what components fail as shorts (vs open)?

[buschfsu]

ive had two devices recently repaired and both were shorts to ground.  i cant think of a component that fails by shorting.  seems like everything i can think of, resistors, caps, diodes, etc would fail open?  this failure mode seems to be a big reason for fuses so i must be missing something...

 

[kvant]

> ...i cant think of a component that fails by shorting...

For instance, tantalum capacitors are said to fail as shorts.

 

[Samuel Groner]

www.groupdiy.com/index.php?topic=15625.0

Samuel

 

[PRR]

They always fail in the direction that does the most damage... -- dale116dot7

An overheated semiconductor diode typically fails short.

We find them "open" when they are in a circuit which has enough POWER to burn-up the short.

Signal transistors may be found short or open. Power supply rectifiers are typically found open.

Film capacitors often "fail gradually", rising leakage; but can fail catastrophically if the paper/plastic insulation has a hole, or gets a hole. The area where a cap can gain a short is much larger than the two points where it could fail open.

Electrolytics sometimes dry-up and go open. Happened in my H-P 200AB. Sometimes they turn sour and leak; this is fairly popular (until they let-loose stinky smoke). Occasionally Al elcos will lose a large area of their tarnish and go very-low ohms, near-short.

Vacuum tubes fail many ways. The last one I saw appeared to be a warped G2 wire which, when hot, touched a support rod and put a dead-short on the rest of the amp. This in turn burned transformer insulation, probably causing a shorted turn. This was left powered-up a while, and the winding finally burned open.

The failed tube before that had a crack in the heater, which was "OK" when cold and opened-up when hot. If it were a glass tube, I might have noticed; this was a metal tube in a chassis so hot that I could not be sure this one tube was "cool".

Incandescent lamps ultimately fail open. But we have all seen one go dark, smack it, it comes on a little bright when the filament-end happens to weld itself to a support. In extreme cases only a little length of the filament winds up across the line and BWOOM! the house fuse blows due to near-short.

Fuses protecting transistors always fail by staying "short" just a little longer than the abused transistor. OK, rarely they will fail open when there was no danger at all.

Joints, connections, and jacks will fail short or open at whim. For about every no-joint I have made, I have made a bridge. A stepped-on cable can be open in some lines and cross-connected on other lines, or both. Even if you use a drill: when I drilled a PBX cable hidden behind a wall, I got 48 opens and 7 shorts where the bit had smooshed two wires together.

 

[buschfsu]

great info guys thanks.  what about transformers?  specifically if one is cooking due to too much current draw?

 

[JohnRoberts]

I have seen lots of both, and my generalities may vary from another's personal experience, so take this with a grain of silicon.

Resistors- typically fail open from overheating.

Capacitors- open from drying out, short from over voltage or mechanical flaws.

Diodes-Typically fail shorted from overheating, unless they vaporize open again.

Transistors (power)- typically fail as short but can open bonding wires. I've seen plenty of old power devices with open bases, and shorted collector-emitter. Shorts are caused by over voltage punch through, and over heating.  MOSFETS I've seen mostly similar to bipolar but they are a little harder to overheat.

ICs- varies somewhat with internal protection schemes in modern ICs but common failure in older IC's was dead short between rails (nasty). Now some input pins are effectively fused so can fail open- or just not working right. Some ICs will still sort of work while partially broken.

Transformers- can fail open or (almost) shorted. Fine wires can fuse open, or insulated breakdown can cause shorted turn inside windings. This will not measure as a dead short depending on where they occur but transformer will be very unhappy with shorted turn when powered up.

Inductors- similar to transformers

FWIW diodes fail as a short circuit reliably enough that I once used diodes in a safety ground path, to provide an open to low voltages, but still a safety path for larger voltages.  While this worked the better practice is to hard bond safety grounds and use differential circuitry to ignore ground potential differences.

JR

 

[bcarso]

My favorite counterintuitive failure was also a dangerous situation, a potential fire in a crowded auditorium, an old orchestra hall in Chicago back in 1972.  I was in the band backing up a Japanese singer, the redoubtable Miss Shimkura Chioko* IIRC.

The guitarist had not been warned against plugging his Fender amp into the convenient outlets in the pit.  The amp didn't come on, and in a short period of time there was the all-too-familiar smell of burning transformer.  Yikes!  And people were filing into the hall.  One did not want to cry Fire.

Turned out the outlets were 115V...DC.  They were for the stand lights, where if you get enough of the typical tubular long-filament ones together, make quite an appreciable acoustical 120Hz hum/buzz.  So they used d.c.

What about the line fuse in the Fender though?  Well, it certainly had well over rated current.  But unfortunately it conducted with a metal vapor arc as the insides of the envelope got nicely plated.  The fuse was not rated to interrupt d.c. of that voltage, nor in most situations would one imagine such to be necessary.


*Whose rendition of Laindlops Keep Farring On My Head was not to be missed.

 

[buschfsu]

Ahahahah nice

 

[PRR]

> it conducted with a metal vapor arc as the insides of the envelope got nicely plated.  The fuse was not rated to interrupt d.c. of that voltage

Fascinating. You sometimes find a standard fuse in the DC B+ rail of a guitar amp, "protecting" the OT and power tubes.

That's 450V-600V across a "250V" fuse, which always makes me wonder.

I had not thought about the steady arc, although I know the difference in power tools. These use universal AC/DC motors, but the switches are AC-rated and won't break DC more thna a few times.

> a Japanese singer, the redoubtable Miss Shimkura Chioko

Also spelled: Shimakura Chiyoko

That's before your time. She's still performing.

Shimakura won a contest as a teenager in 1954. In the 1960s she was doing a Supremes-like shtick:
http://video.aol.com/video-detail/-196812/2761882985
http://video.aol.com/video-detail/-1966/2305843013332362773
The bossa-nova works with the semi-enka.

It's a bit hard to think this act would play in the Swinging Seventies:
http://video.aol.com/video-detail/shimakura-chiyoko-sumida-gawa/3260195041
But Harumi mined the same vein.

Here's a racy set with Fanfan (Okada Masumi, another long-runner with an amazing career, from Crazed Fruit to IZO and Iron Chef):
http://video.aol.com/video-detail/-w-/1212467579

She did age, but mighty slow and gracefully:
http://video.aol.com/video-detail/-tokyo-no-hito-sayonara-shimakura-chiyoko/1708321915
http://video.aol.com/video-detail/shimakura-chiyoko-kazeninaru/3940197294

I was thinking "comfort food", also Guy Lombardo. Indeed Shimakura was featured on the NHK New Year Show 35 times

 

[Andy Peters]

What about the line fuse in the Fender though?  Well, it certainly had well over rated current.  But unfortunately it conducted with a metal vapor arc as the insides of the envelope got nicely plated.  The fuse was not rated to interrupt d.c. of that voltage, nor in most situations would one imagine such to be necessary.

I love situations where the equipment commits suicide to protect the fuse.

-a

 

[bcarso]

I knew my spelling looked wrong for her name, but was too lazy to search.  She was/is a bigger (and older) star than I had thought!

And I exaggerate her L-R difficulties---just makes a better gag.  Another tune she did in Japanese was No Honki Kashira, as I recall something of a tango, which had amusing possibilities for interpretation, but in fact translated as Are You Serious?

The tour was grueling, seven cities in ten days, but had some upsides.  Heard the great trombonist Urbie Green in Chicago; sat in with Elvin Jones at the Village Vanguard, the entree (which brutally violated the Noo Yawk dues-paying protocols for jazzoids) associated with his Japanese wife.  Unfortunately his group at the time was pianoless and doing all originals.  Challenging to say the least when I mistakenly thought the tune called God Bless the Children was in fact God Bless the Child.  I did a fairly unrelated and meandering improvisation until Dave Liebman mercifully cut in.  The crowd was kind nonetheless.