/diy/ - Do It Yourself

>>1040196
>>1040222
This level of safety policing is what kills the DIY mindset. Let this thread die, I won't make another.

>>1039417
>PRECISELY why is that in your informed opinion? Evidence or fuck off.

"Safety first" is just another form of "just buy it". By doing something yourself, you are assuming more risk than you would by paying someone to do it for you. I'm not saying "safety never", it's more like "safety third". You shouldn't let safety concerns stop you from doing something interesting. You should accept the fact that you may be doing something dangerous, and protect yourself as much as possible after the fact. I just thought that refilling O2 bottles would be interesting, so safe-ish methods of doing that should be sought out rather than scrapping the idea altogether.

I've cut and burnt my hands plenty of times while making tubes. That could have been avoided if I just bought them instead. But then what would be the point?

>>1039362
>You don't refill O2 cylinders with a compressor and the suicide warning is correct. The idiot who suggested it should STFU on industrial safety matters he knows nothing about.

Please explain how refilling a disposable O2 bottle to 150 PSI (or whatever its rated pressure is) can cause anything other than a loud bang if it pops. We aren't talking about 2200 PSI welding tanks here, we're talking about the little red cans they sell for 10 bucks at home depot. Regardless, "safety first" has no place in DIY. Take your fearmongering somewhere else.

>>1039345
>What did you change?
Previous tubes were made from soft glass, and I was using titanium for the seals. I was having trouble for the longest time because I didn't think to grind off the native oxide layer on the titanium wire. Apparently that was the cause of the leaks (I assume the layer was porous) because once I started doing that I got perfect seals every time. I've since switched over to Pyrex so I can make larger tubes, using tungsten for the seals. The diode was the first tube I made with this new construction method, and it appears to be holding a good vacuum.

>>1039193
>Didn't the last thread show a maker printing characteristics curves to his handmade valves?
That video was actually what got me thinking about making tubes, when I first saw it years ago. Here you go:
https://www.youtube.com/watch?v=EzyXMEpq4qw

>What are your main goals this time around?
My three big tube goals have always been to make a triode, a tetrode, and a CRT. I've made the triode (and it still works), the tetrode is coming soon, and the CRT is still pretty far off. The idea is to cover all the bases (audio, radio, video) of modern electronics, with completely scratch-made parts. Those won't be the only things I make though, for sure. I tend to get distracted by side projects (the entire month of May was devoted to carbon microphones, for example) so expect some other stuff to show up on my site as well.

Update:
I've pretty much nailed down the socket design, so expect an article on that within a few days once I make a couple of them. After that I'll probably make a few other tubes for practice before attempting a tetrode, since that'll be a fairly complex build. I'm thinking another triode just for fun, and a full-wave rectifier (diode with two plates) because they're interesting and the new envelope size gives me enough room to make a nice-looking one. I'll update the thread when I put something up on the site.

>>1037665
Listen to the radio and record songs on tape.

>>1037731
Thanks.

Just to give you guys an update, the diode still works and the characteristics haven't changed at all, so I think I nailed the feedthroughs on the first try. Going to make a proper base and socket for it, then start working on the next tube.

>>1037651
I love when people come into a thread and post about how <thing they've never attempted> is impossible/will kill you. I've seen this dozens of times, and according to people like you, none of my tubes should work and I should have died in a house fire years ago.

As to the specifics here, I was assuming he'd be smart enough to use an oilless compressor so as to not compress pure O2 to 10 atmospheres on top of some hydrocarbons. Common sense stuff. Other than that, where's the problem?

>>1037643
>>1037645
Thanks man, most of the stuff I've done is just a matter of seeing something cool and wanting to build it. Really, anything can be built on your own. If you think it's too complex, either simplify it yourself or find an older version of the same device which is already simple. Once you have a simple thing (less than 10 sub-components) it's trivial to build it on your own, no matter what it is.

>>1037639
Haha, everyone always asks about nixies. They're really just not my thing. Plus, there's already a guy that makes them from scratch and he does an excellent job:

http://www.daliborfarny.com/

I might make some more glow tubes in the future, since I want to do more experiments with helium in that regard, but I doubt I'll ever make a nixie. I'm just not fond enough of digital things in general for it to be worth the effort. I like it more when the tube is the primary component in the circuit, not just something to be driven by an IC.

>>1037615
>How long have you been practicing?
Just as long as you see on my site, so about 8 months. It isn't really that difficult once you know how glass behaves as a material.

>Also, I know tooling is a bit of an investment to start, but how much does it cost to keep a small stock of materials?
You can start super cheap, honestly. Soft glass (I got mine from Brillite) is $2/lb and can be worked with a blowtorch, and titanium wire is cheap ($5 for a small roll) because apparently people use it for vapes. It seals well into soft glass and is a superb getter material so you won't need a diffusion or turbomolecular pump or any barium or anything. Really, any pump that can get you below 30 millitorr is fine; you can do some bell jar stuff at that level without eating up your filaments too badly, and for making sealed tubes titanium will clean it up just fine.

All of the above will easily get you to the level of making usable tubes for audio/radio. Then if you want to make bigger/more complex tubes you can gear up and start using Pyrex, which is more versatile but also more expensive and less convenient.

>>1037481
Oh man, you've been wasting a ton of money. The large torch tip in the video is only using about half of the maximum output of the concentrator. I've got a larger tip that I'm working on that can liquefy 1" medium-wall Pyrex, and it just about maxes it out. I'd only need to store oxygen if I intended to work with tubing larger than 1-1/2" or so.

On the plus side, now you've got a bunch of empty O2 bottles, so if you get a concentrator you can possibly use a compressor to refill them for portable/small work.

The only downside to the concentrator is that it needs to put out at least 1LPM or so, otherwise pressure builds up on the inside and it starts making warning beeps. So if you want to use a micro torch or something, you'd have to put a leak in the hose on purpose to let out the excess oxygen.

I said I'd be back when I made another vacuum tube, so here it is (pic related).

ITT I'll answer any questions about how I do what I do, and give whatever advice I can if anyone is interested in doing the same. Also if you make tubes as well you should definitely post yours here, even if it's just a lightbulb or something. I know a few guys were interested in doing that in the last thread.

My site is here:
http://simplifier.neocities.org/index.html

If you want to get a good idea of how I got where I am now, start at the bottom. If you were in the old thread, start at the oxy-propane torch article, because that's when I start gearing up to make new tubes.

One thing to keep in mind is that although I make tubes, I'm really not an expert on "modern" ones like you'd find in audio equipment, so this probably isn't the best thread to ask how to fix your guitar amp. My knowledge is more focused on the details of how they work, rather than on the specifics of certain models.

As far as what to expect in the future, I'm going to start focusing on radio, with my long-term goal being a single-tetrode AM transmitter. The tetrode being homemade, of course. Beyond that point I'll start making larger tubes, with my end goal being to make a CRT.

Welp, looks like we've hit the bump limit. You guys pretty much know what to expect with this phone project, so I won't make a new thread until I make another tube. The plan is still to make a tetrode next, and like I said earlier, I have some ideas for making this tube quite large. I'm shooting for multiple watts of power output so I can use it in a radio transmitter.

Thanks for all the posts, and if anyone's working on a tube of their own, good luck. See you all in a few weeks.

>>996036
Thanks.

Updates:
Had to redesign some stuff; the bell mechanism only really works well on 6V, but when I tried to run the microphone on that I got arcing and squealing. I modified the microphone design to have two contacts in series (so each gets 3V) and that seems to work fine, although there's a bit more crackle/hiss than my first microphone had. I've still got a couple more components to make, but I haven't posted an update in a few days so I leaned everything together and took a picture. I should be starting on the final assembly tomorrow, and if it works well then I'll go ahead and build the other one; I've got most of the components made already.

Updates:
Building a relay that's sensitive enough to switch with only 10mA coil current appears to be beyond my current skill level, so I went out and got some 16ga yard light wire from the hardware store, so each telephone can run the other's bell directly.

Started work on the audio transformers as well, got the cores built today (pic related). They're open-core primarily because they're easy to build that way, but also because a closed core would saturate from the large DC bias of the carbon microphone (and a proper gapped core isn't worth the effort). The middle is stuffed with 14ga annealed iron wire. I've used 12ga before in a similar transformer and it worked, so I assume if anything this one should work better due to the reduced eddy currents.

I'll probably wind both transformers tomorrow, build the carbon microphone mechanisms (they'll be adjustable to tune the sensitivity), then start putting everything together.

Updates:
Decided against machining an armature/hinge mechanism and just used thicker steel wire. I also moved the interruptor from in front of the coil to behind it, where there's less overall motion. Soldered on some copper contacts as well and it seems like it'll work fine. No pictures of the finished mechanism because I didn't make a nice looking one yet; I figured I'd do that later once I have the phone enclosure completed, since the layout of the enclosure will determine the size of the mechanism. I also made another electromagnet, since now I know for sure that it'll work.

Moved on to making a reed relay. Each telephone will need one to run the bell circuit, since the bell draws 200mA but the line connecting the two phones is roughly 80 ohms, so one phone's battery can't power the other phone's bell directly. It's turning out to be trickier than I thought. I could just use a higher voltage battery, a lower-resistance line, etc., but my carbon microphone design runs best on 1.5V or lower, and I like the high-resistance line because it simulates a long-distance connection. Based on what I've read about the earliest telephone systems, the line resistance was about 2 ohms per mile.

I'll post more updates as I make progress on the relay. After that, I just have to make an audio transformer, which I've done before so it shouldn't be hard. At that point I'll have all the components finished, and I'll be able to put everything together.

>>992811
Thanks.

>>992816
I have some nickel silver wire, but as for silver itself I'd rather not get into precious metals. Plus I did some experimenting today and found that as I expected, nickel on nickel works fine at low voltages and doesn't have the oxidation problem that iron does.

Updates:
Made a prototype bell mechanism with an electromagnet I made today (pic related). Connections are made to the left-hand copper wire and the overhanging piece of nickel wire. The iron wire gets pulled into the magnet until it breaks the connection and everything resets, forming an oscillator. Putting one of the fencepost caps under the end of the iron wire made a nice ringing sound. It's a little unstable though, and the ringing is a little weak, especially since I'd like to have a double bell on the finished telephone. I discovered that wider pieces of metal are pulled with greater force towards a given electromagnet, however, so I think I'll make a proper armature out of steel flat bar that can deliver more force to the bell. It'll also be more reliable I imagine, since the current setup tends to wiggle out of place.

I'll post more updates as I work through this.

>>992307
Thanks man, you dedicated anonymous bumpers have saved this thread multiple times.

>>992572
I think I'll wait until it hits the bump limit before I start a new thread. Don't want to fill up the board with threads that are inactive 99% of the time. If it dies anytime soon, I'll make a new thread when I have a new article.

Updates:
Telephone may or may not be on hold. I was going to make the contacts for the bell mechanism and relay out of the nickel wire I had, but then realized that thin nickel sheet would work way better (more flexible, bigger contact area). Problem is that thin nickel sheet only comes from China, so it's a 2-4 week wait. Ordered some on tuesday though. Still deciding what to do; I might just build the mechanisms with the wire contacts and design it so I can swap the wire out for sheet when it comes in.

Also, all the material for the tetrode should be here within a few days, so if I stay stuck on the telephone project I can switch over to making a new tube.

Actually now that I've typed all that out, I think I will use the wire temporarily in the telephone. Should have more updates soon.

>>990847
>>991140
Thanks.

Updates:
Finished the woodwork for the first telephone. Pic related, I leaned it all together to give an idea of the final shape. The fence post caps on top are going to be the bell, and the pvc cap on the side is the earpiece. The diaphragm (the milled out part) is 5 inches across, to give an idea of the size. As far as the switches go, one will be to route the battery power to either the bell circuit or the microphone circuit, and the other will be to ring the other telephone's bell. I'll start on the bell mechanism tomorrow.

I've also been doing some research for the tetrode I plan to make next, and I may have come across something that will let me make much larger tubes. I'll let that be a surprise though, mainly because I have no idea how well it'll work.

I'll keep the thread updated with progress on the telephone. This is the most complicated project I've done and I have to make two of everything, so it may take me another week or so to finish it.

Tried a bunch of times to make an encapsulated diode, but the rectifying behavior was extremely pressure-dependent and unstable. I now understand why cat's whisker diodes were a thing. In general I'm much more fond of the electrolytic rectifier I made earlier. I might come back to semiconductors later, but I've decided to move on to building the telephone for now. I made a pair of piezo earpieces yesterday, and I'll start on the enclosure woodwork today. I want this project to be almost 100% homemade, so I'll also have to make pairs of audio transformers, reed relays, and bell mechanisms, which I'll document as well. This article's probably going to be fairly long.

And according to schedule, after the telephone's done I'll go back to tubes for awhile. I still need to make a tetrode and a CRT.

Maybe you could post a picture and I (or someone else) could try to identify it. Really though, if it's part of a working circuit and you don't have much experience, I'd just leave it alone. You can get boxes of old tubes on ebay for cheap if you want to mess around, and that way you won't feel so bad if you accidentally burn out the filament or something.

>>988691
Just find the datasheet for it, that should have all the information you need. That goes for any component really, not just tubes.

Updates:
Rigged up a power diode, pic related is its characteristic curve (H = 2V/div, V = 500mA/div). Like most metal oxide diodes, it has pretty high leakage in the reverse direction, but a 10:1 ratio (excluding where it starts to break down on the left hand side) of forward to reverse current isn't that bad.

I figured out that the best results were achieved by heating the metal to a dull red for a long time, rather than heating it orange-hot briefly. The diode itself is just a strip of burned galvanized steel in contact with a 1/4" graphite rod, with graphite powder at the joint to improve conductivity. It's super jerry-rigged and the powder is only loosely packed, so the final version will probably have less resistance (steeper I/V curve) and be more efficient overall. I'll make a fully encapsulated version tomorrow, and if it works well I'll write an article on it.

>>988383
Yeah, I'm aware of it; I've gotten a headache from burning zinc before and looked up why I felt terrible. It's pretty easy to avoid, just hold your breath and leave the room if you see white smoke.

>>988404
Oh neat. Thanks anon. I thought they would send me an e-mail when it went up.