So, i was experimenting with one of the 3 EOLed GE Indonesian lamps (this was one i was trying to fry for a long time) and today i think i killed it but i'm not 100% sure if i did. The reason i think it might have survived because 1, both cathodes in the lamp are still in good shape even though one of the ends was glowing a bright red when the lamp was being killed. 2, There are no signs of cracks on the glass (i know you don't see this on all fried lamps but considering the cathodes did not melt i would expect to see at least a small hole). And 3, when i hooked the lamp up to a preheat ballast, when "starting", both cathodes would glow but they would not melt (like they would if air got into the tube) but when the starter opens, there would be no discharge in the lamp (no flashing or anything other than the cathodes glowing). One last thing about it is i can't get it to fire up at all on any ballast.

Has anyone else seen lamps do the same thing, as this is the first time i've ever seen this happen to a lamp. And does anyone know how it happens?

Are you able to try the cathode resistance at hot (e.g. measure the current when supplied from a 6V battery) and compare it with a good lamp?
If it is significantly lower (i.e. the current higher), it means some seal broke and the air rushed in and that is preventing the discharge. This could happen even without the filaments breaking...

If there is air inside the lamp, so the filament would burn out after few seconds when heated. This is what happened to my past CFLs when they fell out to the floor and become broken.

I could not get the resistance of the cathodes when hod but the resistance of the cathodes when cold was definitely lower, but when comparing it to
 the "failed" cathode from one of the other 3 EOLed GE lamps it also has the same resistance and these lamps still flicker when connected to a preheat ballast.

I've had other broken lamps that the cathodes would quickly fail when given power.



edit: Thinking about it now, i think i had a cracked PL lamp which took around 30 or 40 seconds for the cathodes to fail.