I think we've all been at least semi-clear about it having little to do with the initial 'power loss' scenario that a certain someone seems to focus on. It's the reactivation of power and the hardware the OP is running that contributes to the potential issue. To claim that his power outages have nothing to do with the problem, and that anyone else is 'shotgunning' to suggest it does (despite the correlation provided by the OP) is laughable in honesty.
Numbers (1990 and Windows 95/98/ME) were correct. A reader interested in learning would then determine why those numbers were correct. Or asked for clarification in a truly respectful manner. If one's attitude is egoistical superiority, then the reply is as posted - accusatory, disparaging, and devoid of knowledge he should have learned. Filesystems that could delete files on power loss were obsoleted after 1990 - as stated. But many consumers continued to buy obsolete technology Windows 95/98/ME. Those who knew computers were using HPFS, NTFS, and other superior filesystems before Windows 95 was released. So, back to damage due to sudden power off. Nobody could list even one part damaged by a sudden power loss. No disk drives were damaged by sudden power loss even when drives were big boxes on the computer room floor or 70 pound monsters whose heads would even through your hand of the drive. One exception existed. An Interdata floppy drive was so defective as to erase track zero on an eight inch floppy. One of so many examples why Interdata died. Even disk drives that moved disk heads with motor oil were not harmed by power loss. Because that standard (no hardware damage due to power cycling) has been standard probably longer than most everyone here has existed. Hearsay and urban myths tend to become facts among computer assemblers. Those fables are why many believe sudden power loss is destructive. Any such statement without reasons why and numbers is usually junk science. A myth of damage due to power loss is settled since no one can even cite one part or component damaged by power off. Move on. More myths. MOVs obviously are not filters. In fact, a typical MOV ignores all electricity until 120 volts rises to above 330 volts. A number even printed on each protector's box. If MOVs are implemented for frequent spikes or noise, then MOVs fail in weeks or months. Please learn from specification numbers before accepting unsubstantiated tales as fact. Or learn why computer power supplies are designed to withstand even 1000 volt spikes without damage. And without using MOVs. MOVs are for an anomaly that might occur maybe once every seven years. MOVs obviously are not filters. And are irrelevant to the OP's problem. Best and robust protection inside a building is inside computers. No matter how 'clean' a UPS makes electricity, a computer first converts that 120 volts into well above 300 volts. And then into high voltage radio frequency spikes. That is 'dirtiest' electricity. Then superior filters/suppressors turn that 'intentionally made dirtiest' electricity into stable, smooth, regulated, and 'clean' low voltage DC. Anything an adjacent protector or UPS might do is already done better inside electronics. An output from my 120 volt sine wave UPS is 200 volt square waves with a spike of up to 270 volts. Ideal power because electronics are so robust. That same UPS may be harmful to small electric motors and power strip protectors. Someone actually recommend plugging a protector into a UPS? Even UPS manufacturers quietly and blunt discourage that. But manufacturers do not like to admit why. A UPS in battery backup mode can be some of the 'dirtiest' electricity in the house. Harmful to power strip protectors. And because all electronics are so robust, that same 'dirty' UPS electricity is also ideal power. Unfortunately many have jumped to conclusions without even trying to understand why original statements were accurate. Apparently many did not know a file system in Windows 95/08/ME was obsolete technology. Apparently many assume a UPS does things even its manufacturer does not claim. Apparently some are searching for emotions in these posts - that are 100% hard facts and numbers bluntly challenging what should have obvious to anyone making a recommendation. OP's problems are not due to power loss. Obviously a CPU, its thermal cream, and disk drive have no relevance to his symptoms. Best way to identify a problem before trying to fix anything are manufacturer comprehensive diagnostics (from better manufacturers). Or third party diagnostics. Test each component or subsystem without adding the complications of Windows. Obviously one subsystem that must be tested is the video subsystem that includes GPU, video display, and other components. Also necessary for an informed reply are numbers using a meter. Once hardware integrity is established (without speculation or doubt), then move on to what might be a software (ie Windows) problem. Power loss did not cause hardware failures. But an anomaly that also caused a power loss might have. Neither a power strip nor UPS even claim to avert that type anomaly. But another and superior solution (that costs about $1 per protected appliance) would have. Of course, we do not even know if that anomaly existed - without first identifying what is actually damaged. By using diagnostics and other tools. Please do not make conclusions by assuming facts and numbers were written wrong. Obviously the file systems in Windows 95/98/ME were obsolete technology. However many did not even know that. Nor the games played by IBM back then to, unfortunately, slow the release of better technology. A story the naysayers should have known before jumping to conclusions. And irrelevant to the OP's problem. How the OP can locate a defect or obtain more useful replies is posted here and previously. With additional information to avert confusion by identifying some posted wild speculation and junk science reasoning. A first step to resolving the problem is collect facts such as from hardware diagnostics.
There's so much irony and outright nonsense in the above post I can't help but laugh. I'm literally only responding at this point to save others time: Don't bother reading it, it's a whole lot of horse piss. Are you even trying at this point? OBVIOUSLY A CPU THERMAL CREAM!! Go back to whatever bar you got drunk in.
LOL. Good one bro. You've set a record on how much can be made up in one post. I left out some stuff, namely the X and Y caps, the fuse, and the coils. But the MOV or MOV's remain the focal point. If you actually want to learn something, try reading this : http://www.techpowerup.com/articles/overclocking/psu/160/3 Funny how PSU's have all that equipment given power outages, surges, and noise don't do any damage, right? Here's a quote from there that specifically squashes the bolded nonsense in your post: "If your PSU does not have an MOV in the EMI/Transient filter then you should always operate it along with a surge suppressor or a UPS, otherwise a spike could damage permanently not only the PSU but your system too." Seriously, where did you go to school for electrical engineering? The Fisher Price academy? Jesus
So this thread obviously got active fo some reason.. Anyways came back to tell ya'll that the power outages stopped and none of the issues that i've mentioned have been repeated, even on purpose if i try to repeat the things i did the issues are not there anymore, i don't know what caused them and i have stress tested ram/cpu/gpu/ssd/hdd, none of them fail and they're up to par so whatever this was, its now gone.
https://www.youtube.com/watch?v=FAozgz3fLQk In games it's apparently popular for helping wounded people. No wonder Weston mentioned it!
In case my cpu has herpes, I switch to an antibiotic cream. Runs a little hotter than ice cream but you gotta do watcha gotta do. Cream cheese is also acceptable. Although I believe herpes is a virus though.
I'm not sure about the cream cheese, always been a little adverse to the culinary computing school. Much prefer a good jar of deep heat myself, as the name infers it's incredible as thermal paste. Especially good for overclockers, just as it is for those who work out. Definitive logic.
IceVip - a best diagnostic tool is heat. Many (without hardware knowledge) see something fail when hot. Then assume heat is destructive. Heat is another superb diagnostic tool. Best used in conjunction with hardware diagnostics. Your problem did not just go away. It has symptoms of an intermittent. A tool for locating intermittents is heat. Two ways to accomplish this. Simply operate a computer in a 100 degree F room. Or use a hairdryer on highest heat to selectively heat suspect components. The naive will excitedly proclaim heat will be hardware destructive. Well, semiconductors must be heated to well over 400 degrees F to be damaged. Any semiconductor must work just fine even when heated to 100 degrees F. As an old engineer once taught me, "If you touch it and don't leave skin, then its not too hot." But heat can identify transistors that are failing today intermittently and will be failing hard months later. Define stress tests. Good reason why stress tests were also called burn-in testing. Because we tested hardware at temperature extremes (well above 100 degree F and well below 40 degree F) to find defects before those defects caused failures. Running diagnostics on hardware in ideal environments (ie room temperature) does not stress test hardware. Selectively heating parts with a hairdryer is one superb technique for finding parts that are 100% defective but are only intermittently creating failures.
If nothing else I've laughed more than once over this thread due to our insane friend. Regardless: Enough is enough, reporting this abject idiocy.
So why do we repeatedly heat and cool semiconductors to well over 500 degrees F to add circuits to them and to not destroy them? Or again, cite a manufacturer specification number that supports your belief. So far, we only have hearsay and speculation. Meanwhile, an Intel Pentium was tested up to 350 degrees. Hardware did not fail. Transistors just no longer worked reliabily. Pentium was cooled and worked just fine. Purpose was to determine how hot a computer in a well drill head could become and still do processing. And still some just know a trivial 212 degrees F must be hardware destructive which proves 100 degrees F must also be hardware destructive. Heat was always a superb diagnostic tool to find defective semiconductors as we used long before the IBM PC existed. We often heated those semiconductors with the tip of a hot soldering iron - to find a defect and do not damage. Even plastic packaged ICs were not harmed. Explain that. A superb tool to locate what is probably an intermittent is heat in combination with hardware diagnostics.
We? Whom exactly would that be? Provide evidence please, if you're going to ask others to do the same. You're spouting a whole lot and not providing any sort of evidence. You're as usual splashing out a whole lot of random non-specific information and expecting it to stand. It doesn't.
Sounds like it was caused by a source power issue. Either the power from the mains was "dirty" or was simply not sufficient to power the system properly.
