On Jul 19, 2:14 am, "Smarty" wrote:
Just for the record, I have to report that fluorescent fixtures do not
***ALWAYS*** perform the same as incandescents even if they are wired the
same:

I purchased 24 fluorescent strip lights from Home Depot, each containing an
electronic ballast and sockets for 2 fluorescent T-8/T-12 style 32 watt
bulbs. These fixtures use the newest electronic ballasts, are extremely
efficient, make no hum or other noises like magnetic ballasts, and were on
sale for $6.99 per fixture.

When they were all installed as three sets of 8 fixtures, each set on its'
own circuit/switch, I found that an extremely high rate of failures
occurred, arising from what I subsequently learned is called "ballast
fratricide", a process wherein the switching transients from the ballasts
all being simultaneously switched results in spikes which the ballasts are
unable to dissipate without damage.

It turns out that strip lights especially are not filtered adequately,
neither in terms of the emitted spikes which damage other nearby/connected
loads, nor in terms of rejecting fast transient spikes which arise elsewhere
and need to be dissipated.

I spent some time with an oscilloscope and a lot of measurements before
being able to find where the problems were, and have subsequently learned
that others have reported the same type of problem with the newest
electronic ballasts in some commercial installations.

If one takes an in-depth look at the schematic and waveforms produced by the
electronic ballast during the start-up cycle, and also looks at the chipset
spec sheets from the ballast ICs involved, it becomes apparent that huge
switching transients are typical, and that external filtering is up to the
fixture designer who incorporates these ICs into their ballast.

The bottom line is that some fluorescent fixtures cannot be simply wired as
if they are incandescents, since they will destroy their neighbors and
themselves, in my case over only a few days of normal use.

My solution was ultimately to replace all 24 ballasts with another, better
filtered design which, unlike the Home Deport brand, have a lot of spike
filtering as well as a 5 year warranty.

Here is another example of this type of problem:

http://lighting.copperwire.org/7.2.php

Smarty

"J.A. Michel" wrote in message

...

Thanks for your posting. Had never heard of this potential for high
incidence of failure with electronic fluorescent ballasts.
Within last year acquired about 7 used (four tube ) fixtures being
taken out of a school being renovated. If I had not taken them the job
foreman said they were on their way to the dump!
I also obtained seven of the T10 type tubes (the thinner one inch
ones) and bought another twenty. Five of the seven are now in service
in our basement workshop; have had one tube failure but no ballast
problems; so far. However also have a bunch of conventional ballasts
and conventional tubes and have often 'rebuilt' older fixtures. So, if
necessary can convert these electronic ballast ones to conventional.
This is in Canada, 115 volt 60 hertz etc. Total load on our string is
5 fixtures, four 34 watt tubes each (total 20 tubes five ballasts) for
approximate wattage (inductive?) of 680 watts.
Maybe the fact that our fluorescents were well used and were working
when removed eliminated any prone to 'early failures'?
However several years ago our local junior school (approx 500 pupils)
did a major conversion (to electronic ballasts and modified many of
the fixtures with better reflectors) and we have not heard of any
problem with frequent failures?
Thanks for the comments.