nixie watch project

249 views
Skip to first unread message

Kevin A.

unread,
Aug 27, 2020, 10:06:13 PM8/27/20
to neonixie-l
Recently started up a longtime nixie project I've had in mind. Quite ambitious (for me at least), and something that hasn't quite been done before to my knowledge. This is a 4 tube, bluetooth enabled (nrf52832 mcu), direct drive (not multiplexed) watch with burroughs b5755r nixies that I'm currently getting out of my head and onto paper. It will include features such as capacitive touch buttons, programmable accelerometer, automatic light sensor, removable 14500 lithium battery, and smartphone app for all sorts of display effects and options. No LED backlights on this one; its all about the neon! 

It's nowhere near complete; the design currently exists in the digital domain. I'm closing in on the first "complete" revision in the next few weeks. Current dimensions are 59mm square by 28mm deep. Its quite large to be hanging off of the wrist.. but are there really any rules when it comes to a timepiece as unconventional?

The software development will also be a tremendous task. Slow and steady wins the race here! 
3d_brd_3.JPG
kicad_3d_2.JPG

Kevin A.

unread,
Aug 27, 2020, 10:07:12 PM8/27/20
to neonixie-l
some more images

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
To view this discussion on the web, visit https://groups.google.com/d/msgid/neonixie-l/766f00c1-571f-4a0c-b7ef-7a1715849e53n%40googlegroups.com.
kicad_3d_1.JPG
3d_brd_2.JPG

Kevin A.

unread,
Aug 27, 2020, 10:07:38 PM8/27/20
to neonixie-l
3d_brd_1.JPG

gregebert

unread,
Aug 27, 2020, 10:46:28 PM8/27/20
to neonixie-l
The DCDC converter to produce the high voltage will be your biggest challenge.

martin martin

unread,
Aug 27, 2020, 11:11:03 PM8/27/20
to neonixie-l
Looking great?  Will it be close to "Cathode Corners" rechargeable?  I think David uses the same Burroughs tubes.  I have a clock or two using those.  They are clear, bright and last long...


Kevin A.

unread,
Aug 27, 2020, 11:38:00 PM8/27/20
to neonixie-l
Yup, I've already characterized the high voltage circuit outside of this project. It is capable of a continuous 5 to 180v @ 15ma boost. Should have a comfortable amount of overhead for these small tubes, which I'm planning to run at 2.5ma per anode. 

Kevin A.

unread,
Aug 27, 2020, 11:44:06 PM8/27/20
to neonixie-l
The tubes are in the same class as those used in the cathode corner watch. 

As far as the tech in the cathode corner watch goes, it doesn't have quite the range of features I'm planning for this project. To be fair, the original schematic is from 2005; and for the period it certainly is an impressive home project: http://www.cathodecorner.com/nixiewatch/firmware/firmware.html

Hopefully the project I'm working on inspires others to use modern, feature rich wireless SoCs to make more cool nixie watches! 

Kevin A.

unread,
Oct 2, 2020, 3:01:42 PM10/2/20
to neonixie-l
Hi all,

Just posted my first update with the initial hardware build, schematics, and more pictures here: 


More to come!

Ian Vine

unread,
Oct 2, 2020, 3:14:23 PM10/2/20
to neoni...@googlegroups.com
Really interesting write up

Look forward to next episode

IanV

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.

newxito

unread,
Oct 2, 2020, 6:33:48 PM10/2/20
to neonixie-l

295/5000
Great project!
Have you had any success building a power supply with the CJ5143? I completely failed trying to build a 15x15mm PSU with this transformer. I have got the 170V but only about 4mA and it wasn't really stable. But to be honest, I haven't tried to improve it at all.
psu.jpg

David Forbes

unread,
Oct 3, 2020, 1:08:48 PM10/3/20
to NeoNixie
Kevin,

That's quite a project! It has a lot of processing power. I can tell that you've put a lot of thought into the design.
I'm curious about the plans for software and features, since you mentioned that you haven't written any of it yet. 
The big limiting factor of Nixie tubes in a feature-laden device, as I see it, is the limited display ability. Not having any letters makes it hard to convey information. The decimal points are better than nothing, but require knowledge of what they're indicating. 
That's why my Cathode Corner Nixie watches only display the time.
Please keep us posted on your progress. 

Kevin A.

unread,
Oct 3, 2020, 2:02:35 PM10/3/20
to neonixie-l
Hi newxito,

Ive been able to get the CJ51-43AL to produce high voltage, but only found that about 5ma @ 170 volts was feasible from a 5 volt input (using the LT-3757afe controller at a swithcing frequency of 100 KHz). Beyond that, the transformer would indeed go into saturation and get very hot. 

If you intend to use the CJ51-43AL for more than 5ma of high voltage, then cranking up the input voltage is the only way. The part just isn't capable of more output, especially at lower input voltages where the primary current needs to be higher for any given power output. 



--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.

Kevin A.

unread,
Oct 3, 2020, 2:33:29 PM10/3/20
to neonixie-l
Thank you David! It is indeed a lot of work for one person, as i'm sure you know. This project is heavily inspired by the successful Cathode Corner watch!

The software features will be focused heavily around power saving when the display is off, in addition to using the bluetooth radio as the primary user interface.

As you've mentioned, the biggest challenge with a nixie tube display is the ability to convey complex information. That is why I've opted to use the nordic bluetooth SoC; the menu system can be built around a nice GUI on a smartphone or similar device, and all of the configuration settings can be broadcasted to the watch over bluetooth. The user can change options related to wakeup modes, display effects (both nixies and RGB leds), touch button setup, power management, or view battery status. 

I have a few modes of operation in mind. In order to save power as much as possible, the power for 3V3 logic devices can be switched on and off using IO from the MCU. The 5V0 devices, supplied by the main 2 amp 5V boost (TPS61230DRCT) which powers the HV supply, LEDs, and logic level shifter, can be shut down by the mcu when the display is not active.

It will be strongly encouraged to use the accelerometer to wake up the device. The ADXL362 is a programmable accelerometer that features a 270 nA motion activated wake-up mode. Using this, nearly all of the other peripherals can be shut off, and the 52832 can be put into a "system off" mode that consumes 300 nA with no ram retention, waiting to wake up from the accelerometer input. If the user instead chooses to use the capacitive touch to wake up, the CAP1293 controller has a 50 uA standby mode which monitors 1 of the 3 capacitive touch inputs. This means that current consumption is 185 times higher than using the ADXL362 and turning the CAP1293 off. The user will have a choice to use both or either as a wakeup and display trigger. 

As far as information to be displayed on the nixies themselves, I wanted to include options for both time (12/24 hr), date, battery status, and possibly an "acknowledge" display routine whenever the user sends a wireless command from a connected bluetooth device. 

For the decimal points, I planned to use them more for visual effect when displaying time. I want to include a few display effect options in the soft menu, such as crossfading, slot machine effects, and using the decimal points as a seconds counter (either blinking, or in other modes such as a "ping-pong" effect where it will bounce or scroll continuously across the tubes). A lot of what I can and cannot include will be limited to the 512k flash on the 52832. Bluetooth stacks use up a lot of memory fast, but hopefully using the Zephyr RTOS (which is supposed to be optimized for resource constrained devices), I can manage to get the most out of the internal storage.

I've already identified a few mistakes on the rev1 board, but I have a lot of firmware development challenges I can still tackle regardless. The firmware is shaping up to be even more monumental of a task than the hardware design. 

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.

Dekatron42

unread,
Oct 3, 2020, 4:04:16 PM10/3/20
to neonixie-l
Looks very nice!

Will you include wireless charging?

/Martin

Kevin A.

unread,
Oct 3, 2020, 8:42:32 PM10/3/20
to neonixie-l
Thank you! 

Wireless charging is a possibility. If I can get this first revision working as intended then im sure there will be plenty of feature creep to come :)

--
You received this message because you are subscribed to the Google Groups "neonixie-l" group.
To unsubscribe from this group and stop receiving emails from it, send an email to neonixie-l+...@googlegroups.com.
Reply all
Reply to author
Forward
0 new messages