One of my NOS F90T17D apparently ran out of gas and quit.
It had very few hours on it, less than 50.
It started acting different than the other bulb in the fixture on startup and one end of the bulb had a odd rusty area show up inside the bulb about four inches from one end.

It looked like rust was spraying on the inside of the bulb.
It was not black, but rust color.
It ran for a while like this and then one day it would not start at all.
It would try to start but would not light up.
Dead bulb.

I have seen other LG members posts about some fluorescent bulbs not starting when they loose their gas fill.
I hope this is not a trend as the F90T17 bulbs are almost impossible to find now.
But I do wonder what the rust like deposit inside the bulb was?

I saved the bulb for display only to show others what these huge bulbs look like and if it should break no loss since it is already broken.

Wow, that's a real shame.  Probably either a defective gas fill, or a very slow leak. It will probably still light on an instant start ballast.

just had a nos ge f15t8/n from the 70's do this.
another had gone to air.
these were stored poorly and i suspect corrosion creeped down the wires into the seal.
many others from this lot were installed and are still ok.
i have seen higher dud rates in new production lamps so i still am not going to worry about this.

From 1971 to 1974 I had a maintenance job that included replacing ballasts and bulbs in a large office building.
The replacement bulbs were GE and it was not unusual to find one or two dead bulbs in a case of bulbs.
They just would not light up.
But using about six cases of bulbs a month I did not think much about it, just grab another bulb and continue on.
Since then I learned that they test each bulb before being packed, so it should have worked before it was out of the factory.
Wire sealing may be on a assembly line may be more difficult than I thought.


 

I have had several to do just as you had happen.  Mine are vintage F90T17/D also.  Two just totally lost vacuum, similar to when you "fry" a EOL tube, but these were working fine then "poof" they lost vacuum. I had one to die in exactly the process you described.  It seems the "sealing" where the wires pass through the glass was less than ideal on some of the tubes.

I have only a few more and so far the replacement F90T17D is working ok.
As you say John, several other LG members have had the same problem with their T17s.
If they were tested ok at manufacture and held up all those years until use it must have been a slow leak.

May be the heat of the bulb in use accelerates the rusting process of the lead wires.
In my case it looked like rust spraying inside the bulb.
But only on one end, as the other end looks like new.

May be rust formed on the lead wires and created a small leak between the wire and the glass.
The lead wire is special wire made for bulbs only, but maybe the wire is prone to rust in high humidity conditions.

I was thinking that a shot of this product below through the small hole in the end caps would stop any rust in the lead wires.
This product is said to stop rust and could be sprayed inside the end cap through the little hole or drill it out some for the plastic sprayer tube.

http://www.cjponyparts.com/eastwood-internal-frame-coating-14oz-aerosol-can-with-spray-nozzle/p/EW12515Z/&utm_source=google&utm_medium=merchant&utm_campaign=shopping?gclid=CIHwmqmhp8ACFYqEfgodMwoA8Q

If I have anymore loss of gas problems this may be the answer.

This is sad news. I always thought an unused fluorescent tube's life was pretty much unlimited (barring accidents, of course).

This is sad news. I always thought an unused fluorescent tube's life was pretty much unlimited (barring accidents, of course).

Even more strange is I just had two NOS GE F96 WM PGs go dead after about 25 min of operation.
They ran great out of the box for 25 min so I shut the fixture down and hung up the fixture the next day and installed the bulbs.
Nothing happened when I turned on the fixture.
Both bulbs dead.

Installed two more NOS Watt Miser F96PGs that I had and so far they both work just fine.
But testing the two defective bulbs again one at a time in the working fixture the result was still the same.
Nothing.
I do have the correct new PG ballast that works just fine with the other two bulbs.
 
How both NOS PG bulbs went dead at the very same time I don't know.
The filaments all test ok, but the bulbs just will not light at all.
Really odd.

That is really odd, Larry. Does anybody have an explanation?

How old were the WM lamps that just went bad on you, Larry? I wonder if quality control went downhill during the final years of production, just because there was such small demand?

Here is one of the NOS bulbs.
I don't know how to read the date code.
May be made after 2005.

http://www.lighting-gallery.net/gallery/displayimage.php?album=3388&pos=3&pid=97936

So far only two bulbs out of eight have failed.
But the really odd part is that that these two new PG bulbs failed at the exact same time.
They both worked fine for about 25 min then the next day both failed at the same time.

I just checked the bulbs I replaced the dead ones with this morning and they still work just fine.
So it is the bulbs that don't work, not the fixture or ballast.

They display the Hg logo, so 2003 at the earliest.

Ugh aur contamination run them on 4 x 32 US ballast
There a chance this while be revived

  I have F90T17 like yours but broken cathode
  Now i have no spares F90T17

   As i used my Sylvania White T17

See if the dead lamps will run on an IS ballast.

It looked like rust was spraying on the inside of the bulb.
It was not black, but rust color.
It ran for a while like this and then one day it would not start at all.
It would try to start but would not light up.
Dead bulb.

I have seen other LG members posts about some fluorescent bulbs not starting when they loose their gas fill.
I hope this is not a trend as the F90T17 bulbs are almost impossible to find now.

Air got in and when you heat up the electrodes, they react with oxygen and burn up just like powering up an incandescent lamp with the bulb broken. The brown thing is most likely whatever that on the filaments that oxidized. Things expand/shrink with temperature and less than ideal seam between one of the leads and glass opened up.