A few viewers complained about the bad frames that had glitched somehow into the upload. I am quite sensitive to optically induced migraines myself so out of an abundance of caution I re-uploaded. You'll have to post your awesome comment again. And @MrFujinko keeps his free vintage electronics oil change award for his previous "First!" comment!
the reupload happened durring my watching of that scene with the glitches, and then the video stopped loading so i refreshed and was private, glad to see its just a reupload and not something else
@@kevmert Yeah I know, re-uploads are so disruptive. I wish there would be a smoother way to do it while preserving the comments and the original link.
@@AUATUWVSH In my case I was trying to watch RUclips on my Switch sending to my TV which is something I've not done before, so I was wondering if it was being buggy. Good to know it's the source video.
The magnetic clutch - combined with the separate, continuously rotating motors, one for forward and one for reverse - is a great example of how IBM used what might be called brute force to jump ahead by twenty or thirty years of technological development. Of course this required an enormous amount of money to accomplish at the time. Those clutches undoubtedly came from some vendor, and were probably custom made to IBM's specifications. Then when you add the costs of assembly and testing, as well as the maintenance chain to keep them working at customer sites, not to mention the chilled air supply that was necessary to keep them cool, and you can imagine how much the total was for each tape drive; many large customers (like the government) had dozens or scores of tape drives running. The vacuum columns are another - and equally impressive - story, in my opinion. I don't know about the drives that were associated with the 1401, but the drives used in the 360 (in the late 1960s) had vacuum columns that were capable of shuttling the tape back and forth at a rate too fast for the eye to follow; the tape in the column was just a blur as it appeared to vibrate over a length of eight or nine inches. When this happened, they made an almost musical fluttering sound, hard to hear over the din of cooling fans and rushing air in the computer room. All this to get a technological advantage over their competition, which was coming on fast. IBM was ahead, and their management knew how to use their financial strength to hold on to their advantage, by heroic means sometimes, as illustrated by this video. As a child, I was taken by my father to see many IBM sites in the 1960s, and worked at one in the 1970s. The actions of the tape drives was one of the most fascinating and alluring aspects of the IBM computers of that time. There was no other mechanical device that moved anything like those tape reels, which (when full of tape) weighed around five pounds.
All of the mainframe companies of the era used similar drives. IBM originated most of the ideas that made them work, but other companies licensed the patents.
It has got to be nerve-wracking to restore mechanical computer components that old. When you opened up that clutch and I first saw all that powder, I thought "Oh poo...A brake pad or seal has disintegrated." The magnetic powder clutch is ingenious. I'm glad you had access to decent maintenance notes and the original technical manuals, and kudos to your troubleshooting/detective work on the felt seals. Thanks for the video and thank you for helping preserve important computer history!
My university lecturer in the 1980s told us that the computer tape drives in the movies were silly props. There was no way that the tape could shuttle backwards and forwards like that without snapping the tape. Of course he didn't know about the vacuum columns and wonderful mechanisms inside the reel drives of these machines, which were very real indeed.
You would think that. I had an engineering professor claim the factor of safety in modern fighter jets was 1.05. Every public resource I could find claims 1.5. Very painful to see people with influence say such things in front of rooms of students who won't check.
I went to the CHM a few weeks ago and dutifully punched my name and date of birth onto a punch card in the 1401 room. You can see the keypunch and pile of blank cards at 0:14.
Hey Marc, long time viewer here. I was actually a very young man when these came out, my first job was in my college working in the datacenter running punch card jobs. When the school upgraded their mainframe I 'finagled' possession of the old one. It was a Control data 6400 full timeshare system complete with 8 of those vacuum column drives (The CDC version). Same design. They used vacum tube logic back then. Had it in my Mom's garage for several years. Fun to work on. Could only run it sparingly because of the power requirements, but fun while it lasted! Nowadays I work on modern enterprise stuff, but loved the days when you could go on a call with rolled up blueprints in one hand, and an O-Scope in the other....
A very similar magnetic clutch was used on the Fiat 131 Supermiafiori racing supercharger. I had one of those in bits and it had the same felt pad issue.
Time to re-load my original comment, then: I always wondered, as a kid, how that captivating tape drive motion was accomplished; now I know. Thanks, Marc, for this latest chapter of Belgian Betterment!
Wow. I learn so much watching these videos. Magnetic powder controlled clutches. What other mysteries surround me that I have no clue about? Although this is older technology, the principle holds.
All John Deere lawn tractors these days include a magnetic clutch, with magnetic powder inside. It's how they bring the mower blades to an instant stop the moment the user does anything dangerous, like going in reverse while mowing. It's an impressive example of electromechanical technology, and has been around a long time, as this video proves.
I had always looked at those electric clutches as something I don't understand yet and would imagine how they work. I even showed my best friend Thais your videos years ago to show that there are electric clutches, I even found a few at a scrap yard I planned to take apart later, but I had no idea they operated on magnetic powder. Magnetic powder? what in the world. I let myself ponder them for years and yet only figured they worked with electromagnetic-actuated-plates not powder. The magnetic powder thing just really surprises me. I love it
I think there are different kinds, and that some are indeed operated with plates. This is kind of the über category, faster, better, more precise, and expensive…
The Computer History Museum is also right down the street from Shoreline Ampitheater! I went there in 2019 and it was a great way to spend the afternoon before a concert. I saw the 1401 demo, played SpaceWar and was blown away by the exhibit hall. It is a fantastic museum for anyone interested in computers.
Finally, a mainframe repair that I'm confident I could accomplish. One of these days I'll make the trip out to the museum and see it in person... And that has to be the chonkiest data plug I've ever seen!
Having turned 61 in early 2022, I have seen machines like these numerous times on television and in movies. I think I have also seen them a few times in person, ages ago. How absolutely _DINOSAURIC_ these things must seem to younger people today.
Great video Marc, just a small observation, at minute 18:22 you state that you are reconnecting the motor that drives the stop clutches; this is actually the motor that brings the heads down and up. The motor that drives a worm gear shaft that engages the stop clutches, to bring the tapes down to the columns, and back up is further to the front, facing up, not using a belt. Hard to see in video.
Fascinating. It's astonishing just how much we were able to miniaturize all of those components to create such commonplace storage media as cassette tapes. It's invaluable to retain such historical articles of computing history so we can show where we've come from. You never know when some young genius engineer will see the future of technology lying hidden and forgotten in such relics of the past.
You are most welcome. IBM 1401 demos are Wednesdays at 3:30p and Saturdays at 11 am. If you come early on Wednesdays you might catch us working on the system!
I'm always up for some percussive maintenance. When seals fail, the rest of the system is thoroughly borked. A clutch is such an important part of the drive mechanism, when it malfunctions all is lost. Doubly so on the IBM 729, a trusty companion to the 7 line of early computers. Glad to see a classic bench instrument (HP power source) to repair a classic computer component!
My curiosity for vacuum column tape drives is the reason I originally found your channel. Love those crazy machines. The clutches are a peculiar design. In cars powder buildup is a bad sign in dry clutches, so when you first opened this one up I thought "that thing is wrecked", but then you explained the whole magnetic powder mechanic and it made so much more sense. I'm not old enough to have ever had a chance to work with these types of drives, but when I worked at MAN in Copenhagen in the late 90s they still had a (much newer model) vacuum column drive sitting in a corner just in case they ever needed to access the old tapes. Of course by then they were running a huge Hitachi mainframe (which as we know was basically just IBM made in Japan) with a massive pile of pocket size harddrives to power their aging CADAM terminals along with being the brains of the SAP and CICS databases. A few years later the mainframe went out and they switched everything to PC servers, but for me who's into the whole history of computers it was nice to work with the systems at the end of their era, and then learn from the likes of you how they started out long before I was born.
This video really makes you realize just how complicated and advanced these machines were, especially for the 50's. Many people I know tend to think of old computers as being almost coal-powered, and the tape drives as being glorified reel-to-reel audio players. They were SO much more than that.
I have so much respect for the early computing engineers - truly amazingly innovative. We are in danger of losing this knowledge so your channel and all the work you do are vital.
Fascinating. I have always wondered how those old tape drives were able to start, stop and change direction almost instantaneously. Thank you for your time and effort to share this video.
I had no Idea how these Magnetic clutches work and that they are filled with magnetic powder. What a genius construction! I'm happy to see the 1401 again. The Fortran video may have been my first of this channel. Good thing that I don't have to repost my comment about flickering.
This is truly amazing video! I love the clarity of explanation and the almost surgical precision used to extract the malfunctioning parts and fix the problem! Thank you for sharing this! ~ VK
It doesn't matter how much time passes, these tape drive cabinets always look so cool to me. Someone should manufacture one for visual effect only.. I'd buy it :)
Next video should be about replacing CRT fluid and floppy drive batteries. Has anyone ever changed the belts on a computer terminal? I hear they stretch out and the bits get delayed because the pulleys slip.
Funny, but I have once repaired a radar screen with a belt (a toothed one). Beam rotation was done by a motor rotating the coils. Lubrication and retensioning was needed as the belt would slide around the gear...
on the micro computers that used compact cassettes for storage a lose or slipping belt could cause unpredictable read errors - simply because they read everything as audio that was then demodulated - if the belt wasn't the correct tension it affected the tape speed which made the tone of the audio change, which resulted in a different data value. If this speed change was subtle enough it would still read as valid data, even if it was wrong, and made for some hilariously difficult bug fixing. So I have had to change belts on a micro computer, but not so much a terminal
Strangely enough, I've replaced many belts on Amstrad PCWs, CPC 6128s and Sinclair ZX Spectrum +3s. The Hitachi 3" floppy drives are belt driven, and they perish with age. The good ones harden and split into pieces - the bad ones turn to a horrible sticky goo that takes ages to remove from the pulleys.
I've de-rooted the logic board on my Datamedia DT80 I pulled from a creek. As in, pulled roots from under the chips. Followed by reverse engineering and fixing it enough to run an ASCII slideshow. Can't find the original ROMs though so it no longer functions as a terminal until I write my own or find a copy.
You joke, but old rear projection televisions and CRT projectors use a fluid filled lens to help cool the phosphor screen. Sometimes the fluid degrades and needs to be replaced.
The most interesting thing to me about the IBM 1400 system was how long some of the related devices were in service. I know customers that used the 1403 printers and 1419 document processors (cheque sorter) well into the 1980s.
@@Rob2 Originally they were only leased, but the anti-trust action forced them to sell some of them. IBM would still to destroy them at end of lease them to prevent them appear from appearing on the secondary market. Somehow these one escaped and were bought used.
One of the hardest Y2K problems apparently was the 1400 series code running in emulation on 360 mainframes. That's "only" the software, but it still is amazingly long lifespan for something that probably wasn't thought of more than something to rewrite later.
Excellent video! It reminded me of my first day as a Sperry Univac customer engineer in the late '70s when I replaced a similar clutch on the Uniservo tape drive. Thanks!
A good opportunity to watch it again xD, fantastic machine, i always wonder how was it possible to move that violently in both directions, now I know, as usual a great video, looking forward for the next, thank you
It's so great to see the 1401 featured again; the earlier videos you made on them introduced me to your channel. As said by another commenter I hope to be able to see it in action myself, one day.
I can’t think of anything cooler in computing than looking at these tape drives, I know how they work (superficially), and I wish I could create a small scale replica for regular tapes.
I love old electromechanical stuff like this. I can follow along with the logic, step by step, and understand the way it's operating. Which makes working on them so much easier than a modern computer
My company uses magnetic powder brakes on the unwinds for our coating lines. Eliminates all chatter, glazing, slip, etc with conventional unwind brakes. But we run our webs slow enough that the units weren't always super happy and struggled to control. Solution? Constantly tumble the powder inside with a separate drive motor.
Fantastic video Marc. I remember seeing these drives in a visit to a computer room in the 1960s and it felt like being transported into the future. The state of the art back then was very visual and the incredible engineering was mind blowing. Today it's SMD and a whole load of code - equally mind blowing but much more difficult to appreciate.
I never thought much about the mechanism behind those drives. It really is quite extraordinary that such a precise device was built out of such non-precise parts.
What can I say having worked on Monotype lead type casting machines, rebuilt and repaired bikes, serviced household appliances and gastronomy equipment, let alone all these hi-fi gear repairs? I love this one :). Filling that clutch is kinda like BGA reballing, isn't it? :) Damn awesome repair job. Makes me wish I had the opportunity to work on tech like that.
21:41 It's not an USB cable with only 4 pins !😅 Slow frequency need more wires for the data bus ?! Right ? Very impressive video. I have immense respect and admiration for your work on this channel. Hello from France.
n the 1990's i had a job of removing various parts from 8 large IBM reel machines like the ones here,just newer models,prob 1980's,for Memorex,they only needed a few pcbs the rest was scrapped,i asked if i could have the DC motors and various parts,had the motor control boards as well,huge aluminium heatsinks,so well made,i still have the motors & Driver pcbs today and they are 19v multi pole brushed,electromagnet poles,weigh about 15kg each,ive made various things with some of them over the years ,they can take 1000 watts for several minutes without getting too warm.I also had 8 vacuum pumps which were really well made,unfortunately all of them were stolen along with aluminium sheets around 1994 from my lock up.I worked till 11-00 at night that day, as the company was getting the chassis collected in the morning, Happy days.
Thanks SKF for creating spherical ball bearings and creating/naming car maker Volvo, which as many of you know means "I roll". Really fitting because of this video and its car innuendos 😀
Marc, what was the manufacturer of the original bearing ? If SKF, the bearing should have a number stamped on the side of one of the races, a number which could help obtaining a new alike one. If not, what was dimensions ? I could ask the local SKF distributor here.
Marc at minute 18:24 you mention that you are reconnecting the motor to drives the stop clutches. This is not correct, that motor brings the head guide assembly and stop capstans into operation. The stop motor connects via a shaft with worm gears, and is located above and to the front, hard to see in the pictures. I thought that I had entered this comment before, by your instructions.
As one of my long-ago professors said: "it's so slow, it looks incredibly fast." This when describing early computer electromagnetic devices, when compared to more modern electrooptics. He made that remark when explaining the operation of a big-screen display device that used a laser in conjunction with a pair of galvo-motor beam steering mirrors, but the principle is the same as one sees when watching the remarkable accelerations of the tape reels and vacuum columns. They are only "slow" when compared with the speed of the digital electronics to which they feed data.
If you do it for long enough, eventually the looping music will be accompanied by explanations of whatever is wrong with what you're working on at the time.
I'm watching Season 5, Episode 3 of the old television series _Mission: Impossible_ , “The Innocent” (1970), where several several tape drives of a mainframe computer are shown. However, these props are shown with the reels continuously rotating instead of the more-accurate stop-stop action.
Strange that solution with the powder. I thought it works like those automotive AC compressor clutches. That does not need rings and brushes either... the coil is stationary there. It wasn't invented that time?
They were, but they tend to exhibit significant levels of snatch when engaging - not a problem for an A/C compressor, but possibly harmful when driving reels of rather fragile tape. Hence the magnetic powder clutch - they have extremely smooth torque takeup, which is why they are still used in material handling systems.
It's truly remarkable what the mechanical engineers of yesteryear were able to accomplish without the high power semiconductors that are freely available today....Geniuses indeed!
I love the IBM 1401 - I use it as my zoom background at work and try to convince the youngsters that it’s actually the computer that runs our database server. 😜
Completely unrelated to the video but; What make and model is the laptop at 10:00 next to the magnetic clutch dust at the top left of the frame? I like its aesthetics haha.
Is there any chance you can just remove the dust shields from the original bearing, wash out all the magnetic powder, and re-lube it and put the seals back in? I'm sure the 2 bearings back to back will be just fine, but I doubt the original is actually "shot" and instead just plugged up with the powder.
Maybe next time you are maintaining this area/piece, you can use the innards of the new SKF bearings, to rebuild the old seized bearing? This way youll have the original bearing??? Just an idea. You seem to have all the tools needed to rebuild the bearing properly. Also you can use contact cleaner or engine degreaser and soak the seized bearing to hopefully remove/extract all the foreign material/magnetic powder that got inside the bearing. Although i would probably rebuild it.
A few viewers complained about the bad frames that had glitched somehow into the upload. I am quite sensitive to optically induced migraines myself so out of an abundance of caution I re-uploaded. You'll have to post your awesome comment again. And @MrFujinko keeps his free vintage electronics oil change award for his previous "First!" comment!
the reupload happened durring my watching of that scene with the glitches, and then the video stopped loading so i refreshed and was private, glad to see its just a reupload and not something else
@@kevmert Yeah I know, re-uploads are so disruptive. I wish there would be a smoother way to do it while preserving the comments and the original link.
oh thats what that was, i thought my GPU was flipping out haha
@@AUATUWVSH In my case I was trying to watch RUclips on my Switch sending to my TV which is something I've not done before, so I was wondering if it was being buggy. Good to know it's the source video.
I was on the verge of thinking it was an effect on how fast the clutches work ! :D
The magnetic clutch - combined with the separate, continuously rotating motors, one for forward and one for reverse - is a great example of how IBM used what might be called brute force to jump ahead by twenty or thirty years of technological development. Of course this required an enormous amount of money to accomplish at the time. Those clutches undoubtedly came from some vendor, and were probably custom made to IBM's specifications. Then when you add the costs of assembly and testing, as well as the maintenance chain to keep them working at customer sites, not to mention the chilled air supply that was necessary to keep them cool, and you can imagine how much the total was for each tape drive; many large customers (like the government) had dozens or scores of tape drives running.
The vacuum columns are another - and equally impressive - story, in my opinion. I don't know about the drives that were associated with the 1401, but the drives used in the 360 (in the late 1960s) had vacuum columns that were capable of shuttling the tape back and forth at a rate too fast for the eye to follow; the tape in the column was just a blur as it appeared to vibrate over a length of eight or nine inches. When this happened, they made an almost musical fluttering sound, hard to hear over the din of cooling fans and rushing air in the computer room.
All this to get a technological advantage over their competition, which was coming on fast. IBM was ahead, and their management knew how to use their financial strength to hold on to their advantage, by heroic means sometimes, as illustrated by this video.
As a child, I was taken by my father to see many IBM sites in the 1960s, and worked at one in the 1970s. The actions of the tape drives was one of the most fascinating and alluring aspects of the IBM computers of that time. There was no other mechanical device that moved anything like those tape reels, which (when full of tape) weighed around five pounds.
All of the mainframe companies of the era used similar drives. IBM originated most of the ideas that made them work, but other companies licensed the patents.
I think IBM was the only company to use vacuum columns instead of spring loaded arms to tension the tape though.
It has got to be nerve-wracking to restore mechanical computer components that old.
When you opened up that clutch and I first saw all that powder, I thought "Oh poo...A brake pad or seal has disintegrated." The magnetic powder clutch is ingenious.
I'm glad you had access to decent maintenance notes and the original technical manuals, and kudos to your troubleshooting/detective work on the felt seals.
Thanks for the video and thank you for helping preserve important computer history!
My university lecturer in the 1980s told us that the computer tape drives in the movies were silly props. There was no way that the tape could shuttle backwards and forwards like that without snapping the tape. Of course he didn't know about the vacuum columns and wonderful mechanisms inside the reel drives of these machines, which were very real indeed.
One would expect university lecturers get their facts checked well, figures.
Presumably not a computer class, or any engineering class.
You would think that. I had an engineering professor claim the factor of safety in modern fighter jets was 1.05. Every public resource I could find claims 1.5.
Very painful to see people with influence say such things in front of rooms of students who won't check.
The controller block pulled out revealing a console is some hardcore sci-fi design. Love it.
I went to the CHM a few weeks ago and dutifully punched my name and date of birth onto a punch card in the 1401 room. You can see the keypunch and pile of blank cards at 0:14.
Hey Marc, long time viewer here. I was actually a very young man when these came out, my first job was in my college working in the datacenter running punch card jobs. When the school upgraded their mainframe I 'finagled' possession of the old one. It was a Control data 6400 full timeshare system complete with 8 of those vacuum column drives (The CDC version). Same design. They used vacum tube logic back then. Had it in my Mom's garage for several years. Fun to work on. Could only run it sparingly because of the power requirements, but fun while it lasted! Nowadays I work on modern enterprise stuff, but loved the days when you could go on a call with rolled up blueprints in one hand, and an O-Scope in the other....
I'm amazed at how fast you can switch magnetic clutches. Added a new level of wow to watching the reels change direction.
Yeah I never even considered there was such a massive mechanical part to rotating those reels, that's quite amazing!
A very similar magnetic clutch was used on the Fiat 131 Supermiafiori racing supercharger.
I had one of those in bits and it had the same felt pad issue.
@@adelestevens so we went from a computer having the same problem as a car to cars having the same problems as computers XD
Time to re-load my original comment, then: I always wondered, as a kid, how that captivating tape drive motion was accomplished; now I know. Thanks, Marc, for this latest chapter of Belgian Betterment!
Wow. I learn so much watching these videos. Magnetic powder controlled clutches. What other mysteries surround me that I have no clue about? Although this is older technology, the principle holds.
The technology is still alive and well today in cars, in magnetically controlled shock absorbers, otherwise known as active suspensions!
All John Deere lawn tractors these days include a magnetic clutch, with magnetic powder inside. It's how they bring the mower blades to an instant stop the moment the user does anything dangerous, like going in reverse while mowing. It's an impressive example of electromechanical technology, and has been around a long time, as this video proves.
I had always looked at those electric clutches as something I don't understand yet and would imagine how they work. I even showed my best friend Thais your videos years ago to show that there are electric clutches, I even found a few at a scrap yard I planned to take apart later, but I had no idea they operated on magnetic powder. Magnetic powder? what in the world. I let myself ponder them for years and yet only figured they worked with electromagnetic-actuated-plates not powder. The magnetic powder thing just really surprises me. I love it
I think there are different kinds, and that some are indeed operated with plates. This is kind of the über category, faster, better, more precise, and expensive…
The Computer History Museum is also right down the street from Shoreline Ampitheater! I went there in 2019 and it was a great way to spend the afternoon before a concert. I saw the 1401 demo, played SpaceWar and was blown away by the exhibit hall. It is a fantastic museum for anyone interested in computers.
I loved my visit, scheduled an extra day in the Valley for it.
Finally, a mainframe repair that I'm confident I could accomplish. One of these days I'll make the trip out to the museum and see it in person... And that has to be the chonkiest data plug I've ever seen!
Having turned 61 in early 2022, I have seen machines like these numerous times on television and in movies. I think I have also seen them a few times in person, ages ago.
How absolutely _DINOSAURIC_ these things must seem to younger people today.
I have to admire the graceful precision the simple mechanical concept achieves with minimal electronics.
Great video Marc, just a small observation, at minute 18:22 you state that you are reconnecting the motor that drives the stop clutches; this is actually the motor that brings the heads down and up.
The motor that drives a worm gear shaft that engages the stop clutches, to bring the tapes down to the columns, and back up is further to the front, facing up, not using a belt. Hard to see in video.
As always you are right Iggy. You are also in the footage if you look closely!
Fascinating. It's astonishing just how much we were able to miniaturize all of those components to create such commonplace storage media as cassette tapes. It's invaluable to retain such historical articles of computing history so we can show where we've come from. You never know when some young genius engineer will see the future of technology lying hidden and forgotten in such relics of the past.
Those drive belts remind me more of an old fashioned V-belt car accessory drive than anything computer related!
Wonderful repair Marc! Thank you for sharing! Hope to see those 1401s in action one day.
You are most welcome. IBM 1401 demos are Wednesdays at 3:30p and Saturdays at 11 am. If you come early on Wednesdays you might catch us working on the system!
This further proves what I already knew... Marc can fix anything that runs on electricity!
As a mechanical engineer following the channel for quite some time this is definitely the most relatable episode yet 😀
I'm always up for some percussive maintenance. When seals fail, the rest of the system is thoroughly borked. A clutch is such an important part of the drive mechanism, when it malfunctions all is lost. Doubly so on the IBM 729, a trusty companion to the 7 line of early computers. Glad to see a classic bench instrument (HP power source) to repair a classic computer component!
My curiosity for vacuum column tape drives is the reason I originally found your channel. Love those crazy machines. The clutches are a peculiar design. In cars powder buildup is a bad sign in dry clutches, so when you first opened this one up I thought "that thing is wrecked", but then you explained the whole magnetic powder mechanic and it made so much more sense.
I'm not old enough to have ever had a chance to work with these types of drives, but when I worked at MAN in Copenhagen in the late 90s they still had a (much newer model) vacuum column drive sitting in a corner just in case they ever needed to access the old tapes. Of course by then they were running a huge Hitachi mainframe (which as we know was basically just IBM made in Japan) with a massive pile of pocket size harddrives to power their aging CADAM terminals along with being the brains of the SAP and CICS databases. A few years later the mainframe went out and they switched everything to PC servers, but for me who's into the whole history of computers it was nice to work with the systems at the end of their era, and then learn from the likes of you how they started out long before I was born.
This video really makes you realize just how complicated and advanced these machines were, especially for the 50's. Many people I know tend to think of old computers as being almost coal-powered, and the tape drives as being glorified reel-to-reel audio players. They were SO much more than that.
I have so much respect for the early computing engineers - truly amazingly innovative. We are in danger of losing this knowledge so your channel and all the work you do are vital.
Fascinating. I have always wondered how those old tape drives were able to start, stop and change direction almost instantaneously. Thank you for your time and effort to share this video.
I had no Idea how these Magnetic clutches work and that they are filled with magnetic powder. What a genius construction!
I'm happy to see the 1401 again. The Fortran video may have been my first of this channel.
Good thing that I don't have to repost my comment about flickering.
At long last a CuriousMarc repair that I can totally understand. Love those electro-mechanical devices.
This is truly amazing video! I love the clarity of explanation and the almost surgical precision used to extract the malfunctioning parts and fix the problem! Thank you for sharing this!
~ VK
Your mechanical skills are excellent.
It doesn't matter how much time passes, these tape drive cabinets always look so cool to me. Someone should manufacture one for visual effect only.. I'd buy it :)
That load/unload procedure is so damn satisfying to watch. Amazing machines!
I really love your pre-historic 'USB' cable and plugs!!!
It's really hard to imagine that modern HDMI connectors have *more* pins than those beasts.
unusually stout bus?
@@Stoney3K I know right! Modern precision manufacturing is crazy intense.
Next video should be about replacing CRT fluid and floppy drive batteries. Has anyone ever changed the belts on a computer terminal? I hear they stretch out and the bits get delayed because the pulleys slip.
Funny, but I have once repaired a radar screen with a belt (a toothed one). Beam rotation was done by a motor rotating the coils. Lubrication and retensioning was needed as the belt would slide around the gear...
on the micro computers that used compact cassettes for storage a lose or slipping belt could cause unpredictable read errors - simply because they read everything as audio that was then demodulated - if the belt wasn't the correct tension it affected the tape speed which made the tone of the audio change, which resulted in a different data value. If this speed change was subtle enough it would still read as valid data, even if it was wrong, and made for some hilariously difficult bug fixing. So I have had to change belts on a micro computer, but not so much a terminal
Strangely enough, I've replaced many belts on Amstrad PCWs, CPC 6128s and Sinclair ZX Spectrum +3s. The Hitachi 3" floppy drives are belt driven, and they perish with age. The good ones harden and split into pieces - the bad ones turn to a horrible sticky goo that takes ages to remove from the pulleys.
I've de-rooted the logic board on my Datamedia DT80 I pulled from a creek. As in, pulled roots from under the chips. Followed by reverse engineering and fixing it enough to run an ASCII slideshow. Can't find the original ROMs though so it no longer functions as a terminal until I write my own or find a copy.
You joke, but old rear projection televisions and CRT projectors use a fluid filled lens to help cool the phosphor screen. Sometimes the fluid degrades and needs to be replaced.
The most interesting thing to me about the IBM 1400 system was how long some of the related devices were in service. I know customers that used the 1403 printers and 1419 document processors (cheque sorter) well into the 1980s.
This system was in service right up into the 90s I think!
@@CuriousMarc Weren't those devices usually leased? That would have cost the user a lot of money over that time...
@@Rob2 Originally they were only leased, but the anti-trust action forced them to sell some of them. IBM would still to destroy them at end of lease them to prevent them appear from appearing on the secondary market. Somehow these one escaped and were bought used.
@@CuriousMarc I was worried that the user leased it for 30 years, but apparently not. Good.
One of the hardest Y2K problems apparently was the 1400 series code running in emulation on 360 mainframes. That's "only" the software, but it still is amazingly long lifespan for something that probably wasn't thought of more than something to rewrite later.
I love these tape drives. They are just so mesmerising to watch
Excellent video! It reminded me of my first day as a Sperry Univac customer engineer in the late '70s when I replaced a similar clutch on the Uniservo tape drive. Thanks!
A good opportunity to watch it again xD, fantastic machine, i always wonder how was it possible to move that violently in both directions, now I know, as usual a great video, looking forward for the next, thank you
It's so great to see the 1401 featured again; the earlier videos you made on them introduced me to your channel. As said by another commenter I hope to be able to see it in action myself, one day.
I can’t think of anything cooler in computing than looking at these tape drives, I know how they work (superficially), and I wish I could create a small scale replica for regular tapes.
I love old electromechanical stuff like this. I can follow along with the logic, step by step, and understand the way it's operating. Which makes working on them so much easier than a modern computer
Had no idea they had clutches, or how they worked. Awesomeness. Thank you
My company uses magnetic powder brakes on the unwinds for our coating lines. Eliminates all chatter, glazing, slip, etc with conventional unwind brakes. But we run our webs slow enough that the units weren't always super happy and struggled to control. Solution? Constantly tumble the powder inside with a separate drive motor.
Fantastic video Marc. I remember seeing these drives in a visit to a computer room in the 1960s and it felt like being transported into the future. The state of the art back then was very visual and the incredible engineering was mind blowing. Today it's SMD and a whole load of code - equally mind blowing but much more difficult to appreciate.
I never thought much about the mechanism behind those drives. It really is quite extraordinary that such a precise device was built out of such non-precise parts.
I’m just impressed that they got that complicated thing apart and then back together with zero swearing 😃
I'm sure the footage is "Politely Curated."
Patreon subscribers should get a compilation of profanity and thumb hammering.
Genius, Marc. You are a wizard.
What can I say having worked on Monotype lead type casting machines, rebuilt and repaired bikes, serviced household appliances and gastronomy equipment, let alone all these hi-fi gear repairs? I love this one :).
Filling that clutch is kinda like BGA reballing, isn't it? :)
Damn awesome repair job. Makes me wish I had the opportunity to work on tech like that.
Finally more IBM 1401 CONTENT!! WOOHOO
I'm watching this waiting for a big digital sign to scroll: THERE IS ANOTHER SYSTEM. I suspect at least a few people will catch that reference.
Well done! You are literately a poly technician.
I'll be coming to the museum from Arkansas in a few weeks. Can't wait to see this beauty!
Crazy that the repetitive spikes of the magnetic field engaging the clutches do not harm the impulses written to the tape hm....
What beautiful machines!! 😍💜 i wish The computers from now were used with tape drives!!
The good old days when an IT engineer could change the head on your Hemi quicker than the garage
21:41 It's not an USB cable with only 4 pins !😅
Slow frequency need more wires for the data bus ?! Right ? Very impressive video.
I have immense respect and admiration for your work on this channel.
Hello from France.
n the 1990's i had a job of removing various parts from 8 large IBM reel machines like the ones here,just newer models,prob 1980's,for Memorex,they only needed a few pcbs the rest was scrapped,i asked if i could have the DC motors and various parts,had the motor control boards as well,huge aluminium heatsinks,so well made,i still have the motors & Driver pcbs today and they are 19v multi pole brushed,electromagnet poles,weigh about 15kg each,ive made various things with some of them over the years ,they can take 1000 watts for several minutes without getting too warm.I also had 8 vacuum pumps which were really well made,unfortunately all of them were stolen along with aluminium sheets around 1994 from my lock up.I worked till 11-00 at night that day, as the company was getting the chassis collected in the morning, Happy days.
Thanks SKF for creating spherical ball bearings and creating/naming car maker Volvo, which as many of you know means "I roll".
Really fitting because of this video and its car innuendos 😀
Chapeau bas Marc, vraiment un travail de pro.
I don’t know why, but this was fantastic to watch. Thanks Marc.
Marc, what was the manufacturer of the original bearing ?
If SKF, the bearing should have a number stamped on the side of one of the races, a number which could help obtaining a new alike one.
If not, what was dimensions ?
I could ask the local SKF distributor here.
This is ‘beyond amazing’
Nice to know you can think beyond paper and pen. Hands on mechanical skills are still worth something.
Marc at minute 18:24 you mention that you are reconnecting the motor to drives the stop clutches. This is not correct, that motor brings the head guide assembly and stop capstans into operation.
The stop motor connects via a shaft with worm gears, and is located above and to the front, hard to see in the pictures. I thought that I had entered this comment before, by your instructions.
You had, but I had to re-upload the video due to a glitch! So you had to reupload your comment too... The good news is that you get another heart. ;-)
Was almost done watching V1, must've fixed the weird glitchy bits lol
Yes, glitchy no more hopefully.
you are brilliant marc!!!
As one of my long-ago professors said: "it's so slow, it looks incredibly fast." This when describing early computer electromagnetic devices, when compared to more modern electrooptics. He made that remark when explaining the operation of a big-screen display device that used a laser in conjunction with a pair of galvo-motor beam steering mirrors, but the principle is the same as one sees when watching the remarkable accelerations of the tape reels and vacuum columns. They are only "slow" when compared with the speed of the digital electronics to which they feed data.
Ah SKF, Svenska Kullagerfabriken (the swedish wheel-bearing factory, which is their name)
Fun fact: SKF also founded Volvo in 1915.
Amazing! Your videos always fascinate and entertain. Thanks Marc
Nice work :) Yes revisions and old diagrams can be a real pain.
awesome piece of computer history, thanks for that
these videos are so nice they are worth the great cost of walking around with that muzak playing in a loop in my head for several days every time
If you do it for long enough, eventually the looping music will be accompanied by explanations of whatever is wrong with what you're working on at the time.
@@FrankGevaerts 🤯
I'm watching Season 5, Episode 3 of the old television series _Mission: Impossible_ , “The Innocent” (1970), where several several tape drives of a mainframe computer are shown. However, these props are shown with the reels continuously rotating instead of the more-accurate stop-stop action.
great content! please more!
Car water pumps use those double row ball bearings, ie, the shaft is only supported on one side so needs the double, and are about that size too.
Well, Marc, after watching the first upload, I'll just say it was awesome. You know what I mean 😉
Much more into the content than any glitches... beneath my notice.
Just brilliant thanks.
Always interesting!
great video for future reference bearing should only be pressed on their outer race or else you could damage the bearing and or reduce its life
HP finished killing off those column tape drives with their "pizza oven", then automated libraries and cartridge tapes ended the reel tapes.
I’ve seen this machine live now. Nothing can prepare you for the noise and the machine oil smell and the heat coming off the machine.
I have to say you guys are my heros.
Wow, just wow!
I feel special with all this content!
Beautiful mechanics :)
Strange that solution with the powder. I thought it works like those automotive AC compressor clutches. That does not need rings and brushes either... the coil is stationary there. It wasn't invented that time?
They were, but they tend to exhibit significant levels of snatch when engaging - not a problem for an A/C compressor, but possibly harmful when driving reels of rather fragile tape. Hence the magnetic powder clutch - they have extremely smooth torque takeup, which is why they are still used in material handling systems.
Holy cow that data connector!
Next content from Marc:
REPAIRING OLD APPOLLO ENGINE Part 1 out of 20
Really cool content, keep it up :)
It's truly remarkable what the mechanical engineers of yesteryear were able to accomplish without the high power semiconductors that are freely available today....Geniuses indeed!
I love the IBM 1401 - I use it as my zoom background at work and try to convince the youngsters that it’s actually the computer that runs our database server. 😜
I know rack mounted/tray loading tape drives are more convenient, but the vertical vacuum column drives look far more impressive.
That is fascinating about the magnetic powder being an integral part of how it works. Were you able to repair the broken tape?
Completely unrelated to the video but; What make and model is the laptop at 10:00 next to the magnetic clutch dust at the top left of the frame? I like its aesthetics haha.
Awesome stuff.
Thanks so much for sharing. 😉👌🏻
Those connecting cables, near the end, are just nuts. Is all the comms parallel?
Is there any chance you can just remove the dust shields from the original bearing, wash out all the magnetic powder, and re-lube it and put the seals back in? I'm sure the 2 bearings back to back will be just fine, but I doubt the original is actually "shot" and instead just plugged up with the powder.
I maintained these back in the day
Amazing video
Maybe next time you are maintaining this area/piece, you can use the innards of the new SKF bearings, to rebuild the old seized bearing? This way youll have the original bearing???
Just an idea. You seem to have all the tools needed to rebuild the bearing properly.
Also you can use contact cleaner or engine degreaser and soak the seized bearing to hopefully remove/extract all the foreign material/magnetic powder that got inside the bearing. Although i would probably rebuild it.