I agree in going to what I was told for DHL lamp bars - replace them all, the lamps for an instant while failing and going supernova - no resistance in
current for a micro second, will
send all the
current to the other lamps in now say in this case 1/4 more
voltage to it and all next. How long can a 120v high output lamp say a
HPL survive with its more robust
filament than the low
voltage lamp in magnitude last given say a 150 Volts in a
spike in comparison to
voltage spike? Yes it will probably survive the
voltage spike, but it will certainly shorten the lamp life.
On a dual ended moving light lamp, a perfectly good lamp installed into a bad lamp
socket will last about 100 hours less each time installed until it fails to
strike. I have seen and tracked this by way of idiot tech people installing perfectly good lamps into bad bases. Bad ignitors in a older
fixture however is harder to figure out lamp hours for and identify - but mostly can. If you look at the tip of the electrode, look for blobbing instead of bubbles and a normal structucture to the tip of the electrode. Blackening of the electrode is also an indidication as with a lack of bubbles about the e lectrode but also not always. A lot of lamps inspecteceted and on a normal moving light lamp, I can guess to within 100 hours of how many hours it has been in use. This even if with a defective
ballast.
That's on a moving light lamp. Never really studied a
ACL PAR 64 lamp failure, but I would expect them blobbing in failed
filament in like a
fuse from overload in also being blobbed instead of blackened from a failure. Believe I remember that description correct and as similar to lamps.
If in series the #4596 lamp has been a very reliable product. Where I work standard is not for them as opposed to the DHL light curtain lamps/gear (probably in dumpster fill by now with only a few saved at best), you do still replace just the 28v/250v individual lamp per policy. 28v x 4 at 112v is still less than even a 118v normal expected
voltage to
fixture at the
plug more at least 1/4 more
voltage for an instant of time. A
spike of
voltage is still a
spike of
voltage especially if to a not pre-warmed
filament.
(Might be another question to ask in if your dimmers are pre-warming the
dimmer circuits so as not to
shock the filaments.)
Overall I agree in replacing all lamps but policy where I work is one for one direct replacement only - and it has not proven to be a problem sufficient to change that policy with further lamps on the bar reflecting a problem with one wired in series of this type of lamp still in use.
This granted I have people every 100 hours and at times with more serieous problems and throwing into them at least 4x the price in lamps into bad bases or ignitors, week to week having to fix the same
fixture and not guessing there was other than a plainly bad lamp problem caused by What? Good you are questioning the
ACL bar, you probably can save future companies in asking about why any lamp you change failed and further concepts of that.
Just wizzles my mind in how tech people at times service the same
fixture time after time and do not get tired of replacing the lamp in it - this same
fixture, but don't have any wonder about why its failing so as to perhaps figure out whats' wrong with it or at least replace the
fixture in not further wasting time on it. Suppose I'm just not of the same ilk of tech person in wanting more work and normally also not resetting the lamp counter. On later upgraded fixtures, them that forgot to reset the lamp counter might attribute a perfectly good lamp in change to "hours"... I have you on a data
base in having changed say last week this lamp, how can it acquire more hours in a week than "over hours" the next lamp changed in tracked is you report? But I digress.
The
ACL's are not used as much, and I'm sure they had the same problems above in why they should, but overall not as much a problem in just changing one lamp in series. Perhaps put a used or new lamp into the second
fixture might be a considered option at least for debate. But the math is that all will soon be going.