Can A Hard Drive Breathe Regular Air Like Us

Hey there! I have here a 40 gigabyte hard drive, and I’m about to do some sensationalism with it It’s a well known fact that spinning hard drives are incredibly delicate devices They’re manufactured in clean rooms, not just because dust getting into the mechanism would be bad, but because it would be utterly catastrophic. Maybe not everybody watching knows how spinning hard drives work And if you want a full explanation I recommend you google it, look on youtube, there’s millions of explanations that you can get in a millisecond that are gonna do a better job than I can do here But, I can show you the basics This obviously is a dead hard drive, per the label so I’ve opened this up so I can show you the parts Youv’e seen this in the icon in the Windows Explorer at some point or another at some point or another [no sound] The drive has a set of platters, in this case only one, coated with magnetically sensitive material and it has a arm, that can swing in and out, with a little electromagnetic coil on the end that can either read what magnetic state is present at a spot on the disks surface, or it can change the magnetic state of that spot In other words, it can either read or write one bit to a particular spot on the disk And to store your data, all it does is as the disk is spinning, it just writes very tiny dots of data very very very close together onto the surface of the platter Now, if you consider that if you have a multi-terabyte hard drive, you are storing trillions and trillions of little dots on this disk, You can imagine that it’s pretty precise. This is incredibly precise work. and it’s actually shocking that we’re able to make these things work at all, not only in our homes, but on the move. We put ‘em in laptops, at least, we used to before we switched to SSDs We even used to put ‘em in MP3 players. In 2000, it wasn’t unusual for people to just walk around, with one of these things in their pockets, spinning away, playing their MP3s This technology requires such close tolerances, that it’ll break if you look at it funny.

01:51 - Obviously this one is dead, but- either one of these would die if i just picked it up and dropped it two inches. Of course, in new computers, even that’s been mitigated. Starting, like, 20 years ago, they started putting accelerometers in laptops so if you drop one, it detects that it’s in freefall and shuts the hard drive off Tells it to brace for impact before it can hit the ground If you’re careful - and lucky - a hard drive can last thirty years. There are hard drives from 1985 that people can still get data off of. It’s incredible, considering how precise everything is.

02:19 - Even through transport, and being put into and out of boxes, and being moved from place to place, getting hot, getting cold, they still manage to survive. This robustness doesn’t actually translate into lower internal tolerances. Just because they can handle all that stuff, they’re still incredibly precise inside They just somehow built them to be so strong that those tolerances are maintained, even when they’re roughly handled. Those tolerances are so close, that our brains can’t actually comprehend how close they are. Listen to this: “It’s generally accepted that the space between the head of a hard drive and the platter is so small, that a fingerprint on the surfac of the drive will hit the head and destroy it.

” 02:57 - Intellectually, we can process those words, and we know they’re true but at the gut level, do you really understand what I just said? Do I? Do I accept and internalize this? I don’t think so. To me, fingerprints are two-dimensional They’re completely flat. They have no mass, no thickness It’s impossible that they could run into the head of a hard drive. It- it just doesn’t make sense. But it’s true. I know it’s true because I’ve read about it in research documents A human hair, or a single dust mote, are also thick enough to destroy a hard drive instantly. I’ve seen electron microscope pictures where a hard drive head hit a piece of dust and it drug a groove into the surface that looked like the Grand Canyon. It’s incredible.

03:33 - It’s hard to imagine that a piece of dust could do that. Now, this is stuff I’ve been hearing since I was a little kid, and when I was a kid I opened up a few dead hard drives - or, ones I was told were on their way out - y’know, looked at the inside, but I never tried putting them back together and using them. I always just assumed they wouldn’t work, but now I wonder- What if I had? Am I certain they wouldn’t have worked? See, people say stuff like: “If you touch high voltage, it’ll kill you” But I’ve touched 110 volts about forty times in my life, sometimes with both hands, and I’m still here. so it’s clearly not absolutely true. It doesn’t mean you should do it, but it is a lesson that people will say stuff will absolutely happen, but the reality of it is much less likely than a hundred percent. How many times WOULD I have to touch 110V before it would stop my heart? We… won’t be testing that one.

04:15 - But I am going to test this hypothesis: If you take the cover off a hard drive in a normal atmosphere, will it destroy it instantly? Could this really be true? I tried looking this up on Youtube and I didn’t find much. I found a bunch of videos of people running hard drives that they said were already dead with the covers off And I found one video where someone said that they were booting Windows 7 off of a drive with the cover off, but I have no proof that that’s true. I did find a video of a guy moving the platters from one hard drive to another for data recovery, and he didn’t use a clean room, he was just under a laminar flow hood, which is a thing that blows clean air down on your workspace to push dirty air away, But that’s a far cry from a clean room, so it seems like there’s some possibility that these aren’t as delicate as I’d been led to believe Now, the air in THIS room is CERTAINLY not clean It’s full of everything from- spiderweb particulates, to mud dust off the rafters, I have a heater fan in here that I run that blows concrete dust up off the floor you know, the air in here is just about the worst environment for a hard drive with no cover But if I take the cover off, how long will it be before a particle actually lands on the platter, in the path of the head? I wanna find out. It’s not gonna be instantaneous, I’m sure of it So I built a computer from spare parts to do this experiment, And- it’s not going to be an utterly unscientific experiment. If I were to just take the cover off the drive, plug it in to the computer, boot it up, and then try and copy some files off of it, I don’t think that would tell us much, because there’s so much time between the two, anything could have happened You know, I could have exhaled and damaged the surface of the disk We’ve gotta get this done in the shortest amount of time possible, so that we can get the drive when it’s still working, with the cover off, and then watch it die.

05:54 - To that end, I’ve prepared this drive by filling it with the entire run of Colombo, the best 60s/70s television detective show, and I’m going to start copying that data off to the boot drive here, then I’m going to stop the transfer, let the drive settle, release the screws, and then I’m gonna restart the transfer. Now, I said “release the screws”, not “take the cover off” because- there’s a bit of an oopsie here. A friend warned me that some hard drives tension the head using the cover of the drive. and when I opened up this 80 gig here, I found that that did seem to be true. This screw right here actually goes right through the head spindle, and if you take it out, the head starts acting kinda weird when you power it up flicking back and forth, as if it’s upset at something.

06:33 - You put it back in and it starts acting normal again. So- that ruins our whole experiment, right? Like, wasn’t the question, “Can you run a hard drive with the cover off?” The answer’s “No” Well, to me it was never a question of whether it can run with the cover off, but rather, whether it can run with dirty air in it. Since this cover will bend up quite a bit, I think I can just wedge it open, and that’ll be good enough. So there’s the plan. We’ll take these screws out, except for the spindle wedge the cover open, and we’ll see how long it takes before it dies. Now, the other element we have to worry about to keep this mildly “scientific” is actually finding out when the drive starts having errors.

07:07 - It’s actually pretty normal for hard drives to have errors in normal operation, so if a hard drive fails to read a spot on the disk, it’ll just read it again and again until it gets a clean read. If it eventually gets a good one, then you don’t ever find out about it. So it can be quite a while from when a hard drive has problems, to when you actually see the problems. Fortunately, modern hard drives have very intelligent onboard firmware, and they have a feature called S.M.A.R.T. that monitors the hard drive’s behavior and health, and can tell you when your hard drive’s about to fail.

07:32 - It monitors dozens of variables to get an idea of the hard drive’s health, like how hot it is, how many time’s it’s been started, how many times it had to reread a sector before it got a good read, -and all sorts of other crap, and from all that, it synthesizes a simple, “Hard Drive Is Healthy” or “Hard Drive Is Going To Fail Soon Now, with the right tools, you can read those variables straight out of the drive instead of waiting for it to tell you that it’s failing. You can just read how many times it’s had an error, and how many times it’s had to reset the head, and so on. So I’ve written some software that lets me do just that. This here is a little Windows utility I wrote which continuously queries the drive’s SMART status and reads out all the variables that it has onto the screen -this really gets in the way of my light- So right now it’s reading off the host drive, which is presumably pretty healthy, since I’m booting off of it. And all these numbers are sort of tuned to this drive, so they don’t mean anything in themselves really, and hopefully we’re gonna see one of them and quicker as the drive dies.

08:25 - The smaller numbers that say “Raw” are the hard drive’s internal variables and you can kinda ignore those unless you know specifically what this hard drive means by those I just thought they’d be interesting to capture for later. So that’s that, I’m just going to point this at the other hard drive, and start monitoring it, and then we’re gonna kill this thing. So, the hard drive’s healthy right now, and here’s all the numbers we’re starting with Hope I get this right - ‘cause I only get one shot [rustling noises of disassembly] Alright - let me get something to wedge it with [loud snapping noise] I broke my tweezers. [no sound] [quiet rumble of hard drive spinning increasing to a low roar] Okay, I can hear the disk howling in there So it’s definitely, uh, definitely open Let’s go ahead and start our copy [mouse clicks] Well, it’s not dead. It’s definitely open to the atmosphere and still going We’re copying files off of here.

09:27 - So, the values I’m looking at here are the “Raw Read Error Rate,” which is still at zero, the “Start/Stop Count”, which happens every time that it puts the head back to zero and flips it back out because it thinks it’s lost it’s place, that’s not going up, although we would hear it if it was, the “Seek Error Rate” isn’t going up, So far, SMART seems to think this drive is healthy, even though the lid is halfway off of it [sound of hard drive platter spinning changing as screwdriver is adjusted] This says it’s gonna take 12 minutes. I don’t wanna stand here for 12 minutes. [over cellphone speaker] Thank you for calling the Safeway Pharmacy I called my pharmacy to transfer a prescription, and it’s still not dying Oh, oh- There we go! Whew! Boy, my hand is tired. There it is - there’s the hard drive head Doing it’s thing And there is the spinning disk - you can barely see it But there it is So there’s no question - this thing is utterly open to the atmosphere right now, and there’s nothing wrong! I promise you, I absolutely assure you, there is filthy air in there, full of particulates, This room is disgusting, compared to this hard drive’s natural habitat I got about 35 gigs of data on there and this is a 40 gig drive, so we’ve covered about half of the surface of the drive so far, we’ve hit every single bit, on half of the drive, no errors I’m astonished - I cannot believe it’s gotten this far. Nice [no sound] I honestly can’t believe this. It’s 83 percent complete. It’s doin’ it! [no sound] I’ll admit - I haven’t been paying perfect attention, but I haven’t seen anything change. There are no read errors.

11:13 - [no sound] 97 percent! 98 percent! 99 percent?! Ohhhh! Wow! I- I honestly can’t believe that worked Just to clarify, that drive is three gigs short of full Y’know, 80, 80-90 percent full So- [sighs] What did we learn here? I’m not sure what I learned here. This was not the outcome I expected. I thought it was going to die. Now, there’s some caveats here- This is a 40 gigabye hard drive, from about 2006, so it is 13 years out of date. Modern hard drives have a density that is unbelievably higher. They have changed the rules of physics in order to make these things possible If I were to put a 6 terabye hard drive down here and do the same thing, I have the feeling it wouldn’t have worked There’s a lot of options for what happened here. Maybe the hard drive surface isn’t as sensitive on drives of this age as I was told? Maybe- even though I have the hard drive cover wedged open, there’s not enough dirty air getting in there to cause a problem.

12:18 - It’s possible, I suppose, that the platters are forming a sort of vortex and just sort of keeping the same bubble of air around them but that doesn’t make a lot of sense to me, because these hard drives actually have vents on them, with little filters so this hard drive should have been sucking in air. I don’t really know why that succeeded. We covered, like, 90 percent of the hard drive’s surface if there were any problems, we would have found them. So, I guess there’s just one more thing to do here, which is: we gotta take the cover the rest of the way off. I mean, supposedly it’s going to mess up the spindle, but we don’t know until we try I’m gonna shut down the SMART monitor for a moment, because it’s making the hard drive heads thrash back and forth and I don’t want them to thrash while I’m taking the cover off. Okay, here we go. [screwdriver clunking] [rhythmic clicking] Whoa! Oof! As soon as I loosened the screw, the hard drive head started thrashing back and forth [more clicking] As you can hear And the hard drive immediately spun itself back down So- let’s see if I can get it to work again by putting it back in [clicking changes tone] Oh, yeah! The sound of the head moving changed completely when that screw bottomed out, so it definitely needs the screw.

13:34 - But I wonder if it needs the entire top, or just the screw? Probably the entire top - but let’s find out. First let’s check the SMART status again The “Raw Read Error Rate” is still at zero And let’s see if we can still read it. [mouse clicking] Sure looks like it. Okay, so, it’s still working. Let’s take this cover off. [screwdriver clunking, hard drive spinning noise gets much louder] [hard drive head starts clicking again] So I’ve got the cover completely off the drive now, and it’s still not showing any errors in the SMART monitor, but it is refusing to do anything so I think it wants tension on that head. [high speed clunking and clicking] It seems that I was right, that the hard drive cover really does provide a very specific tension on the hard drive head.

14:24 - I’d been adjusting it back and forth here and I found that, if I put too much on, the head freezes in place, and if I put too little, the head gets upset that it can’t find it’s position correctly, ‘cause it’s not as heavy as it expects, so it just flicks back and forth, then gives up But what’s astonishing is that throughout this process, as I’ve tightened the screw down and loosened it off and taken it out completely, and taken the cover off, and put it back on, Windows has kept trying to copy the file, and the hard drive has just been a trooper, it just keeps trying! And every time I get the tension correct for just a moment, it starts copying the file again - successfully! Now, I can’t get the tension right, so I’m gonna put the cover back on normally [unearthly whining/beeping noise, changing pitch and pulsing] Hear that noise? I’ve got the screw too tight, and the hard drive heads are fighting to try to get the head to move. That’s pulse width modulation, high frequency signals trying to move the head back and forth. [whining drops off] Loosen it just a bit- Can hear the head moving now [rattling noise] Okay, now I can’t seem to get it to behave again, I could have destroyed it, but let’s try putting the rest of the screws back in Okay, I just finished putting the last of the screws back where they belonged, and I tensioned them all down. I think that’s it, I think maybe I finally did it in. Just to be sure, let’s try restarting the copy.

[sighing] 15:43 - [mouse clicking] No way! Absolutely, no way! It’s working. I can’t believe it - this hard drive has had it’s top open, completely exposed to the elements, with my dirty hands over it, and a screwdriver with, you know, with little metal bits all over it you know, I’ve been breathing over it and everything, and here it is, copying data perfectly fine. I’m not blowing smoke, I’m not- I honestly didn’t intend for this to be a sensationalist experience, but- I cannot believe this is working. I cannot believe it survived this ordeal. Everything I’ve been told appears to have been, at one point anyway, misleading Not a lie, I mean, I’m sure if I opened it up and touched the platter and stuff it would probably stop working eventually, but this is far, far more durable than I expected. This hard drive’s been through hell and back! I’ve taken the cover completely off of it, the head, y’know, came off it’s moorings and no longer was tensioned correctly so I would have thought it would have been you know, dragging on the platters or something like that.

16:48 - I spent a good fifteen minutes messing around with the head tension, back and forth, back and forth, you know, freezing it in place so it couldn’t move at all and was just straining against it’s servos, And you put it back together, and the damn thing just works! This is the Shure SM58 of hard drives, the Technics SL-1200 of hard drives [hubrisly] Would this break, if I dropped it? [loud bang] [hard drive head clicking] Yes. .