Equipment Reliability Institute
ERI News - your reliability newsletter
February 2003 - volume 10

Hello. Welcome to ERI's 10th newsletter. We have two articles, this issue, for our readers. In the first, Dave Douthit shares his concerns about long term reliability of electronics. Dare we wait for field failure reports, before taking action to correct problems?

In the second article, I continue my discussion of ED (electrodynamic) shakers, begun in the last (November 2002) issue.

You should also check on "What is pseudo random vibration?", a recent question from one of our readers.

Just in: I was told to visit http://www.usynaptics.com. I urge everyone, but particularly reliability people, to visit that site, especially if you ever fly in an older airplane.

Best wishes,
Wayne Tustin

Imminent Crisis
COTS parts in service may wear out in three to seven years
by David Douthit

Lloyd Condra from Boeing and Joe Chapman, representing the Program Standardization Office of the DoD, are currently traveling around the USA issuing this warning. It applies to the new nanometer size technology for ICs. This is the first time in the history of the semiconductor/microcircuit industry that such a condition has been admitted publicly. Up to now solid state devices were claimed to not have any wear out modes.

The commercial electronics industry has ceased to produce what is needed for high reliability/harsh environments. Even though this segment only represents about 1% of the component market, it is involved in 100% of essential systems such as traffic control (air and ground), national security, telecommunications, banking, airborne, manufacturing, medical, etc. This "new" failure mode will have a major impact!

The Challenge
"This challenge is as grave as any since the beginning of the solid-state revolution 50 years ago," Condra said in his presentation as scheduled keynote speaker at this year's Military & Aerospace Conference West (with COTSCON), Dec. 10-11 in San Diego. "It must be solved strategically, not tactically." Let's take a strategic look at the issues surrounding COTS. Up until this "wear out" failure mode appeared, electronic components accounted for < 10% of documented high reliability failures (1). In fact generally the largest percentage (sometimes in excess of 50%) of anomalies (an event that does not meet the designed performance requirements) is classified as NFF (No-Fault/Trouble-Found). There is only one guaranteed result from NFF; It will happen again!! It will continue to occur until the problem is solved. Unfortunately, NFF problems won't be solved because the proper test equipment and testing protocols do not exist.

"Why not?" you may ask. Reliability/qualification testing done during design and manufacturing has been based on MIL Specs and Standards. The DoD was concerned with parts surviving "overstress" situations (temperature, voltages, vibration and humidity, etc.) and not with life cycle issues. The prevailing thinking was that there was no "wear out" mode for parts, so no "aging" tests were needed. This led to test equipment and test methods which were designed to only detect "youth" failures. Test equipment designed for field and depot level repair uses the same philosophy and methods. ESS tests were developed to detect and eliminate weak designs and or components. These accelerated stress levels have been in use for a long time. The documented results indicate that <10% failures were caused by faulty components (1). This would seem to indicate the program was fairly successful. But now let's look at the overall picture. First, the efficiency/durability of systems has been steadily declining (1). Second, hardware and software are becoming increasingly complex and integrated (2). Third, these systems are being used in increasingly harsh and difficult situations (1). Fourth, in the past few years the documented failure rate for COTS components has begun to trend upwards (see below).

The environmental stress capabilities demanded by (and created for) today's small-volume high reliability applications are not needed for large-volume commercial products. Commercial product vendors have long since quit building components and assemblies that would pass these stress tests. This left the high reliability equipment suppliers with a serious problem. Suppliers have attempted to do a "work around" by uprating components though third party (such as the University of Maryland CALCE) stress testing. This "uprating" is based entirely on temperature limits. Unfortunately, humidity, contamination, vibration and ESD can also damage components and assemblies. Legal liability issues have become complicated; component manufacturers do not want to be blamed if their components cause costly failures when operated outside of their factory specified limits. Attempts are being made to "harden" or improve the temperature range (ignoring vibration, for example) of these components. These efforts still do not address the majority of system failures. Limiting resources to one minor issue (temperature) wastes time and money.

COTS components go out of production very quickly (generally under 3 years). This rapid turnover also includes passive components, materials, design tools, test equipment, plus manufacturing equipment and processes. This is the source of many problems for high reliability systems. Failure modes such as leakage currents, cross talk, dendrites, CAF (Conductive Anodic Filaments), delamination and solder joint cracks are now common causes of intermittent failures. Field- and depot-level test equipment is incapable of identifying many of these failure modes. Here is the cause of many No-Fault-Found problems. The combination of so many variables can create new failure modes and cause failure rates to vary wildly. We cannot wait for field failure reports to evaluate a design, because to takes years for a high reliability system to go into production and even longer to reach the field. Even the testing of prototype systems under field conditions requires a great deal of time. Without accurate environmental testing based on expected end-use conditions, the idea that we can build highly reliable systems is laughable.

Present designs are losing durability and robustness as the commercial industry quits following DoD requirements and moves towards less durable but more profitable designs. Industry makes money by selling products and services. High reliability designs are less profitable.

Performance-Based Specifications
Currently used testing methods, established under MIL Standards and Specification process, were not designed to determine life cycle capabilities but rather to find infant mortality/warranty failures. They were based primarily on accelerated stresses to stimulate failure modes based on process, materials or design weaknesses. Many, if not most, of the MIL Standards relating to reliability, "how to" build, certify, and qualify components and assemblies, have been canceled. Civilian leaders of the MIL services hoped that Performance-Based-Specification/Standards implemented through contractual arrangements would assure long-term reliability. The main feature of these contractual arrangements is the requirement to maintain various levels of reliability, durability, dependability and maintainability for specified lengths of time. The lack of test equipment and testing protocols capable of proving that the requirements of these contracts have been achieved, means there is no way to enforce "Performance-Based Specifications/Standards"(2). These programs require predictability of the life cycle for systems. The data needed concerning end use conditions (to establish base line testing methods) is incomplete if available at all. It is not possible to determine the "life cycle" of a system without this information!

"Wear out" - a new failure mode
The "wear out" issue may tempt vendors and OEMs to abandon even their present attempts at reliability testing. Present-day ESS may be abandoned or greatly reduced. Why? Because some of these components will "wear out" before other stresses can cause problems.

This "wear out" mode involves metal from traces migrating across or diffusing into the silicon substrate. This issue has been known for years. Today's and tomorrow's reduced geometry, increased speed and densities of IC designs in the nanometer range have moved this failure mode to the front (3)(4). This is a thermodynamic-based failure mode. Even the elevated temperatures used in ESS can shorten the life of components! The possible result is that assemblies/systems are barely turned on to see if they function and then are immediately shipped.

These systems will need regularly timed replacements with newly designed hardware (and possibly software) because the original components are obsolete. Extra components (spares) will have possible storage issues, even at room temperature. This is dangerous for the military, NASA, the FAA, and for any other long term high reliability systems users who hope for 20-year life, but profitable for their suppliers.

Conclusion
All this is based on what is presently known about this "wear out" mode and on current methods of dealing with reliability issues. Remember, as many as 50% of anomalies are No-Fault-Found and <10% failures are attributed to component failures. The reduction in size, the increase in speeds, the lowering of signal strength, the change of materials and new processes have created this new "wear out" mode. These changes are occurring in other segments of the electronics manufacturing industry. It is leading to other new failure modes and to increased failure rates.

The "new" failure modes are based on stress factors other than temperature. There are four environmental stress factors that limit the life cycle of assemblies and systems. They are temperature, vibration, humidity, and contamination. They're usually combined in complicated synergies that are not understood because it has not been required! This ignorance can no longer be tolerated! Without accurate testing done early in the design process, there is no hope of producing a product which will meet the requirements for high reliability operations. Without accurate information about the conditions expected in the end use environment, proper tests cannot be done.

Even if someone succeeds in developing non "wear out " mode components, 90% of the failures and a large percentage of No-Fault-Founds will still exist in MIL warehouses. Sliding back in reliability is beginning to have serious effects on essential mission-critical equipment . Reliability is supposed to increase as knowledge is gained. This cannot happen when there is a lack of proper test equipment, inability to accumulate needed failure data, and unwillingness to share information with all parties involved. Costs for these inadequacies are beginning to mount. The longer we wait the higher the price will become! 2003 will see these issues become a crisis. Running critical/essential systems of the world on untested and unqualified video game type hardware and software is a recipe for disaster.

References

(1) Performance-Based Quality Assurance of Electronic Hardware
Per-Erik Tegehall - May 2002 http://www.ivf.se/elektronik/vi_report_reliability/default.htm

(2) DOT/FAA/AR-01/41
Review of Pending Guidance and Industry Findings on Commercial Off-The-Shelf (COTS) Electronics in Airborne Systems Office of Aviation Research Washington, DC 20591, August 2001.
Final Report: http://av-info.faa.gov/software/Research/01-41.pdf

(3) High Resolution Resistance Measurements Applied to Electromigration Researchers: A. Scorzoni, I. De Munari, M. Impronta, R. Balboni, N.Kelaidis Technician: A. Sardo © 1996-1997-1998 http://www.diei.unipg.it/RICERCA/www_em.htm

(4) Cu Technology
HYNIX Semiconductor
Kyeongkeun.choi@hynix.com
May 2002
http://www.postech.ac.kr/bk21/ece/Kor/Achieve/news/hynix/CU1.pdf

Dave can be reached at douthit@equipment-reliability.com. He is available in both a teaching and a consulting capacity, to discuss these points further. Dave will teach "Contaminants and Moisture can Disrupt your Electronics" coming up Feb. 24-26, in Santa Barbara, California.

(back to the top)

EH, ED or RS? (part 2)
by Wayne Tustin

In the November 2002 newsletter I described EH or electrohydraulic (sometimes called servohydraulic) shakers. Now let's discuss ED or electrodynamic shakers, widely used for sine and random vibration (and for some shock) environmental testing.

Electrodynamic (sometimes called electromagnetic) shakers resemble electrodynamic loudspeakers in principle. A moving coil (driver coil) carrying alternating current (ac) is positioned in a strong radial magnetic field. The result: an alternating vibratory force whose waveform closely matches the current waveform (e.g. sine or random or shock) and whose force magnitude increases with current magnitude.

In small shakers (as with most loudspeakers) a permanent magnet provides the magnetic field, radially directed outward across an air gap. In medium to large shakers, direct current (dc) flows through a coil that generates the magnetic field.

Small shakers are convection cooled. Intermediate shakers are forced-air cooled. The windings of some high-force shakers are liquid cooled.

Table 1 lists some of the automotive uses to which ED shakers are being put. ED shakers are also widely used in aerospace as well as in HALT (highly accelerated life testing during development, as well as post-production ESS (environmental stress screening) and HASS (highly accelerated stress screening) of electronics.

Figure 1 - Cutaway view of electrodynamic shaker

Current designs (as in Figure 1) are "double ended" with half the field windings above and half below the magnetic gap and ac driver coil. Gap flux and dynamic force per driver current ampere are much higher than with early "single-ended" designs. This does, however, lengthen the armature (Figure 2) and somewhat lowers the frequency of axial resonance. Note the base and trunnions, which permit the shaker to be rotated for horizontal forcing.

Figure 2 - Armature construction

Some form of flexures (utilizing rollers or sliding or flexing members) must support armature and load weight and guide the armature along a (hopefully) straight-line path. Any of the shaker manufacturers (see this link for a listing) will provide detailed views showing their construction methods.

In Figure 3, motion (1) is what we want. Motion (2), greatly exaggerated here, of course, can result from unequal dynamic stiffness of the supporting flexures (should be replaced in a matched set) or from unbalanced loading.

Figure 3 - Wanted and unwanted shaker motions

How would you instrument a shaker table to determine if any rocking (2) accompanies straight-line motion (1)?

At least one ED shaker manufacturer equips some shakers with axial hydrostatic bearings for supplemental restraint against table rocking.

The greatly-exaggerated motion at Figure 4 "A" is called table diaphragming or oil-canning. Acceleration at table center might be 50g with 1g at the edge. Now move higher in frequency to anti-resonance. The greatly-exaggerated motion at "B" finds table center isolated at say 1g while the table edge experiences 50g.

Figure 4 - Additional unwanted shaker motions

These motions may not seem possible. Have you seen a welded or cast shaker table from beneath? It can be difficult to visualize such a weldment or casting flexing. The table is almost as rigid as a solid cylinder, but please recall that even solid cylinders can flex.

"A" and "B" definitely compromise vibration testing. Resonant elongation of the armature (not shown here), does not compromise the test, but it shortens the armature's life. Ask your shaker supplier at what frequencies "A", "B" and elongation occur, unloaded or "bare table". The frequencies will drop when loaded.

Unwanted motions "A" and "B" in Figure 4 cannot be avoided. Each will occur at some frequency. But at what frequency? With large table diameters approaching 30 inches or 1 metre, they occur well below 2,000 Hz.

Fortunately, physically large test items need not all be tested to 2,000 Hz. "But my spec requires 2,000 Hz testing" you may be saying. Please - argue with your spec writer. Force him to show you data from the field. Many times you will discover that "someone" added an octave or two.

If your hardware is relatively small, select a shaker with a smaller armature.

Spare Armature
Many laboratories have had to delay tests because their shaker was "down", waiting for armature repairs, usually for a new driver coil to be wound and installed. This can take a week or more. Experienced labs often buy a spare armature assembly. That assembly sits on a shelf in the lab, ready for installation when needed.

In the late 'fifties, a large shaker model that had been sold and successfully used for testing to 500 Hz (in accordance with standards of that era) was re-rated to 2,000 Hz (to appear to satisfy new specs) by the maker's sales department. (The shaker design was not changed.) Major table resonances under 2,000 Hz were apparent to investigators.

One user sprinkled white sand on his shaker table, then slowly advanced the frequency of sinusoidal vibration. At numerous frequencies at which the sand collected in recognizable patterns, he took pictures such as the three in Figure 5.

Figure 5 Sand collects at nodes

Here are his results. Motion "A" (see Figure 4) occurred around 1300 Hz and threw the sand onto the laboratory floor. Motion "B" occurred around 1500 Hz. A "ring mode" occurred at 1800 Hz; probably table center moved up when table edge moved down, and vice versa. The 400 Hz motion has not been explained, but table motion was certainly not uniform!

Possibly you will not be able to use "white sand" for the experiment just described. Yet you might wish to measure diaphragming severity. Quiz: How would you instrument your shaker table?

One user decided to stiffen his shaker's 2 foot diameter armature. He attached a 6 inch thick 2 foot diameter aluminum plate, using a 3/4 inch bolt in each attachment hole of the shaker table. Did that raise the resonant frequencies? No. They actually dropped. His "added-mass" effect was greater than his "added-stiffness" effect.

Think about this when you order a new test fixture. If experienced shaker designers have problems with resonances, you can be sure that new fixture designers will also have resonance problems.

Isolating the shaker
Armature and load acceleration result from force upward; this is accompanied by an equal and opposite force downward, into the test lab floor. And vice versa. At certain shaking frequencies, shaker forcing frequency will equal a building's natural frequency . This may elicit complaints from other departments. And it may structurally damage the building. You may decide to "float" your shaker on air bag isolators.

That helps but may not be sufficient. The next step is to secure the shaker base to a concrete mass (10x or 100x the shaker force rating) that in turn is isolated from the building. That mass usually floats in an excavated pit, so that shaker table height is convenient, possibly at floor level for large test items.

Cooling of shaker
The smallest shakers, using permanent magnets for exciting the field, can dissipate heat into room air by convection. However, with larger ED shakers, the I²R losses generated in the field winding and in the armature require forced heat removal. Some shakers utilize forced air (blower preferably located outside the lab) cooling. Fluid (oil or water) circulation is more effective and makes it easier for the operator to audibly detect test article resonances.

Multi-axis vibration testing and screening
Most vibration testing and most vibration stress screening are single axis-at-a-time. That is traditional, but certainly is not realistic. It necessitates moving the DUT (device under test) with its X, then its Y and finally its Z axis matching the shaker axis. "Real world" vibrations (with very few exceptions) are multi-axis. "Real world" vibrations in the orthogonal axes X, Y and Z are usually different from each other (uncorrelated)*. Sometimes these are identified by such terms as fore-and-aft, lateral or sideways, and vertical. "Real world" vibrations in the rotational axes a, b and g are also usually different from each other (uncorrelated). (The latter are often identified as roll, pitch and yaw.)

*For this reason I object when fixtures resembling Figure 6 are identified as multi-axis. Every few years (the first was about 1954) some test engineer gets this "bright idea": Let's build a fixture which tilts our product so we can shake our product in three axes at the same time. I'm told that a patent has been awarded on such an idea. The patent may or may not be enforceable, but in my opinion the resulting test or screen is not multi-axis.

True, there will be a component of shaker motion (which I assume is vertical in Figure 6) in the product's X, Y and Z axes, but those components are highly correlated. To get uncorrelated (as in the "real world") multi-axis vibration requires several shakers.

Figure 6 "Tilt" (not multi-axis) fixture

Not only X, Y and Z uncorrelated translations but also uncorrelated rotations a, b and g (or roll, pitch and yaw) are also needed. Shakers must operate as pairs to accomplish this, and their thrust axes must pass some distance from the product's cg (center of gravity).

If that paragraph brought the word "expensive" to your mind, I would agree. A three ED system, for example, requires not only a North-South shaker + an East-West shaker + another beneath (thrusting upward), but also three power amplifiers and three control channels. Exemplary systems are located at the Army Research Station, Adelphi, Maryland and at Hill AFB in Utah.

How to get multiple-axis excitation at relatively low cost? By using inexpensive pneumatic vibrators, long used for moving bulk materials. Orient them to give not only uncorrelated linear X, Y and Z motions but also uncorrelated rotations a, b and g. See my final article in this series.

Acoustic excitation
CRIQ in Montreal uses loudspeakers to multi-axis excite bending resonances in printed wiring boards (PWBs). This approach shares with EH and ED shakers the ability to randomly excite the product's known resonances. By contrast, RS machines (that are described next) "waste" much of their excitation energy in frequency regions where there is no modal response.

This article will continue in the next issue. In May 2003, Wayne will conclude with RS systems. Wayne Tustin, ERI's president, can be reached by e-mail or phone (805) 564-1260. Read more about Wayne at ERI's website.

(back to the top)

Questions our readers have asked...

This section of our newsletter was created for you, reader! Feel free to send questions or suggestions to the webmaster. One of our specialists will respond.

Q: What is pseudo random vibration?

A: To understand my answer to this February 2003 question, you need to understand my May 2002 answer to the question "What is random vibration?" Did you understand that "random" basically means "unpredictable"? I assumed that you were familiar with sine and with complex vibrations. When you view these (using an accelerometer for converting motion to a voltage signal) on your oscilloscope, the 'scope sweeps over the same sample again and again, giving your 'scope a stationary, repeated, pattern. Right? Not so with "real" random. Every sweep will be different.

Pseudo random is almost the same, except that the pattern repeats occasionally (though not often). Another term is "quasi-random," meaning "somewhat random". Some claim they can detect a pattern to the sound, which is not surprising, since there is some periodicity. Whereas "real random" or "white random" has a flat spectrum, pseudo random has spectral peaks. "Zoom in" on your spectral display to see them. You may have to borrow an analyzer with "zoom" capability.

When is pseudo random used? (1) That's what you get with pneumatic RS (repetitive shock) hammers. (2) Some random vibration controllers for ED (electrodynamic) shakers utilize somewhat repetitive pseudo random during startup.

Wayne Tustin, ERI's president, can be reached by e-mail or phone (805) 564-1260. Read more about Wayne at ERI's website.

(back to the top)

Were you online with us October 15 and December 3, 2002 and January 21, 2003, for Wayne's three free "live" web-trainings? If by chance you missed them, the illustrations and the "script" are posted at the articles section of our websites. Wayne's three subjects were "What is Resonance all about?", "Measurement and Analysis" and the most recent "Vibration Aspects of Reliability Enhancement via HALT, ESS and HASS." You are welcome to download them for your own use, but please ask before you use them commercially. Thanks to Bruel & Kjaer for donating Webex time and to the Chicago Chapter of the IEST for arranging this event.

February 11th is getting very close. Wayne will be teaching "Fundamentals of Vibration and Shock Testing and Measurement" on Feb. 11-13.. The following week, February 17-19, John Starr will present "Optimizing HALT, ESS and HASS of Electronic Circuit Cards". And the next week, February 24-26, Dave Douthit will present "Contaminants and Moisture can Disrupt your Electronics". Please click on the course links for details and registration. You are welcome to attend your choice. Or all three! Get your registrations in quickly, as the calendar now says February.

ERI is pleased to announce a new course, "Thermal and Random Vibration Stressing for HALT, ESS, HASS and COTS Testing", which meets March 18-20 at DTB. DTB? That's a well-known test lab at Bohemia, Long Island, New York, Dayton T. Brown.

DTB last year invited ERI's Steve Brenner to this year present his short course on "Climatic Environmental Testing." But our plans were changed by an event at Redstone Arsenal in Alabama, where (on an over subscribed and soon-to-be-repeated Army contract) Steve presented "Military Standard 810F Climatics Test Interpretation". His audience was a group of engineers who work closely with Army contractors. During the first hour, these engineers asked that Steve speak about HALT, ESS and HASS. This surprised Steve, because 810F does not mention these subjects. Fortunately, Steve was carrying a number of illustrations concerning these subjects, and he was able to modify his presentation to accommodate participant demand. Many students thanked and praised him for adjusting the contracted course to meet their needs. Now … "fast forward" to the March 18-20 event at DTB and don't miss Steve's course! Click here for registration.

PCB Design Conference
The PCB Design Conference West and HDI Expo will meet on March 10 - 14, 2003 at the San Jose Convention Center. Visit their website to register and receive the special early bird pricing! Wayne will do his presentation on Tuesday March 11, 2:30 - 5pm.

Reliability Symposium
The CRMS Reliability Symposium will meet October 2003, in Ottawa, Canada. Click here to visit their website and get more information.

ESTECH 2003
Plan now for ESTECH 2003, the IEST annual technical meeting and exposition, May 18-21, 2003, Hyatt Regency Phoenix, Arizona. Wayne will present a half day tutorial on shock testing and David Douthit will be teaching a half day tutorial on "Electronic systems, reliability and Harsh environments.

Steve Goldman
We would like to let you know that Steve Goldman passed away at the age of 58 at the end of 2002. He was well known in the machinery analysis field. Steve wrote two books and many papers on the subject and conducted lectures throughout the country..

If you have facts or opinions relating to the following incomplete statements relating to nondestructive testing, please respond on our Message Board and also via e-mail to me.

Are you an NDT (nondestructive testing) specialist? Or can you pass my request to an NDT specialist whom you know? My question: in what ways do NDT specialists use vibration testing and/or measurement? (I'm writing an article for an NDT publication, and I need help.) Buildings can sway at frequencies < 1 Hz but floor and wall vibrations, which can be troublesome in high-magnification X-raying, also in high-magnification video work, in chip manufacture and in robotic surgery, might range perhaps ___ to ____ Hz. Vibration "signatures", much used in preventive maintenance of vital machinery, range perhaps ___ to ___ Hz. Are NDT specialists ever concerned with bridges and other structures, where vibrations range perhaps ___ to ___ Hz? How about seismic disturbances (earthquakes) which range perhaps 1 to 35 Hz? Or with automotive vibrations, which range perhaps 10 to 500 Hz? Or with aircraft and space vehicle vibrations, which range perhaps 10 to 2,000 Hz? Have I missed any NDT activities involving vibration testing or measurement?

Wayne

ERI - Equipment Reliability Institute
1520 Santa Rosa Ave.
Santa Barbara - CA - 93109
Tel: (805) 564-1260
Our fax number:
(805) 966-7875

Wayne Tustin tustin@equipment-
reliability.com

Webmaster webmaster@equipment
- reliability.com

Web sites
http://www.equipment-
reliability.com

http://vibrationand
shock.com

Copyright © 2000-2003 Equipment Reliability Institute. All rights reserved.

Subscribe
If you would like to subscribe to ERI News, go to either website, fill in the form "Free Newsletter" and hit the Submit button. Subscribe now!

Recommend
If you enjoy reading ERI News and want to recommend it to a friend, just hit "forward" on the menu of your e-mail program or tell your friend to subscribe to it at our website.

Previous issues
Missed the previous issues? It is not a problem. Just visit our newsletter archives section and find all ERI's News issues.

Unsubscribe
If you do not want to receive ERI's quarterly newsletter, please send us an e-mail using the same e-mail address that brought you the newsletter, with "remove" as subject.