[sdiy] how to fix a blinking/blank neon power indicator?

[anthony]

Anybody know why sometimes just out of the blue, a neon bulb will stop 
working or begin to blink intermittantly?

My guess is that there wasn't an adequate current-limiting resistor and the 
gap betwwen the electrodes widened. This would seem to explain the blinking 
ones too: the gap hasn't widened enough to keep it from firing, but it has 
to wait for enough charge to build to "jump" across.

I've also noticed that heavy inductive loads, say a variac controlling a 
1000 watt heating element, can cause the neon bulb to go out if the setup is 
(unwisely) connected to a power strip that has a neon power indicator (the 
one I was using didn't have a surge protector so it wasn't anything to do 
with that...). But I don't know what this is about unless it somehow causes 
a voltage drop locally across the system. But it couldn't have to the 
electrical system to the whole building in a modern university.

So I had an idea to "fix" a failing bulb: connecting something like a 
polypropylene X2 rated capacitor across the neon bulb leads. I haven't tried 
this yet: I was wondering if it'd work. I have a HUGE collection of a wide 
range of MKP X2 caps taken from various power supplies.

So does anyone think this'd work?  If so what would be a good capacitance 
rating? Is there a better fix? I'd just replace the bulb, but it is a row of 
switches with the neon bulbs right inside them and these is a number of them 
that do not work. Would something like an oil-filled motor-start cap work 
better? I have a 3.5uF one rated for 450VAC and a 0.5uF one rated for 
2500VAC (probably too small: might as well use a MPK X2...). Or a 10uF, 
160VAC metallized paper motor-start cap maybe?

And gosh what you do with one of those organs that's chock FULL of neon 
bulbs?

aa