OK, that little computer of yours is great and everything but the rabbit falling over when you’re rubbing him is the best thing I’ve seen in a long time! 😂
Wow. This is one of the best channels I have come across in the last decade of TH-cam’s quality decline. Purely fascinating getting to watch your process!
Something my dad taught me. When counter boring, run the mill arbour down to the stop, crank the table up till the cutter touches the part. Let the arbour back up off the stop, and crank up the table the desired depth of the counter bore. And then use the arbour to drill the hole, when you reach the stop you are at the desired depth
@@UsagiElectric it is standard good practice to do it that way. Get in to the habit of following good practice even if you think you don't need to, and it becomes second nature
Ah, this reminds me of the Data General Nova computer we had decades ago at FAU. It had an ASR33 tty with a (slow) paper tape reader as the only input device, and my assignment one weekend was to regen the operating system to include a new device (I forget if it was a high speed paper tape reader or maybe a cassette tape reader). To do that we had to feed about a six inch diameter reel of paper tape into it with the operating system on it. What fun! A single tape reader glitch at any point in the process and we had to start from scratch. After many tries, the final successful attempt consisted of hunching over the tty for several hours doing strain relief on the paper tape by hand so there was zero tension on the tape as it fed into the reader 🙂
Definitely necessity is the mother of all inventions and you have a very imaginative mind!! Ready to watch not only your tube computer running but also all the historic computer pieces that you have in your lab. I am very impress of your videos. Thanks for always guide us through your work in a manner that even a child can understand. Great work!
Soldering for a mechanical joint (like on your gear) is different from electronic soldering. You should heat the shalf and press the solder into the heat only, it should flow almost like a liquid, penetrating the gaps between the shaft and the motor. Also be careful with the parts that contact the paper, so that it isn't damaged or ripped over time while running through the machine over and over. Because, as you point out, normal tape readers read the tape once and were done but yours will run continuously while in operation. Also make sure your lamp isn't going to yellow or weaken the tape, if it's going to be continually flowing.
As someone who subscribed for the UE1, and pre-IC technology in general, this announcement is great for me in particular! Can't wait for Vacuum Tube Computer or Bust!
Hi, without at least one bearing (two much advised - this size can easily be found in scale modeling shops) on the vertical axle of the traction rotor, you will rapidly run into a problem of wearing its aluminum guide. Another thing is the rotor itself, it would have been better if you've used something standardized, like a cassette reader rubber tensioner, to allow an easy maintenance - another solution would have been to use a sewing machine mechanism with the paper band center holes.
Yeah, that shaft is going to eat up the aluminum in no time, especially with the tensioned roller pushing against it. He could have gone with a stepper motor and mount the rubber directly onto its shaft. Any medium sized stepper motor already has two bearings and it would have given him precise position control as an added benefit.
@@vatterger Cheap, but may not enough cheap (can't beat the motor he's using) - but I agree, as this kind of stepper motor is specially made to resist axial pressure (not to mention that with a MCU you can easily program a short stop on each byte if the photo-diodes are a bit lazy).
Also, you can use a fiberglass-brush pen, or a hard-rubber pencil eraser to clean the commutator contacts; Scotch-brite pads may be too abrasive and leave scratches in the soft copper; this’ll just attract more dirt, crud, fine particles of graphite from the brushes, and oxidation - not good. 😢 Also, use a toothpick to clean out the crud in the gaps between the commutator segments; little trick I picked up from SMT Mainline (YT channel)!
Just a little FYI: the term “hard rubber” actually means rubber that is polymerized to the point of being rigid. (Think bowling ball hardness.) 100% agree on the recommendation to use a natural rubber eraser for cleaning contacts. I always have one of those woodcase eraser pencils (the kind used as typewriter erasers in bygone times). I use the ones with ink eraser since it’s just a bit more abrasive than the ones for pencil, but still gentler than the fiberglass pen (which I also love).
I can't speak for everyone else, but I have wanted to see them functional for a long while now, I understand that it's difficult to reverse engineer everything and find parts and information period etc but I'm perfectly ok with a three hour video of you solving multiple issues at once to get things done, or even just making three one hour long videos. I appreciate your time and effort!
This is so cool ! I am glad someone is preserving this old technology and making it accessible to all for free on TH-cam. Thank you Usagi Electric! Good luck with the circuity for the photodiode array, I know sometimes photodiodes can be tricky if they are in photovoltaic mode or reverse current mode and a trans impedance OP Amp circuity needs to be used
Oh, hey a model train motor! Those were also widely used in scale model trains and I've repaired lots of those. They were usually made by Pittmann. Pretty bullet proof. Although it's not always necessary, you can also replace the magnet with Neodymium magnets and the motor will run cooler and have more torque.
Personally, I don't care what you work on, or when you work on it. It's all interesting to me. If it matters, I have no concern about when or how you do anything. I just enjoy the videos. No justification needed on my end.
For people complaining that you're using photo diodes for your reader, the majority of the logic in the vacuum tube computer is diode logic. Same with the Bendix. That's how it was done, back in the day. Semiconductor diodes were available, and were used extensively. The logic is done by diodes, and the vacuum tubes are used to amplify and invert the signals. The same technique was used when transistors became available. Early transistor computers were diode logic with transistors for amplification and inversion. AFAIK, transistors didn't start being used for everything till IC made it easy to put a few hundred to a few million transistors onto a single wafer. And even then, i bet there are still quite a few diodes on IC's, even today.
I know this episode (and probably the next few) aren't Centurion related but I have to say I have absolutely loved watching your entire Centurion build series, I've probably watched the entire video set 6 or 7 times now, and you have sparked a big interest in vintage computers in me. I don't have the space (or sadly, time) to even attempt what you've done, but maybe one day I can make it to a VCF East and finally get hands on with that amazing blue machine. Love your work, love your videos. Please keep going, you do have a lot of fans out here, myself included. Fan who shares your first name.
Just random thought, instead of a 'bucket' to hold the loop, why not a spool like used in old 8-track cassette tapes. pulls from the inside continuously, making the reel turn and takeup on the outside. It does cause the tape to slide over itself which MIGHT be a problem for paper tape, but worked really well back in the day. A loose pile in a 'bucket' might easily get knarled and 'knotted'/ twisted causing all sorts of feed problems. Just a thought though... Always enjoy your adventures and jealous of the manufacturing resources at your disposal. Access to CAD tools, machine shop... get to dream up and make all sorts of things.
Well, when you said that just two computer projects are the ones still to be continued any time soon, I was like: okay, now just sniffing for any notification from David!
Just a thought - but I'd fit oilite bearings - better for friction, wear and fit than a drilled hole in aluminium ! They come in all sizes so you're bound to find ones that will fit !
For adjusting the tension of the strip, you could get some inspiration from old film cameras and projectors, which used something called the "Latham Loop". It acts as a buffer for any vibration or variation in tension while pulling the strip through!
"All your 6AU6 are belong to us." This project is a thing of beauty and a joy for ever. I'm deeply impressed by your brother's skills - I'd never have expected that he'd be a great metalworker with his own machine shop! He did awesome work with that tape drive motor mount. I'm sure Quinn Dunki (Blondihacks) would be impressed too! Watching the reader pull the tape through fills you with determination.
I will also recommend if you want to screw steel screws into aluminum, a good technique is to oversize thread the holes in the aluminum, then put in helicoils. Otherwise your steel screws might gall the aluminum, making the screws either irremovable, or ruining the threads in the aluminum. Both are bad. Having the helicoil gall into the aluminum doesn't matter, as you won't be screwing and unscrewing that.
Wow. It's amazing how far you get just talking about a problem with other people. You find solutions in the most unexpected places! (btw; for pedants it's chAmping at the bit, not chomping).
"I hope you like vacuum tube computers" I do now because of you. :D I was watching 8 bit repair guys until I stumbled upon you some time ago. You're taking computer repair to a whole new level. (You get it? Voltage level?) Carry on.
I hope you don't have a problem with continuous speed, Because usually tape readers are step motors, they move the tape hole by hole like a film projector moving frames one by one to allow the necessary exposition time, Not sure how much the exposition time for those diodes to register a bit, but I hope it will work out with continuous speed.
Hey, if you want to stay vintage-era and not use LEDs, I might recommend an “exciter lamp” from a 35mm projector (these were used for optical sound pickup) as your light bulb. These often have a single linear filament that might be helpful!
Osram/Sylvania BULB Part # 4B10021517OSR 30 day replacement if defective $1.86 From advanced auto parts. This is a 12 volt festoon based bulb similar to the bulb used in the Univac 1532 paper tape reader .
clock escapment mechanism could be operated by cpu clock by a single flip flop and push tape one byte at a time every 1, 2 or multiple of 2 clock pulses basically pengelum made out of 2 latches pushing tape one step at the time
Be very, very careful with what lamps you use to trigger the photo diodes. Color, and temperature of emitted light is VERY VERY VERY critical to high speed running. Check specs on your diodes. You can actually sort of "overclock" phtoto readers by very carefully giving the diode a consistent "sweet spot" from the lamp source. I knew a motorola engineer that explained to me they would go through thousand of lamps and build pretty hefty cap banks to avoid any dip in light. He said everything became easier towards end of product life when they actually used lab microscope lamps as emitters. The rabbit hole of lamp specs is deep, and almost all are still made. Very excited to see where this project goes!
I found that you have to have a couple projects going at once. Inevitably you are awaiting parts for one or the other. So I end up switching back and forth. But in the middle of that my desoldering station broke and is awaiting parts. So now there is 3 projects. So there is always a strangeness factor that comes into play too.
Time for a conversation with Curious Marc about an analogue feedback mechanism to create a frequency to voltage circuit, enabling automatic speed calibration?
When threading with a hand tap, please break the chips now and then. You're just asking for trouble if you ram it in like a regular screw. My dad taught me to do 2-3 rotations forward, then half a turn back to break the chips.
Vacuum tubes, yes. I wonder how you can handle data load, perhaps using index hole photodiode state change? Bendix G15 is a beauty that deserves all your attention.
He will need a motor driver that does step by step ratcheting control due to not using a stepper motor with just right gearing instead. he said he will do direct read off the tape to the CPU, not buffering. It means he needs something, that indepently and accurately drives the motor only so long that next ratchet hole is visible and stops until the read and execution is done and the CPU gives order to step to next instruction. Which to me is weird he didn't mention. Since I don't see that as trivial. He needs atleast few logic gates to do "start running, ignore we are on a hole (the previous current instruction, we want the next one), latch that we moved off that old index hole (reference to compare against for "does light through index whole causing motor stop or not"), keep going, sense that we are again on a hole and stop (because stop on index hole is latched), set signal for the CPU, that next memory position is ready to read". That is non trivial. Not to mention then probably a lot of fine tuning. Since that is non stepper motor. So he has to tune in "how fast can the motor be driven, so that the driving electronics have time to stop the tape dead on the middle so everything aligns and good read happens" and at which level of the index hole lighting up to trip the stopping. Too low and it stops as soon as even the edge starts to come in, meaning still probably out of alingment. Too high and it probably then actually stops after the central alignment, since trip to stop only comes at the central alignment and mechanical things aren't instant. Given such precision timing, I think it has to be pure hardware and then stuff like pot resistors to calibrate the driving. So that reader will need couple tubes itself and maybe some diode logic or something.
Sweet i've been following all your projects but I get most excited for your tube computer projects I've get a big smile on my face for each stage of the bendex start up. I't would only be better if you had to hop on an exercise bike as one of the stages
Electronic solder does not stick well to steel (which is what I assume your motor shaft is made of), so that gear may come loose. Suggest cyanoacrylate or Loctite as an alternative. Nice job, though. Tape may feed better if you rotate the whole system 90 degrees, so that the tape is flat and not trying to twist. See the DEC high speed paper tape reader (although I think that used the feed hole to move the tape) -- I like your method better -- doesn't stretch the feed hole.
The little motors were also used in early HO scale train engines, though those used worm gears on the motor rather than pinion gears. The problems with those motors are; 1) they run too fast for most applications, and 2) the commutators often need cleaning due to arcing, and carbon build-up from the motor brushes. Also, the bearings need a small drop of light machine oil every now and then.
Great stuff ! I can just picture it; The tape containing the code for : Println: "HELLORLD" flying past those photo-diodes at great speed....::...:.:.:::...::.....:
One change I would think of would be to cut the top of the readers so the tape can slide down rather than having to be threaded. This would require some sort of adjustable cap or clip to keep the tape from escaping though that slit. Why? Because when you get down to writting programs it will be easier if the tape is a continuous loop. Otherwise you'll have to 'auto rewind' when you hit end of tape, or cut and re-splice the tape every time you switch programs. Also taking a clue from some of earliest magnets drum memories, a 'jump' forward and backwards is just the same instruction with different values
It's not just me! It's not just me! Hugging a 1950s small drum computer! In my case, when I saw the (non-functional) IBM 650 in the Smithsonian, in 1992. My wife looked on affectionately. (Her choice of adverb) My bachelor's thesis at Princeton in 1987 was all about the 650, including a mostly functioning simulator, written using VAX assembly language for the decimal arithmetic, and the CURSES package for screen handling, making the blinkenlights work in full screen text. I didn't know if I'd ever get to see a real one. I've seen the video of a working restored 650 in Germany from about 20 years ago. .
Hellave! (See what I did here? 😬) As for all your planning going elsewhere remember, what Einstein (allegedly) said: "Planning replaces random with error!" 😊 So just keep on doing the great work - I guess everybody here simply enjoys it and loves your way of publishing it. I certainly do 🤗 So thanks 😊 Axel
Firebottles! Real electronics light and heat the room! (Or maybe I'm just punchy because I've been doing SMT practice boards with my shaky hands.) WD-40 (in the pump sprayer) works well for cutting aluminum. You missed an opportunity for a terrible pun with that unholy rectangle in the aluminum plate!
In case you care, At 16:06 you say skill saw but show a jig saw. Skill saw is normally a synonym of Circular Saws. Not sure if that's a regional thing or just a mouth-typo.
![[LIVE] : ONE ลุมพินี 88 | คู่เอก "ป้อมเพชร vs อัสลามจอน"](http://i.ytimg.com/vi/1ZqTVVbMgbo/mqdefault.jpg)
OK, that little computer of yours is great and everything but the rabbit falling over when you’re rubbing him is the best thing I’ve seen in a long time! 😂
it's a sign he feels safe and content 🥰
She does that about twice a night, but it's always so hard to catch it on camera, I got lucky this time!
@@UsagiElectrici am fascinated that you can look 35 or 55 all depending on how you shave and cut your hair 😅
@@AB0BA_69 I'm glad I'm not the only on who's noticed that! 😁
A bunny going ElFlopso is a thing of beauty.
I smashed like the button when you said all other projects were on hold until both computers were doing something exciting!
Completely off-topic but I made the drawing template that you used in FreeCAD! I hope you found it useful!
If he used the drawing template in this video, it can't be _completely_ off topic!
Thanks for sharing your effort and helping the project!
I haven't seen one of those motors for decades. It's certainly brought back some memories!
It was a brilliant idea to use one (thanks Chris!), and for something designed in the 1960s, they're surprisingly good!
When we were kids in the 90s me and my brother used to be part of a local club that raced those things. Fun times.
Wow. This is one of the best channels I have come across in the last decade of TH-cam’s quality decline. Purely fascinating getting to watch your process!
Thank you so much!
Something my dad taught me. When counter boring, run the mill arbour down to the stop, crank the table up till the cutter touches the part. Let the arbour back up off the stop, and crank up the table the desired depth of the counter bore. And then use the arbour to drill the hole, when you reach the stop you are at the desired depth
That's a great idea if it needs to be super precise. Thankfully, I just needed to be close enough since it's mounting screws.
@@UsagiElectric it is standard good practice to do it that way. Get in to the habit of following good practice even if you think you don't need to, and it becomes second nature
OH YEAH BABY!!!! Vacuum tube computers till the end of the year!!!! Is it christmas already?????
Thank you David, you absolute legend!
David, use glass tubing for your guides - hard wearing and low friction
Or nylon
@@abcpea Would Bakelite, not be more period correct ;-)
Just for the fun of trying (depending on paper tape length) .... an old wind up mechanical clock mechanism would make for a beautiful tape motor
probably not optimal because when its in use it'll be cycling through at a computer controlled rate
@@AsbestosMuffins there's effectively a clock track on the tape, so for this design it would make sense for that to be the master clock.
Glad to see project focus. When its a month plus between episodes on a project, I forget where is it and can’t go back to refresh again and again.
An old reel-to-reel tape deck has reels and a capstan, wouldn’t that be a great starting point, too?
Full speed ahead on the vacuum tube computers and the beard is back!! Success is assured.
Love the bunny flopping over just after petting. So calm.
Ah, this reminds me of the Data General Nova computer we had decades ago at FAU. It had an ASR33 tty with a (slow) paper tape reader as the only input device, and my assignment one weekend was to regen the operating system to include a new device (I forget if it was a high speed paper tape reader or maybe a cassette tape reader). To do that we had to feed about a six inch diameter reel of paper tape into it with the operating system on it. What fun! A single tape reader glitch at any point in the process and we had to start from scratch. After many tries, the final successful attempt consisted of hunching over the tty for several hours doing strain relief on the paper tape by hand so there was zero tension on the tape as it fed into the reader 🙂
Definitely necessity is the mother of all inventions and you have a very imaginative mind!! Ready to watch not only your tube computer running but also all the historic computer pieces that you have in your lab. I am very impress of your videos. Thanks for always guide us through your work in a manner that even a child can understand. Great work!
Hittin" the tubes like it's 1958! Yeah!
Soldering for a mechanical joint (like on your gear) is different from electronic soldering. You should heat the shalf and press the solder into the heat only, it should flow almost like a liquid, penetrating the gaps between the shaft and the motor. Also be careful with the parts that contact the paper, so that it isn't damaged or ripped over time while running through the machine over and over. Because, as you point out, normal tape readers read the tape once and were done but yours will run continuously while in operation. Also make sure your lamp isn't going to yellow or weaken the tape, if it's going to be continually flowing.
As someone who subscribed for the UE1, and pre-IC technology in general, this announcement is great for me in particular!
Can't wait for Vacuum Tube Computer or Bust!
Hi, without at least one bearing (two much advised - this size can easily be found in scale modeling shops) on the vertical axle of the traction rotor, you will rapidly run into a problem of wearing its aluminum guide. Another thing is the rotor itself, it would have been better if you've used something standardized, like a cassette reader rubber tensioner, to allow an easy maintenance - another solution would have been to use a sewing machine mechanism with the paper band center holes.
Yeah, that shaft is going to eat up the aluminum in no time, especially with the tensioned roller pushing against it. He could have gone with a stepper motor and mount the rubber directly onto its shaft. Any medium sized stepper motor already has two bearings and it would have given him precise position control as an added benefit.
@@vatterger True, but my guess is this was out of budget.
@@hankhulator5007 A Nema17 motor + basic driver board cost about 15 to 20$ incl. shipping, thanks to the 3d-printing hype ;-)
@@vatterger Cheap, but may not enough cheap (can't beat the motor he's using) - but I agree, as this kind of stepper motor is specially made to resist axial pressure (not to mention that with a MCU you can easily program a short stop on each byte if the photo-diodes are a bit lazy).
A simple brass bushing would probably be adequate.
The originals used a small sprocket driving the center line of holes and a spring tension wheel holding the tape tight to the reader.
Also, you can use a fiberglass-brush pen, or a hard-rubber pencil eraser to clean the commutator contacts; Scotch-brite pads may be too abrasive and leave scratches in the soft copper; this’ll just attract more dirt, crud, fine particles of graphite from the brushes, and oxidation - not good. 😢 Also, use a toothpick to clean out the crud in the gaps between the commutator segments; little trick I picked up from SMT Mainline (YT channel)!
Just a little FYI: the term “hard rubber” actually means rubber that is polymerized to the point of being rigid. (Think bowling ball hardness.) 100% agree on the recommendation to use a natural rubber eraser for cleaning contacts. I always have one of those woodcase eraser pencils (the kind used as typewriter erasers in bygone times). I use the ones with ink eraser since it’s just a bit more abrasive than the ones for pencil, but still gentler than the fiberglass pen (which I also love).
I can't speak for everyone else, but I have wanted to see them functional for a long while now, I understand that it's difficult to reverse engineer everything and find parts and information period etc but I'm perfectly ok with a three hour video of you solving multiple issues at once to get things done, or even just making three one hour long videos. I appreciate your time and effort!
All other projects in the room on hold until both of these are working... I look forward to the next video being about a brand new project.
This is so cool ! I am glad someone is preserving this old technology and making it accessible to all for free on TH-cam. Thank you Usagi Electric! Good luck with the circuity for the photodiode array, I know sometimes photodiodes can be tricky if they are in photovoltaic mode or reverse current mode and a trans impedance OP Amp circuity needs to be used
Oh, hey a model train motor! Those were also widely used in scale model trains and I've repaired lots of those. They were usually made by Pittmann. Pretty bullet proof. Although it's not always necessary, you can also replace the magnet with Neodymium magnets and the motor will run cooler and have more torque.
finaly
more vacuum tube stuff
great job btw
I enjoy non-transistor computing a lot. Personally need to get my relay computer project going again.
Personally, I don't care what you work on, or when you work on it. It's all interesting to me. If it matters, I have no concern about when or how you do anything. I just enjoy the videos. No justification needed on my end.
Lovett Sundries rocks, and so does their very friendly dog!
For people complaining that you're using photo diodes for your reader, the majority of the logic in the vacuum tube computer is diode logic. Same with the Bendix. That's how it was done, back in the day. Semiconductor diodes were available, and were used extensively. The logic is done by diodes, and the vacuum tubes are used to amplify and invert the signals.
The same technique was used when transistors became available. Early transistor computers were diode logic with transistors for amplification and inversion.
AFAIK, transistors didn't start being used for everything till IC made it easy to put a few hundred to a few million transistors onto a single wafer. And even then, i bet there are still quite a few diodes on IC's, even today.
I know this episode (and probably the next few) aren't Centurion related but I have to say I have absolutely loved watching your entire Centurion build series, I've probably watched the entire video set 6 or 7 times now, and you have sparked a big interest in vintage computers in me. I don't have the space (or sadly, time) to even attempt what you've done, but maybe one day I can make it to a VCF East and finally get hands on with that amazing blue machine. Love your work, love your videos. Please keep going, you do have a lot of fans out here, myself included.
Fan who shares your first name.
12:06 this is how Aperture Science would make aluminum foil.
Just random thought, instead of a 'bucket' to hold the loop, why not a spool like used in old 8-track cassette tapes. pulls from the inside continuously, making the reel turn and takeup on the outside. It does cause the tape to slide over itself which MIGHT be a problem for paper tape, but worked really well back in the day.
A loose pile in a 'bucket' might easily get knarled and 'knotted'/ twisted causing all sorts of feed problems. Just a thought though...
Always enjoy your adventures and jealous of the manufacturing resources at your disposal. Access to CAD tools, machine shop... get to dream up and make all sorts of things.
You beat me to the suggestion.
If the spool is wound loosely, friction shouldn’t be a challenge.
Well, when you said that just two computer projects are the ones still to be continued any time soon, I was like: okay, now just sniffing for any notification from David!
Really enjoying both the repair/diagnostic and green field design work. So satisfying to see that which is old made new again.
YAH project goals!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
Just a thought - but I'd fit oilite bearings - better for friction, wear and fit than a drilled hole in aluminium !
They come in all sizes so you're bound to find ones that will fit !
For adjusting the tension of the strip, you could get some inspiration from old film cameras and projectors, which used something called the "Latham Loop". It acts as a buffer for any vibration or variation in tension while pulling the strip through!
"All your 6AU6 are belong to us."
This project is a thing of beauty and a joy for ever. I'm deeply impressed by your brother's skills - I'd never have expected that he'd be a great metalworker with his own machine shop! He did awesome work with that tape drive motor mount. I'm sure Quinn Dunki (Blondihacks) would be impressed too!
Watching the reader pull the tape through fills you with determination.
I will also recommend if you want to screw steel screws into aluminum, a good technique is to oversize thread the holes in the aluminum, then put in helicoils. Otherwise your steel screws might gall the aluminum, making the screws either irremovable, or ruining the threads in the aluminum. Both are bad. Having the helicoil gall into the aluminum doesn't matter, as you won't be screwing and unscrewing that.
Wow. It's amazing how far you get just talking about a problem with other people. You find solutions in the most unexpected places! (btw; for pedants it's chAmping at the bit, not chomping).
"I hope you like vacuum tube computers" I do now because of you. :D I was watching 8 bit repair guys until I stumbled upon you some time ago. You're taking computer repair to a whole new level. (You get it? Voltage level?) Carry on.
I mean I watch your channel to see tube computers so eat your heart out. And get a few more while you're at it!
I hope you don't have a problem with continuous speed, Because usually tape readers are step motors, they move the tape hole by hole like a film projector moving frames one by one to allow the necessary exposition time, Not sure how much the exposition time for those diodes to register a bit, but I hope it will work out with continuous speed.
Hey, if you want to stay vintage-era and not use LEDs, I might recommend an “exciter lamp” from a 35mm projector (these were used for optical sound pickup) as your light bulb. These often have a single linear filament that might be helpful!
As long as he doesn't use a bare bulb that's a really good idea
Osram/Sylvania BULB
Part # 4B10021517OSR
30 day replacement if defective
$1.86
From advanced auto parts.
This is a 12 volt festoon based bulb similar to the bulb used in the Univac 1532 paper tape reader .
clock escapment mechanism could be operated by cpu clock by a single flip flop and push tape one byte at a time every 1, 2 or multiple of 2 clock pulses
basically pengelum made out of 2 latches pushing tape one step at the time
the vacuum tube computers are my favorite projects of this channel :D
Be very, very careful with what lamps you use to trigger the photo diodes. Color, and temperature of emitted light is VERY VERY VERY critical to high speed running. Check specs on your diodes. You can actually sort of "overclock" phtoto readers by very carefully giving the diode a consistent "sweet spot" from the lamp source. I knew a motorola engineer that explained to me they would go through thousand of lamps and build pretty hefty cap banks to avoid any dip in light. He said everything became easier towards end of product life when they actually used lab microscope lamps as emitters. The rabbit hole of lamp specs is deep, and almost all are still made. Very excited to see where this project goes!
When you have something that must get done, THIS IS THE WAY! No problems whatsoever focusing in on the things that have deadlines on them ;)
Much joy at this video. Tho all your videos bring much joy.
I just noticed in this episode that the Bendix G-15 kind of resembles Rosey from the Jetsons. (Seriously, super cool work on all of this)
I found that you have to have a couple projects going at once. Inevitably you are awaiting parts for one or the other. So I end up switching back and forth. But in the middle of that my desoldering station broke and is awaiting parts. So now there is 3 projects. So there is always a strangeness factor that comes into play too.
Absolutely beautiful, what a unique thing to be working on in 2024, love it! Hope you don't get a burn-out, but I'm here for the vacuum tube episodes!
You might want to check skateboard wheels as rollers. They are large, smooth, and extremely resistant.
I can’t wait to see what the limits of this homemade machine are.
I like seeing measurements in thousandths of millimetres. 👍
Now I'd love to see a crossover between Usagi Electric and Inheritance Machining!
Time for a conversation with Curious Marc about an analogue feedback mechanism to create a frequency to voltage circuit, enabling automatic speed calibration?
Vacuum tubes are whT brought me to this channel - so, yeah 😊
When threading with a hand tap, please break the chips now and then. You're just asking for trouble if you ram it in like a regular screw. My dad taught me to do 2-3 rotations forward, then half a turn back to break the chips.
Great job, keep it up!
It's a 20 minute video that's only been up for 8 minutes
😂
Thank you!
@@charliesybert2683 Patrons get early access :)
I'm super excited for all the vacuum tubes coming my way in these next few weeks~!
Looking good, nice job! I'm glad the slot-car motor worked out so well :)
Vacuum tubes, yes. I wonder how you can handle data load, perhaps using index hole photodiode state change? Bendix G15 is a beauty that deserves all your attention.
He will need a motor driver that does step by step ratcheting control due to not using a stepper motor with just right gearing instead. he said he will do direct read off the tape to the CPU, not buffering. It means he needs something, that indepently and accurately drives the motor only so long that next ratchet hole is visible and stops until the read and execution is done and the CPU gives order to step to next instruction. Which to me is weird he didn't mention. Since I don't see that as trivial. He needs atleast few logic gates to do "start running, ignore we are on a hole (the previous current instruction, we want the next one), latch that we moved off that old index hole (reference to compare against for "does light through index whole causing motor stop or not"), keep going, sense that we are again on a hole and stop (because stop on index hole is latched), set signal for the CPU, that next memory position is ready to read". That is non trivial.
Not to mention then probably a lot of fine tuning. Since that is non stepper motor. So he has to tune in "how fast can the motor be driven, so that the driving electronics have time to stop the tape dead on the middle so everything aligns and good read happens" and at which level of the index hole lighting up to trip the stopping. Too low and it stops as soon as even the edge starts to come in, meaning still probably out of alingment. Too high and it probably then actually stops after the central alignment, since trip to stop only comes at the central alignment and mechanical things aren't instant. Given such precision timing, I think it has to be pure hardware and then stuff like pot resistors to calibrate the driving. So that reader will need couple tubes itself and maybe some diode logic or something.
Looking forward to HELLORLD in some form or another from the UE1!
So many projects mean you have content for future TH-cam videos. That’s probably the hardest thing to think of.
You champ at bits, not chomp
Sweet i've been following all your projects but I get most excited for your tube computer projects I've get a big smile on my face for each stage of the bendex start up. I't would only be better if you had to hop on an exercise bike as one of the stages
It's Tube Time!
Electronic solder does not stick well to steel (which is what I assume your motor shaft is made of), so that gear may come loose. Suggest cyanoacrylate or Loctite as an alternative. Nice job, though. Tape may feed better if you rotate the whole system 90 degrees, so that the tape is flat and not trying to twist. See the DEC high speed paper tape reader (although I think that used the feed hole to move the tape) -- I like your method better -- doesn't stretch the feed hole.
This is petty impressive! Makes me wish I took shop class back in high school instead of thinking it was beneath me.
Makes me wish my hs had shop class 😂
Progress! What a steampunk looking design (needs more brass, though...).
Tip often used by machinists for small stock, instead of Dykem, just use a sharpie :)
And kerosene or WD40 for machining alu :)
Im so exited to see the bendix work! And the homebrew computer!
Hell yeah, I'm excited for a vacuum tube computer end-of-year!
The little motors were also used in early HO scale train engines, though those used worm gears on the motor rather than pinion gears. The problems with those motors are; 1) they run too fast for most applications, and 2) the commutators often need cleaning due to arcing, and carbon build-up from the motor brushes. Also, the bearings need a small drop of light machine oil every now and then.
looks like the channel is gonna go back to it's origins for a bit with every video being about tubes!
Helllllll yeah, I love bendix updates!
"Until both of them are doing something exciting" .... fire is exciting
Oh yeah, full effort November!
use parafin when cutting ally - old timer tip
Paraffin oil or paraffin wax?
@ paraffin oil
Heard WD-40 works too. On paper, it's mostly the same stuff
@@eliotmansfield Thanks!
@@No-mq5lw It’s better than nothing. Of course one can also just buy a can of cutting oil!
Great stuff !
I can just picture it; The tape containing the code for : Println: "HELLORLD" flying past those photo-diodes at great speed....::...:.:.:::...::.....:
Cheering you on from Fort Worth! Had lots of fun playing lander at VCF SW on the Centurion.
We all really enjoy your content! But please don't burn yourself out. Health is the most important.
1:20 it sure needs a hug :)
One change I would think of would be to cut the top of the readers so the tape can slide down rather than having to be threaded. This would require some sort of adjustable cap or clip to keep the tape from escaping though that slit.
Why?
Because when you get down to writting programs it will be easier if the tape is a continuous loop. Otherwise you'll have to 'auto rewind' when you hit end of tape, or cut and re-splice the tape every time you switch programs.
Also taking a clue from some of earliest magnets drum memories, a 'jump' forward and backwards is just the same instruction with different values
It's not just me! It's not just me!
Hugging a 1950s small drum computer! In my case, when I saw the (non-functional) IBM 650 in the Smithsonian, in 1992. My wife looked on affectionately. (Her choice of adverb)
My bachelor's thesis at Princeton in 1987 was all about the 650, including a mostly functioning simulator, written using VAX assembly language for the decimal arithmetic, and the CURSES package for screen handling, making the blinkenlights work in full screen text. I didn't know if I'd ever get to see a real one. I've seen the video of a working restored 650 in Germany from about 20 years ago. .
As to your tension posts, just slip some delrin hollow bar over what you have.
Vacuum tube computers in winter... you can compute and stay warm LOL
Yay tube computers!
Would still love to see a rotating drum (or stacked PCB discs) capacitive memory (regenerative capacitor memory) for the UE1.
Like Atanasoff's memory? Two rotating drums with 1600 capacitors each, mounted around the cylindrical surface.
@@thanbo That's where I got the idea although it wouldn't need to be exactly like that for his system.
Hellave! (See what I did here? 😬)
As for all your planning going elsewhere remember, what Einstein (allegedly) said: "Planning replaces random with error!" 😊
So just keep on doing the great work - I guess everybody here simply enjoys it and loves your way of publishing it. I certainly do 🤗
So thanks 😊
Axel
the choice of motor is beyond strange but if it works it works
Firebottles! Real electronics light and heat the room! (Or maybe I'm just punchy because I've been doing SMT practice boards with my shaky hands.)
WD-40 (in the pump sprayer) works well for cutting aluminum.
You missed an opportunity for a terrible pun with that unholy rectangle in the aluminum plate!
Adding a bearing to better cover the width of the tape and spread the pressure ?
It seems like fabulous beards run in the family
fantastic work, mechanic is always challenging.
I started watching 'cuz of the Centurion but stayed for the UE-1.
In case you care, At 16:06 you say skill saw but show a jig saw. Skill saw is normally a synonym of Circular Saws. Not sure if that's a regional thing or just a mouth-typo.
"There's a LOT of projects going on in this room, maybe too many..."
Me: "Wait, is he in MY garage?!"
😛