Nixie Pendulum Clock - Concept

A pendulum clock with a NixieTube (IN-12b) that plots the time in the air it passes.

Similar projects worth following
It is a pendulum clock with a NixieTube (IN-12b) that plots the time in the air it passes.

A magnet at the bottom of the pendulum gets pushed by an stationary air coil, as soon as it "looks" directly down. The positions of the plotted numbers are acquired by an i²c accelerometer. The pendulum is ball-beard.

If I would build this clock angain, i would use a ESP8266 or ESP32 smd chip so it could be flashed over the air (OTA!) but then you'll need a logic-level-converter from 3,3V to 5V for 74141-Input and IRL540N MOSFET.

You can find here:

I am scrapping this project. I tried to make a new video, but critical timing in the code, did not work out to display the time properly after a while in the book shelve. This was a fun project - but the next are in line. Ill scrap this build for something new. You'll see parts of this in the near future.
Also i made a video with all my old outtakes:


complete wireing. fixed grd & V+ at coil driver. Addet a current limiting Resistor for the Nixietube (Current should be around 4-7 mA for nominal glowing numbers)

Adobe Portable Document Format - 75.55 kB - 05/09/2021 at 11:25



more comments

ino - 10.07 kB - 05/07/2021 at 22:00


  • 1 × Arduino Nano V3
  • 1 × MMA_7455 Acceleratormeter
  • 1 × RTC 1307
  • 1 × IRL540N Discrete Semiconductors / Power Transistors and MOSFETs
  • 1 × DCDC Converter - 5V to 170V

View all 23 components

  • 1
    1. Build the base.

    Connect all the wooden parts exept the arm with wood screws so you get a stand like this:

    (Foot + stand + standof)

  • 2
    2. Mount the arm with 2 ball-bearings

    b)  Drill a hole centric in one end of the arm, so the ball-bearing fits in it.
    The ball-bearing looks like this: 

    and they were stuck in both ends of the hole. A small but long M2 screw goes through the ball-bearings with got tighten by counter nut.

    On the other end of the arm drill a hole for the cables from 74141 to nixie-tube. 

    On the front of the arm i used a Nixietube-Socket ( I stuck in the nixie-tube and and after that i solderd the cables (10 Cathodes, 1 anode with 170 V - The currentlimmiting resitor shouls limit the current from 1-3 mA's). Then the socket gets screwd onto the front of the arm.

    Now find a screw to mount the arm it on the small pice of wood (the stadof), or use a 90° aluminium angle to fixate the arm through the ball-bearing on the "too" long M2 screw. 

    If all went well: after you tip the arm it should swing very long and it will move for minutes!

  • 3
    3. Electronics and correct Magnet placement

    See scheematic for wiring. Build the backside parts onto the backside of the stand onto the wood.

    Build the "Pendulum" parts together and secure them on the backside of the arm. The Accelerometer needs to be as close to the axis (ball-bearings) as possible! Otherwise the centrifugal force will be more present in the meassurement than the gravitational force.

    Find a way to secure the flexy-cable, from the boackside of the stand to the backside of the arm, while the arm still can swing freely, without much dampening of the swinging.

    Now connect Ground, +5V and Coild-Trigger from D13 from the backside of the stand to the flexy-cable and on the arm from flexy-cable to the rest of the electronics.

    Check every module on functionality (RTC, Accelerometer, Nixietube - all numbers)

    Upload and set the RTC.

    Now we need to get that pendulum keep swinging. Check if position: "arm straight down" equals 0° on the Accelermeter (MMA_7455). The Coil should induce a pulse on the magnet. See with some other magnet, witch two sides of the cubic magnet are the poles - one should attracht the coil and the other side should repell from coil - we need the repelling force configuration for this project. Mount the coil on the upside from the wooden foot and glue the magnet on the arm at the backside bottom position WITH: repelling force configuration / orientation!

View all 4 instructions

Enjoy this project?



JP Gleyzes wrote 07/08/2022 at 19:23 point

I do really love this project (being a clock fanatics !)

Thats's a pity that the PoV effect is so limited...

  Are you sure? yes | no

j wrote 07/08/2022 at 19:31 point

Thanks a lot!
If the code would be perfect, it would look better and the frequency of the pendulum could be designed to be higher. 

  Are you sure? yes | no

JP Gleyzes wrote 07/08/2022 at 19:42 point

Did you try to use the dual cores of the Esp32 ?

Hummm sorry I read too fast. You haven't yet used an ESP32... As you said you should, they are fast and powerful

  Are you sure? yes | no

j wrote 07/08/2022 at 21:00 point

The clock speed is/was not the problem. It was my codes semi automatic timing of the "left2right" and "right2left" offset delay. .. at least i think it was.

  Are you sure? yes | no

Ken Yap wrote 05/13/2021 at 00:16 point

At first I though it would be possible to make the duty cycle longer to increase the brightness but then I realised that would blur the digits. Still it's an interesting way to use only one tube.

  Are you sure? yes | no

j wrote 05/13/2021 at 09:50 point

To increase the brightness you cuold give the tube more mA's ny decrease the current limmiting resistor. I think i'll make another video in the dark.

  Are you sure? yes | no

Ken Yap wrote 05/13/2021 at 10:38 point

I think there is less range for overdriving a gas discharge tube compared to say LEDs.

  Are you sure? yes | no

j wrote 05/13/2021 at 10:43 point

true, need to test it with a nixie-tube-tester sometimes. (

..but as i remember its possible to overdrive them a bit, but they warm up. I never pushed it into destruction.

  Are you sure? yes | no

Dr. Cockroach wrote 05/10/2021 at 18:27 point

What cool take on a POV pendulum clock, well done.

  Are you sure? yes | no

j wrote 05/10/2021 at 18:38 point

Thank you very much! The POV effect barely kicks in, but the time is recognizable, due to short (4ms) pulsed numbers.

  Are you sure? yes | no

Mike Szczys wrote 05/10/2021 at 15:50 point

That is a bonkers build! At first I thought this was going to be a bar-graph tube that just lights up pixels for POV display... so much more interesting as an actual Nixie. Do you worry about the Nixie failing due to vibration? I have no idea how these parts are rated and what they can stand up to.

  Are you sure? yes | no

j wrote 05/10/2021 at 16:32 point

Thanks! A previous project tought me, those tubes are fast. So that idea came up by it self. When tapping a tube (ZM1180) you could hear the numbers flapping around just a bit. But here the swinging arm is stored on a ball bearing and the push from the coil on the magnet is barely audible. I dont think the tube it self degrades, but a strong impulse might bend the number-plates.

  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