Cool replacement for nixie tubes

Using microLEDs and clear PETG filament, I have discoverd how to replicate a nixie tube, and the effect is super, cheap and reproducable.

Similar projects worth following
The LEDs are simply mounted on a PCB (or glued to a piece of material. Each LED lights up one digit of the "nixie" display.

The 1.75mm clear PETG Filiment is warmed with a hairdryer and bent around a mould to create each digit. Then, to make it light-up, the appropriate sections of the PETG are lightly sanded , this causes refraction of the LED light that is piped into the end of the PETG - thus giving the glow effect - the 10 digits (0-9) are then glued on top of each other... the ends of the PETG are polished and glued to the top of the microLEDs - 1 LED per digit, Turning on the LED makes that digit light up.

Using an old test-tube, which was cut-off at a good height creates the éffect; of a tube, but it is of course non functional and not really necessary.
You can choose the colour of the elements or even have colour changing "Nixie"displays...
As only 1 LED is on at any one time, the current required is minimal ( depends on the LED)

The first step was just to test if I could transmit light through the PETG transparent 3D printer filament, and I did this in a project with my son. We made a 3D printed diesel locomotive with a string of LED Christmas lights and piped the light from the LED lamps to the appropriate lights inside the model. (see photo).

Although the PETG fiber is not as clear as a real glass fiber, the effect is still quite impressive. (see photos)

Following on from that, I scratched the side walls of the PETG filament and that gave an interesting effect, but not the effect I was looking for.. so I went on to sand the sides of the filament an that produced a fantastic effect, plus I can control which bits of the fiber light up and which not, thus making it easy to invisibly feed the fiber with the LED and only where the fiber is sanded, does the light escape.

more details soon.

  • 1 × Clear PETG Filament (1.75mm) Each digit is made from a short length of filament.
  • 10 × SMD LEDs (color of your choice) choose a color that you like to see in the 'tube'
  • 1 × PCB, stripboard or just a piece of cardboard is good enough The LEDs can be mounted (glued) onto the base material. or soldered to a PCB.
  • 1 × Sandpaper 100 grit
  • 1 × 3D printed base Will be on thingiverse

  • The blue LEDs arrived

    david.reid04/20/2018 at 06:09 0 comments

    So, this is log 5.

    I've been busy over the past 2 days. The new BLUE LEDs arrived ( 3528 SMD PLCC) 500mcd. Wow these things are bright! The  white block in the photo is one LED ( hobby knife to give some scale to it)

     I created a base where it is easy to mount the LEDs on the bottom and slide the filament in the top. It was printed on my 3D printer - and the first prototype worked very well. Took 10 mins to design and 14 mins to print.

    This is the bottom view. 


    So, my idea fort he base seems to work, in this one, I can fit 12 LEDs in the base, allowing me to use 2 LEDs to drive some numbers.

    for instance, the number 6:

    Now let's light the blue touch paper and stand well back!

    Hey, looking good.

    If I make the base in Black PLA, instead of white, the bright-spots from the LEDs will be covered - and then only light will escape via the digit itself.


    I want to experiment with the idea from @Chris Knight - to put foil on the open end and see if I can reflect the light back into the filament.

  • the countdown ...7, then 6...

    david.reid04/13/2018 at 07:06 0 comments

    OK so my 7 worked... now to add a second digit, and see if the effect is OK or the project is a bust.

    So I made a 6 - but this time freehand, just to see if it was possible.

    Then I stuck it to the back of the 7 and lit it up.

    Again, the lighter bit at the bottom is only there in the photo - the 6 lights up, but the very top bit is not as bright as I want.

    I think my next try will be using the mould and bring the ending leg of the 6 back to the bottom and use a second LED.

    If I don't sand the vertical return bit, then it will not emit light - and so appear to be invisible.

    So far, it is looking good,

  • Let there be light! - and there was!

    david.reid04/13/2018 at 06:54 0 comments

    In this log, I first needed to rough up the outside of the PETG, so that the light from the LED would diffuse out and look like a nixie.

    I did this with 100grit sandpaper, rubbing the areas I wanted to shine lightly with the sandpaper. It only takes a few seconds to rough up the surface...

    Then the end of the PETG that I want to shine the light into needed to be polished a bit - I put some toothpaste on a cloth and rubbed the end perpendicular to the cloth to make it smoother and without rough bits. It takes about 30 secs of slow circles on the cloth to finely polish the end of the PETG.

    Now the exciting bit, I used some LEDs that were lying around in the workshop. they are orange 3mm with leads.

    I used a drop of superglue on the top of the LED ( it had a flat top) and glued the PETG to the LED.

    Then the moment of truth... I applied some power to the LED and.....

    wait for it....

    It worked! (the bright effect at the bottom is not really as bright as in the reality the whole digit has an orange glow).

  • Creating the first moulds and numbers

    david.reid04/13/2018 at 06:40 0 comments

    To be able to create multiple nixie tubes ( the ultimate goal is... of course.... a clock) I decided to create some moulds that the PETG can be formed into and then I would get consistency between the tubes.

    I designed them in GIMP 2.8, using the LOGO-Typewriter Font option. as this was the quickest way to make clean numbers.

    Each number was created separately and the additional vertical line added to route the PETG to where the LED will be.

    With these PNG files, I loaded them into the FlashPrint software and it has this really cool function that converts 2D images into 3D objects using the color in the picture to determine the height. So using only black and white, I get a cool mould. Where the height of anything white is 4mm and anything black is 0.5mm - then under that all is a base of 0.5mm. This gives me a mould of 4.5mm where the cutout is 3.5mm deep.

    Each mould takes about 10 minutes to print in High-Res ( it's only 25mm wide and 35 high).

     Here is the mould for the number 7 with the PETG in place:

    I warm the PETG with a hairdryer until it is a bit flexible and then form it into the mould. Wait a few minutes for it to cool and out pops a transparent PETG  '7'.

    I made moulds for all the numbers.

  • The first Steps - playing with PETG filament

    david.reid04/12/2018 at 05:53 0 comments

    The first step was just to test if I could transmit light through the PETG transparent 3D printer filament, and I did this in a project with my son. We made a 3D printed diesel locomotive with a string of LED Christmas lights and piped the light from the LED lamps to the appropriate lights inside the model.

    I made a small block with holes for the LEDs and the fiber, 4 on one side of the block and 3 on the other side of the block. so the fibers could go to the front or the rear of the train. The fibers were then warmed with a hairdryer to make them flexible enough to position correctly through the holes in the train body, when they cool, they keep their shape making final assembly easier.

    For the train, the ends of the fibers were 'polished' using toothpaste and a cloth, This brought the light intensity up by almost double. 

    Although the PETG fiber is not as clear as a real glass fiber, the effect is still quite impressive. (see photos in the details section).

View all 5 project logs

Enjoy this project?



johnjoji wrote 10/02/2019 at 09:09 point

@david.reid I changed on your project while looking up nixie projects. I soon realised that procuring nixies for my personal project and getting them to work might be prohibitive. Your project has caught my attention and wanted to follow up as to it's status. How's the progress?

  Are you sure? yes | no

david.reid wrote 04/20/2018 at 05:45 point

@Yann Guidon / YGDES  This is true Yann, but there is so much light available from these little 500mcd LEDs that the effect is quite amazing.

I was experimenting last night and just sanded down the front side of one PETG number, and that stops a lot of light leaking out to the adjacent numbers - so I have reduced the re-transmission of the light in the other numbers, this enhances the number that is selected,

I've made the first base last night - see the latest log ( log 5) for details and photos. I received my BLUE LEDs yesterday and they are really cool.

Check it out.

PS thanks for the 'like'.

  Are you sure? yes | no

david.reid wrote 04/19/2018 at 21:29 point

@Yann Guidon / YGDES  @Chris Knight  ps don't forget to like the project...Ive only got 4 likes so far... could do with more.

  Are you sure? yes | no

Chris Knight wrote 04/19/2018 at 22:26 point

Ah yes, thanks. And I look forward to when you ramp up production of 10-digit units on a PCB with a glass enclosure. :D

  Are you sure? yes | no

david.reid wrote 04/19/2018 at 21:27 point

@Chris Knight  thanks for the idea. . It is still leaking a lot of light at the open end. But for the numbers like 6 8 4 I now bring the end back to a second led.  Tomorrow I'll post more photos and a new log.

I made  a base which the leds fit into the bottom and the filament goes in the top...14 mins to print on my 3d printer.

Initially i wanted to paint the open end black...but your idea of foil is better- if it works.

I'll see if I can hack something this weekend..

Cheers David.

  Are you sure? yes | no

Chris Knight wrote 04/19/2018 at 18:38 point

Nice work, how much light is transmitted out the end? Would try a bit of aluminum foil or foil tape to create a mirror at the end to keep the light bouncing until it refracts out.

  Are you sure? yes | no

david.reid wrote 04/13/2018 at 11:29 point

Now you can see some photos of the real thing.

Still needs work, but free time is short - hope to test the next bits tonight and get them published in a log, before the competition closes.

  Are you sure? yes | no

Yann Guidon / YGDES wrote 04/12/2018 at 01:15 point

Is it similar to #"Lixie", an LED alternative to the Nixie Tube ?

I'm looking forward to seeing your own drawings.

  Are you sure? yes | no

Yann Guidon / YGDES wrote 04/19/2018 at 21:57 point

Oh I get it now :-)

Diffusion of a lightguide is not easy though, as most of the light will be dispersed near the light source.

  Are you sure? yes | no

Similar Projects

Does this project spark your interest?

Become a member to follow this project and never miss any updates