The f15t8 preheat aquarium light on my fish tank just started flashing on and off randomly. I lifted the cover up, and the lamp's ends would glow red hot, the lamp would strike, and then go out, and the ends would glow again. It seems very similar to a cycling HID lamp. The starter glows when the light is first switched on, but other than that, it doesn't glow. Is this a normal EOL for fluorescents?

Are the ends totally blackened? If so, they're emitter depleted and won't strike as a result.

When an argon T8, short T5 and T12, reaches EOL on preheat circuit, the rectification causes sufficient voltage to flow the glow bottle of the starter, causing it to interrupt current to the lamp to preheat the electrodes and than restrike it in a random periods. This is the normal EOL behavior of a preheat ballast + glow starter. I think since most of your ballasts in the US, are self-starting (Rapid-start, Slimline and trigger start), you never seen an EOL behavior of the more common setup in Europe (Series choke + starter), which the american preheat is the same, with the only difference is that the ballast is autotransformer.

Are the ends totally blackened? If so, they're emitter depleted and won't strike as a result.

The ends have some blackening but not much.

Quote from Dor123 on 9:29 AM
When an argon T8, short T5 and T12, reaches EOL on preheat circuit, the rectification causes sufficient voltage to flow the glow bottle of the starter, causing it to interrupt current to the lamp to preheat the electrodes and than restrike it in a random periods. This is the normal EOL behavior of a preheat ballast + glow starter. I think since most of your ballasts in the US, are self-starting (Rapid-start, Slimline and trigger start), you never seen an EOL behavior of the more common setup in Europe (Series choke + starter), which the american preheat is the same, with the only difference is that the ballast is autotransformer.

Makes more sense. Still was weird how it was similar to an HID lamp cycling.

This is how an EOL fluorescent lamp behaves on a series choke + glow starter: http://www.lighting-gallery.net/gallery/displayimage.php?pos=-110634
EOL F36T8 behaves slightly different on preheat, because of the higher starting voltage of their krypton buffer: http://www.lighting-gallery.net/gallery/displayimage.php?pos=-106469

I saw somewhere where a person here had a vintage fluorescent that had an air leak, and it would cycle like a hid lamp, but not quite like the way you describe, Zarlog.

This is how an EOL fluorescent lamp behaves on a series choke + glow starter: http://www.lighting-gallery.net/gallery/displayimage.php?pos=-110634

That is very similar to what mine is doing, but the delays are slightly longer than they are in that video.

Yes this lamp ran off of emitter when that happens the cathode glows red hot. Also it good way get those starter stuck.

Yes this lamp ran off of emitter when that happens the cathode glows red hot. Also it good way get those starter stuck.

I replaced the lamp as soon as it cooled off (it was also unusually hot, much hotter than it usually is) with another aquarium lamp. The old lamp was 4000k and had a nice color, but the new one is 18000k and makes the fish look weird, so hopefully I can find another in 4000k.

Hot lamp means there could have been an air leak, poisoning the lamp inner atmosphere. It causes the arc voltage to increase, hence the higher losses (so temperature) and then starter cycling.
How long it has been cycling? If for mote than few hours, I think the starter will be quite worn out, so I would replace it too (degrading starter tend to kill lamps)...

Hot lamp means there could have been an air leak, poisoning the lamp inner atmosphere. It causes the arc voltage to increase, hence the higher losses (so temperature) and then starter cycling.
How long it has been cycling? If for mote than few hours, I think the starter will be quite worn out, so I would replace it too (degrading starter tend to kill lamps)...

It was doing this for about 45 minutes at the most, because I didn't notice until I came in the room. The starter looks ancient too, so I plan on replacing it soon. As for the air leak, that seems to be the most likely cause, since it was so hot it hurt to touch it.