When reliability hangs by a "wall wart" thread
The other day, one of the many electronic products in the house went dead. Luckily for me, I soon realized it was not the device itself, but its AC/DC adapter/charger - usually referred to as a "wall wart - that had failed. Even luckier, I went through my assorted collection of wall warts from electronic devices that were long gone and found a direct replacement: same voltage, slightly higher current rating (that's OK), and the same plug (that's the most amazing part).
But the whole incident had me thinking a little, and then worrying a lot, about these ubiquitous, anonymous units, which provide the power to run our routers, adapters, displays, and more (you can detail a list, of course). (Note that I'm not referring here to these aftermarket replacement chargers you might buy for your phone or tablet to have another one, or to replace a lost one; I mean ones that come packaged from the OEM with the device you bought.)
Most of today's electronic devices are fairly reliable in terms of their circuitry, but there's a weak link in the overall product. The adapter that comes with the device may be unreliable; and we just don't know, nor can we. Further, some of them are used near their design maximum most of the time; some are mostly in "idle" mode (less stress on the internal components); some will run for years and years without issues; and some last only a short while.
I did some online research to see if there were any numbers on the reliability and MTTF (mean time to failure) of these units. The data, if you can call it that, was all over the place, between somewhat cloudy and very inconclusive.
These wall warts come under so many names and their operational conditions are so hard to document (i.e., duty cycle, or typical vs maximum load, operating environment), and no one really tracks these. Their end-users are mostly average consumers who obviously can't document them in any statistically valid way, and consider them as replaceable disposables. What I found online were lots of anecdotes spanning "it lasted only a few weeks" to "it's been going fine for years and years;" but anecdotes don't help here.
The problem is that you never know what level of reliable design and manufacture you're getting with these power units. They are sourced from dozens (hundreds?) of suppliers ranging from big ones down to garage shops, and almost all have unknowable and untraceable "provenance."
It's not even practical to go by the track record of the main product's OEM, as they often get them from multiple sources as well. As a result, a product such as a router may have a pretty good adapter in one package and a lousy one in the next one on the shelf.
What can you do to overcome or avoid the problem if you have "mission-critical" devices at home or work? There are some options, none quick or easy:
- Have spare wall warts on hand for critical devices.
- Go the hi-reliability route. There are some units available online that claim to be better ones, complete with detailed specifications for input and output regulation, AC-line transient immunity, MTTF, and more. (You will likely have to change the DC plug end to match, of course.)
- Or "roll your own", by finding or buying a solid AC/DC switching supply. It can be a new one, or perhaps one taken from a defunct PC or, even better, from a quality medical or test/measurement product. Put it in an enclosure with an AC cord and DC output; it may not be "elegant", but you will likely end up with a better unit than that anonymous one you have.
The irony is that for all this reliability we now have in circuitry, there's still this very weak link in the overall chain for which we really don't have a comforting or easy answer. The failure of this humble unit - and it will fail, that's what power supplies do, it's just a question of when - can negate all the good in the product and bring down a complex, sophisticated system such as your home network. You can't even blame a harsh operating environment, since most of these adapters are used in fairly benign settings; the problem is with their design and manufacture.
What's been your experience with reliability, consistency, and failures of these units? How do you deal with their inevitable demise, either before the fact or afterwards?