So my question is about nixie tubes. To skip the whole designing-my-own-controller part I settled for the arduinix (schematics) solution, which some of you might know. I am currently using an Arduino Uno to drive this controller.

The problem

After soldering the kit, I connected one tube to one anode (170V) via a 10K resistor, and the 10 cathode pins to the outputs of one of the two k155ID1 drivers. Because I was in prototyping fase I used jumper wires to connect the anodes and cathodes to the outputs of the driver ICs on the arduinix board.

I used a simple script to alternately send a 0 ( A:0 B:0 C:0 D:0 ) and a 1 ( A:1 B:0 C:0 D:0 ) with a 1 second interval in order to test if the tube is actually displaying the values I want to show.

Here I run into the problem: The nixie tube just seems to ignore these values I set to the IC. Instead, it's is lighting up multiple numbers simultaneously.

I tried disconnecting some jumper wires from cathodes to see if this would make a difference. When disconnecting the cathode corresponding to the number it will turn off, but when reconnecting these numbers light back up again :/

I did a visual inspection of all the soldering connections on the arduinix board; no gates look like they were "accidently soldered together"

Also I replaced the tube with another; same behaviour occurs.

Also I tried the second driver; the same behaviour occurs. It occured to me that maybe BOTH drivers are broken, but it seems rather unlikely..

I also tried verifying the driver IC's behavior by just measuring the voltages from the driver IC. I noticed a very small difference in the voltage (in the range of 0.05V, around a 1.6V base) that corresponded with the 1 second interval that I was sending from the microcontroller. Is this a correct way of verifying the IC's workings? Is this voltage difference enough to properly sink one of the cathodes?

This is the relevant part of the schematic; basically I hooked up a single tube to the driver IC, the high voltage is provided by the arduinix circuit and even the routing from the arduino GPIO pins to the driver IC is taken care of by the arduinix circuit.


simulate this circuit – Schematic created using CircuitLab

I measured voltages while the tube was on and these are my measurements:

  • Only nixie: 130V
  • Nixie + 10K resistor: 152.7V
  • Cathode 0: 3.2V
  • Cathode 1: 27V
  • Cathode 2: 61.5V
  • Cathode 3: 3.2V
  • Cathode 4: 50V
  • Cathode 5: 3.2V
  • Cathode 6: 57V
  • Cathode 7: 3.3V
  • Cathode 8: 33.5V
  • Cathode 9: 26.5V

Looking at these I can't make sense of why this is happening... Hopefully it helps one of you guys to tell me what is going on here..

Picture of the situation:

enter image description here

  • \$\begingroup\$ The voltage could be too high. What is the voltage on the cathodes (with respect to ground) when all the numbers are on? \$\endgroup\$
    – Pentium100
    Commented Apr 13, 2016 at 8:11
  • \$\begingroup\$ A schematic would really help to spot problems. \$\endgroup\$
    – greg
    Commented Apr 13, 2016 at 13:46
  • \$\begingroup\$ Looks like you have outputs as inputs on those pins going to the Nixie Tube. \$\endgroup\$ Commented Apr 13, 2016 at 16:32
  • \$\begingroup\$ I added some info to answer your questions! \$\endgroup\$
    – Pepf
    Commented Apr 13, 2016 at 19:07
  • \$\begingroup\$ I had the same issue, the problem appeared to be not good enough "connection" between the male jumper pins and the breadboard. When I touched/moved a little one of the output jumper pins of the pins the multiple digits enlightened problem suddenly disappeared. bottomline: check your jumper connections and touch them a little \$\endgroup\$
    – joppo
    Commented Oct 24, 2016 at 22:17

2 Answers 2


Could be the IC itself that is faulty or a wrong ground connection.

Have you tried testing each component separately?

  • Nixie tube alone, apply 170V on anode and connect each cathode in turn, checking each digit glow on their own.

  • Once this is ascertained, build a circuit on breadboard with a single driver chip. Set ABCD inputs to GND. Only the 0 digit should glow. Check all possible combinations of ABCD input. Depending on the chip, starting from 1010, all digits should blank.

  • If not working, change the chip.
    If all working fine on the breadboard, then check the PCB for shorts or cut traces.


  • Measure the current through the anode. Don't go above the rated current. Usually, 3mA for smaller tubes and 8mA for large ones.
  • Make sure the GND of the 5V and 170V supplies are actually connected together.
  • \$\begingroup\$ I checked the nixie tubes already separately, they seem to work fine.. Probably the driver chip is broken, I ordered new ones to create this test circuit with. Ill let you know if this improves things! \$\endgroup\$
    – Pepf
    Commented May 3, 2016 at 17:25
  • 1
    \$\begingroup\$ Switched out ICs, everything works fine now :) \$\endgroup\$
    – Pepf
    Commented May 4, 2016 at 7:52

The 74141, and i assume also the k155id1, have zener-protected output drivers, so the voltages you are seeing in the 25v-60v range are somewhat consistent with the pin off state (though i would have assumed them to all have about the same voltage?).

The working principle is that the tube in the "on" state will pull the anode voltage down until the ~40V higher "off" cathodes are unable to neither light nor maintain a glow.

The behavior you are describing seems consistent with the IC not functioning correctly.

This is a rather simple IC, so this can pretty much mean one of three things:

  • The driver is broken
  • The driver is not receiving proper input signals
  • The driver is not correctly powered

Can you measure voltages for VCC and the A,B,C,D inputs on the pins of the driver? Measure also the voltage on the signal pins from VCC so you can see that they are actually low and not just floating. As Tim Spriggs is suggesting in the comments, you may not be correctly enabling your IO pins to drive the signal inputs.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.