book collections email follower instructable user
Picture of Homemade Incandescent Light Bulb - 自作の白熱電球

Recently trend is IoT, AI and etc. But i go against recently trend and made a vacuum incandescent light bulbs!

despite peoples has screaming energy conservation...

anyway, i learned a part of vacuum technology and scientific glassblowing by making light bulb.

Someday it will be useful.





Step 1: Prepare Materials and Tools - 材料と道具

Picture of Prepare Materials and Tools - 材料と道具

1. dumet wire and tungsten filament - デュメット線とタングステンフィラメント

this is a common copper wire...? no. this is dumet wire.

dumet wire is copper coated alloy metal wire of nickel and iron. patented by Byron E. Eldred in 1915.

composition is 42% of nickel, 58% of iron. and copper is coating.

coefficient of thermal expansion of dumet wire almost coincides with that of soda-lime glass.

tungsten filament is coiled shape tungsten wire.

tungsten's molten point is 3380℃. and vapor pressure is 10^-5 Torr in 2547℃. this is very low vapor pressure.


デュメット線は、ニッケルと鉄の合金線に銅が被覆されている線で、1915年にByron E. Eldred によって特許が取得されました。





2. glass tubing - ガラス管

There are various kinds of glass. I used this time is soda-lime glass. because work temperature is relatively low.

diameter of glass tubing is 15mm and 8mm.

15mm tubing is using for envelope. 8mm tubing is using for pinch stem.




3. gas torch - ガスバーナー

gas torch is using for glass work. large torch is using for glass blowing.

tiny torch is using for stem welding and seal bulb and making pinch stem.



4. blowing hose and silicone rubber stopper - ブローホースとシリコン栓

blowing hose and silicone rubber stopper is using glass browing.


5. Other tool and fixtures - その他

・ plier - ペンチ

・ mini lathe - ミニ旋盤

using for making pinch stem. ピンチステムを作るのに使います。

・ metal file

using for polish wire. - 線を磨くのに使います

・ vacuum pump and geissler tube - 真空ポンプとガイスラー管

vacuum pump is using for making vacuum. geissler tube is using for rough vacuum gauge.


・ variable power supply - 可変電源

using for flashing filament and testing light bulb.


・ test tube and test tube stand - 試験管と試験管立て

・ solvent - 溶剤

using for washing. 洗浄に使います。

So cool. I think you should try a vacuum tube (or Valve) Maybe a diode detector or put in a grid and make an Audion.

If you had access to a difusion pump you could pull a deeper vacuum.

2SC1815 (author)  Jack of Most Trades11 months ago
Thank you for watching this project!

yeah, I am preparing to make a vacuum tube now! Induction heater , baking furnace, and diffusion pump. Because diffusion pumps are expensive (mainly shipping costs), I am trying to making from aluminum rods with using lathe.

Thank you!
docman10011 months ago

Amazing :)

2SC1815 (author)  docman10011 months ago

Thank you!

bpark100011 months ago

You need getter to improve vacuum. Put a pinch of red phosphorous on filament it will transform to white phosphorous and combine with residual gases. You can also put 2nd lamp in manifold and seal off the pair. Burn 2nd lamp to failure, then seal off the tube to the 1st lamp. 1st lamp will last longer.

2SC1815 (author)  bpark100011 months ago

Thanks for the tips for improving vacuum!
However, red phosphorus is handled as dangerous goods and can not be obtained easily in Japan. I will use titanium or magnesium instead. (And phosphorus has high vapor pressure. and can not be used phosphorus to vacuum tube to be made next)

Thank you!

jianju11 months ago

Such a deliciously rustic project! Learned a lot just reading it! Maybe I'll do this one some day...

2SC1815 (author)  jianju11 months ago

Thank you!

Someday I am looking forward to the day when your light bulb can be seen!

HarolddL11 months ago


Now it's time to try a vacuum tube. I can't wait to you tackle this. :)

2SC1815 (author)  HarolddL11 months ago

Thank you! I'll do my best!

RaymondR611 months ago

This is more an educational project than practical, since commercialized incandescent bulbs are cheaper to buy already fabricated than the cost of making your own, just as making your own mechanical clock or even making a simple LED digital watch. Opens up the mind to think what can be done to make it better better or cheaper. Thomas Edison would be proud!

2SC1815 (author)  RaymondR611 months ago

Thank you for watching this project!

Japan is unlike Europe and the United States, there is no web site like Instructables. and people who doing DIY are decreasing in recent. On the contrary, there are also people who devalue people involved in manufacturing. It is not good... believe that it will get better. :)

Stuart2111 months ago

Very clever. I recall (40 - 50 years ago) hearing that the Russians used to open up blown light bulbs, replace the filament, & resell.

This is better!

2SC1815 (author)  Stuart2111 months ago

I have heard it before. To replace filaments of blown bulbs, it requires different skills with manufacturing of light bulb. I can not do it. Thank you!

mf7011 months ago

Delightful to read! How much research was needed before you started? Did you have to travel up any blind alleys?

2SC1815 (author)  mf7011 months ago

Thank you for watching this project!

I started a project to make a vacuum tube without research. I bought a burner and a fine glass tube at a general shop, first trying to make a light bulb with Nichrome wire. However, at that time I did not know the existence of the dumet wire, I just used a steel wire. Of course it will not work. When I told this shop to the shop owner of a private shop in Akihabara, he introduced a craftsman who had been building a light bulb a while ago. He offered parts for manufacturing bulbs and data on manufacturing. With this process the process has progressed considerably and has reached the present. and my final goal is a vacuum tube. However, it has been more than 50 years since the production of vacuum tubes .... Detailed data on manufacturing has already been lost, and I am still traveling in a blind alley.

chrisklinger11 months ago

Fantastic job! Thank you!

2SC1815 (author)  chrisklinger11 months ago

Thank you!

kojis9211 months ago

Great work. Amazing manual build.

2SC1815 (author)  kojis9211 months ago

Thank you!

billbillt11 months ago


2SC1815 (author)  billbillt11 months ago

Thank you!

throbscottle11 months ago

Totally amazing! Are you going to try doing nitrogen fill like commercially made bulbs?

2SC1815 (author)  throbscottle11 months ago

Thank you for watching this project!

My final goal is not a long-lived bulb but a vacuum tube that operates as an electronic component. I think that there is no use of nitrogen gas to keep the inside of a vacuum tube high vacuum so as not to disturb the traveling of electrons. Thank you!

berniefra11 months ago
Hi, does someone know where to get this dumet wire in Europe (Germany)?
2SC1815 (author)  berniefra11 months ago

I bought du-met wire from this company.

The nilaco corporation.

And dumet wire is here.

(1500 yen is about 12€.)

The nilaco corporation also deals with tungsten wire, nickel, titanium, covar wire etc.

It's the same wire as the leads going into a manufactured light bulb, so you could break open a blown bulb and re-use the wires.

gm28011 months ago

Very interesting project. Now you have to push the envelop and make a vacuum tube as well. Try a simple diode or triode. Since you have the glass part mastered, next has to be a vacuum tube. Thumbs Up!

2SC1815 (author)  gm28011 months ago

Thank you for watching this project!

I tried to make six in total, three diode tubes and triode tubes. However, the plate current flows only about 10 μA. (The picture shows a triode tube in exhausting. but it is not working properly in later test.)

The possible cause is that the vacuum level is insufficient. The inside of a typical vacuum tube is about 10 ^ -4 to 10 ^ - 7 Torr, but the vacuum level I can reach is 10 ^ -2 Torr, which is not a problem as the light bulb pressure, but as a vacuum tube it is bad. In addition, the vacuum level is further deteriorated by gas release from the plate. In order to solve these problems and to make a working vacuum tube, I need to get an induction heating equipment and an diffusion pump. However, diffusion pumps are expensive, and induction heating equipment using semiconductors are relatively fragile.

In order to solve this problem, we decided to make the diffusion pump from aluminum rod using a lathe. And I decided to make induction heating equipment by myself using Russian transmitting vacuum tube (a vacuum tube for making a vacuum tube :) ).

Perhaps, it seems that either will be completed within this year.


JamesA4111 months ago

Awesome instructable and great detail that is comprehensible. Thanks for the photos and sharing. I wonder what are the details for the specifications to make the 100 yr+ lasting in operation light bulbs? Seems you have identified details to help the light bulbs last long already. Cool other designs also I see in one of the photos. Neato!

I'm thinking you've seen before... though here is a link if not:

Here is another interesting use for a light bulb that I recently read about other than florescent (and I guess neon) bulbs being used as plasma antennas:

2SC1815 (author)  JamesA4111 months ago

Thank you for watching this project!

(This text was translated from Japanese to English using google translate)

> details for the specifications to make the 100 yr+ lasting in operation light bulbs :

I don't own document related to carbon filaments, so I do not know well about light bulbs manufactured in the past, but I will write recipe of long-life bulb as far as I know.

Besides baking the glass at 400 degrees (500 degrees in the case of hard glass) to degas, of course, Enclosing high-purity inert gas in the bulb without making the bulb high vacuum makes the bulb life longer.
This is because after evacuating the bulb, after several cycles of introducing a cleaning gas (generally nitrogen gas) at a pressure of 250 to 300 Torr and evacuating again,
Complete by sealing an inert gas (argon gas) at a pressure of 550 to 600 Torr.

The presence of water vapor and hydrogen and oxygen constituting it,
It is considerably harmful to tungsten filaments. Water decomposed with hot filaments is divided into oxygen and hydrogen. Oxygen oxidizes tungsten, and forms tungsten oxide, sublimes and reaches relatively low temperature glass.
The tungsten oxide that has reached the glass is reduced by hydrogen, the tungsten remains on the glass, and oxygen combines with hydrogen and returns to the water again.
This cycle lasts until the filament breaks while heating the filament.

recrystallization occurs when tungsten is heated, "Slip" occurs and burns out due to local heating,
This can be solved by adding a dopant materials to tungsten or by improving tungsten processing method.
And as stated on this page,
Turning the power on and off also affects the life of the bulb considerably. and life can be extended by lowering to less then supply power of the light bulb's rating.

> plasma antennas :
Although youtube video of plasma antenna could not be seen with errors,
I saw the page of wikipedia and it is a very interesting device and it is interesting also about the future of plasma antenna.
And I might research and make a plasma antenna. (but I don't know if it will succeed or not)

Alex in NZ11 months ago

That is just _amazing_! The number of different techniques and fields of expertise you brought to this is really impressive.

Thank you very much for sharing the process :-)

2SC1815 (author)  Alex in NZ11 months ago

Thank you for watching this project!

It took a while to notice the harmful effects for tungsten filament of residual water vapor. baked light bulb is still alive over 14 hours! (non-baked bulb life is 3 minute, too short!)

seamster11 months ago

Fascinating project! Thank you for sharing your steps - and welcome to Instructables : )

2SC1815 (author)  seamster11 months ago

Thank you for watching this project!
and I am planning to make a induction heater and high vacuum pump for the next project!