You aren't building a museum, you're building a MONUMENT to digital computing's actual history. I'm learning a lot, I've been watching since the Centurion projects started, and I'm in for the adventure! You're also doing a phenomenal job of telling the stories. Please archive every video you've made. Your work and results at this time in history need to be preserved.
Thank you so much for the kind words! All the videos are available here, on Odysee and I have local copies backed up as well, so they won't be disappearing anytime soon hopefully. One of the most amazing things about these machines are the people involved in their creation, and those stories are what I love to learn about and tell. Of course, a story is so much more fun when the topic of the story actually functions, so we gotta get all these machines up and going!
Brings back memories... I wrote my first computer program (1967) for an 40-bit machine with 5 kilobytes of core memory and a drum holding 60 kilobytes including swap area for the core, a small OS, an Algol compiler and a few extras. Files were on paper tape and I still have a few tapes in my belongings. The computer was called "Gier" and the danish company RC built several hundreds of them as far as I can remember.
I need to rectify : According to Wikipedia, they built 56 of these machines over a span of about ten years. Some sources say 40 machines. 15 kHz clock speed. The Algol compiler was quite sophisticated supporting recursive procedures, but compiling of a program of 500 lines could take a minute.
Whoa, programming in the 1960s must have been an experience for sure! Today it's almost second nature, and even when we come across something esoteric, we have the fundamentals that we've all learned to fall back on, but the 60s was pretty much the inception of accessible computing, it must have been wild sitting down with such a wild machine to write programs for it!
@@UsagiElectric One thing that always has amazed me is that they managed to write an Algol compiler that could run in 5 kilobytes of core memory. This was a quite sophisticated compiler featuring everything mentioned in the official report on Algol60, recursive procedures and all. It made nine passes and stored the intermediate version of the program on the drum. The biggest program I wrote for this machine was a simulator of the Intel 8008 microprocessor, 13 pages, all in all. It ran without a glitch. This is one of the "source files" that I still have on paper tape.
I forgot to say that this computer was 100% transistor based, germanium transistors like OC141. I managed to salvage a few circuit boards when one of these machines broke down and was essentially scrapped.
Drum memory! Now I get to retell a tale I read a long while ago (although I'll have to shorten it considerably for the comment and also because I don't remember some of the mundane details). This was in the days of the Real Programmer. This particular Real Programmer worked with several other people, one of whom told this tale. To set the scene, the Real Programmer was so good that he knew exactly how fast the drum would spin and also exactly how long it would take for each instruction to execute, and so in order to speed up his programs, he'd tell the assembler exactly where on the drum to put each instruction. He timed things perfectly so that just as one instruction finished and the CPU began looking for the next instruction, it was just about to pass underneath the read/write head. When there was a printer which required a short pause between each character, instead of programming a delay loop, the Real Programmer repositioned the instructions in the print routine so that when each instruction had finished, the next instruction had just passed under the read/write head and so the CPU had to wait for the instruction to come around again. Due to some events which I forget, the Real Programmer knew he was going to be fired, but he left a legacy. In a program which printed a report, he programmed it to occasionally print a random asterisk, thus spoiling the formatted output. It didn't ruin the results since they could still be read. It was just annoying. The person reporting the story was told to try to find where the code was which produced the asterisk and remove it, so he started trying to trace the program flow. To his astonishment, he reached a point where the instructions reached the end of the drum but it very obviously wasn't the end of the program. After puzzling over it for quite a while, he eventually figured out what was going on. What does a binary address counter do when it reaches the highest possible number and then you add one to it? He went to the start of the drum and sure enough, there was the next instruction in the program. So impressed was he with the Real Programmer's skill and cunning that he stopped looking and just told management that he couldn't find the code which was inserting the asterisk. After all, how were they going to prove otherwise?
@@strayling1 Hmm... it does sound suspiciously similar to the tale which I read, although many of the details are quite different. It seems likely that one story was derived from the other.
@@melkiorwiseman5234 That's probably a safe bet. I first came across it in the days when most stories spread by 10th generation, barely legible photocopies.
I promise I didn't read the story, but I have a strong feeling it's about the programmer who used the rotating drum to find the memory location of "least pessimism" instead of "most optimum."
@@Wizardofgosz That sounds exactly like the story I outlined, although I didn't think it worth the space to talk about that little detail and so missed it in my version of the story.
The Librascope LGP-30 of the late 1950s was also a bit-serial machine. It was designed for scientific applications, and had a 31-bit data length and 15-bit instruction length. It had only 16 instructions but it implemented single-instruction multiply. Its bit-serial nature allowed it to be an exceptionally compact machine for the time. It had 4096 words of drum memory and three general purpose registers which were also on the drum. There is one fully working machine at the University of Stuttgart.
This machine and the LGP-30 share a phenomenal amount in common with respect to overall architecture! I wish I had an LGP-30 here to play with, it'd be so much fun to get the two of them up and running side by side.
@@UsagiElectric One of the reasons for the LGP-30 being so compact is the use of silicon diodes for most of the logic. The tubes- 130 of them if I remember right- are mainly for power, I/O and the drum memory. Oddly enough, the transistor version- the LGP-21- was actually slower.
My god, drum memory! I never thought i would see a computer of that type again, let alone possibly operating! Great score, Usagi. The link between the Litton and the Centurion seriously makes me smile. Your enthusiasm is so contagious I am always grinning ear to ear watching these videos. Cannot WAIT for you to get the beast running and the tape reader going!
If I remember correctly the computer in the first generation of the A6 intruder (aircraft) had drum memory. Wonder if that was a litton product? They did make all sorts of aeronautically-oriented wonders.
@@Sylveste22 Only if Litton was ever a Department of Defense contractor. At that time the military didn't use off-the-shelf components for hardly anything.
@@exidy-yt They made inertial navigation platforms that were used in US combat aircraft, so they were definitely on the approved list. I didn’t know they made computers until I saw this video actually.
@@exidy-yt Well that’s pretty crazy, I looked it up and apparently the first generation a6 intruder computer was the AN/ASQ-61 and yes it had drum memory and yes was made by Litton. I guess spinning things are their area of expertise. This entire chain of thought brought to you courtesy of Willem Dafoe in Flight of the Intruder moaning about having to kick the thing to get the rotor going.
Paper tape is very hacker friendly in that it's very easy to build a reader these days. Jürgen Müller has designed a reader based on an Arduino Pro Micro that looks very easy to replicate and use. A punch would probably also not be too difficult, but definitely takes the amount of work up a notch.
It is interesting how weird stuff persisted into the 1970s. This Litton machine is pretty epic with drum storage rotated by a Bodine motor which would have been at home in the 1930s. I worked for a bank in the proof and transit operation supporting 32 IBM 1260 electronic inscriber machines. The 1260s were full of boards of SLT, used plugboards for setting up the routing program and stored their totals on a acoustic delay line memory. The delay line was a bit like an unwound clock mainspring in about a 12 inch x 12 inch box. Information was stored by an electromagnet torquing one end of the spring, making pulses travel to a pickup at the other end. The picked up pulses were amplified and put back into the spring. As you might guess, environmental vibration such as a cart full of checks to be inscribed bumping into a machine caused angry operators who had to reprocess lost work.
I remember those delay lines hanging around in some surplus stores. Zinc or aluminum box, and indeed some kind of ribbon mounten on soft rubber inside.
@@wtmayhew I amso remember a friend coming home with a type of disc storage taking from a dumpster. It has a belt drive, with a flat belt, and fixed heads mounted at a tiny distance from the disc. I theak each head covered 8 tracks or so, and there may have been 10 of them. The disc has the brown magnetic coating on one side only.
OMG the IC's dated my week of birth! Never have I expected to see the production ICs this old! In the autumn of my life, I dare to say it was a very good time to live in.
I LOVE those 60s/70s (?) carbon resistors... they remind me of my childhood salvaged components collection. A lot of it looks far more modern than I was expecting: ICs and flexible PCB cables, etc.
Those are mostly what I know, and use. My dad worked at Bell Labs and brought a bunch of stuff home in the 60's - including lots of carbon resistors. 5%.
It's definitely interesting how close it was to modern construction here. The flexible PCBs instead of ribbon cables is a bit out of the ordinary, but still pretty close. The sparseness of the PCBs is indicative of the construction techniques of the time (laying out the traces by hand with tape), but the techniques are all there. It really feels like the inception of the modern machine!
Wow, how cool! It sure does look like a washing machine, too, lol. Litton made my family's first microwave oven, and it was a sturdy, long-lasting 1970's beast of a machine. Every time one of our newer microwave ovens has failed, we miss the Litton. We only replaced it because the door latch broke, and replacement parts were no longer available. Alas! Hopefully, your Litton minicomputer will be just as sturdy and reliable! 🙂 Those de-laminating cables look scary for sure, but I would hope that you'd be able to repair any failed connections with sturdier ribbon cables. 🤞
Just made me remember that my first microwave was also a Litton. Back in the 70's. It was the small "Little Litton" version, heavy, built like a tank, still worked after decades. Eventually went to the Goodwill only because I wanted a larger one with the rotating platter.
Ever since I watched the movie "Flight of the intruder" I was fascinated by a scene where the bombardier had to "kick the computer to get it going again.", which was a computer that he used for bombing targets with the A6-Intruder. Low and behold, Litton made that computer that went into the real-life A6-Intruder bomber aircraft. In fact, im willing to bet you have either a close relative of that computer or a copy of the guts of it, timeframe matches. So the reference in the movie is a real life reference to kicking the drum on this computer if it got stuck. Incredible.
That 0.100" flex cable was popular in several machines back then. I have a few HP frequency generators that use it. If you come up with a way to replicate or replace it, let me know. The only thing I could think of was an adapter PCB to convert them to "normal" IDC headers and ribbon cable.. even that tech is vintage now. On the plus side, you get that drum memory spinning up and that cabinet is never gonna tip over.. as the memory turns into a massive gyro :) Cheers,
Modern PCB rapid prototyping companies boast they can make flexible PCBs and flat-flex connectors nowadays, so I assume they could be contracted to produce new ones to the sizes used in that HP gear! Just gotta pull one out and transfer its dimensions into a CAD program.
11:12 So cool to see a Bodine Electric motor - my great grandma worked there! When I was little, she used to tell stories about winding motors and even gave me some broken motors to take apart, to see how they worked.
Balancing this upload with VCF prep is amazing. Can't believe that I managed to speak to you at the fest and then come back home to have another full-length upload waiting! If only there were a way to get drum memory nowadays without harvesting it from an existing computer - or, at least, harvest it from one beyond saving. That'd be so perfect for the Vacuum tube computer
I had a complete working Litton ABS system as shown in the video: CPU, paper tape punch/reader and printing terminal. It used large circular military connectors for interconnection. The terminal had an interesting keyboard which had an interlock mechanism that only allowed you to press one key at a time (all other keys were mechanically locked out and you couldn't press them). It was a rather slow and noisy affair. At the time I was a poor typist, and I could still out-pace the keyboard mechanism. The whole system ran an operating system called OPUS (if I recall correctly). The drum was remarkably quite, although the paper tape punch was very noisy. Someone added a power switch to turn off the punch motor, but it would create a power glitch and crash the CPU every time. The paper tape reader was all mechanical: it had little pins which poked through the holes in the tape. Using a cam, all the pins were withdrawn and the tape would advance one sprocket. The pins were then raised to detect holes in the tape. The PCB's for the tape reader/punch were very large (maybe 20" x 12"), with only transistors, diodes, resistors and capacitors - very sparsely spread out. Sadly, I threw out the Litton, along with an NCR 299 and a Data General MV8000 (yes, the model which was written about in "The Soul of a New Machine"), and yes, I'm regretting it almost every day....
I repair avionics and a few devices I repair use those laminated ribbon strips to connect boards. I fix them by just brushing superglue all over the de-laminated plastic part before pressing them back together and it seems to work well to hold it all together
Very cool - I have never laid eyes on a magnetic drum memory. I was told by fellow employees at GE in 1983 that previous versions of our E2C Hawkeye radar used MDM, but that is the only reference to it I have ever come across. I visited a friend in Colorado around 1988, and he told of ex-Litton computer company employees I believe. I think they had a location there.
Wow! What a find! I had heard of drum memory at my training at Nixdorf Computers. Apparently Nixdorf also had invested in drum memory. However, the manufacturing tolerances were very challenging and IBM had developed the disk drives. Interestingly enough the tracks on the platters are called cylinders to this day and that was according to our trainer based on the drum drives which had several concentric drums, or cylinders. Unfortunately, by the mid 1980s these drum drives were phased out in favour of the CDC drives. So I had only heard of them, but never seen one. About the ribbon cables delaminating. They look like simple straight through connectors. You should be able to design PCBs like that and get them made by JLCPCB, like the utube channel who shows all the PCB motors, I forgot the name. That should allow you to get a stack of them and replace what needs it.
To my eyes they are not really ribbon cables but flexible PCB's anyway and PCBway also do them. I would look to replace the lot. From the video it looks a very simple design.
For the flex ribbon cables used on the bus that are falling apart I would use kapton tape and ever so carefully wrap them up. Also replace the electrolytic caps before powering up and check the bearings on the drum motor.
I suspect the best bet is to replace them with custom flexible PCB's, which are basically the same technology and the likes of PCBway will make them for you in prototype quantities which is what you need here.
I kind of remember seeing a couple of these back in the early days of my Vo-tech training for machine shop. There was also a 'data processing lab' which is what they called the computer room at that time, and when I first started at the Vo-tech in 1972, they actually had a couple of these Litton machines. They later switch to DEC and PDP as they had better support. The only other Litton products I was familiar with were microwave ovens, which were starting to get popular in American homes at the time. It will be interesting to see if you can get this relic up and running 50 years later!😉 *EDIT:* Just a side note: In our machine shop, we had one of the first digitally controlled vertical milling machines ever released. It also used punch tape to operate, and we would spend hours on an old punch tape TTY creating the programming to control the x,y, and z parameters to mill various projects. It was before automated tool change cassettes were popular, so we had to program pauses to allow for tool changes and then hit a button to allow the tape to continue.
All these collections of obsolete old things are what we need, lest we forget how we got to where we are. Like the Sony Triniton collector guy making a Trinitron museum. Love living in a society that allows these things to take place. Love it!!!!
Drum memory! It's always fascinating too see how the old electro-mech stuff is implemented and the hoops the designers had to go through to sync memory access with execution. Mercury delay lines would be another one to see but obviously having tubes of mercury floating around isn't something you want to deal with these days.
In the first 2.5 mins of the intro, I was convinced that there was a bed in the garage background. I was thinking “oho, Mrs Usagi’s patience has finally broken with the arrival of another mini computer and Mr Usagi is now sleeping in the garage!” Luckily, we pan out and see it’s just a cover on the car! Phew! 😅
So, just when I thought you had mastered the craft...you go and UP THE GAME! Yet ANOTHER Forgotten Machine rescued, restored and extremely well-documented by Usagi! I salute you, and I am truly humbled by your accomplishments!
Thank you! Though, I'm so far from mastering the craft, it's kind of hilarious. I'm just hanging on for dear life as the world of vintage machines sweeps me up and takes me on a wild ride!
@@UsagiElectric Well, we're on that journey together, my friend...you just make WAY better videos than I do...and are FAR better organized! So, please receive that in the best nature that it is intended...
Man I can't wait to see some running drum memory, only ever seen pictures. At that speed, probably hear it too! Looking forward to seeing you dig into it and show us more.
If at the very least the drum memory works, you have a storage option for your own vacuum tube computer. You might even use the entire Litten machine as a serial storage device for it.
Interestingly, I've been thinking about going the other direction as far as connecting things together. The Litton loads programs in via paper tape into the drum, and then executes directly from the drum. But, instead of having a paper tape reader feed directly into the Litton, I was thinking to put the Centurion in the middle. The Centurion could then grab a binary file and transfer it over to the Litton to load into the drum. This means that our program storage options are now immense, and we can write new programs on the Centurion and have them executed on the Litton!
@@UsagiElectric nah, keep the Litton original, let it load from paper tape. Or something else. Outside of experiments keep the centurion and Litton separate things. Otherwise you would always have the lug both around. The tube computer is a new development, you can connect that to whatever you want🙂
That's an amazing story!! A piece of computer history I would never have learned about without your channel! Looking forward to seeing this machine brought back to life.🙂
When you opened up the card cage it looked like on of those frames that contained pop posters in music shops. As a kid you always fretted that they would overturn on top of you as you pulled each poster over to view the next. It is always fascinating to see how few components were used in those old machines, on boards the size of a pool table😊
@tradde11 yes I remember those too. Later on they seemed to be mainly horizontal and you had to look down and flip them to you or away from you. Maybe the posters were larger than the vertical ones, I don't remember seeing them in the same place to compare. They always seemed a bit unstable when most posters were on one side.
Wow, when I saw the thumbnail I though: Oh, was he getting bored with the Centurion and where did he come up with this one? 😂 As it turns out, they a computer relatives. Didn't see that one coming! You keep to amaze! Thanks for all the work you do!
I was tangentially aware of the link, but when I saw the machine in person and started reading the manual, it made so much more sense, and I knew I had to get hands on with this one!
I remember reading some old usenet anecdote about a guy who had to work with/on a machine with drum storage/memory which had an infinite loop set up in it's programming, and which confused the hell out of the young tech. After much research they figured out that the program which used the infinite loop managed to make it not so infinite because of the way the read head of the drum mechanism had to physically move, and even though there was no code way it could break the loop, there was something that made it seek long enough to land somewhere "wrong" that got it out of the loop and jump it somewhere is in the program.
Great tie in between the Centurion and the vacuum tube computer. I remember someone suggesting mag. drum storage and you saying something like "where the hell am I going to get a mag. drum from?"... well, hello there!
@@edison700 If it's expecting the holes to be brighter than the rest of the paper, you may need to print everything black except for the holes. Doable but more expensive.
I was at Georgia Tech as the institute itself and the school of computer science each changed computers. When I first enrolled there, the main computer for the institute as a whole was a Univac 1108. And it had no disk at all, only drum storage and tapes.
Back in the late sixties I was a Physics student, and had access to a "Stantec Zebra" computer which also used a rotating drum as its RAM. It used a drum which stored 8k of 33 bit words (I just looked that up) and was the only machine I have come across which kept a teleprinter - not teletype - waiting for the next digit when printing a decimal number. It was huge compared to your Litton.
I had one of these computers in running condition in +- 1980. really interesting to program. I ended up dismantling it for parts, and used the drum memory unit as a tactile/auditory effect in a 2 seat spaceship simulator i built- fully enclosed with a sliding top. i removed the rubber shock mounts on the drum unit and bolted it directly to the frame of the simulator. truly awesome sound and physical effect- sounded and felt just like a jet engine. ahhhh, nostalgia. I later built another simulator that moved in 3 axes using the panels from a discarded IBM 1401.
wow, this is so cool, seeing a company be created to support hardware from a company that stopped making and giving support for their hardware is so wild, and then ending up making their own system with backwards compatibility for that system? just wow, basically this machine existing not only allowed centurion to exist but also made their systems use a very weird filesystem all to add backwards compatibility, this is so fascinating omg cant wait to see more of this, also, spinning ram? it looks like a very fast hard drive does it also keep data after its been powered down? does it have data rn? wild stuff omg. Also backing up those paper rolls into the hawk drive sounds really cool like digitizing it, its so cool to see how tech advanced so fast in the 70s and 80s
I love your TI-99/4a setup on the shelf. I still have my TI that I bought in 1983, which has several of the expansion components, including the PEB. Saving old computer systems is so cool.
Here we go again, Can't wait to see how this unfolds. Huge honor getting to meet you and talk a while at VCF East today, and getting sworn at in Japanese and Yiddish. I wonder if they ever got a sign of life out of that drowned classic Mac. Thanks David, keep the hits coming!
Wow, I thought my brother and I had some old computers, but that thing is older than anything we've got in our collection by like a decade and a half! That's impressive, even moreso that you got it working.
Last time I saw a drum memory was in high school about 1975 . I was in an electronics course and we got old computers donated to us from oil company seismic outfits, the drums were used to record seismic vibrations from dynamite charges set off to find oil formations underground. These things were very heavy with fixed recording heads and pretty slow rotation as I remember . Looks like you found a real winner it Will be amazing if you get it to work again 👍
This seems a lot like the LGP-30 from about 1959 which was a deep-freezer sized tube and semiconductor diode "general purpose computer" that had a 4K word drum memory. It operated bit-serial, with 32 bits per word (though one bit was apparently a "spacer" bit and wasn't writable). It's teletype was a Friden Flexowriter and also had a separate paper tape punch and reader. The LGP-30 had a hardware multiply, though it took like 120 clock cycles to perform. The accumulator was stored as a track on the drum as well, and it's front panel had a small oscilloscope which displayed a square wave trace of the contents of the accumulator and IIRC the program counter. One thing I recall about the LGP-30's drum memory is that after spinning down the drum you needed to wait 1/2 hour before spinning it up again. This was because the drum would heat up and expand, and if you restarted it too soon the heads would hit the drum and damage it. There was a big homemade sign taped to the one I worked with that admonished operators to be sure to wait sufficient time for the drum to cool off before restarting-- apparently, head crashes were common and it could require re-positioning and/or replacing heads so that the new position was over an otherwise unused portion of the drum-- not a job you wanted to have to have done...
That drum really precesses if you try to move it when it is running. I demolished an incomplete one of these back in the early 1980s. I still have the drum itself, and the motor, which has been repurposed to drive my hobby milling machine.
The cool thing is the rotating drum memory. The only other place that I've heard it used was the A6 Intruder's on-board computer during the Vietnam war.
Well this looks like it's going to be a very interesting project! Also, unrelated, but I loved the choice of music for the cleaning montage. It reminds me of the french duo Air.
If it's 1971 you will probably find that what you think are transistors are early ic's with negative logic . I have early calculators with those and they take a lot of thinking to understand what is going on . I am excited to be on the journey with you In your CZcams. Cool and very exciting!!!
Your channel never ceases to keep me amazed and feeling nostalgic. I remember in my shop on the aircraft carrier. The bench that calibrated the inertial navigation unit for the A-7 was manufactured by Singer (yes the sewing machine people). At some point Litton had purchased or taken over that Singer division. The actual Litton bench that maintained the rest of the navigation systems for the rest of the now retired aircraft, is where I saw my first multi platter Hawk drive. It also had a single fixed and removable platter Hawk drive like yours. When you mentioned that the Litton looks like a washing machine. That made me laugh, because it wouldn’t surprise me if there was a Litton link to the washing machine industry somewhere. When we were at our shore duty station. Our Litton tech rep would get calls forwarded to him from unhappy Litton Microwave oven owners quite often. We had a lot of fun with that.
The question is, did they do any money laundering with it? Beautiful machine. The flex PCB ribbons really surprised me! You may try gluing the layers with some flexible polymer glue (no cyanoacrylate as it'll go brittle as the joint is bent). Fortunately you can design new flex (or rigid-flex) PCBs from JLCPCB or PCBWay, and while not original, they will be better quality :) "Miniature nuclear reactor" would be a good name for it, and the Centurion backstory is very interesting too! So bloody awesome that you got it :).
I think a short-term solution on the delaminating ribbon cables would be kapton tape. To prevent any shorts from any dirt coming in. I think I saw something in one of the closeups of the ribbons that looked like steel wool next to one of the delaminated ribbons, which would cause a short.
Yes the peeling flex is hard to solve, as you need a flexible adhesive, which probably will mean using a thin electronics grade silicone to attach the polyamide back to the tracks, after removing the plastic carefully to remove old adhesive, and then apply a thin coat to the polyamide, and hold in place, probably with a soft silicone rubber tubing, covered with some thin PTFE sheet, till it has cured.
@@tobiasfunke8990 I think that "steel wool" is actually foam sound deader that he was removing from the inside of the cabinet. I'm sure it was all rotting and falling off. I do hope he replaces it before starting the drum as I'm sure that changes the sound.
@@KeritechElectronics The *easiest* readily available solution is probably a can of spray clear flex-seal or even plasti-kote, and some creative (and thorough) masking. You wouldn't be affixing the layers, you'd simply be replacing one side or the other.
Although I know almost nothing about these types of machines it is so much fun to hear you talk so enthusiastically about them. I have watched every Centurion epsiode so far, so having yet another piece of the background was really interesting. Looking forward to follow your journey on this Litton minicomputer as you restore it to the same perfect condition as the Centurion! That should add another couple of years of superb content. ;-)
I think the 40bit word length came from the Zuse machines. The Z22/Z23 used 40bit memory with and eight bit tape reader/writer, drum memory in addition to core memory. The Litton looks like a "economy" version of a Z22/Z23 to me. Would be interesting to see the connection there.
That's an interesting connection! I read some books about the Zuse computer history in my youth (I'm Austrian myself) , but it was written without technical details. So that I could read it as well. Or maybe it was in there, but I would have had absolutely no idea, what the words "drum memory" even mean. Let alone the length and specialty of a 40 bit word. Now - with the Centurion restoration and usage (usagi) videos shwon here, I learn about cool technology connections and new (old) technical terms. 🙂
Computer history is so much more interesting than I thought. We're in "the future" where we can put 3-7 billion or more transistors in one CPU, but only thousands back then. They had conventions back in the 60's about the future of computing and AI, which kinda matches up to what we're living today with AI generated art, machine learning, quantum computing, etc. It's kinda cool how far we've come, and the potential for what's to come.
I luv your videos and always look forward to seing more and learning more about such vintage technology, i also like your furry friends :) The Centurion videos have been really cool too..
I am also very curious about the result. They were great times when they came on the market. I am following you closely as I want to see this working, good luck, thumbs up!
WOW....A Litton 1241 ABS Controller Unit. The first program I ever wrote was on this machine. If I recall, the paper tape is encoded as 7 bit ASCII with an extra bit for parity. You should be able to clock the old tapes through an optical array to read them. I remember that the Litton programming language could pretty much be likened to a subset of Assembly language. I was about 15 yo when my mother bought a used one for her accounting practice in or about 1976. It cost around $18k, which was a huge amount of money back then. I hope you find a paper tape reader/punch and teletype. Thanks for the memory.
To everybody who had a C64 growing up: Consider that to even come close to what you had in your bedroom in the eighties, a business in the 60s/70s would have had to have had THREE of these Litton machines AND 4K of actual core memory. Back then, RAM was a limiting factor. It took a while for RAM to become plentiful and cheap.
I was working with a massive dual PDP 10 system that had it’s hard disks replaced by Drum Storage back in the late 70’s. it’s memory was magnetic core.
DEC produced a lower cost version of its 12 bit PDP-8 minicomputer with serial ALU, the PDP-8/S. Drum memory was famous in the IBM 650, which served as IBMs entry level computer for many years in the Fifties and Sixties. Drums were used later for fast random access storage. One Burroughs drum weighed literally a ton and tendended to break free and walk around due the forces generated by its fast spinning drum. The Army and Navy tried to use the Burroughs drum in mobile applications but the the rotational forces were too much for the ship or truck carrying the drum.. The drum in this machine is a relatively minaturized version of the breed. Drum memory machines had to account for drum rotation when placing instructions. The techniques for optimium drum computer performances are similar to those used with RISC machines. Dead space in the instruction stream can be filled in by unrelated secondary operations.
Whelp, I missed you on Sunday, I was beet from the walking around on Saturday, but you Guys/Girls were on stage, and all had to zoom out at the End, Hope the 60's RCA IC made it home with you, it a Pre Production A Version, Just Try and Find some Info.. I wanted to come back Sunday, MY the HVAC unit took a crap on me on Saturday, And Sunday was House Repaire weekend YAY..
To me, 40 bits signal "IAS machine heritage". Many of the direct descendants of "the IAS machine" were 40-bit machines, jsut as the IAS machine was. And they, in turn, gave rise to other machines that were 40-bit. The earliest commerical model I am aware of is the Ferranti Pegasus, but I am also aware of prior machines using it, although they're more "one-off machines". BESK, who begat SMIL, DASK, and SARA (all 40-bit machines on the general IAS architecture spectrum). SARA then begat CK37, and D2, but I do not know if either of the latter were 40-bit machines (CK37 is the on-board computer in the Saab Viggen military jet). That 40 bits happened to also be handy for business computing...
You get all the amazing things! That Litton amazing. It does look like a washing machine! Can hardly wait to hear what the drum memory unit sounds like! Thanks for the history lesson, as well.
You aren't building a museum, you're building a MONUMENT to digital computing's actual history. I'm learning a lot, I've been watching since the Centurion projects started, and I'm in for the adventure! You're also doing a phenomenal job of telling the stories. Please archive every video you've made. Your work and results at this time in history need to be preserved.
Thank you so much for the kind words!
All the videos are available here, on Odysee and I have local copies backed up as well, so they won't be disappearing anytime soon hopefully.
One of the most amazing things about these machines are the people involved in their creation, and those stories are what I love to learn about and tell. Of course, a story is so much more fun when the topic of the story actually functions, so we gotta get all these machines up and going!
Brings back memories...
I wrote my first computer program (1967) for an 40-bit machine with 5 kilobytes of core memory and a drum holding 60 kilobytes including swap area for the core, a small OS, an Algol compiler and a few extras. Files were on paper tape and I still have a few tapes in my belongings. The computer was called "Gier" and the danish company RC built several hundreds of them as far as I can remember.
I need to rectify : According to Wikipedia, they built 56 of these machines over a span of about ten years. Some sources say 40 machines. 15 kHz clock speed. The Algol compiler was quite sophisticated supporting recursive procedures, but compiling of a program of 500 lines could take a minute.
Whoa, programming in the 1960s must have been an experience for sure!
Today it's almost second nature, and even when we come across something esoteric, we have the fundamentals that we've all learned to fall back on, but the 60s was pretty much the inception of accessible computing, it must have been wild sitting down with such a wild machine to write programs for it!
@@UsagiElectric One thing that always has amazed me is that they managed to write an Algol compiler that could run in 5 kilobytes of core memory. This was a quite sophisticated compiler featuring everything mentioned in the official report on Algol60, recursive procedures and all. It made nine passes and stored the intermediate version of the program on the drum. The biggest program I wrote for this machine was a simulator of the Intel 8008 microprocessor, 13 pages, all in all. It ran without a glitch. This is one of the "source files" that I still have on paper tape.
I forgot to say that this computer was 100% transistor based, germanium transistors like OC141. I managed to salvage a few circuit boards when one of these machines broke down and was essentially scrapped.
Drum memory! Now I get to retell a tale I read a long while ago (although I'll have to shorten it considerably for the comment and also because I don't remember some of the mundane details).
This was in the days of the Real Programmer. This particular Real Programmer worked with several other people, one of whom told this tale.
To set the scene, the Real Programmer was so good that he knew exactly how fast the drum would spin and also exactly how long it would take for each instruction to execute, and so in order to speed up his programs, he'd tell the assembler exactly where on the drum to put each instruction. He timed things perfectly so that just as one instruction finished and the CPU began looking for the next instruction, it was just about to pass underneath the read/write head. When there was a printer which required a short pause between each character, instead of programming a delay loop, the Real Programmer repositioned the instructions in the print routine so that when each instruction had finished, the next instruction had just passed under the read/write head and so the CPU had to wait for the instruction to come around again.
Due to some events which I forget, the Real Programmer knew he was going to be fired, but he left a legacy. In a program which printed a report, he programmed it to occasionally print a random asterisk, thus spoiling the formatted output. It didn't ruin the results since they could still be read. It was just annoying.
The person reporting the story was told to try to find where the code was which produced the asterisk and remove it, so he started trying to trace the program flow. To his astonishment, he reached a point where the instructions reached the end of the drum but it very obviously wasn't the end of the program. After puzzling over it for quite a while, he eventually figured out what was going on. What does a binary address counter do when it reaches the highest possible number and then you add one to it?
He went to the start of the drum and sure enough, there was the next instruction in the program.
So impressed was he with the Real Programmer's skill and cunning that he stopped looking and just told management that he couldn't find the code which was inserting the asterisk. After all, how were they going to prove otherwise?
I love that tale. Do a search for "The Story of Mel" to find the original.
@@strayling1 Hmm... it does sound suspiciously similar to the tale which I read, although many of the details are quite different. It seems likely that one story was derived from the other.
@@melkiorwiseman5234 That's probably a safe bet. I first came across it in the days when most stories spread by 10th generation, barely legible photocopies.
I promise I didn't read the story, but I have a strong feeling it's about the programmer who used the rotating drum to find the memory location of "least pessimism" instead of "most optimum."
@@Wizardofgosz That sounds exactly like the story I outlined, although I didn't think it worth the space to talk about that little detail and so missed it in my version of the story.
The Librascope LGP-30 of the late 1950s was also a bit-serial machine. It was designed for scientific applications, and had a 31-bit data length and 15-bit instruction length. It had only 16 instructions but it implemented single-instruction multiply. Its bit-serial nature allowed it to be an exceptionally compact machine for the time. It had 4096 words of drum memory and three general purpose registers which were also on the drum. There is one fully working machine at the University of Stuttgart.
This machine and the LGP-30 share a phenomenal amount in common with respect to overall architecture! I wish I had an LGP-30 here to play with, it'd be so much fun to get the two of them up and running side by side.
@@UsagiElectric One of the reasons for the LGP-30 being so compact is the use of silicon diodes for most of the logic. The tubes- 130 of them if I remember right- are mainly for power, I/O and the drum memory. Oddly enough, the transistor version- the LGP-21- was actually slower.
The anode voltage was higher than Vcc. Thus quicker 😂
My god, drum memory! I never thought i would see a computer of that type again, let alone possibly operating! Great score, Usagi. The link between the Litton and the Centurion seriously makes me smile. Your enthusiasm is so contagious I am always grinning ear to ear watching these videos. Cannot WAIT for you to get the beast running and the tape reader going!
If I remember correctly the computer in the first generation of the A6 intruder (aircraft) had drum memory. Wonder if that was a litton product? They did make all sorts of aeronautically-oriented wonders.
@@Sylveste22 Only if Litton was ever a Department of Defense contractor. At that time the military didn't use off-the-shelf components for hardly anything.
@@exidy-yt They made inertial navigation platforms that were used in US combat aircraft, so they were definitely on the approved list. I didn’t know they made computers until I saw this video actually.
@@Sylveste22 Cool, then it's quite possible! Would take a bit of digging to find out but interesting thought.
@@exidy-yt Well that’s pretty crazy, I looked it up and apparently the first generation a6 intruder computer was the AN/ASQ-61 and yes it had drum memory and yes was made by Litton. I guess spinning things are their area of expertise. This entire chain of thought brought to you courtesy of Willem Dafoe in Flight of the Intruder moaning about having to kick the thing to get the rotor going.
Paper tape is very hacker friendly in that it's very easy to build a reader these days. Jürgen Müller has designed a reader based on an Arduino Pro Micro that looks very easy to replicate and use. A punch would probably also not be too difficult, but definitely takes the amount of work up a notch.
It is interesting how weird stuff persisted into the 1970s. This Litton machine is pretty epic with drum storage rotated by a Bodine motor which would have been at home in the 1930s. I worked for a bank in the proof and transit operation supporting 32 IBM 1260 electronic inscriber machines. The 1260s were full of boards of SLT, used plugboards for setting up the routing program and stored their totals on a acoustic delay line memory. The delay line was a bit like an unwound clock mainspring in about a 12 inch x 12 inch box. Information was stored by an electromagnet torquing one end of the spring, making pulses travel to a pickup at the other end. The picked up pulses were amplified and put back into the spring. As you might guess, environmental vibration such as a cart full of checks to be inscribed bumping into a machine caused angry operators who had to reprocess lost work.
I remember those delay lines hanging around in some surplus stores. Zinc or aluminum box, and indeed some kind of ribbon mounten on soft rubber inside.
@@pizzablender Thanks for sharing. Not too many people have seen that type of storage,
@@wtmayhew I amso remember a friend coming home with a type of disc storage taking from a dumpster. It has a belt drive, with a flat belt, and fixed heads mounted at a tiny distance from the disc. I theak each head covered 8 tracks or so, and there may have been 10 of them.
The disc has the brown magnetic coating on one side only.
OMG the IC's dated my week of birth! Never have I expected to see the production ICs this old! In the autumn of my life, I dare to say it was a very good time to live in.
I LOVE those 60s/70s (?) carbon resistors... they remind me of my childhood salvaged components collection. A lot of it looks far more modern than I was expecting: ICs and flexible PCB cables, etc.
Also, are those old-fashioned germanium diodes next to those carbon resistors?
Those are mostly what I know, and use. My dad worked at Bell Labs and brought a bunch of stuff home in the 60's - including lots of carbon resistors. 5%.
It's definitely interesting how close it was to modern construction here.
The flexible PCBs instead of ribbon cables is a bit out of the ordinary, but still pretty close. The sparseness of the PCBs is indicative of the construction techniques of the time (laying out the traces by hand with tape), but the techniques are all there. It really feels like the inception of the modern machine!
Fascinating. I know virtually nothing about this era of computing but it's slowly coming to light due to your videos. Thanks!
Exactly this indeed, growing older I'm more and more in awe at what engineers were doing in the 60's and 70's.
The 60s and 70s were such a wild time in the era of computing, it's so much fun to look back at the insane things the engineers were capable of!
Wow, how cool! It sure does look like a washing machine, too, lol. Litton made my family's first microwave oven, and it was a sturdy, long-lasting 1970's beast of a machine. Every time one of our newer microwave ovens has failed, we miss the Litton. We only replaced it because the door latch broke, and replacement parts were no longer available. Alas! Hopefully, your Litton minicomputer will be just as sturdy and reliable! 🙂 Those de-laminating cables look scary for sure, but I would hope that you'd be able to repair any failed connections with sturdier ribbon cables. 🤞
Just made me remember that my first microwave was also a Litton. Back in the 70's. It was the small "Little Litton" version, heavy, built like a tank, still worked after decades. Eventually went to the Goodwill only because I wanted a larger one with the rotating platter.
I have a Litton microwave from 1985
Still works
Perfectly, paid $350 CDN for it 👍
Ever since I watched the movie "Flight of the intruder" I was fascinated by a scene where the bombardier had to "kick the computer to get it going again.", which was a computer that he used for bombing targets with the A6-Intruder.
Low and behold, Litton made that computer that went into the real-life A6-Intruder bomber aircraft. In fact, im willing to bet you have either a close relative of that computer or a copy of the guts of it, timeframe matches. So the reference in the movie is a real life reference to kicking the drum on this computer if it got stuck. Incredible.
That 0.100" flex cable was popular in several machines back then. I have a few HP frequency generators that use it.
If you come up with a way to replicate or replace it, let me know. The only thing I could think of was an adapter PCB to convert them to "normal" IDC headers and ribbon cable.. even that tech is vintage now.
On the plus side, you get that drum memory spinning up and that cabinet is never gonna tip over.. as the memory turns into a massive gyro :)
Cheers,
Modern PCB rapid prototyping companies boast they can make flexible PCBs and flat-flex connectors nowadays, so I assume they could be contracted to produce new ones to the sizes used in that HP gear! Just gotta pull one out and transfer its dimensions into a CAD program.
Beware: before starting the drum, ensure that the machine is very well secured to the floor, and check the lubrication.
The front panel is beautiful. Looks like it was designed by Braun. And the Microgramma font is *chef's kiss*
11:12 So cool to see a Bodine Electric motor - my great grandma worked there! When I was little, she used to tell stories about winding motors and even gave me some broken motors to take apart, to see how they worked.
It's been a while since I ran any BSD Unix variants, but I remember the swap device was named /dev/drum.
Say what you will about drum memory, it is much faster than any kind of platter memory. Its perfect for a swap
By the looks of those boards, that Litton may be a transistorized version of a design that used to be a tube computer.
Balancing this upload with VCF prep is amazing. Can't believe that I managed to speak to you at the fest and then come back home to have another full-length upload waiting!
If only there were a way to get drum memory nowadays without harvesting it from an existing computer - or, at least, harvest it from one beyond saving. That'd be so perfect for the Vacuum tube computer
I think PCB Way can make the ribbon interconnects if you need to replace them. They are essentially just flexible PCBs with straight traces.
I had a complete working Litton ABS system as shown in the video: CPU, paper tape punch/reader and printing terminal. It used large circular military connectors for interconnection. The terminal had an interesting keyboard which had an interlock mechanism that only allowed you to press one key at a time (all other keys were mechanically locked out and you couldn't press them). It was a rather slow and noisy affair. At the time I was a poor typist, and I could still out-pace the keyboard mechanism. The whole system ran an operating system called OPUS (if I recall correctly). The drum was remarkably quite, although the paper tape punch was very noisy. Someone added a power switch to turn off the punch motor, but it would create a power glitch and crash the CPU every time. The paper tape reader was all mechanical: it had little pins which poked through the holes in the tape. Using a cam, all the pins were withdrawn and the tape would advance one sprocket. The pins were then raised to detect holes in the tape. The PCB's for the tape reader/punch were very large (maybe 20" x 12"), with only transistors, diodes, resistors and capacitors - very sparsely spread out. Sadly, I threw out the Litton, along with an NCR 299 and a Data General MV8000 (yes, the model which was written about in "The Soul of a New Machine"), and yes, I'm regretting it almost every day....
I repair avionics and a few devices I repair use those laminated ribbon strips to connect boards. I fix them by just brushing superglue all over the de-laminated plastic part before pressing them back together and it seems to work well to hold it all together
Very cool - I have never laid eyes on a magnetic drum memory. I was told by fellow employees at GE in 1983 that previous versions of our E2C Hawkeye radar used MDM, but that is the only reference to it I have ever come across. I visited a friend in Colorado around 1988, and he told of ex-Litton computer company employees I believe. I think they had a location there.
Wow! What a find!
I had heard of drum memory at my training at Nixdorf Computers. Apparently Nixdorf also had invested in drum memory. However, the manufacturing tolerances were very challenging and IBM had developed the disk drives. Interestingly enough the tracks on the platters are called cylinders to this day and that was according to our trainer based on the drum drives which had several concentric drums, or cylinders.
Unfortunately, by the mid 1980s these drum drives were phased out in favour of the CDC drives. So I had only heard of them, but never seen one.
About the ribbon cables delaminating. They look like simple straight through connectors. You should be able to design PCBs like that and get them made by JLCPCB, like the utube channel who shows all the PCB motors, I forgot the name.
That should allow you to get a stack of them and replace what needs it.
To my eyes they are not really ribbon cables but flexible PCB's anyway and PCBway also do them. I would look to replace the lot. From the video it looks a very simple design.
For the flex ribbon cables used on the bus that are falling apart I would use kapton tape and ever so carefully wrap them up.
Also replace the electrolytic caps before powering up and check the bearings on the drum motor.
I suspect the best bet is to replace them with custom flexible PCB's, which are basically the same technology and the likes of PCBway will make them for you in prototype quantities which is what you need here.
No surprise it has a drum, it looks like a washing machine. Thank you for saving these pieces of computing history.
I've heard about drum memory but have never seen it working before. I look forward to the future videos on this!
I kind of remember seeing a couple of these back in the early days of my Vo-tech training for machine shop. There was also a 'data processing lab' which is what they called the computer room at that time, and when I first started at the Vo-tech in 1972, they actually had a couple of these Litton machines. They later switch to DEC and PDP as they had better support. The only other Litton products I was familiar with were microwave ovens, which were starting to get popular in American homes at the time. It will be interesting to see if you can get this relic up and running 50 years later!😉 *EDIT:* Just a side note: In our machine shop, we had one of the first digitally controlled vertical milling machines ever released. It also used punch tape to operate, and we would spend hours on an old punch tape TTY creating the programming to control the x,y, and z parameters to mill various projects. It was before automated tool change cassettes were popular, so we had to program pauses to allow for tool changes and then hit a button to allow the tape to continue.
All these collections of obsolete old things are what we need, lest we forget how we got to where we are. Like the Sony Triniton collector guy making a Trinitron museum. Love living in a society that allows these things to take place. Love it!!!!
Drum memory! It's always fascinating too see how the old electro-mech stuff is implemented and the hoops the designers had to go through to sync memory access with execution. Mercury delay lines would be another one to see but obviously having tubes of mercury floating around isn't something you want to deal with these days.
In the first 2.5 mins of the intro, I was convinced that there was a bed in the garage background. I was thinking “oho, Mrs Usagi’s patience has finally broken with the arrival of another mini computer and Mr Usagi is now sleeping in the garage!” Luckily, we pan out and see it’s just a cover on the car! Phew! 😅
I was thinking nearly the same thing.
*YES, DIGITALLY PRESERVE THE PAPER TAPE WITH THE CENTURION!* This channel is so cool!
Old computers and old foreign cars. Your shop looks like my heaven. The Litton is insane! Congrats.
I love that you are digging into MiniComputers!! They were the first systems I worked on. That I see an overlooked era of hardware. Thanks!
Overlooked indeed. I'm ashamed of how long I went being a computer fan without any knowledge of anything pre-1980. This older stuff is amazing
So, just when I thought you had mastered the craft...you go and UP THE GAME! Yet ANOTHER Forgotten Machine rescued, restored and extremely well-documented by Usagi! I salute you, and I am truly humbled by your accomplishments!
Thank you!
Though, I'm so far from mastering the craft, it's kind of hilarious. I'm just hanging on for dear life as the world of vintage machines sweeps me up and takes me on a wild ride!
@@UsagiElectric Well, we're on that journey together, my friend...you just make WAY better videos than I do...and are FAR better organized! So, please receive that in the best nature that it is intended...
I saw David at ECF East yesterday and I can report that he is very tall.
Man I can't wait to see some running drum memory, only ever seen pictures. At that speed, probably hear it too! Looking forward to seeing you dig into it and show us more.
If at the very least the drum memory works, you have a storage option for your own vacuum tube computer.
You might even use the entire Litten machine as a serial storage device for it.
Interestingly, I've been thinking about going the other direction as far as connecting things together.
The Litton loads programs in via paper tape into the drum, and then executes directly from the drum. But, instead of having a paper tape reader feed directly into the Litton, I was thinking to put the Centurion in the middle. The Centurion could then grab a binary file and transfer it over to the Litton to load into the drum. This means that our program storage options are now immense, and we can write new programs on the Centurion and have them executed on the Litton!
@@UsagiElectric nah, keep the Litton original, let it load from paper tape. Or something else. Outside of experiments keep the centurion and Litton separate things. Otherwise you would always have the lug both around. The tube computer is a new development, you can connect that to whatever you want🙂
@@UsagiElectric That sounds exciting! How is this going?
That's an amazing story!! A piece of computer history I would never have learned about without your channel! Looking forward to seeing this machine brought back to life.🙂
When you opened up the card cage it looked like on of those frames that contained pop posters in music shops. As a kid you always fretted that they would overturn on top of you as you pulled each poster over to view the next.
It is always fascinating to see how few components were used in those old machines, on boards the size of a pool table😊
@tradde11 yes I remember those too. Later on they seemed to be mainly horizontal and you had to look down and flip them to you or away from you. Maybe the posters were larger than the vertical ones, I don't remember seeing them in the same place to compare. They always seemed a bit unstable when most posters were on one side.
@tradde11 I don't know if they still exist. They often used to be in record shops, but those are few and far between these days mores the lity😊
Wow, when I saw the thumbnail I though: Oh, was he getting bored with the Centurion and where did he come up with this one? 😂 As it turns out, they a computer relatives. Didn't see that one coming! You keep to amaze! Thanks for all the work you do!
I was tangentially aware of the link, but when I saw the machine in person and started reading the manual, it made so much more sense, and I knew I had to get hands on with this one!
Reminds me of The Story of Mel, a programmer of a drum machine.
That is so cool. My Dad had a Litton in the early 1970 at his accounting firm, 5mb of storage, paper output and paper tape for storage.
I remember reading some old usenet anecdote about a guy who had to work with/on a machine with drum storage/memory which had an infinite loop set up in it's programming, and which confused the hell out of the young tech. After much research they figured out that the program which used the infinite loop managed to make it not so infinite because of the way the read head of the drum mechanism had to physically move, and even though there was no code way it could break the loop, there was something that made it seek long enough to land somewhere "wrong" that got it out of the loop and jump it somewhere is in the program.
Great tie in between the Centurion and the vacuum tube computer. I remember someone suggesting mag. drum storage and you saying something like "where the hell am I going to get a mag. drum from?"... well, hello there!
Can’t wait to see the drum working. You might be able to build a simple optical reader for the paper tape. Punching is a much harder problem.
Could be done with a back light and a camera and just run the images/video into a script or program to decode it.
he could just use a laser cutter to punch the holes
@@nuthenry2 lasers don't like to cleanly remove the cut section.
If you're using optics or a light sensor, I bet you could just print black dots instead of punching them. Using something like a zebra thermal printer
@@edison700 If it's expecting the holes to be brighter than the rest of the paper, you may need to print everything black except for the holes. Doable but more expensive.
That is so cool, restoring the hardware and the fascinating history of how one system lead to another.
I was at Georgia Tech as the institute itself and the school of computer science each changed computers. When I first enrolled there, the main computer for the institute as a whole was a Univac 1108. And it had no disk at all, only drum storage and tapes.
Back in the late sixties I was a Physics student, and had access to a "Stantec Zebra" computer which also used a rotating drum as its RAM. It used a drum which stored 8k of 33 bit words (I just looked that up) and was the only machine I have come across which kept a teleprinter - not teletype - waiting for the next digit when printing a decimal number. It was huge compared to your Litton.
When you said "Litton" all I can think of is the "Pink Panther". Well either that or a prize package on "The NEW Let's Make a Deal!"
I had one of these computers in running condition in +- 1980. really interesting to program. I ended up dismantling it for parts, and used the drum memory unit as a tactile/auditory effect in a 2 seat spaceship simulator i built- fully enclosed with a sliding top. i removed the rubber shock mounts on the drum unit and bolted it directly to the frame of the simulator. truly awesome sound and physical effect- sounded and felt just like a jet engine. ahhhh, nostalgia. I later built another simulator that moved in 3 axes using the panels from a discarded IBM 1401.
Prehistoric computing! Converting the paper tapes back to digital and viewing it on modern computers is a whole new level of data preservation.
Ohey, my dad used to work at Litton (though the military tech division), and then at Northrup Grumman after!
wow, this is so cool, seeing a company be created to support hardware from a company that stopped making and giving support for their hardware is so wild, and then ending up making their own system with backwards compatibility for that system? just wow, basically this machine existing not only allowed centurion to exist but also made their systems use a very weird filesystem all to add backwards compatibility, this is so fascinating omg cant wait to see more of this, also, spinning ram? it looks like a very fast hard drive does it also keep data after its been powered down? does it have data rn? wild stuff omg. Also backing up those paper rolls into the hawk drive sounds really cool like digitizing it, its so cool to see how tech advanced so fast in the 70s and 80s
I love your TI-99/4a setup on the shelf. I still have my TI that I bought in 1983, which has several of the expansion components, including the PEB. Saving old computer systems is so cool.
Here we go again, Can't wait to see how this unfolds. Huge honor getting to meet you and talk a while at VCF East today, and getting sworn at in Japanese and Yiddish. I wonder if they ever got a sign of life out of that drowned classic Mac. Thanks David, keep the hits coming!
Wow, I thought my brother and I had some old computers, but that thing is older than anything we've got in our collection by like a decade and a half! That's impressive, even moreso that you got it working.
This channel just gets better and better 😁👋😊😊
Last time I saw a drum memory was in high school about 1975 . I was in an electronics course and
we got old computers donated to us from oil company seismic outfits, the drums were used to record seismic vibrations from dynamite charges set off to find oil formations underground. These things were very heavy with fixed recording heads and pretty slow rotation as I remember . Looks like you found a real winner it
Will be amazing if you get it to work again 👍
Cannot wait to see where this adventure get us! Amazing work!!
This seems a lot like the LGP-30 from about 1959 which was a deep-freezer sized tube and semiconductor diode "general purpose computer" that had a 4K word drum memory. It operated bit-serial, with 32 bits per word (though one bit was apparently a "spacer" bit and wasn't writable). It's teletype was a Friden Flexowriter and also had a separate paper tape punch and reader. The LGP-30 had a hardware multiply, though it took like 120 clock cycles to perform. The accumulator was stored as a track on the drum as well, and it's front panel had a small oscilloscope which displayed a square wave trace of the contents of the accumulator and IIRC the program counter.
One thing I recall about the LGP-30's drum memory is that after spinning down the drum you needed to wait 1/2 hour before spinning it up again. This was because the drum would heat up and expand, and if you restarted it too soon the heads would hit the drum and damage it. There was a big homemade sign taped to the one I worked with that admonished operators to be sure to wait sufficient time for the drum to cool off before restarting-- apparently, head crashes were common and it could require re-positioning and/or replacing heads so that the new position was over an otherwise unused portion of the drum-- not a job you wanted to have to have done...
That drum really precesses if you try to move it when it is running. I demolished an incomplete one of these back in the early 1980s. I still have the drum itself, and the motor, which has been repurposed to drive my hobby milling machine.
ooohhh I love the style of this machine.
The cool thing is the rotating drum memory. The only other place that I've heard it used was the A6 Intruder's on-board computer during the Vietnam war.
Those boards are a thing of beauty. What a wonderful machine! You find the greatest gear.
Well this looks like it's going to be a very interesting project!
Also, unrelated, but I loved the choice of music for the cleaning montage. It reminds me of the french duo Air.
The one-bit CPU machines, your favorite!
Fascinating. I know virtually nothing about washing machines, but.. wait. That is not an washing machine? You stole my words.
If it's 1971 you will probably find that what you think are transistors are early ic's with negative logic . I have early calculators with those and they take a lot of thinking to understand what is going on .
I am excited to be on the journey with you In your CZcams.
Cool and very exciting!!!
You have such amazing luck finding this old equipment. I'm glad you do, since we all get to see it. Thanks for sharing all that you do.
the Centurion family is unstoppable!!!! \o/
A new man. You rolled 20 years back with that shave😮🎉
Your channel never ceases to keep me amazed and feeling nostalgic. I remember in my shop on the aircraft carrier. The bench that calibrated the inertial navigation unit for the A-7 was manufactured by Singer (yes the sewing machine people). At some point Litton had purchased or taken over that Singer division. The actual Litton bench that maintained the rest of the navigation systems for the rest of the now retired aircraft, is where I saw my first multi platter Hawk drive. It also had a single fixed and removable platter Hawk drive like yours. When you mentioned that the Litton looks like a washing machine. That made me laugh, because it wouldn’t surprise me if there was a Litton link to the washing machine industry somewhere. When we were at our shore duty station. Our Litton tech rep would get calls forwarded to him from unhappy Litton Microwave oven owners quite often. We had a lot of fun with that.
Push comes to shove PCBway does flex PCBs but i do belive you'll find something off the shelf.
They look like flex that's clipped over an edge connector, if there is continuity from one end to the other it may just be idc connectors on ribbons?
@@timballam3675 looks more like ribbons are pressed onto a PCB with a bracket of some kind.
Yeah I think this would be his best option for those flex cables
The question is, did they do any money laundering with it?
Beautiful machine. The flex PCB ribbons really surprised me! You may try gluing the layers with some flexible polymer glue (no cyanoacrylate as it'll go brittle as the joint is bent).
Fortunately you can design new flex (or rigid-flex) PCBs from JLCPCB or PCBWay, and while not original, they will be better quality :)
"Miniature nuclear reactor" would be a good name for it, and the Centurion backstory is very interesting too! So bloody awesome that you got it :).
I think a short-term solution on the delaminating ribbon cables would be kapton tape. To prevent any shorts from any dirt coming in. I think I saw something in one of the closeups of the ribbons that looked like steel wool next to one of the delaminated ribbons, which would cause a short.
Yes the peeling flex is hard to solve, as you need a flexible adhesive, which probably will mean using a thin electronics grade silicone to attach the polyamide back to the tracks, after removing the plastic carefully to remove old adhesive, and then apply a thin coat to the polyamide, and hold in place, probably with a soft silicone rubber tubing, covered with some thin PTFE sheet, till it has cured.
@@tobiasfunke8990 yes, that would be the easiest and readily available solution.
@@tobiasfunke8990 I think that "steel wool" is actually foam sound deader that he was removing from the inside of the cabinet. I'm sure it was all rotting and falling off. I do hope he replaces it before starting the drum as I'm sure that changes the sound.
@@KeritechElectronics The *easiest* readily available solution is probably a can of spray clear flex-seal or even plasti-kote, and some creative (and thorough) masking. You wouldn't be affixing the layers, you'd simply be replacing one side or the other.
Although I know almost nothing about these types of machines it is so much fun to hear you talk so enthusiastically about them. I have watched every Centurion epsiode so far, so having yet another piece of the background was really interesting. Looking forward to follow your journey on this Litton minicomputer as you restore it to the same perfect condition as the Centurion! That should add another couple of years of superb content. ;-)
I think the 40bit word length came from the Zuse machines. The Z22/Z23 used 40bit memory with and eight bit tape reader/writer, drum memory in addition to core memory.
The Litton looks like a "economy" version of a Z22/Z23 to me. Would be interesting to see the connection there.
That's an interesting connection! I read some books about the Zuse computer history in my youth (I'm Austrian myself) , but it was written without technical details. So that I could read it as well. Or maybe it was in there, but I would have had absolutely no idea, what the words "drum memory" even mean. Let alone the length and specialty of a 40 bit word. Now - with the Centurion restoration and usage (usagi) videos shwon here, I learn about cool technology connections and new (old) technical terms. 🙂
Wow... so exciting to see the Litton at work !!
Great to see you, albeit briefly, at VCF East on Saturday. Good luck with this; you sure have more patient dedication than I do - lol.
An absolute amazing find! I wasn’t aware of the link between it and Centurion, in business or the sector sizing. Just WOW!!!
Computer history is so much more interesting than I thought. We're in "the future" where we can put 3-7 billion or more transistors in one CPU, but only thousands back then. They had conventions back in the 60's about the future of computing and AI, which kinda matches up to what we're living today with AI generated art, machine learning, quantum computing, etc. It's kinda cool how far we've come, and the potential for what's to come.
I luv your videos and always look forward to seing more and learning more about such vintage technology, i also like your furry friends :) The Centurion videos have been really cool too..
I love this machine and the story attached to it! I'm so excited to see what the future holds for it!
I am also very curious about the result. They were great times when they came on the market. I am following you closely as I want to see this working, good luck, thumbs up!
WOW....A Litton 1241 ABS Controller Unit. The first program I ever wrote was on this machine. If I recall, the paper tape is encoded as 7 bit ASCII with an extra bit for parity. You should be able to clock the old tapes through an optical array to read them. I remember that the Litton programming language could pretty much be likened to a subset of Assembly language. I was about 15 yo when my mother bought a used one for her accounting practice in or about 1976. It cost around $18k, which was a huge amount of money back then. I hope you find a paper tape reader/punch and teletype. Thanks for the memory.
To everybody who had a C64 growing up: Consider that to even come close to what you had in your bedroom in the eighties, a business in the 60s/70s would have had to have had THREE of these Litton machines AND 4K of actual core memory.
Back then, RAM was a limiting factor. It took a while for RAM to become plentiful and cheap.
Many of the PCB manufactures (like PCBWay and JLCPCB) offer FlexPCB now. So it should be possible to make new ribbon cables
I was working with a massive dual PDP 10 system that had it’s hard disks replaced by Drum Storage back in the late 70’s. it’s memory was magnetic core.
Holy heck, the Rhodes in that time lapse music is sooo good.
This is awesome. For the papertape, maybe CuriousMarc can hint a help.
The mainboards are basically bound like a book. Neat.
That ROM had a surprisingly late date, 1990?! Paper tape and punch cards were over in the early/mid 80's.
Probably built to retrieve information that was on older paper tape into a newer computer.
DEC produced a lower cost version of its 12 bit PDP-8 minicomputer with serial ALU, the PDP-8/S. Drum memory was famous in the IBM 650, which served as IBMs entry level computer for many years in the Fifties and Sixties. Drums were used later for fast random access storage. One Burroughs drum weighed literally a ton and tendended to break free and walk around due the forces generated by its fast spinning drum. The Army and Navy tried to use the Burroughs drum in mobile applications but the the rotational forces were too much for the ship or truck carrying the drum.. The drum in this machine is a relatively minaturized version of the breed. Drum memory machines had to account for drum rotation when placing instructions. The techniques for optimium drum computer performances are similar to those used with RISC machines. Dead space in the instruction stream can be filled in by unrelated secondary operations.
Very interesting story between those two brands of minicomputers!
Whelp, I missed you on Sunday, I was beet from the walking around on Saturday, but you Guys/Girls were on stage, and all had to zoom out at the End, Hope the 60's RCA IC made it home with you, it a Pre Production A Version, Just Try and Find some Info.. I wanted to come back Sunday, MY the HVAC unit took a crap on me on Saturday, And Sunday was House Repaire weekend YAY..
To me, 40 bits signal "IAS machine heritage". Many of the direct descendants of "the IAS machine" were 40-bit machines, jsut as the IAS machine was. And they, in turn, gave rise to other machines that were 40-bit. The earliest commerical model I am aware of is the Ferranti Pegasus, but I am also aware of prior machines using it, although they're more "one-off machines". BESK, who begat SMIL, DASK, and SARA (all 40-bit machines on the general IAS architecture spectrum). SARA then begat CK37, and D2, but I do not know if either of the latter were 40-bit machines (CK37 is the on-board computer in the Saab Viggen military jet). That 40 bits happened to also be handy for business computing...
You get all the amazing things! That Litton amazing. It does look like a washing machine! Can hardly wait to hear what the drum memory unit sounds like! Thanks for the history lesson, as well.
Very cool. Amazing how things were in the realm of old computers. The history is very intriguing. Thanks for sharing it with us.
It is a crazy turn of events, but so exciting to follow. Thanks a lot.
thank you for saving these examples for the future.