Can A Hard Drive Breathe Regular Air Like Us
![Cathode Ray Dude [CRD]](http://yt3.ggpht.com/3Lm-GuDsob2s_naatda0fzux0WY3AEEzGO_EdVi6wp6TqJjYY3Z_i51y1cLVEuSxiAACN2cAiQ=s48-c-k-c0x00ffffff-no-rj)
Vložit
- čas přidán 30. 12. 2019
- I am an adult now, I can finally do the thing I always wanted to do, and it cost me five dollars.
A couple notes about this: Keep in mind that I couldn't exactly reshoot, for obvious reasons, so that's why the narration or pacing might feel off. I also don't think this result is typical of any modern hard drive, but I wasn't going to destroy a 2TB just to figure out what that sounds like.
Support my channel:
/ cathoderaydude
ko-fi.com/cathoderaydude - Věda a technologie
![Learning to run a broadcast in ONE HOUR [Tricaster TCXD850]](http://i.ytimg.com/vi/oZinM9XFoIg/mqdefault.jpg)
hard drives can have a little air. as a treat.
Ha ha ha! That was the best thing I've read in a while.
My poor lungs. You made me laugh so hard I started a coughing fit. Many thanks.
But they can’t smoke or vape!
can cats have salami
Can a crackhead smoke crack?
"Would this break, if I dropped it?"
I don't know what I expected . meme
that ending was amazing xD
Clatter Platter! :P
I have had drives that I have accidentally dropped that have worked for years after I dropped it. Recently my foot got entangled in computer cords that caused the computer to fall over on it's side and slide about 3 feet on the floor. The computer froze up. I power cycled the computer and it booted right up. Still running.
Rev George
Love it; great experiment and demonstration.
I used to work as a HDD design engineer (I think I worked on that specific drive model you're experimenting with, even), so I could probably talk your ear off about a lot of those subtleties you're playing with, even without getting into any proprietary stuff. They're impressive machines, especially when you consider how cheap they are to buy.
The airflow in the boundary layer on the disk surface helps keep particulates from landing on the spinning disk, and as that recirculates in the (closed) drive it is blown through that white filter pillow to capture any that are present -- next to the actuator magnets in the drive you had open at the beginning, but in other drives sometimes tucked in the other corners around the circumference. Keeping the drive spinning during this test probably helped a lot to keep the surface clean.
The actuator positioning control loop is tuned to expect a specific amount (range) of drag from the preloaded bearings, which is why it failed to work when the preload (tension) was too low or too high. You might be able to cut off just that corner of the cover with a hacksaw or Dremel and leave the rest exposed.
oh! wow!! it's fantastic to hear from an expert!
So then my half-baked theory that the drive was forming "a little bubble of air" was probably correct, then; had I shut it off, the results may have been very different - now I want to do round two, get some rigorous data!
Do you happen to recall what the tolerance on the head spindle bolt torque was? It wasn't terribly hard to remove so I'm guessing it's just a few inch-pounds, but I'm curious how *close* it needs to be to the specific value before the actuator gives up the ghost.
You're absolutely right though - what's amazing is how cheap they are and, as I said, how tolerant they are. I can't *believe* that these devices, which are more precise than almost anything else I've touched in my life, can just be shipped, run in rooms full of chalk dust, next to blast furnaces and everything else, and just keep truckin'.
Thank you again for commenting, I really appreciate it!
@@CathodeRayDude The precision on that torque requirement is probably ±10%, which shouldn't be too hard to achieve. Though I think it's really a tension requirement in that shaft, to preload the bearings as you said, so e.g. if you cut away part of the cover and reduced its stiffness, the nominal torque requirement may change.
At the level of this kind of experiment, doing it by feel is probably OK. The positioning servo loop is pretty robust, and it won't hold a grudge against you for abusing it. (There are all kinds of reasons it might intermittently fail in operation, so it needs to be able to recover gracefully: vibration, shock, temperature to low, temperature too high, someone squeezing the cover, ...)
well you're super awesome. :D
This kind of comment makes reading 1000 crappy youtube comments worth it. Thank you.
I'm not a drive engineer, so take this with a pinch of salt: I have read that the vents on the exterior of a drive are not for significant air flow but rather are for just the tiny amounts of air required to maintain pressure equality between the interior and exterior, for example, as it's moved to different altitudes.
One of my favorite things about (some/most?) computer hardware is if you just, interrupt it and then let it keep going, it just picks up where it left off. Like lifting a hamster off its hamster wheel, but the hamster's still kicking his feet as if he's running, and then setting him back down on the wheel.
I know. Even on 90s machines. You can pull the hdd out of the system and put it back in. After the drive initialized it just continued xD
@@FennecTECH You know, this still happens if you boot Windows directly from a USB drive installation, and then unplug the drive. You get a grace period of about 10 seconds of the system hanging, and if you plug the drive back in within that window of time, the system keeps going as if nothing happened. Only if you keep it disconnected does it eventually give up and throw a BSOD.
@@3lH4ck3rC0mf0r7😂
God the last line was a laugh-out-loud moment, ahaha
i had more stuff planned and then a fadeout but that moment was so funny (I had no idea what was going to happen) that I knew i had to capitalize on it
@@CathodeRayDude i see you've mastered the art of comedic timing as well. you are indeed a man of many talents.
my takeaway from this is that DIYing data recovery by swapping platters with my bare hands is a perfectly fine idea with no issues whatsoever
two days ago i would have chuckled at this obviously absurd idea but now I have no convictions or beliefs left in me
Don't touch the platters! The oil can warp the magnetic signal being read by the heads, or make drive head interference!
Your best option for data recovery is buy a donor drive with the same firmware version, swap the circuit boards.
now i wish i didnt chuck an old external hard drive of mine. i wouldnt say much of anything valuable was lost, but, still stuff i would have liked to archive to some degree.
@@fully_retractable eh, most modern drives have data unique to the drive on the PCB, so you'd need to transfer it over somehow
My first hard drive, from when I was a kid, was a 540MB drive. I ran DOS, and I learned how to program on the machine hosting that drive. When I upgraded to an what seemed an inexhaustible 6GB drive, I decided to open my old drive. It looked very cool, so it hanged, open, on my wall for many years.
Fast forward about two decades, and I was throwing away old junk, and found my old 540MB drive in a drawer somewhere. It must have set there along with nails and screwdrivers and other junk for at least a decade, after being displayed on the wall for the best part of a decade. Just for kicks I decided to plug it back in, to see what would happen. Of course, I had to buy and IDEUSB adapter to do this. I didn't expect much, mostly just some head moving. I certaintly didn't expect to be able to read the data off it, but to much of my surprise, I was able to do just that!
That's remarkable! Wow!
Maybe those old drives had super low tolerances
I've got to say, as someone working in data recovery, this is both one of the most interesting and hard to watch videos I've seen in a while. Hard drives are at the same time so fragile, yet I've seen many drives well older than I am spin up without issue. Cool stuff.
It's a complete crapshoot. Recently, I plugged in a couple-hundred-meg Seagate from the late 80s that used to be in a Mac, and it immediately killed the power supply in my G4 mule. I pulled off the logic board, dug around and found two shorted chip capacitors, which I took out of circuit. Before I buttoned it back up, I noticed the spindle flywheel was exposed, so I gave it a twist; no dice.
I reefed on it pretty hard and it broke loose with a sickening CRACK. The spindle went around 350 degrees, then stopped dead, grinding through the spot it had been frozen at. I forced it through, then gave it another rotation, and this time it didn't resist so much; another rotation, and it spun free. I plugged it back in, it came right up to speed, and I got the file index. Everything loaded fine.
These things are black magic. Hats off to everyone who worked on them.
I didn't see a good place to expand on this in the video, but:
The most interesting takeaway from this experiment, to me, was that drive just kept trying instead of giving up. My impression of these devices was as highly stateful and intelligent machines, and I assumed that after a certain number of critical failures they would stop trying completely, and go into a coma to avoid further damage.
I also assumed that very tiny alterations in their environment would result in self-protection shutdown - when I took the cover off I suspected that the change in air temp, etc. might cause the head to not fly as high, for instance, and it would notice that and trigger an error state, or that if the head positioning coil went over-current it would trigger an error state, and either one would cause a shutdown.
This is, obviously, all entirely false. There really can't be failures worse than what I caused, and, in retrospect, I realize that my assumption wasn't supported by anything I'd seen in my life - hard drives with extreme damage, like a head crash so bad the head wears a visible track into the platter, always continue trying to operate; that's why clicking is such an expected behavior from dead drives, it's the disk trying over and over to start up despite being damaged beyond any possible recovery.
This almost "analog" behavior has changed how I characterize hard drives in my head entirely, from highly sophisticated devices, to machines that know how to do a couple things very well and have little decisionmaking ability beyond that.
As a design consideration, when extrapolating from some limited sensing about the state of your system and its environment, it's often better to avoid adding excessive complexity and prevent false-positive "fail-safe" scenarios -- because those can produce a terrible end-user experience. There are often a number of different conditions that can produce the same sensor values, and keeping track of which regions of this enormous state-space are benign or transient, and which ones are actually critical can be a full-time job, and not worth the effort for "consumer-grade" applications.
Nice job! Really interesting. But I wondered if you could just cut the portion of cover over the screw and tighten it again but just with that little piece from the cover over it to properly tension the heads. Could it work??
@@hugofco2037 I thought about a lot of ways to do that and none seemed practical. It's pretty heavy gauge steel, so I would have had to take it off, go to another room, cut it down with an abrasive wheel and bring it back. This would be incredibly messy and produce a ragged result, which would not only leave the drive open to the world for half an hour while I did it, but I could never adequately clean the resulting object enough to be sure that it wouldn't deposit metal shavings onto the disk surface, which I'm sure *would* kill it.
I also believe the tension would only work if the resulting sheet spanned all the way to the sides of the drive, and possibly only if it spanned all four. So I decided not to try it, but maybe someday I will since I have a better, cleaner metalcutting implement now (drill-powered nibbler.)
Thanks for watching!
Remember: machines have no value for life. They simply continue to do what they're told until they physically cannot, and even then will still try. Watch your fingers, and stay out of the splash zone.
@@CathodeRayDude Well, I just realized I'm late to the show, but besides that. you could get 2 identical hard drives & pre-make a cover for your test drive. great video & amazing ending. I couldn't believe how abruptly it ended. I feel sorry for anyone who didn't watch until the end. ( FYI I had the same thought on how you'd be able to do it. until I thought of a donor drive.)
So glad Technologyconnections reccomended your channel. You guys seem so in sync and I love watching thing I know won't ever affect my daily life lol.
They're brothers
I dunno, he is not pedantic like Alec. I can watch his videos without get bored with his arrogant way of deliver information, CRD is way more down to earth 👍
@@michaellandon1960 legit??
@@JonGallon that's hamming it up, not arrogance. He is rather pedantic tho.
i window modded a hard drive, it worked for a good while after that just not... good.
(window made from a plastic container and clean room made of tape and two plastic shoppingbags)
The good old days of watching Computer Chronicles.
Would've been interesting to see you actually put a fingerprint on the platter and then tried to use it.
haha that is what I was expecting for the grand finale
the ending, i laughed, loudly
How does this man have only 1k followers this is quality content
A few more now! But man I had forgotten. The name of the channel and was on a new CZcams account and I tried finding it by putting in keywords of videos I knew hr had done and I couldn't find it!! It was only weeks later when I randomly remembered the name of the channel that I could find it. I think maybe he isn't good about setting the keywords when uploading videos and is getting crappy discovery.
Have to say man, Loving the content please keep it up. I foresee the CZcams algorithm blessing you at some point next year.
thank you so much, I'm gonna put my back into it!
@@CathodeRayDude hey man, the CZcams algorithms doing its job. Im loving your content.
@@CathodeRayDude The youtube algorithm blessed you at some point this year
Algorithm crew; checking in!
nevek20 was on point.
true
Laminar flow hoods are highly effective. It's a clean room for your hands.
Ah - I had heard of laminar flow hoods being used *in* a clean room, which suggested that they weren't sufficient on their own, and had otherwise heard of them being used mostly in the context of mid-level precision/cleanliness, like optics work. Do you know if they are actually considered equivalent to a given clean room classification?
@@CathodeRayDude I have only researched them for mycology/cloning, most people doing that stuff at home find low-end hoods sufficient for their needs, but as I've browsed the cheapo end of the spectrum I have glanced at some expensive/presumably highly effective examples, probably approaching an actual low level clean room.
This was fabulous and delightful. Thank you for putting this up. Looking forward to a future teardown!
I'm eleven months late, but dang this is exactly what I'm talking about. Very interesting content and fascinating input from an actual HDD engineer
came across your channel via your corridor 8 video and made my way to this video pretty quickly - this is fantastic. i deeply appreciate the hobbyist curiosity and passion for your subjects! please keep having fun with what you do and take it at your own pace so you don't turn a love into a joyless responsibility :) see you whenever you post next!
thank you so much!!
Man, you're my new favourite channel. The detail, delivery, humour and topic choice are just consistently brilliant.
I so enjoy your videos! They're a breath of fresh air compared to your standard smug hyper-kinetic editing video thingy! Good work!
Kinda disappointed you didn't do a fingerprint test on that bad boy, but it was really surprising to see how much abuse it was actually able to take
Love this channel already, don't change the basic layout man, once you become too big and mainstream it's not like relaxing like home anymore especially with big flashy loud videos *ahem* Linustt *ahem*.
Thanks man - Yeah, I'm pretty insistent on going with a pretty plain production style and keeping the flash to a minimum.
Dude your channel is amazing! I’ve been binging
I love that you've got Computer Chronicles playing in the background. It's one of my favorite shows.
I opened up old hard drives and played around with them back in the early 00's. I was amazed at how long they kept on working without a cover on them at all after hearing all the tales about how fragile they are. I think I had one spinning along happily for a week before I got bored with it. Obviously it would eventually die and I would never trust any data on it, but it was neat to watch it work.
OK, I have only started watching you recently. I have to say that I am enjoying your videos, and I found this one fascinating to watch as I have been curious about how well hard drives work when opened. But when you mentioned your love of Columbo, that told me that subscribing to your CZcams channel was the right thing to do.
I just want to say that ive discovered your channel yesterday when linked something about fax machines and have watched at least 3 hours of your videos since. I'm a little tired and a little hungover and each of your videos puts a smile on my face. The quality of your content and you yourself as a host are fantastic and on par with much more professional youtube shows with much higher budget shows, like scishow, etc. I hope you have as much fun making these as i have watching them
😊😊
😊😊
😊😊😊🎉😊😂
2:😂45 😊😊😂😊 2:48 😮 2:49
🎉😅😂😮😂 2:57 😮 2:57 🎉
hi, watched your new years video where you asked yourselfs what you enjoy doing. This video was shown to me again and it was the first video I saw on your channel. I love this because it dabbles in something I asked myself for years. I would watch more videos of you, breaking down old "this will kill hardware" dogmas.
Wish you a great next year and try to remember not everything you see is real and not everything you don't see is unobtainable (except bills, they are usually very real). Have a good weekend.
Many mainframe hard drives had removable disk volumes, because the media was less expensive than the drive. A typical volume had 6 to 11 platters the diameter of a large pizza, stacked four to six inches high, separated by about 3/4 inch on the spindle. Outside the drive they were stored in a two piece “cake caddy” cover, with a handle in the indented clear top piece that turned an inner spindle, to which the opaque bottom piece was screwed. To mount the pack, the operator opened the lid of the drive (which would only open when its spindle was completely stopped), removed the disk already mounted, lowered the disk by the handle on the top cover, turned the handle clockwise until it stopped, releasing the disk pack from its cover and locking it to the spindle, then closed the drive lid and started the drive. When it was up to speed, the heads (5 for 10 tracks per cylinder, 10 for 20), parked under cover, extended between the disks. To remove a pack, the process was reversed: stop the drive (which would retract the heads before turning the motor off), open the lid, lower the top cover onto the pack, turn the handle counterclockwise until it stopped, attaching the pack to the top cover and releasing it from the drive, then lift by the handle, screw the bottom of the cover on with the other hand, and put it away.
Even then, computer rooms had to be semi-clean, particularly no smoking! When storage densities got larger, the manufacturers started using sealed HDA (head-drive assembly) volumes, replaceable only by service personnel. Fortunately, disk and tape drive data speeds by that time allowed entire HDA volumes to be backed up and restored in a few minutes. Critical data was maintained simultaneously on multiple drives on multiple channel attachments.
The storage capacity of those huge disk packs, incidentally, was in the hundreds of megabytes. Later they reached gigabytes with the non-removable HDAs, which were part of the drive.
I love this channel, glad I was sent here, now I learnt a new thing.
Well researched, well exposed and pleasant to watch. Your channel will grow, and you deserve it.
Keep up the good work.
Sorry for my broken english.
Ciao from italy.
Gotta say, you are a criminally underrated youtuber. Keep up the great content, I love your work
I'd call myself very positively rated, just among a smaller audience than I'd like. Thank you!
Fantastic, I especially love your reactions, I was pretty much matching them
Don't know how I missed this vid till just now. Love it. The ending was amazing.
This brought back memories. For a class in high school I attached all the parts from a computer to a board to display on the wall and also removed the top of the hdd. The computer ran for weeks like that before it finally started to BSoD. Granted this was maybe a 2GB drive at most so not nearly as dense, but didn't expect it to survive long at all.
That ending actually killed me, i was not expecting it to just end after you dropped it
Please keep making videos your descriptions and clarity ar beautiful
back in the late 90s at the shop i worked at, we pulled a hard drive apart, used standoffs on each corner to put it back together. then clear tape around the outside left us with a bench drive with a window all around it to look in and see the heads moving. we used it like that for a couple of years on our test machine out front. cool display piece, people always asked about it.
i guess these days getting an acryllic cover made would be the better way to go
Awesome video, awesome ending. Jajaja I still use my 32mb Seagate ST-238r from the 80's. You've got another subscriber. Greetings from Argentina!
As I understood it, the vents with filters aren't to exchange air. They are present so there won't be a pressure difference between the inside and outside. Like, when the drive gets hot, the air expands. You don't want pressure building up in there. So a little hot air escapes through the filtered vent. When it cools back down, you don't want a slight negative pressure, so some air goes in through the filtered vent.
I seem to remember in the distant past, some just had a single pinhole in the lid, possibly with no filter. They were marked with something like "DO NOT COVER THIS HOLE". I know more modern IDE and SATA drives had filters under them. These older drives would have been like 1980s/1990s era hard drives, that I took apart in the late 1990s and early 2000s. Like MFM, RLL, and early IDE. They may have had a single layer filter that looked like a paper sticker. It just needed to relieve pressure, not any large volume of air.
I don't remember handling many SCSI drives from that era. Or at least none that I took apart. Those were serious investments by the company, and they didn't allow for that kind of playing.
That ending was hilarious! Keep up the great content!
Your videos are awesome! Thank you
I was expecting this, really well done and your nice earned a sub
You get a thumbs-up for taking the test to the extreme!! Good job!
Okay this is just amazing! Great video!
I opened an old hard drive and then spun it up with a takeaway box lid over the top. It worked fine for the duration of the experiment.
If I remember right that was a 0.3GB or 0.5GB hard drive.
This is a fun little video, I'm glad the algorithm recommended it to me!
Probably one of the strongest video endings ever uploaded to youtube
Love the Computer Chronicles on in the background. Nice touch.
That was cool! Should've tried some less drastic tests at the end there, like instead of just dropping the spinning drive, could try these first next time:
- Breathing on platters
- Coughing on platters
- Touch platters with clean q-tip
- Touch platters with alcohol on q-tip
- Touch platters with finger wearing rubber glove
- Touch platter with bare finger
- Remove and replace platters &/or head with bare hands
- Yell at hard drive
- Taunt hard drive
- Make cat's paw touch platters
- Scratch platters with fingernail
- Scratch platters with key/screwdriver
these are very good, especially kitty paws, but I felt the video was getting long in the tooth and didn't want to get harder into it - however, I may do a second episode at some point with more experiments
This video with the screw torque changing the head movement may have helped me solve a problem on recovering data off a damaged hard drive i have. Thanks!
I never thought of that until now i need to find a doner hard drive with good heads and try again
Really enjoyed it. Great video. You got my sub mate
This reminds me, I had this one hard drive that failed in 2018 (an absolute chungus unit of a Western Digital drive, I don't know when it was made but seemingly it was in a repair shop in 2013). It did this thing where it would buzz and churn and it just did NOT sound happy. At a random point in 2021 I decided to try recovering the files for the umpteenth time since it failed and, with my own two eyes, I witnessed the hard drive fall off of my bed from a roughly 3 foot drop and BOUNCE once before it lay on the ground motionless. I thought it was absolutely DONE FOR after that.
In some sort of godly 1 in a million miracle, it seems like dropping it actually FIXED IT, with the hard drive hooked up it started displaying errors related to missing windows stuff on boot, I used Hirens to recover and I had some more ridiculous hoops involving recovering the files again back to the hard drive, but I totally saved my 2015- mid 2018 legacy by pure luck and even now I still have no idea what "fixed" the hard drive.. after watching this vid I'm dying to know if anybody has had a similar experience to me in this regard
Cheifet and Kildall in the background talking about early winchesters is the icing on the cake 🤌🏻
The screwdriver on the lid gives me big "Demon Core incident" energy xD
Getting real demon core vibes from you wedging open the drive with a screw driver :D
Cathode Ray Dude, you my only friend. Solid work!
I am so glad I found this channel! Your technical story telling has me binge watching your entire catalogue.
Absolutely perfect ending!
Thank you for your videos 🤘
The sudden end of the video is just *chef's kiss*.
Wow, all that and a head crash killed it. Wonderful video, you've got yourself a new subscriber
Also, if you do this again, I'd like to suggest badblocks for testing as well
Love the ending. Simple, perfect.
Computer Chronicles on the side, makes for a cozier vibe than any fireplace would
finally, a nostalgia-lounge / fun-lab that looks like the one i play in every day. totally smitten with the concurrent episode of computer chronicles mirroring the topic. cathode ray dude forever!
I put a plexi window into an old Toshiba drive, turned the shower on in the bathroom to as hot as it would go, then let the air settle for a bit. Opened it up in there and placed the drive under a clean dish once the cover was off. Ran that drive for months with no issues.
I’ve opened and re assembled multiple hard drives before and I haven’t had any issues with dust causing them to break.
It’s incredibly common for quantum fire ball hard drives to have their drive head mechanism to get stuck to the rubber bumpers. Repairing them is quit easy and I’ve done a number of them and they all work fine.
Great video, thanks!
When I saw the picture of that Seagate ST-225 I laughed out loud! I had a computer at one time that had that same 20 MB drive. One day a lightning strike occurred in the courtyard next to the apartment building where I lived. The PC seemed to work OK, but later found out that the 12 volt output on the power supply was a little weak. When turning my computer on, I had to "jump start" that massive ST-225 with a 12 volt bench power supply!
I am.glad I am not the only one amazed at the relatively old technology and reliability of mechanical hard drives.
I once opened up a Seagate 1TB 2.5" hard drive because I have read that Seagates often get a stuck motor bearing and you can unstuck it in a clean room. It didn't make a spinning sound, just beeping so I thought this is the cause. I had a shower, so that the air in the bathroom gets cleaned from dust a bit by the humidity, immediately after and I opened the disc up. The motor turned out to be actually stuck, so I helped it around with scissors (yes, scissors as my pliers didn't fit) inserted into the screw heads, screwed it together and connected it to my computer. Lo and behold, it worked just fine so I immediately copied everything off it and haven't used it much since. I cecked it now about 2 years later and it doesn't work (spins, so it's not the motor). So most importantly, always make backups. Secondly, if your data is somehow worth thousands of dollars and you didn't back it up for some unfathomable reason despite that, go to a data recovery company. But if it isn't it is either already gone (if the head itself is bad or surface damage happened), or you can give it a shot by opening the disc and seeing what happens. You might fix a stuck motor or head this way for free, or ruin the disc more and make the expensive data recovery even more infeasible. Just don't turn off that disc until you've copied everything off, and don't use it for any new data as it might crap out soon.
The finger print thing is useful because so many people do have the gut instinct that fingerprints are flat. In reality they can be rough enough to feel through a thin glove. So, if you are transferring platters between hard drives it is very important to clean any finger prints off the platters before spinning them up. Really, if you can see the contaminant with your bare eye it is probably big enough to crash into the head, which is a problem when the disk is spinning.
Your test, oddly enough, did succeed in a small part because the platters spent most of their time spinning. Dust cannot settle on the spinning disk because of the air flow crated by the disk's skin friction. Air born contaminants large enough to be dangerous (dust) are usually safe for this reason unless they are made of something that will stick to the disk and hit while the disk is stationary or hit with high enough velocity to penetrate that air barrier.
The cover provides a stable mounting point for the spindle screw. You could've cut the bottom half (or so) of the cover off leaving the top with the upper four screws to secure it in place (and hence still be able to put the spindle screw in) and had all kinds of area open for stale air while still allowing the drive to run.
I've done platter replacements twice in my office, both times worked (at least well enough to recover data to a new drive)
I have HDD 30 years old that still work. I have 4 drives that have been running continuously for 20 years. I have had 10 SSDs fail while in warranty. IMHO SSD's aren't ready for prime time.
Recently a friend brought me two computers that he wanted to get the data off from them with MFM IDE drives from the mid 1990's in them. One of them would not spin up. It made a lot of noise. I powered it down and opened it up. Manually spinned the platters and placed the cover back on. It ran like new was able to create an image and recover all of the files off it.
Rev George
OMG, I love your tangents.
in the late 90s we used to take the hard disks apart and put a window in them. Air must have been a lot cleaner back then because they worked fine afterwards.
One thing to note here is that the drive also contains its own internal filtering, if you look at one with the cover off you'll see the case is actually shaped so that the platter itself works as an impeller, to drive air through a small filter. That way anything approaching the platter rapidly gets pushed away and captured before it can recirculate.
I actually had a sampler's SCSI drive start to do the click of death, so out of curiosity I decided to open it up and power it up with the cover off on the bench. To my surprise the head suddenly started working properly after messing around with it for a bit, so I put the cover back, put it into service, and it's been fine ever since. I think people exaggerate how dangerous particulates are for hard disks - sure they'll kill it eventually, but it's a long-term probability thing, not an instantaneous effect.
Remarkable! Though I assume that was quite an old drive, so maybe the precision has gone up - but I was hearing that "single speck of dust" thing in the late 90s, so yeah, I suspect that it's exaggerated. Consider: if the drive is spinning, unless that dust is INCREDIBLY sticky, centrifugal force is going to fling it away, right? So let's stack up the probabilities:
1) dust particle approaches drive to begin with, instead of brownian-ing itself in a different direction, and
2) dust particle isn't pushed away by spinning disk, somehow (extremely unlikely), or
3) dust particle settles on disk that isn't running, and
4) disk manages to spin up *without* throwing particle off either immediately, or in the potentially many minutes before it has reason to read the spot where the particle is
Assuming that dust *isn't* so sticky that it could cling on to a 5400, 7200 or 10,000rpm disc, it actually sounds incredibly unlikely to me that this failure scenario could ever happen.
I imagine dust could probably get sucked into the boundary layer of the disk if it approaches from the hub, and get pulled under the head, alternatively dust settling on the platter with the disk powered down could probably adhere itself pretty well, just think how hard it is to blow dust off things that have been on a shelf for a while. As with anything involving hard disks though, it's all about getting that MTBF as high as possible - even a tiny probability becomes significant over a long enough period, and when the outcome is essentially device destruction you don't want to take any chances!
Good content, Keep it up!
16 seconds into this video as a first glimpse of your channel and i am SOLD. Especially after realizing your channel name lmao
For voltage to kill a person, it needs a current path. If you're touching voltage alone, and don't have a return current path, it won't harm you at all. So just touching a wire, if you don't have a return current path is harmless. But if, for example, you're holding a grounded piece of metal, or touching a ground wire when your other hand touches touches 120V, you will definitely feel it.
The other piece is that to harm you, the voltage you come in contact with needs to be sufficient to actually cause damage. If you are touching a line that is at, say, 1000V, but there is a large resistance between you and the voltage source, then allot of that voltage will be dumped across the resistor.
Your body has resistance, and if it is the amount of current that goes through your body that can really harm you. For one, it will interrupt the heart rhythm, which is controlled by certain cells that generate the pulses that make your heart operate. Electricity through your heart can disrupt the cycle. This is also why precise electrical pulses from a defibrillator can be used to reset this rhythm at times.
Lastly, 120V isn't considered high voltage. Using the terms 'Low voltage" or "High voltage" to determine safety can get you into allot of trouble. Generally, anything below 600V is considered low voltage. If you look at the insulation on the wiring in your home, it will be rated for 600V max - this is low voltage wiring. Medium voltage goes up to about 69,000V. Distribution lines at high or extra high voltage can reach more than 1,000,000 Volts.
Interestingly, you'll see birds sitting on these lines that are at thousands or even a million volts - and they are unharmed - this is because there is no current path. On the other hand, you will see squirrels trying to transition from the pole to a line, and the resultant arc can take down the power to city blocks - this is because they formed a bridge between the high voltage and ground.
So, please, don't assume that something marked "low voltage" is necessarily safe. While our outlets are technically "low voltage", they are more than capable of causing injury or death. Even the 12V in your car - saying you are wearing a metal watch, and go between the positive battery terminal and ground - is more than sufficient to cause horrible and permanently disfiguring wounds. I've seen accidents with 5V power supplies that caused fairly significant fires. It's really worth respecting electricity.
To me it makes sense that a fingerprint has substance, they are made of water, oils, amino acids, salt, and so on. All the things present in sweat and on the skins surface. Maybe you think of them as flat because they likely absorb into other materials. On a solid liquid-proof surface it makes sense they would stay on top.
Still though it's fun to think such a thin layer of mostly liquid can disrupt a HDD. Apparently it's measured in nanometers (as of 2011), the same size as features in modern CPUs (2024).
The airflow from the spinning players mostly prevents any dust from settling, turn it of, open it and try again after a few hours. Read errors will start showing up. But still possible to run it open for quite a while before it dies.
The LTT durability test = Dropping it.
Damn I laughed so hard at that ending of the video.
the space between a hard drive read head and the platter is about 0.0002" iirc, which is indeed much smaller than hair or smoke particles, but those won't immediately destroy a hard drive. think of a hard drive like a record player. if you play records with dust all over them, you'll damage the needle and the record, and eventually both will sound horrible.
i have definitely seen hard drives being run with the cover off for several years, but i wouldn't recommend it. they're resilient in some ways, very fragile in others. one method of hard drive data recovery is called "the toaster method", and involves taking the board off the hard drive, putting the hard drive over a toaster for maybe ten seconds, and then put it all back together and spin it up to see how much data you can grab before the platter shrinks out of alignment again.
It would be so cool if one day, we could stick a tiny camera or two and a tiny light inside one of these, and record the hard drive working. Maybe it's already been done. But if not, I would love to see that.
When i was younger i had a job destroying hard drives, old hard drives filled with concrete so the best way was with a sledge hammer. I'd put put the drive on-top of two concrete blocks with a gap in the middle then smack the drive in the middle with the edge of the sledge hammer, if you did it right the whole drive folded in on itself and shattered the plates into dust. €2 a drive was a damn good pay and i now have a certain respect for how tough hard drives are.
I had a job destroying hard drives as well for a while; we used an arbor press with a sharpened punch, and boy howdy did I ever find out how many hard drives have glass platters. Punch went through like butter though! Technically a sufficiently motivated attacker could have recovered some data from them but hey, we weren't doing CIA grade stuff.
@@CathodeRayDude I tried using a bench press at the time but it was too slow so the sledge was a good option, if i was having a good day i could get through about 50 drives in half an hour but it was certainly not easy work. If you missed the center of the drive it would hurt like all hell right up to your shoulder.
Quick note, the holes on hard drives aren't exactly for airflow. They're breather holes, so that pressure variances caused by temperature and elevation are equalized with the atmosphere. There are drives that have a channel, that runs from near the head actuator along the side, with a filter in it to catch particles as air moves through, but most of that airflow is at the edges of the platters.
Also, is the program you wrote available anywhere?
Fantastic content, and amazing ending. Would love to see you do a fingerprint test to see how quickly it becomes a problem.
Use a washer that's the same thickness of the lid.
They're harder to get catastrophic failure than you're thinking. The seal and filter help with longevity and reliability, but I've transplanted platters into a new hard drive for data recovery, using nothing but rubber gloves to keep things clean.
I don't know what voodoo the do today, but I have seen platter swaps work with someone just opening the drive in their living room and swapping them out, and the drives still kicking according to that friend. Crazy, wonderful, tech filled times, we are in!
when I was in school, I ran few hundred gigabytes sized HDD in a bucket of water to see if it would survive the water and the surrounding dirty environment. It ran for a while and I could copy and paste data, so it wasn’t tooo detrimental.
After while, though it’s stopped working, but it’s a far cry from instant death as we’re led to believe.
Old old drives would use the air vortex to position the platters and it would lead to more faults. I was just watching this old computer chronicles video from 85 i think it was about the new exciting hard drives at the time that were phasing those out. They'd have little filters and breathing holes tho to get the vortex clean.
My first-and-only MP3 player had a 1.5 G hard disk.
When it finally died, I took it apart and was disappointed that at that scale, it was basically a polished metal washer and little more.
Fun video, got the like and subscribe from me and I'm sure you already know this stuff but just want to point out:
1) Running the hard drive with the platter fully exposed is possible but can be VERY dangerous/deadly in the event of catastrophic failure should the platter shatter.
2) Running the hard drive exposed to the air will work for a while, but I promise, it will not last NEARLY as long as a hard drive that has never been opened. Surely, you have seen the inside of a computer case after a few months of use... Try running that hard drive with even just the corner open for a few months to a year and I can guarantee you will have problems as the dust, hair, etc... began to collect on the surface of the platter.
I swapped the platters from my dead 1TB drive onto a working one (same model) just to recover the data from it which I did on my work bench inside my not so clean garage. A data recovery in my area said it was gonna cost me approx $2.5k to get it done cause there was a scratch on the top most platter in which I told them to forget it and send the drive back to me. I only wanted to recover some of the data not all of it and I was able to do it myself. Total cost I spent was $30 for a used working same model drive and filled slots onto a piece of acrylic I had laying around to use to keep the head comb separated.
I once had a 20GB WD EIDE that had a frozen head that somehow got laminated to the platter. I popped the cover and freed the head, put it back together and it ran for years in a small FTP server as the system disk. It was still running when it was retired and recycled. These things are fragile yet built like tanks.
I took the cover of my hard drive off and did this too, and it took a full 20 minutes of use before I stated having issues, and a full 2 hours before it started clicking and failing more often than working. If you started it from stopped, you'd probably see it fail faster. The Aerodynamics of the disk spinning create an air bearing.
I have a video on facebook where i am writing data from my gaming rig to a hard drive that had a lot of bad sectors, so i made a partition on it and i started the copy, while it was copying around 5gb of data, you could see the head writing some data, go back in the center of the disk (i think updating the file allocation table because she was formatted as fat32), and then it would go back to where it whas, writing some more data, and go back to the center again. It was awesome to see, i show my screen that i am in fact writing to that drive. It whas a ide laptop hard drive, the drive still works to this day. Maby i will make another video for youtube with that same drive
To expose it to atmosphere, you could have just peeled the little seal off the side. It's the access hole for servowrite and dc-erase heads to reach the disk during manufacture