Reviving an 8-inch Hard Drive from the 1980’s!

[Music] thank you hello and welcome back and today we're back to working on the Centurion most notably we're working on this little drive right here this is a CDC Finch Drive and it's not the first time that we've tried to get this drive going on this channel uh the previous escapades with it didn't go well but before we get into the troubles that we're having with this drive and what we're going to do with it let's rewind a little bit because it's been a hot minute and recap what centurion's history with CDC and their hard drives has been like and the very first drive that Centurion used in their earliest systems was the Hawk Drive and that's this drive right here and I have another one just behind the camera over there in the mini Centurion both of these actually work which is kind of crazy these were five megabyte fixed and five megabyte removable so 10 megabyte drives in total they weighed about 135 pounds and they're pretty epic to watch spin up a little later Centurion upgraded to the Phoenix which is this drive right over here this one came in a couple of different sizes although this particular one is a 96 megabyte option that has 16 megabyte removable platters which we have stacked up over there now both of these drives used 14 inch platters looked just like this one right here now if you're worried about me putting my fingerprints on this platter don't worry it's it's already been crashed you never want to try and fly a crashed platter because it'll just exacerbate the problem and crash the heads even further potentially destroying the heads in the process but they used these absolutely huge platters and of course the trend in the industry was to get smaller and lighter so the next logical step was to go from 14 inch platters down to eight inch platters and CDC or Control Data Corporation did that with the Finch Drive there's a little bit of a theme going on here with bird names we've got Hawk Phoenix and Finch the Finch Drive was quite a bit smaller quite a bit lighter this only weighs maybe 20 or 30 pounds as opposed to 130 pounds also these came in a couple of different size options as well there was I believe 20 24 and 32 megabyte options uh might be a little off on those numbers but I believe this one is a 32 megabyte option because it has three platters inside now there is a second Finch over here that only has two platters inside so I think this is a 24 megabyte option but these were still quite big and Centurion used these particularly in the micro plus systems but also use them in the big full-size cabinet systems we got this one with the counterfeit system it was in a separate box that looked pretty neat it was a nice high storage completely sealed option now of course CDC wanted to get even smaller and the next logical step after eight inches was five and a quarter inches which is oh what this guy is right here this is a CDC or uh magnetic peripherals Rin drive that's w-r-e-n Rin again a bird name but this used five and a quarter inch platters and I don't actually know how big this one is I have no idea how much storage is on these uh now the Rin had a couple of different protocols that it used the very early first generation Rin drives used the exact same pin out and protocol as the finch drive so they were actually interchangeable so some models of the Centurion micro plus actually came with Rim drives however the first generation Ren was only Built for a very short amount of time and they very quickly changed over to the second generation Rim Drive which is what this one is and it uses the same format as the Seagate format the st207 or whatever that number is so unfortunately I can't use this drive with our French floppy controller that's in our big Centurion it would be awfully nice to have a somewhat modern five and a quarter in there but well I'm still hunting for the proper first generation Wren so hopefully someday we'll find one this one on the other hand is actually slated for use with the pdp-11 potentially because it does share a format with what that thing can support so those are the types of different drives that Centurion use different hard drives that is centrion went bankrupt and bust long before three and a half inch hard drives were commonplace so it was generally 14 inch eight inch or five and a quarter inch and the five and a quarter inches were really late and they only used them for a very short amount of time but we've got this Finch Drive here and I really want to get it going so where are we on it uh it's broken as a matter of fact this Finch Drive is also broken and where did both of these come from well this one as I said came from the counterfeit system that we got it was in a separate box this one came from a micro plus that is owned by vintage geek over in Tennessee if you want to know more about this drive and the system that it came out of I did actually a full video on my trip to Vintage geek and how I came back with this thing there is a link in the description below but the goal is to get both of these finches going however they're both broken and they're broken in different ways which is good it means we can actually cross-reference across each other because aside from the number of platters they're almost identical this one the original one that we have the 32 megabyte model is mostly working it spins up it does its self-seek test and then it sits over track zero waiting for commands from the controller card and waiting to send data back but the controller card doesn't actually see the drive as existing so what that means is that something is not making it out now this one over here the one that came from Tennessee it actually just has a dead short on the 5 volt rail on it so it never actually Powers up at all so we've got work to do on both of these and we're going to start with this one and I think what is wrong with this one is that we have a bad line driver in the previous episode I took a bunch of scope shots but I didn't get to the point of showing them in the episode so I'm showing them to you now these are the scope shots that I took we have a clean bite clock coming into our line driver which is an sn75110 this line driver is supposed to receive this bite clock and receive enable signals which we can see that it is getting it does have the correct enable signals coming in and it's supposed to supply a differential signal out now if we look at the differential signal here on the bottom you'll notice that it's not differential for some reason the outputs that are supposed to be opposite of each other are actually in Phase so the actual signal that the finch floppy controller receives on the Centurion is bad it thinks that the drive doesn't exist so our first step I think is to replace this SN 75110 and then give it another test so I'm going to pull out the desoldering iron we're going to pull that old chip out put a socket in put a new chip in and spin it up one more time for the first time in a very long time to get to the PCB there are just four screws holding this backing plate in place and with all four of those removed it just pulls right off and that gives us pretty good access to the back side of the PCB but we need to get the PCB fully out and in order to do that there's one more screw right in the center then we can go through and unplug each of the cables that connects the PCB to the motor and heads one by one and this big ribbon Cable in the back is always a pain I'm just terrified I'm gonna rip it in two I hate this design of ribbon cable but it came off and here is the PCB all free from the confines of the finch and if we look a little closer we can see the sn75110ic that wasn't giving us proper differential outputs last time we had the scope on this so in order to remove this IC I start by adding a little solder onto each pin then I use my desoldering iron to suck the solder off of each connection and I love this little desoldering iron it works amazingly well and with all of the pins desoldered the IC needs just a little bit of persuasion and it comes right out then we'll clean up the PCB and solder in a new socket making sure that we've got it nice and flat and finally we just need to pop the new IC in place okay new IC is in and we're going to test it using the real Centurion here I've got the diag card in with the auxiliary test menus up on this test menu I've always been able to pass the seek test but never been able to pass the read test so that's what we're going to do we're going to run the seek test followed by the read test so I've got the lock off of ship and on to operate I'm going to flip the power switch it should spin up do a self-test seek and then once it's done with that it's ready to run the test so fingers crossed here goes nothing [Music] it's spinning up doesn't sound bad that's good sounds like the bearings are nice and smooth [Music] [Music] the heads should have done a self-seek and they didn't that's not good [Music] all right I think I may have narrowed in on another problem this little regulator I see right here is a 78 18 it takes a 24 volt input and gives a regulated 18 volts out and even though it has this heat sink here within seconds it becomes so hot that the heatsink itself almost burns your hand it's a little ridiculous we can check if it's actually regulating though the leftmost pin is the input that should be 24 volts the middle most should be ground and then the rightmost pin the one that's closest to me should be the output at 18 volts this little IC up here is also a 78 18 so we can kind of use it as a reference so I've got my multimeter probes here will pop the power on and if we check this one up here we can see it's regulating at 18 volts perfectly if we check this one here the output is 300 millivolts uh yeah that's not correct and whoo boy that's already getting hot so I think we'll pull that one out and replace it and maybe that'll bring this thing back to life a little bit the first step is removing the nut that holds the regulator IC down to the heat sink then again just like before I'll pull out my desoldering iron and suck the solder up on the three pins but the middle pin was being a little stubborn so I just heated it up with a regular soldering iron and everything pulled right out with ease then finally we'll just solder in the new 7818 regulator IC all right new regulator ICN uh we're gonna spin it up one more time what we want to see is we want to see those heads do a self seek test so here goes nothing yes they did a self-seek test we're ready to plug it back into the machine and see what happens so we'll spin it down that's awesome we'll plug it into the machine and go okay let's try this again we're gonna spin it up and try to run the seek test make sure that still works and then we're gonna try and run the Reed test and fingers crossed everything works although I'm not expecting it because nothing's ever easy with CDC drives but well here goes nothing the drive is spinning up we should see it do its uh self-seek test on here [Music] it did it self seek test it sounds good it's sitting over track zero we'll do a seek test that's going to be uh zero f here [Music] yeah I can see the heads seeking smoothly in and out that works so we're gonna hit Ctrl C it's going to come all the way back to track zero and then the test should finish up there pass that's good we're back up to where we were before all right this is the big one the retest one zero enter oh busy did not clear that's that might be better news okay it's a little noisy in here because I've got both the hawk and the finch spun up at the same time this is so I could get into the question mark dot question mark Finch tests that are stored on the hawk drive itself these are a little more involved and give us a little more control and we've discovered a lot of really interesting things first of all contrary to like pretty much every drive out there track zero is the innermost track on the finch uh whenever you do an rtz it always goes to the innermost whenever you tell it to seek specifically to track zero it also goes to the innermost so that's a little weird but not I guess it doesn't particularly matter all that much all that matters is that we've got to track zero uh however we're not getting any data off of that track zero which is why it always seems to fail now interestingly when when it does fail when it gives us the busy did not clear everything stays asserted so the drive is still acting as if though it's selected and trying to give out data so we can check a couple of signals in this state the first is we can check an index signal that's one of the most important signals that can come off of the drive so that is going to be on pin 20 here so if I count down right there we have our index pulse right here seems to be running at 60 hertz that's I guess I don't know that's that's what the index pulse is at but we have a good index pulse coming in so that's good news next we want to make sure that drive ready is low so that's going to be pin 22 and drive ready is indeed low so this is an active low signal so if it were high that would tell us that the drive was not ready but the finch is telling the FFC controller card that hey I'm ready for command signals now on the second ribbon cable down here we can check a couple of things the first is Servo now the servo clock is going to be a differential signal so it's a little difficult to check with the scope probe here but that's on pin 15 and pin 16. so we should see a little something up here coming off of those pins and it's small it's about one volt Peak to Peak but pin 15 and pin 16 do have something showing up on them and one volt Peak to Peak even though its differential means that it now becomes two volt Peak to Peak and that sounds about right so we've got a drive ready the drive is absolutely ready we've got a good Servo signal going in there we just we don't have any data and we can actually check that on pens 18 and 19. again differential signal but if we check it um so totally dead we're not getting any data at all out of our data differential output now at first I was thinking that this was a fault of the finch but uh talking to some people on the Discord there was one more signal I needed to check and that is read enable the FFC is going to see the index going to see the servo clock and it's going to tell the finch hey turn your read enable on so I can start receiving data if that read enable never goes on the finch will never output data it's dating that data on the way out so read enable is pin 4 on the command cable up here and if we look at that pin 4 is high all the time beat enable never gets asserted read enable should be low when read enable goes low that's going to allow our data to start coming out so we've hit an interesting spot the finch itself might be totally fine and we might have a failure on the FFC so the next step is to spin all of this junk down so I can pull the FFC out put it on the extender card prop it up some way and then start probing around on it all right it's getting a little late in the day and I was kind of procrastinating on pulling the FFC out and starting to probe around on it with the scope but masaka over on the Discord had a brilliant idea she said why not just use one of the other Finch ports the the FFC card has four ports on it you can support one floppy and up to three Finch drives if the finch is actually totally fine but we have a bad receiver chip on the FFC well one of the other ports might actually be totally fine so right now I'm over on Port three I've got both the drive spun up I got into the question mark Finch diagnostic tests I already ran the seek test so I know that I've got everything hooked up correctly and we're right back up to where we were so let's run the read test and see what happens so we'll just hit R here for read test we'll do track uh the drive is actually two uh we'll do disc number zero the seek pattern we'll just do is a our low track and hex will be zero zero zero uh and then our high track will be two hundred we'll hit enter the printer will hit enter on this I can hear it seeking that's all we're passing a seek test oh man we'll see if I can get the camera to focus on this [Music] so there we go we we passed a read test for the first time ever there's only one thing left to do now and that's to get in the OS and see what the OS thinks of the drive so I'm going to bring the camera in a little closer get the nice camera on the screen I'm gonna kick the lights off so you guys can actually see the screen really easy and let's get back into the OS and see what happens okay let's go ahead and uh hit the reset button here we'll type H1 at the D equals prompt here that's going to try and boot us into the operating system we should see a Max disk equals some number other than one here nope all right I've been thinking about this I think what we got to do is we gotta actually register two more Finch drives so let's hop into the sysgen utility here [Music] uh our configuration name is at OSN uh the disk number is one this is our system Configuration utility so we'll do number four to go to disk drives here uh and you can see I've got uh the floppy set up and one Finch but we'll go ahead and add in two more finches so now we have disk zero one two three four and five that should be enough inches for the operating system to see it so we'll get out of this by doing control B for enter uh then end program no okay now we gotta reset one more time I'll hit the load opposite button we'll type H1 uh again now we should see uh something other than uh Max disk equal one Max disk equals five that's good news we're making progress okay uh oh eight twenty three eighty four okay we're in we're into the operating system CRT 0 Let's do a DOT sta foreign oh my God it works like a year I've been thinking that this thing's broken turns out we've got a fault on the f I can't form coherent thoughts oh my God okay there's so much stuff on here there's so much stuff on here oh we stopped oh it took it a second okay oh look at the size use 4156 avail available they stack this thing full of data we gotta figure out a way to dump all this I have no idea how but it doesn't matter because it freaking works yes [Music] so what did we find on the drive well I have a full directory listing here printed out and uh it's really mostly boring files there's there's nothing really exciting on there there are a ton of files in here that start with the letter O and I think that corresponds to it being part of the oil and gas package and we actually can kind of confirm that there's one file on here called OBM one and this is really fascinating this is a purely text based file but I think it's just actually a backup output by the oil and gas package and the reason I'm pretty confident it's oil and gas is because if we look at the actual items over here on the right some of them are really interesting replaced two and a half inch flow line repair leak put belts on unit repair engine number one a haul off oily dirt and replace with fresh dirt that's a pretty interesting in one uh looking at all this stuff in here it all absolutely points to the oil and gas industry and we actually found that in one more file as well the Centurion creates these temporary files whenever you're doing something usually uh and a lot of times those two very files have a name like at scr0 and this is at scr 0 there was actually something still remaining in this temporary file and there is a name it's Commons oil company located here in Dallas and if we do a little bit of Googling we can actually find some information about mccammon's oil company so that's really cool we now know who used this machine so that's really fascinating we figured out a ton about the history of the machine from just a few files but well if it's all just oil and gas stuff and I'm not an oil and gas company that's all kind of boring stuff so is there really anything super exciting on there and there is there was actually two folders on there one of them was called year and super boring don't care about their year-end stuff but there was another one called z-o-l uh and I have the uh directory listing of it right here and it's just pure ASCII files that all start with the letter Z and talking with Ken that is an indication that these are potentially source code files for something now what are they source code files for well that's what I was trying to figure out so I opened up zoa here and started dumping the contents starting to back up these files and uh this is zoa and we were looking through it it was awesome and then all of a sudden on page three it was in the middle of dumping and it aired out we got uh in our space disc five and it just continues to air completely like that so is the file bad well no this is the answer to the big burning question which is why am i showing you all of this information on pieces of paper that I've printed out instead of on the actual machine with the actual Drive hooked up and that's because the drive died while we were backing up files it had a total uptime of about 30 minutes before it kicked the bucket now I don't think that the heads have crashed I am like 95 certain that the platters are still good the heads are still good and we still have good data on there because if I get this scope out I still get a clean index signal and a clean bite clock coming out of the data cable here on the bottom however that index signal never makes it up to here to the command cable which is where the FFC is going to be reading it from and that's because the drive ready signal never goes low so for some reason some I see on this uh Finch PCB lunched itself and the drive never comes up to being Drive ready which is what in our disc 5 stands for means not ready disc five so our Finch Drive is broken yet again all of that celebration maybe I celebrate it a little too hard but uh we know that there's good data on there we know there's some source code that we definitely want to back up the rest of it we don't really care too much about but it would be nice to have a backup of it anyways so we've got to get this drive up and going one more time and that's what we're going to do in the next episode of this series for now I'm going to edit this video back up so you guys can watch it because it's already getting quite long and then I'm going to slowly poke away at this and try to reverse engineer as much as I can to figure out why our drive is never becoming ready and then once we got it up and working again we'll start ripping the data off of it and then we'll use this one as a baseline for the 24 megabyte fence from Tennessee and get it up and going as well I have no idea how far we'll get in that next episode but we're going to keep working on this whenever we come back around to it so I want to thank you guys so much for watching and I hope to see you next time

Channel: Usagi Electric

Views: 63,134

Keywords:

Id: hbc6n3Ln3bM

Length: 28min 21sec (1701 seconds)

Published: Sun Aug 06 2023