There is something about nixie tubes; I don’t know if it the romance invoked by steampunk, my personal fascination with Cold War era Soviet equipment or just the lure of that high voltage orange glow. I have wanted to do a project using nixie tubes for a long time so when my bedside clock failed it was the perfect excuse. Nixie clocks are a staple in the hacker community, with dozens having been featured on Hack A Day over the years. While the common theme is the use of multiplexed displays driven by a microcontroller, I intend to go the analogue route with my design based largely off an application note produced by Texas Instruments in the early 1970’s for the SN7441 – a BCD to Decimal Decoder designed specifically to drive “Gas Filled Cold-Cathode Indicator Tubes” or in English the Nixie tube.
Once I had finished the HV supply, I couldn’t just settle
for a static number. Before I called it a night I had to make the numbers
change.
Given that I have a tray of K155ID1 driver ICs switching the
HV is pretty simple. The K155ID1 (SN74141) is a BCD to Decimal decoder and
display driver designed for nixie tubes. The HV is supplied to the nixie anode
via a 7.5kΩ current limiting resistor and each of the ten numeral shaped
cathodes are connected to the ‘outputs’ of the K155ID1.
The K155ID1 is driven by a MC14510, a 4000 series pre-settable
up/down BCD counter. The clock is provided by a small clock pulse generator my
Dad built 20 years or so ago.
The reader must excuse the untidy breadboard prototype, in
my defense it was late. The isolated HV supply can be seen in the background and
the aforementioned clock generator to the right.
By the time I called it a night, or should that be ‘called
it a day’ given that it was 1AM? I digress. By the time I called it a night I had a single nixie that counted from 0 – 9 and
started again. A single counting nixie wasn’t enough though; I was already
formulating a plan to have a larger counter. This four digit counter would
prepare and test the prototype display driver that would be used when I finally build the prototype clock circuitry.
With parts sorted the next step was to develop a testbed, just a simple High Voltage supply I told myself. Like that ever works, once I started I had to keep going and before I knew it it was 1AM.
Now before we move forward I must remind the reader that High
Voltage is dangerous and deserves your upmost respect and undivided attention.
High Voltage power supplies for nixie operations often contain DC voltages in
excess of 150V and AC voltages at mains potential. The reader must be confidant
in working with these potentials before undertaking a project like this. It
should be noted that some countries, such as New Zealand, consider this to be
prescribed electrical work and thus it must be done by an appropriately licenced
person.
Right, with the appropriate safety notices posted I started
work on my HV supply. Given the IN-12B nixie tubes require an upkeep voltage of
200V, and the New Zealand mains supply is 230V 50Hz AC I had an ideal supply at
my fingertips, all the I need to do was rectify it, right? While technically
correct the last thing I wanted to do was have my HV supply at mains potential,
even if it was just for testing purposes. My solution was to back to back two identical
power transformers, providing full isolation between my supply and mains. These
transformers were originally salvaged from standalone emergency lighting units whose
batteries’ had outlived their usefulness. A Greinacher voltage doubler on the
output of the second transformer proves rectification and the final, stable output voltage.
With a stable HV supply I was able to light up my nixies for
the first time. In a word: Gorgeous. As those readers that have worked
(played?) with nixie tubes in the past will know, there is something mesmerizing
in that orange glow. But static numbers wasn’t enough, and having come this far
I had to continue, I had to make the numbers change.
Upon deciding to undertake this project my first step was to
take stock of what parts I have available and what I need to acquire.
I didn’t have
any nixie tubes, so they had to be ordered. I settled on six IN12B numeric
tubes, six IN1 numeric tubes and 20 INS1 indicators. All the nixie’s were new
old stock components manufactured in the USSR and were ordered off of an ebay
seller that I have purchased from in the past “alexer1” and came from the Ukraine.
I already had about 20 or so K155ID1 driver ICs, the Soviet equivalent of the SN7441,
which I had acquired as part of a bulk lot of surplus New Zealand Defense Force
electronic components. Incidentally, why did the NZDF have Soviet spare parts? A
quick dig through my box of logic IC’s suggests that most if not all of the
required IC’s were present. Passive components were not an issue either with a
large range in stock. Once the nixie’s arrived the first step would be to
establish a high voltage test bed.
I spent a long time trying to decide if I could wire together a simple couter for my Nixie Jewelry project. In the end I decided it was easier to just use a small micro, but it is good to see a counter being used.
Nice looking tubes, I need to find some of them one day. I also liked the HV source, seems to be a simple solution. I suppose using a 2:1 power transformer would solve the issue of 120 VAC mains in North America?
Thanks Eric, Check out SOVCOM (http://stores.ebay.com/SOVCOM) on ebay for the nixies, I have found them to be very easy to deal with and their prices are pretty fair too. The HV supply was 'quick and dirty' so that I could test the nixies, my plan is to build a switchmode boost converter to provide the HV in the final clock. For those in North America a second voltage doubling stage on the output of the second transformer would give the required voltage while still maintaining isolation form the mains.
Another way to get a HV supply is to buy a SMPS converter module for a few bucks from eBay that generates it from low voltage. Also makes your project mains independent. You can also build you own boost converter if you feel buying isn't sporting.
I spent a long time trying to decide if I could wire together a simple couter for my Nixie Jewelry project. In the end I decided it was easier to just use a small micro, but it is good to see a counter being used.