Two failed Hard Drives, let's discover why

[@skjerk]

On the hardcard, one problem with your test could be that the lid is off. This probably prevents the LED-Glass-Detector circuit to work because of the ambient light going into it.

[@tony359]

Yes someone pointed that out, I didn’t think about that!!

[@argoneum]

Very likely, seen that happening with Commodore's 1571 drive also.

[@tony359]

But the 1571 is a floppy drive?

[@argoneum]

@@tony359 Yes, had to go down memory lane to remember exactly: the drive would fail to detect that a disk was swapped if incandescent lamp was shining on a write-protection sensor. Covering it with even a piece of paper made the drive work properly again. Made me wonder, thought I broke it 😁

[@tony359]

ah of course, IR sensor. What was wrong with the good old microswitches :)

[@bentboybbz]

Im fairly sure you want air to be drawn up past you in an updraft configuration in that scenario to help keep anything from falling from you into the drive such as hair etcetera. Not blown which tends to pick up debis and push it towards your project but drawn or sucked like a fume hood. Thank you for your time and effort!

[@tony359]

Well, a proper ‘clean air hood’ will create an airflow from the back of the desk to the front so any debris or hair or anything is pushed off the drive. My approach wanted to do a similar thing with the limitation you spotted: small hole causing turbulence around the drive and not being 100% effective. If I reverted the flow, then the filter wouldn’t do anything (well it would filter air in the room!) and contaminants would be able to fall on the platters with no restriction. Remember it’s not hair causing damage but minute, invisible dust particles which exist in the air. Even smoke particles would be too big. This is why a good filter is paramount. Thanks for your comment!

[@bentboybbz]

@tony359 Thank you for your reply! If someone designed an effective relatively affordable solution that would enable someone to turn a small room into a reasonably "clean room" it would be awesome, giving individuals the opportunities and fair chance of success could allow for repairs and manufacturing on a whole new level...if any of that makes any sense at all...I apologize im walking through my neighborhood, which my city has a very high murhder rate, im in one of the worst neighborhoods, at night, and I need to keep my head on a swivel, some guys were just yelling rude things obviously displeased with the color of my skin, and they are now walking behind me, one of the ladies for sale so to speak just warned the one yelling at me, regardless Thank you for your kind reply! I hope you are having a great day or night!

[@kpanic23]

IIRC hard drive platters have a thin PTFE film over the oxide layer to reduce friction and adhesion when the head touches the surface. If the drives have been sitting for decades, the minute weight of the head slowly squeezes the film out, causing these imprints on the surface, as seen with the hardcard. This is a huge issue with autoparking drives, because the head will always rest on the same track and will also always come to a stop in the "hole" in the PTFE film where friction is the highest. The missing coating also causes high adhesion due to the immense smoothness of the platters and heads, which might be strong enough to keep the drive from spinning up, also knows as "sticky heads".

[@tony359]

Very interesting insight indeed! Thank you!

[@Birdman_in_CLE]

I have a 20MB hard card as well. Mine had a head that fell off and the arm turned into a lathe tool and shaved several deep rings into the platter. It is still amazing to me that these delicate devices can run for years or even decades in some cases. Fantastic engineering and precision manufacturing.

[@tony359]

It is - particularly if you think of when those were engineered! Thanks for watching!

[@williefleete]

Ambient light affecting the servo sensor?

[@tony359]

Probably, I didn’t think about that!!

[@LynxCarpathica]

Hi Tony! Regarding the 2nd HDD: Theese inteligent hard drives (incl. ST-157A), wait for the first command from the controller to initialize. Second, either as you said, the rubber stopper controlled where Track 0 (or the last one) is, and used that to calibrate, or the slow sweeping would explain another, looking for a marker track to initialize with. ST251 ans 225s use that too. If you hear an ST-225 or 251 banging the heads at start, it means it can not calibrate by itself. So it zeroes by headbanging, and calibrates that way anyway.

[@tony359]

Very interesting. Very possible indeed. Thanks for watching!

[@Epictronics1]

It's sad to set yet another PS/2 drive having that weird corrosion on the platters. Those drives sound very cool when they work. Entertaining video as usual!

[@tony359]

Thanks! yes, I think they are all doomed. I'd like to tap into the stepper mechanism and make it move (no heads attached) in sync with the XT-IDE's LED activity. Maybe one day...

[@Epictronics1]

@@tony359 That would be a very cool project! Good luck!

[@grbloopers]

If there was any chance for it to work, you killed it. That plastic disk you broke was creating enough air pressure to blow under the head channel, lift it, and fly them close enough, but over the platter. After the removal, heads don't fly anymore, but touch the platter, scratch it, and destroy head ceramics in seconds. Heads are not flat under them, but are channel shaped so hey can fly when the RPM reaches an average of 30-40% of the nominal. Only then the arm is unblocked to move to outer, and then to middle track. Everything in a drive has a purpose for being there.

[@tony359]

My intention was never to fix those drives. I was just curious to see what happened inside. I’m well aware that opening an HDD - let alone breaking parts inside! - will kill it! Thanks for watching!

[@ovalwingnut]

Score! The "tip" on the heppa filter requiring a hurricane to pass air was a shocker. That never crossed my mind (however short a trip that would have been). It's been said that; "Mechanical hard drives are (were?) the most advanced and complicated devices ever sold to the general population"....Very COoL. Thank you. Who could not sub after this video? Not me. Cheers from So.CA.USA 3rd House On the Left.

[@tony359]

ahah thank you and welcome! :)

[@fft2020]

Tony: I have opened a few hard drives and closed them and they continue to work fine. I also dont have a clean room/box but I make sure: 1. I always cover the platters with a piece of paper when I am working/observing the insides. that way less dust lands on the platters and if I mistakenly touch them with something is not instant game over. 2. Under no circumstances I move the heads or touch the heads directly.. they are very thin and finicky and they can be "bent" out of spec immediately 3. when I do need to move the heads I will do it applying the gentlest possible force horizontal on the opposite side of the arm mechanism near the magnets as to mimic a horizontal arm movement.. and even this is high risk 4. before I close the drive I put it on a lateral position and spray it with compressed air the second I close the cover 5. After closing, I cycle power on a few times but I cut power before the drive arrives at operational speed so this prevents the heads from moving and helps centrifuge as much dust as possible away from the platters. 6. I tend to choose damp days when the dust particles in the air are hopefully slightly less than on a warm summer day... I dont think a slightly higher humidity rate will rust the drive up :p I know those drives are dead but I always try to avoid doing more damage to the drive just with my curiosity .. consider it as a practice run to when you do need to open a possibly salvageable drive. . All that being said I believe there is almost NOTHING inside the drive that a hobbyist like us can fix... except replacing those gooie rubber stoppers and unstuck a stubbornly stuck platter motor (that will probably re-stuck again very soon)

[@tony359]

All very good suggestions, I'd definitely remember them if I ever need to open a drive which has some chances to be re-vived for data recovery! Thanks for watching!

[@minombredepila1580]

It is quite tricky to revive one of those. I wonder if the goo was blocking the heads and even if the spots on the plates were goo. Very ingenious box, though !!. I wonder if one of those could be made to create a "clean atmosphere" where to open HDDs without damaging them with dust, glass top and with a glove on each side to handle the disk properly (and some space for tooling too) just like the ones to handle radioactive material or lunar samples. Another cool video from Tony 🙂

[@tony359]

With proper tools a "clean hood" could be created for sure. If you watch some data recovery videos, that's how they work. I wanted to make a poor's man version but... :) Thanks for watching!

[@alisontelford9339]

Hi Tony -- thank you for another fun and interesting video. As always I like how you come at things from simple first principles. Even unfixable items can provide information and insight! Speaking of unfixable (for me, anyway) I wanted to ask if you'd like an IBM PS/2 model 80 motherboard that I have that has a video fault. The video controller chip (which seems to have a bad address line 1) is a large lidded package chip that I myself probably can't desolder, but you have more expertise with thick planars and awkward large chips than I do. Interesting PS/2 architecture, lots of weird IBM ness about the model 80. Anyway, let me know if you've any interest in a donation! Cheers!

[@tony359]

Thank you for your kind words and for watching! I love IBM stuff but the problem with a model 80 is that I don't have a power supply to run it. I only have my trusted PS/2 model 30. So I wouldn't be able to do much with it I'm afraid? What do you think? Thank you for offering though!

[@alisontelford9339]

Hi tony -- I was indeed thinking about all that. I'd say there would be four challenges to getting it repaired. Power supply, memory modules, beep speaker and sourcing a new VGA controller chip. Power supply could be handled by a homebrew adapter to (eg) an atx supply, though definitely not trivial, the connector is of course unique to the model 60/80s (but also primitive, no power sense, just on-off). The memory modules are also proprietary and unique to the model 80. There are two memory slots and I BELIEVE both must be populated. I have two modules I could loan to you for the purpose if you wish. The beep speaker is the same unit used in all the other machines from the model 50 up; you can probably hotwire in a basic 2 wire speaker (just for the beep codes). The machine has built in vga output and ps/2 connectors, so its easy to hook up to monitor and keyboard. Finally and worst is sourcing a donor board to replace the VGA controller chip. For that I don't currently have a solution, though I can keep an eye out. I've had enormous fun playing around with these microchannel machines, they are pretty neat. If you want to discuss things any further, let me know. I'm not sure how to provide you with a private email via youtube lol. Unfortunately, I'm not a social media person, so I can't resort to twitter etc. Thoughts? Cheers!

[@tony359]

I'd say let's wait until I stumble into a working model 80 - it feels there would be too many variable if I tried to patch up parts, in case one of those is faulty but I don't have a chance to test them! For the future, my email address is on my profile, you can contact me anytime! Cheers for the offer!

[@mihalym.6876]

Thank you your video, especially the second drive part which answers a recent time mistery with my 170MB IBM HDD. The error is the same as yours: drive spins up and shortly spins down. This error came after a surface test for a drive which run about a half an hour in an open cage. I remember the drive outer surface was surprisingly warm. The test was flawless, no bad sector at all. After a couple of days the drive produced the spin up - spind down error at first start. Thank to your experiment it is clear the extensive testing - heat melted the rubber stopper ring. Just for curiosity I will try to replace the rubber stopper with tape as one commenter did. This information is a big help for me because I planned a rescue project for this drive to figure out what went wrong in the electronics. Lesson learned: do not do the surface test without proper cooling!

[@tony359]

Hi there - Glad you enjoyed the video! Bear in mind the rubber "melts" because of age, not because of heat. It's a chemical reaction, those drives were never intended to last 40 years :) I hope you can revive yours! Good luck!

[@richardravich8337]

Tony, Just discovered your channel.interesting stuff. I worked for a disk drive company in the 80's and drives of that vintage have a platter dedicated to being a servo platter and the first drive probably couldn't read the servo track. Probably a bad servo head. Regards, Richard

[@tony359]

Yes, I've been told that, very interesting! Thanks for watching!

[@harvaldi]

From my experience any old hard drive which relies on a stepper motor to move the heads will have mechanical failure with time. In those drives is simply to much parts and every one of them is critical for working. Those mechanisms was not precise enough from the beginning. I had one 20MB 5,25" drive that needed to be push-started manually from time to time!

[@tony359]

It's so sad, I love the noise of those drives!

[@everTriumph]

'Voice Coil' drives have one surface of the 'pack' dedicated to a signal telling the electronics where the 'cylinder' tracks lie. Some drives failed because the signal faded, so the heads seek but do not find. These drives can be destroyed by a powerful magnet.

[@tony359]

interesting, thank you!

[@ejonesss]

if you are willing to put out big $ for data recovery you could send them away to a data recovery house so you can see whats on them* *= you should be safe with the ibm ps2 and xt pcs and others from the same time period(s) because the contents of the drives would just be text and very limited graphics. but todays computers would be very graphics intensive and may contain illegal images and you could get in trouble if there is csam on them. kzbin.info/www/bejne/g33Td3x8mJaofqM is a very good video on what you should do to deal with used computers.

[@tony359]

Absolutely nothing I want to recover from those drives. If I could, I'd like to fix them, but I'd not be interested in their content! Thanks for watching!

[@mrnmrn1]

I'm not an expert in fans but I think your axial fans would never be able to push/pull air through that (or any) hepa filter. You will need a radial blower fan for that.

[@tony359]

Maybe. However I think that that filter is too small, it’s designed for a small airflow, not what that big fan is trying to do. Thanks for watching.

[@2009numan]

surely all mechanical hard drives make mechanical noises, thats why they are called mechanical hard drives

[@tony359]

Absolutely! Maybe a bit less nasty :)

[@ayan.debnath]

Wait a second, that might be the Black Spiderman Alien Goo thing!

[@tony359]

ahahah Oh No, I need to be careful or I'll be turned into.... a.... AHR AHR AHR too late!!!! :D

[@Lilithe]

I wonder if you could block ambient light from getting into the glass sensor. Also: If the smudges stuck on the head, maybe cleaning the head would help or hurt somehow in amusing ways.

[@AlexandroTrevisan]

I thought that too. May be the ambient light is interfering!

[@tony359]

Yes you are probably right, I didn't consider the light interference at all!

[@mima85]

Those IBM PS/2 drives are prone to contamination, perhaps due to the gasket failing over time. I have 3 of them and all exhibited the same failure mode. I tried to transplant into one of them another set of platters taken from another donor drive, after having thoroughly cleaned the heads and the drive mechanics chamber, hoping that this would allow me to low-level format the drive and get it back to work, but I wasn't lucky as the drive wouldn't initialize. And the following day the heads were stuck to the platters, I had to free them manually. Those platters came from a drive which was more recent and they had the modern mirror finish anodized coating instead of the iron oxide substrate typical of the drives from the 80ies, maybe it was this that prevented me to revive the drive. I'm waiting to get my hands on a couple of platters more suitable for this experiment.

[@tony359]

The bottom platters are ok, it's the top one. I still feel it's that plastic bit I showed in the video which slowly but surely attacks the coating of the top platter. Who knows. Well done with the heart transplant even though unsuccessful!

[@mima85]

@@tony359 Yeah that's strange, on my three drives all the platters showed signs of contamination and I don't remember about that white powdery substance on the ribbon cable support piece. Maybe the cause of the contamination between my drives and yours is different. Regarding the heart transplant, I was nearly succesful some years ago with a 3.5 inch MFM drive. After having replaced the platters I was able to get the drive low-level formatted, and after partitioning and initializing it I got just a handful of bad sectors, accounting for around 10-20 KB of unusable space if I remember correctly. The fact is, that also for that operation I employed platters coming from a much more recent drive, and the next day the drive was stuck. When I tried to gently free it up, an head arm slightly bent, putting that head permamently out of alignment, which in fact killed the drive. I guess that platters with the anodized coating are really too smooth for those big (comparatively to more modern drives) heads. That's why I want to wait until I get appropriate platters for the IBM drive to which I'm planning to retry the experiment, because I really don't want to break the heads on those drives as they're quite rare nowadays.

[@tassdesu]

Thank you. Iteresting. I like old hard drives.

[@tony359]

Thank you for watching!

[@Pulverrostmannen]

I think one part of the various behavior on the second drive is that you have the drive exposed to your studio light, as it is a light sensor it depends to be in the darkness of the drive but as you have strong external light coming into it the signal is very likely disturbed. The rubber that become goo is probably made from rubber tree, many drives have this issue and I had several drives die from it which is the Quantum Pro Drives which get the exact same problem. I am not sure if every Prodrive model have this issue but the Fireball does not have this kind of rubber and don´t fail like this fortunately. but it is really sad you can´t even store a drive without it dies on you from age

[@tony359]

Yes, other have mentioned that, I didn't think about that! Thanks for your input and it's sad indeed, those drives are all doomed.

[@Pulverrostmannen]

@@tony359 yep. Not much we can do about it except waiting for it to happen

[@repairitdontreplaceit]

you need a tangential fan they produce far more static air pressure

[@tony359]

Yes, probably!

[@mostrush4849]

I think you should use the filter in suction side for the dust filter.

[@tony359]

The filter has a flow arrow, I am using it following it. I also tried reversing it (and the fan) but it makes no change I'm afraid.

[@ste76539]

Salvage the stepper motors for electronics fun, get the platters out for frisbee type shinanigans (or coasters) and that's your lot with dead drives.

[@tony359]

ahaha maybe! Those IBM drives seem to be all goners so I'd love to have some circuit to spin just the stepper motor, randomly, based to the XT-IDE LED activity. This way I can hear the platter spinning and the stepper motor "squeaking" :)

[@ste76539]

@@tony359 I wouldn't expect you'd have any difficulty making that a reality. Have to admit, that's the entire reason I built an IDE interface for my home brew CP/M computer! I used a 2Gb Toshiba laptop drive with it and it sounds glorious 😊

[@middleway1885]

Cheers!

[@Poppinwheeeeellllllieeeeez]

I thought you were going to be doing a scope video.

[@tony359]

At some point 🙂

[@tomekrv942]

Great video. I wanted to see it. It is very interesting for me

[@tony359]

Thank you for watching!

[@berndeckenfels]

Interesting that it does not park the heads

[@tony359]

Yes, I think those drives were the first not requiring a park command but as you say you'd expect at least the IBM one to go back to the same place when powered off. Interesting!

[@pvc988]

I wonder if opening these drives upside down and working on them in that position could help with the dust problem.

[@tony359]

Spiderman, Spiderman, does whatever a spider can. Or actually: Spider-pork, Spider-pork, does whatever a spider-pork does :) Thank you for watching! ;)

[@pvc988]

@@tony359 🤣

[@aleksandardjurovic9203]

Thank you for the video!

[@tony359]

Thanks for watching!

[@vinitsingh8962]

You can try big filter and big fan together, just purchase a 48v 120mm fan it's available, i have two of these and they are damn powerful. Or there are even powerful one if you want, try those with speed control and you will get what you want, and if you find this helpful let me know.

[@tony359]

Of course. Then you need space to store them though :) Thanks for watching!

[@vinitsingh8962]

Didn't understand what you want to say by store them, if it's about parameters of the plastic body use a bigger one, you can find it easily or maybe if you misunderstood the 120mm, it's maybe the same size of fan you got and if you are talking about me to store the fans that i got, I have a box full of just these fans mainly 80mm and 120mm, 12v to 48v. Whatever it is I have you got the answer. Always a pleasure to watch your video content.👍👍

[@tony359]

I was thinking of the big filter :) I could definitely build a small hood with a proper inline axial fan, a proper filter... but then where do I store it? |Granted, you mentioned 120mm fans so I kind of misunderstood you! To the next HDD project I guess! :)

[@vinitsingh8962]

Many has misunderstood it when I say fan's dimensions. Also good to talk to you.👍👍

[@andream1977]

3 2 1 ... Go and let's rock, great Tony as alaways!

[@general23cmp]

Great video!

[@tony359]

Thank you for watching!

[@TzOk]

Once I had an old SCSI drive from a MAC LC-II, and it had a similar issue with a disintegrated rubber head stop. I've managed to get it working, by replacing this rubber with a few layers of electric tape. I had to roughly adjust its thickness, to make the disk successfully initialize.

[@tony359]

Yes, another point that has been made is that the rubber is a stop to find "track zero" and if the stop is not there anymore, the disk doesn't know how to find track zero. Thanks for your comment!

[@snakezdewiggle6084]

Hi Tony. Looks like you are forcing air Into the box with the fan, and Out thru the filter. You don't need > air flow > in the box. Just a quick wipe-out and seal the box each time, is all you need. I use cling wrap.

[@tony359]

Thanks - the plan was to have filtered air on the HDD, not a clean box though :) Thanks for your comment!

[@snakezdewiggle6084]

@tony359 Oh, I should listen better. ; ) 👍

[@tony359]

Likely due to my broken English :)

[@snakezdewiggle6084]

@tony359 Not at all Tony, your english is fine. I'm in Australia, we speak our own version of english ; D I especially like your, "Hello the Internet". ; )

[@tony359]

Borrowed by The Slow Mo Guys :)

[@homelate1306]

The 1ste HDD, might it be possible that the magnetic layer is reacting on the inner layer of the platter? I think the inner layer might be aluminum.

[@tony359]

Yes, it is very possible! Thanks for watching!

[@donixion4368]

I forgot about hardcards. They were a good idea back in the day. I had a 386 with a lack of places to put an extra hard drive and I considered a hardcard. I ended up just getting a better case because it was just a bit cheaper. The First hard drive looks like a head crash or head scraping the platter.

[@tony359]

I have a feeling it could be that shite powder depositing on the platter which would then damage the platter when the HDD is turned on. Another youtuber has found the same white powder: kzbin.info/www/bejne/fZmzmniwlr5jgKc

[@kuromi_oficialYT1234]

Hard drives have an anti friction coating on the platters to reduce friction and wear. What i think that happens with these old hard drives is that the anti friction coating stops doing its job and the platter gets scratched.

[@tony359]

That is another one yes. It looks like HDDs need to be run every now and then to keep them healthy. Thanks for watching!

[@Thorsten369]

Dear Tony, when i hear the sound of those fans 02:10 it reminds me of when i did refurbish some Compaq servers in the 90's. When you put power on of those servers (they look like a big cube back then) it's like those server gonna take of to the sky. Anyway great video my friend, always love to watch you videos. Have a great weekend my friend and keep up the good work.

[@tony359]

ahh, server fans! Where noise is not a concern :) Thank you for watching!

[@Thorsten369]

@@tony359 Indeed my friend, and if i tell you we had to power up 4 to 5 servers at once? Then my god you really needed back then ear protection. Anyway keep up the good work my friend ;)

[@wimwiddershins]

The variety of head driving mechanisms is interesting.

[@tony359]

Marvellous indeed!

[@askoldmodera]

Hi Tony! Thanks for the video! Regarding first drive, if it really doesn't initialize because it couldn't read scratched top side of top platter, maybe it's possible to rewire its head to good platter, i.e. desolder head for bad side and add badge wires to connect some other head in its place, in parallel to its original connection. Of course it won't function properly that way, but maybe it atleast could initialize :D Not sure if it's okay to parallel heads that way tho.

[@tony359]

That's interesting - do you think the top head is somehow more important than the other ones?

[@askoldmodera]

​@@tony359I don't know for sure, but electronics may require that all heads are able to read the data, otherwise it fails. My guess is, there's some special data on platters that helps electorics to figure out exact position of head in particular moment, and if that data mismatch between different heads it cannot calibrate and fails. There may be caveats like that special data is unique on each platter, or one head wont work properly if connected to electronics twice, only hard disk experts know that, which I'm not :)

[@bogdanbogunovic3995]

Another interesting video. As far as the filtering box I thought that the ventilator should be outside the box and after the filter ... sucking the air and dust particles out of the box. Blowing air directly to the platters will make air turbulence which can cause hads to hit the platters.

[@tony359]

but then where how do I work on the HDD? If I understand your scenario, the clean area would be INSIDE the box :)

[@bogdanbogunovic3995]

Correct, clean area would be inside the box. At least two large filters installed inside two walls and one or two vents, or industrial vacuum cleaner suction hose, pulling the air out of the box. But now after I have realized that your project implies an out of the box scenario ... I would have to think about it a little bit more ... I mean why do we need the box then. One more thing, I'll jump on 14:00 really quick and give a short opinion, maybe the heat gun you used to remove the cover melted that plastic which has already started to degrade.

[@tony359]

Yes, mine was definitely a non-idea filtering system. Better than nothing I guess. On 14:00, you mean the gooey stuff inside? No, I can assure you the hear did not get so further into the drive, that degradation is very normal with those drives unfortunately. Thank you! :)