The Policy

The use of Plastic Encapsulated Microcircuits (PEMs) is permitted on NASA Goddard Space Flight Center (GSFC) spaceflight applications, provided each use is thoroughly evaluated for thermal, mechanical, and radiation implications of the specific application and found to meet mission requirements.  PEMs shall be selected for their functional advantage and availability, not for cost saving; the steps necessary to ensure reliability usually negate any initial apparent cost advantage.  A PEM shall not be substituted for a form, fit and functional equivalent, high reliability, hermetic device in spaceflight applications.  

Due to the rapid change in wafer-level designs typical of commercial parts and the unknown traceability between packaging lots and wafer lots, lot specific testing is required for PEMs, unless specifically excepted by the Mission Assurance Requirements (MAR) for the project.  Lot specific qualification, screening, radiation hardness assurance analysis and/or testing, shall be consistent with the required reliability level as defined in the MAR.  

Developers proposing to use PEMs shall address the following items in their Performance Assurance Implementation Plan: source selection (manufacturers and distributors), storage conditions for all stages of use, packing, shipping and handling, electrostatic discharge (ESD), screening and qualification testing, derating, radiation hardness assurance, test house selection and control, data collection and retention.  Use of PEMs outside the manufacturer’s rated temperature range requires written approval from GSFC.  Specifically, PEMs must be:  

• stored under temperature controlled, clean conditions, protected from ESD and humidity,
• traceable to the branded manufacturer,
• procured from the manufacturer or their approved distributor,
• tested to verify compliance with the performance requirements of the application environment over the intended mission lifetime,
• tested using practices and facilities with demonstrated capabilities sufficient to handle and test the technologies involved.  

Testing in accordance with 311-INST-001 shall be performed as necessary to qualify and screen the devices, in order to verify compliance with the application requirements and project risk level, defined in the program Mission Assurance Requirements (MAR).  Radiation evaluation shall address all threats appropriate for the technology, application and environment, including Total Ionizing Dose (TID), Single Event Effects (SEE) and displacement damage.  Existing radiation data can be used only with the review and approval of the project radiation specialist.

PEMs with manufacture dates older than 3 years before the time of installation shall not be used without GSFC approval.  Derating of PEMs must be addressed with consideration of specific material, device construction, device characteristics and application requirements.  

Use of PEMs with pure tin plated terminations requires special precautions to preclude failures caused by tin whiskers.  GSFC approval of mitigation strategies is required.  

Exceptions to testing required by 311-INST-001 may be permitted by GSFC on a case-by-case basis, where it can be demonstrated that either existing lot specific test data show acceptable results, or the use of high risk PEMs represents low risk of functional loss should the part fail.  All rationale for such exceptions shall be documented.

NASA will use part performance data collected in accordance with this policy to evaluate the policy’s effectiveness and to develop recommendations for future improvements and streamlining.  

Discussion of Policy Content

Allowance of the Use of PEMs

Goddard Space Flight Center  (GSFC) is adopting the policy to allow the use of Plastic Encapsulated Microcircuits (PEMs) in spacelight applications, in order to have access to the latest technology advances, which are offered in PEMs and only rarely in hermetic high temperature packages.  The use of PEMs in space applications can bring advantages to the application in terms of performance and part availability but thorough evaluation of the thermal, mechanical and radiation environment, together with qualification testing and screening, are generally required to ensure reliability.  

Environmental Factors

In space environments, the advantages of PEMs are often accompanied by additional risks of failure.  This is especially true in applications where PEMs encounter stresses (temperature extremes, vacuum conditions, radiation, etc.) outside the operating conditions for which they were designed.  PEMs are typically designed to operate within two temperature ranges: -40°C to +85°C (industrial) or 0°C to +70°C (commercial) in contrast to military hermetic styles which are generally rated from -55°C to +125°C.  The PEMs manufacturer designs, selects materials and tests parts to meet the needs of their primary customers (commercial high volume) and principal end use environments (industrial or commercial) NOT for high reliability, spaceflight.  PEMs are generally intended for use in benign environments, where failure in service can be mitigated by replacement, maintainability or repairability.  

The sheer variety of materials and fabrication techniques that may be used in making PEMs represent a number of application concerns and reliability risks in space environments.  Often, these materials and manufacturing techniques have little or no spaceflight heritage.  

Typical environmental concerns for spaceflight include:

• Optical components are at risk of contamination from outgassing of the molding material in vacuum environments.   
• Molding materials have varying glass transition temperatures that can affect their maximum testing and use temperature.     
• Molding material dimensional creepage with respect to temperature and time may place stress on internal bonds and wires.     
• Thermal cycling can generate mechanical stresses on bonds, die paddles etc. as a result of Coefficient of Thermal Expansion (CTE) differences between materials.  
• Molding material in PEMs is usually in contact with the die and may contain ions or be the source of secondary emissions that can influence radiation susceptibility of the die.  These factors can result in PEMs packaged die having lower radiation tolerance than the same die, hermetically packaged.  
• Operation outside the device’s nominal rated temperature range may result in rapid deterioration of package properties, weakening of bonds, mechanical overstressing of the die, etc.     
• PEMs are all susceptible to moisture ingress and absorption but to varying degrees dependent on design, materials and processing.  While not a risk for corrosion or other deterioration in space vacuum, moisture can promote corrosion during storage and ground level processing as well as “popcorning” during soldering, if strict moisture controls are not enacted.

Other, PEMs Specific Reliability And Application Issues

High volume PEMs parts have steadily decreasing time periods between their introduction and obsolescence.  Various reports put the current time to obsolescence in the range of nine to eighteen months.  Rapid obsolescence has a number of impacts for spaceflight applications.  Reprocurement to cover shortfalls or test fallout may not occur before the parts are no longer available.  Good experience on one project is not useful for another project with similar needs, if the parts are no longer available.  Multiple spacecraft programs with builds spaced years apart, require a one time, multi spacecraft buy or may require different parts for successive spacecraft.

Not only is rapid obsolescence an issue but this has also led to corresponding reduction in designed operational life of the die design.  Reports say that PEMs designers are now operating on as short as a five year life expectancy.  Design compromises, such as reduced metallization and oxide layer thicknesses reduce costs but also reliability and life expectancy.  It has been reported that mask changes can occur as frequently as once a month.  Die shrink changes are known to have dramatically impacted radiation tolerance.

PEMs manufacturers generally utilize continuous improvement philosophies that result in frequent, unannounced changes to designs, materials and processes.  While improving performance or cost for the target commercial applications, such changes can have unconsidered, negative impacts for space applications.

For high reliability products, MIL specifications define the assignment of lot date codes and the composition of the corresponding lot.  There is no such recognized definition for a PEMs lot, it is up to the manufacturer and the needs of their intended market.  This situation results in traceability and lot sample testing concerns for the spaceflight user.  It has been reported that as many as five distinctly different dice of the same function have been found in a single PEMs “lot”.  In addition, a single lot of PEMs can be processed in a number of die fabrication facilities, packaging and test houses throughout the world.  PEMs of the same nominal device type but manufactured through different flows in different facilities can be intermixed and marked with the same lot identification.  

Counterfeit parts are an increasing problem in the semiconductor market, particularly PEMs.  Such parts have an identical or near identical appearance to genuine parts but are known to have substandard quality, reliability and performance.  It is essential to evaluate each lot of parts and to use reputable distributors for PEMs, in order to reduce the risk of purchasing counterfeit parts.

Rationale for User Imposed Qualification Testing and Screening

PEMs are not governed by strict military standards that require inspection of the die for workmanship flaws and the performance of burn-in on each device to remove early random failures.  In lieu of piece part testing, PEMs vendors typically employ various sample based techniques for calculating reliability.  Testing may include proprietary testing regimes and employ unique rules governing sample sizes or the exclusion of failures from reliability calculations.  These variations can make it difficult to compare PEM reliability data from vendor to vendor, let alone from PEM to hermetically sealed part.  

For these reasons, it is not prudent to rely solely on unvalidated reliability data from PEM vendors.  Screening of PEMs is essential before they are inserted into most flight hardware.  The most important element in screening for reduced reliability risk for PEMs is burn-in.

Burn-in at the piece part level addresses infant mortality, which represents a significant problem for space applications, and provides some insight into lot reliability and quality.  If burn-in at the parts level is not performed, these needs must be addressed by another test approach agreed to by the project, such as board level burn-in, or board/box level Environmental Stress Screening.  

The argument that burn-in of PEMs should be avoided as it reduces the Total Ionizing Dose (TID) resistance of PEMs should be rejected unless solid evidence is produced to support the claim.  Most studies have shown burn-in to have an impact of 500 rads or less.  To properly evaluate TID of burned-in parts, the TID test samples must be burned-in prior to testing.  

Radiation Lot Acceptance Testing (RLAT) of PEMs should be performed independently of any data that may exist for equivalent or similar hermetically sealed devices, and should be performed under the direction of the project radiation specialist.  This is necessary as market conditions may drive unannounced process changes, creating differences in radiation response.  It may be possible to dispense with single-event qualification of the PEM if data exist for the hermetic device.  However, because PEMs are passivated with nitride layers, which are known to be responsible for TID sensitivity to Pre-irradiatiation Elevated Thermal Stresses (PETS), TID characterization should always be independently performed .  

Testing and Qualification

Testing and qualification of EEE parts for space applications are usually performed to requirements specific to the risk level desired for the application.  Three risk levels are currently defined for NASA GSFC Projects.  Risk Level One has the lowest inherent risk and is intended for critical applications such as single-string, single-point failure and mission essential functions.  Risk Level Two has an increased risk and is intended for general-purpose spaceflight applications, although use in single-string and single-point failure applications may be permissible with Project approval.  Risk Level Three has an unknown risk due to the lack of formalized reliability assessment, screening and qualification, and due to unreported and frequent changes in design, construction and materials.  These inherent risk levels can be modified by additional testing such that Level 3 parts can be elevated to Level 2, and Level 2 to Level 2+.  Upgrading to Level 1 is theoretically impossible due to lot specific controls imposed during Level 1 manufacturing that cannot be imposed once the part has been finished.

Testing of PEMs should be tailored to the application and based on such factors as manufacturer history, analysis of materials, flight history, technology maturity, application criticality and redundancy.

Basic Screening

Minimum additional testing for PEMs is established in GSFC 311-INST-001 for mission risk levels 1, 2 and 3.  A flow chart of the evaluation process is shown in Appendix I of this document.  

Basic process flow:  

• External visual inspection for workmanship defects such as, bubbles or voids in the plastic package, separation of the package from the terminations, lead corrosion etc.
•   Radiography and C-mode Scanning Acoustic Microscopy (CSAM) to inspect for swept bond wires, delaminations, voids, damaged or displaced die, mixed die sizes and shapes etc.
• Functional testing to ensure parts meet requirements over the full application range of temperature, power, frequency, voltage etc.
• Destructive Physical Analysis (DPA for PEMs) on samples from each lot to inspect for internal workmanship, bond pull, step coverage, die passivation, metallization voids, corrosion, contamination etc.
• Radiation testing on samples from each lot.  Each lot needs to be characterized for Total Ionization Dose (TID), Single Event Effects (SEE) and Displacement Damage from charged particles.

Qualification

Qualification is required on a lot-by-lot basis unless objective evidence is provided that qualification data for a previous lot of the same or similar devices is applicable to the lot in question.  When qualification testing is required, GSFC 311-INST-001 defines risk level specific requirements.

Typical qualification flows include:

• Operational life test to simulate performance under application conditions and duration; may also be used to estimate life failure rate.  
• Highly Accelerated Stress Testing (HAST), subjects parts to high levels of temperature and humidity to accelerate destructive processes such as corrosion, delamination and die attachment failure, detectable by post-HAST Destructive Physical Analysis (DPA).  Preconditioning of the samples that includes solder exposure, is recommended.

Additional Testing to Lower the Risk of PEMs 

In addition to the basic process flow described previously, the following additional tests shall be performed as required by GSFC 311-INST-001 and as tailored to the PEM for its application, based on project requirements:

• Temperature Cycling to excite material CTE mismatches and stress wirebonds, die attachment etc.
• HAST or Temperature Humidity Bias to evaluate package integrity
• Burn-in for longer duration or at higher stress levels than the basic requirement
• Post test analysis consisting of PEMs specific DPA.  CSAM may also be required.

Exceptions to Testing 

Reductions to the testing listed in 311-INST-001 may be permitted with project approval on a case-by-case basis, where it can be demonstrated that:

• Existing test data for the delivered lot date code demonstrates acceptable results.  
• Use of PEMs represents low risk of functional loss should the part fail.  Low risk is defined as low application criticality or low potential for loss based on such things as:

          - light duty cycle

          - benign environment (minimal temperature extremes, radiation exposure, etc)

          - more than one redundant circuit

          -  short mission life

          - and low mission cost.  

All rationale for such exceptions must be documented.

Age Control

Due to molding material creepage, risk of corrosion, material aging etc., it is necessary to limit the age of PEMs to no more than three years from date of manufacture to date of installation, unless otherwise permitted by the project. Exceptions for age control may be granted by the project based on a need for the performance characteristics of older codes, or to use PEMs in inventory that are no longer in production.  

Recommended Processing for Storage and Use of PEMs 

• PEMs must be baked out prior to storage and prior to use in order to drive out absorbed moisture from the plastic molding material.  Storage should be in dry environments (Nitrogen purged) at room temperature.
• The developer shall clean and dry boards using solvents and baking methods that will not risk compromising the reliability of parts or boards.
• The terminations of PEMs should be pretinned using tin–lead solder to reduce the risk of tin whisker growth or to remove gold plating.  PEMs typically have pure tin plated terminations, which are a risk for tin whisker growth and subsequent system failure due to shorting or plasma arcs.  Alternatively PEMs may be available with gold plated terminations, which are at risk for failure due to gold embrittlement. 
• After installation and cleaning, the application of conformal coating to the devices is recommended to minimize re-absorption of moisture and to further reduce the risk of tin whisker growth.  

Use of Off-the-Shelf Assemblies Containing PEMs 

Use and function of off-the-shelf units or assemblies that contain PEMs should be analyzed for mission criticality.  When loss of “off the shelf” units does not compromise mission success, on a case-by-case basis, these units may be considered exempt from additional PEMs testing requirements, subject to approval by the project.  However, additional unit level testing such as thermal cycling or thermal vacuum testing, may be directed by the project in lieu of additional part level screening.  

When failure of such units represents significant compromise to mission success, an analysis of the parts used within the units shall be performed.  The parts shall be evaluated for screening compliance to GSFC 311-INST-001, and include a radiation analysis.  Pending the results of this investigation, units may be required to undergo modification for use of higher reliability parts, additional shielding or replacement with radiation tolerant parts.  When no high reliability parts are available, additional testing of the unit may be required.  All parts upgrading or additional testing shall be subject to project approval.

If a “high-risk” designation is not acceptable for the application, then additional screening must be performed to ensure that the PEMs are consistent with a medium-risk or low-risk level as defined in the project MAR and GSFC 311-INST-001.

NASA REFERENCE DOCUMENTS 

Goddard  311-INST-001   Instructions for EEE Parts Selection, Screening, and Qualification

Rose, Virmani and Kadesch  (Goddard)   Plastic Encapulated Microcircuit (PEM) Guidelines for Screening and Qualification for Space Environments

S-311-M-70   Specification for Destructive Physical Analysis

NEPP Document TR04-0600   PEM Derating, Storage & Qualification Report

IPC-SC-60A   Post Solder Solvent Cleaning Handbook

  Appendix I: PEM evaluation process

1/ High risk may be acceptable if the impact of part failure on achieving mission goals is minimal.  

2/ Additional screening performed in accordance with GSFC 311-INST-001, or at project direction for the appropriate risk level.