Westom, Thank you for your input. It was an Antec sp500 power supply but had been in service for several years. It was plugged in to a Belkin UPS, also several years old. I replaced the battery once. She lives in an older home, built circa 1950, without updated wiring. I have not had a recent look at the setup as I live 200 miles away. I assembled three computers using the MSI K9N6SGM mother board. My daughter's, my wife's (no problems yet), and mine that I had to re-cap two years ago. My daughter's machine was running Vista, my wife is running Win7 and I am running Win8.1 with an SSD. I am tempted to upgrade but these are adequate for Spider, Freecell, Facebook and Skype. Thanks again to all.
Neither age nor manufacturer nor software adds anything to better explain the damage. A supply must define, in specification numbers, functions that make that damage impossible. Functions that are specifically required by ATX Standards. Your damage is classic of a supply missing essential functions. Power supply manufacturer is not responsible for providing those function. A computer assembler is responsible for selecting supplies that meet those standards. Therefore some power supplies do not always meet those standards - are missing essential functions.
Originally Posted by cranerw3156
Nothing on its power cord (ie UPS) is relevant. Even 1930 two wire AC circuits should not cause hardware damage. In fact, a UPS in battery backup mode is typically some of the 'dirtiest' power a computer will see. Due to other functions that must exist in every supply, even a dirtiest UPS is sufficient power for a computer. Those functions are irrelevant to what would explain that hardware damage.
Another way of causing that damage is to disconnect parts when a computer is off but still connected to AC mains. Reasons why should be obvious.
Not true. Specs are one thing, but only when the components continue to meet them. I spent many years debugging hardware faults to component, internal gate and ultimate cause of failures. And I have been present at the often explosive end of many power supplies. Power supplies are full of capacitors and all have some sort of transformer or other coil in them. Capacitors are made in many different ways, but nearly all of them age and there are many failure modes. Since they store charges, they can unleash quite a bit of energy when they fail. And since they frequently are used to slow voltage transitions and in circuitry which limits voltage swings, a failed capacitor can allow high voltages to pass into other parts of the supply and even out into the equipment. Copper windings are covered in a thin layer of varnish as their only insulation. Coil wire insulation is much better than it used to be, but only due to improved quality control, not to any basic change in how a coil is constructed. Heating/cooling can cause cracks in the varnish allowing shorts to form. Age can dry the varnish making the cracks more likely. Shorts reduce transformer output voltage causing the downstream components to draw more current which exacerbates the problem. Same with flyback coils. Etc., etc.
Originally Posted by westom
Component engineers constantly strive to improve the reliability of their components, but cost is always an issue (just compare the price of a commercial grade cap and a military grade one) and so consumer grade equipment does not have the most reliable components. Sorry, it just doesn't.
And we all make many other choices based on cost. There is always something more we could do if we were willing to spend more time and money. In a sense, we are all engineers of our own little worlds. (That's what an engineer spends most of his time doing, trading off function for cost.)
I personally like the idea of using an external USB case with a replaceable drive with its own power supply. If the external drive is taking its power from the host computer, then host power supply problems will affect it directly and could damage the drive. If the external drive is not taking its power from the host computer, then a catastrophic PS failure in the host computer could damage the USB transceivers in the external drive, but then you could move the disk to a new external drive housing and it will probably still work fine. If the external drive PS fails catastrophically it could take out the USB transceivers in your PC, but probably would not damage the data on the internal drives. If your MB has multiple USB controllers on it, you might not even have to replace it, though you might have one or more USB ports that no longer work. And if you were really concerned about protecting the PC you could use an add-on USB controller for your external drive. That way damage from the external drive would probably be limited to the replaceable USB controller.
But these are all just probabilities. Such is life. This proposal only protects from one failure mode which you probably won't see again. The next time it will be something different. Other suggestions have been made which will improve your data security even more. Distributing your backups, for example. Backing up to the "cloud" sounds attractive, but I don't have the network bandwidth to fully backup even my personal PC, much less my other six business computers. I backup only my most important corporate data to an offsite computer. The rest I distribute around to other computers on my LAN. Your situation might be such that you can backup everything to offsite storage. Or you can do local backups to a USB drive and then unplug it. That's too much for my old brain to remember, but it obviously works for some. We're all different. I'm sure you can find something that works for you in all of these postings. It's a great group here in the lounge.
Good luck. I hope the data on your drive is recoverable. It might be, you know. Try it in a different PC (or install it in an external USB housing, just in case...) And there are data recovery services if that doesn't work. I have used one once and they saved my bacon.