Diode Failure Mechanism & Mode

With nickel leads, the thermistor range is extended to 300degC.

RL

Hello Robert,

True. You can't fully grab the middle lead with regular needle nose pliers because they'll mostly lock onto the outside lead. I took an older pair and ground down the tips except for the very end. That provides a gap so it really only grabs at the tip.

And then everything went SMT...

I think I've seen some similar items (bulk silicon thermistor?) rated for intermittent 1200F. Could be wrong. Target application was domestic self-cleaning ovens.

Best regards, Spehro Pefhany

Oh God! have mercy. Justin Case A. Shuttle on launching pad, don't tell me that some un-inspected and un-certified parts were included, and the system cannot be refuelled because a no-return valve was assembled into piping the wrong way, inspected and certified. After that, and only after that! NASA speck specified one way valves to have different diameter threads on "in" and "out" ends.

Justin Case B. Idiot-proof design in which a push button is positioned behind a metal revolving plate with a hole in "zero" position so the button can be pressed. The "newer, better, upgraded" idiot got a very thin long screwdriver and succeeded to press the button in HV position. QED.

Justin Case C. Old, very old, mobile logical circuitry for support of avionics, few racks of eqipment on vehicle and another vehicle sporting a 3-phase generator for powering it. The "smartie" responsible noticed that the main power plug has some screws wobbling, disassembled it, put everything back, tightened the screws and connected. Result, IC's exploding with such force that they created holes in adjacent prints. On inspection we found that the ground connection was inserted into one of the phases and one phase was grounded.

My favorite Justin Case D. Avionics inspection of military planes system, the security IS verything as some of the equipment includes various "nasties". So a _continuous_ wire is running between ALL parts of the equipment to let the safety part to be able to check existance of proper connection of all parts for powering the system. Safety refuses to connect, gives us about a week of hair tearing till we find on "desperate" "crazy" full disassembly, can you visualise this with equipment distributed behind panels in all parts of the "bird" till we found it. Properly produced and inspected, proofs in papers! cables were connected (Mil-Speck multicontacts) male to male and the connecting technician did not notice but worked with such force or impatience that the pins were bent everywhich way.

I rest my case of Justin Case and ask the court to call Mr.Murphy for his expert opinion.

Have fun

Stanislaw Slack user from Ulladulla.

Hmmm..have yet to crank up SMTs above 200C but suspect problems in the 250 to 300C region in the following order: 1) die attach solder melting - most likely an open circuit as the result; 2) failing that (meaning the solder is a liquid but the whisker wire is holding the die in place) then diffusion of the P and N dopants resulting in a short. That is, if you have a *furnace* or *kiln* to get up that high... Oh yeah; next to go is that the glass enclosure melts and then the die melts and then the metal melts and then the furnace liquefies and Pele has a good laugh...

Mister Ed???? Is that you typing instead?? Or is it a neigh-bor?

Good greif! And i would not be surprised to learn that those were *mild* cases.

I have seen something like this once, but it was a replacement cable for a customer, who later called back and said it would not work. I could hardly believe it when it was returned and I looked at it. The male pins were put in the female housing and vice-versa.

Of course, this cable was made by the same technician who once was struggling, for what seemed to be hours, to drill a few small holes in a steel panel. The drill was making an awful noise, and so was he. When I finally investigated I found he had the drill in reverse. I think he was still able to drill one or two holes.

I also once found, in a customer's facility, some drill bits that were sharpened backwards. I thought they were left-handed, but the spiral was right-handed. Maybe, in this case, two wrongs make a left?

Paul

Just samples of "Failure Mechanism & Mode" as presented by designs of experts in technology and Quality Control. "The best definition of infinity is human stupidity!"

Have fun

Stanislaw Slack user from Ulladulla.

Well, then, put it in an oven, do the experiment, and then tell _us_ what the results were!

And howcome you still haven't learned to bottom-post?

Good Luck! Rich

But, did the glass break before, or after, the diode failed? i.e., was the diode's failure the cause of the broken glass?

Thanks, Rich

I used a blue one too, but it belonged to my employer at the time. I do have a couple of WSU-30Ms, however. ;-)

Cheers! Rich

Interestingly, the four out of the first eight images that are the thing we're talking about show red ones!

Must be a political thing. ;-) ("Is your lead bender red, or blue?") (mine were always blue, or maybe aquamarine, as well.)

Cheers! Rich

"Sheep LIE!!!"

Cheers! Rich

The king sentenced some peasant to death for some triviality. The peasant, at his sentencing, said to the king, "Your majesty, if you postpone my death for one years' time, I will teach your horse to talk. Yours will be the only kingdom in the realm whose king's horse can talk! How about it?" So, the king says, "OK, one year." About six months later, he's in the stable, still babbling at the horse, and the poop-shoveler guy asks, "Why did you make such a stupid claim?" And the peasant says, "Well, I got a year of life. The king could die, there could be a revolution, I could die of natural causes, and who knows? The horse may talk."

Cheers! Rich

Makes me want to design a jig - The tips of my needle-nose are tapered, so it's hard to get a nice right-angle bend. But a sort of a cradle, that clamps the leads but lets you bend closer to the epoxy could be useful.

Cheers! Rich

I should have made one clarification that is applicable: Mil-HDBK-338 pertains to military class parts, and not commercial parts. So Joerg, and the rest of you, are all probably correct that the "other" percentage is much higher in commercial parts. IMHO mechanical failures are still the lowest probability mode of failure compared with the other modes.

I looked at other military-type standards for failure modes, but the rest of them showed "open" and "short" as the only 2 modes.

I doubt there is a "commercial" standard for failure modes, because rigorous industry-wide FMEAs are usually not performed for commercial components/assemblies.

Tom

Hello Tom,

Maybe true. However, some of the broken glass in diodes I have seen were when I was serving. We weren't supposed to repair this stuff ourselves but the maintenance guys used to be quite happy when we presented the diagnosis on a silver platter. A lot of times it might have been simple mechanical trauma that cause it.

Oh yeah, there is. If we didn't do FMEA in the medical device industry the Federales would come swooping in and shut the place down, and for good reason (FDA is known to occasionally do that...).