DIY Electron Accelerator

Reader [Xellers] sent in his newest instructable: DIY Electron Accelerator: A Cathode Ray Tube in a Wine Bottle. While not exactly what you might think of a cathode ray tube, the basics are in place. A wine bottle is used as a vacuum chamber and a 9kv neon transformer is attached to a stopped in the top. A cathode is placed mid way, the air is sucked out with a pump, and high voltage is applied.

Naturally as more air gets pumped out the electric arc intensifies into a pretty solid plasma filling the space between the two contacts. While mixing drilled glass with a vacuum and high voltage sounds like an awesome hospital story [Xellers] does cover some safety points including the possibility of this thing putting out some nasty waves.

One thing that is not mentioned (that I saw)  is this is very similar to how florescent light tubes work and without florescent material lining the chamber it will spit out quite a bit of UV light (notice germicidal UV lights are clear). So you will want to watch your eyes!

Join us after the break for a quick video.


26 thoughts on “DIY Electron Accelerator

      1. It depends mostly on the metal that he used for the anode. In some metals (I don’t remember which ones) when an electron is knocked loose, an “x-ray” will be created when the electron falls back into place.

        But you’re right, the electrons probably don’t have enough energy to create x-rays anyways.

  1. You laugh BiOzZ, but if the voltage were sufficiently high enough, then yes, there would be some stray x-rays! I don’t see too many 25+Kv transformers out there ready-made, but older TV’s improperly repaired could achieve the same thing. Rotting your brain would only be ONE of the many side-effects in that case :P

  2. agreed, 9Kv should be quite safe, 20-25Kv is likely to give some trouble though.
    even so, you Should be able to get some neat effects using Magnets around this plasma, and maybe even modulate it with an electro-magnet too, wouldn`t really cost you anything and would make it a little “Interactive” :)

  3. it will produce soft x-rays which will be absorbed by the glass, around 18 kv there will be hard x rays which will require shielding

    color tvs required higher voltages for brightness, and had to use leaded glass for shielding

    and while mercury florescent tubes do emit uv light it is limited to uva and uvb bands, germicidal tubes use flashed quartz tubes to allow uvc light to pass though

  4. I’d think the thick glass would absorb the vast majority of the UV light produced, which is (like dcroy already mentioned) the reason UV tubes are made of expensive quartz glass. They wouldn’t do that if it would work with regular glass, would they?

    I’m by no means an expert, though!

    This sounds like a very neat project, I might try it sometime, maybe with some coils to modulate the plasma. This belongs in the study of a self-proclaimed evil mad scientist.

    1. longer wave light will still get though, mostly uva and some uvb will get though

      still dangerous when the intensity is high enough, when i made by eprom eraser i made absolutely sure there was no leakage because even a 15 watt germicidal tube would produce enough uvc light to cause permanent retina damage and skin burns

  5. The biggest problem with this is a vacuum pump, that is it.

    You need to get quite a good vacuum pump which can go below 50 microns to have some fun vacuum fun with electrons.

  6. Peeps, modulate the glow tube to music! Imagine if ya can find a Old CRT , take the deflection yoke off the neck and put it on your bottle. Drive with a Low Z amplifier and have fun playing. no idea what it will exactly do without a formed E beam, but worth a try.

    1. You can get hard x rays off of stainless at a pretty low voltage (5 to 10kv?) Cant remember for sure.

      The glass will absorb virtually all of the nastier UV. Germicidial UV lamp envelopes are made from silica glass (Quarts) which passes UV.

      Tachikoma, 50 microns is 50 milli-torr. Pretty easy to get with a generic vacuum pump. My systems will get down to 1e-8 torr on my system with a 6″ diffusion pump and my little turbo pumped residual gas analyzer setup.

  7. Great project!

    Technically, this is closer to recreating a Plasma Globe than a CRT. The goal in a CRT is to have absolutely no air (or very close to it) and light up a phosphor coating; here the object is intentionally ionizing the gas (at partial vacuum) inside the bottle. The light bluish-purple glow is characteristic of atmospheric air – the main (pretty much only) difference from a commercially produced plasma globe is that they use different / more exotic gas mixes (typically noble gases) to produce different or more intense colors.

    A high-quality vacuum pump is NOT needed to recreate this – the compressor from an old fridge or dehumidifier will pull adequate vacuum to play with this. In my tests a fridge compressor pulled a jar down to about ~20 Torr (~.026 atmospheres), which is not too shabby!


    1. No, more like a crookes tube. I am sure if he put phosphor on the bottom of the bottle it would light up pretty good from electron acceleration. This is the proper two terminal tube that runs on HV DC.

      Plasma globes are a different critter altogether. They are a single terminal display. They use high frequency pulsed dc or ac. Rarely do they even use an electrode. As you say they use rare gases, often a mix of helium, neon, krypton. and xenon.

      There is an old article in the Amateur Scientist section from Scientific American where they detail the use of refrigeration compressors. 20 torr alone is pretty useless for things like this, but put two pumps in series and you will get a much better ultimate vacuum, though your throughput will be awful with the small lines. For the time spent is better use to get a cheap HVAC service dual stage pump.

      For those that are interested in vacuum science it seems one of the best places to find info on the net is the forum since these guys need vacuum systems for the fusors. Also has a good section on vacuum with some members that build particle accelerators for a living.

  8. What’s missing is the mercury, which has a characteristic emission in the UV range. Whatever mix of compounds are incidentally in the bottle are not going to emit a whole lot of anything in particular.

  9. im a glass blower whos been screwing around with this for 206 years…u can use almost any vacuum pump if ….you wash the vessel with argon basicly…suck down to 30 torr…fill to 700 torr with argon or helium,..suck back down to 30 torr..repeat…do it a 1/2 dozen times…and then do your final fill with a bit of neon,argon, krypton or xenon…a bright pure neon sign finishes at like 27 torr…but most plasma art runs 50-150 torr..any questions drop me a line at

    1. and if u use a borosilicate vessel (pyrex) it will absorb all the uv-a and alot of the uv-b. 9-15kv around 12 watt power supplys are very cheap, and more often free from junk plasma globes from china..and will light a 2 inch 2 foot long pyrex tube VERY well….even better if its a cavity vessel.
      a nice oil rotary vacuum pump is great if ya got one..but a compressor out of a water cooler is currently makeing nice plasma bonsai trees for me.

  10. My 17 year old nephew thought I was lying to him about old TV’s putting out X-rays and needing lead shielding to protect viewers. Up into the attic I went o retrieve a 25in TV weighing more than my nephew and luckily the warning sticker was still there. Then he switched to how could they make something so dangerous, while he cleaned the rifles from our range day🤨

Leave a Reply

Please be kind and respectful to help make the comments section excellent. (Comment Policy)

This site uses Akismet to reduce spam. Learn how your comment data is processed.