What voltage/current to run tubes at

921 views
Skip to first unread message

kitehman

unread,
Feb 7, 2017, 1:22:36 PM2/7/17
to neonixie-l
finally got a power supply and am wondering what to run my IN-1 tubes at

IIRC your supposed to use current to change the brightness right?

i set the supply to 140v and am using 10K or 20K ohm resisters on the cathode and get 1.01-1.24ma at 10Kohm and 0.65-.083ma at 20Kohm

what the target? lowest voltage or lowest current?

SWISSNIXIE - Jonathan F.

unread,
Feb 7, 2017, 4:14:32 PM2/7/17
to neonixie-l
If you have pwm available, i would use the duty cycle for adjusting the brightness.

Roddy Scott

unread,
Feb 7, 2017, 4:42:22 PM2/7/17
to neonixie-l
Dieter's http://www.tube-tester.com/sites/nixie/nixie-tubes.htm has data sheets on all common tubes available.

The data sheets on this tube recommend 2.5mA. Tubes should be operated within the values supplied otherwise you will have either low levels of brightness or if exceeded, short tube life. The typical strike voltage is 170v and the maintaining voltage typically 133v at that current value from the data sheet.

kitehman

unread,
Feb 7, 2017, 10:58:57 PM2/7/17
to neonixie-l
with the adjustment turned all the way down (137v it strikes just  fine, and maintains at 121v) and takes 0.67ma is there anything wrong with running at lower voltage levels other than brightness?

threeneurons

unread,
Feb 7, 2017, 11:21:30 PM2/7/17
to neonixie-l

See what voltage you need for the tube to strike in a dark room.

If you run the tubes at too low current, you risk cathode poisoning. It won't appear immediately, but takes time. Weeks or months. What you'll eventually see is only partial illumination of the digits. Positive neon ions hit the lit cathode, sneding that cathode's metal flying. That metal covers its unlit neighbors. When those unlit cathodes have their turn at being ON, they can shake of that crud, but only if the current is high enough. If too low, that crud builds up, causing dark unlit splotches on the cathodes.

So, if the current is too high, you just ear out the cathodes too quickly, but also if the current is too low, you won't allow them to clean themselves. Operate in the recommended current range for that tube model.

Other tube model thoughts. If the nixies are to be ON 24/7 such as in a clock, then only use long life tubes, containing mercury, for those. Of the Russian devices, those are: IN-8-2, IN-12 (A & B), IN-14, IN-16, IN-17, & IN-18

Save the non-mercury tubes (IN-1, IN-2, & IN-4) for items that are on transiently, like meters and games. 

kitehman

unread,
Feb 8, 2017, 12:50:59 PM2/8/17
to neonixie-l

so i should shoot for just under the 2.5ma current rating and go for lowest striking voltage? higher voltage for a brighter display?

threeneurons

unread,
Feb 8, 2017, 2:05:24 PM2/8/17
to neonixie-l
On Wednesday, February 8, 2017 at 9:50:59 AM UTC-8, kitehman wrote:

so i should shoot for just under the 2.5ma current rating and go for lowest striking voltage? higher voltage for a brighter display?

Yes, I would aim at 2.5mA. Exact value +/- not critical. You will find, that once they strike, the maintaining voltage, is relative constant, as you change the supply voltage. They are gas filled cold cathode tubes, and just like neon bulbs, and tubes like the 0A3 and 0B2, they act as voltage regulators. Very similar to a zener diode. So if you increase the supply voltage, the neon voltage will increase ever so slightly, while most of the voltage change will be across the anode limiting resistor. Since the current is the voltage across that resistor divided by the voltage, then the current will increase, and so will the brightness.

Its been suggested earlier, that you should look at pulse width modulation (pwm). to adjust brightness. That is use a fixed value for supply voltage and anode resistors, to get roughly 2.5mA per tube. But add a pass transistor to your supply so you can turn it ON and OFF at a rate faster than your eye can perceive flicker. For most people this is faster than ~30Hz. What exact frequency is not important. On the low end, visual flicker is your boundary. At the high end, its the turn ON time of nixies. It takes time for a nixie to ionize. I ran experiments on speed a tad over 10 years ago. The fastest I could get a nixie to turn ON, was a ZM1000, with its primer electrode always ON, and a 200V supply. I could get it to turn ON in 7uS. 20 to 30 is typical for most nixie tubes. 50uS nominal worse case.


So, in practical terms, I would set the minimum ON time for about 200uS. So if you pick a PWM frequency of 100Hz, you'll get max brightness at 10mS (0.01sec, 10000uS), and minimum at say 200uS, for a 50:1 brightness range. Your eyes don't see linearly, though, so as you view it, that 50:1, will look more like 5.6:1 (almost 6:1). 
Reply all
Reply to author
Forward
0 new messages