🚀 go-pugleaf

In article <1993Jan3.145525.28378@osuunx.ucc.okstate.edu>,
martin@datacomm.ucc.okstate.edu (Martin McCormick) writes:
|>
|>     We recently had a fluorescent lamp ballast go bad at our
|>house which got me wondering about what actually happened.
|>The lamp in question is an instant start type, using a
|>12T/20CW tube.  This tube has two pins at each end and can
|>operate in either the instant start-type fixtures or the old
|>style ones that use the starter.
|>     The trouble started innocently enough.  The lamp didn't
|>light when switched on.  One could see a flash in the tube
|>when turning it off, but there was no other light at all.  I

sounds like the ballast failure at this point..

|>had a spare tube which had, at one time, been in an old
|>starter-type fixture so I decided to use that one until we
|>could go out and buy a new one.
|>     The light came on with the installation of the spare
|>tube and we went on about other things.  A few minutes later,
|>my wife said that she had seen a flash.  We weren't really
|>sure, though, so we didn't take any action.

dying ballast took out the old tube -- new one works OK for a
bit but the ballast is still weak..

|>     About five minutes later, it was obvious that something
|>was terribly wrong.  The ballast was sizzling and filling the
|>house with the aroma of melted pitch and burning paper.  We
|>are, now in the market for a new fixture.
|>     Just for fun, I did an autopsy on the original
|>fluorescent tube and the spare.  The original one shows that
|>both cathodes are open, (infinite resistance at each end.)
|>The spare tube that had bin in the ballast when it smoked
|>shows one open cathode and one which is still good, (about
|>4.5 Ohms.)
|>I am not really sure how the instant start ballasts work, so
|>I am wondering if the presence of one open cathode might have
|>actually been what smoked the ballast since it hadn't smoked
|>until I installed the spare tube.  One open cathode would
|>cause the load to be half-wave rectified DC instead of the
|>full-wave load which normally occurs.
|>Since the ballast got severely overheated, it must be
|>replaced, but I would like to know what really happened.
|>Maybe saving the old tube turned out to be false economy.

Probably the ballast quit partway first, then died the final death
after the spare tube was put in.  FYI, the instant-start ballast for
a two-pin (filament type) tube goes like this:

           120VAC HOT O------() || ()-------O
                             () || ()        > 2-3VAC CATHODE  (high current)
                             () || ()-------O
                             () || ()        \
                             () || ()         > 1.9kVAC GAS DISCH (low current)
                             () || ()        /
                             () || ()-------O
                             () || ()        > 2-3VAC CATHODE  (high current)
           120VAC NEU O------() || ()-------O

The cathode bridles are high-current windings added to the high-voltage
excitation winding -- probably the unbalanced high-current loading caused
the core to overheat and burn out the varnish, which caused primary and
secondary winding shorts, etc., especially if the core is grounded like it's
supposed to be.

The most economical course is to start over with a new lamp, tubes, ballast,
and all, from the hardware store.  Other than that you could put in new tubes
along with your new ballast, but be sure to replace them _together_ if a new
ballast is going in.  They're interrelated in the extreme..

<BGB>
lihan@ccwf.cc.utexas.edu        /   The only reason the world hasn't
(really Bruce Bostwick)        /   dissolved into total chaos is that
from the great state of TEXAS /   Murphy's Law also applies to Murphy.