4:08 Another Forgotten Machines Tshirt sighting! Thank you! What an amazing overview...So glad to be here, and so glad you're doing this and documenting it SO well in video!
As I watched this, I got curious again about your methods of circuit board fabrication, so I went back and tagged your video about how you use a CNC mill to cut these out. I'mma go back and watch that again (it's been a couple of years), and see if I might be able to come up with a dedicated device for cutting circuit boards for myself. I love the "steampunk" look of the UE-1, and the bench power supply you were using for this video (and others, I'm sure), and I'm looking forward to the video where you fire up the whole shebang. Be careful in the workshop and everywhere else, and I'll see you in the next one.
If you are at all worried about that 1R resistor cooking up simple enough to crimp ( not solder, as that will kill it) a 105C one time thermal fuse to some wires, and put in series with the power rail. That will then be attached to the resistor body, simple enough by using a fibreglass sleeve (again heatshrink not usable with 105C melting it before it will shrink) for insulation, and then a simple fabricated clip using the grooves on the resistor to hold to the top. 105C surface temperature, and the thermal fuse will melt and stop operation long before that resistor gets hot enough to burn the wood or PCB anyway. Commonly done on old TV sets with dropper resistors, though there it was often just a spring held with solder, so that if the resistor got hot enough, over 183C at the solder point on the spring, it would soften and the spring would open and break the circuit. Simple, reliable, and even if the resistor got on the way to red heat it would trip before it got hot enough to do more than start to smoke gently as it cooked the dust accumulation on it.
About a year ago I had no tubes, and no tube devices. Always been 'interested'. Anyway, a KG-635 oscilloscope, Jackson 607D HV power supply w/filament supply, grid dip meter, all vacuum tube powered and recapped by me - I now own well over 70 tubes not including the tubes inside my equipment. I also enjoy radios, so I have a old tube CB radio and some ham radio 'pieces'. Working now on a tube tester for 7 pin Soviet tubes. I have 40 6J1P, 15 6J1P-EV, some 6Z4's, OA3's, 6N2P's, and others. I've just started to experiment building circuits with the 120 volt 6J1P's. Anyway - thanks for all the videos I hope to see some more tube videos in the future from you.
6:47 lol! Thank you for incorporating the classic Signetics 25120 WOM chip data sheet. Very much a rarity in IC design, in that it includes a filament supply pin, making it perfect for your valve/tube computer. ;)
I have been on a marathon iteration cycle on my homebrew computer I definitely understand how great seeing the finish line feels 😊 well done! I can't wait to see this finished
UE-1 looks fantastic, David! And you are within shouting distance of having the most steampunk-esque computer on TH-cam. We need to get you a woodgrain keypad with brass keys to complete the look, or something like that. 👍
Hey David! Ever heard of M^2 logic, also called Mickey Mouse logic? It uses diodes and common CMOS chips to build lots of quite complex gates, potentially even oscillators or counters. Maybe it is adaptable to tube circuitry, and could simplify your design! Look it up, if you already haven't! I'm so excited for all this progress you're making! I can't wait to finally see the whole system together!
Ha! Lovely. I wonder if you'll be taking it to VCF again. Lessons learned and hopefully no rapid unscheduled discombobulation this time. If so, I'd suggest sandwiching the boards together with rigid links between them, all electronics inside, protected from any outside objects.
The funny thing is, for most industrial controls, a lot of the work IS single bit. Ladder logic (AKA relay logic) is still to be found in a LOT of machines, even modern ones. Because for some tasks, a single bit is all you need. Granted, modern PLC's aren't using single bit processors, that would be silly, but a lot of the logic is still single bit. So back in the day, i can easily see a single bit processor for industrial automation would be a smart thing to make.
Quoting Bob Cratchit from the Alistair Sim version of A Christmas Carol: "Another triumph!" 😄 BTW, you may have some ppl attempting to contact you via YT because I've occasionally been mentioning you in a FB group for old computer equipment. Whenever someone has some pretty old stuff and wants to find a home for it, you're the direction I send them in, knowing of your connections to at least one computer museum. Hope you don't mind. 😉
Have you considered putting some tinted film on those VFD tubes to give them more contrast, so you can see which indicators are lit more easily? I'm thinking some car window tint film would be perfect, and should be cheap and easy to source.
The resistors only need a heatsink if the are expecting power that would cause them to overheat. The metal is already a heatsink. It's like saying TO-220 package needs a heatsink to function, just because it's designed to have an OPTIONAL heatsink fitted if load requires it. Anyway, that's my rant on it lol. I love how the UE-1 is coming along. Look forward to it being completed and the UE-2 coming along :)
I'm wondering if a design like this should consider full on max current, in case of some worst case failure just so the thing does not catch fire or melt or something?
I really like the look of the curved traces on your PCB boards. Guessing you designed it that way since you're milling the traces rather than etching them, and a curve is going to be faster to mill than a corner because it avoids a stop and start to change directions. My current electronics project is very retro, and I've sourced a bunch of vintage electronic components, so now I'm designing the PCB boards and I'm thinking of etching them myself. Perhaps I might try designing them using a more curved fluid design like they'd have used in the late 70s/early 80s, which would match the era of most of my components. The enclosure I'm also designing will probably have a transparent lid so you can see the vintage goodness, which might also show the curved traces if I use single sided boards like you use.
6:47 in many circles, "write-only-memory" is considered a joke, because supposedly there is no such thing, what would be the point! BUT, here it is, you've proven that it IS real...provide the DATASHEET to prove it, and you're demonstrating the use of it. Totally blows my mind!
Single bit CPUs... The company that I first worked for in the 1970s made a high-performance minicomputer, built from Motorola 10k series ECL, where the primary goal of the architecture was programmable interfacing. It was essentially a single bit machine, although it had 16-bit extensions which were normally included. Data RAM was also customer option. Power consumption was high, due to the ECL, but performance was very respectable for the time at 8 MIPS. The MC14500, of course, was at the other end of the power and performance spectrum.
I think jumps can be implemented on paper tape if the motor can be controlled precisely. Just step forward or backward "n" times (indicated by the jump instruction)
Since those are cathode-followers (current amplifiers) for increased fan-out, I’d check the to see if they can - in fact - provide the required current. Also, since the output voltage was higher than expected, alarm bells should be going off in your head. Either the B+ is higher than you thought, or you’ve got a short somewhere (cathode to grid, maybe?) , or you hooked the DMM to the wrong point, or the switch goes to the wrong point.
About "one bit computer" there is soviet calculator, "Электроника МК-54" (must be some clone of western one). It is programmable, scientific calculator, with polish notation. And internally, it is ONE bit computer- all chips are connected in "circle" only by one data line.
So... thoughts about the final program you'll run? My vote is for that traffic light state machine you showed near the beginning. Some switches representing road-sensors and colored LED outlets, would be a great little demo.
Just a little addition how 1 bit computers work: Let's think about 8 bit computers first. What is the biggest number to process directly in such a computer ? It's 256. The apple 1 had a 8bit processor and it was used for lots of processing/ math situations. It was of course not limited to numbers smaller than 256, that would be really bad. Bigger numbers were split in 8 bit parts and send through the bus. The same is true for a 1bit machine! The big disadvantage is that even for a simple 8bit integer, 8 CPU clocks are needed to transfer the data. It just slows everything down. The MC14500 has 16 instructions, therefore even the most simple command takes 4 cycles to be send.
Your last sentence would be correct if this was a von Neumann architecture, where the instructions go through the data bus. In the MC14500 however, the instructions come through a dedicated 4-bit instruction bus, so it does not need 4 cycles.
Did you cut the other two boards? I remember those being relegated to "if I have time/resources", but they're not mentioned here at all. In any case, it's nice to see this thing again! I can't wait for it to run DOOM!
If only there were some way to miniaturise this, it might just sell. :-) Have you seen the MegaProcessor project ? Same idea, but discrete transistors and lots of LED's
LOL: COOLING The 25120 is easily cooled by the employment of a six foot fan 1/3" from the package. If the device fails you have exceeded the ragings. In such cases, more air is recommended.
I know this is probably way outside the realm of your intentions with this project. But since you did take it to a show at one point, how cool would it be if once it’s tested and working you rebuild it in the Bendix style. Couple backplanes in an open chassis, so we can see all the tubes, they’re protected and it’s somewhat portable
Am I getting this right. The PCB has a fully conductive surface and then you cut breaks in the conductive layer to make separate pathways instead of cutting and then filling traces in a nonconductive layer as done in more modern production methods?
I wonder if you the computer would be fast enough to make a 1-bit Digital to Analog Converter (DAC)? It would be a Delta/Sigma increment/decrement type of DAC. I think I saw a clock frequency of 100 HZ. It that's the max speed, then it probably wouldn't work. But, it's a thought.
as someone who has a lot of knowlage about tube logic. i have a question thats nagging me for some time. would it be possible to use a VFD type transistor logic ... not the fvd itself but with slight modification. or maybe a porpuse build VFD but instad of lighting up pixels it gets a output transistor like? woulnd that decrease the size alot?
Absolutely! I actually use a VFD as half of a flip flop for my memory boards. The entire thing could be built from small VFD triodes, like the 6977, to dramatically reduce size and power draw. But, those are definitely harder to find in large quantities than 6AU6s, and they would present interesting fanout problems. But, if I have 500 of them hanging about, I'd certainly try!
Why not just wire-wrap across the little board interconnect jumpers rather than cutting out little boards? You'd be able to disconnect individual wires if needed for testing and then replace them one by one with the boards if desired afterwards.
Boringly trouble-free episode. I'd say you have one coming!! 😀 Could you make the episodes LLLOOONNNGGGEEERRR? We have to wait a whole week for our next fix.
"The voltage I'm reading is exactly what the theory says it should be, I must have done something wrong." Are you also a C++ programmer by any chance? 😆
Most of us, if we were sent back in time or survived into after a catastrophe type of scenario, would be very useless in trying to recreate let’s say a computer. This knowledge would actually be useful. Still needs a way to reproduce a tube still though.
An RP2040 64 bit chip could be described as a "one bit" computer if both the input and output streams were serial. That does not make any clear sense to me.
I counted 285 valves, (excluding the two 2D21 thyratrons)…I know that thay are not ALL 6AU6’s, but for the purpose of this “back of the envelope argument” I will assume that they are all 6AU6’s….rated at 300mA Heater current at 6.3v. I going to round the total number of valves to 300 just to make the maths a bit easier. Initially I will assume thst ALL the heaters are in paralell, (although I did see the circuit where you show that thay are actually in series groups of four)… So one 6AU6 = 300mA Ten = 3A One Hundred = 30A three Hundred = 90A! Yikes! With groups of four in series, thst reduces the current by a factor of four, so about 22.5A…snd take off sn Amp for the fifteen that I assumed were there but aren’t….so 21.5A! That’s still a pretty big ask from your 24v D.C. supply…I assume it is regulated…poor regulator! Generally, indirectly heated valves like 6AU6’s are run of A.C. Supplies, just another (pretty solid) winding on the power transformer…or in this case I’d probably consider a separate heater transformer. When I was a kid, I went to Catholic Mass…and at the church they had this organ that had one 12AX7 based oscillator for every one of the 88 notes on the keyboard…there was a chassis deck in there four foot square and just carpeted with 12AX7’s! The amplified and power supply was down in the bottom, ther was an 0D3 regulator, a few 12AU7’s, a few GZ34 rectifiers and two EL34/6CA7’s in push pull, class AB as the output to the speaker….but, there were three iron cored things there, one would be the power transformer, one was probably a filter choke snd I reckon the remaining one was just a dedicated heater transformer for all those 12AX7’s. The two secondary wires coming out of it were bloody thick! Cheers, The Globe Collector,p Tasmania
This channel is worth watching just for the sheer excitement over a vacuum tube warming up.
I mean, they're just so cool. They essentially use fire to do math, that's awesome!
You are the best at keeping ancient technology alive.
Thank you so much!
You've done so much in 2023 it's amazing you've found any time for this on top. It's great to see it almost complete, well done.
And I feel like I'm sitting still looking at some of the amazing things others in the community are building!
I am so looking forward to seeing it fully functional.
That is the cutest little HP DMM!
4:08 Another Forgotten Machines Tshirt sighting! Thank you! What an amazing overview...So glad to be here, and so glad you're doing this and documenting it SO well in video!
You guys are awesome too ❤
Gotta represent!
Only need to build the paper tape reader...NO PROBLEM 😂
I have a handful of ideas I think will work... I hope, haha.
Bletchley Park can supply drawings for a high speed paper tape reader…
As I watched this, I got curious again about your methods of circuit board fabrication, so I went back and tagged your video about how you use a CNC mill to cut these out. I'mma go back and watch that again (it's been a couple of years), and see if I might be able to come up with a dedicated device for cutting circuit boards for myself.
I love the "steampunk" look of the UE-1, and the bench power supply you were using for this video (and others, I'm sure), and I'm looking forward to the video where you fire up the whole shebang. Be careful in the workshop and everywhere else, and I'll see you in the next one.
If you are at all worried about that 1R resistor cooking up simple enough to crimp ( not solder, as that will kill it) a 105C one time thermal fuse to some wires, and put in series with the power rail. That will then be attached to the resistor body, simple enough by using a fibreglass sleeve (again heatshrink not usable with 105C melting it before it will shrink) for insulation, and then a simple fabricated clip using the grooves on the resistor to hold to the top. 105C surface temperature, and the thermal fuse will melt and stop operation long before that resistor gets hot enough to burn the wood or PCB anyway.
Commonly done on old TV sets with dropper resistors, though there it was often just a spring held with solder, so that if the resistor got hot enough, over 183C at the solder point on the spring, it would soften and the spring would open and break the circuit. Simple, reliable, and even if the resistor got on the way to red heat it would trip before it got hot enough to do more than start to smoke gently as it cooked the dust accumulation on it.
Or use a couple of light bulbs as ballast. Which obviously give a convenient visual indication of the load flowing through them.
It's taking shape nicely. Can't wait to see some paper tape going into it.
About a year ago I had no tubes, and no tube devices. Always been 'interested'. Anyway, a KG-635 oscilloscope, Jackson 607D HV power supply w/filament supply, grid dip meter, all vacuum tube powered and recapped by me - I now own well over 70 tubes not including the tubes inside my equipment. I also enjoy radios, so I have a old tube CB radio and some ham radio 'pieces'. Working now on a tube tester for 7 pin Soviet tubes. I have 40 6J1P, 15 6J1P-EV, some 6Z4's, OA3's, 6N2P's, and others. I've just started to experiment building circuits with the 120 volt 6J1P's. Anyway - thanks for all the videos I hope to see some more tube videos in the future from you.
Incredible project, looking forward to it getting over the finish line.
6:47 lol! Thank you for incorporating the classic Signetics 25120 WOM chip data sheet. Very much a rarity in IC design, in that it includes a filament supply pin, making it perfect for your valve/tube computer. ;)
I have been on a marathon iteration cycle on my homebrew computer I definitely understand how great seeing the finish line feels 😊 well done! I can't wait to see this finished
UE-1 looks fantastic, David! And you are within shouting distance of having the most steampunk-esque computer on TH-cam. We need to get you a woodgrain keypad with brass keys to complete the look, or something like that. 👍
Hey David! Ever heard of M^2 logic, also called Mickey Mouse logic? It uses diodes and common CMOS chips to build lots of quite complex gates, potentially even oscillators or counters. Maybe it is adaptable to tube circuitry, and could simplify your design!
Look it up, if you already haven't!
I'm so excited for all this progress you're making! I can't wait to finally see the whole system together!
This was interesting to watch as the VTC comes together. I can’t wait to see how you make the paper tape subsystem
Nice to see that old WOM chip!
It's looking awesome. Just the simple case of paper tape. 👍 lol
Thanks for the Signetics 25120 data sheet... 😂 That took me down an internet rabbit hole that I wasn't planning on... 😊
Ha! Lovely. I wonder if you'll be taking it to VCF again. Lessons learned and hopefully no rapid unscheduled discombobulation this time. If so, I'd suggest sandwiching the boards together with rigid links between them, all electronics inside, protected from any outside objects.
The funny thing is, for most industrial controls, a lot of the work IS single bit. Ladder logic (AKA relay logic) is still to be found in a LOT of machines, even modern ones. Because for some tasks, a single bit is all you need. Granted, modern PLC's aren't using single bit processors, that would be silly, but a lot of the logic is still single bit.
So back in the day, i can easily see a single bit processor for industrial automation would be a smart thing to make.
It's a Turing Machine! I will be fascinated to see what kind of programs you can write and run on your little system!
Quoting Bob Cratchit from the Alistair Sim version of A Christmas Carol: "Another triumph!" 😄
BTW, you may have some ppl attempting to contact you via YT because I've occasionally been mentioning you in a FB group for old computer equipment. Whenever someone has some pretty old stuff and wants to find a home for it, you're the direction I send them in, knowing of your connections to at least one computer museum. Hope you don't mind. 😉
Just a though about the design of the tape reader: Use a wind up motor so you don't need extra current.
But can it run Doom? 🤭 Sorry, had to do it!
You are a pleasure to watch. Always positive and never giving up.
Excited to see some paper tape in action!! I'm not sure why I like it so much but paper tape is awesome!! =P
Have you considered putting some tinted film on those VFD tubes to give them more contrast, so you can see which indicators are lit more easily? I'm thinking some car window tint film would be perfect, and should be cheap and easy to source.
The resistors only need a heatsink if the are expecting power that would cause them to overheat. The metal is already a heatsink. It's like saying TO-220 package needs a heatsink to function, just because it's designed to have an OPTIONAL heatsink fitted if load requires it.
Anyway, that's my rant on it lol. I love how the UE-1 is coming along. Look forward to it being completed and the UE-2 coming along :)
I'm wondering if a design like this should consider full on max current, in case of some worst case failure just so the thing does not catch fire or melt or something?
I look forward to see it completed, Good luck!
I really like the look of the curved traces on your PCB boards. Guessing you designed it that way since you're milling the traces rather than etching them, and a curve is going to be faster to mill than a corner because it avoids a stop and start to change directions.
My current electronics project is very retro, and I've sourced a bunch of vintage electronic components, so now I'm designing the PCB boards and I'm thinking of etching them myself. Perhaps I might try designing them using a more curved fluid design like they'd have used in the late 70s/early 80s, which would match the era of most of my components. The enclosure I'm also designing will probably have a transparent lid so you can see the vintage goodness, which might also show the curved traces if I use single sided boards like you use.
6:47 in many circles, "write-only-memory" is considered a joke, because supposedly there is no such thing, what would be the point! BUT, here it is, you've proven that it IS real...provide the DATASHEET to prove it, and you're demonstrating the use of it. Totally blows my mind!
The Signetics 25120 (9046 WOM) was released on April 1st 1973 using the Signetics Extra Secret (S.E.X.) process.
Of course it is real, how else would computers implement their /dev/null device?
You need to read that Signetics 25120 datasheet and look up its history (early 1970s stuff). It's a good read.
Single bit CPUs... The company that I first worked for in the 1970s made a high-performance minicomputer, built from Motorola 10k series ECL, where the primary goal of the architecture was programmable interfacing. It was essentially a single bit machine, although it had 16-bit extensions which were normally included. Data RAM was also customer option. Power consumption was high, due to the ECL, but performance was very respectable for the time at 8 MIPS. The MC14500, of course, was at the other end of the power and performance spectrum.
Did I see the (in)famous Signetics 25120 datasheet go past? I would love to see this computer used in a post mortem weapon system!!
I imagine the power-up of the full thing like the Christmas light scene in Christmas Vacation. Nuclear auxiliary activated!
im so excited for the paper tape reader because i have no idea how that works and i would love to see that built
I think jumps can be implemented on paper tape if the motor can be controlled precisely. Just step forward or backward "n" times (indicated by the jump instruction)
9:05 You forgot carrier pigeon.😂
Great job!
It would be interesting if the reader was mechanical instead of optical with diodes.
Since those are cathode-followers (current amplifiers) for increased fan-out, I’d check the to see if they can - in fact - provide the required current. Also, since the output voltage was higher than expected, alarm bells should be going off in your head. Either the B+ is higher than you thought, or you’ve got a short somewhere (cathode to grid, maybe?) , or you hooked the DMM to the wrong point, or the switch goes to the wrong point.
can't wait for the tape reader
nice work
YAY I was waiting for this!!!
About "one bit computer" there is soviet calculator, "Электроника МК-54" (must be some clone of western one). It is programmable, scientific calculator, with polish notation. And internally, it is ONE bit computer- all chips are connected in "circle" only by one data line.
"MY goal was to... And WE did not make that deadline"
You sound like my manager XD.
So... thoughts about the final program you'll run? My vote is for that traffic light state machine you showed near the beginning. Some switches representing road-sensors and colored LED outlets, would be a great little demo.
I love this idea!
Having a little traffic diorama next to the UE1 would be an awesome demo!
Been a minute since we looked at the tube computer
Hello brother. Death, I had forgotten.
Q: How do you create your circuit boards? Do you etch your own?
Yeah, there is a reason why you start looking like Dr. Reinhard from the Cygnus again!😮😉
Excellent, I'm really looking forward to how you handle the paper tape reader. Are you taking some tips from the Bendix g15 ?
👍
Well it's got a series bus !
You should just set up a teletype on a modem that we can dial into instead of dealing with proprietary crap like discord.
3d print some bending templates for those wire jumpers
What if any of the thyratrons get their filament open? Fire?
26:46 A boy named Sue. Meow!
Just a little addition how 1 bit computers work: Let's think about 8 bit computers first. What is the biggest number to process directly in such a computer ? It's 256. The apple 1 had a 8bit processor and it was used for lots of processing/ math situations. It was of course not limited to numbers smaller than 256, that would be really bad. Bigger numbers were split in 8 bit parts and send through the bus.
The same is true for a 1bit machine! The big disadvantage is that even for a simple 8bit integer, 8 CPU clocks are needed to transfer the data. It just slows everything down. The MC14500 has 16 instructions, therefore even the most simple command takes 4 cycles to be send.
Your last sentence would be correct if this was a von Neumann architecture, where the instructions go through the data bus. In the MC14500 however, the instructions come through a dedicated 4-bit instruction bus, so it does not need 4 cycles.
Did you cut the other two boards? I remember those being relegated to "if I have time/resources", but they're not mentioned here at all.
In any case, it's nice to see this thing again! I can't wait for it to run DOOM!
If only there were some way to miniaturise this, it might just sell. :-)
Have you seen the MegaProcessor project ? Same idea, but discrete transistors and lots of LED's
I'm just waiting for it to say 'Hello Dave' when turned on
Well, once you’ve repaired the paper tape reader for the Bendix, you’ll be better equipped to build one from scratch, yes?
Why no tube sockets? Isn’t it a major pain if you get a bad tube?
They are all socketted, don't worry. See previous episodes.
Yes. You don't solder tubes anyway since the sudden heat would damage the metal to glass seal.
LOL: COOLING
The 25120 is easily cooled by the employment of a six foot fan 1/3" from the package. If the device fails you have exceeded the ragings. In such cases, more air is recommended.
My current project had a target completion date of Christmas 2022!!!!! :)
Hey david can you turn a cidco mail station into a terminal using the printer parallel port?
I know this is probably way outside the realm of your intentions with this project. But since you did take it to a show at one point, how cool would it be if once it’s tested and working you rebuild it in the Bendix style.
Couple backplanes in an open chassis, so we can see all the tubes, they’re protected and it’s somewhat portable
Am I getting this right. The PCB has a fully conductive surface and then you cut breaks in the conductive layer to make separate pathways instead of cutting and then filling traces in a nonconductive layer as done in more modern production methods?
Ah, another explanation applicable to telephone switching methods. Cool
I wonder if you the computer would be fast enough to make a 1-bit Digital to Analog Converter (DAC)? It would be a Delta/Sigma increment/decrement type of DAC. I think I saw a clock frequency of 100 HZ. It that's the max speed, then it probably wouldn't work. But, it's a thought.
I would totaly pop the cover off thaat relay so you see the coil!
as someone who has a lot of knowlage about tube logic. i have a question thats nagging me for some time. would it be possible to use a VFD type transistor logic ... not the fvd itself but with slight modification. or maybe a porpuse build VFD but instad of lighting up pixels it gets a output transistor like? woulnd that decrease the size alot?
Absolutely!
I actually use a VFD as half of a flip flop for my memory boards. The entire thing could be built from small VFD triodes, like the 6977, to dramatically reduce size and power draw. But, those are definitely harder to find in large quantities than 6AU6s, and they would present interesting fanout problems. But, if I have 500 of them hanging about, I'd certainly try!
電気ウサギさん、すごいですにゃ～
Why not just wire-wrap across the little board interconnect jumpers rather than cutting out little boards? You'd be able to disconnect individual wires if needed for testing and then replace them one by one with the boards if desired afterwards.
Boringly trouble-free episode. I'd say you have one coming!! 😀 Could you make the episodes LLLOOONNNGGGEEERRR? We have to wait a whole week for our next fix.
It happened again. How do you do this?!
23:18 resistors are crying lol
"The voltage I'm reading is exactly what the theory says it should be, I must have done something wrong."
Are you also a C++ programmer by any chance? 😆
If you set out to livestream an empty piece of wall while explaining what you were going to do with it, I would probably hang on (pun intended) 😀
Most of us, if we were sent back in time or survived into after a catastrophe type of scenario, would be very useless in trying to recreate let’s say a computer.
This knowledge would actually be useful. Still needs a way to reproduce a tube still though.
If anyone wants a challenge, drink a shot everytime you hear the words "Cathode follower buffer" in the video
I'm pausing for a minute to see if you noticed the jokes on the 25120 data sheet. I don't think I've seen anything like that before!
This guy playing redstone computers IRL
Does it run Doom?
You know, with all that unused ceiling space, it seems like the UE1 needs a KB of RAM installed......
This is why it's only version 1...the next will have 4KB
So does the cat thank you for the gravel in his gut and the spit in his eye?
So much so that if he ever has a kitten, he might just name him... Bill! Or George! Any dang thing but Sue!
23:56 but "Have You Tried Turning It Off And On Again?"
th-cam.com/video/rksCTVFtjM4/w-d-xo.html
An RP2040 64 bit chip could be described as a "one bit" computer if both the input and output streams were serial.
That does not make any clear sense to me.
Next video: Today I try to recreate ENIAC…….
Shoulda used shellac. Poly coat looks like a cheap vinyl wrap. 😢 Where's the love.
I counted 285 valves, (excluding the two 2D21 thyratrons)…I know that thay are not ALL 6AU6’s, but for the purpose of this “back of the envelope argument” I will assume that they are all 6AU6’s….rated at 300mA Heater current at 6.3v. I going to round the total number of valves to 300 just to make the maths a bit easier.
Initially I will assume thst ALL the heaters are in paralell, (although I did see the circuit where you show that thay are actually in series groups of four)…
So one 6AU6 = 300mA
Ten = 3A
One Hundred = 30A
three Hundred = 90A! Yikes!
With groups of four in series, thst reduces the current by a factor of four, so about 22.5A…snd take off sn Amp for the fifteen that I assumed were there but aren’t….so 21.5A!
That’s still a pretty big ask from your 24v D.C. supply…I assume it is regulated…poor regulator!
Generally, indirectly heated valves like 6AU6’s are run of A.C. Supplies, just another (pretty solid) winding on the power transformer…or in this case I’d probably consider a separate heater transformer.
When I was a kid, I went to Catholic Mass…and at the church they had this organ that had one 12AX7 based oscillator for every one of the 88 notes on the keyboard…there was a chassis deck in there four foot square and just carpeted with 12AX7’s! The amplified and power supply was down in the bottom, ther was an 0D3 regulator, a few 12AU7’s, a few GZ34 rectifiers and two EL34/6CA7’s in push pull, class AB as the output to the speaker….but, there were three iron cored things there, one would be the power transformer, one was probably a filter choke snd I reckon the remaining one was just a dedicated heater transformer for all those 12AX7’s. The two secondary wires coming out of it were bloody thick!
Cheers, The Globe Collector,p Tasmania
Hmm... time to start on that DOOM port I guess?
I could run model rail lights? Lol
But can it run
Mr Spock's brain.exe?
You know, if you build a network switch out of rubes, you could claim the internet really is a bunch of tubes...
Hi!
periodic video and thoisoi 2 and some other youtube channel
Ballocks can't get discord to do email verification.
Oi se tu traduziu esse comentário responde 🧭
reply
'X (formerly Twitter)' is properly pronounced "the river of bile and internet nonsense".
Yes but can it run DOOM?