Essay Writing Service

1 Star2 Stars3 Stars4 Stars5 Stars (No Ratings Yet)
Loading...

Short Term Isothermally Aged Doped Lead Free Solder Reliability

Short Term Isothermally Aged Doped Lead Free Solder Reliability subjected to Mechanical Testing with Semiparametric Estimation

ABSTRACT

This paper discusses the performance of lead (Pb) free alloy qualifications through mechanical vibration and drop testing. With the environmental directives prohibiting the use of hazardous substances, electronic packaging industry is mitigating away from alloys containing Pb. Environments in the aerospace and automotive industry are particularly harsh and the products must withstand those conditions and perform with high reliability. Through collaboration with the high reliability solder manufacturers and governmental agencies, a wealth of knowledge in lead free solder reliability was gained. A sound methodology has been developed to carry out reliability comparisons of different Pb-free alloys. This paper discusses the statistical methodology for evaluating solder alloy performance for high reliability products using survival analysis and gives a detailed description of the technique for vibration and drop experimental setup. The vibration and drop test failure results are based on 15 mm CABGA 208 components and show failure modes that might give an indication of what material combinations fail early. Both vibration and drop testing results showed a similar failure progression based on resistance measurements. The lowest failure rate and highest failure rates are highlighted in this paper. The SAC305 alloy is the baseline material in this paper.

KEYWORDS: BGA, PCB, Lead-free, Reliability, Isothermal Aging, Flip Chip, Doping, Weibull Analysis.

NOMENCLATURE

AIC  Akaike Information Criterion

ASTM  American Society for Testing & Materials

BGA  Ball Grid Array

CABGA  Chip Array Ball Grid Array

FR  Flame Retardant

IPC Association Connecting Electronics Industries

JEDEC  Joint Electron Device Engineering Council

KM  Kaplan-Meier

OSP  Organic Solderability Preservative

PCB  Printed Circuit Board

PH  Proportional Hazard

QFN  Quad Flat No-Lead package

RoHS  Restriction of Hazardous Substances

SAE  Society of Automotive Engineers

SBC  Schwarz Bayesian Criterion

SEM  Scanning Electron Microscopy

TV  Test Vehicle

Symbols

Ag  Silver

Bi  Bismuth

Cu  Copper

Ni  Nickel

Pb  Lead

Sb  Antimony

Sn  Tin

SAC105  98.5%Sn – 1.0%Ag – 0.5%Cu

SAC305  96.5%Sn – 3.0%Ag – 0.5%Cu

SAC405  95.5%Sn – 4.0%Ag – 0.5%Cu

Greek Symbols

β Slope

η Characteristic Life

ρ Probability plot

Hazard rate

INTRODUCTION

Aerospace and automotive companies are being restricted to provide RoHS compliant solders. Their products require new innovative solders that withstand the higher vibration, drop/shock performance, and combined-environment reliability compared to the conventional SAC305 alloys. The NASA-DoD Lead-Free Electronics Project confirmed that pad cratering is one of the dominant failure modes that occur in various board level reliability tests, especially under dynamic loading[1]. The mechanical properties (and related failure behavior) of lead free solder joints vary on different materials in electronic assemblies. Prior work by the Center for Advanced Vehicle and Extreme Environment Electronics (CAVE3) at Auburn University has shown rapid deterioration in the material behavior of SAC105, SAC305 and Innolot solder joints for varying elevated temperatures and aging durations[2].

Binary Eutectic Tin-Lead solders known for their low cost, low melting point, good wetting and mechanical properties have better reliability than that of lead free solder interconnections when subjected to harsh environments. The mechanical and thermo-mechanical properties of solder joints can be improved by possible selection of the Pb-free solders with lower process temperatures. Several ternary and quaternary alloy compositions that melt about 10°C lower than that of the SAC305 alloy. These solder compositions have shown great mechanical and thermos mechanical reliability[1].

From an initial list of 200 alloys studied for metallurgical performance, a total of 24 lead free solder alloy and paste combinations divided amongst 3 reflow profiles and 2 stencil thicknesses were chosen for experiments on mechanical and thermo-mechanical harsh environment test. The results demonstrated a significant degradation of the performance of the solder joints after aging for 6 months at 125°C. The degradation were charted through the survival analysis methodology using SAS software.

Test Vehicle

The test vehicle (TV) printed circuit board is a standard 4 layer FR4 material composed of copper vias, glass epoxy covered by a thin solder mask on both sides and an overlaid silkscreen for labeling. The test boards had 16 BGA’s, 20 Resistors and 6 QFN’s respectively. The plating material used for the test vehicles is Organic Solderability Preservative (OSP). CABGA 208 from Amkor were the primary components chosen for the study. The test vehicle is designed with the capability to house multiple packages of CABGA 208 to provide an opportunity to investigate the effect of material alloy, reflow profile and solder paste of components combined with solder paste volume.

Fig. 1 Test Vehicle Design

The BGA alloys used in this test consists of SAC105, SAC305 and doped alloys. Since the SAC 305 alloy is the most widely used industrial standard, the test is categorized as 75% SAC 305 alloys and 25% of either SAC 105 or special doped re-balled alloys. The test board is grouped into 4 different zones as the stress levels from the experiments are different across the different locations of the test boards. The zone and special alloy locations are highlighted in the following diagram.[3]

Fig. 2 Test Vehicle Zonal Mapping[2]

For evaluating the performance of various doped solders alloys used with the BGA components, we chose the 15 mm CABGA 208 package. It consisted of 208 interconnections with a 0.8mm lead pitch. The solder balls were 0.46mm diameter in 17X17 matrix with standard daisy chain wiring scheme. Figure 3 gives a detailed description of the same.

Fig. 3 CABGA 208 Component Design[3]

The other components in the board design are the 2512 resistors as it is the most common and the largest size used in the electronics industry, and an Amkor MicroLeadFrame QFN, it

contained 20 leads at 0.65 mm pitch with total size 5mm2.[3]

After the components are assembled, the test boards are placed

in the Vitronics SMR-800 reflow oven. Three different reflow

profile have been used such as Low, Best and High. The set

point temperatures in different zones and conveyor speed for

each profile.

Table 1. Reflow Profile Set Point Temperature

The reflow profile limits are shown in Table 2. The values used in Low and High profile are shown under Low limit and High Limit. The values for Best profile is set between Low and High Limit. The thermal profiling equipment used here is KIC 2000.

Table 2. Reflow Process Window Limits

The test build matrix and solder composition for paste and alloys are listed in the Table 3.

Table 3. Test Build Matrix

EXPERIMENTAL SETUP

Vibration Test Setup

In order to vibrate the test vehicles, a fixture system had to be designed that would be capable of holding the boards perpendicular to the electro-dynamic shaker table along the X and Y-axes while vibrating in the Z-direction.  Each test vehicle was affixed to one side of the aluminum fixture using a set of four screws, four washers, and four either large or small spacers. To determine the natural frequency of the fixtures and test vehicles, a laser source and an oscilloscope is used upon the boards to measure the natural frequencies of the test assembly. In this setup, the laser source records the frequencies induced with the highest distortions. Only the first natural frequency that induces more damage to the components is measured in this test. The setup test assemblies were mounted on an LDS LV217 electro-dynamic shaker table and subject to a 4.6 Grms constant stress vibration profile. According to the results, one major natural frequency appears at between 350 to 400 HZ. The magnitudes are very consistent with close peak magnitudes.[4][2]

Fig. 4 Aluminum Fixture Bode Plot

Fig. 5 Vibration Stress Profile 4.6 Grms

The test component resistance was hand probed with a hand-held ohmmeter for electrical continuity once every 60 minutes.

D:DropboxSolder Doping Project 2013ImagesIMAG0344.jpg

Fig. 6 Vibration Test Setup

Drop Test Setup

Lansmont M23 shock test drop tower was used to conduct the drop test experiment. SAE F5/ASTM 26Rl standard felt was used to absorb the generated shock in this experiment. JEDEC BS 111 test standard was followed, the maximum peak acceleration of 1500G and half-sine impact pulse duration of 0.5 milliseconds was maintained throughout the experiment. Test vehicles were dropped to a maximum of 300 times. The test component resistance were hand probed with a hand held ohmmeter for electrical continuity once every 20 drops.[5]

Fig. 7 Drop Test Profile 1500G, 0.5ms

Failures were recorded as per the IPC-SM785 standard, i.e., the solder joint failure can be defined as an interruption of electrical continuity greater than 1000 ohms.

STATISTICAL ANALYSIS OF TEST RESULTS

Vibration Test

First, it is important to understand the quantities necessary for survival/reliability analysis. Consider T to be the time until certain event. The survival function S(t), also known as the reliability function, is described as the probability of an entity surviving beyond time t. Survival function is given by

St=P(T>t)           (1)

S(t) is a monotone non-increasing function. When T is a continuous random variable with probability density function f(t), the Cumulative Distribution Function is given by

Ft=PT≤t                               (2)

The survival function can be found by

St=∫t∞fudu=1-F(t)          (3)

ft=-dS(t)dt            (4)

For example, the probability density function of T following Weibull distribution is

ft=βηtηβ-1e-t/ηβ

And the reliability function S(t) is,

St=e-tηβ

,

t>0

Where the slope,

β>0and characteristic lifetime, η

>0are the parameters. Another quantity known as Hazard function is denoted as the conditional failure rate in reliability or instantaneous rate of failure,

ht=lim∆t→0⁡P[t≤T<t+∆t|T≥t]∆t                                  (5)

From the above equation, h(t)∆t is the approximate probability that an entity at time t experiencing an event in [t, t+∆t). The restriction on hazard function, h(t) is that it is nonnegative, h(t) ≥ 0 [23].   The hazard function relates to the survival function and pdf in the following way,

ht=ftSt.  In Weibull case,

ht=βγtγβ-1,t>0.

Cumulative Hazard Function is probability of event up to time t.

Hx=∫0xh(u)du                               (6)

In the Weibull case,

Hx=xβ, x>0, β>0.

Consider the data of sample size n consists of triple {(Tj,

δj, Zj (t)), j = 1 to n} where Tj is the time on study for the jth subjects,

δis an event indicator for jth individual (

δj = 1 if the event occurred and

δj = 0 if censored) and Zj(t) is the vector of

pcovariates, which may depend on time. Let h (t | Z) be the hazard rate for an entity with risk vector Z at time t. The proportional hazard rate model proposed by Cox [11] is given below.

htZ=hotexp⁡(βTZ)=hotexp⁡(∑kβkZK)       (7)

where

hotis an arbitrary baseline hazard rate, β is the vector of p parameters. This is known as Semiparametric model because the parametric form is assumed only for covariate effect. Cox model which is also known as proportional hazard model for two entities with covariate values Z and Z*, because the ratio of their hazard rates is

htZ htZ*=hotexp⁡(∑k=1pβkZk)hotexp⁡(∑k=1pβkZk*)=exp⁡[∑k=1pβk(Zk-Zk*)](8)

which is a constant. Hence hazard rates are proportional. Equation (8) is called Hazard ratio of an entity with risk factor Z experiencing the event as compared to an entity with risk factor Z* [23].

This experiment, presented in this paper, consists of five categorical variables such as Paste combination, Stencil thickness, Profile and Aging Time. The censor indicator used in this experiment is delta. The Time to Failure measured in hours is considered as response variable.

Table 4. Data description

Variables Description
Paste Combination Solder paste with Solder alloy combination
Reflow Profile Reflow Profile – Low, High and Best
Aging Time Aging Time – 0months and 6months
Delta Status at endpoint (1=Failed, 0=Censored)
Failure Time to Failure (Hours)

The proportional hazard model has been used to fit the above data. The model is implemented using SAS®. Initially the global hypothesis is tested such that all coefficients are 0. The data set also consisted of tied observations. The Breslow method is used to calculate the log likelihood with tied observations [23]. Let t1<t2<…. <tbe the distinct and ordered event times. Let di be the number of failures at ti. Di is set of all individuals which fails at ti. Let sbe the sum of vectors Zj over all individuals which fails at ti where si=

∑j∈DiZj. The set of all individuals at risk prior to tis Ri.

Lβ=∏i=1Dexp⁡(βtsi)∑jϵRiexp⁡(βtZj)di

(9)

The log likelihood is calculated by taking logarithm of above equation and multiplying by -2. The criterion -2logL is obtained using above method. Akaike Information Criterion (AIC) shown in Table 4 examines likelihood and number of parameters included in model.

AIC=-2logL+2k                   (10)

Where k is number of covariates. Here the smaller the AIC, the better the data fits into model. Schwartz Bayesian Criterion (SBC) gives more penalization for additional covariates, where SBC is found using Equation (11).

SBC=-2logL+klogn

(11)

Table 5. Model Fit Statistics

The hypothesis is tested to check whether all the coefficients (βi’s) are 0. The likelihood ratio statistics is calculated by finding the difference between -2logL with covariates and without covariates as shown in Table 5. The Likelihood ratio statistics is shown to be 547.1201. The null hypothesis is rejected indicating at least one of the coefficient is not 0.

Table 6. Test of Global Hypothesis

The Table 5 shows the effect of covariates and its interaction which contribute to distribution of time to failure response. It is shown that there is significant difference in time to failure among each levels of Paste combination and Aging time at 5% level of significance. Also, variation is observed among each level of Reflow profile and Aging time. The effect of Aging is also significant. In this paper, it is important to understand the effect of aging for each solder material combination. In terms of interaction, there is difference in time to failure for interaction between Paste combination, Reflow profile and Aging with Wald Chi-Square as 26.9532 and p-value 0.0026.

Table 7. Overall Test of variables

The effect of interaction in terms of Reflow profile, paste combination and Aging are calculated using Semiparametric estimates based on Maximum Likelihood Estimate.

Table 8. Analysis of Maximum Likelihood Estimates for 15mm BGA package

Table 8 shows the calculation of estimates for 15mm BGA package. It is found that the Baseline SAC305 follows proportional hazard assumption and its p-value is 0.0137. The parameter estimates for the Baseline SAC305 is -0.69076. The Hazard ratio is derived by taking the exponential of the parameter estimates. The hazard ratio or relative risk is defined as the ratio of hazard rate in treatment group versus control group. In this case, for the baseline SAC305, after taking the exponential of parameter estimate we get 0.501. It means the relative risk for the SAC305 material with SAC305 solder alloy after 6months of isothermal aging at 125°C will increase the failure by 0.501 times compared with after assembled material.

From the Table 8, it is shown that all the solder pastes except SAC405-Dopants+SAC305, SACX+SAC305, SAC Doped Sb+SAC105 (6mil Low profile) and SAC Doped Sb+SAC305 (6mil Low profile and 4mil Best profile) follow proportional hazard assumption.

Fig. 8 Weibull graph for vibration reliability of SAC405-Dopants paste with SAC105 solder alloy

The lower relative risk is seen in Baseline SAC305 with SAC305 and SAC105 solder alloy. The relative risk is higher for Sn-3.8Ag-0.7Cu-3Bi-1.5Sb-0.02Ni and SAC doped Sb+SAC105 after aging.

The vibration reliability for SAC405-Dopants paste with SAC105 solder alloy between No Aged and Aged is shown in Figure 8. This graph is converted into Product limit estimator as shown in Figure 9.

Fig. 9 Kaplan Meier Survival curve comparison for SAC405-Dopants paste with SAC105 solder alloy

The nonparametric estimation of survival function is performed using Kaplan Meier method which is commonly known as Product Limit estimator [24]. The hypothesis is tested to determine whether the survivor function, S1(t)= S2(t)=…=SK(t) for all t,  are the same for all paste materials. It is defined such that for all values of t in the range where there is data:

Ŝt=1                                if t<t1 ∏ti≤t1-diYi          if t1≤t

(12)

Where di is the number of events and Yi is the number at risk. The survival probability is calculated at each event. The Product Limit Estimator is a step function with jumps at the observed event times. The size of these jumps depends not only on number of events observed at each event time ti, but also on pattern of censored observations prior to ti. The log-rank test is used to compare the survivor function between groups. It is well known and widely used to test the difference in survivor function [23]. For the two group comparison, the log-rank test is capable of comparing survivor function of the form

S1(t)=S2(t)γ                                     (13)

Where

γis a positive number other than 1. In this paper, the Product Limit survival curves are compared to test the hypothesis that there is a difference in survival function among each solder pastes using log-rank test.

Fig. 10 Kaplan Meier Survival curve comparison for SAC405-Dopants paste with SAC105 and SAC305 solder alloy

Figure 10 shows the comparison for 15mm BGA package using SAC405-Dopants with SAC105 and SAC305 solder alloys. After assembly, it is seen that SAC105 solder alloy performed better than SAC305 alloy under vibration condition. Looking into 6month aged performance there is a crossover between two alloys at 8 hours. After 8 hours of testing, SAC305 alloy did better than SAC105 alloy. There is a difference in survivor function among the pastes with different alloys.

The Product Limit survival curve for Sn-3.8Ag-0.7Cu-3Bi-1.4Sb-0.15Ni paste with SAC105 and SAC305 is shown in Figure 11. The survival curve for No aged alloy is substantially higher compared with its aged. The solder paste Sn-3.8Ag-0.7Cu-3Bi-1.4Sb-0.15Ni with SAC105 solder alloy outperformed SAC305 after assembly. But after 6months of aging, Sn-3.8Ag-0.7Cu-3Bi-1.4Sb-0.15Ni with SAC305 solder alloy did better than SAC105 solder alloy. There is a strong evidence that survivor functions are different for at least one of the solder pastes after assembly and post aging.

Fig. 11 Kaplan Meier Survival curve comparison for Sn-3.8Ag-0.7Cu-3Bi-1.4Sb-0.15Ni paste with SAC105 and SAC305 solder alloy

The vibration performance of 15mm BGA with SAC Doped Sb paste is shown in Figure 12. The SAC105 solder alloy is significantly better than SAC305 solder alloy at No aged. The survival curves for aged materials is close to each other which is opposite to what appeared in after assembly performance. The p-value is less than 0.0001 suggesting that there is difference in survival function.

Fig. 12 Kaplan Meier Survival curve comparison for SAC Doped Sb paste with SAC105 and SAC305 solder alloy

Figure 13 shows the comparison for 15mm BGA package using SACX with SAC105 and SAC305 solder alloys. After assembly, it is seen that SAC105 solder alloy performed better than SAC305 alloy under vibration condition. Looking into 6month aged performance, SAC305 alloy did better than SAC105 alloy. There is a difference in survivor function among the pastes with different alloys.

Fig. 13 Kaplan Meier Survival curve comparison for SACX paste with SAC105 and SAC305 solder alloy

After assembly, it is seen that Sn-3.8Ag-0.7Cu-3Bi-1.5Sb-0.02Ni solder with SAC105 solder alloy is more reliable than SAC305 under vibration testing. After aging, it appears that SAC305 performed well until 7 hours and later SAC105 did better than SAC305 solder alloy. With p-value<0.0001, hypothesis testing indicates there is a difference in survival function among at least one of the solder mixes.

Fig. 14 Kaplan Meier Survival curve comparison for Sn-3.8Ag-0.7Cu-3Bi-1.5Sb-0.02Ni paste with SAC105 and SAC305 solder alloy

The results shown above indicates a significant difference in survivor function in at least one of solder pastes among all combinations. The graphs compare between each solder pastes. It is important to find the difference in survivor function between SAC305 solder paste and all the other pastes. Also, it is necessary to analyze the significant variation between pastes after assembly and after aging.

Pairwise comparison is being done between pastes and p-values are calculated for each comparison. Table 9 shows the p-value for each solder pastes compared with the baseline SAC305 solder paste with SAC305 solder alloy among No aged and 6month aged group as seen in 5th and 6th column of Table 9.

Table 9. Pairwise comparison of solder pastes compared with SAC305 solder paste (SAC305 solder alloy) and after aging

For example, the table shows there is significant difference between vibration performance of Sn-3.8Ag-0.7Cu-1.4Sb-0.15Ni-3Bi+SAC305 solder alloy (4mil stencil thickness) and SAC305 after assembly and after aging with p-values 0.0171 and less than 0.0001. The last column in the above table indicates the significance test for survival curves for each solder paste post-aging compared with its own after assembly. For instance, there is a strong evidence in difference in survival curve between No aged and 6month aged for all paste except SAC Doped Sb+SAC305 (4mil Best profile, 6mil Low profile), SAC405-Dopants+SAC305 solder alloy and SACX+SAC305 solder alloy.

Fig. 15 Characteristic lifetime comparison for 15mm BGA Best profile

The Characteristic lifetime summary of 15mm BGA package built with Best profile is shown in Figure 15. From that summary, Sn-3.8Ag-0.7Cu-3Bi-1.5Sb-0.02Ni solder paste is most reliable after assembly compared with all other solder paste built using Best profile. But they did not perform well after aging and SAC305 paste outperformed other solder pastes after 6 months of isothermal aging at 75C.

Fig. 16 Characteristic lifetime comparison for 15mm BGA High and Low profile

SAC doped Sb with SAC105 solder alloy did well after assembly compared with other paste as seen in Figure 16 but have shown severe deterioration after aging. Also, it is shown that Sn-3.8Ag-0.7Cu-3Bi-1.5Sb-0.02Ni solder paste is reliable for both no aged and aged.

Drop Test

Similar analysis is done using Product-Limit estimator approach. Solder materials using SAC305 solder alloy and Best profile after assembly are compared in Figure 17. SAC305 solder paste with SAC305 solder alloy outperformed other solder alloys. There is significant difference in survival curve for at least one of the solder paste with p-value less than 5%.

Fig. 17 Survival curve comparison for 15mm BGA package with SAC305 solder alloy subjected to drop shock after assembly

In terms of materials built using SAC105 solder alloy, SAC doped Sb have better performance than the other solder pastes as seen in Figure 18. There appears to be variation in survival function between solder pastes with SAC105 solder alloy with p-value less than 0.0001 as shown in Figure 18.

Fig. 18 Survival curve comparison for 15mm BGA package with SAC105 solder alloy subjected to drop shock after assembly

Figure 17 and 18 have shown the comparison of solder paste with SAC305 and SAC105 after assembly. Figure 19 shows the drop test survival curves for pastes with SAC305 solder alloy shown in Figure 20 after aging. With SAC305 solder alloy, Sn-3.8Ag-0.7Cu-1.4Sb-0.15Ni-3Bi solder is highly reliable after 6months of aging as seen in Figure 19. SAC305 solder paste is better after Sn-3.8Ag-0.7Cu-1.4Sb-0.15Ni-3Bi solder paste. The test of hypothesis that survival function of the solder pastes are the same is rejected at 5% level of significance.

Fig. 19 Survival curve comparison for 15mm BGA package with SAC305 solder alloy subjected to drop shock after aging

For the materials built using SAC105 solder alloy, SAC305 solder paste have better drop test performance along with 3.8Ag-0.7Cu-1.4Sb-0.15Ni-3Bi. SAC doped Sb with SAC105 solder alloy used to be reliable after assembly but the same paste with SAC105 solder alloy did not do well after 6months of aging as seen in Figure 20. The survival curve trend is significantly different from each other having p-value less than 0.0001.

Fig. 20 Survival curve comparison for 15mm BGA package with SAC105 solder alloy subjected to drop shock after aging

The Lifetime summary is provided in Figure 21 for all solder pastes with SAC105 and SAC305. SAC305 with SAC305 solder alloy have shown higher lifetime after assembly but severe degradation is seen after aging. For SAC105 solder alloy, SAC Doped Sb is highest for No aged.

Fig. 21 Lifetime comparison for 15mm BGA package subjected to Drop test

The interesting trend to note is that there is significant improvement in drop test after aging for 3.8Ag-0.7Cu-1.4Sb-0.15Ni-3Bi especially. There is slight improvement for SAC305 solder alloy with 3.8Ag-0.7Cu-1.5Sb-0.02Ni-3Bi solder paste after aging. After aging, it can be seen that 3.8Ag-0.7Cu-1.4Sb-0.15Ni-3Bi showed better reliability.

REFERENCES

[1] S. Subramaniam, P. Snugovsky, J. Kennedy, E. Kosiba, Z. Bagheri, and M. Romansky, “VIBRATION TESTING OF LEAD-FREE ALLOYS FOR HIGH RELIABILITY,” SMTA J., vol. 27, no. 4, pp. 31–47, 2014.

[2] S. Thirugnanasambandam, T. Sanders, A. Raj, D. Stone, J. Evans, G. Flowers, and J. Suhling, “The study of vibrational performance on different doped low creep lead free solder paste and solder ball grid array packages,” Thermomechanical Phenom. Electron. Syst. -Proceedings Intersoc. Conf., pp. 920–923, 2014.

[3] S. Sridhar, A. Raj, S. Gordon, S. Thirugnanasambandam, J. L. Evans, and W. Johnson, “Drop impact reliability testing of isothermally aged doped low creep lead-free solder paste alloys,” Proc. 15th Intersoc. Conf. Therm. Thermomechanical Phenom. Electron. Syst. ITherm 2016, pp. 501–506, 2016.

[4] D. Stone, “A Study of the Vibrational Reliability Performance of Different Doped Low-Creep Lead- Free Solder Paste and Solder BGA Packages,” Auburn University, 2015.

[5] S. Thirugnanasambandam, N. Vijayakumar, J. Zhang, J. Evans, D. Ph, F. Xie, and D. F. Baldwin, “DROP RELIABILITY TEST ON DIFFERENT DIMENSIONAL LEAD-FREE WAFER LEVEL CHIP SCALE PACKAGES,” in SMTA International, 2012.



Recommendation
EssayHub’s Community of Professional Tutors & Editors
Tutoring Service, EssayHub
Professional Essay Writers for Hire
Essay Writing Service, EssayPro
Professional Custom
Professional Custom Essay Writing Services
In need of qualified essay help online or professional assistance with your research paper?
Browsing the web for a reliable custom writing service to give you a hand with college assignment?
Out of time and require quick and moreover effective support with your term paper or dissertation?
Did you find someone who can help?

Fast, Quality and Secure Essay Writing Help 24/7!