psu logo back to main page

Nixie Tube Clock

Browsing Mike Harrison's Nixie Clock page on his website and looking at the examples of clocks that other people have built inspired me to build my own.

What is a Nixie Tube? See Mike Harrison's site and Wikipedia.


Photos + Construction

Front view of the clock. The long actuator time setting buttons are within the recess on the front. The holes in the top were drilled and filed out, luckily the ABS plastic is easy to cut!
Rear view of the clock. Old DC connector hole was carved out and a figure-8 mains connector was fitted. The holes on the rear panel were blocked up with insulation tape.
Inside of the clock before assembly. White wires go to the front panel setting buttons. Bank of MPSA42 transistors can be clearly seen on the PCB as well as some of the logic drivers peeking out from underneath.
Homemade valve PCB. Drill template was made by measuring the valves with my vernier calipers and using Microsoft Visio to draw a 1:1 template. This was stuck to the PCB and drilled out. The anode resistors can be seen hanging off the PCB. Copper was hand carved with a craft knife and the remainder peeled off using a hot air gun.
My usual construction underneath the PCB - matrix board with tinned copper wire (component legs) for power supplies and transformer Enamelled Copper Wire for signals. Some of the copper pads around the mains input area were removed for safety.



  • 4 x Hivac XN-11 Nixie tubes from Hollow State Electronics on eBay UK. Very satisfied with the service as they replaced one of the broken tubes quickly and with no fuss. Recommended!
  • A redundant set top box plastic chassis from work, 150mm x 100mm x 25mm
  • Various logic ICs, semis and passive components as per the schematic from Mike's site
  • Bits of protoboard and copper clad board for building the circuits on.



I used Mike's default circuit to start with, but it is wired for 12-hour display and not the 24-hour which I prefer. So I modified it as per his instructions but it wasn't working particularly well. It seemed that the passive component circuit (R10, C2, D5) that Mike was using to reset the 1/hours and 10/hours counter ICs wasn't working properly.

I swapped it for a 4011 quad NAND gate (pinout) and used 2 of the gates to provide the reset signal when the display ticks over to showing "24" on the hours. It now reset's flawlessly for the price of an extra logic chip. I could have used the other half of the 4011 to buffer the 50Hz signal and remove the need for the 4013 but I'd already built that part into the circuit, and the 4011 was nicely piggybacked on top so I couldn't be bothered.

I've drawn my ammendments onto the original schematic and will make it available here shortly.