GE Halogen "modified spectrum" A19s

[Lumalux]

I was at Walmart yesterday here in the Richmond, VA area and found some GE reduced wattage halogens in the 4-pack with "modified spectrum" in relatively small type on the box.  The box also describes them as being GE's best bulb.  What caught my eye was the faint pink color of the bulbs' inside coating.  All of the boxes were like that.  I bought a 4-pack of the 53 watt (75 watt replacement), good for 2,000 hours.  These bulbs are made in Mexico and appear to be well-made.  To the naked (and my own discerning) eye, the light emitted does not appear to be any different from a regular white bulb, and if the light is indeed pink it is so faint that it is undetectable.

I can't find any information about these bulbs.  What is the purpose of the pink coating?  It's much paler than the regular Soft Pink bulbs.  Is it just to make the light warmer and more pleasing?  This is the first time I've ever seen a halogen not marketed for its whiteness.

[Keyless]

I would think to make the light warmer. The new energy saver halogens are over driven halogen capsules which produce a very white light that some find harsh. I will admit the bare halogen energy saver bulbs produce a lot of glare and the light is crisp white rather than the soft yellow we are used to.


My guess GE is trying to replicate a regular bulb in the color department. The label modified spectrum I think comes from the fact the coating reduces the lumens per what required by law, so the label somewhat lets the lamp slide as a specialty bulb. 

[Mike]

I have sensitive eyes (i sometimes find myself squinting on cloudy days because it's too bright outside) so i don't like clear lamps too much. I love the whiter look to halogens though and use the 53W soft white A19 halogens to replace 60W incandescents. More light, energy is saved, and the light is whiter and more pleasant.

I suppose i wouldn't mind using clear lamp in fixtures where the light is behind a diffuser but i just buy coated lamps so that i can use them in any fixture.

[lights*plus]

Usually "lower watt" tungsten lamps have krypton as the inert gas instead of argon. The higher molecular weight allows the filament to be operated at a higher temperature (by making the tungsten coils thinner) hence whiter. They produce about 5-15% more light than argon filled lamps. Evaporation and failure of the filament is reduced because of the heavier krypton gas.

But true halogen lamps, with iodine or other halogens as the vapor surrounding the coiled tungsten filament, have an additional 10-20% improvement over the amount of light and color. The peak of the light output is shifted towards the blue, hence more lumens in the visible part of the spectrum per watt.

If you can somehow see a quartz tube within the bulb, then it's a halogen lamp. The pink coating may be there for further color correction. If no quartz tube is seen, then it's an ordinary tungsten filament with krypton.

[Medved]

The halogen lamps offer longer life, but due to positive temperature coeficient of the tungsten resistance the filament tend to heat up unevenly during the startup and some say even overheat, causing the tungsten to recrystallize, so loose the strength and for the next time make the unevenness worse, so with frequent switching the lamp life becomes very short. Mainly the lower wattage 230V (100W and below, so with a thin and long filament) are the most affected.
The cooler running "classic" argon or krypton lamps have higher temperature margin, so do not suffer from that effect.

And with halogen lamps, the halogens tend to attack the rather cold tungsten wire emerging from the quartz seal, so that part of the filament get's  eaten out by the halogen cycle - the same effect, what keep the bulb wall clean and moves the tungsten onto the filament.

[lights*plus]

Medved for very frequent on-off, does a dimmer switch help to preserve the filament in a halogen lamp...turn up the light by turning up the dimmer slowly rather than with a straight on-off switch?

[Mike]

I'd guess that with any filament-containing lamp, starting it at a low voltage (via dimmer) and then building up the voltage would help the lamp last a lot longer since there's not that sudden "rush" of voltage that could break the filament as it ages. I've noticed incandescents on dimmers seem to last forever (firguratively speaking of course, but even the standard 1000HR soft white lamps used  on a dimmer last longer than the extended life lamps on a standard switch.

[Medved]

A dimmer will help, the "touch sensor" controller IC's (e.g. SLB0587 or so) always soft-start the lamp, although there the aim was to prevent current surges into the triac and fuse.
Otherwise with the manual dimmer, you can never turn the knob so fast it won't soft-start the lamp either...
It's just sufficient to prevent the filament to approach too close the melting point.

As the regular GLS operate farther away from the critical tremperatures, so even the temperature overshoots at the hot spots do not reach it, they do not suffer from this problem (there are lot of flashers in use, where the lamp goes complete cold between light periods and yet the lamps sustain there their complete life rating of net-time burning, so no extra degradation from the ON/OFF cycles at all).

It is practically just the halogens, whose filament is designed to operate at the higher temperature, so with the overshoot the hot spots may reach the dangerous levels.

But still I'm not sure, this mechanism is, what make the mains halogens last so short time in the real life, it is just a hypothesis.

How the dimmers help with regular GLS life is, most of them are not able to exceed 90..95% of the full power, what mean the lamps run colder. But that mean the consequence of lower efficacy, so for the same light output, more expensive electricity bill...

[merc]

Linear halogens (such as Sylvania 240 V / 400 W) have their rated switching cycle only 8,000 times. (That's a switching cycle of an average/slightly better CFL.)
People used them in PIR sensor started floodlights at their houses before LEDs took their place. They were not seldom started by passerby persons or cars.
This way their lifetime was probably very short, especially in winter when the filament was being heated up from freezing temperatures.

[Medved]

@merc:
The "switching cycle" rating is based on 15minutes ON / 45minutes OFF. That means during the 8000 rated cycles the lamp burns 2000 hours, what equals the rated life.  So that means all rated wear goes on the account of the burning hours.

In a similar way, regular 1000hour GLS would bear "4000 cycles" rating figure.

So the rating claims no influence of the switching alone at all.

The fact is, the real influence of the switching cycles alone could be statistically hidden behind the normal burning wear with the test cycles designed as they are.

One could wonder if those tests were not artificially designed so, the CFL's "perform" always better than the GLS...

[merc]

Not in all cases though.

Osram Classic Superstar (halogen GLS retrofit) has "Number of switching cycles" = 50,000 and the balloon tip says: "At 1 min on / 3 min. off".
This is definitely better than most of CFLs.

The linear halogen Sylvania 240 V / 400 W in question has "No. of switching cycles before premature failure (min)" > 8,000 and they don't say anything about the switching scheme.

[Medved]

The switching scheme is in the standard and the mandatory rating information (after 2009 in the EU for all lamps for general lighting) is based on that standard. The official statement is "to have unified, but simple rating system for all lamps", how that "works" you know, I guess...

I know the "1min ON/3min OFF" makes way more sense for PIR sensor applications, but the EU regulations just ordered the 15min of ON time.

Or better to sat: All that applies for the mandatory information required to be printed on the lamp package (rated lumen output with the largest font, lamp power, lamp life in 165/15min scheme, number of cycles in 15/45(?)min scheme, color rating and rated voltage and socket).
Other information must be published in a lamp catalog (few spots from the mortality curve, ...) and there could be some other, non mandatory information, like e.g. the application specific life rating (like the 1/3min for the PIR's).

I know Osram was very "creative" with specifying 500000cycles or so on a plain PTC controlled preheat CFL's, but after the regulation came in force, the same lamps suddenly became 10000cycles/16000hour rating, that is quite strange...

[merc]

Here's another example I've just posted. It's definitely not a pre-2009 product and they specify 100,000 switching cycles both on the box and on their website.

The EU legislation may require something but might not enforce it (by penalizing producers, for example) so they specify more realistic (or better looking) values...?

But back to:

The halogen lamps offer longer life, but due to positive temperature coeficient of the tungsten resistance the filament tend to heat up unevenly during the startup and some say even overheat, causing the tungsten to recrystallize, so loose the strength and for the next time make the unevenness worse, so with frequent switching the lamp life becomes very short. Mainly the lower wattage 230V (100W and below, so with a thin and long filament) are the most affected.

If this value is true (and they specify the same value even for a 230V / 28W lamp with a very thin filament), the problem with recrystallization might not be such a big problem. Could be interesting to prove these switching cycle values in an independent lab.

[Medved]

Beside the official rating suggest no influence of the switching to the lamp life, real life tells something different.
The recrystallization effect will be strongly dependent on the impedance of the mains connection: Higher impedance means softer startup, so less severe overheating. Here I suspect the way such tests are usually performed: A test population is mounted in a rack, many bulbs in parallel and for the setup simplicity controlled by one common cycler switch. But that means all the lamps start at once.
While when the lamp is alone, the inrush current (6A or so) is not able to influence the mains voltage, when you start 100's lamps at once, the current becomes in the kA range, so becomes pretty potent to briefly reduce the mains voltage at the point of connection. The result is, the lamps are started "softer" during the test, compare to the real life operation.
And in the real life, the problem strongly depends on the mains impedance: If two lamps are connected in parallel, the inrush current of 12A could cause quite significant drop on the wiring (10V is already a difference here), so the lamps then start more softly , so last longer than in other installation, where the mains voltage is way more rigid.
And other aspect: The larger power lamps have way higher inertia of the filament, so tend to warm up more evenly, so the problem affects mainly the low power models...

And for the Emos lamp I haven't found any cycle life specification on the box (I have few just in front of me - used as a "fuse" for small ballast tests)
That is acceptable, when no degradation from switching cycles is anticipated (and that is the general belief about all incandescents)
In the web offering they specify the 100k cycles, but without reference to any standard nor test scheme, so it could mean anything...