Alright, I feel like I need to address some things here because, whew, y'all are riled up! First, keep in mind, that I'm a car guy, I'm very familiar with dead or loose wheel bearings. The reason we thought it was potentially a head crash was the initial squeal. We never ruled out a bad bearing, as a matter of a fact, we were quite confident the bearing was unhappy. Our main worry though was that the bearing had maybe lost some tolerance and that that may have caused a head crash, which was that initial squeal we heard. And if that were the case, diagnosing the bearing would have been pointless, so we cracked the case to confirm whether heads had crashed or not. Next, after reading everyone's comments and having a good long think on it, I'm pretty sure the bearing noise and subsequent "repair" was due to three things: 1. The drive was on its side for the better part of 40 years and ran on its side for who knows how many thousands of hours. Then it gets here and I not only spin it up on its back, but also let it sit on its back for 6 months. No matter how good a bearing is, it will experience some wear, and the wear pattern made by this bearing was with it on its side. It's not impossible to think it slipped out of a very specific worn groove and got into some position it was unhappy with. 2. Temperature. Last I spun it up it was nice and hot out. Now it's nice and cool out. The grease is probably already starting to harden, but being cool just exacerbates the problem. Furthermore, I never spin it up and just let it ride, it's always on for just a few minutes at a time, so the grease never really has a chance to get warm and lose some viscosity. So a combination of being out of its groove and the grease not doing what its supposed to do just compounded the problem. 3. Some percussive maintenance. In order to get the clips back on the enclosure, I had to essentially hammer them on. This probably jostled the bearing just enough so that the next time I spun it up on its side, it slipped back into its groove. Which is why it now sounds totally normal again. The bearing is back in its happy place. So, what's next? The number one comment I've gotten is that I absolutely must repair or replace the bearing immediately, which I'm not going to do. Let me explain. With regards to replacing, I have no clue where on earth I'd ever find replacement parts. The tolerances for the drive bearing are very specific, and CDC has been out of business for 20+ years. I've never even seen a picture of another Finch drive. Furthermore, that requires a level of disassembly that all but guarantees the platters have to come out. This means the platters will become misaligned and the data will be 100% irretrievable, whereas right now there's still a chance we can get the data off the drive. Now, let's talk about "repairing" the bearing. The bearing is worn, there is no real repair that can be done here. The best we can do is maybe work some light spindle oil down the shaft and hope it penetrates the seal and works its way into the bearing itself. But, this in itself can be quite dangerous. If the spindle oil is too light, it could seep past the seals on the bearings and into the enclosure, which would definitely crash the heads. Even though it's sitting on its side, I've seen oil creep along a horizontal shaft before with ease. Also, I have no clue what type of grease was originally used by CDC, probably something extremely specific. If I get the wrong type of oil in there, it could react in an unpredictable way. It may make things better in the near short term, but could make things far worse in the long term. Getting oil into the bearings may be something we have to do in the future, but for now, it seems just a little too risky. So, what are the next steps? Well, as long as the bearing stays quiet, it'll live on its side and continue to be used. Keep in mind, I'm a hobbyist, not a business. This drive will maybe see 30 hours of runtime a year, 60 if we're feeling spicy. The bearing that's in there may have 200 or 300 hours of usable life left in it. That's unacceptably low for a business, but that's like five years of use for us. So, for now, mechanically, we're not going to do anything more to the drive until the bearing speaks up again. Hopefully it won't now that it's back in its groove. But we'll cross that bridge when we get to it. We're not out of the woods yet though, there are definitely failures on the PCB though, so that's the real next step, start trying to track the data signal back and figure out where we're losing it!
I think opening the drive to check the platters was a good idea since the motor shaft felt ok. Why did CDC use plastic, but not transparent plastic? You should make a nice enclosure for the drive to sit on its side and put a thermometer in there so you can always run in a consistent environment.
Very good video! I recognize that noise. That is the sound of ball bearings which are suffering from unloading or skidding. The balls are not rolling between the two bearing races but are sliding. They don't slide smoothly and set up a resonance where the shaft has a whirl. The lubricant can play a big part in this. If the lubricant is too dry or firm, it can exacerbate it. I expect that the vertical shaft orientation allowed one of the two bearings to support all the weight of the platter, while the other bearing was just floating without any load on it. That can lead to a slide or whirl building up. Manufacturers often include thrust springs or wave washers in designs with two ball bearings, to eliminate any chance of unloading. That works when the lubricant is on-spec and new. Your Finch drive has decades on it for the lubricant to dry up and become harder. Also, as you said the last time it ran with a vertical shaft was in summer, when the drive was warmer and the bearing lubrication softer and more capable of operating without sliding or whirling. Now that it's cooler, that will be stiffer. If this happens again in spite of it being in a vertical shaft orientation, stop the drive and place it in a warm environment. Heat it up to about 90 or 100°F for an hour or two, then try again with the lubricant warm. Looking forward to your future videos where the data is finally read from this drive! Thanks again for making these videos for us!
If this is your case, then this suggestion may be your only hope as getting access to the bearings probably means removing the platters, I doubt you'd ever get them aligned again, an alignment that is more critical than the factory alignment, as you want the data, the factory had no such concerns
imo you've hit the bullseye on this one. The only thing I would add is once in that warm environment, keep it there with the temperature as stable as possible before & during operation.
Yep , you hit the nail on the head :-) I don't have that much experience of HDDs but that was the sound of a dry bearing (replaced many in other equipment over [too] many years)
I think you're spot on! I was thinking about the temperature difference between summer, but was thinking along the wrong effect from it. Having the grease be softer when warmer makes way more sense. And it took banging on it a bit and running it while vertical to kind of get the bearing to settle back in to the shape it wanted to be. Hopefully it won't happen again now that it's in a vertical orientation (and will be until it fully gives up the ghost), but if it does, I'll put it on the back burner until summer and then let it sit in the Texas heat for a while to think about what it's done wrong!
@@paulstubbs7678 This right here is something that I think a lot of people haven't realized yet. Being soft sectored, the moment the platters are loosened up to give access to the bearing, the alignment gets out of whack and the data is completely unrecoverable. Furthermore, I don't have any of CDC's alignment tools, which means no way of writing new alignment data to the servo platter. I suppose we could get lucky and the old alignment data could be reused once the platters are cinched down, but the rest of the data on the platters will never align to that again and be totally gone. It's a last ditch effort if we want to restore the drive but already have the data off of it.
Get a HEPA fan filter. I have one on the bench for stuff like this (and biotech stuff). Not too terribly expensive. Expensive diffraction particle counter in front showed nothing, not a thing down to 0.3 micron (lowest it goes).
I am amazed none gave you the advice to build a DIY clean-room box, similar to sandblasting booths or spray painting boxes, they are similar to the fume cupboards used in chemistry but in reverse, you push air trough an array of some fine filters and into the box, and then you have a fixed size aperture that ensures a certain positive pressure is held in the box compared to the outside, then you could either use air-tight gloves, or rubber seals for your arms, or adapt the air-exit aperture to be the holes for your arms to access.
To be honest, I've believed half of my life that HDDs are brittle little touchy devices that basically will die if you ever open one up. But after being way too curious I've actually tried running a couple of older HDDs in opened state. And lo and behold, actually nothing bad happened even after having run them for a couple of days. And by "older" I mean way more modern than those monster HDDs you work with, so actually devices with really low tolerances. I mean, I understand that you CAN crash the head with some dust or a fingerprint, but the chances for that are really really low. So, usually, if an HDD platter looks clean, you are actually pretty safe.
It depends on construction and thus data density. Very old hard drives can tolerate some dust. HAWK is not even sealed. Oldish HDDs may run for a few days or weeks before dusty head scratches the surface enough. Modern HDDs are generally to be presumed dead if the capsule was ever unsealed outside of clean room
Oh yeah, back in the day, I had a bunch of obsolete drives and before taking the neodymium magnets out I opened them up and played around with them and they were almost bullet proof. I'd spin them up write some data and read it, then stick my finger all over the platter to stall the motor and then let go and they would just spin back up again and still read and write just fine. ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-yISqCAnROh8.html
I remember doing that to a Quantum Fireball after it had a lot of bad blocks and I got like only 25% of the original space left usable. The thing just kept working just fine, I was like WTF this thing is indestructible. It would even boot Win98 with lots of data written in bad blocks not yet marked if you turned off "turbo". And yes, Win98 in a 486DX2 , stretching the machine to the very limit (perhaps that's why I was getting bad blocks, who knows) I also remember using that machine with Autocad R14, took its sweet time to boot, but I had a 8MB VGA and the machine had 32MB of RAM, so It ran fine after 30min of boot. (I think my bad blocks was coming from the "scratch file", aka, virtual memory, literally scratching the disk dead with excessive use (write cycles of that thing might have been totally used)).
Louis Rossmann just moved to Austin, TX. They regularly offer public repair workshops and have a decontamination cabinet for working on hard drives. Perhaps you can ask them to use this.
Definitely a bearing going out. I have heard that exact sound in tools with bad bearings. Simply handling it probably knocked the bearing balls back into place but it will very likely do it again and you will have to take it apart to replace the bearings. So don't feel bad for opening it, it will have to be opened again. That motor looks like a motor that might use standard ball bearings so you actually might be able to replace them. But if it comes to that, shelf it until you have the means to do it properly (In a cleaner room) :-)
Either that or the stepper motor for the heads. I have a 500MB Maxtor drive that spins up fine and then makes the most awful banshee screeches instead of loading the heads until you either heat it up slowly or give it a little wack, then it runs perfectly fine for another year.
Sounds like the a ball in the bearing got out of its cage/cage failed. The orientation may certainly be impacting this. it may work on its side but certainly its life is very short. If he gets it working then like he said always run it on the side and avoid using.
@@tomlindo2863 The heads will probably be aligned in the orientation that the data was written/formatted in anyways. I would guess a new orientation would cause problems.
No "Hellorld"? :( About the dust - while certainly not ideal, I think it's unlikely that you've ruined the drive by opening it up. I've done the same to a couple of 3.5" hard drives from the late 90's before, also in very much not a clean-room setting, and they still work fine. I'm no physicist, but my guess is that any loose specks of dust are sucked right off the platters by the airflow and centripetal force inside the drive.
It may be built with the principles of a modern drive but the tolerances are probably 1000x looser. A tiny spec of dust won't kill the drive like it would a modern one
I think you had the right of it at the very end. Since that drive had spent its life on its side, the bearing had worn unevenly, so they weren't perfectly round anymore. When you laid it flat, it got screamy. But once you set it back on its side, those mishapen bearing fell back into their old configuration. I guess what I'm saying, when its on its side the bearing actually follow more of an oval than a circle (gravity pulling down on bearings with a very specific wear pattern). So when you laid it flat, centrifugal force forced them to follow a circle instead of their worn in oval. SCREEEE! LOL
Depends on the bearing style. Anything with balls or rollers will not care about orientation. A bushing will. I can't see them using a bushing for those speeds.
@@SeekingTheLoveThatGodMeans7648 Highly unlikely. If it did... then all positions will be affected. The hardware was seriously over built in those products. Other electromechanical items are going to be suspect far sooner than a simple bearing.
Back in the 80's we had a huge number of HDD that had a "grounding strap" on the motor bearing, and that would sometimes (inevitably, usually) wear down to a hard spot in the graphite, then the drive would scream and howl the entire time the spindle motor was running at speed. I don't believe that is a head crash sound. It didn't change at all when the heads did their test seek, and they should have, as the disk is spinning at different linear speeds at the various radii. If it was dragging a head on the platter, the tone should have shifted. And to my ear, it did not. I firmly believe you have an issue in the spindle motor somewhere.
You can build a very simple proto cleanroom enclosure from some solid, dust-free material and one of those room air cleaners with HEPA filters. The entire point is not to remove all dust from the room but have clean air blowing constantly over the thing you're working on to carry any stray dust away.
The surface in this case is actually a full on textured pattern cast into the plastic. Water definitely would have made it clearer, but not enough to see deep enough to the spindle. Excellent idea though!
@@UsagiElectric as long as the texture is on the outside of the case, you can put scotch tape or clear packaging tape on it and you should be able to see right in. It works on glass with this kind of texture on it, but I don't know if it will work on plastic. Here's a video of someone doing it to glass: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-IRMGXS8UqzM.html
That second spin up at 10:50 sounded exactly like dry hydrodynamic bearing. It could also be worn out enough that it has lost it's preload. Changing the orientation would definitely affect the distribution of any lubrication which might have otherwise been dampening the vibration so it runs smoothly. This can make a dramatic difference if it has a hydrodynamic bearing instead of ball bearings.
When I heard the finch noises first thoughts were initial squeal were motor bearing noise and the zing ping scouring was a head crash. So glad the latter didn't happen (up to where I've watched). Even though the disk drive doesn't have a fan or purge pump there will be a fair bit of air circulating around inside the enclosure. The plates stir things up quite a bit. Opening the enclosure up isn't necessarily a death sentence but you never know how long things will run after that. I resurrected failed Conner hard disks that had erased the servo track or bump stoppers had gone gooey on by taking the lid off, powering up, and before they could give up and spin down again giving the head assy a little nudge to move it out of the landing zone and then copying everything off pronto. Afterthought, could the squeal have been the ground tab on the motor? I've heard the grounding tab on head drums for VCRs and DAT drives make squeals.
I have my doubts that air pressure is the cause of the noise. What is the temperature like? I'm thinking the lubricating oil got stiff in the colder temperature of winter. To my ears, the screech sounded like bad bearings.
@@russellhltn1396 - It has been more than 30 years since I worked on equipment from this era, but I believe that the heads did unload. In the "park" position, the heads rode up on ramps that lifted away from the platters.
the sound did not change when the heads did their seek @11:25, not a crash. its a bearing. still bad because you'll have to split the drive shield off and pull the motor to change the bearings, the platter will be exposed to the environment.
Wow, man. First thing I thought of with the screeching sound was a head crash! I programmed for Control Data when I was 19. I was the youngest programmer they had ever hired up to that time. My team worked with Nova, Phoenix and Eclipse systems on the Tools project, which means we wrote the tools the programmers used to write programs. We did not use this computer or OS. We were programming in a language called TAC BB2, which didn't even have floating point math! (ʘ_ʘ) I remember the hard drives we were using looked like washing machines that had the nice, distinct, multi-platter drive cartridges. Your stuff is always entertaining, even when you have your share of challenges. It's a great journey! All good wishes.
Yup, the initial screech that then settled into a gravelly nightmare sounding like a head crash to us too! Either way, we don't have any spare parts for the drive, hell, we've never even seen another one out there, so head crash or bearing, both put the drive out of service. That's awesome that you programmed for CDC, and even more wild that y'all used a language that up to this point, I had never heard of before! Thanks for the kind words on the videos!
I have seen this before from the motor grounding strap. They get an almost invisible wear spot that starts interacting with the motors shaft, causing it to vibrate, squeal, chatter etc. Try cleaning off its bearing surface with a fine abrasive to remove that wear spot. Otherwise, give it a bit of a bend so it's tension is different and does not interact the same, or see if there is enough slack in the retaining screw so you have a slightly different spot for it to 'bear on', actually as you did pull it apart and put it back together you may have inadvertently done this and given yourself some more time. Years ago. where I worked I came across a 4 head drive that had a really bad crash on one head, the drive had run many hours like this and there was a track in the platter and heaps of aluminium dust in the case. For a joke I blew out all the dust, cut off the dead head, then proceeded to format it up as a 3 head drive - I got it to boot! ('we' had many a computer that booted up and just ran some system, only using the drive to boot. If one of these had a head crash it could be 'ok' for months until a reboot was needed - they were just noisy, but in an already noisy 'equipment' room, so just leave it alone, the system is up)
I'll bet there's more airflow inside the drive then you think. The platters act as a centrifugal fan. All that's needed is a air channel to the hub and the spinning platters will pump air.
This is true, I chose the wrong words here. The spinning platters do create a bit of airflow within the enclosure, but I'm so used to working on the Hawk, which use a large amount of positive pressure to purge the drive, it's a relatively tiny amount of airflow. But, the small bit of airflow within the enclosure will hopefully carry away any dust that may still be in the enclosure.
My friend's Dad used to work for CDC (or at least claimed to, as he was poor and smelled like salmon), but he told me that when starting a Finch drive you need to drunkenly shout "Here Finchy Finchy Finch!". He gave me this advice after I caught him inhaling all of the Freon from my moo-moo's fridge.
I know a guy who's dad worked at CDC and he came up to me and asked "do you know how to get rid of canisters of freon". I think his dad kept them in their barn, he worked on the mechanical side of the mainframes, cooling, drives, that's sort of thing.
Yeh. Contrary to common belief, "clean rooms" don't necessarily need to be full sized "lab rooms" with lots of big air treatment machinery attached. And actually, commercial data recovery ones have almost all them the size of a table. If is for one use and for a personal project, you can construct a simple one in the size of a shoe box as you said.
Where can people find more information on how to construct those, what precautions to take, etc.? Wouldn't one effectively have to build a glove box in order to do anything with whatever is placed in the box? Or am I missing something?
@@ropersonline Look for laminar flow bench or clean bench as keywords. The box is half open on one side, so you can put your hands inside and move tools in and out. There is a constant air flow out that opening and the air speed is slow enough to be laminar, which avoids whirls that could take dust inside. That air flow is created by fans with HEPA filters. Such boxes are common for working on optics to avoid any dust.
with regards to the bearings, i would say that's what it is. i thought originally it was a head crash, but then hearing it again changing with speed the way it is, definately a bearing. the same sort of thing applied to ST225 hard drives back in the day. had one that spent its life on its side, when i put it in a new machine flat, had a similar issue. As regards no clean room... I used to mod hard drives with clear plastic/acrylic covers cus it was a fad, we used to do that in a bath room with a high humidity and opened the drives inside a ziplock bag.
Takes me back seeing that. Worked in the clean rooms at IBM in the mid 80's, line i worked on was rebuilding their Gulliver drives which were getting on for 10 years old by that point. Massive things they were. I remember hearing the odd head crash on testing, never a happy day when that happened. I can't remember how they used to fit the platters and head actuator assembly but do remember soldering the head wires back onto a interface board. They used to coat the disc platters in house as well, that was quite a setup for that.
Sweet memories eh? I worked on developing of drives in the 80s in former east Germany. We had IBMs as benchmark models. Unlike the Russians we didn't copy technology but developed all from scratch with the means that were available.
When opening a hard drive outside of a clean room, I'd probably have put a piece of double sided tape inside the platter housing to collect any stray dust. Maybe worth a shot.
Years ago (like 30ish years,) I had an old full-height 5 1/4" 80 MB drive. Being the curious type, I decided to remove the cover so I could see how it worked. I hooked it up to my machine and let it run. It had a stepper motor mechanism for the heads, and I was fascinated watching that thing zoom around the platter when I did various things on my computer. It actually ran for quite some time -- I can't remember for how long -- but at some point the coolness wore off and I just started running off another HD I had. I wish I still had that thing. My point is that old HDs are tougher than you'd think as far as dust particles and such getting on the platters.... (or I got VERY lucky, because I left the cover off for quite some time.) Excellent video -- I love this whole series!
Looking at the big picture - the Finch drive isn't critical to the overall project now that you've restored the Hawk and the floppy drive. At worst you've lost another copy of the operating system and potentially some interesting data (applications?) that you didn't have before. Hopefully you can extract the data from the Finch before it dies, but if not, there and many other worthwhile projects waiting for you in Centurion-land.
Hey, did "Hellorld" get old? Maybe you can lay your hands on a glovebox? Way less hassle than a fully fledged clean room... And I agree - the worst malfunctions are those with no identified reason, you can never be sure. Nice result - winner winner, chicken dinner! Oh, nice to see kapton flex cables here; I'd expect that stuff in aerospace or military applications, but computers back then? Nice surprose. Pet the neko-chan from me :)
Since we know something is bad with this drive (most likely spindle) I think the best course of action is to abandon trying to make the machine itself read it, and construct the interface to dump its contents. Good thing "byte clock" still works, this means it is probably possible to copy whatever is still left there bit-by-bit. The data rate seems to be quite high (6MHz?) and there is a lot of data, so unfortunately a circuit needs to be constructed. Actually, a binary counter + latched shift register + arduino (or better - STM32) + SD-card should probably be able do the trick. The firmware should not be too tricky or complicated either. Raspberry pi gpio might be just fast enough to be able to do it, although it would need either RTOS or some buffer RAM.
"I'm gonna quit while I'm ahead. No, I'm gonna quit while I got back up to where we were at the beginning" reminds me so much of xkcd #1994! In all seriousness though, I love your channel and appreciate you sharing this experience. After all, it's as much a part of electronics repairs as the times when with a bit of effort everything just starts working seamlessly. Great video!
Reading the cdc finch drive specification document on bitsavers, I noticed the following The disk and actuator chamber is environmentally sealed. No outside air is drawn into the unit. Air is recirculated within the disk/actuator chamber and passes through the nonreplaceable absolute filter to ensure the maintenance of a con- tamination free disk / actuator environment. So it sounds like any dust you introduced will have been flung off the platters and should be captured by the internal filter, so running it up with the heads locked was a wise choice.
When cleaning my camera sensor for astrophotography, I create my own "clean room" by steaming the bathroom with hot water running for a while. Then steam settles down along with dust etc.
How the fart did anybody identify this as a head-crash though? If the sound is caused by rubbing against the platters then the sound should be extremely dependent on the speed of the platters and the position of the head on the platter, but instead this seems to just come at a certain RPM, get louder as speed increases and go away when the RPM gets below a limit. That's the definition of what a dry bearing does. As soon as it gets above a certain speed it starts to wobble out of control making a horrible noise and that noise doesn't go away or change in pitch(!) until the speed drops below a limit under which the bearing can.... get it's bearings, and selfcorrect. 15:40 "there is no airflow" With that platter spinning, it's an absolute tornado in there. 21:00 "pressure differential" No it can't be, because of the port and the fact that the only thing that can bend under the pathetically minute pressure differential is the plastic cover and that does not touch the spindle. It's a worn bearing.
My money is on a failing bearing. A worn bearing would create noise through the vibrations of its worn parts while rotating and they can be sensitive to orientation where being noisy is concerned. Like if your bathroom extractor fan or laptop fan gets noisy, you can sometimes shut it up with a well placed whack. That might explain whey the drive was noisy one minute, and quiet the next? I would imagine that a head crash would make noise due to the constant collision of the damaged heads and damaged areas of the platters, but if you watch the video back at 11:00mins you see the drive already spinning (noisily) , but when the heads start to seek, there is no change at all to the noise. Would it not be the case that there would be a noticeable change to the sound as the heads swept over the damaged/undamaged areas of the platters if a crash ware the cause?
I agree with the others, as a previous mechanic, that sounds just like bearings going, or starting to seize up, if its not that internal break making contact or dragging. Not to mention the second time you spun it up, the sound started, then the heads did their test, and the sound didn't change at all when they were testing. I will put my money that it not the last time your going to hear that sound. Also, if you didn't you should have wiped the whole drive clean with alcohol before opening it.
sounds like a bloody dry bearings to me... bearing probably not worn out "on the side" but all the grease may move to the side when bearing is hot (at working temp).
Can really tell this episode rattled you, you are talk/laughing far more then usual. Personally I think it was ballsy as hell to expose those platters to the open air! Glad you seemed to get away with it. I think I would take a closer look at that drive motor, it seems to my ear to be spinning too quickly. I had a 1983-vintage SCSI HDD for my Amiga 500, and although it was as noisy as a vacuum cleaner, it never got NEAR that pitch. I think it's spinning too fast. Either way, besof luck, I can"t wait to see what is next!
This was actually a super stressful video to make! I had no clue how the narrative was going to go and even if the the heads hadn't crashed, if the bearing had fully died, we don't have any spare parts for the drive, so there was a lot of teetering on the edge of losing the drive. But, it seems to have just been set to the "let's screw with David" setting, haha. This thing spins pretty slow at about 3500 RPM and the sound is actually a little deeper than the microphone picks up, but it is a bit a howly, which is probably the bearing being upset. Hopefully, now that its on its side, it'll behave itself for a while longer!
@@UsagiElectric I bet it was! Glad to hear the actual cause of the problem has most likely been discovered, and I hope the drive at least lasts long enough for you to get the data off it before you possibly have to rebuild it.
From my early years of working with HDD's in the way past, this sounds like a bearing issue. It also kind of sounds like when the bearings of a fan begins to act up.
Even though it itself is a bearing problem, this sound is typical of that, it makes a hell of a vibration that alone can, and likely is to cause a head crash. So, if it didn't this time, you are insanely lucky, but it will definitely do it again, the bearengs not replaced. Will there be another instance of this great luck, that's the question…
I would recommend taking it to a data recovery centre and getting them to do a full rebuild of the drive so that you can guarantee the drives future!!!!
@@SeekingTheLoveThatGodMeans7648 they have all the tools to do every drive even sd cards and usb sticks and have usually had experience with older drives as a lot of the tooling and stuff is the exact same for a modern hard drive the only physical difference is the size and the storage on these drives. Clean room they would have and all that would be needed for this drive is a tear down and rebuild with fresh oils and a proper cleaning of the disks and housing for the drive, they would have all this on hand and access to old factory specifications for any drive as it’s there job.
@@SeekingTheLoveThatGodMeans7648 also to back up my previous reply since It was a tad too long for RU-vid, a lot of high quality drive repair shops have taken water submerged drives for repair for years and done data recovery on them and been successful with there process. I would think it would be worth the attempt before any more damage happened to the drive.
@@timbersrcadventures you know they generally operate by having parts for the drives right? They take the good platters out of your drive put them in another identical drive with working mechanicals and electricals and read the data off. I don't think they will have many spare giant disks from the late 70s on hand.
@@zyeborm what I was on about wasn’t that it was to lubricate the parts and to do a complete clean in a clean room and make 100% sure that the drive will live on for years to come they may be able to fix the bearing if it’s not to heavily damaged.
See if you can send it to someone who has a laminar flow workstation to clean the dust out of the drive. You NEED to get that dust of of there, or else it will end up messing up the drive with time.
You can roll your own "clean box" that's "good enough"(or better than nothing) with a bunch of HEPA filters and some fans made into a glove box type arrangement. Not saying you should do it, but I have a feeling you may need to at some point. Pretty sure you're going to be up for some bearings. Regarding alignment, I do wonder if the platter stack as a whole might be able to lift out rather than taking them out individually. If you can take the stack spindle out, you may be able to replace the bearings and reload the stack without upsetting the alignment of the platters.... Possibly.... Maybe.... On a good day.... If your tongue is in the right spot when you do it.
About dust, one of the main sources is clothing, so a good way to reduce the amount of dust is to remove your clothing. It is also worth making some sort of tent or enclosure that sits on the bench. Finally, wipe all the work surfaces frequently with something like moist cleaning wipes, or tissue paper soaked in isopropyl alcohol.
I've heard on some of the PC retro channels here on RU-vid that you can successfully open up even some of the early PC hard drives and get them to work afterwards so I guess these things might not be quite as sensitive as the modern ones.
Yup, the data density on these older drives isn't as high as modern drives, so in theory you don't risk much by cracking open a older drive like this. A newer drive though, if you open it up, it's going to be toast.
I've had many apart, no probs. I even fitted a plastic lid to one and stuck it in an indefinite test mode as a demonstration piece. Maybe I should video it and pop it up on YT. For a joke, I fitted a window to another drive, then half filled it with water and dishwashing liquid, it looked quite spectacular when powered up, trouble was the motor/controller soon burned out (with lots of smoke) as it tried to get up to speed. No head crash though!
On older drives the heads are physically much larger (because the sizes of bits they write are much larger) and they fly much higher. The former somewhat enables them to potentially survive impacting small particles, and the latter to potentially fly over tiny ones entirely. A modern r/w head flies less than 5nm - 50 atomic diameters - above the deck, and writes patterns with feature sizes of about 50nm. No margin for much of anything.
As a former repair tech there are few things as frustrating as something randomly starting to work without a reason why. How long do you test it before returning it to the customer?
I wonder if that really thin, really sticky 3m tape around the inside of the plastic case would catch residual dust spun off by centrifugal force, Like a getter for hard drives.
Addition to that. You're a car guy, right? Do you know about crank walk? Same deal. When the sleeve bearing wears the spindle can "walk". Like its trying to roll around the surface of the bearing instead of spinning in the middle of it. When you think of the relative sizes, that's a lot of "laps" per revolution. Trying to do this makes the motor bog down and the bearing screech as the contact points grind over each other, unable to "walk" at that ridiculous speed slipping against each other. I have this trouble a lot with old smaller electric motors.
as far as being concerned about tippynes maybe some of that angle "metal" that usually holds up garage door openers and a base of wood just to bolt it to something that increases its footprint
I've discovered that if I jam a folded up ribbon cable under the enclosure, it actually stops it from being too tippy. You can actually see it at 10:45 in the video! Ultimately, it will be bolted to a tray just like the floppy though.
Man that sucks. Maybe we can all look for another finch drive to get parts from. But I think its a bearing that needs lub. thermal differentials can make a lot of difference on those precision bearings.
I'd say after the second startup I am even more certain it's the bearing - if it was the heads, the sound would have changed on the second power-up during the initial seek test.
Nice video! Two things: the drive cover is translucent rather than opaque. Opaque means no light or vision passes. Trasparent means one can see clearly through. Translucent is between, light passes, but vision is not clear. The noise you hear is a dry bearing. Grease hardens over time, and no longer lubricates. A single drop of premium synthetic lubricant on each bearing will solve the problem, but it would be best if the old grease were cleared away first.
I think this is an example of the hardware engineers and the PCB engineers disagreeing on stuff! The PCB engineers probably had to stick with a two layer board and the hardware engineers dictated the physical size, so the PCB engineers were left trying stuff 10 lbs. of ICs in a 5 lbs. bag so to speak!
It's not OPAQUE. That would mean you couldn't see through it. It's TRANSLUCENT. You can see light through it but exact details are obscured. Otherwise it'd be TRANSPARENT. #TheMoreYouKnow
I share your concerns on dust, but I have seen even modern hard drives being opened and closed again on the desk and they still worked afterwards, so it is not impossible. An easy way to improve the situation is working with hands and drive inside plastic bags, because those are dust free when fresh and their static charge collects dust. More work is building an improvised laminar flow bench, which is basically a box that has a constant positive air flow pushed in by fans with HEPA filters at a rate that causes laminar flow to avoid whirls sucking in dirty air. That is typically used to work on optical systems to avoid dust particles.
Definitely ball bearing sound. Also, sound didn't change as the heads moved. I've taken apart tons of hard drives and put them back together and most(that were recoverable) came back to life no problem. I think the hard drive crash thing is overblown to be honest. I also think the cleanroom thing is a warranty thing as well.
I love how much engineering went thru 70 years since first hard drive to current 20TB ones still in the same 3.5'' form factor since 1980's - much more engineering than flash storage
I think it was something to do with the bearings, i've heard that sound many times before as an electrician I've also heard a headcrash on a 10 disk pack On the disk pack the squeal carried on while the disks were spinning because they were bouncing (one head crashed and sent vibrations down the common head carrier), that ruined at least 6 disks But with yours, the heads did the seek and came to rest, then the squeal started before you did anything.
it's funny how even modern computers 'get used' to being in a certain configuration and/or physical position. It's maddening how many times I've 'fixed' a pc just by putting it back how it was. Complicated systems are complicated
I need to look into more of your older videos.. the short recap of the Hawk just there reminded me how extremely similar it looked to the one in my Nixdorf 8850 system (also mid-70s made, 10MB fixed and 5MB removable platter that look identical to yours, including the dust cover "design"), so now I'm wondering whether Nixdorf rebranded CDC drives, or they had some sort of joint-venture business. Even the boards look quite similar, except for the connectors, where Nixdorf uses 100-pin 5-row staggered monstrosities. Sadly I can only go off pictures of the drive, as the physical one was trashed in a basement flood a decade ago. I've still got about a dozen of the removable cartridges left, though.. ;-)
Oh well, good I've read the comments!! I have 11 years old 2.5" WD laptop hdd and it sounds very weird, but still is working almost 24/7 and keeping data in my desktop. But I have another 3.5" WD HDD which has only 2 years of being online in vertical position in a PC case, after which it started vibrating quite noticeable. So had to reposition it horizontally.
If you have ever had an older vehicle _and_ a (front?) wheel bearing "went", then you would have recognised the sound. And yes, bearings over time will bed in based on the "verticality", and the "static-ness" of said position - FWIW this is why you _never_ "run-in" a new "vehicle engine" at a static/fixed RPM (especially if you want it to perform across the whole range) - It appears the CDC Drive Gods looked favourably on your "incursion in the space which only Hell occupies" .. :)
Nice Drive! Never seen this kind of hdd before. Sounds like a bad spindle bearing, know this from a IBM PS2 Model30 HDD wich generated a noise quite similar to this. Also, some Seagate ST225 suffered the same pronlem. Flipping over by 90 degrees fixed the issue for some time. If it was a head crash, the sound should have changed while the heads were moving over the platter.
Awe man, what a rollercoaster!?! I'm glad I'm not in your shoes, no offense, just seems nerve wracking. You've provided me with a lot of enjoyment from these videos. Great at work as background noise that isn't just dumb or distracting. Thank you, thank you. Good luck moving forward 👍
Once lubricated the screaming bearings of a old 5 1/2" drive with a few drops of motor oil from a Volvo 172! Ran it open for very long time without problems! =) Not recommended though! =)
I tend to agree with the ball bearing theory . . . but FWIW, in the days of servicing VHS video recorders, a very similar noise was caused by the phosphor bronze, head drum static discharge arm (same style as that which you removed to get to the fan) going dry. A tiny spot of graphite grease would effect a 100% cure.
If it was a case of head crash, there'd be a groove gouged into the surface of the platter, more than likely it's an issue with the bearings, you could try spraying the bearings with a little bit of WD40 and see if that improves things, my guess is that the bearing grease/lubrication has most likely dried-out by now.
After a certain point. Give or take. Some do have air in them. They also have a filter and a piece of pad? Looks like a small pillow that catches all the stuff flung off the platters. This is why even 5 years ago on a drive that was say 5+ years old. I was able to take the top off, unstick and put the cover back on. Then I Cloned the drive. After that it did die. I heard stories of old full height drives say 40 megs being unstuck and continued to be used. Providing it wasn't a head crash. They continued backing up data and since drives was $$$$ back then. It would be used for years until it would die.
All hard drives have some airflow internally, generated by the spinning platters. Modern drives are designed to channel this airflow through a small filter to catch any debris that may have been present at manufacture. For cleaning these out after contamination, use filtered compressed air, not any chemical stuff out of cans, unless it's rated for use with camera optics. Chemical air cans can leave residues on the surfaces. But yeah, it's bearings. This drive may be serviceable if you can get it apart in a cleanroom environment, unlike modern drives where they're all using FDBs and/or are epoxy-sealed and there's no way to get them apart without destroying them, and even if you could disassemble them, there are no new spares available anyway. There may be some DIY cleanroom box instructions on YT. I've seen a few online. They usually look like media blaster boxes without all the media blaster bits and with some fans with filters attached to provide positive pressure to keep dust out.
Sounds like a bearing and it needs oil or grease. I will guarantee its a bearing, same sound as my show model engine also without oil and the bearing will die from it. Thanks for sharing, love those old computers :)
I opened a seagate ST-225 once and it survived the non-cleanroom air.... but ST-225s are "indestructible" (mil. spec. ???) drives... oh good luck... and take it easy. ;)
20:56 No, you have an equalization port, remember? You got bit by grown tolerances and cold lubricants. A cold bearing will have greater internal tolerances than a warm one; it can also get sticky if the lubricant is degraded. When you powered it up cold on the flat, the bearings started moving without a side moment and started "wobbling" around the races with a resonance. This goes away when lubricant warms, and when you're operating in a mode (sideways) where to resonate the shaft has to climb in the race against a LOT of weight. Keep it on it's side and warm, it should be good for a few hundred more hours without any intervention with regards to the bearings. They will fail, though.
the odds that you've ruined that drive by opening it are actually low. we used to do it quite a lot back in the days with MFM and IDE drives, and it was never a problem.even the modern very high capacity drives usually survive long enough after being opened to have the data copied from them somewhere else. btw, the correct way to do it is in the bathroom, after leaving the shower running for 15-20 minutes. that's probably the most decent clean room environment one can get in amateur setting.
I used to manage a PDP11 based GenRad circuit testing machine in the late '80s that had a Pheonix drive and it was a beast! There was a strict spin-up and spin-down procedure for the drive and I remember having to slide out the drive and manually lock the heads and platters before the machine could be moved. When in use you could feel the whole machine shake every time the drive was accessed!
Dude! When you reassembled and retested the drive, it totally sounded like a bearing with worn races and/or improper pre-load. I'm not sure if you are aware, but bearing are designed for either the vertical or horizontal positions, sure you can take a vertical shaft bearing and use in the horizontal plane, but never the other way around. The axial load will cause excessive wear and premature failure......
Another vote here for the spindle grounding strap vibrating, the early 5.25" PC drives would do exactly the same thing and make an awful noise. Just loosen the screw and move the pad slightly so it's rubbing on a different spot. Or wait until it starts making the noise again and either press lightly on the strap or pull it back to prove it's that. I won't tell you on the PC's we just used to bend the strap away from the spindle, didn't seem to cause any problems.
If you have to take the cover off, do it with the platters spinning; this’ll throw any dust particles off by centrifugal force, though - likely - the thin cushion of air near the disk surface will blow away any dust particles before they even get to the surface. Likewise, put the cover back on before spinning down the drive, and store the cover face-down while it’s not on the drive, so any dust doesn’t settle on the interior portion. Also, take the cover off slowly so you don’t kick up a dust cloud. Ditto for putting it back on. HTH. 😌
Ground straps usually have a carbon tip, so it shouldn't be a worn strap making the noise; if it were, then the copper of the strap would be visible and there would be loads of carbon fragments. A CDC engineer once showed us that dust will usually not give you a head crash, it's such large particles that they are thrown off. Smoke, on the other hand, is sticky.
