This made my morning. I recall many weeks in my time as the IT guy in our large CPA firm where I was trying to recover clients data from proprietary legacy systems. Many in the office never really understood why I would get so excited when victory came. It's hard for non tech people to fully grasp the amount of failure modes that make a simple concept of recovering data so nuanced. Most of us watching were feeling it. Thanks
Thank you! It's so easy to think of things as big large collections. Like, we only really needed four things to work - the computer, the two finch drives and the hawk. That doesn't sound like much, but when stopping to really think about everything that was involved in just getting a little bit of data from one place to another, it was almost overwhelming!
That's the same thing as having a videocamera not working as intended and opening it up from the easiest parts/side and finding the fault, and hoping that it's the only defect, and upon switching it back on performing as intended... We once had someone who just had one of the worst days of life come into our office completely heartbroken because the camera this person had been using had written all the video files (of a big house fire) to the SD card in a corrupt fashion due to someone who used the same camera earlier pushing in an SD card the wrong way in, which bends a CRC pin from what we can find, it records without errors, but corrupts the whole card. Due to some other things this person just had it for the day and sat in our office crying, and I decided to let a recovery program (SanDisk's program was, as usual, useless for that task) work on the cards. Just as the person had gotten themselves back together a bit, I asked them to check the files it had already recovered, and when seeing what was coming back the person broke down crying yet again. I never felt like hugging someone but oof, that time I absolutely thought about doing so.
I'm a tech person and I still dont undertand his excitement and feelings, even tough it is a hard task. He got some data out of a disk, and can now delete it since he has no use for it :D Was the journey the reward?
@@marsovac Conversely, we have a ton of use for the data we saved versus the data we lost. My Finch drive was full of Oil and Gas backups for a local Oil Company in Texas. There weren't any applications, software, operating systems, etc. Just legions and legions of personal data. The MicroPlus Finch drive does have personal data on it, but it also has a ton of applications that we've never seen before, a full operating system, software packages, linkers, and just a ton of stuff that is extremely useful! Although, I think the misunderstanding is coming from potentially a lack of knowledge of the history of the Centurion computer. I've been working intimately with this machine for two years now, and we only know of four remaining in existence. The Centurion uses a completely custom CPU, architecture, storage system, and software. For example, the entire machine uses 400 bytes per sector to remain backwards compatible with data from Litton minicomputers from the 1960's. So, we found a bunch of software that was nearly lost to the sands of time. For some of these applications, the MicroPlus Finch drive quite possibly contained the last remaining copy of that piece of software in the world. That's why I was so excited! Also, this is my hobby, and my hobby brings me huge amounts of joy. Let people have fun!
On the "plugged the connector in the wrong spot" event. When I was stationed in Japan in the 1980s, we had a Data General Micro Nova we used with a radio system. We started having problems with it booting up from the floppy drive intermittently. We struggled with it for about a month with intermittent fail to boot problems. We swapped all the cards, and tried going through the troubleshooting guide several times. Keep in mind we were radio techs, not computer techs and we were not supposed to get too creative with the system. We finally called the main depot back in Sacramento, CA. They walked us through the same steps we had done ourselves to make sure we covered everything and finally determined they needed to send someone to Japan to assist us. a short while later they arrived, went to the machine, pulled it out of the rack, looked at the connector and moved it one position over. It turned out the guide didn't show which connector to use, and we were using the wrong one. Apparently one connector had a parity bit, and one didn't. We were using the non-parity bit connector so errors didn't get found and that was why it was intermittent. We ended up having to people travel from CA to Japan to have us move the connector over one spot.
Hah, well they're paid for the knowledge of knowing that the connector needed to be moved over one. Itemized bill: Moving connector: $1.00 Knowing where to move the connector: $9,999.00
You might consider finding a museum or archive as a place to store that personal data. Many of them will hold that kind of stuff for a hundred years before releasing it. You're right at the peak of the digital dark age, lots of stuff went digital but it has all become inaccessible. That kind of every day stuff about people's pay, shipping volumes and the like is invaluable to historians.
@@fred_brunell True, but that drive contained (as far as I understood) a lot of company data about bills, deliveries and even court cases that you don't want to end up at places.
@@Rob_65which is why I suggested it going to a museum dedicated to the purpose. The 50 year old data can sit for another 50, everyone mentioned will be long dead and then historians will be able to take a glimpse at the day to day of life a hundred years ago.
I always archive every drive I come across. These archives will be transferred to new media as new systems come out until I'm gone. Then they will go to a larger archive for the future. I have resurrected old hard drives that looked like they had been stored in a trash can in the garage for 20 years. I had to do a major physical cleaning before letting them into my office. My recovery software managed to get most files, leaving only a few unidentified files. I could tell by file sizes what some of the files might be, so changing the file extension allowed me to few a lot of them. One person was so happy I was able to get back their old photos and music.
Well done David, a tenacious, dragging the Finch drive kicking and screaming to a positive result, I am jealous, you transported us back to early 80's for 37 minutes, thanks alot David !
I worked as a 'field engineer' (mainframe service tech) in the late '70s; Immediately on hearing that HDA, I said to myself, "wow, those bearings sound crunchy!" Somehow, I knew it wasn't a head dragging - although I mightn't have risked flying the heads with the spindle in that rough a shape. Glad you got the data off!
Yup, that's a common sound for a ball bearing that's beginning to fail. I've heard plenty of bearings making that sort of noise on various machines. They can last quite a long time sounding like this. Very often when you take the bearing out it will look and feel fine, even though it sounds terrible at operating speed. If it progresses to a lower pitch rumbling or grinding noise you know the bearing is not far from spitting it's contents on the floor and needs to be replaced ASAP.
The 'warm it up' is a well known trick for very old drives with potentially dried out lubricants. I generally leave them on a 'hand warm' heater for like 8+ hours. This also reduces thermal stress on the electronics btw.
@@derpmellow2 You generally do not want to take a harddisk apart beyond possibly removing/replacing the pcb. Harddisks are assembled in an area with a very clean atmosphere and are typically sealed because even small bits of dust can cause a head crash. There are ways to do this without having a clean room, but those are still risky enough to avoid them if you can.
@@derpmellow2 Opening a harddrive and closing it up again and having it still working IS possible - then again you might as well wreck it completely due to dust contamination - impacts with grains of dust can actually physically destroy the heads themselves, at the very least losing all data stored on that disk surface, and quite possibly making the drive totally inoperable. Then you need to realize that the drive spindle bearings aren't necessarily exposed for you to lube them up either, and if you need to physically tear the drive down to get at them then you're never going to get any data out of it again, because you can't put it back together and have it still function. Not just a matter of contamination, it's also tolerances. Harddrives - even ones as old as these - are high precision instruments, and you just can't tear it down and put it back together and keep all the tolerances within spec. So consider the possibility of lubing up a prehistorical HDD a "no", for all intents and purposes. Harddrives are essentially consumables - always have been. You buy 'em, and they'll wear out eventually, since they're partly mechanical systems. These ancient ones even have carbon brushes for the spindle motor, so they're definitely going to wear out eventually - you can see all the carbon dust residue on the underside of the drive chassis after the PCB has been removed. That stuff is electrically conductive by the way and can cause shorts or other problems if it amasses in the wrong spots over time. Just one reason why modern drives use brushless motors... :)
If you are ever tracking down shorts in the future, a thermal imaging camera is a great tool to have. You can just power the board up with a current limited supply, and the offending component will stick out like a sore thumb because it will take all the current compared to all of the other components. That saves you from playing hide and seek with parts.
A thermal camera is definitely one of those things I need to invest in! I did think about just cranking the current up and whatever starts to glow or let out smoke is a pretty good indication, but I figured I would try probing around with the DMM first, haha.
@@UsagiElectric problem is other parts could be destroyed by other bad ones in a chain reaction destruction, i've had it happen in colour portable tvs in the 90s....
That is true. It could be a usable technique but it's dangerous. Some dead parts (i.e.e diodes) may fuse into a dead short so the part would have near-zero resistance, capable of sinking a few amps. The traces however are not, causing them to literally burn off the board. In this case the capacitor had
The tools you really need for tracking down shorts is an old school HP 546A Logic Pulser, and an HP 547A Current Tracer. You trace shorts with these with the power off. No FLIR required!
Even though it was all extremely sensitive personal data, I still hope that the original data on your Finch was backed up. It'd be an excellent example of the kind of data these machines would have been used for. In 300 years, the data won't be personal any more, because all the people and likely most of the businesses will be long gone, even possibly a few of the addresses themselves. But it can never reach that point if it isn't preserved now. Still extremely pleased that the data on the Vintage Geek Finch was backed up, as well. Backing up more data will never hurt!
@@KameraShy And of course, I would never want it released any time soon, not while that data could still affect people's lives, but it won't affect people's lives forever. At some point, it's all just history, as are we all. So I really, truly hope that, even though it's not on the Finch any more, it's somewhere backed-up. Only David really knows, I suppose.
@@KameraShyI had todo something similar on my 2009 Mac Pro when the 12 hour video during rendering ran out of room during rendering so many games and videos I did in the past had to be erased just to finish rendering that video. It bare squeaked by with15gb of had to spare.
So to answer your question directly, no, the data on my Finch was not backed up. Let me explain why I didn't. With regards to the Centurion and what it would have been used for, we've successfully backed up multiple Hawk platters, some with customer data and some with applications. While I don't agree with keeping personal data, we have specifically set some aside for preservation sake. At this point, storing more and more personal data doesn't help the project any at all, and conversely, makes me extremely uncomfortable. There was literally a workman's comp case followed by the company taking the injured worker to court to pay them less on the drive. It was all outlined with dates, names of the lawyers, and money amounts. That type of stuff makes me extremely nervous, and I do not want to have that information in my possession. So, if the goal was to have some information backed up so we can learn more about the system and how it was used, then we've already got that. If the goal was to just hoard hundreds of people's personal information for the sake of informing some archeologist hundreds of years in the future, well, I'll let that anxiety inducing task fall on someone else.
@@UsagiElectric Well, I am grateful that the stuff on the Hawk platters was backed up. I'll admit, however, that I don't understand why retaining that data makes you nervous. It's not the data, it's how you use it, and I, for one, would definitely trust you with stuff like that, because I know that you never release it or use it in any way. You seem the furthest type of person from that. I don't believe that just possessing the data can get you in trouble, but so long as there is something saved, then it's your machine and you can do what you want and I won't be upset. Thanks for all your hard work to lift the scene up.
I thought for sure that the heads crashed. It had that grindy, scraping sound. I didn't see a line on the platter though, so when you said it may be something else, I held out some hope =).
I have got to wonder if it wouldn't be useful to attempt to read the disk with a modern flux imaging system. That could attempt to read past the errors detected by the OS and allow the full platter image to be analyzed, and potentially discover if there are any more data hiding on the disk.
Regarding the tower of complexity, yes, we stand on the shoulders of giants who stood on the shoulders of giants, who stood on the shoulders of giants, etc. This episode reminded me of the times I've whacked an IDE hard drive, or balanced it at a certain angle to get it to spin up and copied data from it before it never worked again. Well deserved triumph there :)
Oh dear, me too. My mentor sysadmin when I was a wee PFY taught me to lift a Mac II cabinet and give it a good twist to unstick the drive bearings of the original Mac II harddrives - I think they were Apple-branded 20 or 40 MB Seagate 5¼" SCSI. IIRC he said they had been mounted wrong side up from the factory, which caused them to stick. I later improved the procedure by taking the lid off, unscrewing the drive, and just jerk the drive. Much safer for my back. I once rescued a disk containing lots of user data using the same method and immediately making a track by track SCSI copy to a good disk, then fixing it using Disk First Aid
Don't worry, that's on the docket to get fixed! I have a completely new front panel made of steel that I can weld proper mounting tabs to. Once I get some paint matched up to the original color, I'll spray it and replace the scratched one.
I did stick around and I did enjoy the journey! Next time (surely there will be…) please make a backup of the data onto modern media, even if contains personal identifiable data, because old software and data is almost as important as old hardware.
Agree. Destroying that data was reckless. It could easily have been held securely for future generations. (In fact it pretty much was, as who else could possibly read/copy it!)
It's so wild to me to think that the Centurion was this great unknown that we never thought we'd get going, and now it's become the most reliable (knock on wood) tool in the room!
The mighty old Centurion, the old man in the crisp, sharp blue business suit and tie, cane in one hand and a raised victory-fist in the other, "I'm being used for work again, at last! Hah, let's show those modern chump computers what an old feller can do! Just... make sure you look after me, and keep me maintained, eh? And... careful with me birds, they're delicate..."
I have literally no idea what is happening in this video, but it was a rollercoaster. Nice to see people's enthusiasm for topics outside my understanding.
@UsagiElectric As qualifier, I once helped fix one of these drives when just getting started in computing. When working on a very similar(if not identical) drive, we were hearing a very similar grinding sound. We thought it was a spindle motor problem until one of the company engineers walked in, told us it sounded like the platter-brake dragging and showed us. He locked the heads, disconnected the data cabling and turned the drive on it's side. He powered it up and sure enough the platter-brake shoe was not pulling away from the rotor surface. We took the solenoid off, took off the brake shoe, put the solenoid back in and powered it up again. No more grinding sound and the drive seemed to get right up to speed like it was supposed to. We left it that way and put it all back together. Powered it up and everything was back to normal except, no platter-brake. At that time replacement solenoid units were not available so the company just replaced the whole system with something new. So, I think it might be worth it to check the platter-brake on your unit. The solenoid might be stuck or have failed. Couldn't hurt as long as you lock the heads first. Keep in mind, my experience was 40ish years ago. Still, if I can help, it seemed like it was worth mentioning.
It was a long, but good day anyway. You are real techno-necromancer, bringing dead stuff to life again. These computers are about my age (born in late '60) and I'm still working with tech.
Race against the clock time again, eh? Nice cap fix. A nice thing I learned from watching @DiodeGoneWild is converting radial caps to axial by wrapping a cap with one of its leads bent back using kapton tape. Does the job really fine, lets you read the markings, keeps the cap nicely lying in place and provides insulation in case you don't have solder mask on the board. Nice meta-programming there! A shell script writing a Centurion script to the serial - thing of beauty, joy for ever. I love the storytelling in your videos. It's like reading a good book or watching a movie, with ups, downs and plot twists. You're obviously very good at it - and very entertaining too! Stay determined and Keri on :)
Man this was such a joy to watch, these days it's hard to be so enthusiastic about how a machine actually works, but you're rekindling that feeling a lot. Thanks!
Just looked up the original specification manual for the model 9410 Finch drive, it says the service life is 5 years or 30,000 hours, I think you might have exceeded that slightly lol. It also states, "Following an initial period of 200 hours, the Mean Time Between Failure shall exceed 5000 hours for units manufactured in the first year of production and 7500 hours for units manufactured in the second year. For units manufactured after the second year, the MTBF shall exceed 10,000 hours.". It's incredible you have managed to get any of them working, congrats!
Service life for electronics means hours of operation, and for hard drives it means how many hours the platters have been spinning. These Finch drives are old, but we don't know how much use they've actually had. They could well be within the service life (or alternatively far past it), but age is not kind to grease as we've heard haha.
I love it when a plan comes together. :-) Congratulations, bro. Feels great when it all works out, especially when you had to go to such extreme lengths as ripping apart 3 different drives to get the sketchy Finch going. I don't blame you for geeking out so hard at the end at ALL. Great stuff.
Thank you so much man! It was a wild ride, but all in, I started working/filming on Sunday, and on Thursday morning we had the data successfully backed up. I never would have guessed we could have gotten it going in less than a week!
Cannot believe I used to do all of this 35+ years ago and I was swapping components as an engineer to make it all work for customers. It was so natural and normal for me as my day job. Well actually i used to do this night and day for customers that paid for 24hr cover.
I give you a lot of credit. That was challenging. It sure is a great feeling when you finally succeed. Amazing to think this vintage hardware still works. It takes us back in time to when this was just modern hardware they were working with daily.
You did sooooo well!... I've had my own drives from the 1980's, stuff I'd long forgotten about, not these big drives, just drives from old IBM286's and Acorn BBC's and such, and honestly, to see some of it come up fully intact, programs I'd written was just amazing!... I can honestly relate to how you felt making this video...
I can barely follow what's going on with most of your videos, but your enthusiasm and excitement come through perfectly, and that's what keeps me coming back. Another great video.
Awesome! My heart was in my mouth several times during that. Sometimes when people ask me ‘why do you do this’ I have difficulty explaining that feeling you get when things go exactly right and you get a beep or a READY or something like that. Now you’ve encapsulated that feeling in a video I can use this to show them why I do what I do :D
Just found this channel and your collection reminded me of a '90s Eastern European comedy show ("Imanta-Babīte nepietur"). In one sketch, a GP recieves a patient and asks if he made the urine sample for testing. Patient confirms and takes out of his bag two full 3l jars full of pee. Doctor, eyes like dinner plates, asks if he brought some suitcase of fecal samples too, which patient confirms and slaps a big old suitcase on the table. So this '80s tech have similar vibes today for some..)) However, without experiencing much of it myself, i too have respect and some phantom nostalgia for it. As a normal guy born in USSR, i still prefer the miniturization, as long as it's sturdy design. We didn't have commercial consumer PCs back in the '80s (unless you risked paying 10× the price to some smugglers, same in beginning of the '90s). My dad did get the B&W soviet IBM equivalent in the first half of '90s, with 8in(?) floppy and i think 84K of RAM from the national TV he still works at. A bit bulky and heavy as hell. If i had it still, i'd send it to this guy. Unfortunately it got lost after someone cleaned out the parishe's storage space where it was sitting after apparent catastrophic failure. ["ПКП" on the screen - not in any manuals, without any other signs of life or bios, though no visible damage inside. No specialist could tell what's up at the time, i'd be surprised if anyone knew here].
Amazing. I've never heard of any of this equipment, but it sure looks well built and will probably last another several decades! Good for people like you who can figure these systems out and show people like me how they worked, what they were for, and why they matter!
I manage storage systems with hundreds of SSDs (and/or HDDs, depending on the system) operating in synchronicity. The mind boggles that any of it works. It's basically magic. Congrats on getting the data!
The modern hard drives these days are incredibly complex machines. The hard drives from the early 80's were straight forward and basically much simpler.
I don't think my anxiety has ever gotten so high by just watching a clip As a trucking computer tech with 2 bachelor's degrees that just loves old tech this made me a kid in a candy store but make me feel like a dad watching his kids play in the middle of the highway at the same time. Thanks for the peefect video. If Hollywood videos had as much excitement and suspense as this clip does, it would be sold out for weeks.
Putting the drive and PSUs on the wood brought back memories. When I was a kid without money, I would often get PC parts, but wouldn't get a case and, since I couldn't afford one, I'd end up building PCs with everything attached to a piece of plywood. A lot of times, I didn't have switches either, so I'd turn them on with a screwdriver. My friend used one as a dialup proxy machine for 5 years before it needing a reboot. Although I've worked on much bigger stuff since then, I always consider those machines my greatest builds because I really did scrap them together from nothing and they lasted quite a while. Really love that your setup got that data off the drive. Cool stuff!
15:57 "it suddenly got quiet": same thing happened once with my PC fan. It was making horrible noises and then noises suddenly stopped. I thought it is dead. But nope, it continued spinning, but quietly.
New to the channel... this was great, thanks. I shared your sentiments at the end throughout the whole process. It is impressive just how robust this old tech really is for the most part.
Insanely long? lol I watch my man, Adrian Black, so this was nothing! I love his long vids and loved this one as well. His disbelief and joy of bringing old machines to life again like you did is my favorite part. Y’all are such treasures and I know the original engineers of these processing machines would cry tears of joy if they saw y’all’s work. Also, love from another Texan. I’m in between Austin and San Antonio so maybe somewhere somehow I’ll meet you.
This randomly came up on my suggestions and is completely different from anything else I watch on TH-cam. However, I gave it a try and I'm absolutely hooked. I've even subscribed because this brings back distant memories of a misspent childhood in front of old computers (I'm not quite old enough to have used these systems though!). Excellent stuff!
Mate I wouldn't have wiped that original drive but sadly it's done now. Your dealing with Electronic Archaeology here & the more you can save the better no matter what it is including data from any machine. I on the other hand have gone back in time quite a bit further to the Valve era looking at doing repairs to some ancient valve gear going again which is usually dried out caps & sometimes a resistor or 2. The valves are usually live by the amount of hours it's been run. It's lucky you have the skill to save these antiques. Keep up the good work
This brings back some memories of recovering data from hard drives. I could even spin up frozen 5 1/4" disks by twisting them during power on to release the stick slip on the spindle. It was quite common for disks that had been used in servers (running 24/7) to just freeze up as soon as they were stopped for a few days.
Congratulations, some awesome work here, very glad to see more and more of the legacy of this small branch of computing being saved! Your retrospective at the end puts into perspective what we sometimes miss, just how much has to work to make all this stuff (that no matter how well made is well past its design lifecycle), do its thing again.
Amazing video, the fact that data was able to be copied and there was cool programs is important for the archival of the machine itself. But I also thing that user data is also interesting. I still think old data should be saved and left in some corner on some disk, no need to put it on the internet for all to see yet. Maybe in 50y the PII won't be a problem, everyone in the list would be dead. I was seeing dates like 1995, people it was referring to might still be alive. For archeology reasons in 500y if it survives, it would be funny to see the logs of a company. It's sad private data about companies don't get left anymore , there won't be anything for archeology to see about how companies in 2020 actually ran. Just like clay tablets said a lot about the first corporations in the old civilizations, but once they where, with old HDDs we have a glimpse of 1995, but after that it's going dark. Imagine in 2100 wanting to do archeology on this old defunct company called Google. I for sure am donating every data I collected when I was alive if I die, I don't even care about the origins and that it was private data of mine or others. You can have it as long as you promise to keep it for archeological reasons. And I have a lot, like a lot of copyright material that's probably not even in the internet anymore. Like at least 20TB of "proprietary" things. It's going to be all public when I die. I can't be sued if I'm dead. Why would I care about my PII data if I'm dead. I think everyone's data should become archived when they die and then become public a couple of decades after they died. But I do understand the liability of holding PII data nowadays. Not saying you shouldn't have deleted it, it's not very useful for content on this internet television website. I'm just saying what I usually do when I find fun things on disks.
I agree about saving the data anyway. Just encrypt it if you're concerned about accidental leaks. I also agree that some people might still be alive considering I was born in '95 and I'm only 28 lol
Video's like these make me appreciate modern computers in a way: Not having to explicitly know what disk to target or what partition and type out commands is a whole lot less cumbersome, but compared to my first computer (A 486, not sure what kind, it was a business-line HP Vectra from my dad's work that passed it's financial lifespan) I weirdly enough still miss the days of being on an offline computer and having to make sure your IRQ's were set properly for something as seemingly simple as a sound card. The idea to now just have to plug in a disk that is having problems and clicking a recovery program is wild: the amount of technological progress we made with things like this is insane when you think about it.
Ikr. It's so crazy to me as somebody who grew up with XP and 7. By then PCs were pretty much just plug and play. That's probably a reason why many people buy microcontrollers and SBCs, to get that feeling of being in direct control of the machine. It's the equivalent of driving a manual transmission.
Thank you! Now that we have the data backed up, the big question is where do we go with the drive from here. To do the level of repair required for the bearings is really risky as we can potentially lose alignment with the soft-sectoring servo track, and if we lose that, there is no recovering. So, I need to talk with Aaron a bit and see where he wants to head next with it!
@@UsagiElectricI wasn't suggesting attempting any mechanical related repairs. The drive to be rescued had at least one fault preventing it from working. Perhaps you already fixed it with the incremental head seek test. You could replace the PCBs with the original ones and see if the drive works fully. If not, you can compare signals between those on the PCBs of the drive that needed backing up with those of the signals of the donor drive PCBs as they seem to be similar sets of boards.
While I don't understand everything said and done in this video, I am glad my start in computing, in 1988, began with finding a way to make a 286 with 30 meg hard drive running DOS 3.1 useful in a video production office. Many books were bought and read along the way, and enough of that now useless information gave me enough insight to follow along with what you did in this video.
I remember that mainframes and desktops split due to physically being too big. I see now that desktops need to do a split. We can do cell phones, laptops, and tablets; but the workstation desktops are facing a forced breakup right now.
In large enterprises there is a shift back towards "terminal" like computers by using virtual desktop environments. A low powered computer is used to log into a virtual desktop server and all the processing and storage is done remotely while the screen information is sent back.
Great video , i,m only 27 years old but i love the older elektronics/computers i still cant belive you got that old drive to work !. i hope one day i have the space and time to take on similar projects, i know the feeling of getting something to work thats basically dead and its the best feeling ever
Amazing as usual! I just had to go through some headaches with a dying drive in my personal PC, fun to see the parallels between what I went through in 2023 vs what you would've had to go through 40 years ago, the more things change the more they stay the same
I love the age we live in today. Look at the size of that beast and 32Mb. What a small, small capacity for such a big drive. It's just wonderful how we have teeny weeny NVME cards with 4TB and ssd drives with 8Tb and even thumb drives with 256gigs of space. Love the video, thank you for letting us see how far we have come since 1980. WOW 1 hour to format a 32Mb drive boy we really have come along way since 1980. One last comment I'm glad someone else needs to check weather or not they are coping the right disk from the right source I'm always double checking when I do it too.
A surprising result, especially considering how much of the process can be summarized as "I know it's erroring, but we're gonna try to read it anyway." Well done!
Congratulations for this achievement well beyond your own expectations and a big thank you for keeping me so well entertained watching you fighting your way through this. I love old computers and weird operating systems and well, this one gets a high grade in the weirdness class! My favourite retro computing TH-cam channel for sure.
Great job! It's a good day when all the hard work pays off. Although errors on the "good" 32MB finch still trouble me. We really. really need some kind of SD-card drive emulator or similar for Centurion. I think this should be doable. If only I had enough spare time to help...
Thank you! Whenever S.NEWDSK runs, it makes a bad sector table that lives on the drive, so all the bad sectors are known by the OS and it can write around them. We're working on emulation options for the Centurion. It took about a month of work, but we did manage to get a floppy emulator working (the Centurion uses very custom encoding and timing that did not play well). We can now use that as a springboard for Finch emulation, as it's mostly like a Floppy with much faster data rates.
@@UsagiElectric Thanks a lot for replying. it SHOULD mark the bad blocks, but if there are many in same spot the whole area is likely marginal and likely quite unreliable - at least this is how DOS era hard-drives are. Floppy emulation is excellent news! I applaud you and everyone who made that possible. I think all of your viewers would really appreciate if you could give us occasional updates on community development progress - emulation and all. Thank you and stay awesome!
Probably not useful for this occasion... but if you ever come across some damaged media like an HDD and can connect it to a machine which can boot Linux, do this. There's a tool called dd_rescue which is by far the best option to get the most data from a damaged device as gentle as possible. It basically copies data with large requests, until it encounters an issue. Then it just jumps in the center of the unread data and continues there, avoiding stressful rereads of the damaged data over and over again. This will continue until it's on the end of each unread block, then it jumps in the center of the largest remaining unread block. And repeat and repeat and repeat. This way it avoids the damaged areas as long as possible. After which it reduces the request size down to just a single sector and repeats the reading of the previously large areas with read issues, but only tries each sector once. After this pass most readable data will be read. But you can let it run continuesly. It will repeat the reads one damaged sector at a time, and I found it's often lucky and will get most of the damaged data anyway from the disk. But those operations tend to get stressful for the drive, as it needs to seek a lot. So it may die in the process.
To give a perspective: I tried to copy a hard drive which had some damaged areas with a windows tool, and it got to the first damaged area and got stuck there for two days. Then I terminated it. dd_rescue got more than 99% of the data from the harddrive within around 2 hours and then started the first retry pass. In just a couple of hours it got everything out of the hard drive, except ~10 sectors. After this experience I've never looked back and saved a couple of dozens or so broken harddrive data with it. Worked great every time.
I don't know how I stumbled on this video but it's certainly top-notch nerd pr0n. There's nothing like spending days or even weeks cursing every minute of the mountain you are climbing, and then finally through unbreakable persistence, you get the win. As I get older and earn more and more patience, the wins feel even better when they come along. Thanks for sharing this one... some of us know how long this 37 minute video REALLY took. hehe
I love data archival, and I hope you can get that error fixed on that source drive now that the data has been mostly preserved and you are free to rewrite it.
I think just running S.NEWDSK on the Source disk would totally solve the error! Most likely, it's developed a bad sector or track somewhere, and S.NEWDSK detects those and creates a bad sector table that lives on the drive, so the OpSys can write around them. But, the bigger issue is the bearings. I'm not sure how we can get the bearings serviced without losing alignment with the soft-sectoring servo track on the disk. If we can get the bearings repaired, then I think we could totally bring the drive back up to reliably working!
@@UsagiElectric As weird as it is, see if you can find a Data Recovery Center or HDD repair place that has a clean room; maybe they can totally pull the drive apart, clean and service it, clean and relubricate the bearings?
Amazing stuff, I really enjoy this channel, never heard of centurion or finch drives before. It's nice to learn about this stuff and how it worked before modern era of computing.
Your videos are the absolute highlight of my day! We exhibited across from each other at VCF east. I had the silly mac classroom setup. This journey has been just so much fun!
Absolutely love archival recovery projects. I remember many moons ago working on a project to recover data from a machine I can only remember being called an "Ego". Was a sort of single box minicomputer thing with a couple terminals, booted from ROM and had a removable 5Mb (I think) disk. Went through a process of dumping data out the centronics parallel port, via a parallel to serial converter to an RS232 port on a PC hahaha.
i don't know what bearings are in those drives, but if they are sintered bronze bushings, then it sounds like they are dry or worn If the bronze is worn, the shaft rattles around, if it's not too worn, a little oil will work (at leats for a while)
i do enjoy the videos on this channel, despite 40 years of computer use, starting with spectrum in 1982, for my 8th birthday, i only started to learn electronic stuff earlier this year, with a box full of components and a soldering iron i have had a wonderful time, ive made a driver to power laser diodes, and set up an optical bench, ive made a zvs circuit to drive an old LOPT, ive chopped the deflection coils on an old CRT b and w tv to make a vector-scope. despite not having a treasure trove of ancient computers per se, i still get so many general tips from watching this bloke. thanks. and btw way is it called USAGI?
I once had to copy an 8-inch floppy with good data onto an UNFORMATTED 8-inch floppy to back it up. This was on a CPM machine. I wasn't paying attention and put my floppies in backwards in the drives. Since I was doing a sector-by-sector copy, the system dutifully copied the unformatted floppy over top of my good floppy. Now I am crazy paranoid about deleting data of any kind 🙂
I think the type of harddisks my dad used to work with might be in this video, he once deleted the whole company's several years worth of archiving of theater documentation (song-texts, information about where to find decor pieces or clothing items, he worked at a conservation company) due to a faulty bulb he tried troubleshooting by exchanging bulbs. In the meantime, he also tried something else, and mistook the good disk for the wrong one, and destroyed the backup, with the original work-disk being non-functional at the time. When my dad was with the company for 25 years and it was celebrated, the then-head of the company had a little speech about how he was not easy to be moved about big changes in the company, but "I was told that one time he made a big mistake and destroyed years of work due to a faulty bulb, and it was the first time that he decided to take a walk, looking slightly more pale than he usually was..."
We all do this exactly once. It may not be in the same physical format, but there's always exactly one clone job every PC tech that gives them this lesson.
I knew someone who did the same with one of the CDC hawk like 500mb drives. He was supposed to do an overnight backup, but backed the empty or"scratch" drive onto the full working drive, wiping all of the data. He was called scratch from then on😊
Fantastic video, and you really took us along for the emotional ride. Well done on recovering that data. I do a lot of searching for short circuits as part of my job, and here are my tips, in case they're useful. You can use a Toneohm, a sensitive low-ohmmeter with an audible output so you don't need to look at the screen, to measure the resistance at various places and you very quickly get an idea of where the fault is. Or if it's a power rail, apply power from a current-limited power supply (current depends on the circuit, but somewhere between 0.1 and 10 amps may be appropriate) and then measure the voltage across the power rails in various places with a high-resolution (5 or 6 digits) voltmeter. You can also measure the voltage drop from supply to various points along each supply rail. That will also quickly give you an idea of where the shorted part is.
21:49 Argh! It sounds like evil! I cringed my way through this video whilst hugging my self in hope. Well done for thrashing this through to a great result!
Well, it was the transistor that killed vacuum tube computers. Vacuum tubes just didn't last long enough, and when you have hundreds to thousands of tubes, and a single tube failing can kill the machine, well there you go. Transistors just last longer (when they got perfected, early transistors were a nightmare). Especially in low density IC's. Those 74 series were workhorses, that's for sure.
This made my morning. I recall many weeks in my time as the IT guy in our large CPA firm where I was trying to recover clients data from proprietary legacy systems. Many in the office never really understood why I would get so excited when victory came. It's hard for non tech people to fully grasp the amount of failure modes that make a simple concept of recovering data so nuanced. Most of us watching were feeling it. Thanks
Thank you!
It's so easy to think of things as big large collections. Like, we only really needed four things to work - the computer, the two finch drives and the hawk. That doesn't sound like much, but when stopping to really think about everything that was involved in just getting a little bit of data from one place to another, it was almost overwhelming!
That's the same thing as having a videocamera not working as intended and opening it up from the easiest parts/side and finding the fault, and hoping that it's the only defect, and upon switching it back on performing as intended...
We once had someone who just had one of the worst days of life come into our office completely heartbroken because the camera this person had been using had written all the video files (of a big house fire) to the SD card in a corrupt fashion due to someone who used the same camera earlier pushing in an SD card the wrong way in, which bends a CRC pin from what we can find, it records without errors, but corrupts the whole card.
Due to some other things this person just had it for the day and sat in our office crying, and I decided to let a recovery program (SanDisk's program was, as usual, useless for that task) work on the cards.
Just as the person had gotten themselves back together a bit, I asked them to check the files it had already recovered, and when seeing what was coming back the person broke down crying yet again.
I never felt like hugging someone but oof, that time I absolutely thought about doing so.
I'm a tech person and I still dont undertand his excitement and feelings, even tough it is a hard task. He got some data out of a disk, and can now delete it since he has no use for it :D
Was the journey the reward?
@@marsovac Conversely, we have a ton of use for the data we saved versus the data we lost. My Finch drive was full of Oil and Gas backups for a local Oil Company in Texas. There weren't any applications, software, operating systems, etc. Just legions and legions of personal data. The MicroPlus Finch drive does have personal data on it, but it also has a ton of applications that we've never seen before, a full operating system, software packages, linkers, and just a ton of stuff that is extremely useful!
Although, I think the misunderstanding is coming from potentially a lack of knowledge of the history of the Centurion computer. I've been working intimately with this machine for two years now, and we only know of four remaining in existence. The Centurion uses a completely custom CPU, architecture, storage system, and software. For example, the entire machine uses 400 bytes per sector to remain backwards compatible with data from Litton minicomputers from the 1960's.
So, we found a bunch of software that was nearly lost to the sands of time. For some of these applications, the MicroPlus Finch drive quite possibly contained the last remaining copy of that piece of software in the world. That's why I was so excited!
Also, this is my hobby, and my hobby brings me huge amounts of joy. Let people have fun!
@@UsagiElectricWell said. That's the difference between an Engineering mind set and one that isn't. Great Video.
On the "plugged the connector in the wrong spot" event. When I was stationed in Japan in the 1980s, we had a Data General Micro Nova we used with a radio system. We started having problems with it booting up from the floppy drive intermittently. We struggled with it for about a month with intermittent fail to boot problems. We swapped all the cards, and tried going through the troubleshooting guide several times. Keep in mind we were radio techs, not computer techs and we were not supposed to get too creative with the system. We finally called the main depot back in Sacramento, CA. They walked us through the same steps we had done ourselves to make sure we covered everything and finally determined they needed to send someone to Japan to assist us. a short while later they arrived, went to the machine, pulled it out of the rack, looked at the connector and moved it one position over. It turned out the guide didn't show which connector to use, and we were using the wrong one. Apparently one connector had a parity bit, and one didn't. We were using the non-parity bit connector so errors didn't get found and that was why it was intermittent. We ended up having to people travel from CA to Japan to have us move the connector over one spot.
Definitely not the first time somebody has had to travel just to move a plug. Happens a lot in IT lol.
@@SockyNoobwith this case covering a distance of around 5500 miles it might hold the record for something that simple, though
I bet the tech from Sacramento told that story for the rest of his life.
Hah, well they're paid for the knowledge of knowing that the connector needed to be moved over one.
Itemized bill:
Moving connector: $1.00
Knowing where to move the connector: $9,999.00
You might consider finding a museum or archive as a place to store that personal data. Many of them will hold that kind of stuff for a hundred years before releasing it.
You're right at the peak of the digital dark age, lots of stuff went digital but it has all become inaccessible. That kind of every day stuff about people's pay, shipping volumes and the like is invaluable to historians.
I was thinking the same thing, he's working so hard to retrieve data from one drive, and totally flushes another drive.
@@fred_brunell True, but that drive contained (as far as I understood) a lot of company data about bills, deliveries and even court cases that you don't want to end up at places.
@@Rob_65which is why I suggested it going to a museum dedicated to the purpose. The 50 year old data can sit for another 50, everyone mentioned will be long dead and then historians will be able to take a glimpse at the day to day of life a hundred years ago.
@@zyeborm Yeah probably a good idea, its a common thing finding old drives with data on them, probably good idea to archive it.
I always archive every drive I come across. These archives will be transferred to new media as new systems come out until I'm gone. Then they will go to a larger archive for the future. I have resurrected old hard drives that looked like they had been stored in a trash can in the garage for 20 years. I had to do a major physical cleaning before letting them into my office. My recovery software managed to get most files, leaving only a few unidentified files. I could tell by file sizes what some of the files might be, so changing the file extension allowed me to few a lot of them. One person was so happy I was able to get back their old photos and music.
Well done David, a tenacious, dragging the Finch drive kicking and screaming to a positive result, I am jealous, you transported us back to early 80's for 37 minutes, thanks alot David !
Thanks!
It put one heck of a fight, but we managed to come out on top!
Yeah, we're gonna have to start calling David Tenacious D! 😄
Nice one, don't know if David remembers the D @@horusfalcon
@@horusfalcon This is not the greatest hard drive in the world, this is just a tribute :D
@@UsagiElectric Hahahahahahaha!
I worked as a 'field engineer' (mainframe service tech) in the late '70s; Immediately on hearing that HDA, I said to myself, "wow, those bearings sound crunchy!" Somehow, I knew it wasn't a head dragging - although I mightn't have risked flying the heads with the spindle in that rough a shape. Glad you got the data off!
Yup, that's a common sound for a ball bearing that's beginning to fail. I've heard plenty of bearings making that sort of noise on various machines. They can last quite a long time sounding like this. Very often when you take the bearing out it will look and feel fine, even though it sounds terrible at operating speed.
If it progresses to a lower pitch rumbling or grinding noise you know the bearing is not far from spitting it's contents on the floor and needs to be replaced ASAP.
The 'warm it up' is a well known trick for very old drives with potentially dried out lubricants. I generally leave them on a 'hand warm' heater for like 8+ hours. This also reduces thermal stress on the electronics btw.
could you not just regrease the drive?
i know nothing about these old drives, but they put it together so could you not take it apart?
@@derpmellow2 You generally do not want to take a harddisk apart beyond possibly removing/replacing the pcb.
Harddisks are assembled in an area with a very clean atmosphere and are typically sealed because even small bits of dust can cause a head crash.
There are ways to do this without having a clean room, but those are still risky enough to avoid them if you can.
@@c128stuff understood, so it is doable but not recommended.
@@derpmellow2 Its something to avoid at all cost unless there really is no other option, you know what you are doing, and have the setup to do it.
@@derpmellow2 Opening a harddrive and closing it up again and having it still working IS possible - then again you might as well wreck it completely due to dust contamination - impacts with grains of dust can actually physically destroy the heads themselves, at the very least losing all data stored on that disk surface, and quite possibly making the drive totally inoperable.
Then you need to realize that the drive spindle bearings aren't necessarily exposed for you to lube them up either, and if you need to physically tear the drive down to get at them then you're never going to get any data out of it again, because you can't put it back together and have it still function. Not just a matter of contamination, it's also tolerances. Harddrives - even ones as old as these - are high precision instruments, and you just can't tear it down and put it back together and keep all the tolerances within spec.
So consider the possibility of lubing up a prehistorical HDD a "no", for all intents and purposes. Harddrives are essentially consumables - always have been. You buy 'em, and they'll wear out eventually, since they're partly mechanical systems. These ancient ones even have carbon brushes for the spindle motor, so they're definitely going to wear out eventually - you can see all the carbon dust residue on the underside of the drive chassis after the PCB has been removed.
That stuff is electrically conductive by the way and can cause shorts or other problems if it amasses in the wrong spots over time. Just one reason why modern drives use brushless motors... :)
If you are ever tracking down shorts in the future, a thermal imaging camera is a great tool to have. You can just power the board up with a current limited supply, and the offending component will stick out like a sore thumb because it will take all the current compared to all of the other components.
That saves you from playing hide and seek with parts.
A thermal camera is definitely one of those things I need to invest in!
I did think about just cranking the current up and whatever starts to glow or let out smoke is a pretty good indication, but I figured I would try probing around with the DMM first, haha.
@@UsagiElectric problem is other parts could be destroyed by other bad ones in a chain reaction destruction, i've had it happen in colour portable tvs in the 90s....
That is true. It could be a usable technique but it's dangerous.
Some dead parts (i.e.e diodes) may fuse into a dead short so the part would have near-zero resistance, capable of sinking a few amps.
The traces however are not, causing them to literally burn off the board.
In this case the capacitor had
The tools you really need for tracking down shorts is an old school HP 546A Logic Pulser, and an HP 547A Current Tracer. You trace shorts with these with the power off. No FLIR required!
@@UsagiElectric Get one of those phone InfiRay ones, fairly cheap, cheerful and quite useful.
Even though it was all extremely sensitive personal data, I still hope that the original data on your Finch was backed up.
It'd be an excellent example of the kind of data these machines would have been used for.
In 300 years, the data won't be personal any more, because all the people and likely most of the businesses will be long gone, even possibly a few of the addresses themselves.
But it can never reach that point if it isn't preserved now.
Still extremely pleased that the data on the Vintage Geek Finch was backed up, as well. Backing up more data will never hurt!
I thought it was a HUGE mistake to delete that data. It is (now was) an historical time capsule.
@@KameraShy And of course, I would never want it released any time soon, not while that data could still affect people's lives, but it won't affect people's lives forever. At some point, it's all just history, as are we all.
So I really, truly hope that, even though it's not on the Finch any more, it's somewhere backed-up.
Only David really knows, I suppose.
@@KameraShyI had todo something similar on my 2009 Mac Pro when the 12 hour video during rendering ran out of room during rendering so many games and videos I did in the past had to be erased just to finish rendering that video. It bare squeaked by with15gb of had to spare.
So to answer your question directly, no, the data on my Finch was not backed up.
Let me explain why I didn't.
With regards to the Centurion and what it would have been used for, we've successfully backed up multiple Hawk platters, some with customer data and some with applications. While I don't agree with keeping personal data, we have specifically set some aside for preservation sake. At this point, storing more and more personal data doesn't help the project any at all, and conversely, makes me extremely uncomfortable.
There was literally a workman's comp case followed by the company taking the injured worker to court to pay them less on the drive. It was all outlined with dates, names of the lawyers, and money amounts. That type of stuff makes me extremely nervous, and I do not want to have that information in my possession.
So, if the goal was to have some information backed up so we can learn more about the system and how it was used, then we've already got that. If the goal was to just hoard hundreds of people's personal information for the sake of informing some archeologist hundreds of years in the future, well, I'll let that anxiety inducing task fall on someone else.
@@UsagiElectric Well, I am grateful that the stuff on the Hawk platters was backed up.
I'll admit, however, that I don't understand why retaining that data makes you nervous. It's not the data, it's how you use it, and I, for one, would definitely trust you with stuff like that, because I know that you never release it or use it in any way. You seem the furthest type of person from that.
I don't believe that just possessing the data can get you in trouble, but so long as there is something saved, then it's your machine and you can do what you want and I won't be upset.
Thanks for all your hard work to lift the scene up.
What a ride, I was on the edge of my seat with that finch sounding like it was a box full of rocks
Thank you! The first time it spun up, I was just sure that we wouldn't be able to rescue it, but against all odds, the drive hung in there!
I thought for sure that the heads crashed. It had that grindy, scraping sound. I didn't see a line on the platter though, so when you said it may be something else, I held out some hope =).
I have got to wonder if it wouldn't be useful to attempt to read the disk with a modern flux imaging system. That could attempt to read past the errors detected by the OS and allow the full platter image to be analyzed, and potentially discover if there are any more data hiding on the disk.
I think the controller and interface are proprietary, so you would need to make a custom device to perform the flux imaging.
I know nothing about computers, but it's fun watching you work on those old machines
Regarding the tower of complexity, yes, we stand on the shoulders of giants who stood on the shoulders of giants, who stood on the shoulders of giants, etc. This episode reminded me of the times I've whacked an IDE hard drive, or balanced it at a certain angle to get it to spin up and copied data from it before it never worked again. Well deserved triumph there :)
Oh dear, me too. My mentor sysadmin when I was a wee PFY taught me to lift a Mac II cabinet and give it a good twist to unstick the drive bearings of the original Mac II harddrives - I think they were Apple-branded 20 or 40 MB Seagate 5¼" SCSI. IIRC he said they had been mounted wrong side up from the factory, which caused them to stick. I later improved the procedure by taking the lid off, unscrewing the drive, and just jerk the drive. Much safer for my back.
I once rescued a disk containing lots of user data using the same method and immediately making a track by track SCSI copy to a good disk, then fixing it using Disk First Aid
That scratch in the floppy front panel is driving me crazy
Don't worry, that's on the docket to get fixed! I have a completely new front panel made of steel that I can weld proper mounting tabs to. Once I get some paint matched up to the original color, I'll spray it and replace the scratched one.
There are car repair products that can help you buff out the scratch on the existing front panel.
I did stick around and I did enjoy the journey! Next time (surely there will be…) please make a backup of the data onto modern media, even if contains personal identifiable data, because old software and data is almost as important as old hardware.
Agree. Destroying that data was reckless. It could easily have been held securely for future generations. (In fact it pretty much was, as who else could possibly read/copy it!)
As a backup fanatic, I can only agree, and add that some of the backup should be off-site.
Exactly, though purging sensitive data is the best thing to do. But never purge software and other types of files!
Seems your mighty Centurion has become your lab assistant. So many of its brethren will most certainly be saved thanks to your excellent work here!
It's so wild to me to think that the Centurion was this great unknown that we never thought we'd get going, and now it's become the most reliable (knock on wood) tool in the room!
The mighty old Centurion, the old man in the crisp, sharp blue business suit and tie, cane in one hand and a raised victory-fist in the other, "I'm being used for work again, at last! Hah, let's show those modern chump computers what an old feller can do! Just... make sure you look after me, and keep me maintained, eh? And... careful with me birds, they're delicate..."
I have literally no idea what is happening in this video, but it was a rollercoaster. Nice to see people's enthusiasm for topics outside my understanding.
@UsagiElectric
As qualifier, I once helped fix one of these drives when just getting started in computing. When working on a very similar(if not identical) drive, we were hearing a very similar grinding sound. We thought it was a spindle motor problem until one of the company engineers walked in, told us it sounded like the platter-brake dragging and showed us. He locked the heads, disconnected the data cabling and turned the drive on it's side. He powered it up and sure enough the platter-brake shoe was not pulling away from the rotor surface. We took the solenoid off, took off the brake shoe, put the solenoid back in and powered it up again. No more grinding sound and the drive seemed to get right up to speed like it was supposed to. We left it that way and put it all back together. Powered it up and everything was back to normal except, no platter-brake. At that time replacement solenoid units were not available so the company just replaced the whole system with something new. So, I think it might be worth it to check the platter-brake on your unit. The solenoid might be stuck or have failed. Couldn't hurt as long as you lock the heads first. Keep in mind, my experience was 40ish years ago. Still, if I can help, it seemed like it was worth mentioning.
Thank you Ren & the Centurion team.
Well done! Have been through board swaps on hard drives back in the 1980s, I can identify how good that “YES!” moment feels when it works.
Very well done! You and everyone who has pitched in are legends getting all this working.
Thank you!
It was a long, but good day anyway. You are real techno-necromancer, bringing dead stuff to life again. These computers are about my age (born in late '60) and I'm still working with tech.
Race against the clock time again, eh?
Nice cap fix. A nice thing I learned from watching @DiodeGoneWild is converting radial caps to axial by wrapping a cap with one of its leads bent back using kapton tape. Does the job really fine, lets you read the markings, keeps the cap nicely lying in place and provides insulation in case you don't have solder mask on the board.
Nice meta-programming there! A shell script writing a Centurion script to the serial - thing of beauty, joy for ever.
I love the storytelling in your videos. It's like reading a good book or watching a movie, with ups, downs and plot twists. You're obviously very good at it - and very entertaining too! Stay determined and Keri on :)
Man this was such a joy to watch, these days it's hard to be so enthusiastic about how a machine actually works, but you're rekindling that feeling a lot. Thanks!
Just looked up the original specification manual for the model 9410 Finch drive, it says the service life is 5 years or 30,000 hours, I think you might have exceeded that slightly lol. It also states, "Following an initial period of 200 hours, the Mean Time Between Failure shall exceed 5000 hours for units manufactured in the first year of production and 7500 hours for units manufactured in the second year. For units manufactured after the second year, the MTBF shall exceed 10,000 hours.". It's incredible you have managed to get any of them working, congrats!
Service life for electronics means hours of operation, and for hard drives it means how many hours the platters have been spinning.
These Finch drives are old, but we don't know how much use they've actually had. They could well be within the service life (or alternatively far past it), but age is not kind to grease as we've heard haha.
Love your personality and energy, this was a fantastic ride! Congratulations, excellent work!
Thank you so much!
A Trip down memory lane
This just reminded me of my early days at HP, early 80's working on HP-7910 Winchester Drives
Made me think of Jacob Geller's "how can we bear to throw anything away"
I love it when a plan comes together. :-) Congratulations, bro. Feels great when it all works out, especially when you had to go to such extreme lengths as ripping apart 3 different drives to get the sketchy Finch going. I don't blame you for geeking out so hard at the end at ALL. Great stuff.
Thank you so much man!
It was a wild ride, but all in, I started working/filming on Sunday, and on Thursday morning we had the data successfully backed up. I never would have guessed we could have gotten it going in less than a week!
@exidy-yt "I love it when a plan comes together. :-)"
Yes, Mr. Smith. 😎
@@pianoman4Jesus If there was a cigar smoking emoji I would have used it. 😁
@@exidy-yt For sure! 🤣
Cannot believe I used to do all of this 35+ years ago and I was swapping components as an engineer to make it all work for customers. It was so natural and normal for me as my day job. Well actually i used to do this night and day for customers that paid for 24hr cover.
I give you a lot of credit. That was challenging. It sure is a great feeling when you finally succeed. Amazing to think this vintage hardware still works. It takes us back in time to when this was just modern hardware they were working with daily.
REN is like the Centurion whisperer.
It's wren I believe edit: apparently I'm mistaken, never mind. Must be confused.
Ren is an absolute legend and has saved my butt more than I can count!
You did sooooo well!... I've had my own drives from the 1980's, stuff I'd long forgotten about, not these big drives, just drives from old IBM286's and Acorn BBC's and such, and honestly, to see some of it come up fully intact, programs I'd written was just amazing!... I can honestly relate to how you felt making this video...
I can barely follow what's going on with most of your videos, but your enthusiasm and excitement come through perfectly, and that's what keeps me coming back. Another great video.
Awesome! My heart was in my mouth several times during that. Sometimes when people ask me ‘why do you do this’ I have difficulty explaining that feeling you get when things go exactly right and you get a beep or a READY or something like that. Now you’ve encapsulated that feeling in a video I can use this to show them why I do what I do :D
Just found this channel and your collection reminded me of a '90s Eastern European comedy show ("Imanta-Babīte nepietur"). In one sketch, a GP recieves a patient and asks if he made the urine sample for testing. Patient confirms and takes out of his bag two full 3l jars full of pee. Doctor, eyes like dinner plates, asks if he brought some suitcase of fecal samples too, which patient confirms and slaps a big old suitcase on the table.
So this '80s tech have similar vibes today for some..))
However, without experiencing much of it myself, i too have respect and some phantom nostalgia for it. As a normal guy born in USSR, i still prefer the miniturization, as long as it's sturdy design. We didn't have commercial consumer PCs back in the '80s (unless you risked paying 10× the price to some smugglers, same in beginning of the '90s). My dad did get the B&W soviet IBM equivalent in the first half of '90s, with 8in(?) floppy and i think 84K of RAM from the national TV he still works at. A bit bulky and heavy as hell. If i had it still, i'd send it to this guy. Unfortunately it got lost after someone cleaned out the parishe's storage space where it was sitting after apparent catastrophic failure.
["ПКП" on the screen - not in any manuals, without any other signs of life or bios, though no visible damage inside. No specialist could tell what's up at the time, i'd be surprised if anyone knew here].
I especially like your reaction when things work. I can feel it for you. Great job.
Amazing. I've never heard of any of this equipment, but it sure looks well built and will probably last another several decades! Good for people like you who can figure these systems out and show people like me how they worked, what they were for, and why they matter!
I manage storage systems with hundreds of SSDs (and/or HDDs, depending on the system) operating in synchronicity. The mind boggles that any of it works. It's basically magic. Congrats on getting the data!
The modern hard drives these days are incredibly complex machines. The hard drives from the early 80's were straight forward and basically much simpler.
I don't think my anxiety has ever gotten so high by just watching a clip
As a trucking computer tech with 2 bachelor's degrees that just loves old tech this made me a kid in a candy store but make me feel like a dad watching his kids play in the middle of the highway at the same time.
Thanks for the peefect video.
If Hollywood videos had as much excitement and suspense as this clip does, it would be sold out for weeks.
Putting the drive and PSUs on the wood brought back memories. When I was a kid without money, I would often get PC parts, but wouldn't get a case and, since I couldn't afford one, I'd end up building PCs with everything attached to a piece of plywood. A lot of times, I didn't have switches either, so I'd turn them on with a screwdriver. My friend used one as a dialup proxy machine for 5 years before it needing a reboot. Although I've worked on much bigger stuff since then, I always consider those machines my greatest builds because I really did scrap them together from nothing and they lasted quite a while. Really love that your setup got that data off the drive. Cool stuff!
I missed this video a few months back. The work you did is absolutely amazing. Taking the time to clear it and copy a finch is so cool. You rock.
.
15:57 "it suddenly got quiet": same thing happened once with my PC fan. It was making horrible noises and then noises suddenly stopped. I thought it is dead. But nope, it continued spinning, but quietly.
New to the channel... this was great, thanks. I shared your sentiments at the end throughout the whole process. It is impressive just how robust this old tech really is for the most part.
It's really cool that the Finch drive has a clear cover. Glad you got things to work in the end.
Insanely long? lol I watch my man, Adrian Black, so this was nothing! I love his long vids and loved this one as well. His disbelief and joy of bringing old machines to life again like you did is my favorite part. Y’all are such treasures and I know the original engineers of these processing machines would cry tears of joy if they saw y’all’s work. Also, love from another Texan. I’m in between Austin and San Antonio so maybe somewhere somehow I’ll meet you.
I think this is my favorite series across all of TH-cam! So much vintage computing drama over the months.
Thank you so much!
24 minutes in and the excitement is breaking through my screen.... Squeals of utter joy!!!! So awesome!!!!! 🎉🎉
This randomly came up on my suggestions and is completely different from anything else I watch on TH-cam. However, I gave it a try and I'm absolutely hooked. I've even subscribed because this brings back distant memories of a misspent childhood in front of old computers (I'm not quite old enough to have used these systems though!). Excellent stuff!
Mate I wouldn't have wiped that original drive but sadly it's done now. Your dealing with Electronic Archaeology here & the more you can save the better no matter what it is including data from any machine. I on the other hand have gone back in time quite a bit further to the Valve era looking at doing repairs to some ancient valve gear going again which is usually dried out caps & sometimes a resistor or 2. The valves are usually live by the amount of hours it's been run. It's lucky you have the skill to save these antiques. Keep up the good work
Good point. History is history. You never know what secrets could of been in there for future archeology research.
Haha Michigan here..
"Its chilly here.. its 60 degrees" 😂😂😂
Meanwhile up here where the windchill has been like -30°F 😅😅
It'll be interesting to see you guys come up with another drive type for these old machines, the old drives are on borrowed time IMHO.
Yeah, an adapter to either SD card or CompactFlash really is the best bet for long-term preservation.
This brings back some memories of recovering data from hard drives. I could even spin up frozen 5 1/4" disks by twisting them during power on to release the stick slip on the spindle. It was quite common for disks that had been used in servers (running 24/7) to just freeze up as soon as they were stopped for a few days.
Congratulations, some awesome work here, very glad to see more and more of the legacy of this small branch of computing being saved! Your retrospective at the end puts into perspective what we sometimes miss, just how much has to work to make all this stuff (that no matter how well made is well past its design lifecycle), do its thing again.
glad to have you as part of my sunday morning routine, your channel rocks
Thank you so much!
Amazing video, the fact that data was able to be copied and there was cool programs is important for the archival of the machine itself.
But I also thing that user data is also interesting.
I still think old data should be saved and left in some corner on some disk, no need to put it on the internet for all to see yet. Maybe in 50y the PII won't be a problem, everyone in the list would be dead. I was seeing dates like 1995, people it was referring to might still be alive.
For archeology reasons in 500y if it survives, it would be funny to see the logs of a company.
It's sad private data about companies don't get left anymore , there won't be anything for archeology to see about how companies in 2020 actually ran.
Just like clay tablets said a lot about the first corporations in the old civilizations, but once they where, with old HDDs we have a glimpse of 1995, but after that it's going dark.
Imagine in 2100 wanting to do archeology on this old defunct company called Google.
I for sure am donating every data I collected when I was alive if I die, I don't even care about the origins and that it was private data of mine or others. You can have it as long as you promise to keep it for archeological reasons.
And I have a lot, like a lot of copyright material that's probably not even in the internet anymore. Like at least 20TB of "proprietary" things. It's going to be all public when I die. I can't be sued if I'm dead.
Why would I care about my PII data if I'm dead.
I think everyone's data should become archived when they die and then become public a couple of decades after they died.
But I do understand the liability of holding PII data nowadays.
Not saying you shouldn't have deleted it, it's not very useful for content on this internet television website.
I'm just saying what I usually do when I find fun things on disks.
I agree about saving the data anyway. Just encrypt it if you're concerned about accidental leaks. I also agree that some people might still be alive considering I was born in '95 and I'm only 28 lol
@@clonkex I said it for fun, I'm just 35 .
Great work! I felt your excitement through the screen. Can’t believe everything actually fell into place the way it did.
Video's like these make me appreciate modern computers in a way: Not having to explicitly know what disk to target or what partition and type out commands is a whole lot less cumbersome, but compared to my first computer (A 486, not sure what kind, it was a business-line HP Vectra from my dad's work that passed it's financial lifespan) I weirdly enough still miss the days of being on an offline computer and having to make sure your IRQ's were set properly for something as seemingly simple as a sound card.
The idea to now just have to plug in a disk that is having problems and clicking a recovery program is wild: the amount of technological progress we made with things like this is insane when you think about it.
Ikr. It's so crazy to me as somebody who grew up with XP and 7. By then PCs were pretty much just plug and play. That's probably a reason why many people buy microcontrollers and SBCs, to get that feeling of being in direct control of the machine. It's the equivalent of driving a manual transmission.
Your enthusiasm always makes my day better
Good job on getting the data off the Finch drive. The next job will be attempting to repair the original PCBs from the drive.
Thank you!
Now that we have the data backed up, the big question is where do we go with the drive from here. To do the level of repair required for the bearings is really risky as we can potentially lose alignment with the soft-sectoring servo track, and if we lose that, there is no recovering. So, I need to talk with Aaron a bit and see where he wants to head next with it!
@@UsagiElectricI wasn't suggesting attempting any mechanical related repairs. The drive to be rescued had at least one fault preventing it from working. Perhaps you already fixed it with the incremental head seek test. You could replace the PCBs with the original ones and see if the drive works fully. If not, you can compare signals between those on the PCBs of the drive that needed backing up with those of the signals of the donor drive PCBs as they seem to be similar sets of boards.
@@UsagiElectric Compare the windowed PROMS at 5:32 vs 20:16 , I'd guess the first PCB's firmware "state machine" is all 1s at this point.
While I don't understand everything said and done in this video, I am glad my start in computing, in 1988, began with finding a way to make a 286 with 30 meg hard drive running DOS 3.1 useful in a video production office. Many books were bought and read along the way, and enough of that now useless information gave me enough insight to follow along with what you did in this video.
Best sunday start is with a new Usagi video 😊
Good stuff! Glad you got it backed up and also are going to copy to a new PC. Well done Usagi and REN
I remember that mainframes and desktops split due to physically being too big.
I see now that desktops need to do a split.
We can do cell phones, laptops, and tablets; but the workstation desktops are facing a forced breakup right now.
In large enterprises there is a shift back towards "terminal" like computers by using virtual desktop environments. A low powered computer is used to log into a virtual desktop server and all the processing and storage is done remotely while the screen information is sent back.
@@volvo09 Funny how it goes full circle.
Great video , i,m only 27 years old but i love the older elektronics/computers i still cant belive you got that old drive to work !. i hope one day i have the space and time to take on similar projects, i know the feeling of getting something to work thats basically dead and its the best feeling ever
Time, patience, and perseverance, and a methodical approach mark the solution to another problem for a Centurion. So very well done.
Nail biting stuff 😳 But WOW what a treat to join along… The screeching of the drive made this big time suspense 😀
Amazing as usual! I just had to go through some headaches with a dying drive in my personal PC, fun to see the parallels between what I went through in 2023 vs what you would've had to go through 40 years ago, the more things change the more they stay the same
I love how invested you are and how animated you get when things get squirly lol.
seeing you begin so happy after successfully backed up the drive just made me so happy as well! you made my day!
I love how hype you are about computers. So wholesome.
I love the age we live in today. Look at the size of that beast and 32Mb. What a small, small capacity for such a big drive. It's just wonderful how we have teeny weeny NVME cards with 4TB and ssd drives with 8Tb and even thumb drives with 256gigs of space. Love the video, thank you for letting us see how far we have come since 1980. WOW 1 hour to format a 32Mb drive boy we really have come along way since 1980. One last comment I'm glad someone else needs to check weather or not they are coping the right disk from the right source I'm always double checking when I do it too.
A surprising result, especially considering how much of the process can be summarized as "I know it's erroring, but we're gonna try to read it anyway." Well done!
This was so much fun to watch. I was a programmer in the 80s. Your excitement was contagious. 😅
I do not remember any video that I commemorate so much like this one. Congrats on the success!!
What a win. Thanks for the supper video. Away great to see the recovery of history, and thanks again for the video.
Congratulations for this achievement well beyond your own expectations and a big thank you for keeping me so well entertained watching you fighting your way through this. I love old computers and weird operating systems and well, this one gets a high grade in the weirdness class!
My favourite retro computing TH-cam channel for sure.
Your energy and exuberance is so infectious. I even raise up my arms and let out a loud YESSSS when you win such a battle 😀
I have a lot of patience, but I would gave up after the read head jumped around, Good job! Glad you stuck with it till the end.
Great job! It's a good day when all the hard work pays off.
Although errors on the "good" 32MB finch still trouble me.
We really. really need some kind of SD-card drive emulator or similar for Centurion.
I think this should be doable. If only I had enough spare time to help...
Thank you!
Whenever S.NEWDSK runs, it makes a bad sector table that lives on the drive, so all the bad sectors are known by the OS and it can write around them.
We're working on emulation options for the Centurion. It took about a month of work, but we did manage to get a floppy emulator working (the Centurion uses very custom encoding and timing that did not play well). We can now use that as a springboard for Finch emulation, as it's mostly like a Floppy with much faster data rates.
@@UsagiElectric Thanks a lot for replying.
it SHOULD mark the bad blocks, but if there are many in same spot the whole area is likely marginal and likely quite unreliable - at least this is how DOS era hard-drives are.
Floppy emulation is excellent news! I applaud you and everyone who made that possible.
I think all of your viewers would really appreciate if you could give us occasional updates on community development progress - emulation and all.
Thank you and stay awesome!
9:07 Still being used in 1996. I'm impressed.
That is 10 years after Centurion (ZTRON) went bankrupt in NOV-DEC 1985. The assets being sold to ex-Centurion dealer in the spring of 1986.
I love your determination and I can't believe any of this works, yet it all works! Congrats you got my thumbs up, cant wait for more.
Bravo! So happy you managed to pull her off; These old gurls were in a class of their own!
> (Maybe today will be an easy day!)
Shout out to Editor David. 😆
This was epic. Well done man
Haha, editing David was having a bit of a laugh!
Probably not useful for this occasion... but if you ever come across some damaged media like an HDD and can connect it to a machine which can boot Linux, do this.
There's a tool called dd_rescue which is by far the best option to get the most data from a damaged device as gentle as possible.
It basically copies data with large requests, until it encounters an issue. Then it just jumps in the center of the unread data and continues there, avoiding stressful rereads of the damaged data over and over again.
This will continue until it's on the end of each unread block, then it jumps in the center of the largest remaining unread block. And repeat and repeat and repeat.
This way it avoids the damaged areas as long as possible.
After which it reduces the request size down to just a single sector and repeats the reading of the previously large areas with read issues, but only tries each sector once.
After this pass most readable data will be read. But you can let it run continuesly. It will repeat the reads one damaged sector at a time, and I found it's often lucky and will get most of the damaged data anyway from the disk. But those operations tend to get stressful for the drive, as it needs to seek a lot. So it may die in the process.
To give a perspective: I tried to copy a hard drive which had some damaged areas with a windows tool, and it got to the first damaged area and got stuck there for two days. Then I terminated it.
dd_rescue got more than 99% of the data from the harddrive within around 2 hours and then started the first retry pass.
In just a couple of hours it got everything out of the hard drive, except ~10 sectors.
After this experience I've never looked back and saved a couple of dozens or so broken harddrive data with it. Worked great every time.
I don't know how I stumbled on this video but it's certainly top-notch nerd pr0n. There's nothing like spending days or even weeks cursing every minute of the mountain you are climbing, and then finally through unbreakable persistence, you get the win. As I get older and earn more and more patience, the wins feel even better when they come along. Thanks for sharing this one... some of us know how long this 37 minute video REALLY took. hehe
Great job on getting the data off the drive! I like your enthusiasm in working with this old computer gear!
I love data archival, and I hope you can get that error fixed on that source drive now that the data has been mostly preserved and you are free to rewrite it.
I think just running S.NEWDSK on the Source disk would totally solve the error! Most likely, it's developed a bad sector or track somewhere, and S.NEWDSK detects those and creates a bad sector table that lives on the drive, so the OpSys can write around them. But, the bigger issue is the bearings. I'm not sure how we can get the bearings serviced without losing alignment with the soft-sectoring servo track on the disk. If we can get the bearings repaired, then I think we could totally bring the drive back up to reliably working!
@@UsagiElectric As weird as it is, see if you can find a Data Recovery Center or HDD repair place that has a clean room; maybe they can totally pull the drive apart, clean and service it, clean and relubricate the bearings?
Amazing stuff, I really enjoy this channel, never heard of centurion or finch drives before. It's nice to learn about this stuff and how it worked before modern era of computing.
Your videos are the absolute highlight of my day! We exhibited across from each other at VCF east. I had the silly mac classroom setup. This journey has been just so much fun!
Absolutely love archival recovery projects. I remember many moons ago working on a project to recover data from a machine I can only remember being called an "Ego". Was a sort of single box minicomputer thing with a couple terminals, booted from ROM and had a removable 5Mb (I think) disk. Went through a process of dumping data out the centronics parallel port, via a parallel to serial converter to an RS232 port on a PC hahaha.
i don't know what bearings are in those drives, but if they are sintered bronze bushings, then it sounds like they are dry or worn
If the bronze is worn, the shaft rattles around, if it's not too worn, a little oil will work (at leats for a while)
love how upbeat you are about everything, Dave!
i do enjoy the videos on this channel, despite 40 years of computer use, starting with spectrum in 1982, for my 8th birthday, i only started to learn electronic stuff earlier this year, with a box full of components and a soldering iron i have had a wonderful time, ive made a driver to power laser diodes, and set up an optical bench, ive made a zvs circuit to drive an old LOPT, ive chopped the deflection coils on an old CRT b and w tv to make a vector-scope. despite not having a treasure trove of ancient computers per se, i still get so many general tips from watching this bloke. thanks. and btw way is it called USAGI?
You can't script this level of drama and pacing. I love these hard drive recovery/restoration episodes.
Beyond interesting, I can just watch for your excitement alone. Nice work!
I once had to copy an 8-inch floppy with good data onto an UNFORMATTED 8-inch floppy to back it up. This was on a CPM machine. I wasn't paying attention and put my floppies in backwards in the drives. Since I was doing a sector-by-sector copy, the system dutifully copied the unformatted floppy over top of my good floppy. Now I am crazy paranoid about deleting data of any kind 🙂
I think the type of harddisks my dad used to work with might be in this video, he once deleted the whole company's several years worth of archiving of theater documentation (song-texts, information about where to find decor pieces or clothing items, he worked at a conservation company) due to a faulty bulb he tried troubleshooting by exchanging bulbs.
In the meantime, he also tried something else, and mistook the good disk for the wrong one, and destroyed the backup, with the original work-disk being non-functional at the time.
When my dad was with the company for 25 years and it was celebrated, the then-head of the company had a little speech about how he was not easy to be moved about big changes in the company, but "I was told that one time he made a big mistake and destroyed years of work due to a faulty bulb, and it was the first time that he decided to take a walk, looking slightly more pale than he usually was..."
We all do this exactly once. It may not be in the same physical format, but there's always exactly one clone job every PC tech that gives them this lesson.
I knew someone who did the same with one of the CDC hawk like 500mb drives. He was supposed to do an overnight backup, but backed the empty or"scratch" drive onto the full working drive, wiping all of the data. He was called scratch from then on😊
Fantastic video, and you really took us along for the emotional ride. Well done on recovering that data. I do a lot of searching for short circuits as part of my job, and here are my tips, in case they're useful. You can use a Toneohm, a sensitive low-ohmmeter with an audible output so you don't need to look at the screen, to measure the resistance at various places and you very quickly get an idea of where the fault is. Or if it's a power rail, apply power from a current-limited power supply (current depends on the circuit, but somewhere between 0.1 and 10 amps may be appropriate) and then measure the voltage across the power rails in various places with a high-resolution (5 or 6 digits) voltmeter. You can also measure the voltage drop from supply to various points along each supply rail. That will also quickly give you an idea of where the shorted part is.
21:49 Argh! It sounds like evil! I cringed my way through this video whilst hugging my self in hope. Well done for thrashing this through to a great result!
Absolutely enjoyed watching you nerd out on this..😊😊😊
Well, it was the transistor that killed vacuum tube computers. Vacuum tubes just didn't last long enough, and when you have hundreds to thousands of tubes, and a single tube failing can kill the machine, well there you go. Transistors just last longer (when they got perfected, early transistors were a nightmare). Especially in low density IC's. Those 74 series were workhorses, that's for sure.
The longevity of the vacuum tubes was a "solved" problem. It was their physical size and power consumption that killed their usage in computers.