rose colored fluorescent lights

Sounds like the discharge gas has run low on mercury and you are seeing the argon discharging instead. There could be a number of reasons:

1. The tubes were manufacturered with too little mercury.
2. They are low mercury content tubes which have reached the end of their life. (Mercury is slowly lost as it gets absorbed into the phosphor and other parts inside the tube. A low mercury content tube might actually run out of mercury for the discharge, before failure of the electrode emission coating, which is the most common failure mode of a tube.)
3. There is a DC component to the current, which has forced the mercury ions up to one end of the tube (usually makes one end more pink than the other, at least for a while).
4. The tube (or just a small part of the tube) is cold, causing the mercury to condense out of the discharge.

It would be interesting to know how old the tubes are, how long they worked properly before going pink, and how quickly they then went pink.

I don't know the age of the tubes as it is in a office environment, but I haven't seen the maintenance guy for quite a while. There are two U shaped tubes in the fixture and I expect that they would be 4 or 5 feet in length if they were straight. One tube went pink first and 4 or five weeks later the second one went. The color is fairly uniform down the length of the tube with the exception of the moving bands that almost look like a beat frequency of some kind (perhaps a 60 Hz phonemena). The bulbs are also in a main hallway and are on 24/7. They are not a subtle color and I was curious about them. Thanks for your insight Andrew (Inquiring minds wanted to know)

Ron

I have a tube that has gone very pink.

It is a miniature tube installed under the kitchen units last Christmas in a fitting with a high frequency ballast and is used intermittently. Other tubes installed at the same time have not done the same.

Seemingly it suddenly went pink very quickly about a fortninght ago.

DG

I would go in favor of a badly designed ballast. Which might had degraded the lamp itself, though it still was a ballast problem.

Unlikely. I would go with reason #1 based on the Derek's report, although more information would have been useful. Only #3 is caused by ballast fault, and there was no mention of the pink being more at one end.

Thought someone mentioned that the lamps went pink early. Another option could be that the lamp is started often with short turn on periods. Any information is available about the electrodes?

I am not too > > > > >

The pink is about the colour of Blush Zinfandel. ;-)

And uniform along the tube.

I've hauled the tube out now, it's a completely anonymous T5 18 watt

Ring lighting's depot is only a mile or so from here so I went along to their factory shop and picked up a replacement, but it turned out to be a 15 watt and smaller so not useable, they don't stock a longer T5 . However inspection of the crimping and knurling of the tube end fittings seemed to indicate it had the same origins probably in China.

Quite a difficult tube to source locally BTW. Both RS and Farnells don't stock it. I know I could easily buy it in bulk off the internet.

DG

Are you sure you mean T5 tubes? Both the power ratings you mentioned are standard for T8 (1" diameter) tubes in Europe, but not for T5 tubes.

Certain I 've got them here clenched in my clammy claws.

DG