Reverse-engineering an IBM Mainframe Vacuum Tube Pluggable Module
Vložit
- čas přidán 30. 08. 2020
- In this episode, we get our hands on a pluggable module out of an IBM vacuum tube computer. IBM used these awesome modules that could be pulled out as a unit, and each module consisted of some type of circuit. For example, some were cathode follower buffers, some were dual inverters, and so on. I’m not really sure what this one is, so today, we’re going to see if we can’t figure out how it works!
Thanks for watching!
Wiring diagram: i.postimg.cc/R02jHFm3/Wiring.jpg
Schematic: i.postimg.cc/cHYPMS1F/Schemat...
IBM 604 Engineering Manual: www.bitsavers.org/pdf/ibm/604/...
IBM 650 Engineering Manual: www.bitsavers.org/pdf/ibm/650/...
Computer picture 1: commons.wikimedia.org/wiki/Fi...
Computer picture 2: commons.wikimedia.org/wiki/Fi...
Pluggable module picture 1: commons.wikimedia.org/wiki/Fi...
Pluggable module picture 2: commons.wikimedia.org/wiki/Fi...
Music: Suco de Melancia by Red Tailed Fox - ocremix.org/remix/OCR01671 - Věda a technologie
This was an amazing trip down memory lane for me. My dad was a field engineer for IBM for 36 years. He started in 1955, 9 years before the S/360 was introduced. He actually worked on this type of equipment. He regularly brought "dead" parts home that he didn't have to turn back in and gave them to me to tear apart. Between that and the occasional trip to one of the data centers in his territory sparked a lifelong love of computers and electronics that eventually led to a 39-year career in mainframes, which I truly love. In this era of terabyte cell phones and 5nm VLSIs, it's refreshing to work on projects where you can actually see and touch the individual components. Thanks for helping to keep the curiosity about vintage computers and electronics in general alive!
Back in 1964-66 when I was in high school in St. Paul, a friend of mine and I discovered a junkyard that was scrapping out mountains of IBM tabulating and early vacuum tube machines - primarily RAMAC 305 and 650's. They would sell us pieces of machines and I grabbed a bunch of these modules, along with power supplies and front panels. At one time I must have had a barrel full of these modules. Wish I had saved them, but I left home for school and later the Army and my dad basically gave them away! According to my dad (who worked for Sperry Univac at the time), IBM would rent or lease all of their equipment and thus retained ownership of the machines. When they became obsolete they were sent to a facility for decommissioning where cables were cut, heads were welded to disc surfaces, screw drivers were punched thru core memories and front panels were destroyed. Then the machines were sent to scrap yards like the one I had found. Thus it is doubtful that you would find a complete machine unless it came out of a museum. IBM did not want the old machines hanging around and competing with the new ones. I am enjoying your vacuum tube and computer series very much. Well done! I am considering doing something similar to your vacuum tube computer but with relays instead.
Reverse engineering is such a wonderful journey I think students really miss out on. Great work. A pleasure to watch.
You made my day. I haven't done any vacuum tube work since 1970 but as soon as I saw your wiring diagram I remembered pins 3,4 and 4,5 for heaters and that 4-5 connected meant a split heater. When I began teaching electronics technology at Ferris State College (now University) in 1980 one of our early tasks was to remove tube tech from the curriculum. The reasoning was that the few graduates who would need it would pick it up quickly by individual study. Other comment suggested Paleo-electronics term. Love it!
This reminded me of the time back in the very early 1960s when my dad worked at Corning glass and they made resistors and capacitors. Thanks for the memory.
Corning Glass making resistors and capacitors seems an odd thing for them to have been doing as a side business.
For computer applications special versions of vacuum tubes were made
to avoid cathode poising when tubes were in idle states for extended
periods.
Cathode poisoning was a serious issue in Nixie tubes, too.
Level shifting NAND gate. One of IBM’s “signal restoration input” series.
I have a pair of IBM 60x modules too. As a collector of anything old and computery, they were always a bit of a bucket list item for me.
I've just found this video (and your channel) and loved every second of it. Subscribed and will be watching quite some more for sure. I've just turned 50, so I've had almost no contact with tubes, none at all in the digital realm, besides some museum pieces, so your sleuthing and excitement are appreciated and shared, Thanks!
I love your videos, I've always been a vintage computer geek, when I was younger (and had my parent's basement to put things in) I'd collect old Hard Drives, racks, servers, and even weird stuff like VCR/BetaMax Editors. I got most of it from my school, they were happily offloading their vintage hoard onto me. However, I never could get into vacuum tube stuff for so many reasons. Your channel really reminds me of those great days. With rent being so expensive these days I can't collect as I used to, not if I want to have space to live that is. Thanks for sharing your adventures.
Your videos are so good quality, you deserve more subscribers
Thank you so much! I'm super happy that you guys are enjoying the videos!
I agree so I just became a Subscriber videos are awesome
I didn't know I needed this video today. But since it found me and I enjoyed it, I sent me sub!
Oh wow, I have seen modules like this online from time to time and as I experiment with homemade logic circuits I just have to find one of these and try it out. I have not worked with tubes for many years but after watching your videos I am catching the bug again 🙂
The Teleprinter sound in your intro brings back some memories! I used to run several different Creed machines (Model 3, 75, 444 and 2300) in my radio shack 30 years ago.
Thanks for the 'forensic' work with that module. Very educational. 🙂👍
What an enormous effort for two simple NAND gates.
Thanks for the video and thanks for 74XX chips ;)
I really enjoy this dive into classic computing tech.
What do you even call this field of interest? Paleo-electronics? Regardless, it's a very interesting video and I appreciate you making it. I am old enough to remember every piece of electronics having vacuum tubes. It's great to see someone young keeping the art alive. Thank you sir.
Fascinating work you're doing with sleuthing out the Neanderthals of computers. Being a child from the 50's I also work with tubes but in the audio realm. Working with plate voltages in the 250-500v range is normal. I've also worked on RF transmitters where the plate voltages of the output driver tube was in the 1500v range. That will make your hair stand up, literally. Lol. Great work you're doing keeping tube technology and knowledge alive. 👏
The tube tech is actually Old World tech, and we will end up studying what we had (whats left of it) and going completely backwards in coming years, our engineering will never come close to what was here before colonists arrived.
What took me so long to find your channel? This is great content - VT logic! Have been working with vacuum state technology in AF and RF applications for years. I have seen very little experimentation with logic on CZcams until your channel. Subscribed/All notification - and, I must be cautious not to binge! Keep up the great work!
P.S. I've had 320 V through a breadboard rail with 1 or 2 ma of current for short duration tests with little problem. Key phrase: "short duration tests"!
Another great video. I'm still catching up on your excelling work.
I did want to mention that you should consider 9v batteries in series for your high voltage tube tests. The low current and high internal resistance might even be a good thing.
Oo love the new intro. Such an underrated channel
When I was growing up (early 60's) you could buy NOS IBM flip flop assemblies at Radio Shack for $1.50 each. They also had grab bags for $5.00 that were all IBM computer hardware.
Having viewed many of you other videos in a short period of time (very bingeable LOL), whenever I see those milled circuit boards, I immediately think "Usagi Electric." Between their color and the way the traces look, they're kind of like a signature. :D
Genial, me ha hecho recordar los inicios en mis estudios y primeros trabajos en electrónica.
25:58 your low voltage is exactly one third of the 24-volt. So if you compare this to a hi of 150 volts and a low of 50 volts, the low is one-third of the high.
Thank you! I learned so much from this video!
Thank you so much for checking the video out!
wow, I didn't even know those tubes existed. Great Time watching this!
That Uni-T meter is a beast. Glad to see it becoming more common
I have that same model in my toolbox for mainly current metering. My old GwInstek 396 multimeter has bit better continuity check beeper thou (detects crackling contacts).
Extremely interesting and well done. I'm a subscriber.
Cool. When IBM had Server Group Headquarters in Somers, NY, the central building had a 701 in the lobby. The tubes were exposed, and only 3 or 4 were "missing." I don't know where the machine is located now, but IBM sold Somers in 2016. Perhaps it's in the "archives" in Kingston, NY. I can probably find out.
Also, when I was a kid, Olsen Electronics in the late 50's and early 60's (a retail store) would sometimes sell single and multiple tube IBM "pluggable units." I bought a few, but they are lost forever.
Finally, I worked with many IBM engineers who developed the 70X series and many other mainframes. IBM named many of its Poughkeepsie buildings after the 70X series, i. e. 701, 702, 703, 705, 706, 707. When you asked the engineers about these machines, their eyes would light up like tube filaments.
This is like archeology but with technology. You are amazing. It hurts my head trying to imagine how smart you are.
Techeology? :p
Nice, logical reverse engineering. I was also waiting for the breadboard to melt. LOL.
Thank you!
Not gonna lie, I was a little nervous running higher voltage through the breadboard, but fortunately, everything in use here was at extremely low amperage, and I wasn't at high enough voltages to cause arcing, so it all worked out alright in the end!
nah, the traces is probably far enough, 150V is not that high for the arc to jump on air that distance. as long as you have a good enough bread board. perhaps if you tried to pull more than 10A it would blow up
Great video man, high production value and actually interesting. Personally, I've always had an aversion to vacuum tubes because of the not-easily-attainable voltage levels and high impedance outputs, but it's great to watch someone passionate about them. I have a question, but it feels a bit weird to ask because it's not about the actual module that is the topic of this video but rather about your breadboard. I see you have jumpers on your breadboard's power bus, but the bus markings are unbroken, indicating there's no break in the bus itself. Did you do it as a precaution because you got burned with this bus discontinuity stuff before (I know I have) or is there an actual error in the markings (which I have to say I've never come across)?
Thank you so much!
That was actually my primary aversion to tubes for a long time as well. High voltage scares me because I have a bad habit of sticking my fingers in circuits. But, we had some spare tubes from salvage in a box, so I decided to see if I could get any reaction out of them at all at lower voltages. I tried a variety of voltages from 6V to 30V and settled on 24V because it worked well for stacking heaters in series.
No worries on asking about the breadboard! The best part of making these videos is interacting with people, so feel free to ask any questions, whether related to the video or not! I don't know why, but these Miyako breadboards I'm using don't have the broken markings, but the bus itself is indeed broken. It took me the better part of a day to figure out why half my circuit wasn't working the first time I built something big on the breadboard. Now, I chose to go with the Miyako brand because I have another Miyako breadboard I picked up years ago in Japan that I love and the bus isn't broken on it. So, I have no clue why they didn't used to be broken but now are, and it's even stranger that they didn't mark it.
Here's the Miyako breadboard: www.amazon.com/gp/product/B014LEA2DU/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&th=1
Thanks again for the kind words!
for a display you should mount it on a piece of hardwood, and put a couple fancy switches on it, with the neon bulb! and make it a functional tabletop conversation piece. very cool video :)
Hello, this was AWESOME! I learned the biggest amount of information about vacuum tubes from your videos. I was curious how did you made that socket, what were the things you put the socket in? Keep up the good work, you're doing great videos!
Thank you so much!
The socket is actually just an extension of the 9-pin sockets that are similar to the sockets I made in the previous episode where I make the circuit board for the SR flip flop. I have a custom template in DesignSpark that draws a large pad and machines a large hole in the center of that pad that little 1mm PCB headers (www.mouser.com/ProductDetail/Harwin/H3161-46?qs=p%252BybrH12zCFIuSSHm88cAQ%3D%3D) can be pressed into.
The IBM module uses the same pin spacing and diameter as a 9-pin vacuum tube, except it has an extra 10th pin. That made it pretty easy to modify my 9-pin template to fit the IBM module!
Thank you again for all the fantastic support!
Thank you for the video and for the explaination
Thank you so much for checking the video out!
Good job!
I never thought a second about the breakdown voltage of a breadboard, until I saw this video.
Excellent video. In my mind, what is the fastest you can get the dual NAND gates to flip/flop? Of course that is base on the inputs but at what speed does this become useless?
My guess is that you've got it bang on. Dual Nand gate. Great bit of sleuthing.
Nice work! Lots of combinational logic typically use NAND gates. I worked on a early computer Univac 1219B that had tons of little transistor flip-flop cards (and also tons of neon displays!) Maybe this was a 1 bit register (SR Flip-Flop)?
Thank you!
That's awesome that you got to work on an old Univac! Transistorized computers were insane. Even with transistors, they were still absolutely massive units.
This very well could have been used as a 1-bit flip flop. Though, I think IBM's goal with these little modules was to have them as universal pieces that could be interchanged and used wherever necessary. So, I imagine that things like this module also got used for the glue logic that helped connect larger sections of the processor with each other.
Yippe, just random found you. Subscribed, liked and ready to go back to the beginning. Great video.
It's good practice to turn on the heaters and let the tubes warm up first. When they are stable you can turn on the supply voltages safely.
Do you happen to know how the tube based amplifiers that were used on the undersea phone lines worked?
very informative
Would have been interesting in seeing you make a voltage ladder for a higher levels.
Really nice video! I remember seeing modules like this when I was much younger (in junk stores), but having no idea what they were. One thing that stuck in my mind were some of them having RCA Metal encapsulated tubes - which I've never seen since. Have you come across them? Maybe I've imagined it! I'm guessing they were just to make the tubes less fragile.
Thank you!
Interestingly, I was just talking with someone a few days ago about the RCA metal encapsulated tubes! I've never come across any, but there are some interesting theories as to why RCA built them like that. The most popular reasoning is that it was a shield to keep interference down, but my favorite theory is that Germany was producing full metal tubes, but RCA didn't quite have the manufacturing capability to do that just yet, so they encapsulated some glass tubes in metal as a stop gap to compete.
Who knows what was actually going on, but there are certainly some interesting theories out there!
@@UsagiElectric Mr. Carlson's Lab runs across metal tubes occasionally in his old test gear/radio restorations. I always enjoy Paul's content, and he is a true tube wizard. I've just discovered your channel, and have next to none as far as electronic knowledge, but hoping to pick up bits a pieces following along with big brains such as you.
Thanks for your great vids!
DT
What do you use to make the pcb?
You're a cool dude!
So one of my Crazy Questions: Could you use this vertical Tube design (the hardware harness-- U-bracket with 2 PCB boards) to shrink your Tube R-S Flip flops foot print for your memory and would it be worth trying ???
Absolutely!
But it's kind of counter to the style that I've been building on. The tube computer is honestly more of an art project than a legitimate effort to build a capable tube computer, so a lot of the inefficient use of space is actually kind of intentional to keep the aesthetic looking good.
Try using space charge tubes 12 volt on filament and plate. 12G8 is a dual triode.
These were used in car radios.
Great! Thanx!
Question can you Put a zener Diode into the Output? IT would make the Output much clearer.
The only additional data point that might have been interesting to measure is the switch time. Of course, that would have required some sort of wave generator of the appropriate voltage for one of the inputs and a dual-trace oscilloscope as desirable to display the input and outputs together as you varied the frequency. That all said, I admire the work and knowledge you put into this. Being born in 1957, I missed experiencing the vacuum tube computer era, so I always wondered about the circuitry I see. Cheers!
Somehow I saw a vacuum tube computer at a flight simulator on Long Island when I was in the 8th grade, my Sunday School teacher ran it or something. It was really amazing, big room full of racks with those 6h6 tubes or whatever they were. Sinclair I think was the company.
Your a genius !!!!
非常有意思的内容，订阅了
You the man.
Since you could not find this circuit in the IBM manuals, I am wondering if perhaps it could have been made by one of the other early computer manufacturers (Remington Rand, Eckert & Mauchly, RCA, Ferranti, Engineering Research Associates, etc). There were also some vacuum tube computers that were built in-house by universities, government agencies, and militaries of various different countries. The reason I think that very few of the tube computers appear on Ebay, is that I suspect that many of them were dismantled, and parted out a long time ago. I also think portions of them were donated to museums. Some may have been destroyed completely as some of them were used for very specialized applications such as encrypted message code breaking, missile firing tables, math research, aerospace, census, income tax processing, research, and a variety of other very specialized applications. Computers would not appear in widespread home use until the late 1970s and early 1980s.
I'm fairly confident it is from an IBM because in the IBM 700 Customer Engineering manual they show the shape and pinout of the pluggable modules they used and it's pretty much identical to this one: i.postimg.cc/PxDPZ3WL/700-module.jpg
My guess is that this particular unit is from the 704 or 705, but I just don't have good documentation on those machines to confirm with.
You're absolutely right that a lot of the computers got dismantled or parted out long ago! I kind of see them like a 10 year old Dell desktop. I love salvaging old stuff, but I'd be hard pressed to pick up one of those, haha. In the 70's, when the transistor was bringing computers down to manageable sizes, these old tube mainframes and computers were often returned to IBM and pulled apart. It's a shame because they were such instrumental pieces of history, but that just makes it that much more special when you come across a piece from one!
Too Cool!
Around 1980 I bought a literal 5 gallon bucket of those. Most of them had a few 1 watt resistors and a 2D21 thyratron. So probably drivers for printer or card punch solenoids. Some had 5965 tubes, so probably flip-flops.
Oh man, I wish I could get my hands on a 5 gallon bucket of these! The IBM 604 is a bit of a dream machine for me. Someday, I hope to find and restore one, though that's certainly one of those pipe dream things.
The ones with the 2D21 thyratrons were most likely used as solenoid or relay drivers as you guessed! The ones with the 5965 tubes were actually probably used as power tubes. IBM actually preferred to use the 6J6 7-pin dual triode (with internally shared cathode) as their primary "trigger" tube (they tended to call their flip flops/multivibrators "triggers"). The larger 5965 dual triode was hotter and physically larger, so it was used primarily where they needed more current. They would even run the two triodes in parallel to get even more current driving capability. Sometimes this was to combat fan out but also, sometimes they would use this when the logic signal had to cover a long distance to make it to another part of the machine.
Of course, that was just for the IBM 604. IBM used these types of modules in many different models and each one model is pretty wildly different, so it's really hard to tell exactly what each module was used for!
@@UsagiElectric I have a 5727 sitting somewhere which is a ruggedized 2D21 with a higher withstanding voltage. 2D21s were also used in electronic metronomes.
This guy is very intelligent !!
Very interesting and instructive test. Survival of regulators without real heatsink surprised me too although it probably didn't run for long enough. I'm seeing lots of tubes in semi pro audio running at around 24V because they distort a lot at such low voltages, which is what many people like.
Anyway, test shows tubes are very bulletproof and of course much, much bigger than the same thing made with transistors. My guess is tubes will be used in audio for many years to come because there are no solid state counterparts when it comes to harmonic distortion, headroom, etc. Transmitters are another application for them.
Thank you!
Growing up, I always heard about how tubes were super unreliable and broke all the time, but my experience has been that tubes are, as you said, pretty bulletproof! Out of the hundreds and hundreds of tubes I've been messing with so far, I've only had one tune that had an internal short, and less than 10 that I've broken the glass on due to dropping them, which is actually not all that different of a failure rate compared to cheap ICs and digital components. Conversely, it's hard to find an old Commodore that doesn't have a bad chip, but almost every old tube equipment I've come across is still rocking the original tubes, despite being often times 60 to 70 years old!
@@UsagiElectric Old circuits are pretty simple, of course depending on application, transformers needed to match impedances also help protect them and give them big part of the sound. Microphones, limiters/compressors, preamplifers, power amps and similar often sound much better with tubes than transistors. Of course it depends of what one wants. Russia is continuing making tubes for 100 years or so, China joined too so we have supply of well made components comparable to old production. I'm interested in knowing about your stock of old tubes, if you want to share information about mine i can send you contact. Btw, modern synthesizer modules are often made from receiving tubes, rectifiers are solid state because they don't influence the sound or function in most applications.
Some tubes are actually quite clean sounding at 24V.
@@douro20 Still not nearly good enough for pro audio or even serious guitar amps. I'm talking about in production receiving tubes, not NOS stuff that is getting depleted.
@@UsagiElectric I was born 1960 in Helsinki, Finland. I remember that once we (me and my father) went to a tube shop with a tube from our B&W TV to check if it has failed. They had somekind of test bed where you inserted the tube and it told if it was working. The tube weren't and so my father bought a new one. Can't remember that we went there more than that one time. What I mean that by my recollection the tubes weren't breaking willy nilly. We also had a tube amplifier for our record player and it worked until 80's when my father bougth a stereo amp. Can't remember that we changed tubes in that ever.
In 70's, when we had bougth a colour TV, I took the B&W and dismantled it studying how it was made. It had full circuit diagrams folded like a map and put into plastic holder behind the TV. Whenever I had some problem to figure out I asked my father to help and we figured it out togetrher. My father was an electrician. I'm an telecom engineer but studied later to became an automation electrician. Love watching these.
Great!
Thank you!
I'm trying to make a replica model of this. Just as a display piece. Could I trouble you for some measurements or even better a picture with a ruler next to it?
Absolutely, but I'm out of town for the next week-ish. If you have Discord, shoot me a reminder and I'll get it when I get back
(discord.gg/sQGCFSJeXc)
IBM: "We need to call this guy, pronto!"
**dials phone**
Phone: BUSY SIGNAL - LINE BEING USED FOR 1980's ACCOUSTIC COUPLED MODEM
Can you tell us what that 6V power supply module is?
Sure thing, it's just a little buck converter I found on Amazon (link below). I just soldered header pins to the Vin and Vout pins and it was just serendipitous luck that it lines up perfectly with the voltage rails of the breadboard!
www.amazon.com/gp/product/B018JPGKQK/ref=ppx_yo_dt_b_search_asin_title?ie=UTF8&psc=1
So thats what these things are!
I have hundreds of them and never knew what they were. I also have quite a few boxes that have numbers 0-9 on them with 9 12AU7's in them, and the box uses a single octal 8 pin socket to plug into whatever it did. Wonder if they were also used in this application?
Ah...good work....figured out...nice.
When I see two NANDs together, I think… a flip flop. A single bit of storage… So maybe this wasn't for computation but for storage?
Now I really, really wonder if the naming convention for the PowerPC 603/604 CPUs had anything to do with the much earlier IBM 603/604s.
UE reverse engineers a dual diode by memory. Gets excited about a neon bulb. :) I'd be fascinated to see how much fan out a person could get at lower-than-spec voltages compared to the 150V. With modern output devices, could a usable adder be done on lower voltages? I hope UE can find more modules!
Thank you! I do tend to get excited by the simple things, haha.
It's funny you mention a usable adder as that's generally my go to circuit when experimenting. It took me a lot of experimenting and revisions, but, that's something we're ultimately working towards with this channel/ It's quite surprising what tubes are capable of at just 24V!
Here's a sneak peek: i.postimg.cc/856TBWwm/IMAG3002.jpg
I'm always keeping my eye out on eBay and other places for more modules, if I can find some more, you can be sure I'll be snatching them up and figuring out what they do!
@@UsagiElectric The photo definitely piques my interest. If I find anything I will send it to you. About all I've got left myself now, though, is a falling apart RCA vacuum to manual. Those were the days...in the basement as a kid checking for linearity on an old Simpson.
yeee so compact tube circuit and small
I like it
IBM were really masters at design I feel, this is such a nice, compact solution!
Costa Del Sol!
I think you have something with more potential than you imagine. Let's say you have 100 units identical to this one, all wired exactly the same. Without changing anything on the unit, you should be able to get all 5 types of gates. And, or, nand, nor, invert... the way i would do it is plug all of these units in the front side of a 2 sided rack panel. The back side has access to the same parallel socket. Into the back socket, you plug a personality module that wires it into the type of gate you want. If you want a AND gate, you plug in the AND personality module. With only an AND gate and Inverter, you can make any of the gate types listed above. Does this sound like a workable solution?
Hmmm. A NAND gate should be able to make any of the other gates I listed above.
It’s crazy the lengths they went to for logic back then.
150 volts ! It's for science b**** :-D
This is a really nice piece of hardware that you got here, enjoyed it a lot !
I am wondering what you are planning for the next video, we'll see it soon :-)
Thank you so much!
Fortunately, DC is a lot less scary than AC, and while 150V can make some pretty sparks, it's a lot less likely to really hurt me. Still, not recommended though, haha.
The next video has me pulling out a piece of old equipment and fixing it up, so I can test some theories and ideas!
@@UsagiElectric Yep, AC is amazingly efficient but a lot more hazardous ! However, you know, I managed to see a little arc with the power supply of Ben's 8 bit computer. I don't know if it was that dangerous for it was tiny but I fixed it anyway ^^'
Ohhh this sounds cool, can't wait to see it soon :-)
@@antoinestevan5310 DC is more efficient than AC.... Thats why very high power transmission lines across states/countries are DC like Basslink (between Victoria and Tasmania)
What a beautiful thing, half a 7400 with valves, wonder how fast it can switch?
Probably up to 1MHz or so. So it offers 4011 CMOS speed & output current with the power consumption of several 7400's. But the glass package is so much more appealing than epoxy packaged silicon
Also, being a dual-nand, the module can probably be configured to be a latch.
Thanks for checking the video out!
There's actually a couple of different ways to make a latch using tubes. IBM liked to use an incredibly simple multivibrator setup with a single dual triode and a handful of resistors that works surprisingly well!
I go into it in a lot more detail in this video: czcams.com/video/JmNRxVi2y-g/video.html
Would love to see you make your own 24v version from scratch
I've actually built quite a few NAND gates inspired by this design! Whenever I needed a NAND or AND gate in the tube computer build, it was often a design not all the different from here.
Here's a short 1-minute video I did on building a NAND and AND gate at just 24V: czcams.com/video/WRzzEaXgjag/video.html
@@UsagiElectric ah yes, but I meant an exact copy of this double decker arrangement
Молодец информативнное видео !
The figure II-11 of the IBM 650 manual shows a "Diode AND circuit" with "projected" voltages - which also confirms standard logic level of +10V ("signal level", i.e. 1/H/high) and -35V ("no-signal level, i.e. 0/L/low) - keeping kind of standards during transition (mixture) from tube to semiconductor diodes - and a pullup "load" resistor - which may apply in your case? Just to make a really "educated" (no) guess - does this make sense to you?
The pinout on the module is like a 10 pin relay connector.
But the notch on the center pin is missing.😁
I love it, but You know You look like dr. Brown from "Back To The Future"? :D
Thank you!
And I'll take that as a compliment, Doc was my favorite character!
Nice alanyslis. 150v is easiest. Take AC 120 caoacitor couple in series use full rectifier, voltage fuse voltage drop resistor dexouple capacitor and blerder trsistor... same rig with half wave rectr. two dexouplimg capacitors in series centertap to ground you have + - 35v use voltage drop reistor or set of resistors i seties to capacitor for adjusting voltage divider. Thats old axhool. DC load B+
Thank you!
I'm not quite sure I follow your comment, but thank you regardless!
@@UsagiElectric its based on " capacitor.loaded AC" a.k.a Marshall bias circuit. Also two diodes head to head with 10.000 uf 10v capacitors in paralel to each diodes do the trick. ( diodes clip AC mains voltage to few volts). Rest is voltage divider.. note: first circuit is galvanicaly isolated from mains by capacitor so depending on how you hround it it can ne B+ B- or both with centertap grounded capacitors in series.
Get another, you could make a SR Flip flop, and have 1 byte of storage 🤓
Why?
Wow man! Your channel's logo is from old russian animation)
He's a comunist. If he wants to know how it works, why doesn't he just call IBM and ask them?
I don't know why IBM didn't use sub-miniature valves instead of miniature valves as the modules would've been a great deal more compact especially if used in conjunction with PCBs.
my high school electronics teach is screaming at those resistor schematics. Was drilled into me "3 spikes".
Either a NAND gate ... or a 32-bit ALU with Barrel Shifter.
You decide. 🤔
Transistor cu
ALU logic oproming unit
Next recommended video: Gigatron 🤔
NAND gate (double/dual)
Knowing a thing or two about serial computers, (czcams.com/video/Wa7KVU_e8U8/video.html) I wouldn't expect this circuit to be used as a "NAND" but my guess would be it is a "gating circuit" with one input supplied with a bit stream and the other input with a control signal that either lets the bit stream through or blocks it. There are other possibilities as well.
Thanks for the insight!
That makes total sense. Actually, I often use NAND gates for gating in my own designs, so that seems like the most logical use for them (I use NOR gates/multivibrators for making flip flops and all the other logic in my tube computer).
That Elliot 803 is a super cool computer! I always forget just how big transistorized computers really were. They were definitely an improvement over vacuum tubes, but it wasn't until ICs and minicomputers that computers started to get down to sizes that were a bit more manageable. Man, I'd love to play around with an old tube computer or transistorized computer one of these days!
@@UsagiElectric If you are ever in the UK drop me a line and we can meet at TNMOC for a day of old valve and transistor computers !
This it looks like a simplfied pare of your 555 video
before I start watching this..... can't you just reverse engineer it by looking at em, I assume can also break em in the same way sometimes but yeah
Old computers did it all first. Really, the only thing that happened in the past 50 years is everything got smaller. Get rich, get workers, buy a warehouse or two, and you can build an old computer that does everything a new computer can do, and more, because where new innovations are done through a microscope, old computers just get bigger so you can literally walk around inside each component and see in giant detail what can be improved. Got a small part that you want to change its function? Then just blow that box up to room-size and stick another large computer inside it. It's the difference between growing life in a petri dish and growing life on an actual planet.
When I first started learning about the old IBM machines like the 604 650, etc., I was blown away at how the fundamentals of computing haven't changed since the 1940's! You're right that really the primary change has been size, though speed has also been an area of pretty impressive improvement. Vacuum tubes definitely have a speed limit. The IBM 604 had a clock speed of 50kHz, which was impressively fast for the time. For my tube computer though, I'm aiming for a clock speed that I can vary from about 1Hz to maybe 1kHz peak.
Still, if I got rich and got myself a warehouse, I very much so would try to build an 8-bit vacuum tube computer that can run Basic, haha!
Well, actually.... I am no expert on old computers, but using binary logic was a not the obvious choice from the start. If I remember correctly a handful of computer pioneers used trinary logic, that is - 1, 0, and +1. I am not sure why or what the advantage is/was, but it was used.
Damn, Son...Don't you have any TRANSFORMERS? (I used to have a good sized BOX full of them!) You need to find some really old TVs ...Not too many of those still around. How much current would you need for the 10V? You could build a voltage regulator to drop 12V .It should be able to provide enough current. you might want to check an older (pre-1990) ARRL handbook for a quick and easy bias power supply (maybe even using vacuum tubes!) Yeah, You have to be VERY CAREFUL around HV. I have been bitten a few times ... NOT fun!!!. REMEMBER the ONE HAND RULE! Only ONE hand anywhere near ANYTHING with HV! Do NOT touch anything grounded with the other hand!) Hell, I would of used a non-conductive tool like an "IF stick" to turn those switches off and on. REMEMBER, "VOLTS JOLT - mills KILLS!"
You could have built a latch.
😂 just wait until the fallout series becomes cancelled.. everyone will use this video for homebrew vacuum tube pocket watches.😂
First generation
Third generation transistor cpu