Multiplexing Resources

90 views
Skip to first unread message

Xavier Burton

unread,
Aug 26, 2024, 4:38:18 PM8/26/24
to neoni...@googlegroups.com
I have been working on a modular nixie tube display for the last few months as a learning exercise, and I would like to learn muiltiplexing next in order to reduce the number of pins used for controlling them. What are some good resources I can look at? 
The pictured unit with the energized tubes is my original prototype, hence the garbage wiring and excess number of test points. The other pictures are the current pre multiplexing version.
image001.png
2024-07-28_16-28-24.png
20240731_201608.jpg
e6713a93-18b9-494e-bcc0-814ddd5d44e1.jpg

gregebert

unread,
Aug 26, 2024, 5:17:39 PM8/26/24
to neonixie-l
Try experimenting with LEDs. I've only completed 2 multiplexed display projects, and both were with LEDs. The basic principles are the same between LEDs and nixies. Where things get trickier is that nixies are subject to ghosting if the timing / sequencing isn't right, and the voltage is a lot higher. Once you have the LEDs working nicely, move-up to nixies, or perhaps VFDs. It's pretty hard to kill an LED unless you really beat it up with ridiculous current, and they are low voltage so you wont zap yourself or your scope when you are debugging it.

I have a multiplexed nixie project in the works (13-character / 9-segment Burroughs display), and I'm going very, very cautiously with it because I only have 1 of these rare displays and I cant risk damaging it.

Since this is a learning exercise, I suggest you use cheap/expendable nixies. Multiplexing requires higher current to get the same brightness as direct-drive, and tube-wearout is exponentially related to current, so you will likely reduce the lifetime of the tube. Others have reported audible noise from their multiplexed or PWM-dimmed tubes.

Senk Ju

unread,
Aug 26, 2024, 5:26:33 PM8/26/24
to neoni...@googlegroups.com
I unfortunately don't have any resources but what I would recommend doing is the following:

Use a decadic counter IC (like the CD4017) for controlling the active tube. If your clock has four digits, connect the fifth output of the chip to its reset pin. Also connect the reset pin to your microcontroller so you can reset it after the clock has been enabled. Pulse the clock pin of the chip to switch to the next tube.

Use a BCD to decimal decoder (like the CD4028B) for the cathodes. This way, you will only need four IO pins of the microcontroller to display set the current digit. 

Connect the output pins of the decadic counter and BCD decoder to high voltage transistors like MPSA 42 to control the high voltage. Then connect the same cathode of every tube in series through the corresponding transistor connected to the BCD decoder. This way you can control which anode gets HV and which cathode is grounded.


I have attached a simplified schematic of how I would do it. Hope that helps.


Xavier Burton <angrycr...@gmail.com> schrieb am Mo., 26. Aug. 2024, 22:38:
I have been working on a modular nixie tube display for the last few months as a learning exercise, and I would like to learn muiltiplexing next in order to reduce the number of pins used for controlling them. What are some good resources I can look at? 
The pictured unit with the energized tubes is my original prototype, hence the garbage wiring and excess number of test points. The other pictures are the current pre multiplexing version.

--
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/CAOt79-vM7Zudj-ts%2B%3DtH4mAogqU01VuMsrB3uH0UugRNzJbHzQ%40mail.gmail.com.
IMG_20240826_232353.jpg

Adam Piórko

unread,
Aug 27, 2024, 6:57:07 PM8/27/24
to neonixie-l
Controlling with ground (cathodes) is correct. However, controlling with positive (anode) is not correct. Controlling with the positive is more challenging. 
Attached are two options—using a phototransistor (on the left) and using two transistors (PNP and NPN). Of course, there are more options...
Of course, instead of connecting the anode control to a microcontroller, you can use a 4017.
R2 and R5 are current-limiting resistors.
nixie-anoda.png

Senk Ju

unread,
Aug 27, 2024, 7:02:02 PM8/27/24
to neoni...@googlegroups.com
Interesting! Thank you for the input.

Senk Ju

unread,
Aug 27, 2024, 8:22:00 PM8/27/24
to neoni...@googlegroups.com
Adam, to add to my previous response, you are of course absolutely correct. I don't know what I was thinking when I gobbled together that schematic. Xavier, please ignore this part of my schematic. Sorry for the confusion.

Adam Piórko <adam.p...@gmail.com> schrieb am Mi., 28. Aug. 2024, 00:57:

Mike Mitchell

unread,
Aug 28, 2024, 10:20:08 AM8/28/24
to neonixie-l
I built a clock using ZM1032 Nixie tubes and multiplexed their anodes instead of their cathodes.  The ZM1032 uses a 9-pin base, with five cathode pins, two anode pins, and a screen pin.  See https://www.tube-tester.com/sites/nixie/dat_arch/ZM1032_philips_02.pdf  The odd digits use a front anode, the even digits use a back anode.  There is a screen in between the even and odd cathodes, tied to 60v.  The screen blocks some of the light from the even cathodes, so I increased the "on" time of the even anode.  It's just about double the "on" time of the odd anode.
A similar circuit is describe at https://www.dos4ever.com/nixie1/nixie1.html#zm1030 however they used different current-limiting resistors for the even and odd anodes where I used the same value but used different "on" times for even and odd anodes.
Reply all
Reply to author
Forward
0 new messages