Classic Nixie Tube Clock with Bluetooth/WiFi capability
The nixie tube displays are fascinating to look at and were used in many Cold War era test equipment and various other devices. I was inspired by many nixie tube clock designs to design my own as well. On the surface, this project seemed very easy. However, the deeper I got into the design phase, I quickly realized I had to learn more about designing switching power supplies to get a high voltage rail for the nixie tubes from a much smaller wall adapter. As well as wanting to learn more about transmission lines and impedance matching for implementing a Printed Inverted F Antenna (PIFA) or in general terms, a microstrip antenna with a resonant frequency of 2.4GHz for WiFi capability.
- WiFi-enabled for configuration and additional features such as a remotely accessible timer or scoreboard.
- RTC module or integrated RTC to allow time-keeping during failures
- A low voltage power supply such as 5V to 12V wall adapter
- Rotary encoder w/ momentary push button for user control
- Direct drive for Nixie Tubes
- 160V to 180V power supply @ 10mA to drive 4 IN-12A Nixie Tubes (Each tube sinks about 2.5mA at 180VDC)
- Chose nRF52832 MCU from Nordic Semiconductor w/ integrated RTC and Quadrature Decoder for encoder
- Serial to Parallel Shift Register w/ open-drain latched outputs (also 50V Zener clamped common, discussed later) for tubes
- LM3478 Boost/flyback controller for generating 180VDC
- 3.3V boost regulator for MCU
- 5V boost regulator for shift registers and other IC's
- Barrel jack input for wall adapter allowing about 5V to 12V input (protected by TVS diode and Schottky diode, as well as fused)
- I2C Temperature sensor for monitoring overall system temperature
- Later decided to implement this as a multi-board project with a front panel PCB for Nixie's and a control board for MCU for modularity
Just a picture of one of the nixie tubes while testing out the power supply section.
Purchased a power supply module designed by Paul Andrews on Hackaday to learn more about designing and understanding switching power supply design. Ultimately, I decided to implement most of Paul's design with minor changes such as a different switching FET due to availability of parts. I'll have to monitor the temperature of the new FET although the FET I chose has similar paramters such as Rds(on), Coss, Vds max, gate drive voltage, etc.
Also spent alot of time in this month learning more about Nixie Tubes and how they work and how they are driven. I learned that they could be driven by a "lower" voltage shift register if the digits that aren't being used at the time are clamped to a common 50V zener on the driver. With this configuration, a dedicated HV driver chip or discrete HV transistors aren't needed. Since the maintainence voltage to keep a Nixie lit after ignition is around 120V, we'll have to lower the supply voltage to around 160V. This will prevent other numbers that aren't being selected from lighting up partially since when they are clamped by a 50V zener, they'll be sitting at around 110V, which is not enough to maintain ignition in a Nixie. By the way 160V is enough to ignite the Nixie's as well, although a common figure tossed around is 170V to 180V.
Finshed most of the schematics, just refining and adding additional features
Finished PCB layout of the control board and figured out how to create a Multi-Board project within Altium Designer.
No ground or power planes added yet, but looking good so far! Next up is adding planes and connecting 3v3 through via's to the power plane, as well as adding a via fence for the microstrip antenna ground connections.
I was also thinking about how I could test the antenna S-parameters upon assembly, but this may not be possible due to lack of access to a network analyzer. Unfortunately, the cheap mini VNA's on Amazon and such do not go up to 2.4GHz. Regardless of this shortcoming, I did include a pi-network filter just in case this testing becomes possible.
After finishing the control board, I'll start working on the Nixie Tube board or front panel of the clock. The tricky part with this portion will be finding an adequate source for IN-12A Nixie tube sockets, since the pin dimensions are not consistent enough for a direct IN-12A footprint.
Completed layout of the control board with via shielding along microstrip transmission line, and via stitching GND planes together:
