| Forum Home | ||||
| Press F1 | ||||
| Thread ID: 126726 | 2012-09-15 03:31:00 | Computer wont boot after power surge | antares (16890) | Press F1 |
| Post ID | Timestamp | Content | User | ||
| 1301098 | 2012-09-21 16:43:00 | One of my responsibilities was to analyse common failures and report back to manufacturers via my employer, with recommendations for changes. Exactly what revisions were implemented was entirely up to the manufacturer, but sometimes it was simply a change of output device manufacturer, a change of the device specification, or revisions to the control circuitry in relation to over-current conditions. Many facilities could not bother or already had engineers who did that. However, your job was to find the most common reason for failure - manufacturing defects. That point made repeatedly. Some work with switchers involved finding a few that were defective by design. As you apparently know, a switcher should output rock solid voltages even when incandescent bulbs dim to 50% intensity. When no longer able to output a stable voltage, then the switcher powers off. In any design, this function was always tested to confirm numeric specs. Were we destroying prototypes by applying a low voltage? No, never. If low voltage caused a failure, the design was 100% defective. That reality contradicts widely believed myths. And applies to all electronics. For example, one switcher periodically went 'strange'. So the tech would simply replace it rather then discover a problem - that festered. To maintain galvanic isolation, a feedback circuit featured an opto-isolator. But opto gain was too low. Opto needed a gain of 150. Manufacturering could not obtain them. So manufacturing used one with a gain of 100. Since observation (also called junk science) 'proved' it worked, they shipped them. Then we observed strange and intermittent failures. Supply would be rapidly power cycled until it would power up. Techs were even doing that rather than find the problem. So was the failure due to low line voltage? Of course not. Like most failures, it was traceable to manufacturing defects. Sometimes components are manufactured defectively - ie the famous electrolytics with counterfeit electrolyte. Sometimes manufacturing itself changes a design. But most failures are due to manufacturing defects. Most who just 'know' it was a surge or low voltage have no idea what caused a failure. That is the point as demonstrated even by wainuitech. Many who fix things don't want to learn why failures occur. Don't want to avoid future failures. Only want to fix it, as he said, ASAP. Therefore never really know why things fail. Most only know what soundbytes and hearsay tell them to believe. Reading walls of text to learn how things work is hard. This dicussion started with speculation that "Power Surge takes out the power Supply. ... can get 240V into 12 V components,". Can? Yes. Probable? No, it is virtually impossible. Obviously impossible IF a switcher design is understood. More likely, protection inside a supply is bypassed. Others assumed damage elsewhere - ie the disk drive. Chaz even recommended a UPS as if that would avert a drive falure or failure never defined. A UPS will somehow accomplish something that the UPS manufacturer does not even claim? But again, the point. A UPS was recommended even when no one knows what failed or why. Why would anyone recommend a UPS when "Nobody can conclude hardware damage due to woefully too few facts."? A point made often. So many recite solutions that are only "usual suspects". As if damage is created by power off. Or as if damage means a surge existed. Advertising and hearsay says to blame those "usual suspects". Not blame most common source of failure: manufacturing defects. As noted earlier, surges occur maybe once every seven years. Even transistors power cycle so many times without damage that the spec number is no longer listed. Most who know why it failed are only speculating wildly about power cycling or surges. And then recommending solutions (ie a UPS) to avert that mythical failure. UPS recommended even though it does not even claim to protect from surges. But popular urban myth says a UPS will protect from surges ... that the OP probably did not have. Most failures are manufacturing defects. Locate those defects because manufacturing defects (not sudden power loss, surges, etc) are the most common reason for failure. Your post even demonstrates that important concept. What does that mean for the OP? Execute diagnostics to discover what is defective (hardware or software) before fixing anything. Solutions are not obtained by fixing the "usual suspects" or curing symptoms - ie fixing something as fast as possible. |
westom (16792) | ||
| 1301099 | 2012-09-21 22:52:00 | Well to stray back towards the topic slightly, Nobody claimed low voltage or loss of power damaged the hard drive. Rather they were referring to the possibilty of it corrupting software which can happening when the system has cached file operations in memory not yet written to disk or when a file operation is in progess when power is interrupted. This data loss is what a UPS mainly protects you from and is the reason one is often reccomended to protect from power outages, not for any protection from hardware failure as you seem to think was said (it wasn't). Most decent UPS devices do also have some surge protection built in but it is not the main purpose of the device (also I personally think surge protectors are a waste of time but that's a different topic). If the OP was a real problem (and not an alias of yours starting a thread so you could answer it) then we should be focussing on suggestions to help the situation. If you think diagnostic tools are the best Idea then you should link to them and explain their use, which ones to use and why. Some of what you post makes sense, but it's lost among a sea of pointless arguments. If spares are available swapping hardware is a valid diagnostic tool which can quickly eleminate or Identify what area has a problem, it doesn't necessarily prove what the problem is but can help point us quickly in the right direction. No one is advocating buying new hardware without first attempting to prove if there is a fault. Several diagnostic suggestions have been made in fact. First, the PC isn't booting so any software that requires a running system is immediately made useless 2nd the OP describes a strange noise that doesn't sound like a software error, identifying the source of that would be a good Idea Booting from a Live Linux CD or a diagnostic software disk (very often linux based as well) is a first step to establish if the hardware is still functioning Running memtest is non-destructive and worth trying Running scanning tools on the hard drive in an attempt to repair corrupted files is worth trying The OP stated that they had tried and failed a repair from the boot menu and asks if a linux live CD is a good Idea to recover files and is the hard drive faulty, the answer is yes linux is a good Idea, and no we don't yet have enough information to establish if the hard drive is faulty. We can tell windows is corrupted so it's worth attempting a repair based on the premise that it's simply software corruption, once the data is backed up reinstalling or repairing windows is probably the quickest way to see if there is a fault. Arguing about what caused the fault is pointless, especially as we haven't established what it is. In any case the OP hasn't returned, so either they have found help elsewhere, the tangent the thread went off on has scared them off, or there was never a real problem to begin with. |
dugimodo (138) | ||
| 1301100 | 2012-09-21 22:59:00 | Chaz, Nick, you're not helping with your little digs. Post count isn't everything... | Chilling_Silence (9) | ||
| 1301101 | 2012-09-22 00:40:00 | In any case the OP hasn't returned, so either they have found help elsewhere, the tangent the thread went off on has scared them off, or there was never a real problem to begin with Very true, sometimes as we all know people post and never come back for what ever reason. I hate to say this, as I'm kind of expecting the reply I'll get from a certain person ;) But yesterday a workstation I had a PC on the UPS was beeping like crazy when I got back from a job, the PC that was connected to it had shut down as the battery was too low to keep it going, the output voltage was down to 0, instead of the normal 230 -237 it displays. There were guys working on the power lines down the road when I came home so maybe they had something to do with it. The fuse on the main board had tripped, and the UPS clicked in. What caused it - no idea, I wasn't here. When starting the PC it came up with the exact same error the original post was here - selected start normally, but it refused, it took 4 goes to run through windows repair startup then it ran normally again. |
wainuitech (129) | ||
| 1301102 | 2012-09-24 05:39:00 | I followed up Wainuitech's suggestion of Googling westom, and it is quite obvious he is a technical troll nutter, and best to be completely ignored. At least two other forums of the hundreds he subscribes to reckon he suffers from Aspergers Syndrome. He warrants being banned. "People with Asperger syndrome often display behavior, interests, and activities that are restricted and repetitive and are sometimes abnormally intense or focused. They may stick to inflexible routines, move in stereotyped and repetitive ways, or preoccupy themselves with parts of objects.[27] Pursuit of specific and narrow areas of interest is one of the most striking features of AS....." en.wikipedia.org He typically searches out electricals posts, preferably those involving 'surge protection' and lightning strikes, and PSU failures etc and repetitively says the same things over and over in each of the hundreds of posts. He is obssessed with talking about 'the numbers', yet when he got involved in a thread about cars colliding head on, although he rubbished nearly everyone else and claimed the problem was high school physics he did not set out the mathematics of the problem, just rabbitted pages of screed. Here are just two out of thousands of posts, good for a laugh: forums.moneysavingexpert.com www.quartertothree.com As one poster said in 2011 if you put westom + elecrical + question into Google search you get 11,000 results. now in 2012 you get 12,300 :) |
Terry Porritt (14) | ||
| 1301103 | 2012-09-24 05:50:00 | :D :D I feel sorry for you westom...AS sounds real bad. | ChazTheGeek (16619) | ||
| 1301104 | 2012-09-24 05:52:00 | I take it you didnt read chill's post. | icow (15313) | ||
| 1301105 | 2012-09-24 05:53:00 | Whoops... | ChazTheGeek (16619) | ||
| 1301106 | 2012-09-24 06:18:00 | Such people need kindness and understanding. ;) | KarameaDave (15222) | ||
| 1301107 | 2012-09-24 06:22:00 | Such people need kindness and understanding. ;) Indeed, as long as he refrains from trolling. |
Terry Porritt (14) | ||
| 1 2 3 4 5 6 | |||||