Hi /diy/, I make vacuum tubes. I make other things as well, but the tubes are by far the most interesting thing. I've been documenting my projects here:
http://simplifier.neocities.org/
If you have any questions, ask away. I don't typically browse here, but it seems like the best place to put this. If you're wondering why the site has almost a year's worth of projects and only ~500 views, it's because I've mostly been using the site to show things to my friends and family what I do in my spare time. I never considered my projects to be novel enough to be worth publicizing. Yesterday I made a working triode though, and I got so excited about it that I had to share it with other people. I'd suggest you start at the bottom, or at least at the vacuum chamber article if you want to have a good idea of what's going on.
>>946133
Radical dude
>>946133
Very interesting, taking DIY to a whole new level.
Anyone got a link to that old tranny who makes vacuum tubes on YouTube? Dudes a badass...
>>946166
Thanks man.
>>946194
It's really not that hard to do, that's what surprised me the most about the whole process. The hardest part was finding a source for the glass.
>>946201
https://www.youtube.com/user/glasslinger
Ron Soyland is my hero. I've watched every video on his channel multiple times. Here are some other tube makers if you're interested:
Claude Paillard:
http://paillard.claude.free.fr/
Robert Hunt:
http://teralab.org/
Nyle Steiner:
http://www.sparkbangbuzz.com/
>>946133
Hello, /diy/ glassblower and tinkerer here.
I did some simple scientific glassblowing, simple condensers and vaccum aparati. Also i used to make christmas ornaments, usually spherical, but experimented with woodme moulds - preheated metal ones would be much better.
Then made some neon bulbs and diodes.
But then... i had to stop having fun with glass. because:
>>946206
>The hardest part was finding a source for the glass.
Couple years ago, the only glass manufacturer who produced glass rods and tubes in my country, ceased to produce them... I have couple tubes left but not enough to start having fun again.
It was a great hobby, wish i could advance myself in that field further.
>>946210
>Then made some neon bulbs and diodes.
Hell yeah. I wish I could get my hands on some neon for cheap, seems like a fun gas to play with.
>Couple years ago, the only glass manufacturer who produced glass rods and tubes in my country, ceased to produce them
Look for neon sign tubing. I'm sure someone in your country (or a nearby country) makes neon signs. I overlooked it as a potential source of glass early on because I figured neon signs are really thin. Turns out they make neon tubing up to 20mm and it's really nice to work with.
>>946133
I'm fascinated by the site. Thanks for sharing.
I'm curious what your background is because you seem to have a pretty decent grasp of a variety of subjects.
>>946214
I still have small tanks of neon and xenon laying around.
I bought them 7 years ago, Linde was selling them in small 200ml aluminum bottles. They were ridiculously cheap (but still pricey as for a mere mortal). Good that i bought them back then because Linde is all dicks now, don't want to deal with mortals...
>>946214
>Look for neon sign tubing.
Tried that, i could import it from abroad for too much money, and ones available on local market are too thin... Also only bulk orders...
>>946221
Dropped out of college twice, lol (CS and EE).
Most of the time I just see something I want to build, get obsessed with it, and end up picking up every skill required to make it. Over time the skills accumulate.
>>946223
>Linde is all dicks now, don't want to deal with mortals
The fact that you can buy radioactive isotopes as an individual but you can't get a lecture bottle of neon is mind-boggling.
>>946226
>Tried that
Lame :(
If you're really desperate you could try test tubes I guess. That would be hell though, speaking as someone who tried that.
>>946237
>speaking as someone who tried that.
Been there, done that either.
Crappy quality ones tha ti got for free required annealing before any work could be done. Also they are short and requrie attaching stems to them in order to work easily and some fiddling.
>>946133
Let me preface this by saying I don't know shit about glasswork.
With the difficulties of trying to keep stuff in place doing the crimp closures once you started adding more than the two pass through wires...would it work to make like a small glass bead, arrange your wiring around that, then heat it, press the wiring into the glass bead (not necessarily fully), then stick that assembly into the tube's hole, heat all of that up and sort of use the bead like a cork?
>>946244
The pinch seal really isn't that hard to do. I've got the procedure down and I haven't had a seal failure in a long time. You could pre-bead it like you said, but with titanium wire you really want to minimize the amount of heat it gets in open air. It sucks up oxygen like a sponge and gets brittle really fast. I'm sure it would still work, but the pinch method is just really easy on its own.
So long as the feedthrough assembly makes a friction fit inside the tube when it's cold, when you heat the tubing up the wires will stick to the hot glass and stay in place as you pinch it off.
>>946249
>titanium
You use titanium for feedthroughs? Interesting.
I mainly used tungsten (acquired a lot of very thin TIG electrodes) worked like a char. I couldn't obtain copper cladded NiFe but made some copper cladded steal wires, they seal excelently, but may crack glass due to different thermal expansion coefficients.
Have you tested your seals with helium/hydrogen? Mine leaked quite subtantially, except for pure copper ring seals.
>>946257
>You use titanium for feedthroughs?
Yeah, it's a great material for anything vacuum-related. Has the same coefficient of thermal expansion as soft glass, and the oxide layer dissolves in the glass and makes a hell of a strong bond. It also acts as a getter if you heat it (and maybe to a limited extent without heating, I've found). I don't have any fancy leak-checking equipment, but I made a diode a little over a week ago and its characteristics haven't changed since, so I assume the seals are good. Really, all the tube failures I've had so far were either due to poor glasswork or deliberate overheating.
I really think I lucked out by picking titanium at the start, everyone who makes tubes seems to have a hell of a time with the feedthroughs, but it's been smooth sailing for me. The only downside to titanium is that it can't be soldered without some super nasty fluxes.
>>946262
>Has the same coefficient of thermal expansion as soft glass
That's interesting. I never knew that, but I've also never worked with soft glass before, only borosilicate. Have you ever tried sealing copper and pyrex? I've read it's possible despite the massive difference in expansion coefficients because the hot copper is malleable enough to bend under the pressure of the cooling glass instead of cracking it. Supposedly it works best with very thin sections of copper, like ribbon, flattened wire, or a copper tube whose open end is filed to a knife-like edge.
>>946293
I've never done it, but I know what you're talking about. It works on soft glass as well. Here's a video of the process.
https://www.youtube.com/watch?v=OWw32BLodjY
I've considered doing this with two copper pipe caps at the ends of a length of tubing (outside of the cap facing into the tube) to make a big Geissler tube. It could act as a crude vacuum gauge (the characteristics of the glow discharge of air are a pretty good indicator of its pressure), and it would be good practice for making this kind of seal.
Now that I think about it, that sounds fun. I might try it tomorrow.
>>946301
Whoops, forgot my name. I usually post anonymously, but I figured for this thread a name would be relevant.
Bump for awesome
>>946293
>>946301
>https://www.youtube.com/watch?v=OWw32BLodjY
I've done but with smaller diameter tubing (6mm turned to really thin wall).
Wordked like a charm, i've done it on my metla lather, spinnign aloows more even heating and nicer seal. I was talkign about that seal here: >>946257
>Mine leaked quite subtantially, except for pure copper ring seals.
>>946302
>Whoops, forgot my name. I usually post anonymously, but I figured for this thread a name would be relevant.
Well it makes sense, let's call myself MSG, i'm not the OP.
>>946416
>except for pure copper ring seals
Ohh, gotcha. I wasn't sure what you meant by that. Good to know it's a reliable type of seal. I'm waiting on parts in the mail to make more triodes, so I'll mess around with some copper today instead. Probably test the triode I made earlier as well. Here's hoping it isn't full of gas by now. I'll be back with updates later this evening.
>>946133
If you have higher-res photos of your projects, you should update the website with them. This is some good content.
This is so fucking cool.
Updates!
RIP triode. It can still amplify a signal somewhat, but there's a lot of leakage current that can't be modulated by the grid no matter how hard I pull it negative. Also the signal cuts in and out as I vary the filament power. Tl;dr half-dead via (most likely) outgassing.
Tried to make a Houskeeper seal between a test tube and a copper pipe cap. It almost worked (bright red color at the joint), but I think the copper was too thick so it cracked apart as it cooled. Didn't really pursue it further.
Working on a second triode at the moment. My induction heater is coming along and can put out enough power to heat a loop of titanium wire, so I copied the first design but gave it a halo as a getter. Pic related. Just sealed the base into the envelope and it's cooling right now. Should be finished later tonight. If it works, I'll have an article within a few days.
>>946741
Oh shoot, forgot to mention that I also managed to spot weld 0.5mm nickel wire to 0.8mm titanium wire. That's why the leads look different from the guts of the tube. Wasn't sure if it would work; couldn't find any info online about joining the two. I figured since Nitinol exists it shouldn't be a problem, and turns out I was right. Couldn't do it with 0.8mm nickel because the nickel is too conductive compared to the titanium and wouldn't melt.
>>946741
>Tried to make a Houskeeper seal
I first drilled as much material as possible form the inside of the tube and then turned a vry mild taper on the outside, at the edge it was less than 0.3 mm. If that helps you anyhow. Also i failed 5 times before successfully making gas-tight seal.
Btw, why your triode's intestines look so oxidized even before sealling it?
Is titanium so prone to oxidation even when just spot welding it?
>>946753
>Btw, why your triode's intestines look so oxidized even before sealling it?
If you're talking about the dull finish and gray-ness, the wire looks like that when I get it. But yeah it's super prone to oxidation when it's hot. Anything smaller than 0.5mm can be lit like a strip of magnesium.
>>946753
Titanium oxidises immediately with air but it is passive and the layer generally prevents it from oxidising further. It's so good at resisting further corrosion that we don't bother plating or coating it in the aerospace industry.
wouldn't a titanium mesh be better?
cool stuff by the way, post more of this kind of stuff
Awesome stuff man, kinda pleasant knowing that some people have the skills to make old technology.
Cool stuff, op
What metal are you using for your glass to metal pinch?
Looks to me that shit would fail a lot.
Reminder that this guy started few years ago from scratch.
http://dalibor.farny.cz/
http://www.daliborfarny.com/
>>946890
Titanium or nickel mesh would be ideal for the grid (the squiggle part) but I wasn't able to find any in small quantities for cheap. I found a square foot of nickel mesh for 40 bucks on ebay, but that's a little pricy for early prototypes. Making a wire squiggle is fine for now. Eventually I'll be making tubes with cylindrical anodes so the grid will just be a spring anyways.
>>946894
>>946895
>>946921
Thanks guys.
>>946964
Titanium. It's surprisingly reliable. I've made pinch seals with 1.0mm titanium without a problem, and could probably go bigger if for some reason I needed to. Which makes sense, the expansion coefficients of soda-lime glass and titanium can be an exact match depending on the composition of the glass and the alloy of the titanium.
>>946969
Dude does come cool stuff. Wish he would make posts more regularly though. One thing I want to avoid that seems to have happened with him though is tool accumulation. While having a turbomolecular pump and a glass lathe would be hella rad, it would also remove my projects from the realm of amateur tinkering. I don't want my page to be "look at the cool stuff I did because I'm awesome", I want it to be "look what you can do too because it's easy and cheap". Nothing against Dalibor though, making nixies seems fun as hell. I've considered making a helium-based nixie (hexie?) in the future. Would be a hell of a pinch seal though, not sure if my skills are there yet.
As for updates, I finished the second triode late last night, and will be testing it over the next few days. The titanium ring getter seems to work well; I heated it with my induction heater (article soon) until it glowed orange-hot, and when it cooled down the metal was shiny (the oxide layer had absorbed into the metal). Initial tests looked good, I could block 400µA of plate current by pulling the grid to -6v, which translates to a transconductance of roughly 67 µmhos. Time will tell if it's stable though.
>>947037
>Would be a hell of a pinch seal though, not sure if my skills are there yet.
I was considering trying to do it properly like in real tube factory, with mould and pressed glass pedestal, but this again collides with:
> tool accumulation
The bigest pinch seal i've done was 5 wires, i tried 6 but failed couple of times and didn't have enough incentive to try more. At one point i thought about layering the seal like making a sandwich with glass bologney inbetween wire lettuce. It would reduce width of whole pedestal.
>>947046
>I was considering trying to do it properly like in real tube factory, with mould and pressed glass pedestal
It could probably be done with a graphite mold and a small microwave kiln. Wouldn't be too expensive that way. Graphite's cheap and you could make the mold by getting creative with a drill press. It'd be tricky to seal into a tube without melting the pins out of alignment though. That's why I like the pinch seal method, you can really butcher the envelope seal and still have a working tube in the end.
>The bigest pinch seal i've done was 5 wires, i tried 6 but failed couple of times and didn't have enough incentive to try more.
I've done 5 myself and that does seem like the reasonable limit with 0.8mm feedthroughs. I'd have to drop down to 0.5mm to make a nixie. The layered idea seems neat, but I'd be worried about getting the glass too thick and having it crack as it cools. Either that or pinching too hard and shorting the layers together. I'll probably just stick with regular pinch seals.
>>947057
>crack
Never tried soda-lme glass, but with pyrex i never had problems with seals breaking unless i haven't cooled them slowly and annealed.
I wish i had some more expendable money so i could rent oxygen tank and have some fun again...
>>947060
>Never tried soda-lme glass
It's really not that bad, you just have to make sure everything is of a fairly uniform thickness, and that's just if you don't have an annealing kiln. I like it because you can work it with a regular blowtorch.
>oxygen tank
Look on craigslist for oxygen concentrators. I've seen them for $50, and one of them will make enough oxygen for small-scale pyrex work. IIRC you only need two if you're working with tubing over 2" in diameter.
there was a video I saw once of a guy going through the whole processes
it was like and hour long
there was no narration or talking
if there was any background music it was quiet and subtle, maybe classical
it might have been a couple decades old
anyone know what I'm talking about and can find it?
>>947070
https://www.youtube.com/watch?v=EzyXMEpq4qw
That video was my inspiration. I basically follow the same method he does, except with smaller tubing and cheaper equipment.
>>947064
>Look on craigslist for oxygen concentrators. I've seen them for $50
Well, i live in Europe (much closer to daliborfarny than you i suppose, haha). I searched for O2 concentrators but with no luck, usually they needed new zeolite packs or were sold without this crucial element...
Also since Linde became ultra-dicks refilling my own tank is very expensive (they charge twice the price of O2 plus some extra charges for handling). Only option is renting tanks. But it hurts when 11m^3 of O2 costs like $10, and they charge $0.5/day for the bottle, so i end up paying sugnificantly more for the tank than the gas (i usually use whole bottle within 2~3 months during fiddling in my free time)... Oh and there's recognisance for the tank - $200 (new tank costs $100 - i love you Linde...).
Enough of me bragging!
>>947070
>there was a video
I could only find this one in my YT favs:
https://www.youtube.com/watch?v=dErsA6wFmlM
Although it contains narration.
there's this gap in diyable things and it's growing ever bigger ever faster
tubes are awesome, but what about transistors?
will we ever be able to diy or home 3D print ICs?
>>947079
>but what about transistors?
Transistors aren't that hard, you need just monocristaline silicon wafers ad some basic kitchen equipment.
Just take a look at her amazing work:
https://www.youtube.com/watch?v=PdcKwOo7dmM
Making ICs, that's the challenge. Hard to get ingredients, machinery, clean rooms, etc.
Updates!
Tested the second triode (pic related) today, and turns out it improved overnight! No erratic behavior, and the tube was able to switch up to 1.4mA at 90V (albeit at very high filament power). Grid current seems very low, if it's there at all. I was able to modulate the plate current with a megohm resistor in series with the grid lead, and could use my body as an antenna if I left the grid lead floating and touched it with my hand. All very good signs. Heated the getter again after testing. I'll check up on it tomorrow and the next day, and if it still works I'll write an article on it.
>>947278
Congrats! Any plans to move on to something like a pentode once you finish your work with the triode?
>>947301
I've got some titanium tubing coming from china that I plan to use to make some triodes with cylindrical anodes, but once I get the hang of that I'll probably make a tetrode. Specifically because with a tetrode you can make a single-tube AM transmitter, which is pretty neat. Pic related, sorry about the small size.
I can't see myself making a pentode; I think I'd rather deal with the non-linearities of a tetrode than try to stuff a filament ,three grids, an anode, and a getter, all in a piece of 18mm neon sign tube.
Beyond that, I think I'll try seeing if I can get any of my tubing to fuse to recycled glass (beer bottles, etc). I had success joining it to pieces of a broken light bulb a few days ago. Being able to use bottles would allow me to make massive tubes though, which could be fun (big nixies, fancy lightbulbs, CRTs, etc).
>>947309
I built a series modulated one with commercial tubes. Worked ok apart from the FM content.
This is very cool, keep it up man!
>>946133
nice buttplug.
Second triode is still alive, and I wrote an article on it.
http://simplifier.neocities.org/triode2.html
>>948276
Sorry for asking but i never laid my hands on titanium. Which metal would you compare it to in terms of malleability? Does it work harden easily like silicon rich aliminum or chromium alloys?
>>948284
Really depends on the alloy. It can be like spring steel or like soft copper. The soft stuff is nice to work with, the hard stuff is awful. Work hardens really quickly. Typically I can bend something and unbend it once or twice before it starts to get crumbly, even the soft stuff. Can't heat treat it in air at all, either. Really nice strength to weight ratio though. Feels like really light steel, like it's got air in it or something. Haven't machined it much, just faced off the ends of an inch of tubing on my lathe, but it cut like aluminum with a sharp HSS tool.
Hope that helps.
>>948292
Oh, forgot to mention, the work hardening is only if you really bend it out of shape. It has a fatigue limit like steel, so you can make springs etc out of it and they won't fail over time.
I've been ignoring my glassblowing for a while and only working on electronics projects, and this thread gave me a little motivation. I was able to seal a few thin, flattened copper wires through some tubes. Not sure if they're vacuum-tight; don't want to wake up roommates with the vacuum pump. I want to attempt to make a light bulb. To keep it on the cheap side I was thinking of using pencil graphite as filaments, but I'm not sure how to connect them to the copper leads.
Camera tripod for webms when?
>>948466
> I was able to seal a few thin, flattened copper wires through some tubes.
Nice!
>To keep it on the cheap side I was thinking of using pencil graphite as filaments, but I'm not sure how to connect them to the copper leads.
Just wrap the copper around the graphite a few times. There's really no good way (short of the Edison method of carbonizing some organic compound in a blob around the connection) of connecting metal to carbon in a vacuum. I'll warn you though, pencil leads are a pain in the ass. They have a negative temperature coefficient of resistivity, so they eat more current the hotter they get. Expect to put a lot of power into a seemingly small lead. I bought some 0.2mm leads from Japan to experiment with and IIRC they still needed a couple amps. For sure keep experimenting with them, but in the meantime order some .004" to .002" tungsten wire (thinner will burn up too quickly in a poor vacuum) from ebay. It's only like 10 bucks for a 30ft roll, and 30ft is way more than you think it is.
>>948482
I've considered it, but I really hate being in pictures, and there's no way to avoid that with a video. Plus, while videos would be entertaining, the site has all the info necessary to reproduce my experiments as it is. I tried to make sure nothing was missing.
ever work on lasers ?
I'm attempting to build a co2 laser bulb
or maybe a He,Ne bulb (pic related)
I would like to pick your brain :^)
>>949024
Nah, haven't tried that yet. Seems like it would require a lot of precision. Closest thing to that that I've seen DIYed is this air laser:
http://www.sparkbangbuzz.com/tealaser/tealaser7.htm
While I'm here, I'll update the thread on triode #2. Characteristics have been completely stable since I wrote the article, so it looks like the titanium getter is viable long-term (if 5 days can be considered long term). Got a small single-ended output transformer in the mail today and had the tube drive a little speaker, which was neat. The speaker was fairly quiet, but using headphones with the transformer was downright loud, even with only a 1-2 volt signal on the grid of the tube. Probably due to the reflected impedance of the headphones being greater than that of the speaker, providing a better match for the tube, which has a plate resistance of around 200K at best. Future tubes will have lower plate resistance, so they might be able to be used in an audio amplifier of some sort.
Also got some $2 high-voltage silicone-potted mystery-modules from ebay to mess around with. I'm guessing they're taser parts because they make a huge loud spark at the end. Turns out they can run continuously, so I'm gonna use them to drive Geissler tubes.
>>949141
>titanium getter
I'd like to know how does it compare to plain magnesium getter. I deposited a few times a magnesium getter, using induction heater. It seemed to work. I flushed the discharge tube under power, with pure oxygen at around 1 torr till it started glowing and sealed. After checking every day the glow seemed to lighten just as if the getter was absorbign oxygen. I couldn't get my hands on any source of relatively pure barium so couldn't even try higher class getters.
>The speaker was fairly quiet, but using headphones with the transformer was downright loud
Sounds like impedance mismatch - wrong transformer.
>$2 high-voltage silicone-potted
>huge loud spark
>can run continuously
I like it, for my test with discharge tubes i usualy use potted ignition modules for oil fired furnaces. They run straight from 230VAC and produce around 10kV (with no load ofc).
High voltage voltmetter is still in the planning, so couldn't check the drop on my discharge tubes.
Is this a dildo?
Wow, it's really cool to do your own triodes. I always want to try. But i don't have most of the tools to build and test them.
Probably in the future i will give it a try.
Yesterday i was watching a youtube video about the "flame-triode" amplifier. Seems like a lot of fun to mix high voltage and fire.
>>949350
>"flame-triode" amplifier
Oh right, i nearly forgot about that guy. He inspired me to experiment with thermocouples. I remember making oneusing nickelsilver and copper, it worked poorly, but worked. I couldn't repicate his copper oxide thermocouples though.
itt: op is a god damn hero
>>949226
>I'd like to know how does it compare to plain magnesium getter
It doesn't "flash", i.e. it stays solid and doesn't coat the inside of the glass. It also absorbs a wider variety of gases; IIRC everything but noble gases and methane, while Mg only absorbs oxygen.
>Sounds like impedance mismatch - wrong transformer.
You're right. The triode has a plate resistance of ~200K, and the transformer has an impedance ratio of 5K:8. I got the transformer because it was cheap ($13) and because it would be a better match than connecting a speaker in series with the plate directly.
>>949350
>I always want to try. But i don't have most of the tools to build and test them.
It really doesn't take all that much. Bare minimum you would need:
For testing:
>some form of power supply (can use batteries, honestly) ($20)
>some passive components ($10)
>a multimeter ($10)
For actually making the tubes:
>a 2-stage rotary vane pump capable of getting below 10 millitorr ($100)
>alternatively, a cheapo single-stage pump ($50). it'll probably work but expect frustration
>some vinyl tubing from the hardware store ($10)
>a hot glue gun and lots of hot glue ($10)
>assorted tungsten and titanium ($50)
>neon sign tubing ($50 can get you 15 pounds from FMS/brillite, use them if at all possible, they're great)
>propane torch ($15)
>spot welder (free if you're clever and have a scrap microwave)
And that's it, really. If you use a filament type getter like I did in my helium discharge lamp you don't need an induction heater, and Ti absorbs water vapor so you don't ~really~ need a bakeout oven if you're willing to run the getter every day for a week until things stabilize. As far as the physical connections, hot glue and plastic tubing can go a long way. You'll need a vacuum gauge as well, but my next article will be on how to build one for like $5. Tl;dr everyone thinks tubemaking is fancy but it's not fancy.
>>949438
Thanks man.
>>949524
Whoops, forgot my name again.
>>949350
One more thing, if that list is overwhelming just start with a pump, some tubing, and some hot glue, and suck the air out of a pickle jar. Also all the stuff I do with a lathe is just done that way because I have a lathe. Without a lathe it's entirely possible to do it by hand or by getting creative with a drill.
>>949524
>Mg only absorbs oxygen.
Yea it reacts with oxygne but i read in some old documents that it adsorbs N2 an H2, especially the later one.
>transformer
Have you tried making your own impedance matching transformer? Especially whne audio quality isn't its key feature EI cores from power transformers are pretty good for this up to around 2kHz. I'll look around, maybe i still have such DIYed one laying around in the trash pile.
>>949226
Barium nitrate can be obtained from "sparkler" fireworks. I'm not sure if there's an easy way to get pure barium metal from something like that but it doesn't seem impossible. You may even be able to plate a section of the glass with barium. I know that you can plate out silver from silver nitrate onto glass.
>>949698
>I know that you can plate out silver from silver nitrate
Precious metal vs. one of the most reactive metals.
Electrolytic method might work using molten salt or some unusual solvent, but certainly not with water. Thermite-like reduction of oxide with aluminum might also work.
You can buy small amounts of barium (grams) from eBay, though, so unless playing with chemicals is your thing, there's no point in fucking with barium compounds.
>>949627
>Have you tried making your own impedance matching transformer?
I could do that, but I think I'll wait until I've optimized the tube design. Hoping to lower the plate resistance to something more reasonable.
>>949698
>Barium nitrate
Wew lad. Not gonna mess with that stuff. As a general rule I try not to work with anything toxic because I'm a dummy and will rub my nose with it on my hands and end up bioaccumulating nasty metals over time. Same reason I won't mess with mercury and I try to avoid lead when I can.
Really, titanium seems to be an excellent getter and I see no reason to ever mess with anything else. Maybe magnesium if I want to go oldschool and use only ingredients that are commonly found in nature. But even then, titanium is pretty common, just energy-intensive to produce.
In other news I was messing with one of those HV modules and my bench power supply had a stroke and gave it 50 volts on the input and turned it into a firecracker. Couldve been that I turned the supply on while it was under heavy load, or it could be that the failure mode of those mystery modules is to short the HV end into the input end. Either way both things are hella dead so I'm waiting for a new (bigger, 30V 10A) power supply in the mail while I gut the old one for parts and clean up my mess of a shop. Expect possibly a slow week article-wise.
>>949524
The pump is probaly the hardest thing to find for me. At least at a reasonable price.
I guess that two fridge compressors won't do the trick.
>titanium wire
Probably not, i only know one supplier here in Argentina
>tugsten
I guess i can use TIG electrodes.
>Neon Sign tubing
I think i have laboratory glass tubes, 8mm dia.
>Spot welder
I'm definetely going to make one, but also i can borrow it.
Yeah, is not much tools.
Did you build amps or circuits with your tubes?
>>949892
>I guess that two fridge compressors won't do the trick.
One fridge compressor goes easily to aroudn 60 Torr (8kPa). It is hardly enough for degasing silicones and resins. For tube work it won't do the trick.
>Probably not, i only know one supplier here in Argentina
Try ebay. You might find some free shipping from china. It will take up to 40 days but you'll get it eventually.
>I guess i can use TIG electrodes.
That's what i'm using, smallest diameter that i could find 0,5 mm IIRC.
>I think i have laboratory glass tubes, 8mm dia.
Lab glass is pyrex (borosilicate), it requires oxy-propane torch to work with.
>Did you build amps or circuits with your tubes?
I planned to, but i run out of oxygen and coulnd continue my work. I only managed t test my triodes but didn't go as far as building actual proper working amp. However it is feasible.
>>949892
Do you not have ebay in Argentina? It might take longer for things to get to you, but that's where I get most of my equipment. Also when I said tungsten I meant <0.1mm wire for the filaments, not thick tungsten wire like TIG electrodes. If you can't get access to the parts I use though, here are some substitutes:
Titanium wire (for glass-to-metal seals):
>dumet seals from light bulbs
>kovar seals from reed switches
>MIG wire, as a poor man's dumet (will seal, not sure if it's airtight though)
Titanium wire (for getters):
>magnesium (strips from scientific supply stores, or shavings from a magnesium firestarter)
>barium getters from old vaccum tubes (you'll need an induction heater to use them though)
Tungsten wire for filaments:
>any type of light bulb filament
Glass:
>anything made of glass, thinner glass is easier to work with
>you could use old bottles but you'd need a hell of a torch
>really, just try to find a neon sign supply company, it's the best option. i'm sure someone makes neon signs in argentina
The pump might be tricky. They're pretty common around here to service air conditioning units, and there's plenty on ebay, but if you absolutely can't find one the cheapest pump I've seen is this:
http://www.tinkerhack.com//vac4.htm
No idea what kind of vacuum you can achieve with it; the guy who wrote the article got below 1 Torr, but that could definitely be improved with better seals and better grease. For the tubing, it'll need to have a wall thickness at least half the internal diameter or it'll collapse under vacuum (e.g. for a 0.25" ID you'd need a 0.50" OD) and it'll need to be made of soft rubber so it can squish together with no gaps. Good luck, man.
>Did you build amps or circuits with your tubes?
Not yet, still improving the design. I have a few things I plan on making once I have a more powerful triode though.
>>950066
No, there is no official ebay in argentina. There is a second company form ebay called alamaula.com but is more like criaglist than ebay.
Due a change in politics in the last years we can't buy things from abroad. Is really hard to buy from china for example.
If an air conditioning pump will work, i may find one. A friend use to work with them.
Tugsten and titanium is still tricky to me. Argentina forbbid the use of incandescent lightbulds in 2010 so it's hard to find one tha still uses a tugsten filament.
I'm going to ask in the engineering school, they may know where i can find some.
>>950080
If it's anything like here in Eurofagistan, the ban does not cover the small indicator lamps and such. Halogen lamps aren't covered either.
>>950066
>Also when I said tungsten I meant <0.1mm wire for the filaments, not thick tungsten wire like TIG electrodes.
FYI 0,5 tungsten TIG electrodes work well wth pyrex ich you first make a glass bead on them and only thne create a seal. Pinch sealing raw tungsten doesn't work reliably ot this diametr.
>>950102
>Eurofagistan
It's getting worse here with every month...
>>950080
>Due a change in politics in the last years we can't buy things from abroad. Is really hard to buy from china for example.
That's a shame. Probably 90% of my tools come from china.
>If an air conditioning pump will work, i may find one.
Absolutely. That's what I use. You can see my original setup here, it hasn't changed much since I first built it:
http://simplifier.neocities.org/vacuum.html
> Argentina forbbid the use of incandescent lightbulds in 2010 so it's hard to find one tha still uses a tugsten filament.
Like >>950102 said, halogen lamps might not be banned, and they typically have fairly thick/long filaments. For titanium, you don't really need it for anything but the getter, so if you can find anything made of titanium (a bolt, a washer, etc.) you can cut shavings from it on a lathe and use the shavings as a resistively heated getter inside your tubes. Like I said before you could also try magnesium for the getter, it's probably easier to find but in my experience it doesn't work nearly as well. Apparently an aluminum/magnesium alloy works very well too, if you can find it or make it.
>I'm going to ask in the engineering school
This might be your best resource. I'm sure they have lots of everything you need, it's just a matter of getting access to it. If you're near a university, the chemistry department might have some scrap glass you could use as well. It would be pyrex though, so you'd have to use a different construction method and different tools than what I use. >>950112 has the right idea as far as using tungsten and pyrex together, and Ron Soyland has mastered it. His youtube channel and website are here:
http://www.youtube.com/user/glasslinger
http://tubecrafter.com/
Again, good luck. With the limitations you're talking about, you're essentially DIYing in hard mode.
>>950125
>http://www.youtube.com/user/glasslinger
https://www.youtube.com/watch?v=ppj3gqUTt9E
Woa... this channel just took a pretty weird turn lately...
>>950136
Haha dude's always been like that. He's wearing a dress in the video where he makes a 50KV power supply. Doesn't bother me, he's still smart as hell.
>>950140
>dude's always been like that
Hah, i've never noticed and i watched most of his vids a couple years ago.
The more you know.
But yeah, he's smart.
>>950125
>DIY hard mode
Hahahaha.It's always been like that here.
I study Aerospace Engenineering, you won't belive how we make satellites hahah
For some reason, we don't have most of the material that are really common in other countries and if we do have them, they're at least 3 times expensier.
Same with tools. It's quite rare to find modern CNC machines and other cool tools (i'm not saying that we don't have them, but, they're 20 years old, one of the two strenght test machine that we have still uses floppy disk to collect data)
Maybe next year the import politics changes and i be able to buy tools in china again. Meanwhile i may start making small neon tubes (an pure UV tube will be usefull for my screenprinting exposure unit)
>>950112
Copper and MIG wire doens't seal well against glass?
>>950320
Pure copper doesn't seem to bee too good for feedthroughs, low melting point, doesn't bind with glass very well on its own. Copper clad wire is just a bit more manageable.
I haven't tried MIG wire, only DIYed copper clad, which i think had thicker copper layer than MIG wire had and it worked. However i had to heat treat wire to oxidisie it properly and then heat the seal in order for glas to proprly react with Cu2O layer and make a good seal - on the other hand using NiFe wire i could just pinch and aneal.
Keep in mind that almost all glasswork that i have done was using pyrex glass. I don't have much experience with soda-lime glass, other than christmas ornaments (which i've made out of pyrex mainly either).
>>950320
fine copper wire can be sealed through glass if it's flattened first.
Updates:
Still waiting on parts and equipment for more triodes, so I'm making a Geissler tube in the meantime. I'm experimenting with 0.7mm MIG wire for the feedthroughs and it seems to seal well, with little to no stress in the joint when viewed with a polariscope. It forms bubbles around the wire though, which may cause the seal to leak. If all goes well, I should have an article on it by tomorrow.
>>950746
cool man!
Decided to just marathon it and get the article done tonight. Pic related.
http://simplifier.neocities.org/geissler.html
>>950963
Well done.
Btw, why do you convert your photos into GIF format? You're loosing quality this way.
You should only resize them to fit into content, and mabe leave link to a bigger phot, all in JPEG @ >90% quality setting.
>>946133
How about adding a paypal address for donations? Awesome work btw.
>>950963
Keep up the good work, really admiring your glass work and how you're able to make electrical equipment from it.
>>951009
I'm using the free hosting option on neocities, so bandwidth is limited. Plus, and this might seem silly, but I really like the web-1.0 dithered GIF look. I try to make sure I don't lose any necessary information in the cropping/crompression though.
>>951012
You know, I made a patreon a while ago (paypal had a bunch of rules about donations), but I was just really hesitant to put a link to it on my site. Felt like begging. I have no idea about the etiquette of things like that so I'd like some opinions on whether it would be okay. If you guys think it's alright I'll put a link up when I make a more useful triode.
>>951061
Haha thanks, whenever I look at my glass work all I see are the mistakes (like the wiggly evacuation stem joint on the geissler tube) so it's nice to hear an outside opinion.
>and how you're able to make electrical equipment from it
This is really the biggest joy of it all. I went into this never having used a tube in a circuit in my life, aside from messing with old CRTs. I had no idea how many different things you can make just by sucking the air out of some glass. It's almost like making my own ICs, except better in a way since ICs don't make pretty lights.
>>951119
The whole pay per month thing on patreon is something i dislike, it may be good for people who depend on it. That's just my opinion though.
I thought you just have to unlock your paypal account for receiving payments and then people can donate by sending money to that address.
>>951120
Yeah, that's another reason I didn't post it. I'm looking into paypal again and it seems like I just overlooked something and it's easier than I thought. If anything, I'll do that instead.
>>950963
Neato. Why does evacuating the gas introduce the possibility of xray generation?
>>951324
Below a certain point, the less gas you have the more voltage it takes to ionize said gas to conduct across the gap. The higher the voltage, the faster the electrons will go when they finally do conduct. The faster the electrons go, the harder they hit the anode, and if they hit the anode hard enough, they make x-rays. IIRC this starts at around 10KV.
>>951377
Cool, thanks for the good explanation
LLLLLLLLLLADIES AND GENTLEMEN do ya EVER feel like "hey, I want a glass dildo??? NOW YOU CAN!
Man I wish there was such a thing as a low voltage nixie tube. I've looked everywhere but I'd say it's limited to physics of the type of display it is.
>>951463
What's wrong with high(ish, ~100-200v isn't that high) voltage?
>>951463
use a 34063 to make that voltage easy peasy.
>>946226
>from a broad
its not nice to call a women a broad. get some class.
>>951499
problem is when actors start running around with 100-200v in props it becomes a pretty big safety hazard
>>952393
Not much worse than electrostatic headphones.
>>952393
Honestly, 150v isn't a problem if you have current limiting resistors (Which you should, nixie tubes don't like a lot of current).
>>946133
can you make a light bulb that lasts eternal
>>952877
Not OP but it is a matter of having an oversize filament on an underpowered line IIRC. Look up Centennial Bulb.
>>952877
Nothing's eternal, but you can get close. Basically this: >>953127. Incandescent light bulbs are a huge scam and are calibrated to last exactly 1000 hours. Look at the curves on this graph. With only a modest reduction in power and efficiency you could easily have a bulb that lasts hundreds of times as long. IIRC filament life is inversely proportional to the 12th power of the voltage applied, which is ridiculous. Planned obsolescence is the devil's engineering.
Any tube I make, lightbulb or otherwise, will have an oversized filament. As far as I'm concerned, an inefficient device is always more efficient than a broken one.
While I'm here, might as well post updates.
Still waiting on titanium from china and a power supply from california. Both should be arriving next week. Been messing around with bulb shaped envelopes (pic related) but haven't put anything in one yet. Might use one for triode #3, still deciding. If the new triode design gets above 1mA plate current I'll start making circuits with them. Planning on making an audio amplifier first, then maybe a radio.
If I get the power supply before the titanium, I might do a little article on a joule thief variant I've been working on. Otherwise expect triode III.
>>946201
Well, don't know about the "old tranny" part, but this video made the rounds years ago
https://www.youtube.com/watch?v=EzyXMEpq4qw
fascinating.
>>946133
I would like to know your opinion on ouija boards
because clearly you are a wizard.
Just found some interesting info, about titanium used for pumpng in UHV, that i wasn't aware of.
https://en.wikipedia.org/wiki/Titanium_sublimation_pump
>>953175
>bulb shaped envelopes
They use those on some high power stuff I work with. Not sure what advantage they are to small tubes.
>>953715
There are also ion pumps, which use titanium too. It is generally very wasteful to run titanium pumps in pressures above high vacuum.
https://en.wikipedia.org/wiki/Ion_pump_(physics)
>>953175
Me and some friends messed around with what for now is basically a vacuum diode (pic related), and had some issues with getting emission going without burning up our cathodes. There seems to be a rather fine line between getting emission going and the cathode wire burning up (especially if you're a dummy like me and use constant current instead of constant voltage). Do you have any suggestions for materials/coatings that may emit at lower temperatures? We used nickel and konstantan filaments in our tests, plus scavenged cathodes from T15/T30 tubes
Oh, and having access to a glass vacuum bell is super useful for figuring these kinds of things out
>>953647
Never used one. Not a necromancer, sorry ;_;
>>953715
Yep, that's basically how my getters work. Only difference is that I typically don't heat them to the point of evaporation, because at that point pure Ti starts to melt. Commercial TSPs use a Ti/Mo alloy. I actually built a mini TSP for one of the electrodes in my helium discharge tube, though. I found that even a small amount of Ti (1-2cm of 0.3mm filament) can absorb multiple torr of air. Very useful.
>>953725
I was thinking I might need them because I plan on using a section of Ti tubing as a combination anode/getter, which will put out a lot of heat when I "flash" it with the induction heater. With a normal envelope, the glass would only be 2-3mm away from the hot metal, and I'm not sure if it could stand it without cracking. Plus they look cool.
>>953727
Those are neat, but so far I haven't found the need to get into real high-vacuum stuff. Maybe in the future I'll cast and machine a turbomolecular pump out of soda cans or something. At that point I might make an ion pump, because they look fairly simple. I've seen one that's basically a honeycomb of titanium tubing in a box, between two titanium plates with a big magnet on either side and HVDC between the honeycomb and the box. I have all the parts already.
(1/2)
>>953737
If you're just doing this in a bell jar, you can get emission from a pencil lead quite easily, although it will have to be white-hot. I'd recommend 0.3-0.5mm leads. 0.2mm are hard to find and are tricky to work with, and 0.7mm leads take way too much current (IIRC north of 5 amps).
If you're looking for a coating, you'll want to look into oxides, particularly Ca, Ba, and Sr. Not necessarily oxides to start with, but compounds that will decompose to oxides with heat. I've used calcium hydroxide (pickling lime) with some success in the past, just making a slurry of it in water and painting it onto a pencil lead. Got emission at lower temperatures (red-orange heat) but it also tended to ionize the gas in the tube much more easily. The coating was also mechanically fragile and would flake off easily. Ron Soyland uses a mixture of strontium and barium carbonates and seems to be very successful. Be careful though, barium is a neurotoxin. Doesn't accumulate like lead and doesn't make you retarded, but a gram will kill you. Plus, I'm no expert, but IIRC poor vacuum does bad things to barium oxide coatings.
In my experience, it's easiest to just make uncoated "bright emitter" tubes with either carbon or tungsten. If you want to go in an interesting direction where I haven't experimented yet, you could try carbon fiber. Seems like if you used steel for the leads (MIG wire?), you could spot weld it and have the fiber bond to the steel for a solid connection, since carbon dissolves in iron. Just guessing, though.
(2/2)
>>953931
OK, I'd read the barium/strontium carbonate thing elsewhere but those do feel a bit dangerous. Carbon is an interesting idea. My ideal filament would be etched or laser cut on a substrate, possibly with parts of the substrate somehow etched away too (leaving tiny supports)
What do you recon is the tiniest
>>953932
Nah, you're overthinking it. Don't get too high tech, a vacuum tube is just a hot wire in a bottle with the air sucked out. Do as much by hand as you can. As far as small filaments go, the smallest I've used was 0.0006" but that was absolute hell to work with; had to hold it in front of a white background otherwise I couldn't see it. In general, it seems like 0.1mm (0.004") is a good size to start out with. A little power hungry (1-2 amps is typical) but resistant to poor vacuum and easy to handle.
>>953936
Mkay. This is all good fun anyway, hopefully good things will come of it
>One thing I want to avoid that seems to have happened with him though is tool accumulation. While having a turbomolecular pump and a glass lathe would be hella rad, it would also remove my projects from the realm of amateur tinkering. I don't want my page to be "look at the cool stuff I did because I'm awesome", I want it to be "look what you can do too because it's easy and cheap".
I guess that you are fully aware of it, but I dig your website man, it simple, and conveys exactly your presentation method.. I really apreciate your work, even thought I would not dabble on it myself as of now, but I know you'll end up making me want to soon enough... Keep it up!
If you have this vacuum pump have you tried a radiometer?
>>953964
Thanks, you seem to get the idea. If you ever do decide to get started, keep in mind you don't have to do things as fancy as I did. A pump and a pickle jar is an adequate starter kit, and don't be afraid to use hot glue.
>>954698
I'd be more interested in radiometers if they actually worked the way people thought they did (light momentum) instead of by convection of residual gases. Can't see myself making one, to be honest. A photocell, however...
In other news, it's time to get started. The titanium tubing got here yesterday, and my new power supply is getting delivered today. Gonna go ahead and make triode #3 as soon as it gets here. Probably won't have an article done today, but expect an update on triode #2 and pictures of the new one.
This, all of this, is awesome and YOU OP, are a frickin wizard.
Have a gratuitous bump and my never ending interest.
>>954837
Thanks, glad you like it. I have all of my equipment ready now so expect lots of content in the coming weeks.
Updates:
Power supply came, works great, read the manual this time and turns out you're not supposed to start those things under load. Whoops. Explains why the last one died so young ;_;
Triode #2 still worked when I lit it up today, but tended to ionize at high plate current. Speaking of which, the plate current was triple what it was before for any given value of filament power, but after reheating the getter it dropped back down to its normal range. Maybe Lee de Forest was right, and a small amount of gas does make the tube more sensitive. Interesting.
Triode #3 is in the oven, gonna seal it off in an hour or so. Pic related is its insides. Much more tidy than previous audion-style tubes. I'm using a smaller filament this time; I hope I can still get close to 1mA plate current with it. Decided to use a bulb envelope, hope it doesn't collapse in the oven.
More updates coming later tonight.
Triode III is a success. Has better emission than triode II with less filament power required. Plate resistance is around 60K at 90V; I can get 1.5mA of plate current with the grid tied to ground. If I tie the grid and the plate together, I get 4mA. Kind of want to try higher voltages on the plate but I also don't want to blow it up. Might experiment in that area once I'm sure the tube has a stable level of vacuum. Grid response is somewhat nonlinear, but this might change as the tube stabilizes. I made a little audio amp with it and while it can definitely produce audible sound from a speaker, it's still limited to around 100-150mW, which is tiny. I might order some thicker tungsten to see how much power I can fit into this size of tube.
Construction-wise I'm really liking this new design. It's easy to assemble, and the 10mm titanium tubing is spacious enough that a tetrode is definitely possible if I'm careful. The spring-style grid seems like it might cause some microphonics though; I'll check that out tomorrow to see if I need to add more support there in future tubes.
I'll probably keep checking on it over the next three days and have an article up by the weekend.
>>954986
sexy
I was thinking about grid materials. How would colour CRT shadow mask work?
>inb4 mah residual charges in the CRT
If he can build and use these I think he is smart enough to deal with that.
>>955154
Seems like it would work alright, but honestly winding a little nickel spring like I did for this last tube seems the easiest; pic related. That way I have control over the gap spacing as well, which to a certain extent determines plate resistance and transconductance (tighter spacing = less electrons get through but more controllable).
>If he can build and use these I think he is smart enough to deal with that.
I'm not as smart as you think, I've discharged a CRT through my chest before (leaning on the metal chassis while discharging the anode cap with a poorly grounded screwdriver). Not saying CRTs are particularly scary, just that I'm dumb with HV and get zapped pretty often.
In other news, testing of triode III continues today, and since that'll only take a half hour or so I'll do a little side project as well. If all goes well I might have an article on that today or tomorrow. Triode article is still scheduled for the weekend.
Seems like thsi thread reached its limit? It's slowly sinking to page 9 :(
>>956188
Thanks for the bump, looks like it's not at the limit yet.
Side project hit a snag, so I'll work on it later. Wrote the triode III article a day early because the tube has been stable since day 1, no point in waiting an extra day.
http://simplifier.neocities.org/triode3.html
I also figured out how to use paypal and added a donation link at the bottom of the index page, since one anon was interested and nobody objected. There's also an e-mail address as well in case this thread ever dies. I'll try to remember to check it.
>>956213
Thank you for documenting the work you do. It's always awesome to have resources like these online as a source of learning for others.
>>946133
>http://simplifier.neocities.org/
You deserve a medal dude! I am proud of DIY for once.
Don't you 404 on me!
>>956371
>>956990
>>957965
Thanks for keeping the thread alive. I've been working on a few side projects and I should have an article or two up by the end of the day. I'll probably do 3 or 4 non-tube projects and then go back to tubes. Once I get back to that, the schedule should be:
- Power triode
- Tetrode
- CRT
Assuming all goes well.
>>954986
oh fuck that would make the loveliest looking headphone amplifier
>>958107
I was thinking that, but I don't have anything that needs amplifying on that small of a scale. I'm assuming a headphone amplifier would be for a record player or something, where the output is low voltage and high impedance.
In other news, the side project is done and the mystery is revealed!
http://simplifier.neocities.org/battery.html
It's a battery-based medium-voltage supply. Nothing very exciting, but maybe someone will find it useful. Pic somewhat related, it's a little part I made to connect the switch to the box.
>>958320
you can connect it to your computer and use it to drive your headphones. there are quite some people who do this because your onboard soundcard can drive the headphone but often does a poor job at it because it isn't really made to drive heavy loads.
or just do it because you can
For the love of god, make a headphone amplifier with two of those tubes
>>958444
>>958575
I'll consider it in the future, but for now the goal is to make a tube that's strong enough to drive a speaker (1W+). I don't have any speakers for my computer, so it would be something I actually have a need for.
Did another mini-project today. Made a full-wave rectifier out of metal and glass, but not in the way you might think.
http://simplifier.neocities.org/rectifier.html
Would love to follow you on facebook mate. tubes have always facinated me, and at one point considered building my own untill I saw how complex it is to start up.
>>959493
Don't have a facebook, sorry. 4chan and youtube are pretty much the only websites I go to anymore. You aren't missing out on anything, if you're wondering. Everything significant that I do either gets posted here or on my site.
Daily updates:
Working on one last non-tube project; it's an improvement to the Joule Thief, if anyone's familiar with that circuit. I'd explain more, but I don't want to ruin the surprise :^). Should be done tomorrow.
>>959525
boost converter to suck the last life out of dying batteries, ammirite?
>>946133
>http://simplifier.neocities.org/
How old are you OP?
![book1[1].jpg](https://i.imgur.com/mwVdNGE.jpg "book1[1].jpg")
you are truly an inspiration.
I'm an electronics tech that's been dabbling with tubes on my free time.
I've built a few amps and stuff, but this so far beyond any of that.
i got pic related in the mail yesterday, from the valve wizard.
what sources would you recommend for learning more/design blocks/referances/schematics?
so far the valvewizard.co.uk has been great.
I'm into instrument amps, but also want to mess around with synth stuff.
got lot of sub-mini tubes for cheap some time ago.
cheers!
>>959493
>follow you on facebook
[cringes internally]
>>959535
Right, that's what the standard joule thief is for. Mine's a little different, though. A little more general-purpose.
>>959538
I'm 24.
>>959539
I don't really have a go-to place for schematics, usually I just mess around in a simulator until it works right, sometimes LTspice but usually this:
http://www.falstad.com/circuit-java/
When I'm doing research I typically don't have a certain project in mind, I just read about different circuits that do different things on wikipedia or wherever, and come back to them as I need them. I tend to sacrifice quality/efficiency/performance for simplicity though, so if you're more into professional-tier circuit design you'd probably have to do things differently.
Article's done. It's fairly text-heavy, but I'm going back to tubes after this so I'll have more pretty pictures soon.
http://simplifier.neocities.org/joule.html
>>959719
>http://simplifier.neocities.org/joule.html
damn I didn't have glassblowing supplies and a lathe at 24
>>960085
I got the lathe about 5 years ago from harbor freight, it was on sale for $350. My "glassblowing supplies" are a $10 propane torch from home depot and a $15 kaowool blanket from ebay. It really doesn't take much to get started. You don't even need a lathe for most of the stuff I do, I just use it because it makes things look nicer.
OP
what do you use to draw your schematics? If it's just a component template would you post it here? I love the simplicity of your site.
>>960240
I just use paint. I used to draw each schematic by hand, but I wanted to start having a uniform style and component size so for this last article I made a little tile map. It's far from complete, but here it is.
>I love the simplicity of your site.
Thanks, that's what I was going for. Too many websites are full of stuff that isn't content, which makes the actual content harder to access. I was trying to avoid that, and at the same time integrate a little bit of a 90s-internet feel to it. Glad you like it.
>>960085
In fact it is much easier to aquire expensive or fancy stuff when you're young, especially when still living with parents. You don't have to manage your own finances, all of te money you earn you cna spend on your hobby. All it takes is not be be an idiot and save your money instead spendign it on stupid stuff/booze.
>>960275
Apparently the supplies are much cheaper than I thought.
Maybe I missed it but what kind of vacuum pump did you use.
I probably have a lot of this stuff within reach due to my work, even a glassblower but those guys are so overloaded with work that they are not going to do fun projects for me
>>960348
I use a two-stage rotary pump, that's it. No diffusion or turbomolecular pumps or cold traps or anything fancy. My setup is here:
http://simplifier.neocities.org/vacuum.html
And the pump is here:
http://www.ebay.com/itm/1-3HP-3CFM-2Stage-Rotary-Vane-Deep-Vacuum-Pump-AC-R410a-R134-Freon-HVAC-Air-Tool-/121505638978?hash=item1c4a4cea42:g:BXMAAOSwjVVV3z6m
The pump is the only expensive part that's required. You might even be able to get away with a single-stage pump (~$50) if you're low on funds. Titanium is an excellent getter and can take a lot of abuse, pressure-wise. For a gauge you can make a geissler tube, which is also on my site. Other than that the costs are just titanium (~$6 for 50ft of wire), tungsten (~$10 for 30ft of filament) and glass (~$2/lb from FMS/Brillite). I use an induction heater but it isn't necessary, you can heat the titanium other ways if you need to, and a spot welder can be made for about $20.
The whole point of me making these tubes is to show that it can be done cheaply. If you're interested in getting started, start with the vacuum chamber article on my site and read through my progress. I explain a lot of the details and issues involved.
>>960356
>Mandatory reply
Thank you mr.Tublar
I've browsed /mu/ for the last few years and I'm so sick of the very, very apparent decline in quality. I've also been hit really hard with the urge to start working on electronics and learn about music gear on a much more intimate level. I absolutely love the fact that there's a community for this sort of stuff on this board. Cool shit OP.
>>960356
>>960356
>Thank you mr.Tublar
>>960399
As a sidenote, about /mu/, what kills me is the fact that many music equipments offer quality than simple stuff made in the breadboard at home using readily available cheapest components... It just shows how much value is shave dof off this stuff in design and production stages that they give you such shitty performance...
>>960257
To draw schematics either use Latec (beautiful but difficult)
Or use LTspice, and alter the schematic settings so the symbols is thin lines and text is times new romance..
[OFFTOPIC]
OP, can you explain this to me.
why is there a short between Base and Emitter.
I looked up joule thief in wikipedia[en] and they dont have this there
https://en.wikipedia.org/wiki/Joule_thief#Description_of_operation
>>960733
Haha oh man, good catch. I made the last schematic and worked backwards. I'll fix it.
>>960353
I do have access to both oil diffusion pumps and turbomolecular pumps. This seems like a fun project
Do you expect to get a usable tube, as in it works properly and can last a long time in an amplifier?
>>960841
You're good to go, then!
>>960861
Hopefully, yes. Currently I know for sure that the tubes will last at least a few weeks, but they may need to be "refreshed" (by reheating the getter) later to trap any gases that may have been released over time. I'm assuming they will reach a point where no more gases are being released from the components of the tube, at which point the tube will be stable indefinitely (until the filament burns out). Either that, or perhaps the titanium getter does have some functionality even when cold, and they won't need to be refreshed at all. Time will tell. As for working properly, they already do; they're just low-powered (~100mW). The next tube I make will be a power triode, and I'm shooting for 1W which I think is definitely achievable.
Updates:
Once again waiting on parts from china. I have some 8mm OD, 6mm ID titanium tubing coming in the mail that I want to use in the next triode (plate closer to filament = lower plate resistance); according to the tracking page it should be arriving any day now, possibly monday.
>>960356
Thank you Mr. Tublar
>>960356
Thank you Mr Tublar
>>959539
Merlin and Randall Aiken are two of the best. I've gone through most of that book multiple times and I'll still dive in and take notes if I'm fine tuning a piece of equipment. Aiken's site, aikenamps, has white papers and technical q&a. All the information is outstanding though kind of scattered around. Some of it is not even linked on the page, but I find redirects from random searches or forums.
Once you're through Designing, you'll have a good footing on how to determine how a preamp valve is going to react. You'll want to grab another reference to design your synth front end, but making the Hammond-like preamp and power amp should be pretty smooth sailing from there out. To boot you can grab a near functional chassis for fifty bucks any day of the week.
>>961944
>aikenamps
haven't seen this before, thanks mang.
>>959719
you said earlier that you had some hard time to get tungsten wire, it is very common in my country . Wouldyou liike some ?
>>961966
Best theme song ever
https://www.youtube.com/watch?v=g3ggYGn8ssw
>>962140
I think you're thinking of the guy from Argentina. I have no problem getting tungsten, I order it on ebay and it gets here in 3 days. Thanks for the offer though, maybe if the argentinian anon is still here he could take you up on it.
Updates:
Still waiting on the new titanium tubing, but I've been doing other /diy/ related things in the meantime. Went for a walk in the woods and found a pair of long-handled pruning shears; judging by the degree of rust and rot they must have been there for a decade or so. They cleaned up nicely though, and I made new handles since the old ones were 90% mush. Nothing article-worthy though. If the tubing still isn't here tomorrow I'll start working on making a table saw from an old circular saw I have; I need one for my home shop so I don't have to drive across town to use my dad's. If that happens I'll document it, since being able to make a table saw on the cheap seems like something that would be useful for /diy/ers.
Updates to avoid a 404:
Decided to start making the table saw on tuesday, and I should be finishing it up today. The new titanium tubing came yesterday, and it's the perfect size. I should be able to use a standard (non-bulb) envelope this time. If I have time I'll start on the power triode tonight, after the saw is done. Expect multiple articles within a week.
You really motivated me to look into this. All the stuff seems easily available and seems extremely cheap compared to my other projects (which are running for years because it's hard to find cheap UHV parts) and i already have the needed vacuum pump. Maybe someday i can make a gauge for a controller i already have.
>>963507
If you need a quick-and-dirty gauge, read the geissler tube article on my site. It should be adequate if you're just doing what I do (low-end tube-making).
Updates:
Spent today working on the table saw, and it's almost done. Taking a break for an hour or so, then I'll probably start working on the power triode, since it's dark out and I don't want to get sawdust everywhere in my workshop by doing wood things inside. Next three articles are probably going to be (in this order):
>power triode
>table saw
>cheap HVDC power supply (induction coil equivalent, for use with the geissler tube gauge and possibly x-ray tubes)
Beyond that point I'll probably make a few useful devices with triodes before moving on to tetrodes.
>>963571
I have a pirani gauge (like in the pic) which should be good to 10^-3 torr. I have another controller (i think it's for penning type gauges) which goes as low as 10^-7 torr for which i'd like to have a gauge someday.
>>963580
>I have a pirani gauge (like in the pic) which should be good to 10^-3 torr.
Oh, that's plenty then. My thermocouple gauge has the same range.
Updates:
Assembled the power triode (pic related). However much power I can get out of this tube will be the maximum I can achieve with current envelope sizes, at least with a plain tungsten filament. I don't think I can use any smaller tubing for the plate; the 8mm tubing was a tight fit. I also don't think I'll be able to use any bigger of a filament; this one (0.15mm) should draw 3-4 amps, which is about the maximum I can put through the 0.8mm feedthroughs without them overheating. I also used 0.3mm nickel for the grid spiral, which is about as small as I can go without it being too weak to handle without bending it. Going to evacuate it tomorrow.
Rough day. Lots of casualties.
Built two power triodes. The first failed as soon as I lit the filament; I didn't have the filament in sufficient tension so it warped as it lit up and shorted to the grid. The second I was able to test briefly, but the large filament generated a ton of heat which caused it to go gassy very quickly (I'm assuming due to trapped gases in the grid). I reheated the plate with my induction heater to absorb the gases, but overdid it and the heat warped the grid, shorting it to the filament support. Triode #3 also died; I tested it without reheating the getter first, and it lit briefly but then the filament burned out. Tested triode #2 after reheating its getter, and it was fine. Seems like I'm having longevity issues, which is to be expected with the quality of equipment I'm using.
On the plus side, both power triodes easily reached 10mA plate current, and the second one reached 30mA when I ran the filament up to 4 amps (destructive testing after the grid had shorted). It also had a transconductance of over 1,000µmhos, for the brief period of time when it was stable. Pic related.
I'm going to spend some time rethinking the design, and in the meantime I'll get back to finishing up the table saw.
I think i just found myself a new hobby. Like i didn't have enough of them already.
Also bumb.
>>965459
Thanks for the bump, and good luck.
Table saw is done, pic related. Now I can make more things out of wood.
http://simplifier.neocities.org/table.html
Going back to tubes tomorrow. I'm thinking of experimenting with an oxide-coated titanium filament. I'm hoping it will start emitting before the titanium starts to sag (if loose) or stretch (if under tension). If it works, the filament will double as the getter and I will have solved the problem of the vacuum degrading over time, at the same time eliminating the need for an induction heater. I might also get more emission at lower filament power, improving the efficiency and maximum power of the tube. This is all wishful thinking though. If it doesn't work I'll go back to tungsten filaments and make another triode similar to triode III (aesthetic bulb and all), with the grid parameters tweaked to improve power output.
>>960356
Thank You Mr. Tublar
I hope you realise if you can make tubes reliably and make them look pretty enough, you could sell them to audiophiles for 5-figures each? Some of them would go nuts for a unique, hand-made, tube. Because, you know, they are nuts.
>>965834
Nah they want the old tubes. This is exactly how it sounded to grandpa sort of nostalgia shit they are going for
>>965806
Thank you Mr. Bumper
>>965834
I'd rather not bother with that. I don't want to get tied down to one thing, and becoming a tube manufacturer would take up time that could otherwise be spent experimenting in other areas. Besides, the whole point is that they're easy enough to make on your own. Me selling them would make people less likely to make them and more likely to buy them, which is counterproductive to my goal of helping DIYers.
Updates:
Ordered some 0.15mm titanium wire and some barium and strontium carbonate for making coated filaments, funded in part by a guy who donated (thanks man). I previously said I wouldn't work with barium due to its toxicity, but turns out barium carbonate is only poisonous orally, and the antidote is epsom salt which I have on hand. I doubt I'll accidentally eat any, though. While all that is being delivered, I'll experiment with titanium lathe turnings (it makes nice spring shapes like aluminum) coated in calcium hydroxide as a substitute. I don't expect much success, but it's something to do.
Tested some titanium filaments in the vacuum chamber, with very interesting results. I decided against using lathe turnings; I had some 0.3mm titanium wire and used that instead. Turns out due to the high resistivity of titanium it only needed 2 amps to start emitting, with a maximum of 3 amps where it started to sag. I made a simple diode and tried two filaments: one with a coating of calcium hydroxide, another that was just plain titanium. The coated filament had excellent emission at the start, but it declined over time and when I opened the chamber I noticed it no longer had a coating, although I never observed it flaking off. I think I may have run it too hot. The plain filament had stable emission, and I could get about 1.5mA plate current at 40V. I'll do more experimentation with oxide coatings, but even plain titanium looks promising as a filament material.
On the other hand, I noticed I could easily reach 5 millitorr with the chamber attached, but I usually only get to 10-20 millitorr when I'm evacuating a tube. Either my o-ring adapter is leaky, or my feedthrough seals are. If it's the seals, that may explain why my tubes tend to require re-gettering after a week or so, and I'll have to improve my pinch seal construction, perhaps by grinding the wires at the sealing area to remove the thick oxide layer which may be porous enough to let air through. If it's the o-ring adapter then it's an easy fix. More updates tomorrow.
>>960356
DenkYu Masta Tabulor
Still waiting on the 0.15mm titanium wire, so I made a little portable high voltage supply in the meantime. Used it on some old dead tubes and turns out the majority of my tubes are leaky. Going to have to work on my seals.
http://simplifier.neocities.org/spark.html
If the wire's still not here tomorrow, I think I'm going to try to make a coherer (pic related), since I now have something capable of generating a spark for it to pick up.
That was actually really easy.
http://simplifier.neocities.org/coherer.html
Short article, but it has some pictures and useful information if anyone wants to build their own. Really straightforward method of remote control on the cheap, if you just need to turn something on.
>>946133
Anon, you're making me feel like a very lazy person
good stuff. thank you for sharing!
pic unrelated.. just a glimpse of my country's future
>>968245
Thanks, anon.
Updates:
Just got the 0.15mm titanium, barium carbonate, and strontium carbonate I ordered. Going to finish repairing my signal generator (accidentally fed 40VAC into the output and cooked the power transistors), then experiment more with oxide cathodes and possibly make another tube.
Signal generator is perma-dead, RIP. Replacing all the transistors didn't work so I gutted it for parts.
In other news, made a barium carbonate coated filament and tested it in the vacuum chamber, with excellent results. Could get emission (only 0.1µA, but it's something) with the filament barely glowing a dull red. At higher temperatures I could get multiple mA, but this was tricky since the vacuum chamber runs at a higher pressure than the inside of a typical finished tube so the residual air would ionize at high currents. Titanium seems to work well as a filament material when coated, but has the tendency to evaporate at high filament power, coating nearby surfaces. Also, the 0.15mm wire turned out to be more like 0.05mm, so for now I'll probably be sticking to 0.3mm since with the barium coating it only requires a little over an amp for emission. I also made a triode configuration in the vacuum chamber using the coated filament, and it worked well.
I think I'll make a tube tomorrow; expect an article within a few days.
Made two triodes yesterday, with 0.3mm oxide-coated titanium filaments. Both failed, but I learned from it so no biggie.
In both cases, the filament behaved strangely. It would start with no emission, then after a minute or so of being brought to a dull red heat, the emission would start climbing and would plateau at half a milliamp or so. Then this would decrease to nothing (I assume something was degrading the oxide coating) at which point I would raise the filament temperature to an orange-white heat, and would get around 1-2mA of emission. I believe this is just from the plain titanium. This was barely usable (I could only get 1-2mA if the grid was tied to the plate) so I increased the filament power until it evaporated and went open circuit. This made a thin semi-reflective coating of titanium on the inside of the tube (pic related is the second triode I made yesterday). In future tubes I will experiment with this as a getter.
I also believe I solved the leak problem. I had been using the titanium wire as-is for the pinch seals that ended up leaky, so this time I ground off the native oxide layer and didn't pre-oxidize the wire before sealing. I also heated the seal until the glass began to gather together and wet to the titanium. As a result, I was able to get both tubes down to 5 millitorr on the pump (the minimum achievable with my setup) and noticed no change in manifold pressure after sealing off the tube, meaning whether or not the feedthroughs were present made no difference, i.e. the seals were probably good.
I think I'll go back to tungsten filaments now, but with better sealing technique and perhaps with an integral titanium filament-getter instead of using the induction heater. In the future I might do more experiments with oxide coatings, using a thicker mix of barium carbonate so it doesn't degrade as quickly.
So tube anon, you ever think about making nixies?
>>970331
I might make some glow lamps eventually, which operate on the same principle, but I don't think I'll make any sort of numerical display. It would be a ton of wires, and I'm not a big fan of digital stuff.
>>946133
Do you have any experience making Nixi Clocks?
Every university chemistry department has a glassblower and supplies, FYI
>>970433
Nah, the only tube experience I had prior to making my own was messing with old CRTs. Never really understood why nerds love nixies so much, to be honest. If I need to tell the time I look at my phone or computer, and if I wanted a clock purely for aesthetics I'd get a mechanical clock.
Updates:
Working on another triode; went back to using a 0.1mm tungsten filament, but with a much wider spaced grid (about 80% open space) than triode III so it'll be optimized more for power than sensitivity. I used my improved sealing method which I'll describe in detail in the article if it works. I decided against using a filament getter; I didn't want any evaporated titanium making internal shorts, and triode II (which used a non-evaporating getter) seems to still be working well, almost 50 days since I made it. If this new tube seems like it'll last, I'll make an audio amplifier with it. I just put the internals inside the envelope, and it's annealing now.
>>970262
It is often said that the oxide cathodes are easily killed by the residual gases, particularly if you draw so much current that the cathode emission sets the current limit. Apparently all the other filth has an effect too, including the purity of the filament wire and the oxides themselves.
Instructions for making them emphasize cleanliness and rather long bake-outs at low pressure. No, I don't know if this is the reason behind your problems. Might be just old-fashioned adhesion problems. Or something totally different.
>>970468
Not anymore.
I now have my first working power triode (pic related).
Sealed it off, heated the plate/getter, and instead of testing it quantitatively I immediately made an audio amplifier circuit, using just an input transformer (to step up the input signal to drive the grid), the tube, and an output transformer (to drive a speaker). I played Ave Maria through it and it was quite an experience. No obvious distortion, and definitely loud enough for multiple people to listen to, given that the room is quiet. Measured the plate current while it was running and it averaged around 6mA at 90V. Maximum plate current (grid tied to plate) was a little over 11mA, which hits my 1W goal.
Assuming it doesn't leak and die overnight, I'll write an article on it tomorrow with more quantitative data.
Whoops, forgot to reply.
>>970812
>It is often said that the oxide cathodes are easily killed by the residual gases
Yeah, I read this too. Didn't have high expectations for the coated filaments due to my crude setup, but I figured I'd give it a shot anyways. I might do some more work with them later, but for now plain filaments definitely seem far easier to make and operate. From what I've read, even properly made oxide cathodes are fairly delicate and don't like variations in filament power. Plain filaments on the other hand can have the power adjusted all over the place so long as they don't evaporate or melt.
>>946133
Congrats, OP, you're doing the Lord's work here.
Especially interesting to me are your discharge and Geissler tubes and the bulb-style triode.
>>971015
Thanks. I might go back to gas-filled tubes later as a side-project. The two helium tubes failed pretty quickly from leaks, but I know more about making good seals now so I might try that again; I've still got a helium tank and a bunch of 6mm tubing sitting around.
In other news, new article:
http://simplifier.neocities.org/power.html
>>971241
Haha, should've waited until after april fools' to post.
Hey OP, recently at an auction I picked up 3 pretty sophisticated vacuum pumps, two large (as pictured) and one small. Would you be interested in one of the larger ones?
OP here.
>>971325
Oh wow, that does look pretty nice. Looks to be the type that can get down to 10^-4 torr. If I had the space I might be interested, but my shop is pretty small (about 10x10 feet of working area) and I just don't have the room for it. Plus, my current pump works fine for what I'm doing, and I like working with low-end tools to keep my projects realistic for amateurs. Don't want to bring the tool requirement list up too much. I'm sure there are other anons in this thread who would be interested though.
>>971334
Yeah, I picked up two of those, plus a smaller one(pictured) for a total of $70. YMMV however while trying to buy these elsewhere, from what I remember you spent 90ish on yours. Just letting anon's know that if you wait around you can pick them up for quite cheap sometimes. I passed on one at a pawn shop after talking him down to 90, mainly because the oil fill level was way too high. Right now I'm trying to set-up my own homelab, was really inspired by your work. Farnsworth fusor is the end goal.
>>971344
Hot damn, that's quite a find. That first pump you posted is worth at least a grand new, and seems to be in good condition. Would be great for a fusor. I tend to just order things off ebay because I'm a hermit and don't leave the house much. Plus I don't live in a big city or near any research facilities, so the most I'd find in a pawn shop is old power tools.
Do you plan on trying to improve the fusor design at all? I know standard fusors require more energy in than they put out, so they aren't useful for power plants. It would be neat to try to improve the efficiency. I don't know much about fusion though, so I could be suggesting something very difficult.
>>971355
>so I could be suggesting something very difficult.
Different anon here. Yeah, that's an understatement. The main problem with fusors is excessive losses when particles hit the grid. The polywell concept attempts to fix that flaw but I remain very skeptical about it.
>>971355
I'm sure that if you were capable of actually net gaining energy from a fusor, your idea would be worth billions. It's more than difficult atm. I do plan on playing around with it, as it seems like an awesome way of getting started with fusion. Which state are you in? I made a 3 hour drive to pick up those three vacuum pumps, an oscilloscope(pictured, purchased for $15, second one on bottom rack doesn't count, I've had that for a while and display functions, won't input or modulate, do plan on fixing it), and two HV Op Amps (Burleigh pz-70)
>>971363
While I'm here, does anyone know what I can use the Op Amps for? I bought them on a whim as I do some stuff with high voltage, but I have no clue how'd they be useful for me as I don't do anything piezoelectronic.
>>971369
I just missed out on an auction in Spartansburg, SC because it was a 6 hour drive, there was a nice 8"x28" machining lathe that went for $160. I wanted it soooo bad, they also had a vacuum pump that went for $10, same as mine. I'm in Norfolk, VA.
I don't think I'll make an electron microscope, as my interests are more around medium-high energy plasma physics(math major). I'm sure I can find some use for it, although the manual says it is for low power applications(relatively).
>>971374
Where do you find these auctions? I don't have any equipment I need right now, but if those prices are typical it'd be nice to know where to look in the future.
>>971378
rasmus, yellowtagauctions, governmentliquidation. Those are the ones off the top of my head, you should be able to google-fu your way into finding more however. Yellowtag is based in SC, so I'd predominantly check them for stuff you might want to get.
On the topic of vacuum pumps, do you plan on making a diffusion pump in the future? It might help with reduction in the need of a getter in your vacuum tubes.
>>971383
>rasmus, yellowtagauctions, governmentliquidation
Thanks.
>do you plan on making a diffusion pump in the future?
I've never been a big fan of the idea of a diffusion pump, to be honest. I've had some unpleasant experiences with oil spills, so a big tube of boiling oil doesn't sound like something I'd want to have under my work table, especially if it's homemade. If I do make a high vacuum pump, I'll probably try to make some kind of molecular drag pump; Claude Paillard made one from scratch for his tubes and it seems like it worked well. It'd be fun to cast the parts from melted down aluminum cans.
Haven't found the need for one yet, though. If I do hit the limit of my current pump/getter system, it will probably be if/when I try to make an x-ray tube, since those require an extreme level of vacuum. Titanium getters seem to work impressively well, though. I've done some preliminary testing of the level of vacuum in a freshly gettered tube by connecting unpowered tubes to my spark box, and in all cases the spark prefers to jump between the leads rather than through the inside of the tube, indicating there isn't much of anything floating around in there to conduct through. Plus, even with a higher level of initial vacuum, a getter of some sort is still necessary, so the only benefit of having a fancy pump would be the ability to do more bell jar stuff.
>>971391
Thanks for clarifying with the second part. I do plan on going through most of what you have described, right now I'm trying to figure out the best way of getting a glassblowing setup to work nicely. I do have a woodlathe to use as my workspace. Do you think beer/wine bottles would make decent vacuum tubes out of? I have managed to pinch the tip off with a lot of heating and effort, and avoided cracking the beer bottle whilst cooling it down.
>>971398
Haha strangely enough that's been in my project backlog for awhile. I've been wanting to see if my neon sign glass could be fused to a beer bottle, but the problem is I don't drink beer. It'd be trivial to find one, I just haven't gotten around to it. Regardless of whether it works or not though, bottles shouldn't be your staple material if that's what you're getting at. There's too much variance between manufacturers, and even between batches from the same manufacturer. You'd be giving yourself glass problems in addition to the vacuum and sealing problems you'll already be encountering.
If you need a source of glass, Brillite has the best prices I've encountered ($2/lb + $20 shipping) and their glass is wonderful and can be worked with a propane torch. I'd suggest 5lbs of 5/6mm, 5lbs of 10mm, and 5-10lbs of 18mm to start with (5lbs is the minimum order for each size of glass).
A wood lathe might work, if you can run it at a low enough speed. For glass stuff I typically run my lathe at a 1-3 revolutions per second (60-180 RPM). Any faster and the glass slings off to the side when it's molten. Since there's no torque involved for glasswork, you could probably just rig up the belt to a little DC gearmotor and that would be fine.
>>971402
Oh wait, there is one exception. I do have to run the lathe at normal speed when I'm making flares.
Also, I forgot to note that you will need a chuck for your lathe, and if your lathe doesn't have a spindle bore you'll need to make a jig for scoring the glass. Something like this (at the bottom of the page):
http://www.teralab.co.uk/Glass_Blowing/Hot_Wire_Tube_Cutter/Tube_Cutter_Page1.htm
>>971402
Yeah, I was planning on turning another set of pulleys for really slow turning. I'll try my luck at working with the salvaged glass first, I'll post my results here. I don't drink much beer myself, but it helps that my renters do. I don't recall if this book was posted on your website or not, but I would recommend this as a read to anyone https://ia800304.us.archive.org/0/items/ProceduresInExperimentalPhysics/Strong1938ProceduresInExperimentalPhysics.pdf
It does seem that copper can be used as your through wire if you flatten it nicely.
>>971406
I devised a clamp to secure odd sized objects that is suitable for slow turning. Flares detailed in the book can be made at slow speeds. Scoring the glass shouldn't be too big of a hassle.
>>971407
Never seen that book before, but it seems extremely useful. I'll skim through it tomorrow.
>It does seem that copper can be used as your through wire if you flatten it nicely.
I've considered that before, but titanium just seals so nicely that I haven't bothered with it. Plus since I'm using titanium for the getter, and in my experience titanium only welds to nickel, using copper for the feedthroughs would require a copper -> nickel -> titanium weld chain, and I'm not even sure if I'd be able to spot weld copper to nickel due to the low resistivity. I'd probably have to make a capacitive discharge welder. Plus making the seal would soften the copper, causing things to possibly shift inside the tube. It would just over-complicate things in general.
I'm out for the night, good luck with your work.
>>971418
I plan on using other getters (carbon based) in my tubes, you can electroplate copper with other materials if it doesn't possess the right properties, or solder to it (EEK, obviously not a good idea for high temps).
I do plan on having the thread open in a tab, but if you would like to contact me my alt-email is [email protected]
>>971426
>getters (carbon based)
I wasn't aware that this was a thing. The closest I've heard of is using activated charcoal as a sorption pump. Care to elaborate?
>>971823
http://accelconf.web.cern.ch/AccelConf/e98/PAPERS/TUP03E.PDF
>>971889
>http://accelconf.web.cern.ch/AccelConf/e98/PAPERS/TUP03E.PDF
Oh, so carbon nanotubes and stuff. Seems tricky.
Updates:
The power triode shows no signs of leaks or outgassing whatsoever. This may be the first properly sealed tube I've made. I'll wait until it's been a week and then test it more thoroughly. If all goes well, at that point I'll start building an audio amplifier with it.
On a side note, I have a lot of thinner filament (.001") that I haven't used much of. I might order some even smaller (6mm) titanium tubing and try to make a miniature triode that fits in my 10mm glass for the envelope. That way I can make little preamp tubes, or tubes for portable devices since the .001" filament only draws around 150mA.
>>972836
Ordered some 6mm OD 4mm ID titanium tubing for making miniature triodes, should be here in a few weeks. Made an experimental seal with 4 leads through some 5mm tubing (pic related). Turns out it's not as hard as I thought it would be; since the whole thing is smaller I can get the glass really hot, so it wets to everything really well and gets between all the feedthroughs. Even if two feedthroughs are close together they don't end up shorting.
I also found a really good deal on some carbon steel piano wire, so I ordered some of that as well. I was thinking about some uses for a miniature tube, and a pocket-sized wire recorder seemed like a fun project; sort of like a 19th century mp3 player.
The power triode still seems to be in good condition, so I'll start on the audio amplifier in two days after I do some final testing. All in all, lots of projects coming up.
>>973366
jesus fucking christ
If you were Tom Hanks in Castaway, how many years would it take you to make a radio?
>>973376
Years?
Just kidding. I really don't have any exceptional skills; what I do is just a matter of taking theoretical things people already know, and just doing them. Like, everyone knows that to cast metal you just melt it and pour it in a hole. So I do that. Likewise, everyone knows vacuum tubes are just glass and metal with the air sucked out, so I just make them.
Most things are exactly what they look like, and go together exactly how you'd expect. The more you overthink things, the less you end up accomplishing. Technology is generally simple, or can be made simple by generalizing (removing case-specific add-ons) and de-optimizing (trading performance for simplicity), so just build stuff and figure out the details as you go.
OP: ever consider making an integrated circuit tube?
>>973459
Not really. I'm a big fan of general purpose components, so an IC tube doesn't seem like something I'd want to make. Probably the closest I'd get is something like a 12AX7, where there's two triodes in one tube. I might be able to do something like that with the smaller titanium tubing that's coming in the mail. That would take at least 6 feedthroughs though, and currently the most I've done is 5.
Another reason I don't think I'll make an IC tube is just due to the fact that the circuits I'll be making are so dead simple that there'd be nothing to integrate. For example, my amplifier probably won't even have any resistors, just two transformers, the tube, and the power supply (which may just be batteries). Tubes, especially homemade ones, really aren't picky like transistors are about what's fed into them. Sure you could keep adding components until everything's 100% optimized, but that's not my thing.
Was looking at the power lines outside today, and decided to make one as a side project.
http://simplifier.neocities.org/singlewire.html
Should be starting on the tube amp tomorrow, if all goes well and the power triode isn't dead. This is the first time I've let a tube just sit for a week instead of messing with it constantly. Testing it with high voltage it seems okay, but I haven't lit it up since I first made it, so I don't know for sure.
>>973882
Micrwave transformers are rather inefficient due to the flux shunts.
http://members.westnet.com.au/page3/single_813_G2DAF_RF_amplifier.htm
>>973891
Wasn't sure what effect they would have at low power; I don't have much knowledge of magnetic circuits. I know I knocked out the shunts when I made my spot welder because I was under the impression that they were some sort of current limiter. Plus I just wanted to keep things simple; no point in chopping up two transformers for a day project. Good to know, though.
>>973900
Some of the guys who use them on high duty cycle run them at half voltage. They are designed to have a squareish output being run into saturation under normal circumstances.
The triode still works exactly as it did a week ago, so I'm pretty sure the new sealing method was a success. Starting on the audio amplifier. Made a little boot for the triode (pic related) so it can be mounted on a piece of wood. Just spray painted it black and will clearcoat it to try to get that old tube base look. I'll take pictures as I make progress on the amp as a whole.
>>974416
Looks really cool. Good look with the Amp.
>>974613
Thanks.
Updates:
Finished painting the tube boot, it's drying now. Did some layout work for the amp and decided I'll use silicon diodes in the power supply. I initially planned to make two electrolytic rectifiers for the plate and filament supplies, but they would've taken up quite a lot of space. More importantly, they produce hydrogen and oxygen gas, which when enclosed in the case of the amplifier would pose an explosion hazard. Sure I could've added ventilation holes and a fan, but I'd rather keep things simple for my first tube-based device. I'll make entirely-homemade purist devices later.
Going to a flea market tomorrow to look for a suitable power transformer for the amp; if I can't find one I'll order one off ebay and move on to wire recording as a side project, since the piano wire came in the mail earlier this week. I figure I'll make an echo loop with a gearmotor and some pulleys and mess around with different recording and playback heads.
I also did a minor repair to the spark box. I replaced the sketchy high voltage module; the new one seems to work better. I updated the article with a picture.
http://simplifier.neocities.org/spark.html
>>973366
Hmmm, that gave an idead.
Why not just coat the inside of glass envelope with metal to act as anode instead of putting tubing.
>>975651
Different expansion rates.
>>975651
some super high power tubes (and other special purpose glassware) use metal tubing sealed directly to the glass tubing. Coating glass has problems like getting a sufficient thickness of metal to stick to the glass and the problem with expansion rates as noted above.
![Glass_to_metal_seals_1922[1].png](https://i.imgur.com/V1ffI1z.png "Glass_to_metal_seals_1922[1].png")
>>975709
forgot pic
>>975715
when you deposit metal onto the glass envelope as a getter, it's only requirement is that it has enough surface area to absorb any gases left in the envelope, so it can be made extremely thin. If you want to carry current through the metal, it will need to be much thicker. I would guess that the thickness required to match the current capacity of a normal lead sealed through the glass is beyond what is practically doable due to the constraints mentioned above. Even if it worked, I doubt it would be economical.
>>975727
If you make carbon nanotubes on the surface of the glass you would have no problem. It would also serve as a getter. That was the method I was going to use, please refer back to
>>971889
as a reference material.
However, as far as I'm aware, it might be impossible to grow carbon nanotubes parallel to a given surface, also you might run into the hairy ball problem.
>>975651
That would put the anode pretty far away from the filament (increasing plate resistance), unless I were to use really tiny tubing. I'd also have to have some internal titanium fuse to do the coating, since I can't get titanium hot enough to evaporate with my induction heater. The coating might also have a high resistance if it isn't thick enough, and it might coat the glass between the feedthroughs (shorting them together) unless it was obscured somehow.
Tl;dr there are some potential issues, but it could possibly work with a tiny envelope or at very low plate current. It's been done before, by the way. This guy made a photocell that way.
http://www.teralab.co.uk/Glass_Blowing/Mg_Photocell/Mg_Photocell_Page1.htm
>>975706
Not with titanium and soda lime glass, they're nearly identical. And for thin coatings it barely matters anyways.
>>975727
So far all of my tubes have only needed to conduct 10mA or less through the plate, so the plate feedthrough isn't the limiting factor for current. I think 10mA through a thin film of metal is reasonable.
Updates:
Looked for a suitable filament transformer at the flea market, couldn't find one. Picked up some nice DC gearmotors though, for wire recording experiments. They fit the voltage range of my bench power supply perfectly, and have the exact speed range I was looking for. Also picked up a big AC motor for $3. It needs a capacitor but runs nicely if I start it by hand. No idea what I'll use it for yet.
Then when I got back home I realized I had forgotten to account for transformer loading, did some more calculations/simulations, and turns out one of my parts box transformers will work perfectly. I'll start building the amp tomorrow.
Amplifier is coming along. Cut the wood for the enclosure and milled holes in the front panel (pic related) for the speaker and for a cubby where the triode will sit. I figured it would look nice having the tube exposed with the filament glow and all, and it'll also let me stick the induction heater over it if it starts to go bad. Popped a knot out of the corner of the board though, so I'm waiting on the glue to dry before I drill holes for screws and put it all together. Should be done tomorrow.
>>976355
Hey, I found something that might interest you.
Everyman's turbo-molecular pump: http://hackaday.com/2016/04/09/everymans-turbomolecular-pump/
>>976495
My first impression is that I don't think the boundary layer effect works at 10e-7 torr. I don't mean to be negative, but I just can't see it working. IMO the tesla turbine is a meme engine and people think it can do everything that spins. Not saying they aren't cool or useful, it's just that they're a magnet for the non-technical big-ideas arduino crowd.
Updates:
Turns out that even after a decade of collecting scrap parts, I never collected a 3.5mm headphone jack or a male-to-male audio cable. Ordered both from an east coast ebay seller, so they should be here in a few days. Amplifier won't be complete until they come in, but I'll try to get as much done as I can. If I get stuck, I'll switch to working on either a wire recorder or a tetrode until the parts arrive. Pic related is the current state of the amp.
Aaaaaaaand, bump
>>978248
Thanks for the bump.
Updates:
Still waiting on parts for the amp. Did some experiments with wire recording in the meantime; made a read/write head out of a ferrite ring that I cracked in half and superglued back together with a gap spaced by paper. Wrapped about 100 turns of wire around it, and tried to record to/play from some .010" steel wire with my mp3 player. Had no success, so I assume I need more power. Still deciding whether to just rig up a little 2n3055 power amplifier or to get really authentic and make a carbon mic, which I wanted to do anyways in the future. Leaning toward the carbon mic, that way I can write another article before the amp parts come in, then get into wire recording in depth once the amp is done.
>>946133
You try making a nixie tube?
Updates:
Spent a couple days working on carbon mics; no article because they universally sounded like garbage. I might come back to them in the future, but it's getting close to when the amp parts should be arriving so I'll go back to working on that for now. Pic related; I made one carbon granule based mic, which could pick up mechanical vibration but not sound, and one imperfect-contact type mic which could pick up voice but with a lot of distortion and resonance, and I had to yell at it through a pvc pipe to get it to even do that.
Overall I'm not terribly impressed with the technology, seeing as historically even the best carbon mics had a fair amount of distortion. The fact that they don't need an external amplifier is still very appealing though.
Updates:
Amp parts still haven't arrived and I couldn't stand giving up on something, so I spent more time on carbon microphones. Made a working Reisz microphone (pic related), but the output power is pretty low so I'm going to try to improve it before I write an article on it. It sounds nice though, despite having some crackle and hiss, so I retract my earlier statement on carbon mics. I thought they all sounded like old phones.
>>980309
Neat. What's the principle of operation?
