Are power outtages harmful for modern PC hardware?

[John Solar System]

Hey everyone. I had a power outage today and I have a high-end PC that is connected to a power strip with surge protector (it is an Schneider APC one). Normally, this power outtage was the first that happened for years and I am just curious about it's effects on the hardware. Does it harm anything or modern PC hardwares have built-in protectors that deals with this kind of stuff? Should I be worried about something? I have turned off the power strip after the power went and waited to switch it back on till the power came back. I also have turned off my PSU and took out the modem cable aswell.  I've already decided to buy an UPS just for the peace of mind but I am curious about this question. Thank you and have a nice day.

[Motifator]

A power outage can indeed break your PC, especially if you're running a PSU without a MOV, no surge protector, old house without grounded outlets in bad condition & no form of protection, no nothing at all.

But in your case, it likely won't do much.

[da na]

It can cause PSU issues, but if anything it'll corrupt OS or delete unsaved work. Hard drives are more robust now and don't crash on power loss like they did 40 years ago.

[Donut417]

1 hour ago, John Solar System said:

just curious about it's effects on the hardware.

Ive had a few power outages in my day. My Plex server which runs hardware from 2013 isnt on a UPS and I haven't had an issue with the few outages we have here in Michigan. That being said, Id be more concerned with data, as a sudden loss of power can cause data corruption or data to go missing, for example if the data is in a cache and not written to the disk yet. 

[Pickles von Brine]

3 hours ago, Mel0n. said:

It can cause PSU issues, but if anything it'll corrupt OS or delete unsaved work. Hard drives are more robust now and don't crash on power loss like they did 40 years ago.

I had that happen to a customer. They had a 6TB drive running fat32... Corrupted the file system pretty bad. 

[da na]

5 minutes ago, Pickles von Brine said:

I had that happen to a customer. They had a 6TB drive running fat32... Corrupted the file system pretty bad. 

Not just an unexpected power loss, I mean the head actually hitting the platter as the cushion of air disappears. Even with landing zone drives I've never seen something like this happen on anything past 1982. I'm sure it still can, it's just not a guarantee anymore as it once was since drives are designed to automatically park on power loss nowadays (instead of just leaving the head in place over the platters as they once did, allowing for it to crash and destroy the platters on next spinup).

Power goes out pretty much weekly on my school campus, and as a hard drive enjoyer myself I have had none of that happen. Granted, I use high end drives so they're designed to be more robust, but I'm sure that no consumer grade drive would just die like that either. 

Unless we're talking Barracuda 7200.9 series. 

[Pickles von Brine]

1 minute ago, Mel0n. said:

Not just an unexpected power loss, I mean the head actually hitting the platter as the cushion of air disappears. Even with landing zone drives I've never seen something like this happen on anything past 1982. I'm sure it still can, it's just not a guarantee anymore as it once was since drives are designed to automatically park on power loss nowadays (instead of just leaving the head in place over the platters as they once did, allowing for it to crash and destroy the platters on next spinup).

Yeah those super old drives would do that.... Hasn't been a problem since I think about the early to mid 90s. 

You know what is more fun? Having the drive spindle controller fail while the drive is in operation. Drive still thinks it is on but the spindle is spinning down. I will let your imagination go from there. 

[da na]

5 minutes ago, Pickles von Brine said:

Yeah those super old drives would do that.... Hasn't been a problem since I think about the early to mid 90s. 

You know what is more fun? Having the drive spindle controller fail while the drive is in operation. Drive still thinks it is on but the spindle is spinning down. I will let your imagination go from there. 

Ooh... that sounds terrible.

I imagine a controller board swap would fix that, but the data... ouch. I would think it could at least park the heads, were it built competently with a failsafe in place? 

I've had to replace a few drives with that issue. Best example would be some very worn out Caviar 1200 drives run 24/7 in a server. 15 years of flying hours on the spindle motor - and it sure did sound like it. Eventually scrapped all the drives since they made the most horrible sound imaginable. But, one of them had its motor controller fried somehow. And it would have been with the machine running, since it wasn't shut down pretty much ever. (I've found the foam on the controller board on those older WD drives makes a great thermal insulator, it's often discolored from heat. In this case, the chip had released the magic smoke and burned all the way through that foam, likely a death from overheating.) Anyway, swapped controller board and it worked again. 

[Pickles von Brine]

8 minutes ago, Mel0n. said:

 

Unless we're talking Barracuda 7200.9 series. 

The barracuda drives were known for manythings. 7200.6-7200.11 were known for many things. Spindle failures, medica cache corruption, translator issues, multiple head failures and my favorite, spindle power dying while the heads are not retracted. 

[Pickles von Brine]

4 minutes ago, Mel0n. said:

Ooh... that sounds terrible.

I imagine a controller board swap would fix that, but the data... ouch. I would think it could at least park the heads, were it built competently with a failsafe in place? 

I've had to replace a few drives with that issue. Best example would be some very worn out Caviar 1200 drives run 24/7 in a server. 15 years of flying hours on the spindle motor - and it sure did sound like it. Eventually scrapped all the drives since they made the most horrible sound imaginable. But, one of them had its motor controller fried somehow. And it would have been with the machine running, since it wasn't shut down pretty much ever. (I've found the foam on the controller board on those older WD drives makes a great thermal insulator, it's often discolored from heat. In this case, the chip had released the magic smoke and burned all the way through that foam, likely a death from overheating.) Anyway, swapped controller board and it worked again. 

Lets move this to dms. 

[da na]

Just now, Pickles von Brine said:

The barracuda drives were known for manythings. 7200.6-7200.11 were known for many things. Spindle failures, medica cache corruption, translator issues, multiple head failures and my favorite, spindle power dying while the heads are not retracted. 

70% of the 7200.x drives I get in from between 2008-2018 have at least 20 reallocated sectors. 23% of those have a head crash of some sort and do not work at all. 5% of that 70% have click of death syndrome. 

I keep very few on hand, for good reason

although the Barracuda ES from that gen wasn't too bad

[Dabombinable]

1 minute ago, Pickles von Brine said:

No, there is no fixing that. Heads get ripped off, broken armature tears the drive apart. 

And it'd be even worse if the drive uses glass platters...such as my 40GB IBM Deskstar 120GXP.

[Pickles von Brine]

Just now, Dabombinable said:

And it'd be even worse if the drive uses glass platters...such as my 40GB IBM Deskstar 120GXP.

Older 3.5 IBM and Hitachi (when hitachi bought IBM's drive division) used glass platters. Circa late 90s early 2000s. 
All 2.5" drives use glass platters. 

[John Solar System]

Thanks for your answers guys. This may sound weird but does it matter which applications you run when you have a power outage? For an example, let's say that i got a power outage while playing a demanding game. Would it increase the risks compared to a normal power outage while I am on the Windows or?

[Dabombinable]

35 minutes ago, Pickles von Brine said:

Older 3.5 IBM and Hitachi (when hitachi bought IBM's drive division) used glass platters. Circa late 90s early 2000s. 
All 2.5" drives use glass platters. 

The 120GXP were one of IBM's earliest (if not the first) glass platter 3.5" HDD.

Edit: Far from the earliest - 75GXP

[Pickles von Brine]

20 minutes ago, John Solar System said:

Thanks for your answers guys. This may sound weird but does it matter which applications you run when you have a power outage? For an example, let's say that i got a power outage while playing a demanding game. Would it increase the risks compared to a normal power outage while I am on the Windows or?

It doesn't matter what app you are running. Think of it this way. Imagine having someone randomly smash you in the head with a baseball bat and knock you out. Whatever you were doing is now junk and you likely don't remember what you were doing. That is essentially what a power outage is to your computer. 

 

It could be writing REALLY important stuff for the file system, OS, a program, etc. and that can cause corrution. The file system has fault tolerance so it will generally recover. However, other file systems dont (fat32 and others). Either way, if bad data gets written/partially written depending on what and where it can cause some serious trouble.