Why Are Tiny Light Bulbs So Inefficient?

[Seun Osewa]

After a bit of research, I discovered that very small halogen lamps (50W or so) and smaller incandescent lamps (flashlight bulbs) are orders of magnitude less efficient than their moderately sized or large equivalents.

Why is this so?

It would have been nice to be able to assemble 256 flashlight bulbs to make a pseudo-soft light source running on batteries (no inverter required). Unfortunately, given how inefficient these bulbs are, such a fixture would be horribly inefficient compared to diffused incandescent lighting.

What do you think?

[Chris Soucy]

Er, I don't think incandescents differ that much in their lumens/ watt ratings, but I could be wrong.

As for the rest of your question - have you investigated high efficiency Led's?

They're incredibly expensive for good quality ones, but their efficiency is pretty (very) good.


CS

[Seun Osewa]

According to this page:
A 100W incandescent outputs about 17 lumens per watt while a flashlight bulb outputs less than 6 lumens per watt. So the flashlight bulb is roughly three times less efficient than the 100W bulb.

Wish I knew why.

As for the rest of your question - have you investigated high efficiency Led's?

To simulate a large soft light with lots of small hard lights, I guess LEDs are the only way. That's not cheap, though. My problem with softboxes/flos is that they absorb so much light and are reportedly to weak to be used from a distance unless one buys very many of them; which would be hot/expensive.

[Boyd Ostroff]

I have no scientific data, but I assume it has something to do with the operating temperatures of flashlight bulbs, which must be kept low for obvious reasons.

I don't know the technical details, but a low voltage halogen flashlight bulb is a very different animal from higher wattage halogen lamps like the ones used in stagelights. The lamps in stagelights are made of quartz instead of glass because the temperatures get so high that glass would melt. And that high temperature is part of the tungsten-halogen cycle which allows the filament to actually "renew" itself, leading to longer life than a comparable incandescent lamp.

This stuff is all black magic to me though... maybe someone else can explain the different technologies to us?

[Richard Andrewski]

Some common lumen per watt ratings:

14-17.5 lpw 60 to 100 watt incandescent bulbs
16 to 21 lpw for most halogen lamps rated to last 2,000 hours or more
average 30 lpw for LEDs

A quote from a site explaining why larger wattage tungsten are more efficient than lower wattage:

http://members.misty.com/don/lede.html

"The most obvious disadvantage of lower current incandescents is that the thinner filament must be operated at a slightly cooler temperature to reduce the evaporation rate accordingly. The lower temperature disfavors radiating visible light, while increasing the already-high percentage of radiation that is infrared. To some extent lower voltage incandescents can be more efficient than higher voltage ones of the same wattage since the filament is shorter and thicker and can be operated at a higher temperature for a given life expectancy. The most efficient incandescents of a couple hundred to a few hundred watts are designed to operate at 20-28 volts, the most efficient ones of around 10-100 watts have design operating voltages not far from 12 volts, and the most efficient ones (of a given life expectancy) of around a watt to a few watts have design operating voltages not far from 6 volts, and the most efficient ones of a fraction of a watt generally have design voltages close to 5 volts. But lower-than-optimum design voltages have the disadvantage of heat conduction from the ends of the short thick filament.

Another advantage of thicker filaments is the advantage of a fill gas. The fill gas slows down filament evaporation because gas atoms "bounce" evaporated tungsten atoms back to the filament. This permits a higher filament temperature for a given life expectancy. A disadvantage of the fill gas is that the gas conducts heat from the filament, and that is energy that cannot be radiated and this is a loss.

A thinner filament has a thinner "boundary layer" of hot gas around it, and a higher temperature gradient within this "boundary layer". For this reason, the amount of heat conducted from a filament by the gas is nearly proportional to the visible apparent length of the filament wire, and hardly varies with filament diameter. A thin filament has almost as much loss per centimeter as a thick one, despite the thin one having much less input power per centimeter.
The cheapest fill gas, sometimes argon but more usually a traditional mixture of argon and nitrogen (nitrogen impairs destructive arcs which form excessively easily at over approx. 12 volts in pure argon), generally is better than no gas at all (vacuum) if the wattage exceeds approx. 8 to 10 watts per centimeter of visibly apparent filament length. Premium gases such as krypton and xenon are better, but are still worse than a vacuum if input power is less than a few watts per centimeter of visibly apparent filament length.

In addition, with ratio of power consumption cost to lamp cost lower with lower wattage ones, incandescents of lower wattages tend to be designed for longer life - which further impairs the efficiency of lower wattage incandescents.

If you only half-followed the previous few paragraphs, don't sweat it. But a 4 or 7 watt 120 volt night light bulb usually has an efficiency (more accurately, overall luminous efficacy) of only 4 to 6 lumens per watt. Most 15 watt 120V incandescents get around 8 lumens per watt. This is obviously much less than the 14-17.5 lumens/watt of most "regular incandescents."

So the basic issue is that higher wattage tungsten halogen lamps (like those commonly used in stage / studio instruments) are more efficient than smaller ones for the following reasons:

thin filaments vs. thick filaments
less halogen vs. more halogen
thinner boundary layer vs. thicker one
long life vs. short life

[Boyd Ostroff]

I kinda thought you might sort this all out for us Richard -- thanks!

[Jim Andrada]

I was also thinking along the lines that the physically smaller the bulb the closer it approximates a point as opposed to extended source so the faster the light would fall off with distance. Also,designing an optimal reflector for a lot of small bulbs might be a challenge.

[Richard Andrewski]

Originally Posted by Boyd Ostroff

I kinda thought you might sort this all out for us Richard -- thanks!

You were definitely on the right track!

[Richard Andrewski]

Originally Posted by Jim Andrada

I was also thinking along the lines that the physically smaller the bulb the closer it approximates a point as opposed to extended source so the faster the light would fall off with distance. Also,designing an optimal reflector for a lot of small bulbs might be a challenge.

Yes! This is why it will be tough to have a CFL or LED fresnel. Neither of those are capable (today) of a single concentrated and intense small point of light. Otherwise, put a collection of them together with a huge mirror behind them and you have a large source that someone might try to call a fresnel but they'd be wrong. You'd have to make up a new name for it.

[Peter Ferling]

As much as I like CFLs, I still break out the tungstens for backlighting and effects. Eventually they will heat the entire room (almost negating the advantage of CFLs). Therefore I am careful to leave the hot lamps on only as needed. It can be managed.

I have experimented with various LED lightbulbs (a good excuse to visit the hardware store and buy tools :) Mostly for my education. Yes, generally they are much brighter their inefficient incadenscent cousins. However, you have to modify the lens and gel them, which robs them of that xtra brightness.

[Bob Grant]

And now that you know the answer spare a thought for those of us using 240V, our inkies put out even less light and don't last as long as the 110V equivalent. On the upside though the power cords are easier to wrangle.

[Richard Andrewski]

Actually I wish the whole world operated off 240v 50hz. It would make things so much easier for manufacturers and its also more logical when you consider that you have less amps drawn for a given load and therefore more capacity to run larger loads. For instance, it's no big deal for many to run a 2000w fresnel in their house in other countries but that's pushing it in the USA. 18amps in the USA but 9 amps elsewhere. Household circuits everywhere are pretty much 15a breakers or fuses on the average no matter what voltage so you can see you have better capacity outside of the USA (and Japan has it even worse than the USA as they operate off of 100v).

[Marcus Marchesseault]

It strikes me as ironic that we power-hungry Americans have the lowest capacity in our electrical systems.

"As much as I like CFLs, I still break out the tungstens for backlighting and effects."

I'm eagerly awaiting the release of the coolights.biz 150W HID. With the brightness and focus like a 650W tungsten fresnel, it should be quite handy as background lighting for soft fluorescent keys. It even has the option for either tungsten or daylight color. Fluorescent and HMI/HID are the perfect compliments to each other and neither are so wasteful as tungsten. Living in a tropical environment, I can't stand hot lights on a cramped set with too many cast and crew trying to keep out noise by shutting the windows. I'm assuming that in temperate zones that hot lights overwhelm air conditioning in the summer just like they do here. To top it off, finally getting rid of the last reason to use tungsten will be good for the environment.

[Jim Andrada]

Not sure if you're aware of it, but Japan is more of a nightmare re power than you might think.

They're 100V allright, but frquency differs depending on where you are. Southern/Western Japan is 60 Hz, Eastern/Northern Japan is 50Hz, including Tokyo.

Even worse - some buildings in Tokyo run on 60Hz even though Tokyo is supposed to be 50Hz. When I lived there I was working for IBM Japan and the big banks in Tokyo all ran 60 Hz systems with their own motor generators to power the computer rooms. In the early days, all the computers were imported from the US and were 60Hz machines, so the banks, which were early adopters, standardized on 60Hz.

Of course, when a system was ordered, the IBM ordering system knew that Tokyo was a 50Hz location so all the mainframes came wired for 50Hz and had to go back to the factory to be converted to 60Hz. And if the customer wanted to move the machine to Osaka, it had to go back to the factory again to be reconverted.

Thank God for modern switching power supplies!

[Richard Andrewski]

Originally Posted by Marcus Marchesseault

I'm assuming that in temperate zones that hot lights overwhelm air conditioning in the summer just like they do here. To top it off, finally getting rid of the last reason to use tungsten will be good for the environment.

That's the whole reason I got into Cool Lights in the first place Marcus. I lived in Houston and it doesn't get much more tropical than there. In the middle of the summer, my little backyard studio air conditioning just couldn't keep up so that's when I started investigating other solutions and back then fluorescent was still very expensive.