In the past couple of years, I’ve had some strange stuff happen to some of my electronics. It has started to make me think (not seriously) that electronics can get better, just like living tissue repairs itself. I know this is totally impossible, but are there any explanations for the following events?
My 19" computer monitor started to have problems, when I would turn it on, everything would have a green tinge to it. This seemed to happen more and more often, till it was an all the time occurrence. Then after seemingly nothing changed (not my use habits or anything near the monitor) This stopped happening, and now my monitor works fine and has for a year or more.
About two years ago, the backlight on the display on the radio in my car stopped working. If you shined a bright light or the sun on the display you could see the display, otherwise it was black. Yesterday, when I got in my car after work it was working again.
So what is at work in these situations? Do electronics spontaneously start working again?
As for the monitor with the green tinge, it’s possible that the cord was not plugged in all the way, or that there’s a loose connection somewhere which managed to get jostled just the right way and hasn’t been loosened again.
For the monitor it may have been a loose connection as suggested or you may have had a magnet (like in a fan or something) near the monitor that you moved. It could also be the monitor needed to be degaussed and someone pushed that button when you weren’t looking (assuming it has such a button).
As for the car stereo I’d say it was a loose connection that bounced back into place.
Just guesses. You are right that they cannot repair themselves.
In some cases, a “component” or “trace” in an integrated circuit can “heal” by random migration of a few atoms through the lattice. This may bring it close enough to spec for a previously nonfunctional circuit to work. I’ve seen dead devices “come back to life” after long periods on the shelf, even though extensive debugging has pinned the problem down to the inside of a specific IC (not a loose connection).
I can’t prove that the mechanism above was responsible, because that would require atomic level examinations (probably destructive examination) but lattice drift mechanisms have been observed in hundreds of published papers
Of course, lattice drift is more likely to make your stuff break rather than heal. There are only a few cases [e.g. heavy current on certain chips) where “helpful drift” helps even 10% of the time, and the results are usually unstable: the chip remains on the edge of failing, it’s just working for the moment.
In rare cases a discrete electronic component (e.g. transistor) may actually get “better” on its own. But in most cases, when an appliance appears to “heal” itself, it’s usually the result of a connection getting better.
The least reliable electronic component is the connector. Connectors are notorious for being intermittent, and can cause all kinds of problems brought upon by thermal cycling, vibration, too much current, too little current, arcing, fatigue, etc. This can cause oxidation, pitting, corrosion, loss of spring pressure, brittleness, excessive resistance, arc burns, loss of surface finish, misalignment, etc. (Connectors suck. I only include them in a design if I have to.) And so not only will connectors fail, they will also get better.
Supporting anecdote: When racinchikki moved to Texas, her computer monitor lost its red somewhere along the way. We went to the local computer shop and got a used monitor, and put the broken one in the closet, intending to dispose of it in a creative manner (i.e., .45ACP through the screen) but never got around to it. About a year later, I bought a computer, and since the store with the cheap monitors had closed for the weekend, we hooked up the broken monitor to check out the computer. The monitor came on in full 24-bit color, and I’ve used it ever since (that was in July).
Check the connector on the monitor where it plugs into the video card. One time I was working on a friend’s computer and plugged it in. All the red was missing from the image unless you held the cord just right. Apparently one of the pins was slightly bent to begin with and when I plugged it in, part of the pin went into its hole and the rest of it folded leaving only a short part of the pin sticking in the hole. That’s one thing it could be.
I used to be cursed with magic fingers back in the day when I earned money by fixing stuff. Customers would bring me faulty kit, and just from my very presence it would spring back to life and never misbehave again. I could never charge for this service, so the customers were delighted.
Crafter_Man is right about connectors. They’re difficult (and expensive) to make so that they work reliably, and so generally they don’t. A metal-metal contact will also need a minimum wetting current flowing through it to make a good connection. As the contact surfaces oxidise and get dirty the wetting current required to make a good contact increases, and they start to fail.
Anyone who drives a Citroen in a damp climate can tell you about connector reliability. At the moment I have fluctuating reliability in the contacts in my dashboard dot matrix display, external temperature sensor, radiator water level sensor, left tail light, and fuel gauge.
I think rackman’s monitor feeling off-colour was due to someone waving a magnet (possibly a loudspeaker) too close to the CRT. This will give a lovely sunset effect as the electrons are bent off course and hit the wrong colour phosphors, but will gradually fade each time the monitor is switched on and the tube is deguassed by the degaussing coil wrapped around the back of it.
Dry solder joints can also be a problem with monitors (or indeed anything electronic, but in particular things that go through temperature cycles). If this is causing an intermittent colour balance problem then the shading will be uniform over the whole screen. A degaussing problem will look more arty.
Semiconductors can show some weird effects too as dopants migrate about and suchlike, but it’s rarer. Chips have connectors too, and sometimes the thin bondwires that connect the legs to the silicon can get a bit intermittent. Also most IC packages aren’t as hermetically sealed as they might appear, and they can allow ingress of water and air. Semiconductors also age. Anyone wanting to reproduce exactly the characteristics of a 1960s transistor amplifier (say) can’t, because no-one makes those germanium transistors anymore, and any you might find gathering dust on a shelf will have aged to some extent.
I once had a car that when you turned the headlights on, the radio would start playing. Even with the radio off. It did this for about a week, then never did it again.
My monitor has a flaky capacitor in its vertical amplifier. A couple days a year, the temperature and humidity get just right, and the thing oscillates, providing me a screen view that stretches and shrinks several times a second. When the weather moves out of this narrow range, everything goes back to normal.