel 3 (2.76 Grms) Channel 4 (1.11 Grms) Spectrum Analyzer Channel Assignment Channel Axis Location or Description 1 Z Mounted underneath the vibration table at center. 2 Z Base Board Mounted on the PCB next to J27. 3 Z Base Board Mounted on the PCB between U17 and the battery. 4 Z COM840 Mounted on the Heatsink. Note: The X-axis runs through the chamber doors, the Y-axis runs from the access portal to air plenum, and the Z-axis is vertical to the vibration table. Page 33 of 37

Vibration Plot - 5 Grms 2 Hz 10 khz 1.0E+00 Power Spectral Density 1.0E-01 1.0E-02 Grms2/Hz 1.0E-03 1.0E-04 1.0E-05 1.0E-06 1.0E-07 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 Frequency (Hz) Channel 1 (5.80 Grms) Channel 2 (6.49 Grms) Channel 3 (3.20 Grms) Channel 4 (1.14 Grms) Spectrum Analyzer Channel Assignment Channel Axis Location or Description 1 Z Mounted underneath the vibration table at center. 2 Z Base Board Mounted on the PCB next to J27. 3 Z Base Board Mounted on the PCB between U17 and the battery. 4 Z COM840 Mounted on the Heatsink. Note: The X-axis runs through the chamber doors, the Y-axis runs from the access portal to air plenum, and the Z-axis is vertical to the vibration table. Page 34 of 37

Vibration Plot - 50 Grms 2 Hz 2.5 khz 1.0E+02 Power Spectral Density 1.0E+01 1.0E+00 Grms2/Hz 1.0E-01 1.0E-02 1.0E-03 1.0E-04 0 500 1000 1500 2000 2500 Frequency (Hz) Channel 1 (41.31 Grms) Channel 2 (56.80 Grms) Channel 3 (23.54 Grms) Channel 4 (6.14 Grms) Spectrum Analyzer Channel Assignment Channel Axis Location or Description 1 Z Mounted underneath the vibration table at center. 2 Z Base Board Mounted on the PCB next to J27. 3 Z Base Board Mounted on the PCB between U17 and the battery. 4 Z COM840 Mounted on the Heatsink. Note: The X-axis runs through the chamber doors, the Y-axis runs from the access portal to air plenum, and the Z-axis is vertical to the vibration table. Page 35 of 37

Vibration Plot - 50 Grms 2 Hz 10 khz 1.0E+02 Power Spectral Density 1.0E+01 1.0E+00 Grms2/Hz 1.0E-01 1.0E-02 1.0E-03 1.0E-04 0 1000 2000 3000 4000 5000 6000 7000 8000 9000 10000 Frequency (Hz) Channel 1 (63.83 Grms) Channel 2 (62.50 Grms) Channel 3 (33.07 Grms) Channel 4 (9.20 Grms) Spectrum Analyzer Channel Assignment Channel Axis Location or Description 1 Z Mounted underneath the vibration table at center. 2 Z Base Board Mounted on the PCB next to J27. 3 Z Base Board Mounted on the PCB between U17 and the battery. 4 Z COM840 Mounted on the Heatsink. Note: The X-axis runs through the chamber doors, the Y-axis runs from the access portal to air plenum, and the Z-axis is vertical to the vibration table. Page 36 of 37

Appendix D Screen Captures Figure D.1: Horizontal Black Lines Observed During Rapid Temperature Cycles Page 37 of 37