Hard Drive Failure and PSU Discussion

[jeffryfisher 2,176]

On 30/01/2021 at 10:20 AM, RobertaME said:

a hard drive crash and a faulty backup (what are the ODDS!)

If a power-supply goes bad, then anything taking power from it can fail (or seem to fail). Therefore, if multiple components fail improbably, and if they're all powered by the same supply, then it's time to suspect the supply (and possibly resurrect "failed" components by attaching them to good clean power).

  Edited February 2, 2021 by Cyclone Boom  
Split to new topic.

[RobertaME 1,376]

8 minutes ago, jeffryfisher said:

If a power-supply goes bad, then anything taking power from it can fail (or seem to fail). Therefore, if multiple components fail improbably, and if they're all powered by the same supply, then it's time to suspect the supply (and possibly resurrect "failed" components by attaching them to good clean power).

Separate power supplies... the backup was done on a networked PC and I was able to pull a backup, but the most recent ones that had all my SC4 work in them were corrupted. (just a fluke... :^/ )

Now my PC is using multiple RAID 1 drives and I do manual backups to several external drives on a regular basis. (one is even not at my home, but kept at my father's house... and that gets brought over every 6 months to refresh it and sent back the same day... that way if my HOUSE burns down I have less than 6 month old backups)

I'm taking no more chances! :^Þ

[rsc204 21,808]

26 minutes ago, jeffryfisher said:

If a power-supply goes bad, then anything taking power from it can fail (or seem to fail).

Perhaps 20 years ago, but modern electronics don’t tend to work like this. Unless you are using real junk. Not to mention modern main boards have safety cut outs, this is a really unlikely scenario.

13 minutes ago, RobertaME said:

I'm taking no more chances! :^Þ

That’s exactly what I did after my first HDD death and the £500 bill to recover the data thankfully worked. Of course, now it means I have to pay 3x the price for the same storage, but as you say, if it’s important take no chances. 

[Matrim Cauthon 344]

3 hours ago, rsc204 said:

Perhaps 20 years ago, but modern electronics don’t tend to work like this. Unless you are using real junk. Not to mention modern main boards have safety cut outs, this is a really unlikely scenario.

It does still happen, but it's far more rare nowadays. The biggest culprits are usually extended brownouts, they overwork the PSU as it tries to draw more current, and can cause DC surges when power normalizes.

A supply with good PFC shouldn't suffer much from this, but anything not in a custom rig probably falls short. MFG's love to cut corners on PSU's, I still run into some that weigh barely an ounce.

[rsc204 21,808]

1 hour ago, Matrim Cauthon said:

The biggest culprits are usually extended brownouts

Is that a problem where you are? It would be unthinkable here in Germany and I've never known of it in the UK either. Granted my only experience is in big cities, but outside of the main power line going out twice, causing total power failure, I've never had to deal with such problems.

1 hour ago, Matrim Cauthon said:

MFG's love to cut corners on PSU's, I still run into some that weigh barely an ounce.

Yeah I hear you on this. Even though it's been a long time since building a PC was cheaper, you still get a better system IMHO. If nothing else you can pick and choose parts, warranties are a good guide to quality.

Had to fix many stupid problems over the years caused by poor/cheap designs, including vertically mounted HDDs in Dell systems and the cheapest crappy plastic mechanism used instead of a spring, on the power button of an expensive HP's case. Geez I remember the bodge job I had to do with hot glue and cardboard on that, ironically it lasted longer than the original did!

[Matrim Cauthon 344]

31 minutes ago, rsc204 said:

Is that a problem where you are? It would be unthinkable here in Germany and I've never known of it in the UK either. Granted my only experience is in big cities, but outside of the main power line going out twice, causing total power failure, I've never had to deal with such problems.

Semi-rural, heavily forested coast adjacent Island(Fidalgo).  I can tell when nearby areas lose power because our browns for a second everytime it does. (Thank god for UPS systems).
Extended browns are pretty rare, but windstorm related PSU replacements used to be significant business for me in the winter. More people use Laptops and tablets now though, so it's quite a bit more rare.

We got quite lucky with the recent windstorm that took out about 700k people's power in the PNW. Ours stayed on, despite many surrounding zones being without power for 3 to 5 days.

[rsc204 21,808]

11 hours ago, Matrim Cauthon said:

Semi-rural, heavily forested coast adjacent Island(Fidalgo).  I can tell when nearby areas lose power because our browns for a second everytime it does. (Thank god for UPS systems).

Fair enough, if you have to deal with that, a UPS pretty much becomes mandatory, although it's not really typical consumer gear. As you say, a laptop essentially becomes a UPS, given they are almost cheaper than desktops these days, for the average user it probably makes more sense. Although I'm an advocate for proper desktops myself, unless you really need to work on the go, they are just better in almost every respect.

[City_Slider 2,716]

13 hours ago, rsc204 said:

Is that a problem where you are? It would be unthinkable here in Germany and I've never known of it in the UK either.

Few Days ago, Europ stands for an Completly Blackout, cuz some of the Nuclear Power Plant in East Europ went down. And if they sell more and more off this freaking Eletro-Cars, we are going from one Black out to the next, trust me.

[rsc204 21,808]

27 minutes ago, City_Slider said:

And if they sell more and more off this freaking Eletro-Cars, we are going from one Black out to the next, trust me.

Electric cars are a (bad) joke and only a temporary solution until we can work out hydrogen fuel cells anyhow. The only environmental improvement they offer is that they don't pollute where they are driven, but over their lifetimes, they are about 2.5x worse for the environment overall than a big gas-guzzling SUV with a V8 or bigger engine! It's just that the environmental damage and emissions occur elsewhere, where they are put together or as part of the ridiculously complicated supply chains for the materials they use.

If you really care about the environment and need a car, a 3 cylinder petrol engine is about the best you can do right now. But, regardless of how dirty your current car is, because making it is far worse overall than running it, keeping an old car is almost always better for the environment than buying a new one. Assuming you bother to look at the full picture, rather than pick and choose your facts.

But all of this is meaningless, because the bulk of the population lives in places like I do, streets. So unless we're going to install power sockets every 5-10m on every street where people live/park, how is this going to work? Yet, many governments are subsidising this crap, in other words giving the Car manufacturers money. Wait for the inevitable, oh but they are making these batteries ever-quicker to charge, soon you'll be able to do so in a matter of minutes like a petrol car. But, at the expense of the life of the battery, which is the dirtiest part of the production *fail*. Not to mention, the cost of replacing the batteries pretty much makes a 5-year old model a write off financially, meaning cars will be used for less time, in other words more get produced than now.

Electric cars are an environmental disaster, but seemingly common sense has gone out of the window on this one.

[City_Slider 2,716]

The Electro Hype can´t work. Have a look at the cooper Production. It went down since Years. The hole World needs Cooper for Electric Connections... that are Billions of Miles. Who is at the Adsministration? Only Idiots. Let SimCity Player Manage the System ^^

[jeffryfisher 2,176]

23 hours ago, rsc204 said:

Perhaps 20 years ago, but modern electronics don’t tend to work like this. Unless you are using real junk. Not to mention modern main boards have safety cut outs, this is a really unlikely scenario.

You must have misunderstood what I was trying to say. I'm not talking about cascading burn-out but rather symptoms that rise in one component because of weakness in another. Also, I'm speaking from personal, hands-on experience with desktop PCs like I have for legacy gaming (more than once in the last ten years). My wisdom may not extend to laptops.

My legacy gaming rig is built from used components (3-8 years old), and it runs an older version of Windows. I've given it a graphics card that's a bit more powerful than the original system might have expected, and it has some extra HDDs as well. When I upgraded that graphics card, the generic power supply, as calculated, handled the load... for about a year. So when I started getting system error messages, I initially took them at face value.

I first thought the graphics card (bought used) had died, so I downgraded to the previous one that happened to draw less power. Then a couple months later, I thought the system HDD was failing. Fortunately, at that point I cottoned to the possibility of aging power-supply, so I upgraded it to both higher quality and greater power. Then I was able to reinstall the more powerful graphics card, it and the hard drive both magically springing back to life -- I had resurrected two important components (and recovered all my data) by buying one relatively inexpensive refurbished power supply. That PC has been rock solid since.

Analysis: As a PC's power supply ages, its capacity to deliver steady, in-spec voltages gradually declines. As that capacity titrates down toward the existing load, marginal/unsteady voltages can trigger system messages that look exactly like some component is dying, leading to misdiagnosis and unnecessary data-loss, expense etc. Further confusing the issue, removing any one component (e.g. the one that complains first) can reduce load enough that all errors go away, possibly tricking one into deciding that the removed component was the problem when it was only the messenger. I believe that many such messengers have been "shot" this way without their owners ever learning that their "dead" components were still perfectly good.

Other users here have similar legacy rigs, so they could run into the same problem, so I stand by my wisdom: Before you give up on a hard drive that seems to have failed (unless it's smoking/squealing), investigate the power supply. For a desktop PC anyway, it's a quick, easy step that could avoid painful expense and/or data-loss heartache, so please don't dismiss my personal experience as if it can't help anyone else.

[rsc204 21,808]

On 31/01/2021 at 7:11 PM, jeffryfisher said:

Other users here have similar legacy rigs, so they could run into the same problem, so I stand by my wisdom

Wisdom is not taking an example that happend to you and repeating it as if it applies to all. I don't personally consider a machine of this vintage legacy equipment, especially since it would need a modern PSU capable of handling the C-State power requirements, which are intrinsic to my argument.

On 31/01/2021 at 7:11 PM, jeffryfisher said:

Before you give up on a hard drive that seems to have failed (unless it's smoking/squealing), investigate the power supply.

Why not simply remove the drive and mount it externally before assuming it's failed? Isn't that a better approach? What does a SATA/USB adaptor cost?, a lot less than a PSU that's for sure.

On 31/01/2021 at 7:11 PM, jeffryfisher said:

You must have misunderstood what I was trying to say. I'm not talking about cascading burn-out but rather symptoms that rise in one component because of weakness in another.

Your PSU did not break your Hard Drive or Graphics Card, it led to safety cut-outs kicking in. Yes, eventually that could cause failure, if you were willing to ignore the totally unreliable system that's constantly Freezing and bury your head in the sand.

On 31/01/2021 at 7:11 PM, jeffryfisher said:

I first thought the graphics card (bought used) had died, so I downgraded to the previous one that happened to draw less power. Then a couple months later, I thought the system HDD was failing. Fortunately, at that point I cottoned to the possibility of aging power-supply, so I upgraded it to both higher quality and greater power. Then I was able to reinstall the more powerful graphics card, it and the hard drive both magically springing back to life -- I had resurrected two important components (and recovered all my data) by buying one relatively inexpensive refurbished power supply. That PC has been rock solid since.

What you describe is a classic example of a PSU that's simply not rated high enough for the components being attached to it. You replace said PSU with a more powerful one and now everything works, but does it necessarily follow that the PSU itself was faulty? I don't believe that it does. Am I saying it's impossible, no, but is it the most likely explanation, I respectfully disagree if you think it is.

On 31/01/2021 at 7:11 PM, jeffryfisher said:

For a desktop PC anyway, it's a quick, easy step that could avoid painful expense and/or data-loss heartache, so please don't dismiss my personal experience as if it can't help anyone else.

I'm not dismissing your experience, but it does not equal an understanding of how things work, it's simply one experience. If I only flew once on a plane and it crashed, does that mean all planes are likely to do so?

If you find me pushy in responding to you, it's simply because I think it's bad advice to suggest everyone start pulling out their PSUs at the first sign of problems and replacing them. PSUs are NOT cheap, at least not these days, and I never replace anything unless I'm very certain it has failed. The danger is, someone comes into a forum, who doesn't understand things well, sees what you wrote and thinks that's the best course of action. You'd find my response quite different if someone had described the symptoms of a faulty PSU and you advised them to consider it's replacement.

On 30/01/2021 at 11:20 AM, RobertaME said:

a hard drive crash and a faulty backup (what are the ODDS!)

A hard drive crash (sorry to drag others into this by quoting posts)...

Does this sound like a power failure to you? By the sounds of things, the drive broke, which is sadly all too common these days, it was then replaced. If the PSU was faulty, doesn't it follow that the replacement would have suffered the same fate?

This is my problem with your logic, you just don't seem to considering the specific circumstances whenever you bring this up. OK, it's an assumption, but I suspect a reasonable one, that the person in question would have checked said drive worked, you know by trying to mount it externally, before assuming all their data was lost forever? I mean, they have since setup a RAID system, that suggests a modicum of knowledge in this area.

I just worry making rash conclusions which don't fit the circumstances might be somewhat reckless, with the potential effect that someone might inadvertently waste their money. I'm sorry if that makes you feel bad, but I'm not going to simply pretend the facts stated here are any different to spare your feelings. This isn't about trying to prove who's better or more technically adept, this isn't personal either. But your singular experience as described doesn't count for much, especially when it's not even contextually relevant here.

  Edited February 5, 2021 by rsc204  
Edited my response to hopefully be more clear on the exact nature of my disagreement.