On Wed, 8 Nov 2017 09:04:42 -0800, "David Farber"
wrote:

https://www.nachi.org/forum/attachme...-schematic.jpg
This is the correct diagram except the resistors are not 47k, they are 68k.

The exact value isn't important. There's some effect on the
brightness, but not much.

On page 81 of the GE Glow Lamp Manual, there is an interesting test circuit.
It shows the neon light in a box with another light source. I think it
explains why my ~20 year old battery powered fluorescent book lamp doesn't
work very well at night when I use two rechargeable AA batteries. The lower
voltage isn't able to turn on the lamp when there is no ambient light. I
have to shine a flashlight on the lamp to get it to come on. When the
batteries get a bit drained, it can take almost a minute for the light to
come on with the flashlight trick. Once it's on, it provides very nice, even
lighting.

Yeah, that makes sense. The test setup shows that the light source is
0 to 500 ft-candles which converts to about 5400 lux. That's rather
bright, especially when measured 1 ft away. (I just tried it with my
LED flashlight and lux meter).

A variation on the same test is to add something radioactive inside
the neon lamp. The idea is for the neon to ionize at a lower voltage,
thus making a better voltage regulator, memory device, or whatever.
Radioactivity is also used to help fluorescent tubes start:
http://www.irpa.net/irpa10/cdrom/00740.pdf
Half-Life Emits
Kypton-85 10.8 years Beta
Promethium-147 2.6 Beta
Tritium H3 12.3 Beta
Thorium-232 14 billion years Alpha

My guess(tm) is that your 20 year old fluorescent tube had some kind
of starting help from some radioactive material, which is long gone by
now.

--
Jeff Liebermann
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558