I got some IN-14 gas discharge tubes and I want them to live as long as possible. For that I need to dim them to the point that the digits just about to start to partially lose glow.

One way is to adjust the dropping resistor, which lowers the ignition voltage (~170V) to operation voltage (~120V). However this method is very sensitive to current (Ohm's law) and I need one digital pot for each tube.

Another way is to utilize the blanking function of the driver chip to PWM dim the tubes. However I have no idea what frequency is optimal for the tube lifespan.

I'd like to avoid audio frequency (32Hz-32KHz) and stay above "human eye frame rate" (some where 20-60Hz). High frequency is easily achievable but I wonder If this will shorten the lifespan as the tube ignites tens of thousands times per second.

Any insight on the frequency selection is much appreciated, or is it altogether a bad idea to dim the gas discharge tubes with PWM ?

  • 3
    \$\begingroup\$ Why use PWM in the first place, instead of driving them with a constant current driver set to a lower than normal current? \$\endgroup\$
    – Hearth
    Commented Oct 11, 2021 at 0:17
  • 1
    \$\begingroup\$ "I want them to live as long as possible" - in that case just keep them turned off when nobody is looking at them. \$\endgroup\$ Commented Oct 11, 2021 at 0:31
  • \$\begingroup\$ @Hearth There is no constant current driver at that voltage \$\endgroup\$
    – 7E10FC9A
    Commented Oct 22, 2021 at 9:27
  • \$\begingroup\$ @BruceAbbott I cannot make SCP-096 with electronics. However a PIR sensor is possible. \$\endgroup\$
    – 7E10FC9A
    Commented Oct 22, 2021 at 9:29
  • \$\begingroup\$ @7E10FC9A I didn't say use something someone else made. Make your own. \$\endgroup\$
    – Hearth
    Commented Oct 22, 2021 at 13:41

2 Answers 2


PWM frequency for dimming nixie tubes

You cannot use a very high frequency to PWM a Nixie tube. Since the Nixie tube can take up to 20-30us to strike, the upper bound for PWM is probably about 20-25kHz. The time to strike a neon element depends on gas mix, voltage, temperature and incident light. Most Nixies are particularly problematic (high strike voltage) at low temperature and in the dark.

This might help you:

enter image description here

Prior to striking, the light output is obviously very close to zero, so taking say 30us to strike from a PWM repetition rate of 50us (20kHz) leaves you with a very small time frame (20us) to set the tube current. If you are multiplexing your digits then you are already increasing the pulse current by x4 or x6 depending out the number of digits you have. This higher pulse current is likely to shorten your tube life considerably.

If you are driving the Nixies directly with a latched signal you can use multiple anode resistors and simply switch them in as needed to set the current for each digit. Or you can modulate the voltage supply with a fixed anode resistor to reduce the tube current.


There is often the requirement for at least 20 kHz PWM and a load which might dislike fast actuation altogether. In that case, choose a frequency as high as possible (e.g. 100 kHz to 1 MHz) and post-filter with an LC for almost constant current output.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.