IN-12B decimal point

[newxito]

I was told several times to use IN-12B instead of IN-12A with neon bulb as decimal point. I finally tried that but I don't like it. Is anyone using the decimal point of an IN-12B in a project? When combined with a 3 or 8, the point is barely noticeable. Another problem I'm having is that it glows much brighter when no digit is lit. As soon as a digit is lit, it gets dimmer. Is this normal or I’m doing something wrong? I use a 15K anode resistor and an 82K cathode resistor for the decimal point with a voltage of 170V.

[Peter Hall]

From🚨Peter Hall😁 insidiousnixies🦘Utube

On 12 Sept 2025 4:45 pm, Peter Hall <hall...@hotmail.com> wrote:

One off me clocks with dot, being used and same brightness..

From🚨Peter Hall😁 insidiousnixies🦘Utube

On 12 Sept 2025 4:08 pm, newxito <axt...@gmail.com> wrote:

I was told several times to use IN-12B instead of IN-12A with neon bulb as decimal point. I finally tried that but I don't like it. Is anyone using the decimal point of an IN-12B in a project? When combined with a 3 or 8, the point is barely noticeable. Another problem I'm having is that it glows much brighter when no digit is lit. As soon as a digit is lit, it gets dimmer. Is this normal or I’m doing something wrong? I use a 15K anode resistor and an 82K cathode resistor for the decimal point with a voltage of 170V.

--
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, visit https://groups.google.com/d/msgid/neonixie-l/613d7d66-dcab-48ae-bd34-a5f714071e72n%40googlegroups.com.

[Michail Wilson]

Something is wrong.

 

My project (Bitcoin Ticker) uses 8x IN-12B and they (hundreds of tubes worth) don’t have that issue).  I used the 12B since it has a decimal point on both sides.

Have you tried another tube?  What’s your project schematic?

 

M1

[newxito]

I have the problem with all my IN-12B...

 

Schematic of the display: https://github.com/axtar/Nixie-Calculator/blob/main/Schematics/Schematic_Nixie_Calculator_IN-12_Display_Rev2.pdf

If using the B version I populate the 82K resistors. If using the A version, I populate the neons and the 270K resistors.

Schematic of the driver: https://github.com/axtar/Nixie-Calculator/blob/main/Schematics/Schematic_Nixie_Calculator_IN-12_Driver_Rev2.pdf

Just the usual HV5622 thing...

What resistor values are you using for your Bitcoin Ticker?

[newxito]

I just tested this with an IN-16 and a B-5870. They behave exactly the same. When only the decimal point is lit, it glows brighter than when lit together with a digit. This is probably just how a nixie tube works. I never noticed it because I always used only the decimal point on the right. So when entering a decimal point, a digit is always already on. The IN-12B, however, only has a decimal point on the left side.

And in terms of visibility, I like the decimal points of the IN-16 and B-5870 much better. After a calculation, you see the decimal point immediately, whereas with the IN-12B you have to search for it. I think I'll switch back to the neon bulbs.

[Jarod Findley]

Its looks like your voltage or amperage is high try cutting it back a few volts or a miliamp 

To view this discussion, visit https://groups.google.com/d/msgid/neonixie-l/c9cbf447-021e-4b3f-bf5a-b1871d3c94e6n%40googlegroups.com.

[newxito]

I took a few current measurements, all with 170V, 15K anode resistor and 82K cathode resistor for the decimal point.

The first measurement was just to check if the anode resistor is correct. With the digit 3 turned on, the current was 2.13 mA. Should be 2.5 mA, but I think 2.13 is still ok.

Then I measured the current through the decimal point cathode with only the decimal point turned on: 0.401 mA. This is kind of ok, should be 0.5 mA

Then the same but with the digit 3 also turned on. The current dropped from 0.401 to 0.143 mA.

[newxito]

Is there a better solution than removing the anode resistor and using separate resistors for all cathodes? 

[Dekatron42]

Have you checked circuit diagrams for clocks or calculators how they have designed the circuit around the decimal point?

Burroughs used 180k for the decimal point with their B5750 Nixies so twice what you have tested.

/Martin

[Sture Nystrom]

Try and use a cathode resistor for each cathode and one resistor for the decimal point. 

You will not have any anode resistor, it’s swapped to the cathodes instead. 

To view this discussion, visit https://groups.google.com/d/msgid/neonixie-l/99a6b8db-7e7d-491c-b7e0-2df8ffc5eba9n%40googlegroups.com.

[newxito]

Thanks Sture, you probably missed my question a few posts above. I've tested it and the method works fine.
But that's a major change for a board with 16 nixies, and I'm trying to avoid that.I was aware that the anode resistor causes the current to drop for the decimal point when a digit is turned on, but I didn't expect the big difference in brightness.
As Martin suggested, I'll try using larger resistor values ​​for the decimal point so the brightness difference might be less noticeable.

Andy

[newxito]

I will leave the IN-12B option on the board and in the firmware but I'm going back to neon bulbs so there is no way to overlook the decimal point...

[Jon]

It's a current-stealing effect. If you search the old archives of the group you'll find that Pete Hand identified it and also proposed the use of a single transistor constant current sink driver as an alternative to the decimal point cathode resistor. I encountered the problem when developing my Niximatrix clock that uses IN-12B, and experimented with the current sink driver per his recommendation - it works a treat.

Jon.

[newxito]

Thanks Jon, I will give it a try on the breadboard.