Reviving a 1970’s Hard Drive for the Mini Centurion!
ฝัง
- เผยแพร่เมื่อ 2 มิ.ย. 2024
- VCF Southwest is barreling down on us real fast, and none of my junk works, haha. Time to hunker down and start getting stuff over the finish line. The Mini-Centurion is specifically meant for shows and right now, it doesn’t have a hard drive. So, in this episode, we dive in deep to try to bring the old 14” Hawk drive that’s bolted to it back into the land of the living.
If you want to know more about the Centurion, the wiki is full of just about everything we know:
github.com/Nakazoto/Centurion...
Come hang out with us at VCF Southwest:
www.vcfsw.org/
Check out John’s shop, Kad Industries, here:
kadindustries.com/
Check out Butler Tech here:
www.butlertech.org/
If you want to support the channel please hop over to Patreon: / usagielectric
Also, we now have some epic shirts for sale!
my-store-11554688.creator-spr...
Come join us on Discord!
Discord: / discord
Intro Music adapted from: Artist:
The Runaway Five Title:
The Shinra Shuffle ocremix.org/remix/OCR01847
Thanks for watching!
Chapters
0:00 Carry on my wayward son…
5:56 Power supply shenanigans
12:30 Head cleaning and disk inspection
15:08 Loading the heads and failing to boot
18:18 What went wrong?
20:37 Repairing the drive and aligning the heads
26:14 Test drive
28:08 The good, the bad and the ugly
31:20 ニャラリー - วิทยาศาสตร์และเทคโนโลยี
I found a drive similar to this one *STILL RUNNING* in a rack at an AT&T wire center. I'm sure it hasn't done anything useful in 40 years, but if you've ever worked at/with AT&T, you'll understand why it's still there. "Not mine. I don't know what it does. I don't want to be responsible for what happens if it doesn't do that anymore. Just build a new rack beside it and leave it alone."
I see a lot of older people here talking about their experiences with these computers, and that's awesome :3
I'm a gen z (20 years old) but find older tech like this really cool, especially love the big bulky monitor aesthetic, always glad to see people making efforts to preserve tech history ^w^
I'm also gen z and intrested like you (I'm 16)
I find it funny that gen z think this is cool. Why? Why do you think it is cool? I'm of the same era as this hard drive and to be honest, my generation are a bit disappointed in the majority of your generation. Not all of them but most.
@@WhatALoadOfTosca This hard drive is cool as a piece of computing history. The engineering and design that went into making something that could store what is now a very small amount of data is fascinating in and of itself, and stands as a testament to how far we have come. As for your last statement, rest assured that equally as many young people are just as, if not more disappointed in people from your generation.
I like their miniaturization. I bet one could store a lot of song lyrics on it for their Centurion pod. Imagine rollerblading down by the beach, with your mini centurion in tow, with a battery powered terminal, reading off all of the era's best song lyrics. 😅😊
maybe could also store the scores, so you could have been singing them as well :)
you might need to add in a white cargo van, but that's not suspicious in any way. At least not for 70's tech. Maybe throw in a waterbed and a Boom BOX.
@@MikelNaUsaCom what you're describing is a much better world than the one we live in now man!
I know you are joking, but 10MB will only store about 5 minutes of audio, even at moderately low quality (8 bit 32Ksamp/sec). So you would probably need that white van with a dozen or so more drives in it to get enough audio to attract the girls. :-)
@@lwilton"lyrics" means text only 😅
I remember seeing my first 10mb hawk drive and thinking, it’s insane how could you ever create 10mb of data without backing up everything 100times. I remember also looking at the ibm 360k floppy disk and wondering what would you be doing moving that much data. How things have changed
It's very weird to take a packet capture and see modern devices generating megabytes of data per second that are just screaming into the void.
Never underestimate the bandwidth of a container ship full of 9 track.
David, that head retract gave me a PTSD! So did reusing that filter! But I do understand...
The Hawk drives are very forgiving. We would go in for a preventative maintenance and find a scuffed platter that was still working perfectly. The customer was able to back up what they needed, generally on floppy, and we'd replace the platter and heads. Phenix drives aren't so forgiving!
I believe the base configuration was a 2.5 meg fixed config without the removable. Not sure if that had a different bowl or some kind of insert. By the late 80s the hawk based systems were end of life, so if the customer was running out of space, we'd flip the switches and give them the whole 10 meg.
Remember, back in those days 10 meg was a lot of floppies! We worked on a lot of systems that didn't have hard drives. So having 10 meg was awesome! Seems like 8 inch floppies were about 160 k!
Hooking up an Oscope would have been a good idea on your alignments. It would help to see the maximum signal. And cool to look at! We used to align drives to a customer's pack when the drive wasn't properly aligned before a crash. Of course we'd align it properly after the data was backed up.
You need a Kennedy tape drive! Still my favorite thing I've worked on!
Great video! Thanks!
Even modern hard drives are forgiving, at least some of the better quality ones.
I only have an old computer (Core i5, 10ish years old) because I am poor. Several weeks ago, it was getting hotter and hotter (the fan was spinning way faster than it used to do before) and getting slower and slower. One day (I am running Windows 10 which is basically rock solid - for a Windows) I saw a blue screen for the first time in the 4 1/2 years I have owned this machine. After that, it won't start anymore. I opened it up and the hard drive was COOKING. I let it cool down and after that it would sometimes hit the windows boot screen and would fail with a hardware error screen. Pretty sure then it was the hard drive.
Bought an SSD, and after setting up a new Windows on it, out of curiosity I hooked up the old hard drive onto the second SATA port - and it was readable! I was able to get back ALL my data, only some times it put read errors but would eventually read it after hitting "Retry" several time. Wow!
So, only thing what really annoys me: why on hell the S.M.A.R.T didn't warn me in time that the hard drive is gonna fail?
+1 on the Kennedy drive! I remember our system vendor using a Kennedy tape unit to backup the drives on a Perkin-Elmer 3205 🤓
@@Oli1974 Did you have that feature turned on in the BIOS? Even so, I'm not sure if it's going to warn you about excessive temps. It's criteria for "impending failure" may be different.
@@russellhltn1396 Yes, that was my first question too, but I checked and yes it was turned on.
I think the hard disk probably still is perfectly fine but due to the heat it started to have problems magnetizising the platter properly, causing the errors in the most written sections first. Pretty sure after a reformat it won't be showing issues anymore.
The drive was so hot I couldn't touch it, so well over 80 degrees C for sure and that must have been way out of spec.
Yes. Can confirm that 8" floppies were 160kb at the time.
This brings back lots of nightmare memories. In the early 1980's, we paired the CDC Hawk drives with either a Texas Instruments Ti/99/10 or a Ti/99/10A CPU. Being forced to run one of these drives on an active construction site with lots of drywall dust was a challenge. I got really good using a small microscope to re-polish crashed heads and a CE Pack to realign the heads. At that time, new heads were $750 a pop, so there was a bit of an incentive to refurbish heads whenever possible.
That's so interesting, what was the use for one of these on a construction site?
@evanbarnes9984 A building automation system. We were integrating HVAC, fire and security systems for two 26 floor office towers and a large retail shop area all under one roof.
11:13 The LM339 is not an OPAMP but a comparator with open collector output. Connecting the outputs together forms a so-called "wired and". The output is high if and only if all four conditions are met (= all outputs are high). It any of the comparators is low, it's output transistor pulls the signal low.
Yeah, that's on me, after having built OpAmps in tubes so much using long-tailed pairs, my brain just defaults that terminology. It was only during editing I realized what I had said and I thought "No one will notice." Boy was I wrong, lol.
That entire hard drive it too small to contain a single Amazon webpage.
Ah yes, the modern web and it's insane data sizes!
@@UsagiElectric Not only the web, almost all modern software is huge. Faster computer results in lazier developers, more layers of frameworks and code and thus way more storage usage than necessary.
Manually aligning the heads on a spinning hard drive has got to be one of the more metal things I've seen on tech hobbyist TH-cam.
That was part of the annual maintenance procedures to service these drives.
A trip back to the good old days of field service, I used to fix those in the early 80's, particularly the Hawks and Tridents, my workmate left his pack of cigarettes inside the cabinet under the Hawk used by ERNIE (the random number generator used by the premium bond system), he wasn't a happy camper that day as we had a tight window to perform a PM - fortunately they were still there when we did the next PM a year later.
This episode means so much to me man. This is exactly what my dad used to do at his first job. He'd go on service calls and repair/align disk drives, although I believe they were something more like IBM 1311s. Originally an electrical engineer, taught himself to program and followed that path. Still bummed he got rid of his SWTPC 6800 sometime in the early 00s. But his working TRS-80 sits behind me as I work in my office.
Thank you for all that you do.
Thank you so much, and I'm glad to hear that this episode brings back good memories!
Hi, David!
Сarbonized textolite conducts electric current. Before installing a new thyristor, it is worth cleaning off the burnt layer between the board tracks. Otherwise, an unpleasant surprise may occur.
is the same for modern PCBs? then again I imagine if theres burnt PCB material on modern stuff its done in most cases since its probably going to be multi-layer PCB for modern electronics and computers.
@@nekomasteryoutube3232 Yes, modern PCBs being burnt causes them to become conductive. Though with modern multi-layer boards, you're usually in for a bad time of grinding down through the board layers and removing all of the conductive material. You can't have layers shorting to other layers, or traces in layers shorting to other traces.
Certain older models of Apple's Macbooks were known to have a fault where inner layers of the board had a short, and it would blow the board apart, either leaving a bulge or blowing off either the top or the bottom of the board.
I definitely did! I just didn't film it is all. I went through with an exacto knife and got rid of the charring. The fiber underneath was horribly discolored, but the charring itself was properly removed.
I would consider a heat sink for that replacement part you soldered in.
Running hot is what killed the original SCR.
@@FlyMIfYouGotM Did CDC change the spec on the item as time went by (or they discontinued completely as computers became smaller form factors)
@highpath4776 The last drives we used actually had sealed HDD's installed on the chassis with a tape drive for backup. I can't remember for sure, but I think they were Sugart drives. Within a year or two, we began using PC's.
I mean, the original part lasted at least 30 years, I'd say it did pretty good!
@@UsagiElectricYes, and it probably even wasn't designed for that lifespan. So why not add a small heatsink, and make it another 60 years? 😊
There may be more issues with that burnt-up SCR. It's supposed to be, I think, a crowbar circuit to trip the breaker if the regulator voltage gets too high. It's supposed to only conduct for like a tenth of a second until the circuit-breaker trips. I used a circuit like that around 1968 but with a 1 ohm current-limiting 1/4 watt resistor in series with the SCR. When I tested it, the resistor exploded and sent resistor shrapnel all over, including near my eyes!
But in your example there seems to be a different problem-- the area is so charred up, one might deduce that the breaker did not trip quickly enough or maybe not ever. So I'd probably test or just replace the associated circuit breaker. It's always something!
I wonder the same thing, 2 of them, both blown, what’s the cause.
Yep, I thought the same... An SCR can only be turned off by cutting the current through it.... It's certainly being used to protect against an "oh f##k" scenario....
@@realnutteruk1 Yeah. My guess is a small electrolytic in the 5-volt regulator got leaky due to sitting around for a few years, so on the next power-up the regulator circuit thought the +5 was waay too high so it tripped the SCR crowbar. But tripped it continuously, and the circuit-breaker is stuck and won't trip, so we're talking SCR-deadly current in a second or two. Not a great design.
Long ago I fixed a $4900 Fluke voltage standard with a similarly-jinxed power supply. They may have put a new-hire electrical engineer on that one.
The SCR successfully carried out its secondary function of acting as a fuse to protect the circuit breaker.
+1 for the Crowbar circuit, have not checked full schematics there, but an SCR behind a bunch of LM399 comparators that seemingly shorts out the supply is like 99.999% a protection device. In this case it was seemingly to weak to trip the breaker. BUT the core issue must be found or the PSU may burn up the drie electronics.
Great episode. I was surprised you didn't check the rubber bumper from the first. You might put that on your checklist. All the drives you work on after sitting for 40 years will need new bumpers.
I was having a problem with saturating both of my brain cells while working on this one, haha.
But you're totally right, checking the rubber bump stops should definitely be a step on the bring up check list!
@@UsagiElectric All important checklist items are generated by not thinking about them previously.
In age of a punch cards - 10 MegaBytes? -_-... For REAL???!!! GIMME TWO!!!
But how damn expensive these were...
This reminds me of the NCR 8250 mini computer, my wife worked on. She did the data enter for a accounting office on a terminal similar to those. She could key very fast an accurately it amazed me. When they switched to micros in mid eighties we got the old computer and drives and discs to play with. My buddy still uses the cabinet in his shop got a mig welder in it, slides out nicely lol.
This is such a great benchmark to learn how hard it was reading and writing 10mb data in that era. Kudos to the creator and the channel for this awesome video 🎉
That BASF label on the lever of the removable platter gave me warm fuzzy feelings of decades long gone, my first own floppy disk I bought for school (we had Apple IIe machines donated by a local bank when they upgraded to something else) to save my work on... Was a 2-pack from BASF. Sweet blast from the past, haha. Had been a computer nerd for a while before we finally got computer science available as a course. I literally knew those machines a lot better than the teacher (back then, there weren't any actual CS teachers around, they were physics or mathematics teachers that took some training to qualify as a CS teacher).
I loved the take before the intro rolled xD "RIGHTTTT?!!!"
Another very interesting episode, David! Enjoyed it a lot!
Am I totally out of my mind when I‘m thinking: “That head doesn’t look mechanically too complicated - can’t they ‘just‘ build a new one from scratch?“ 😳
Thank you!
I mean, that is something we want to do! Though, not quite making one from scratch, more I want to dig in deep into rewinding one that is already blown. But if we could essentially repair bad heads into full working condition, the only that's then keeping us bringing all of these old drives up are platters!
@@UsagiElectric I imagine that at some point someone will be setting up a CVD machine to deposit a fresh coating of magnetic material onto a platter. Should be a lot easier to do today than in the 1970s, that's for sure :)
Same with building a new RW head, considering how primitive those early HDDs were compared to even what we saw in the late 1980s, I'd be shocked if they couldn't be made any more today by a serious hobbyist. Then again I do not have the design specs in front of me, so what do I know :)
@@MayaPosch I mean, we're already looking into how to re-coat a rotating drum, it's not an insane leap to get to spin coating aluminum platters! I have a bunch of crashed ones hanging out, someday we may just dig into them and see if we can get to a point to where there's nothing on these drives we can't restore!
@@UsagiElectricyes, I think repairing blown coils on the head may be the most viable solution. I think an orthodontics or jewellery laser welder might give you a chance to reconnect broken coils.
If that does not work, you may have to drill down into the head material to the coil and use a solder to re-connect the break.
I think winding anew coil around the ferrite might be a but challenging because the you need to build up the whole cover material around the coil. The tolerances for the ferrite positioning are pretty tight.
@@UsagiElectricthat is a sporty challenge. In my experience with these disks they rarely just crashed. Usually by the time the drive was shut down the head had been grinding into the Aluminium. The tolerances are right enough that such ridges could eliminate all the. Work you had put in.
But I think the head supply is the main issue. I think there are still a fair number of disks to be found on the 2bd had matkets, but heads ..., that is not so easy.
When PCB material is charred like that it turns to carbon and becomes conducive. It can be a low enough resistance to heat the PCB and cause it to glow red if there is sufficient power connected to it or it can short out signals. Best to cut it out and glue some replacement material over the hole. Drill it and use wires to remake the traces or perhaps make a new tiny new PCB.
Did I see that the write-protect lamp was still on (28:08) for the drive? Maybe that explains why you could not write anything to the disk?
It is in that shot, but the write failing was confirmed in the memory monitor with the write protect off. It actually illuminates the fault light when trying to write to that specific lower head, which means that the Hawk sees a current problem trying to power either erase coil or the write coil. Thankfully, it still reads just fine, so there is an avenue for usage there.
I think it's getting to be time to have someone dissect a bad head and figure out what it'd take to wind and pot one. Great video, as usual!
That's on the list of things to do! I just gotta get my hands on a proper trinocular microscope so I can bring y'all along on the ride.
@@UsagiElectric Awesomeness!
I remember scrapping one and kept the huge powerful magnets !
You have impressive perseverance and patience.
IPA has an additive that leaves a residue. Most places that I have been, you can buy 100% ethanol at the drug store by asking at the pharmacy counter, and ethanol leaves no residue. We used to use it to clean internal elements on broadcast camera lenses, and after it evaporated, there was no film left on the glass. You might want to consider using it for cleaning platters and heads.
It's actually available on Amazon. It was hard to find during Covid.
IPA sold in the US generally will have no additives. There are variations with respect to the amount of water; "rubbing alcohol" generally will have 70% IPA, with the rest water, while "pure" IPA is sold at 91% or 99%. Either of the latter would be fine for cleaning electronics or disk platters.
In some cases, rubbing alcohol will have other additives, like menthol or other fragrances, but since those will all be 70% concentration anyway, they wouldn't be suitable for cleaning in the first place.
Isopropyl alcohol should not be confused with "denatured alcohol" (as it's sold in the US), which is methyl alcohol that does have various additives, all meant to render the alcohol non-drinkable. One wouldn't want to use that sort of thing for cleaning electronics or disk platters, but it's fine for other kinds of workshop cleaning uses.
Ethanol has a nasty habit of melting plastics. Maybe a good idea for glass optics and/or metal parts but I would be careful around vintage computers.
Ethanol is hygroscopic. It likes to mix with water to such a degree that it will pull water out of the air to dilute itself down to ~95%. That's why Everclear is 190 proof, not 200. You can get it more pure than that, but once it's opened and exposed to air, it's going to regress toward that 95% level.
brake cleaner works fine atleast the brand i use cleans platters very well and no residue
Brake bleeder dust covers would be perfect for bumpers
It's crazy to think that just over 10 years later, the IBM Type 1 MFM hard disk would come out in full height 5.25 inch size, weighing about 10 lbs with a capacity of 10 Mb. I had one somewhat recently in a 286.
With those fancy new IBM drives, my TV would only weigh slightly more than the RMS Titanic
"The removable pack sector transducer is still attached"
yeah man for sure!! no idea what that means but absolutely you're right
Power Supply - when you see blackened / carbonised pcb under, in this case, the scr, the board will almost certainly have become conductive due to the carbon build up. You can check this with a multimeter set to ohms. Where you have 120v mixed in close proximity with ttl control signals this can blow the logic. Also, the board can and will slowly heat up and eventually make smoke :¬). Please go back and check, I used to do this type of repair work and this is a common issue, repair requires milling out the burned area and filling with a mix of resin and filler material - glass beads work well, or you can cut out a patch using layers of fine glass fiber mat.
When I started out 40 years ago I was a field service engineer servicing Hawk and other drives. Not all happy memories - head crash would mean hours of work - careful cleaning followed by a head alignment. First load always an adrenaline rush....
Man, this reminds me of working on the old Burroughs 5-platter drives back in the day. Big blowers before they knew how to leverage the Bernoulli Effect. We had to open them up in a clean-room. Mind the solder joints, make sure none are cold on that high-current PS or you'll regret it!! Test that PS under load before connecting up a drive or you may be buying new heads and platters. SCRs usually don't die without some other cause.
Spent lots of time in the 70's aligning drives like these. The Ampex ones had a special tool used with a 'Cats Eyes' disk. If you got it wrong, it would auto-retract and try and take a couple of fingers with it. Happy days :)
"COPYLEFT, ALL WRONGS RESERVED"
Fascinating - appreciate the work that goes into these videos and - keeping this equipment in working order.
I knew nothing of the Centurion, but I worked for DEC in the UK in the 80's and spent a lot of time servicing RL01 and RL02 drives on pdp-11s - I think they uses a very similar disk pack. They only had the removable pack, no fixed disc like these. RL01's were 5 MB RL02s a massive 10 MB on one platter!
They were incredibly reliable, and very rarely did they suffer from head crashes, unlike the bigger drives. However, one of my customers almost lost a pack to a head crash - and he did not have a backup - something like 5 years of research work! Somehow we managed to recover most of the data using a set of 'sacrificial' heads before the pack gave out completely.
Glad i re watched it... the Bump stop material is SURGICAL TUBING, Sliced with a razor to height, Glue with contact cement.
31:25 "Be vewy vewy quiet, I'm hunting wabbits"
I guess you will have to learn how to fix blown coils on these heads. Unfortunately, the coil repair was not covered in the maintenance course I had in 1988. We just installed new heads from CDC.
I guess there might still be some old CDC engineers around, who might know the production process. The size of the actual magnetic heads on these heads was not yet ludricrously small like on later disk drives. So it may be possible the fabricate them.
I am not sure if the heads are actually covered in glass or if the top material was epoxy.
In any case the coils are only a few windings around that special Japanese ferrite material used in magnetic tapenheads.
Maybe it would be possible to repair some of the coils under a microscope with a master welder. There might be some jewellery companies in your region, who have laser welding equipment and could try to weld the broken coils on these heads.
Alternatively, you would have to strip down the heads and apply new windings then encapsulate the heads again in the cover material and create the right geometry for the spoiler hole on the head to allow it to float on the air vortex over the platter.
I don't think you will get lucky enough to find any new old stock of these heads, as they were extremely fragile and prone to failure. I would think that any still available stock was used up by the mid to late 1990s, as long as the drives were still somewhat in use. I guess that the unavailability of these spare parts led to the last minicomputer to be decommissioned.
So, ultimately only fixing up broken heads can fix this supply problem.
These heads will fail, sooner or later. We exchanged at least 2 heads per customer per year at the time. The heads were as much consumables as the disk packs. So you need a back up solution!
That is why I think your best course of action is to build a μSD card adapter that you can be used instead of the fixed platter, maybe even tie into the read/write path of the CDC drives. Then you still have experience of the drive spinning up, but not the problems associated with the lower platter.
I can see that the visceral attraction of experiencing the spin up of a CDC drive. It is a unique experience, but without a source for these heads it will be a time limited experience.
Good luck!
I wonder about the extent to which these corporations retained such manufacturing data and procedures. Or whether all that stuff just got tossed along the chain of buyouts, mergers, etc. Or if some other supplier made the heads that still has records.
@@KameraShy I think it was probably tossed. The CDC Disk drive manufacturing fiction seems to have been closed already in the 1980s and the company was sold in bits and pieces several times over. I also know that the German IBM disk company was dissolved an the subcontractor making the heads here has moved on from that. All the IBM managers/technologists in that division have been retired for over 20 years.
Nevertheless the production process is still used by tape head manufacturers in Japan and China. The current HDD manufacturers like WD and Seagate are also constructing magnetic heads with a similar technology, but much smaller structures.
The heads got reduced in size to allow for denser writing and reading, there was also a change of the way heads work with the introduction of the Giant Magnetic Resonance technology that upped capacity by a factor of 1000 in the late 1990s.
So, I would think that some Japanese and Chinese companies still make tape heads that have a similar structure to those old HDD heads, although the magnetic tape heads of that time were already bigger than the disk heads, but I understand that the spoiler holes in the heads are still used in modern heads as well and that is the black magic that keeps the heads from crashing on the platters.
Maybe a gofundme campaign could get a few batches manufactured, as technical drawing could be made from existing heads and the electronics in them is no magic. I remember only thst the head contained an inverter chip that allowed the noise reduction ofbthe signal in the amplifier by another inversion and signal subtraction. Naturally all of that analogue.
You are very lucky to access to this computers. Congratulations, great work!.
I wonder if that blown SCR is part of a crowbar circuit to shut down the 5V supply hard, tripping the breaker, in the event of a 5V overvolt. Perhaps the 5V rail has tried to overvolt intermittently in the past but, for some reason, was unable to trip the breaker and cooked.
Oh further reading of the schematic, in fault conditions SCR Q3 actually applies power to a solenoid in the mains breaker to trip the mains power off. Perhaps there is something wrong with the breaker/wiring taking out the SCR in the event of over voltage on the 5V rail.
Gotta love tech history!
It's almost like you're off by one track on the fixed disk on one of the heads.
Awesome video. It reminds me on the early 80's, when I toured the Burroughs Corporation many, many years ago with my grade school classmates.
The platters make great clock faces !!
This hardware - and you videos - are fascinating to me cuz you go "one giant step" farther back from the gear I'm experienced with. I collect and restore PC hardware from the original 1981 IBM PC up to the Pentium III stuff. I've owned a little bit of older gear, a TRS-80 and a couple Apple II's - but I've never had the chance to play with anything like this! We met at VCF East in '23 and hopefully I'll catch you again if you go in '25. Can't make it to Texas though, so I appreciate the YT videos!!
I wonder if you could just use a modern HEPA filter cloth as replacement air filter.
_Probably_ so.
There's a world of different hepa filters out there. No doubt something would work.
@stephendouglas684 The best HEPA media would be the filter media that's used for Bio Safety lab hoods. That media is a high efficiency filter media designed to trap even viral material.
@@FlyMIfYouGotM thar sounds good. With my less than perfect eyesight, I saw several potential spots of debris on thar platter. I think...
@@stephendouglas684 I saw the same thing. Fortunately, his process of running the drive for an hour without allowing the heads to fly, would most likely purge everything off of the platter. This is exactly what I would do when cleaning them in the field without benefit of a proper clean room. When I did this, I would make my own clean room by setting up a sheet plastic lined room around the drive and CPU. I pressurized the pseudo clean room with HEPA filtered air from a small fan. Everything in the room was vacuumed and wet wiped down before opening up the drive. i really didn't miss it when these drives were replaced with sealed HDD's. The funny thing was, at that time, all the corporate gurus told us you couldn't do this in the field. I figured out all of this out without the benefit of any company training. A year later I was finally sent to the company school. I wound up showing the instructors how to field clean, repair and replace heads in these drives. Fun times.
A few years ago I got a Compaq Portable II still rocking its original Miniscribe 20MB hard drive. I well remember my DOS days juggling floppies in a HDless system so it was nostalgic (and more than a little therapeutic) to go back to being efficient with data. The things you can do with mere megabytes if you drop fancy graphical interfaces and media-rich stuff...
I'm wondering about the filters. If they are HEPA, it should be possible to obtain a new cartridge somewhere. If they are not HEPA filters, then a good fine grade roll media might suffice, something like MERV 13 or better? I'd use such media to improvise a "pre-filter" if opening the filter box and replacing the element is not an option. You might also consider using a 3D printer to make a new filter box with a replaceable element.
Some have already mentioned the SCR heat sink that is needed to keep the device cool. Even a simple aluminum angle bracket could serve, along with some thermal paste. Might not be "period correct", but it might also save you an unexpected failure (which, odds are, would happen at a VCF somewhere...)
I love the cat. "SOON!!!"
An awesome piece of tech. Thanks for sharing this and doing such a great job on troubleshooting this...
So fun to watch this. You are inspiring! Did that SCR need heat sink compound?
That Cat Named Sue is eyeing that Koma-san, wondering why he can't have his lunch yet.
Wouldn't you get a better alignment by checking the head signal with a scope?
Just saw the familiar 4 letters:
BASF
Grew up 10ml away from the plant they made them here in Germany. Still remember the impact on the region, when the German chemic giant stopped magnetic storage media in the 80s.
BTW, LM339 is a quad comparator. Not exactly an opamp.
Amazing job!
Heh. When I saw the title of this video what came to mind was some full height drive of maybe 5 or 10MB capacity and with an ST506 interface. I actually have run across drives like this, one where a guy actually had one with legs on it and brought it in to me wanting to know if there was any way that I could interface it to his c64. :-) Then there was a GRI mini in several racks that had a couple of these type of drives in it, I think in that case the capacity was 10MB + 10MB rather than 5MB for each platter such as you describe here. The power supplies were different, being completely enclosed and connecting to the drive with a thick cable terminating in an oddball multi-pin connector. I used to have a platter hanging on my wall that had a head crash on it, not sure whatever happened to that. It's been a really long time since I had to troubleshoot anything like that! Worst I can remember doing is the external HD box that hooks up to my Osborne Executive computer, it had a linear power supply in it and a brief power glitch wiped out the first track on the drive. After putting a switching power supply in the box I laboriously re-built that track using a sector editor, and recovered all of my files. That drive had a whopping 20MB of capacity. There's also a Kaypro 4 around here with a different external drive box that has a pair of 40M drives in it, plenty of capacity for when you're running CP/M.
Great work 🎉
Flawless work !
Ok.... Here may be a fix for that head.... ISO clean the wiring of the coil set on the back of the head.... With the Solder connections. Once clean you load it with flux in the well an do a reheat cycle in an oven...~240 few minutes let it cool and the thermal failure SHOULD be fixed. The Thermal of the write cycles loosens the wire in the solder blob and eventually it goes cold solder joint and breaks from the thermal stress... But if you carefully get the coil set cleaned off, ISO may not be enough to clean it up.. Once all the old flux is gone... you replace it with new and reheat.... Poof new head. Try your hardest to NOT to disturb those coils position. I find a soak with ISO and a blast of compressed air over a series of passes will clear it out. The head is still fine if the insulation you can visually examine is not scorched... Think: Video Card Reheating.😉
These systems are ideal against data theft. Try to lift one of these drives!!!
16:52 it’s WHISPER,QUIET1!
Having schematics of your hardware sounds amazing when used to modern electronics. Imagine, fixing stuff by yourself ? Can't have that in the 2020's.
Are any of the engineers/design personnel who designed the head/drive still living? If you could track one down, maybe they could help reproduce one.
You are the king of keeping of ancient business minicomputer technology alive
I saw a video with surviving engineers getting an ibm 7000 series mainframe and tape drive working. They are in their 80s. Worth a shot
Put an ad in the newspapers for Sun City and The Villages. Somebody may know somebody.
@@KameraShy Though tongue in cheek, survivors are likely to live in both communities.
I have a platter like that, along with a spindle and a hub. Just too cool to toss out.
Man, you are genius!
Beautiful. It looks similar to DEC RL02.
Mr. Hall. The man. The Legend!
Awesome restoration! I wonder what really burned that SCR. It's well worth looking into the root cause.
The mechanical engineering is beautiful here... the rubber is not. Nice fix on those bumps.
The special alignment tool will always remind me of the Monotype casters I worked on, and the enormous number of special single-purpose tools necessary for adjusting and maintaining them.
Interestingly, of the five Hawks I've worked on so far, three of them have popped that very same SCR. That seems to a common failing point among them. I think it's less of an electrical fault and more of a heat thing. That SCR runs hot and after 30 years of heat soaking, it seems to let loose in a wonderfully violent way!
@@UsagiElectric That is true!!
@@UsagiElectric Still - next time you see a dead regulator please double-check whatever it was powering is fine. I.e. there may be marginally bad capacitors which exacerbate the problem.
If those regulators tend to go bad replacing one with a smaller package is not the best idea. I would add an off-the-shelf heatsink if one could fit there.
Charred board segments are also not just an aesthetic eyesore - they can in fact conduct electricity! And it's a nighmare to fix that. In your case the leads are probably far enough for it to matter that much but please be aware of it.
On an unrelated note, could you please update us on the state of emulation, software development, HW/SW reverse engineering, etc the community has done to this point? Thanks!
@@jwhite5008 yes, I was thinking of suggesting cutting the charred part off and soldering cables to an out-of-board SCR placed on a heatsink. This would have to be done after careful analysis of the circuit as for voltages, interference, thermal considerations etc.
@@jwhite5008Glad to see someone else mentioning that the charred PCB material can become conductive. It's always a good idea to remove as much of it as possible. Learned this myself the hard way, although on a much less important piece of equipment - a car audio amplifier.
I've actually aligned floppy heads just by feel - in DOS, keep reading the directory until you get from General Failure to Sector not found, then when you can get a directory, read/write a large file (execute it) and to be sure, format a floppy in it and see if it'll boot on another machine. I don't know how to do it properly TBH but it looks like the way is very similar in this Hawk Drive.
The proper way of aligning the heads on a floppy drive requires an oscilloscope and a practically unobtainable alignment disk.
@@rocketman221projects With a scope? Interesting.
@@the_kombinator CuriousMarc has a video showing how it's done. The alignment disk has an analog waveform that you view on the scope to see when the heads are aligned. The disks can't be copied and there were not very many made. If you don't have an alignment disk, you can use a known good disk and just look for the strongest signal to get close enough.
for text based storage... 10 mb is alot.
Oh totally! It's a massive amount of space!
14:30 It's like working a Demon Core!
I worked in Mil. Ind. (BAE) on one of the projects used these monster discs and boy were they delicate!
A hair would upset them.
Well back in my Hayday when my repairshop was running full steam I was rwepairing a lot VTRs and VCRs . I bought chamois in bulk . Anyway that SCR got cooked I can't believe none the traces got damaged .
I absolutely love these bulky machines. Designed to one only thing, but to do it very well.
@ 12:10 - I would suggest a heat sink as a mod: I think I'd have mounted it vertically with a clip on HS.
You should put a small makeshift radiator on the SCR. It seems to be always in pain.
I recognize that brown colour :)
Those trace stubs on a PCB indicate the pin 1 for chips, and orientation for diodes (anode or cathode - whatever their marking convention is).
You know, if you got one of those AC filters, I'm fairly confident you could find one of that size, or close enough you could shave the excess off without compromising it, that should work as an effective replacement.
Love your videos, I date to the dawn of the PCs, worked on a number of TRS 80s. I was looking forward to seeing you in Baltimore but as fait would have it someone close to my wife and I had a major life event the Saturday the 18th. Oddly I was customer of Bob's in the Computerland era, so wold be good to catch up with him as well. In any case great video as always.
One bad coil don't spoil the whole bunch girl haha you are just as much nerd as me love the content
As interesting as ever, hope your search is successful... Ryn
You and me both!
Thanks!
@@UsagiElectric Heads - If P/N 70599901 NSN 7025-01-155-5352 there is a pair on buy now old s/h stock look good, they are CDC but might not be for Hawk
@@highpath4776 Thanks for the lead. Checked and they are not for the CDC / MPI Hawk drive.
@@kenromaine2387 OK, they )or someone similar) had a 10 head device for one of the other CDC multi disk units. There was a complete Hawk that appeared to be from a Harris system (same colour blue?!) but that has sold now.
WOW, thank you for saving these machines for posterity.
Cute rabbit and cat 🤗🤗🫠🫠
I had two questions
1. have you tried moving the lower head from the removable disk to the fixed disk to get the write completed?
2. did you try realigning the head to get the data off the original bad sector? Theoretically, if you can get a complete image you can write it all back no matter what the head alignment is.... providing the heads can write/erase data
OMG This "Hellord" T-Shirt!
Facinating.
LM339 is a quad comparator. The outputs are connected in a wired or configuration.
Man, usually when i think of computers from the 70s i think of the massive room sized computers. I wasn't expecting how relatively small some of it was.
Note- i know near nothing about older computers past 2000s/ late 80s computers
For a man that hasn't used anything older than Basic 2000/Lambda 8300(ZX-81-clone, my 1st computer back in 1984), this all is works of Black Magic. :D
I am sure I sent you a link for NEW OLD STOCK CDC HEADS (though I didnt realised Top and Bottom heads were different)
just incomprehensible that 10mb was a huge amount back then, and i even lived through that time
The tails on the pads are polarity / pin 1 indicators
Do some livecoding on Twitch or on a second channel since you're doing dev work on those beasties anyway.
I'd watch.
As someone who codes for fun sometimes, it floors me that there's a market for people wanting to watch other people do that. It seems to me like, watching me rewrite a first-draft conditional block into something more elegant and organized would be like watching my SO load the dishwasher, or watching someone play Tetris.
@@nickwallette6201 For me, I learn better if I can watch someone do something a few times first before trying it myself. For someone who wants to learn to code who learns like that, watching someone else do it in real time would be SUCH a valuable resource.
I get that it’s not everyone’s thing though.
@@nickwallette6201 Me too when I first heard of it. But these days I only code for fun and at some point I watched a bit of one.
Now I watch them all the time, usually sped up in the background when I'm coding or when I feel like coding. There's plenty that are terrible but watching people 10x better than me or working on stuff I always thought was too hard or just never got around to can be pretty inspiring. The different thought processes and problem solving techniques can also be fascinating. And it's also refreshing to see people who can code much better than you still struggle with stuff or miss obvious bugs, and to find and fix problems I might've given up on.
10 MB on 2 big platters. My HS in 1981 had a Corvus HD hooked up to half a dozen Apple ]['s. It was partitioned into 60+ 143k Apple floppy disk images, allocated to students rather than giving them easily-damaged floppies.
Yeah heads are going to be very difficult unless it's something like a Diablo Series 30 or an IBM 2315 where the heads are still relatively common.
Ouch! Best of luck!
I wonder how possible new fabriaction of a head is...
Welcome to Control Data Institute.
3:57 : that moment when you're reading Fallout terminal logs...
"format c: /s" of course, lol. to write the MBR.
that is the grandson of the programs to do that job on these particular systems
Don't forget "fdisk /mbr" -- the sneaky one that seems to catch a lot of retro YTers from getting a random old drive to boot successfully.