eHam.net - Amateur Radio (Ham Radio) Community

Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net



[Articles Home]  [Add Article]  

Sherlock Investigates Radio Failures

from Paul Signorelli, W0RW on April 14, 2017
View comments about this article!

Sherlock Investigates Radio Failures
By W0RW

If you don’t ask yourself, “why do things fail?” You are missing the most exciting part of your electronic hobby.

When you have a piece of equipment that fails, do you just throw it away?

Analysis of the failed item can be the most interesting part of your electronic experience.

It can lead to new designs or modifications to improve your equipment.

Most of our current electronic equipment is designed to be replaced at the Subassembly or Printed Wiring Board Assembly level, but that doesn’t prevent you from trying to find the root cause of a failure at the piece part level.

Sometimes your investigation will lead to a simpler repair action that will save the subassembly or unit.

Design and construction of Electronic Components can be studied on Wikipedia at:

http://en.wikipedia.org/wiki/Electronic_component.

It has a large data base on how these components work.

Sherlock’s Science of Deduction:

It is keen to apply the precepts of advanced detective procedures in these matters:

•“It is a capital mistake to theorize before one has data. Insensibly one begins to twist facts to suit theories, instead of theories to suit facts.” (A Scandal in Bohemia.)

•“When you have eliminated the impossible, whatever remains, however improbable, must be the truth.” S.H. (The Sign of Four, Ch. 6.)

•"The world is full of obvious things which nobody by any chance ever observes," S.H.

Statement of the Cases:

I have left out all the flummery and verbose descriptions that Dr. Watson would have added if he were recording these stories, so you will have imagine your own settings, However some of these stories did begin “On a Dark and Stormy Night.”

“Many of these singular happenings have never before until now been fully dealt with in any public print,” S.H.

The Episode of the ‘Black Widow’ Transistor:

One of the most devious failures ever encountered involved the ‘Black Widow’ transistor in a push pull switching power supply. Female Black Widows have been known to eat their mates. The male spiders still have not figured this out yet.

This failure occurred on a TO39 power transistor (2N3553) that had a big copper slug inside the case that was crimped in place. The case had three dimples on the side of the case that grabbed the internal copper header.

In this singular case one transistor in the circuit burned up and shorted, it looked like secondary avalanche breakdown, collector to emitter short. There was significant melting of the die. The transistor was replaced and after a few temperature cycles the newly installed transistor failed again in the same spot. Ahaa! - The spider got another one.

The root cause of the failure was traced to the ‘good’ transistor in the push pull pair. It was opening up and latching the other one ‘on’. The crimp on the header can was the only electrical connection to the collector and when it opened up the other transistor was turned ‘on’ with full DC current that just melted it. So remember in any push pull circuit failure the ‘good’ part might really be the ‘bad’ part, however improbable it may seem.

The Midnight Flight of Denny Dendrite:

It was a dark and stormy night, really, when the intrepid Denny Dendrite electronic sleuth made his way carrying a dendrite to a secret laboratory in Washington D.C. for analysis.

Arriving at 3am, after a transcontinental flight in the winter, a dozen government investigators stood ready at their SEM and Auger equipment’s.

The defect had just been confirmed on an integrated circuit that had used leaded (Pb) glass sealing material. The surface of the sealing glass had provided a platform for pure lead dendrites to grow from one IC lead to another under high impedance electrification.

These are easily visible using a microscope. Indeed, many youtube videos show the growth progress. See https://www.youtube.com/watch?v=COMTYzqe_XE for an example.

These types of failures are frequently induced by taking a unit from a dry, cold environment into a warm humid environment causing the unit to go through the dew point. When that happens, water condenses out of the air and can cause dendrites.

There is usually enough carbon dioxide (CO2) in the air or acidic residue on a circuit board to start a dendrite growing. So do not apply power to a unit that has gone through or is going through the dew point until it is completely dry. Remove any memory battery back-up battery or it will cause new dendrites.

The use of ‘Freeze Mist’ for trouble shooting is highly discouraged because this causes frost to build up on the components and they all get wet as they come back through the dew point. Obviously it can cause dendrites that will mask the clue being investigated. Elementary isn’t it? If you must lower the temperature of a part, use an ice cube in a plastic bag.

The Intermittent Green Radio:

This radio performed perfectly for many years, but it was infrequently used. After field testing in hot weather, the unit would fail after a few hours of operation. The radio has many internal protection circuits so it just shut down and no internal damage was found. The radio would work fine the next day.

Trouble shooting began at the power supply circuit and using a heat gun, the unit could be made to fail quickly when it was hot. Switching to a soldering iron to gently heat the components, the failure was isolated to a Tantalum Capacitor. The capacitor was physically reversed in its circuit and after replacement there was no more intermittent high temperature failure. The part had simply been placed in the unit with reversed polarity. Tantalum capacitors typically can operate for many hours in the reversed direction, but will eventually short out.

The printed circuit board did not have any polarity markings on it and it escaped visual inspections. This expensive radio was repaired for the cost of a two dollar part.

Many articles about capacitors can be found on the Kemet.com Website which might require log in.

http://www.kemet.com/TechnicalArticlePage

Another similar failure occurred in a similar radio that was caused by an over-heating TO-220 Transistor. The transistor was mounted with small insulated plastic bushing to keep the mounting screw from shorting to the chassis. During assembly the bushing/washer became smashed under the transistor body (See Photo A) and this prevented the TO-220 transistor from fully seating on its heat sink. The transistor was not able to dissipate the heat on its case and that caused the radio power supply to shut down. Reseating the transistor flush onto the heat sink fixed the unit.

This episode was further detailed in my other monogram "Death on the Trail".

The 3.5% Solution:

It was a warm sunny day on the Island of St. Lucia when a rogue wave crashed over the rocks and drenched the HF operating site. The main radio was drenched and was inoperative. You could almost hear the dendrites growing. The famous (Old Goat) operator, rushed the radio to a clean water source to rinse it and flush the salt water.

Ocean salt water is about 3.5% saline. After drying the radio, fortunately it began to work after a few days. The normal lab procedure would be to wash it with distilled water, then an alcohol rinse followed by a vacuum bake. If a vacuum bake is not possible the radio could be dried in an oven with MIL-D-3464 desiccant or a bag of rice.

The Curse of the Tin Whiskers:

If you have a good microscope, with a magnification of at least 15X, you can find these tin whiskers in your old parts bin. Just look for an old USECO stand-off. These old insulating turret stand offs where tin plated and have over the years grown tin whiskers. You can see them growing off the top of the tin plated terminal. They are extremely small (Perhaps 1/20 the diameter of a human hair but up to ¼ inch long) they are straight and shiny. If you catch them in light reflection, the whiskers are easy to see. There are several Tin Whisker videos on You Tube:

http://tinyurl.com/o8whamx.

You can find them on Wikipedia: http://en.wikipedia.org/wiki/Tin_Whiskers and at https://nepp.nasa.gov/whisker/.

The whiskers can cause shorts and Multipactor discharges. Using parts that have a small percentage of lead mixed with tin eliminates this problem. It occurs on silver and zinc plated parts also.

The Scintillation of a Tantalum Capacitor:

Have you ever seen the ‘scintillation’ of a Tantalum capacitor?

You can be ‘Sherlokian’ yourself and find these defects and see what makes these things fail.

Wikipedia is a great place to learn about the construction, application and failure modes of components. Check out: http://en.wikipedia.org/wiki/Tantalum_capacitor for a briefing on tantalum capacitors.

You can see the capacitor scintillation on a transistor curve tracer. Just slowly raise the voltage up on the capacitor to three times over its rating, with about 1000 ohms in series and it will scintillate right before your eyes. The Manganese dioxide (MnO2) melts down and plugs the Tantalum (Ta) dielectric performing it’s well known ‘Self Healing’ effect. If you don’t have enough series resistance in your circuit this capacitor can permanently short or explode.

If you don’t have a curve tracer you can see pictures at:

http://www.kemet.com/Lists/TechnicalArticles/Attachments/90/2006%20CARTS%20Scintillation.pdf

Sometimes the capacitor manufacturer puts the wrong tantalum slug in the case or marks the case with the wrong rated voltage. You can strip the Manganese Dioxide (MnO2) off of a failed slug and tell by the color of the slug what the real tantalum oxide thickness is and what its real voltage rating should be.

One of the best ways to insure that your replacement tantalum capacitor doesn’t fail again, without redesigning the circuit to make it more reliable, is to ‘surge test’ the capacitor.

There are dozens of papers on this technique at the NASA Electronic Parts and Packaging (NEPP) Program Website:

https://nepp.nasa.gov/DocUploads/1701FE5C-6D72-4E40-A5A25FDC7F1F670B/Effect%20of%20Inductance%20NEPP%20report.pdf

The Electrostatic Discharge League:

There isn’t enough room in this monograph to cover the many ESD failure analysis techniques. A curve tracer is all you need to see MOS FET input/gate damage. See one example of an ESD strike in Photo B.

Go to the Electro Static Discharge Association web site for more training information.

https://www.esda.org/about-esd/esd-fundamentals/.

The main ESD Rules are:

Keep your wrist strap on (your body grounded) and your soldering iron grounded.

Remember “The contact and separation of unlike materials causes static”.

The Case of the Open Case:

Cold temperature failures of the venerable (1975-1980) JANTX2N2222A and 2N2907 transistors led to the discovery of this major manufacturing defect. One qualified manufacturer was making the TO-18 headers by filling the bottom with a glass sealing powder. You probably have some in your junk box. The collector lead connection to the header and thus to the transistor die was a pressure point contact (See Photo C) that was held fast by the contracting sealing glass during formation of the seal.

The reliable construction design of other manufacturers had a silver pellet on the tip of the collector wire (See Photo D) and it made a reliable metallurgical ‘nail head’ weld to the header during the glass seal operation. This problem was also found on ground leads on TO-5 cases also.

When part replacement was not an option, nickel wire was welded to the cases and soldered to ground.

The Cracked Solder Joint Mystery:

Solder joints are our friends but they can be defective if they are ‘cold solder’ joints or if they are mechanically pulled apart. Any part that is fixed/mounted on one side of a PCBA and has leads that go through and are soldered on the other side are candidates for solder fracture.

Long periods of temperature cycling will also cause solder fractures. Don’t leave your gear out in your unheated shack over the winter. These fractures cause open circuits and can be seen with a microscope. Many solder joint failures can be induced by subjecting the item to very cold temperatures for a long time and then applying circuit power while they are still cold.

Multiple Cases, Same Solution:

Many cases have the same defect at the root cause. This can provide many benefits to the electronic sleuth making detection and solution easier.

• Many TS-440 radios have been found with similar symptoms. They frequently loose VCO lock on 28 MHz and just display dots. This is caused by a materials defect in the VCO cans. The manufacturer used foam potting material to keep the components stable for mobile operation. This foam fooled a few NASA Space engineers too. The foam reverts into conductive sticky goop after 10 or 20 years in humid environments. The conductive foam shorts out all the high impedance circuits. The repair is to simply remove the goop and clean the VCO cans out. It is called ‘Reversion” but it is really ‘Degradation’.

• Many foreign engineers use polarized aluminum capacitors as audio coupling capacitors. Audio is AC and the negative part of the audio voltage can de-plate the capacitor dielectric causing audio distortion. The symptom is usually reported as “Transmitter works good but receiver has distorted audio”. I have purchased several radios with this same symptom for junk prices and by changing one capacitor, I now have a good radio. These radios usually come out of Packet systems where the squelch is left off and the audio runs at maximum around the clock. This design technique works for a radio that is always squelched and used infrequently, but used in full time service they may only last 10 years.

Elementary Equipment needed for these Analyses:

Only the most basic equipment is needed for these analyses, but it should include:

Multimeter, curve tracer, microscope or a small magnifying glass and a small hacksaw.

Danger:

• Some components are made of hazardous materials.
• Don’t grind up any Beryllium Oxide (BeO) which is white looking ceramic. This material is used on many power semiconductors.
• Some capacitors have carcinogenic or acid electrolyte.
• If it smells bad throw it out.
• Don’t mess with mercury relays or 866 mercury vapor tubes.
• Don’t dissect radioactive sources.

The Epilogue and Additional Clues:

• Wikipedia http://en.wikipedia.org/wiki/Electronic_component
• The Jim Williams Papers, EDN 1991.
• Trouble Shooting Analog Circuits, Robert A. Pease, National Semiconductor Corp, 1989.
• IEEE Reliability Physics Proceedings, Read all you can find.
• International Symposium for Testing and Failure Analysis (ISTFA) Proceedings, Read all you can find, 1970 to 2000.
• MIL-STD 1580A Destructive Physical Analysis
• The Adventures of Sherlock Holmes, Sir A. Conan Doyle

“I can’t give you a more detailed account of these cases without referring to my notes.

I do not know anything essential that has been left unexplained.” S. H.

Sherlock, W0RW

The game is afoot.

Member Comments:
Add A Comment
 
Sherlock Investigates Radio Failures Reply
by W0WCA on April 14, 2017 Mail this to a friend!
Well done Paul!
 
Sherlock Investigates Radio Failures Reply
by AA4LR on April 14, 2017 Mail this to a friend!
I personally experienced radio equipment passing through the dewpoint. I took my camping trailer out in the field and was set up for field day Friday night. I ran my A/C in the early evening, but then cut it off and opened all of the screened windows for the night.

Next morning, my virtually new Elecraft K2/100 wouldn't work -- kept giving an INFO 080 error, which means that the main microprocessor was having difficulty talking to the other microprocessors in the rig via the AUXBUS. Turns out, there was plenty of condensation in the rig from the high humidity mixed with the super-cooled radio.

By the early afternoon, it dried out enough to start working again. Didn't have a further problem, so long as the radio stayed on, and the internal heat kept the condensation away.
 
Sherlock Investigates Radio Failures Reply
by DL8OV on April 14, 2017 Mail this to a friend!
An addition to the list of links:

http://www.sherlock-holmes.org.uk/

Recommended for all fans of the great detective.

Peter DL8OV
 
Sherlock Investigates Radio Failures Reply
by K6UJ on April 15, 2017 Mail this to a friend!
Wow, thanks Paul.
Very interesting information.


Bob
K6UJ
 
Sherlock Investigates Radio Failures Reply
by K1DA on April 15, 2017 Mail this to a friend!
"Don't leave your gear in an unheated shack." Does that mean I have to take my mobile radios out of the car every winter night?
 
RE: Sherlock Investigates Radio Failures Reply
by W0RW on April 16, 2017 Mail this to a friend!
The idea is to avoid "unnecessary" environmental stress.
Vibration and temperature cycling are the worst environments but your radio was made for 'necessary' mobile operation.
I bring my mobile radio inside when we have those -20F winter times.

Sherlock
 
RE: Sherlock Investigates Radio Failures Reply
by K9MHZ on April 18, 2017 Mail this to a friend!
"An addition to the list of links:
http://www.sherlock-holmes.org.uk/
Recommended for all fans of the great detective.
Peter DL8OV"

Well, there's a value-added comment to this discussion. Good grief.

Anyway, to the OP....great topic.

 
Sherlock Investigates Radio Failures Reply
by DL1MEV on April 20, 2017 Mail this to a friend!
Nice article, Paul. I also prefer repairing over discarding when possible and with better spare-parts if necessary. In some cases the failure after the time of warranty seems to be built in, especially in lower priced consumer goods.
 
RE: Sherlock Investigates Radio Failures Reply
by AI4WC on April 20, 2017 Mail this to a friend!
Great, inspiring article. I always get a great deal of pleasure in repairing as it increases my understanding of my equipment and it's design. I guess I'm just a "tinkerer" at heart!
Thank you for helping me justify my efforts.
 
Sherlock Investigates Radio Failures Reply
by WB0RXL on April 21, 2017 Mail this to a friend!
When Jim Williams was teaching at MIT, he writes in an article entitled, "The Importance of Fixing" that he could pretty much order all of the lab equipment he wanted. There was only one budget item that was denied. Instrument repair and calibration was set to zero by department head Jerrold Zacharias. He scribbled across the request, "You fix everything." Mr. Williams tried his best to reverse the decision, but Mr. Zacharias held firm with "You fix everything." Mr. Williams said he learned more about electronics during that time of his career than at almost any other. It even became a game. He would have others purposely disable a piece of equipment so he could time himself fixing it. He wrote, "The inside of a broken, but well designed piece of test equipment is an extraordinarily effective classroom." Williams said that learning to "dance with the problems" helped make him a much better circuit designer.
 
Email Subscription
You are not subscribed to discussions on this article.

Subscribe!
My Subscriptions
Subscriptions Help

Other Humor Articles
Early April