Digital potentiometer for high voltage?

Thread Starter


Joined Oct 30, 2019

I am hoping to limit the current into a Nixie tube tube via this device.. (LR8)

View attachment 191535

I am really hoping I can make it digitally adjustable too. My project goal is to be able to cycle through the numbers and change the brightness via a smartphone/bluetooth. So my nixie tube is an IN-17 which has a current rating of 1.5mA to 2.5mA. With this device, that means R needs to be between 480 and 800. The first thing that comes to my mind is a 1k digital potentiometer, since at 0.0025A there would only be 2.5 volts at those pins (where the maximum is ~5.5v). Unfortunately, all these pots are 20% tolerance and not ideal for this situation. The 1% tolerance digital pots seem to go no lower than 20k either.
I could perhaps use the 1026 step, 1% 20k d-pot from steps 13-29 to keep the pot in the 380-800 realm. But then I need to ensure that the pot is in "performance mode" and it just seems like it's becoming way overly complicated to do this.

Is there a better solution? This is meant to be a battery powered application so I primarily wanted to minimize wasted energy over the typical anode resistors. I am super open to suggestions, such as perhaps a high-side BJT setup? I keep reading how those can be unreliable though.. I could toss the brightness adjust-ability idea, but I am hoping not to lose sight of my design goals.

Thank you for any input you have on this!! I'm at a loss at what to try next.


Joined Jan 8, 2017
You could use a number of resistors selected by PNP transistors or P channel mosfets feeding the anodes. For example with 4 transistors selecting 4 resistors you could obtain 16 different brightness levels. You would also need a fifth resistor that was setting the lowest brightness. (Unless you wanted the lowest level to be off.) This would be like a D to A converter.


Thread Starter


Joined Oct 30, 2019
Yes! I can't seem to edit my original post though. It showed up just fine when I wrote the post, I'm not sure what happened. My apologies.

This seems like a perfect component for the job and I am stoked by that. It is unfortunately not a jelly bean component though and if sources run out, the design is obsolete.

R would be hard to adjust digitally too, as far as I'm aware.

An alternative is something like this,

(taken from

This uses PWM of course, but I think it is debated whether or not PWM is bad for nixie tubes? I could be wrong with that.


Joined Aug 21, 2008
You could consider this configuration. Note the capacitor C2 between the op amp output and its inverting input. Apparently the stage made of Q1 and Q2 had too much gain and significant phase shift that may have caused oscillation. With different parts you might need to adjust C2. I would probably try an emitter resistor on Q1 first to lower the gain of that stage.


Joined Sep 17, 2013
I think it is debated whether or not PWM is bad for nixie tubes?
I haven't researched nixie properties, so can't say how they'd be affected, but I can't think of any obvious reason why a low PWM frequency of, say, ~100Hz would be harmful. After all, nixies have been used for ages in instrument displays with rapidly changing digits.

Thread Starter


Joined Oct 30, 2019
That's true, it's probably fine. I thought that maybe if the gas is "ignited" constantly because of PWM, then that would cause a reduced life. It's possible that this makes no difference and perhaps the gas doesn't instantly de-ionize/sputter or whatever and it doesn't actually get re-ignited during each period of the PWM.

Anyway.. Hard to say. There's no apparent reports of PWM destroying tubes so I'm sure it must not be a killer at least.

That's interesting! I'll look into that versus PWM for the sake of power losses. There are 2 primary goals with my current limiting in this project, to minimize power loss and make it digitally adjustable. Minimize power loss definitely comes first. Thank you for posting this schematic, I'll try playing around with it in ltSpice


Joined Mar 14, 2008
I thought that maybe if the gas is "ignited" constantly because of PWM, then that would cause a reduced life.
If that's a concern, you could perhaps add a parallel resistor to give a minimum current that keeps the tube slightly on when the PWM signal is low, avoiding the need to be constantly "ignite" the tube.

But that may not be a problem.
Before LEDs, small neon bulbs were often powered from line AC as indicator lamps, which turns them on and off at twice the line frequency, and they have a long life in that application.