@@locutusofborg Apparently more than just you! I thought it would be a pretty safe joke in this crowd. (Safe mainly in that the creator of the UE-1 probably would see it and take no offense.... I want the kit version of the UE-1, and the sawhorses to help me put it together!)
I've been working on developing the core theory for building a real sentient Strong AI for a long time. It would be difficult to get it to work on a UE-1 but it would be so much fun..
When we did computer history at school and I learned the "first generation" used thousands of valves, I thought "I'll never ever see one of those working!" To my younger self, it's pretty much a miracle that you've got 2 to play... erm... experiment with. :)
Ehhh... The true "first gen" machines were relay-based. Very slow, but much more reliable and affordable. Vacuum Tubes made computers faster and more practical, but appeared about a decade or two after the first relay based machines.
@@the123king Yeah, Conrad Zuse didn't really get a look in in our syllabus back at school. And Collosus was still secret back then. So for us, back then, things all started with the likes of Eniac and the Manchester MK I.
I was just thinking the same thing. Tommy Flowers did some remarkable things with that design - like applying his GPO knowledge that valves are (relatively) reliable if you just leave them on all the time to make the Collusus machines using valves when others though it couldn’t be done. One day I’ll make a pilgrimage to TNMoC to see the replica - and other goodies they have.
@@timothyp8947 I was there before the coof. Do it. The history of Bletchley Park alone is worth it (GCHQ) but TNMoC is the cherry on top. Make sure you go on a day when they have the full tour, not just Colossus!
Interesting! You hit exactly the same problem as the first pioneers of stored program computers (SPC) - memory using just tubes (valves) is too expensive/bulky/energy-hungry! The very first SPC to run was the Manchester ‘Baby’ SSEM that was built as a demonstrator to show how a Williams Tube memory could work in 1948. It had 32 words of 32 bit wide memory (1024 bits). 75 years later and Usagi is on the job!
wow, building 1024 bits of ram in VVL logic just as a proof of concept sounds a bit masochistic... Usagi, vacuum tube memory is so 1948. All the cool kids use delay lines nowadays.
@@yuvalyeru The Williams Tube memory used spots of light and dark on a CRT as a kind of recirculating delay line memory, but much faster than acoustic delay lines. They were used in one of IBMs first commercial computers too, before they got magnetic core memory to work.
Those sockets (or something like them) are used in custom mechanical keyboards. They're incredibly resilient and resistant. The most popular for keyboards are 3305-1
I saw the switches in a newer video, couldn't believe my eyes and went to this video. Soviet micro toggle switches (in this case МТ1), which I absolutely love. They are damn reliable. They are still produced and are quite expensive (compared to similar Chinese switches), but in post-Soviet countries you can find used ones much cheaper, they work no worse than new ones. So I also use them in my projects.
This brings back some great memories when I was in college: They only had one lab that I got to play with vacuum tubes. I had another lab where we built a 4-bit computer using TTL, it needed a beefy 5V power supply for that! I had fun designing my own instruction set and architecture for fun... but we got to use EPROMS for our logic and 4K x 1 RAM chips I got much of the system working, it had a branch instruction, ALU ADD, and logic like bitwise AND, etc. I can't recall, but I think it had 16 instructions... In the summer 1986 I worked as a tech working on C64's and VIC-20's (I also wrote two VIC-20 games that were sold on the market in 1982). So I just love the "retro" computing bringing back the earlier days in computing history... It's just so cool to see you working on these projects!
The original Colossus code crackers at Bletchley Park during WW2 used a long loop of paper tape in a similar way! They were read in with photocells rather than mechanically, allowing it to run much more quickly. Although I'm kinda disappointed you didn't consider making your own drum storage :)
Hey! If you need a good paper tape reader I have got a teletype model 28-8 bit (same as a model 35) paper tape reader you can use. its made of tougher stuff than the model 33 reader and is rated for a %100 duty cycle!
its good to see progress on this after so long. the combo of paper tape, magnetic tape, and tubes is just cool to look at even if its as power hungry as a super computer while doing basic math lol i know the tape reader needs darkness to work but somewhere you need a peephole so people can see the tape flying along.
You should do a collaboration with youber Mr. Carlson's Lab. He is an absolute electrical engineering genius, and a vacuum tube savant! He could help you get that thing up and running as if it just rolled off the factory floor in the 50's. No joke, that Collab would be a massive hit!!!
Hey I love your videos. I can't say I understand next to anything I'm seeing, but your charisma and enthusiasm on this subject, especially tube computing is something that I admire about your channel. Keep up the great work :)
Oh! You did the looping tape thing! It actually might be a good idea to take the clock from the tape too (otherwise you need some way to sync the tape transport to the cpu). Just let the tape clock it fromt the sprocket hole👍
Hey, damn awesome intro there :) It's a wonderful idea to add some memory, and the quaint paper tape would be a suitable one. All the best at VCF! Overshadow the last one. Knowing I'll see you and CuriousMarc together on livestreams/videos... it fills me with determination. "All your 6AU6 are belong to us."
6:57 till 7:41 - open discrete circuit with a many potential fault points and humungous power demand. That is why computing skyrocketed only when computing logic transformed to chips.
If I may suggest -- perhaps not "Kevin", but how about "Calvin"? Aside from the fact that it will likely have a temperament invoking the eponymous character from so, so many newspaper "Funny Pages" strips, simply because vacuum tubes :P it actually works as a bit of a backronym, and in a way that fits the era of UNIVAC, EDVAC, the WITCH, etc -- C.A.L.V.I.N. -- Computative Arithmetical-Logical Valved Integrator and Numerator Let me know what you think, Mr Usagi :) also, love the pet break at the end of each episode, it's lovely!
I've always wanted to build a vacuum tube computer. The closest I've come is a ring counter based desktop calculator design from 1966 that uses little neon tubes. I've had success with it, but I made the mistake of using vintage capacitors and diodes. They were shite. If I rebuild using modern diodes and caps it'll work just fine. My breadboard version used modern components with the tubes and worked perfectly.
Years ago i helped put a vacuum tube computer in the basement of our high school. Im pretty sure there was a drum type memory. There was a big adding machine type printer with rows and rows of the same number Also i remember there was a plug board to create some programming. Maybe nothing more than a calculator But man there were thousands of tubes. I think it might still be there
You may have answered this already, but a couple of questions: 1. Are you against all solid state devices? For example, what about diode based logic with the tube as the amplifier? 2. For memory, have you considered "dynamic" memory? Use capacitors to store the bits, and a little tube logic for the refresh logic and the address decoding.
For reference, except for you, I've never heard anyone say each number individually. It's always been "emsee fourteen five hundred bee". To me, "youee fourteen five hundred" rolls off the tongue pretty easily.
I think one of the limiting factors for an all tube memory design is financial, i.e. can you afford to pay the electric bill! The memory design just really doesn't scale up. You would need to purchase an industrial size air conditioning system to keep the whole thing cool and that doesn't include the cost of the tubes that will degrade or fail due to the heat.
This is an awesome project. I'd really only thought about tubes recently in amplifiers. It's cool to see them used this way again. One question though. Do you think this will ever support a microservices architecture? 😊
Elsewhere in the comments I replied to @thomashowe855 concerning his idea to use the eighth tube socket as a tester for 6AU6 tubes. I just want to boost that idea here. I think he's hit on a way to make that eighth socket truly useful to your design. It is so crazy and yet so cool that you would build a vacuum tube computer. That crazy/cool factor is what drew me to the channel originally, and I stayed because of your knowledge and boundless enthusiasm for your work. (I need to capture more of that in my own work, I think...) Keep on keepin' on, my brother. Your channel makes TH-cam a better place.
If you loop the paper tape and add seek commands to it you could somewhat simulate jumps - that could make the computer much less limited. And with two you can have subroutines which you will need a lot of. Very slow though. Given this is a 1-bit ALU, there should probably be some dedicated circuitry for string operations - do same operation multiple times somehow streaming input and output data. This way a single punch of paper tape can execute like a dozen operations which would make it slightly less slow. And you still need some kind of scratch memory. I feel like 6 tubes per bit is way too excessive. There should be a better way. After a lot of thinking here is what I might have done: For fast access, intermediate calculation results and flags use existing 2 octets of memory since that already exists. For somewhat slower numeric variable storage - a bunch of ultrasonic delay lines with self-refreshing and timing. They could store a dozen bits each and need same 6 tubes per delay line plus synchronization/refreshing that can service all DL at once If the operations can be timed precisely enough you could use string operations so a memory cell would contain a word of data rather than one bit. Possibly a whole array with circuitry to store and extract exact elements. For even more storage but even slower - 8-track mechanism with very short tape loop and electro-mechanical track selector and auto-cueing circuitry. This will give the computer 8x1xtape-length of storage that should be enough for working memory it may need. If larger permanent storage is needed than a spooled tape with electro-mechanical seeking could be used.
Hmm player piano computer. I seem to remember one of my old profs talking about such things. A competitor to IBM in the 1920~30s used a tape loop to enable subtraction in a tabulator, In the late 1920 some boffin came up with a way to do subtraction on a 80col punch card and IBM when with that.
Seems to me you could still jump if the tape reader can move backwards and forwards, though to jump very much you'd need to support multi bit operands, like a jump offset register (which one of your existing memory registers could be used to support). Jumps are needed for if/then constructs and loops, and your program capability will be pretty lame without them. But it was good to notice how the Bendix did things, with so much stuff driven from the drum -- even the clock itself. Is this something you're now going to support, clock from the paper tape? It wasn't clear from the cursory going over everything that is going to go on.
It's essential to remember that the most important part in any design is what you're going to call it. I don't think they teach that to beginning engineers...
A rotating (serial drum) 2023 edition would be awesome for ram , since it's just 1 bit, shouldn't be that difficult, I'll have to think on this, but I think that would be awesome, btw the idea of using paper tape for ROM, I loved that, very 1950😊😊
You don't need a better name than Kevin. You need a backronym for KEVIN. EV could be Electronic Valve. K could be Kilo- if you thought there would be a thousand of them.
If Kevin ever achieves self-awareness (a couple more valves, sorry tubes should do the trick), 'he' will probably be pleased as Punch with 'his' name. Although as I remember; Punch was never pleased by anything much except walloping poor Judy with a plank of wood so where that old UK saying came from I'm unsure. Any ideas?
Stay with "Kevin" can't change it now
Hear hear!!!
I second this
Definitely on team Kevin.
Me too.
KEVIN!!!!
I watch this series with my bunny Karen. Thanks for hosting the only rabbit friendly vacuum-tube-based retrocomputing channel on TH-cam.
I got my little squirrel friend on my lap watching together attentively as the tubes warm up and flip the bits. :)
I suspect that, worldwide, there could be a market for maybe five UE-1s, but what do I know?
LOL!!! Wonder how many would get that reference.
@@locutusofborg Apparently more than just you! I thought it would be a pretty safe joke in this crowd. (Safe mainly in that the creator of the UE-1 probably would see it and take no offense.... I want the kit version of the UE-1, and the sawhorses to help me put it together!)
@@locutusofborg "Big things have small beginnings." -- Lawrence of Arabia (1962).
@@AlanCanon2222 my wife would be pleased to hear that then :)
I've been working on developing the core theory for building a real sentient Strong AI for a long time. It would be difficult to get it to work on a UE-1 but it would be so much fun..
When we did computer history at school and I learned the "first generation" used thousands of valves, I thought "I'll never ever see one of those working!" To my younger self, it's pretty much a miracle that you've got 2 to play... erm... experiment with. :)
Ehhh... The true "first gen" machines were relay-based. Very slow, but much more reliable and affordable. Vacuum Tubes made computers faster and more practical, but appeared about a decade or two after the first relay based machines.
@@the123king Yeah, Conrad Zuse didn't really get a look in in our syllabus back at school. And Collosus was still secret back then. So for us, back then, things all started with the likes of Eniac and the Manchester MK I.
It's good to hear about Kevin after all that time!
We need to talk about Kevin.
Using a paper tape for the address/program counter reminds me of the Colossus that has a 30MPH tape loop whizzing around all times 😊
I was just thinking the same thing. Tommy Flowers did some remarkable things with that design - like applying his GPO knowledge that valves are (relatively) reliable if you just leave them on all the time to make the Collusus machines using valves when others though it couldn’t be done.
One day I’ll make a pilgrimage to TNMoC to see the replica - and other goodies they have.
@@timothyp8947 I was there before the coof. Do it. The history of Bletchley Park alone is worth it (GCHQ) but TNMoC is the cherry on top. Make sure you go on a day when they have the full tour, not just Colossus!
Interesting! You hit exactly the same problem as the first pioneers of stored program computers (SPC) - memory using just tubes (valves) is too expensive/bulky/energy-hungry! The very first SPC to run was the Manchester ‘Baby’ SSEM that was built as a demonstrator to show how a Williams Tube memory could work in 1948. It had 32 words of 32 bit wide memory (1024 bits). 75 years later and Usagi is on the job!
wow, building 1024 bits of ram in VVL logic just as a proof of concept sounds a bit masochistic...
Usagi, vacuum tube memory is so 1948. All the cool kids use delay lines nowadays.
@@yuvalyeru The Williams Tube memory used spots of light and dark on a CRT as a kind of recirculating delay line memory, but much faster than acoustic delay lines. They were used in one of IBMs first commercial computers too, before they got magnetic core memory to work.
@@Richardincancale ok, that sounds much more sane.
8:59 bravo for quietly including the classic Signetics datasheet. Well worth a freeze frame for anyone not familiar with it.
Idea: You could turn that useless tube (for the heater string) into a testing apparatus for 6au6s!
Hey, that's not a bad idea at all. It would surely speed troubleshooting suspected tube problems with the unit. Good thinking!
@@horusfalcon why yes, and it would have good/bad lights.
@@SeekingTheLoveThatGodMeans7648 yes
Those sockets (or something like them) are used in custom mechanical keyboards. They're incredibly resilient and resistant. The most popular for keyboards are 3305-1
UE-1 is its official name but Kevin is 100% its nickname
I saw the switches in a newer video, couldn't believe my eyes and went to this video. Soviet micro toggle switches (in this case МТ1), which I absolutely love. They are damn reliable. They are still produced and are quite expensive (compared to similar Chinese switches), but in post-Soviet countries you can find used ones much cheaper, they work no worse than new ones. So I also use them in my projects.
This brings back some great memories when I was in college: They only had one lab that I got to play with vacuum tubes. I had another lab where we built a 4-bit computer using TTL, it needed a beefy 5V power supply for that! I had fun designing my own instruction set and architecture for fun... but we got to use EPROMS for our logic and 4K x 1 RAM chips I got much of the system working, it had a branch instruction, ALU ADD, and logic like bitwise AND, etc. I can't recall, but I think it had 16 instructions... In the summer 1986 I worked as a tech working on C64's and VIC-20's (I also wrote two VIC-20 games that were sold on the market in 1982).
So I just love the "retro" computing bringing back the earlier days in computing history... It's just so cool to see you working on these projects!
The original Colossus code crackers at Bletchley Park during WW2 used a long loop of paper tape in a similar way! They were read in with photocells rather than mechanically, allowing it to run much more quickly. Although I'm kinda disappointed you didn't consider making your own drum storage :)
Hey! If you need a good paper tape reader I have got a teletype model 28-8 bit (same as a model 35) paper tape reader you can use. its made of tougher stuff than the model 33 reader and is rated for a %100 duty cycle!
So cool! Tubes fascinate me and mystify me too. I just bought a 7 tube AM radio. Homebrew it uses 7 1T4 tubes and batteries! Great video!!
its good to see progress on this after so long. the combo of paper tape, magnetic tape, and tubes is just cool to look at even if its as power hungry as a super computer while doing basic math lol
i know the tape reader needs darkness to work but somewhere you need a peephole so people can see the tape flying along.
A level of engineering above my head!
You should do a collaboration with youber Mr. Carlson's Lab. He is an absolute electrical engineering genius, and a vacuum tube savant! He could help you get that thing up and running as if it just rolled off the factory floor in the 50's. No joke, that Collab would be a massive hit!!!
Hey I love your videos. I can't say I understand next to anything I'm seeing, but your charisma and enthusiasm on this subject, especially tube computing is something that I admire about your channel. Keep up the great work :)
Yes! I've been waiting for one of this series for a while! Keep up the good work!
The UE-1 Model Kevin VTC, that sounds like a proper name for it.
Kevin-1 is coming along nicely. More valves!
Oh! You did the looping tape thing! It actually might be a good idea to take the clock from the tape too (otherwise you need some way to sync the tape transport to the cpu). Just let the tape clock it fromt the sprocket hole👍
Came for the tube computer, stayed for the bunana!
Hey, damn awesome intro there :)
It's a wonderful idea to add some memory, and the quaint paper tape would be a suitable one.
All the best at VCF! Overshadow the last one. Knowing I'll see you and CuriousMarc together on livestreams/videos... it fills me with determination.
"All your 6AU6 are belong to us."
I'd be just as fascinated if you made it out of transistors, and it would be much easier for you. Man, I envy your knowledge.
My hats off to you and all those vacuum tubes. Guess I will stick with Light Logic and continue with my 4-bit processor build with Leds and CdS cells.
You have a paper tape collection. Wow. So cool.
You hould've called the computer KV-1. Because Kevin!
20:26: I love that you have a supporter whose username is a single underscore.
yes! a new vaccum tube computer video! i love this project!
That is one beautiful bodge... Stuff of dreams
6:57 till 7:41 - open discrete circuit with a many potential fault points and humungous power demand. That is why computing skyrocketed only when computing logic transformed to chips.
That editing with the advertising cracked me up
If I may suggest -- perhaps not "Kevin", but how about "Calvin"? Aside from the fact that it will likely have a temperament invoking the eponymous character from so, so many newspaper "Funny Pages" strips, simply because vacuum tubes :P it actually works as a bit of a backronym, and in a way that fits the era of UNIVAC, EDVAC, the WITCH, etc --
C.A.L.V.I.N. -- Computative Arithmetical-Logical Valved Integrator and Numerator
Let me know what you think, Mr Usagi :) also, love the pet break at the end of each episode, it's lovely!
oh yeah, love it. Lookin good, keep it up!
Ever consider creating a backplane for Kevin? Perhaps making it modular as well? Just a thought; making it more condensed.
Amazing, I barely understand a word of what you are doing (normal for me), but I’m loving this series anyway.
Looks stunning as always!
KEVIN - Knowledgeable Electronic Vacuum Instruction Network
I've always wanted to build a vacuum tube computer. The closest I've come is a ring counter based desktop calculator design from 1966 that uses little neon tubes. I've had success with it, but I made the mistake of using vintage capacitors and diodes. They were shite. If I rebuild using modern diodes and caps it'll work just fine. My breadboard version used modern components with the tubes and worked perfectly.
Years ago i helped put a vacuum tube computer in the basement of our high school. Im pretty sure there was a drum type memory. There was a big adding machine type printer with rows and rows of the same number
Also i remember there was a plug board to create some programming.
Maybe nothing more than a calculator
But man there were thousands of tubes. I think it might still be there
You may have answered this already, but a couple of questions:
1. Are you against all solid state devices? For example, what about diode based logic with the tube as the amplifier?
2. For memory, have you considered "dynamic" memory? Use capacitors to store the bits, and a little tube logic for the refresh logic and the address decoding.
I believe most of his logic is done using solid state diodes.
"I have a UE-1". Ah yes the Kevin, a good model.
look at that
more tube stuff
AWESOME!!!!
There's a glowing gas stream phenomenon in the sky of the Arctic called "Steve". I think it would be funny to name it after the guy that named Steve.
Wartime British "proto computers" used paper tape loops quite happily... well, kinda... stretching and snapping aside.
For reference, except for you, I've never heard anyone say each number individually. It's always been "emsee fourteen five hundred bee". To me, "youee fourteen five hundred" rolls off the tongue pretty easily.
Please do a subtitled (English) intro in Japanese for a future video 😂❤
I think one of the limiting factors for an all tube memory design is financial, i.e. can you afford to pay the electric bill! The memory design just really doesn't scale up. You would need to purchase an industrial size air conditioning system to keep the whole thing cool and that doesn't include the cost of the tubes that will degrade or fail due to the heat.
It look good so far.
If "six bytes" of memory doesn't sound like a lot, on this particular machine you could always call it "forty-eight words" of memory!
If an unused tube is labeled "spare" it is planned and not unnecessary. 😎
This is an awesome project. I'd really only thought about tubes recently in amplifiers. It's cool to see them used this way again. One question though. Do you think this will ever support a microservices architecture? 😊
"Cooling: The 25120 is easily cooled by employment of a six-foot fan, 1/2" from the package..." 😂
Series string heaters make a lot more sense here, you can run 20 tubes off the 120v line.
It _is_ series string. One of the design decisions is he wanted everything to be safe-to-the-touch low voltages.
Elsewhere in the comments I replied to @thomashowe855 concerning his idea to use the eighth tube socket as a tester for 6AU6 tubes. I just want to boost that idea here. I think he's hit on a way to make that eighth socket truly useful to your design.
It is so crazy and yet so cool that you would build a vacuum tube computer. That crazy/cool factor is what drew me to the channel originally, and I stayed because of your knowledge and boundless enthusiasm for your work. (I need to capture more of that in my own work, I think...) Keep on keepin' on, my brother. Your channel makes TH-cam a better place.
Utterly and completely mad! Magnificent though.
i can't believe no one suggested Tubey McTuberson
Kevin looks great.
Marvin… he often referred to his exceptionally large mind albeit he was more down than up ;)
If you loop the paper tape and add seek commands to it you could somewhat simulate jumps - that could make the computer much less limited. And with two you can have subroutines which you will need a lot of. Very slow though.
Given this is a 1-bit ALU, there should probably be some dedicated circuitry for string operations - do same operation multiple times somehow streaming input and output data. This way a single punch of paper tape can execute like a dozen operations which would make it slightly less slow.
And you still need some kind of scratch memory. I feel like 6 tubes per bit is way too excessive. There should be a better way.
After a lot of thinking here is what I might have done:
For fast access, intermediate calculation results and flags use existing 2 octets of memory since that already exists.
For somewhat slower numeric variable storage - a bunch of ultrasonic delay lines with self-refreshing and timing. They could store a dozen bits each and need same 6 tubes per delay line plus synchronization/refreshing that can service all DL at once If the operations can be timed precisely enough you could use string operations so a memory cell would contain a word of data rather than one bit. Possibly a whole array with circuitry to store and extract exact elements.
For even more storage but even slower - 8-track mechanism with very short tape loop and electro-mechanical track selector and auto-cueing circuitry. This will give the computer 8x1xtape-length of storage that should be enough for working memory it may need.
If larger permanent storage is needed than a spooled tape with electro-mechanical seeking could be used.
Hmm player piano computer. I seem to remember one of my old profs talking about such things. A competitor to IBM in the 1920~30s used a tape loop to enable subtraction in a tabulator, In the late 1920 some boffin came up with a way to do subtraction on a 80col punch card and IBM when with that.
That was friggin cool
proposal to name it 1-UE so it translates to 1UP :P
like the new cut
I also vote UEVAC as a name
"I have this terrible pain in all the diode tubes down my left-hand side" - Kevin... Or was that Marvin?
I still think Kevin is a great name. Kevin = Kevin's Electronic Vacuum tube Instruction Negotiator.
Seems to me you could still jump if the tape reader can move backwards and forwards, though to jump very much you'd need to support multi bit operands, like a jump offset register (which one of your existing memory registers could be used to support). Jumps are needed for if/then constructs and loops, and your program capability will be pretty lame without them. But it was good to notice how the Bendix did things, with so much stuff driven from the drum -- even the clock itself. Is this something you're now going to support, clock from the paper tape? It wasn't clear from the cursory going over everything that is going to go on.
That's the style of computer from space patrol!
Kevin-1-Bit! ❤❤❤
Save the out of spec 6AU6 tubes for the heater resistor position. Turn trash into gold!
The name is not merely, "Kevin", rather it is "KEVIN!!!!" aka McCallister.
Kooky Electron Valve Integrator Numeric - K.E.V.I.N.
You're going to need to build a vacuum tube manufacturing machine at this rate.
Actually now that I've written that I kinda want to do that 😅
Just call it the UE-K or UE-K1, keeps the joke alive and still works for the versioning.
I'm fine with calling it 'Kevin.' Kevin is a fine name.
Nice project Love the Rabbit!
I saw you at VCF Southwest!
In the dark the UE1 looks Eastern Orthodox
We need a good backronym for Kevin.
UE-1: Kernel Executive Vacuum Integer Numeric Unit (KEVINU)
I think "Kevin" was a superior name.
It's essential to remember that the most important part in any design is what you're going to call it. I don't think they teach that to beginning engineers...
They say there are only 2 difficult problems in computer science: naming things, cache invalidation, and off-by-one errors.
A rotating (serial drum) 2023 edition would be awesome for ram , since it's just 1 bit, shouldn't be that difficult, I'll have to think on this, but I think that would be awesome, btw the idea of using paper tape for ROM, I loved that, very 1950😊😊
@ 7:29 - do you have a part number/supplier for the hollow pins you are using for the tube contacts on the PCB (or MCB - Milled Circuit Board)?
UE1 - Maybe should have an interface to a delay line type memory using the Litton's drum memory subsystem. Regards, Ken R.
Gooooood Morning from the MorotorCity!
Best show on the BoobTube!
Btw I liked Kevin as name for the machine, but ok UE1 makes sense
Kevin, Kevin, Kevin!!!
If you use a piece of hardwood for the PCB material to sit on while it is being machined might be the way to go.
You don't need a better name than Kevin. You need a backronym for KEVIN. EV could be Electronic Valve. K could be Kilo- if you thought there would be a thousand of them.
Maybe the I could be a 1, for 1-bit.
As many others will say: YAY!!!!! :)
Imagine the size of the machine needed to replicate an AMD Ryzen using vacuum tubes.
A24EVR/KansasFest 2023
If Kevin ever achieves self-awareness (a couple more valves, sorry tubes should do the trick), 'he' will probably be pleased as Punch with 'his' name.
Although as I remember; Punch was never pleased by anything much except walloping poor Judy with a plank of wood so where that old UK saying came from I'm unsure. Any ideas?
Mister Punch was very pleased with himself when he walloped the policeman who came to arrest him for walloping Judy. "What a clever fellow."
Awesome stuff!
UE. Call it Huey.