Rarely during product development do you get it correct on the first design iteration. Something always goes wrong or just isn’t perfect. However, like trying for a hole-in-one on a par 3, you always try for perfection but expect to need a few extra strokes. So, while I almost hit a hole in one in the first version of the Nixie Tube HAT (Part 1), a few improvements were required. In this blog, I describe the few improvements found from Part 1 and complete the design to reach stage 6 of the surfncircuits defined development flow. As with the other projects in the blog, the complete design files in Kicad, schematics, layout, BOM, are available at GitHub for use in your own projects. You can build it yourself and the PCB can also be ordered directly from Oshpark.

Design Improvements
Increase bandwidth of the display
As mentioned in Part 1, R2 and R9 were reduced to 1kohm to increase the display bandwidth. The bandwidth is plenty for the water meter application I’m building and should be good enough for a music VU meter. Lower resistances on R2 and R9 will increase the bandwidth further but should be tested to verify.
Updated Silkscreen
The PCB silkscreen was improved to help attach the Nixie tubes and power supplies to the PCB. For this, the Nixie tube connectors J3 and J4 now show the A, C, and Aux_C for the anode, cathode and aux_cathode pins of the IN-9 and IN-13. While the IN-9 only has an anode and cathode connection, the IN-13 also has an aux_cathode connection.

The connector for the power supply silkscreen was corrected to show the +70v, +5v, Enable, and GND pins. This will help eliminate the chance the Nixie tube power supply is connected incorrectly.

Added a 1k resistor to the enable pin
When designing the Nixie tube power supply I learned to add a 1k resistor to the enable pin connection. During development of the Nixie supply, I destroyed a Raspberry Pi when I accidentally connected the enable pin to 5v. By adding a 1k resistor in series with the enable pin, accidentally touching the EN pin connection to 5v won’t destroy the raspberry pi. This is a good safety measure.

Do you see an improvement?
In reality, no project is really ever complete. There are always improvements to be made. If you see something that can be changed or modified, or new design improvement added, please send me a note below. I hope you enjoyed this project and can use it for your next linear display for the Raspberry pi.




I’m looking at this for a temperature bar in conjunction with also building an integral in-12 Nixie tube clock. I see from the photo that you’re using your power supply to power your hat as well as the raspberry pi. I’m assuming your power supply’s input voltage is 5v (amps?). So here is the few questions: Can I split the power supply’s output to also drive the Nixie’s? through two HV5622PG’s? Can I use your header pins for 5v supply, gpio10, 11, and 17 with 4.7k ohm resistors to drive the HV5622PG logic?
Hi,
Yes, the Nixie tube power supply can use a 1amp 5v supply to power a couple of Nixie tubes. For a string of 6 nixie tubes and a Raspberry pi, I use a 5v (2Amp) supply. The 5v to the Nixie tube power supply has a feed through to power other 5v electronics.
Mark