Introduction: Bottle Light Bulb AKA Bacardi Bulb

Picture of Bottle Light Bulb AKA Bacardi Bulb

This will show you how to turn an ordinary bottle into a light bulb.

This is an experiment and there may be flaws but that is all part of the learning process.

I used a Bacardi bottle for the bulb but any glass bottle will do,

This in not a LED instructable. This is a direct response to the death of the incandescent bulb.

I harvested a tungsten filament from a 40 W bulb since I did not have a local source for tungsten wire,

The build is 40 W and I love the glow. I t has been running for a couple of weeks now with no problems. I tested the bulb with a constant run of over 24 hours in the shop'

There is a slight discoloration of the bottle near the filament (the glass turned blue) but the light is bright and constant...

Step 1: Materials

Picture of Materials

You will need a 40 W Incandescent bulb.
A bottle
A wiring solution.
Some brass rod and screws
Some wood for a base
The ability to braze
High Temp Silicone
glass drill
a vacuum pump
Free will
And some imagination

other materials will be required as presented.

Step 2: The Bottle

Picture of The Bottle

You will need to drill 2 holes onto the base of the bottle.

I used a tile bit for a Roto-Zip tool.
This creates 2 holes that are equally spaced.

I used a small section of black PVC plumbing pipe in a vise for a holding tool for the bottle.

Step 3: Harvest the Filament.

Picture of Harvest the Filament.

Ideally I would make my own  but I chose o harvest a filament form a 40 W light bulb.

Wrap the bulb with a paper towel and gently tap with a hammer. Increase the force until the bulb breaks. If the filament breaks, begin again.

Trim all the connection points until the filament is free and supported by 2 sections of bulb wire.

Step 4: Make the Trodes

Picture of Make the Trodes

The electrodes are made from section of Brass.

You will; need 2 section that will place the filament in the center of the bottle.

In this case I welded 4 inch brass rods to 10-24 bolts then to thin steel wires for positioning.

I used a silicone baking sheet to make washers with hole punches.

Mark the bottle for position and weld the filament legs to the brass rods.

Weld the filament to the brass rods ( i brazed but the process is the same.

Assemble the rods by placing a silicone washer on the bolt then feed the entire assemble through the holes on the bottle.

Take care as the filament is delicate. Try to not touch the filament as this will cause premature burn out.

Feed the bolts through the holes on the bottle, align and tighten with the appropriate size nut.

I used a 10-24 nut with another silicone washer and liberal amounts of high -temp silicone.

Tighten the nuts then trim the steel rod off. You should be left with a brass bolt protruding through the base of the bottle. These are the connection points for 110V to power the bulb.

Step 5: The Base

Picture of The Base

The base is a couple of 2X3 sections left over from another instructable.

You need a stable base that is  larger than the base of your bottle. You will need a hole cut into it to allow the bottle to sit flat

You should have a cord that allows for switching on/off.

Step 6: The Plug

Picture of The Plug

You need to make a stopper for the neck that will allow you to remove the air. The amount of air in the system is directly proportional to the speed of the filament burn out.

The vacuum port is made from a refrigeration schrader adapter brazed to the 5/16 washer.

TIG welding is not possible here so I braze-welded the washer to the adapter. There is a rubber washer for  mating and I used High temp silicone fore the final seal.

The adapter is placed on the neck of the bottle and the rubber/silicone creates an air tight seal. This is important since any air in the system will cause the filament to burn out quickly..

Step 7: Suck Out the Air

Picture of Suck Out the Air

You will need to use a vacuum pump to suck out the air. A vacuum cleaner will not work.

Any contaminants will cause problems.

I made sure I had a tight seal and then ran the pump for a 1/2 hour prior to removing the vacuum source.

I tested the bulb under full vacuum by powering the filament with 110V.

The glass turned blue to indicate some contamination in the bottle. Possibly due to residual cleaning supplies.

Step 8: Enjoy

Picture of Enjoy

Remove the vacuum source, Seal the system and enjoy the bulb.

The vacuum port can be disguised with a bottle cap to hide the port.

I left it visible to allow for future vacuum since this was an experiment.

Recently my bottle bulb has been lit for over 72 hours straight before the filament burn-out. The glass had some discoloration but the filament was replaceable and I can get up and running in less than 1/2 hour. I suspect that I may have been a vacuum issue.

I want to make my own filament but need some tungsten wire for the experiment.

Yes the bottle did get hot like a standard bulb. No problems just don't kiss it...


russ_hensel (author)2012-01-26

This is a very cool instructable. I have some ideas that might help others or might be bogus.

A refrigerator compressor can be used as a vacuum pump, good enough who knows.

Old vacuum tubes used to use a getter, this was a ( i think ) a second filament that was deliberately burned out to consume remaining gas, perhaps you could add one.

Nicrhrome wire from a toaster might make a filament.

For quite a time edison used carbon filaments. He made them from bamboo that he carbonized by heating.

Good ideas... Thanks.

Refrigerator compressors produce good vacuum but I amnot sure if they can get as good as the 2 stage ones like I used. if you have to buy a compressor, for roughly the same price you can get a decent vacuum pump.

I inadvertently included a getter in this build as part of the filament wires. zirconium is used on the filament leads. check out my second bottle light...

I use nichrome wires in my glow lamps. They may be too thick to produce more light than heat.

where did you manage to find zirconium?

pwnag3 (author)2012-07-17

Was thinking about doing this with a mason jar the second i saw the pic XD

fjchorn (author)2011-12-05

Nice one, what would happen if I filled the bottle with an inert gas such as Argon instead of using a vacuum pump?

Random_Canadian (author)fjchorn2011-12-05

Depending upon who you talk to the standard bulbs may be charged with Argon. It prolongs the life of the filament.

Rowen27 (author)Random_Canadian2012-03-17

also with the right set up and argon, you can turn it into a 'plasma' bottle. I plan on making one this summer to use as a 'lightning in a bottle' stage prop. there is another instructable on here that explains the setup for turning a light bulb into a plasma ball, but I forget the name...

Argon would be maybe easier than vacuum (assuming you can get some) because it is denser than ambient air, so you would "fill" the bottle with argon gas that would displace the air. Then you could avoid using the schrader valve in the neck if you put the bottle in an argon "tank" stoppered it then screwed on the original bottle cap to finish it off.

This is an awesome instructable!

I used to work at a large wine store. they sold bottles of inert gas to fill wine bottles with so as to displace the air and preserve the wine longer. I'm not sure which inert gas it was though... just a thought

My mom received one of these for Christmas. The one she got says argon right on it.
Don't know how pure, but it's sure better than plain air.

sbeltz (author)the walking stick2011-12-06

IIRC it is nitrogen gas.

the walking stick (author)sbeltz2011-12-07

ah no dice. Would that work>?

Any inert gas should work. (welding supplies).
It is intended to keep the high heat from causing a reaction between the tungsten and other elements like oxygen, burning it up.
One might try sealing a burning piece of something (or other chemical reaction) inside to consume all oxygen creating a non-reactive environment.

Orngrimm (author)2012-02-06

Nice one!
I admit: I didnt read all the comments, but i thought of something:

You evacuate the bottle. Why not fill it with an inert gas? like CO2?
Longer lasting (Vacuum will not be hold forever) and simpler to create (since CO2 is heavier than air, simply pour it in the bottle and let the Air escape upwards & out of the bottle). Also a CO2-Cartridge is cheaper than a vacuum-pump (if you have to buy one).

My 0.02$ :)

Random_Canadian (author)Orngrimm2012-02-24

thanks for the idea... check out my other light bulb for more on this.

bearuangpanda (author)2012-02-19

the best way is you just put the small bulb inside the will light more longer...

Where's the fun in that?!

Random_Canadian (author)2012-01-19

Thank you all who voted and a special thanks to the Judges for selecting my project for this prize.

canucksgirl (author)2012-01-18

Congratulations on winning First Prize in the "Make It Glow" Contest. You got my vote, and I'm not surprised to see you as won of the winners! - Great Job. :D

Thanks. This was a fun build.

neffk (author)2012-01-02

The dark band can be prevented by adding a halogen to the bottle. Or you can slow it by connecting the bulb to a lower potential.

I was reading about incandescent bulbs the other day and was surprised by how many factors contribute to the life of the bulb.

Random_Canadian (author)neffk2012-01-02

In my other light bulb instructable I have addressed and corrected many of the flaws in this one.

I am sure that was some interesting reading, I spent many hours myself in search of proper methods and answers..

bmiller11 (author)2011-12-08

Great Idea.. I would run the vacuum pump on the bottle and bring the bottle down as much as possible. This will evaporate the water. then use the same whole to fill the glass bottle with nitrogen. The nitrogen might stop the tungsten from evaporating.. I don't know it's density compared to argon. I work in HVAC and we use vacuum pumps to remove moisture from an a/c system. We pull the system down to a 500 micron vacuum. A good pump can do this in 15 minutes if you don't have any leaks. I don't think those rubber gaskets would hold a vacuum so deep though.


I definitely learned a lot from this. A low pressure inert gas will be needed.

My 6cfm pump is rated down to 50 microns and I have seen as low as 12 on newer refrigerator systems. tested with an inficon pilot gauge. the bottle would do 100 but leaks were evident overnight.

more work is needed on this one but I am pleased with the results so far.

sconner1 (author)2011-12-08

A high quality vacuum is extraordinarily difficult to achieve and maintain on Earth.
I'd say 72 hrs of burn is pretty good.
I like the tungsten deposits on the glass. It looks great.
What would happen if you placed a piece of aluminum tape on one side of the bottle and put a + or - charge to it? Could you steer the tungsten deposits toward one side?
How about a third electrode inside the bottle.
Home brew tube rectifiers anyone? Sweet!!

Random_Canadian (author)sconner12011-12-15

Thanks. I will be definitely playing around with your suggestions soon. Been having a blast with this one...

zmiller1 (author)2011-12-14

I love this concept. I am working on a similar project. I was thinking of using helium for the air replacement since it is cheap and readily available. It's not argon or xenon, but hey, we're not light bulb manufacturers either. I found the filament here",-Graphite,-EDM,-Casting,/Detail
It's expensive ($75), but at 100 meters you could make about 600 light bulbs.
I was also thinking about using the neck as the entrance point for the filament and leads. This would cut down the number of points that need to be sealed. I realize that would also make for a tight fit with the hardware.

Random_Canadian (author)zmiller12011-12-15

Thanks. There is some good information in the comments section here. I was going for helium as well but the comments from another user here made me do some digging and the balloon helium that I had sourced is not pure. I am looking for another source now. I chose the bottom to get a specific look.

There is a source of inexpensive wire on eBay that is located in Canada but definitely not local.

good luck with your project.

signalelektronik (author)2011-12-08

You may find Tungsten Wire From Aliexpress. Write your name in bottle.

TY Sir for the addy.

This place sounds like an electrical hardware store online(not had a chance to really check it out as of this writing, but saved the addy). If that's the case, I'll be spending lots of money at this place. LOL :)

The Green Gentleman (author)2011-12-05

That's really cool, and honestly makes a nicer light than most store-bought bulbs.

The dark band might be the result of tungsten evaporation/condensation. You may also have some water vapor in the bottle. I don't know how you would remove all water vapor, honestly. Wikipedia claims that store-bought bulbs use zirconium to remove oxygen when the water is split by the high temperature. If water vapor is the problem, then the blue compound might be tungsten pentoxide. It might also be tungsten hexachloride, which might (???) have been introduced by a chlorine-containing brazing flux or chlorine-based cleaner. This is pure speculation though - I'm not a chemist, so I don't know the feasibility of creating these compounds in situ.

I think you could simplify this considerably by using Helium as an inert gas. It's lighter than everything but hydrogen, so you could invert the bottle and fill it from the neck. The helium will displace the air from the bottle. Then, just put the electrodes in through the neck and attach it to the base upside down, or hang it right side up.

Finally, if you look at the filament under a magnifying glass, you will see that it's tightly coiled to increase surface area and efficiency. There's no way you're going to be to replicate that at home. I think your approach to getting a filament from a store-bought bulb is probably the best one.

This is just blue-skying, but you might also be able to scavenge one from a "burnt out" higher wattage bulb. Grab the biggest piece of the broken filament and try stringing it between your electrodes. Of course, I could get off my butt and try it myself, but it's much easier for me to suggest that you try it in your very cool and beautiful creation.

I like your ideas and will definitely be playing around a little on this one.

I thought of putting the electrics through the neck but the cord hanging out the top presented a problem with aesthetics. I will be making the top stopper more covert so that it can be hidden beneath the bottle cap. The bottle shown was a proof of concept and more work is needed to make it pretty.

Final washing in extremely hot water then inverting the bottle to dry gets rid of most of the water. Running in a deep vacuum for a half hour tends to boil off the remaining moisture. My pump is a two stage oil bath type capable of producing a deep vacuum.

The discoloration must be from contaminates in the silicone or from the brazing process or as you mentioned from the filament itself..

Thank you again for your input.

acti (author)Random_Canadian2011-12-08

Note that Helium tends to leak out through the walls of the glass.
Most people don't realize that true helium is a bear to store,
even in tanks. It is so small of a molecule that even steel
walls are "porous" to it, to some extent.

BTW, don't even bother with department store "helium" tanks.
That is NOT true Helium, but "Ballon Gas", a MIX of normal
air and helium most people use to inflate balloons.

That's why over a few days your balloons will slowly deflate to a
"medium inflation", and sink to the ground. The helium in the
mix has left the balloon, leaving only normal air.

In our case, trying balloon gas MAY expose the filament to oxygen!
In reality though, the Helium will leave your glass within days,
leaving only whatever other gasses they used to dilute the mix.
This may or may not include Oxygen.

TRUE "pure" Helium MAY work. The partial pressure will descend
until it is similar to the outside pressure. (I'm not sure what happens
after that...) But it is very expensive to buy. You're probably better
off (cost wise) with other, cheaper, more commonly available
welding "oxygen displacing" gasses, like Nitrogen or Argon.

- Keith

The Green Gentleman (author)acti2011-12-09

I did not know that about helium. I knew hydrogen had gas permeation problems, but I thought Helium was less unpleasant. I was going to suggest Argon originally, but I thought it would be easier to displace air downward with Helium, and more effectively, because the only thing lighter than it is hydrogen, which isn't really a component of our atmosphere in a pure state. I don't know about the feasibility of nitrogen. I thought there were problems with reactivity with the filament due to high temperatures.

odalcet (author)2011-12-08

very nice, very diffcult :)

MrRodrigez (author)2011-12-08

what if you vacuumed out the air, then filled it with welding argon/co2?

rgershey (author)2011-12-08

An argument in favour of using an inert gas is one of safety. Pulling a vacuum on an old liquor bottle may create a potential hazard of implosion, flying glass, etc. Precautions should be taken.

heathbar64 (author)rgershey2011-12-08

Pressure from your inert gas could also cause a problem with explosion if the bottle was weak. Another reason I was thinking of a small canning jar. they are made to take the pressure/ vacuum and heat. one of the old blue glass ones would look pretty cool me thinks.

heathbar64 (author)2011-12-08

So! Been thinking about this some more. I think what I will try is using a small canning jar for the bottle. that way the large lid would allow easy insertion of the filiment and and easy changeout.

beta1072 (author)2011-12-05

Interesting, this is definitely going on the list of things to do, but not tell the wife. Is it possible that the discoloration is caused by sputtering like you get on vacuum tubes.

Random_Canadian (author)beta10722011-12-05

Always tell the wife... even if it it after the fact.

Still looking for an answer for the bluing though. I will try some different methods to isolate the source...

sconner1 (author)Random_Canadian2011-12-08

My best guess is sputtering tungsten on the glass.

Random_Canadian (author)sconner12011-12-08

I recently spoke with a friend of mine and he showed me a bulb that burned out is a funny manner. The inside of the glass was smoky and multiple shaded of blue.

I believe that it is a byproduct of tungsten burning in air.

sconner1 (author)2011-12-08

I like the bluing

Random_Canadian (author)sconner12011-12-08

So do I but the life of the filament is nowhere near what I would like it to be.

sitearm (author)2011-12-08

@Random_Canadian; I like that you got this to work. I kept thinking of you and Thomas Edison in your respective labs.

Looking at your prototype, it looks like it's almost as much work to replace the filament as to assemble the bulb initially. Have you learned any tricks to make that easy?

Cheers! :)

Random_Canadian (author)sitearm2011-12-08


It does take about 20 minutes to change the filament but I am more concerned with getting a longer life out of the filaments right now. The short trick would be to leave the steel wires attached to the electrodes and placing the bottle on a pedestal for testing until the life issues can be worked out. That would greatly shorten the rebuild time.

I think I will be trying this on a wide mouth jar in the near future

heathbar64 (author)2011-12-08

This is way cool. I want to try it.
My first thought on this is, Why not insert the filiment through the neck of the bottle instead of drilling holes? One less source for leaks. I would like to try the cotton filiment just for fun. if edison could make it work, why couldnt we?

I drilled the holes since my ultimate goal here was to have the bottle sitting in a table. What you see is a proof of concept. I have learned that I need a better filament and a better way of sealing the bottle. I may also make use of an inert gas for preserving the life of the filament.

I have been having difficulty repeating the results and i suspect that I have a vacuum leak somewhere. I will update this instructable with a tweaks section once I get a little more information.

Thanks and good luck

abizar (author)2011-12-08

Might be an incredibly stupid idea, but what about using an inert liquid (don't know if mineral oil would qualify) instead of a vacuum or argon gas? Might give it a shot over the weekend with mineral oil.

About This Instructable




Bio: Bit of a background in various electrical and mechanical fields, obscure sense of humour and typically willing to help...
More by Random_Canadian:Miniature LED TARDIS ThrowawaysPlywood Icosahedron  Coconut Cracker
Add instructable to: