Rajeshuni Ramesham, NASA JPL/Caltech, Rajeshuni.Ramesham@jpl.nasa.gov

Ashok Sharma, NASA GSFC, Ashok.K.Sharma.1@gsfc.nasa.gov

Alexander Teverovsky, 3 QSS Group, Inc./NASA GSFC, Alexander.Teverovsky@gsfc.nasa.gov

Abstract

Analog Devices ADXL 250 and 202, dual-axis, surface micromachined accelerometers were subjected to extreme environmental thermal cycling, beyond the manufacturer’s specifications of -65 oC to 150 oC, to evaluate the robustness of devices and packaging. The extreme environmental thermal cycling conditions were established as follows: (a) Total number of cumulative thermal cycles = 61; (b) temperature range of -125 oC to 90 oC; (c) ramp rate of 7 oC/minutes; and (d) dwell time of 10 minutes on each of the cold and hot sides. A non-destructive X-ray evaluation technique was used to image the accelerometers before and after thermal cycling. Functional tests were performed during temperature cycling, and failures were observed in some of the accelerometers. Analysis of the failed devices is still in progress, and results will be reported later.

Introduction

Micro-electromechanical systems are some of the most recent advanced technologies in microelectronics, which have paramount importance to military and aerospace applications. Accelerometers are the earliest and the most developed representatives of MEMS, which have been under use for air bag deployment applications in automobiles for more than a decade. The accelerometer sensitivity required for aerospace or military applications varies from 80,000 G to 10-6 G, and the applications include ballistic munitions launches, deep space probes, and attitude and position control. The presence of moving parts in MEMS, which are fabricated by using a surface micro-machining process on the surface of a silicon substrate, dictates a reliability assessment towards cycling fatigue of micro-mechanical components under extreme temperatures or beyond manufacturer specifications. Fatigue damage of micro-mechanical components will certainly influence the performance of the MEMS devices. Environmentally induced failures generated via thermal cycling are of significant importance to NASA if one can understand and mitigate the failures by improving the design.

Temperature Profile

Figure 1 shows the temperature profile employed to perform the extreme temperature thermal cycling of the ADXL 250 and the boards with ADXL 202 accelerometers.

ADXL 250 and 202

Analog devices ADXL 250 and 202 are dual axis surface micro-machined accelerometers rated for ±50 G and packaged in a hermetic 14 lead surface mount cerpack. The operating temperature range of these accelerometers is -55 oC to 125 oC and storage temperature range is from -65 oC to 150 oC. Figure 2 shows optical photographs of the some of the test vehicles with ADXL 202 and ADXL250 accelerometers.

Electrical Tests

ADXL accelerometers have a limited number of parameters specified, including sensitivity for X and Y channels, self test for X and Y channels measured as output change, and quiescent supply current. Tables 1 and 2 show the electrical characteristics of the several accelerometers after 27 and 61 cumulative thermal cycles, respectively. The sensor number 38 has failed after 27 thermal cycles as per data in Table 1. Failure analysis of this is still in progress. No further failures were observed after 61 cumulative thermal cycles as noted the results in Table 2.

X-ray Imaging

Fein focus X-ray imaging technique has been used to image the sensors and boards before and after thermal cycling. Figures 3, 4, and 5 show the X-ray images of the ADXL 202 and 250 before and after thermal cycling. No noticeable changes were observed in the X-ray images. X-ray imaging of sensor #38 will be performed to determine whether any failures internal to the package can be detected. No external cracking was observed as a function of thermal cycling. No damage was observed at the interconnects as a function of thermal cycling as per images in Figures 3 through 5.

Conclusions

Analog Devices ADXL 250 and 202 accelerometers were subjected to extreme temperature thermal cycling such as -125 oC to 90 oC. X-ray images of the sensors were taken before and after thermal cycling for 61 cycles. There were no physical failures observed externally or internal to the package. Only one sensor failed after 27 thermal cycles. At present we do not have destructive failure analysis (DPA) on the failed sensor device. This work is still continuing, and any noticeable findings will be reported in the future.

Acknowledgements

Thanks are due to Mr. Steve Bolin for his help in X-ray imaging prior to and after thermal cycling of the accelerometers. I would like to thank Dr. Reza Ghaffarian for his encouragement and support. This work is supported by NEPP to assess the reliability of the package.

Figure 1. Thermal cycle profile employed for thermal cycling.

 

Figure 2. Optical photographs of the some of the test vehicles with ADXL 202 and ADXL250 accelerometers.

 

S/N	Sensitivity X                        Y		Iq	Rin	Output change	Zero G bias level output bias voltage X                   Y
16	37.9mV	37.9mV	2.88 mA	53 K	0.340V	2.4558V	2.5064V
17	37.4 mV	35.0 mV	2.81 mA	54K	0.324V	2.4173V	2.4516V
22	36.2 mV	34.0 mV	2.71 mA	55K	0.358V	2.5351V	2.4849V
28	37.5 mV	35.3 mV	2.78 mA	55K	0.345V	2.4985V	2.5462V
32	39.0 mV	36.2 mV	2.61 mA	56K	0.325V	2.3664V	2.5047V
33	37.0 mV	35.8 mV	2.78 mA	55K	0.335V	2.4419V	2.4397V
35	37.3 mV	39.0 mV	2.89 mA	53K	0.337V	2.3719V	2.4350V
37	39.6 mV	37.0 mV	2.85 mA	53K	0.338V	2.4212V	2.4446V
39	37.2 mV	34.4 mV	2.80 mA	54K	0.335V	2.4126V	2.4084V

Table 2. ADXL 250 Analog Devices Dual Axis Accelerometer electrical
characteristics after 61 extreme temperature thermal cycling (-120 oC to 90 oC).

Figure 3. Typical X-ray images of the sensors shown in Figure 2B.

Figure 4. X-ray images of the sensors shown in Figure 2B after 23 extreme thermal cycles.

A	B
Figure 5. X-ray image of the ADXL250 accelerometer A (before) and B (after) 61 extreme temperature thermal cycles.