RPi 6 Digit IN-4 Nixie Clock Project

[DSeabra]

Hi all, this is my first post, so please let me know if it's not kosher or if I should move it elsewhere. I would like to share my summer project with you, including the open hardware design files. I would also like to get your suggestions for further improvements on this project.

Yes, yes, I know it's just another nixie clock, but I'm really enjoying having it in house so far. Here's a few quick pictures of the clock as it is now:

The PIR sensor is used to shut off the tubes when no motion is detected (to preserve the lifetime of the nixie tubes). The nixie tubes are driven by separate BCD decoders which are in turn driven from the RPI's GPIO pins. I've programmed the clock to alternate between the current time, current external temperature (using openweathermap API) and internal temperature, and current AQI (air quality index, using AirNow API). I implemented the AQI screen because of the fire and smoke situation that we currently have in California, and it's been very handy to look at before going on walks or runs outside. The whole project was programmed in Python for simplicity.

Block Diagram

The board I designed is a rectangular 230mm x 75mm board with a 40 pin (standard 2.54mm pitch header) connector for connecting up to a Raspberry Pi and with footprints for all the parts in the above block diagram. The enclosure is a laser-cut design which can be made in a variety of materials (I chose a cheaper black acrylic material). The schematic and layout were done in KiCAD and the enclosure was designed in FreeCAD. All these files are provided in the repository below.

Design Files Repository
Bitbucket Repository

I've written about and continue to write about this project on my blog. You can find the landing page for that project **broken link removed**. You can find more posts there as well (including **broken link removed**) as the PDFs for the schematic and pcb layers (also on the bitbucket repository).

Future Improvements
There's a few things I'd like to add to this project going forward:

1. Improve the nixie footprints by rotating them clockwise slightly and make all of the holes significantly larger; this should allow the person soldering the tubes to rotate them in position and then solder them.
2. Move the power supply onto the main board
3. Remove the 40pin ribbon cable and allow a RPi Zero W to be directly connected instead (via a 90 degree header).
4. Clean up and fix a few things on the PCB
5. Try out multiplexing the tubes; if this looks good, I’ll do so on the final clock to reduce the number of necessary IOs. That will allow me to add a SPI/I2C temperature sensor to put on the board for a better sense of the room temperature (rather than using a fixed offset from the die temp of the Broadcom processor on the RPi board).

I considered swapping out the RPi for something a little cheaper that still supports WiFi such as an ESP32, but I think I'll hold off on that for now (there's a couple of other projects I'm trying to work on as well). I'm curious to know if you have any further suggestions or ideas to make this clock project more interesting.

 

[Pommie]

Hi and welcome to ETO. Nice project, looking forward to your future contributions.

Mike.

 

[gophert]

A belated welcome to ETO.

I think it looks great. I like the idea of mixing new and old school technologies - I guess you have no option if you want wifi.
I'd be interested in more info about the Nixies. How long do they last when powered 24/7? Can you limit current a bit and make them last significantly longer? Where did you buy them, ...

 

[KeepItSimpleStupid]

Don't multiplex in the traditional sense because of the time it takes the nixies to start.

Add a set of latches that you would mux or what digit to change and value.

You forgot the colons made from neon bulbs. The guy below makes a special colon tube.

There is a guy that makes new Nixie tubes by hand and they look really nice. Take a look here: https://www.daliborfarny.com/

Keep the power supplies separate. Especially the HV ones.

You can always do something wierd. You can supply one voltage always to the tube and then raise it briefly to turn on the digit and lower it briefly to turn it off.

 

[schmitt trigger]

Excellent idea using a PIR sensor to turn off the Nixies when no one is watching.

 

[DSeabra]

Pommie - Thank you! Excited to be here!

gophert - Thank you as well! From what I remember, most nixie tubes are spec'd on the order of about 6000 hours of operational time on their datasheets, although from what I've seen online (anecdotally), they tend to last quite a bit longer than that on average. Current limiting helps (I've seen about 2-3 mA as a common value that's on the low side), but below a certain current they won't turn on at all (or you might run into gradual cathode-poisoning problems as seen here: https://surfncircuits.com/2019/04/0...ry-digit-ghosting-and-heavily-oxidized-leads/). Most people use sockets or socket pins (like in this ebay link here), but I'm pretty confident I can desolder them OK. I might come to seriously regret taking that stance when it comes to replacing them, though, haha.
I bought the IN-4 tubes on e-bay; they're sold individually ($6), in a set of six (which is what I did, $15-20), or in bulk. They're all old, though - the last tubes were made in Russia in the 1970s or maybe 1980s. Unless you're getting the really fancy ones like the ones KeepItSimpleStupid pointed out (very cool, they look awesome, but your wallet takes a big hit...)

KeepItSimpleStupid - Yeah, I noticed the start-up time now that you pointed it out. Not all the tubes wake up at the same time when the clock has been "sleeping" (display off), and it's noticeable. Some take about a second to light up. I might go with latches idea if I want to reclaim some time from my GPIO pins. And those tubes look awesome!
I'm not finding the missing colons to be too distracting so far - I considered doing orange LEDs or something like that, but then I thought it'd be a nightmare to try to match the colors...

schmitt trigger - Thank you! Excellent username I'm trying to implement a feature that allows me to continuously wave at the thing to get it to transition to the next screen (for example, right before going outside if I want to know what the temperature is and I'm too lazy to take out my phone). It works but the PIR is a little finicky for that. Works great for turning it off though, I logged the wake-up times for days and didn't find any evidence of "noise" waking it up when no one was there.