Through The Past Darkly - Horrible Headlight History

[Junkman]

3. 1911-1920

In 1911 some U.S. passenger cars had electric headlamps installed as standard equipment, an industry first. However, these were high end luxury cars, hence volume was extremely low. In 1912 the first time comprehensive wiring harnesses including circuits for electric starters and an electrical system were installed in Cadillacs. This allowed an increased use of electric light sources and significantly increased the installation of electric headlamps on vehicles thereafter. It was also at about this time that the U.S. driver position was established on the left side of the vehicle. Consequently, all further lighting developments in the U.S. occurred with the same basic driver location as today.

Initially already available industrial light bulbs were used, but due to their filaments being made from sintered tungsten, they were very fragile and thus not really suitable for automotive application, especially when considering that country roads were still very bad and town roads often paved with cobblestones at the time. A solution had to be found and GM's Charles F. Kettering discussed several of the intermediate steps in going from industrial electric light bulbs to electric light bulbs specifically for cars in 1912. After several attempts, he discovered drawing or extruding as a means of making reasonably ductile tungsten wire.

At this time, most headlamps used a 21-candlepower bulb with a shallow parabolic reflector. It was claimed that this light source would light the full width of 15 m wide roads from the front of the car to a point over a quarter of a mile away, I wish they would employ such technology today. But no matter how you look at it, America had taken the undisputed lead in automotive headlight technology, at that point.

In Europe, it took car makers until 1913 to install their first electric headlamps as original equipment but in all fairness this can be considered being on par with America. Two years later, they started to fill tungsten filament bulbs with nitrogen gas (why did the Europeans always have such a knack for gases?) for use on cars. The gas vastly reduced the tungsten evaporation, consequently the filaments lasted a lot longer.

On both sides of the Atlantic, the headlamp glasses were just that. They did not have any purposely designed optical elements, so there was no bending, or spreading the beam pattern. While this was initially still acceptable due to the likelihood of one car encountering another on a nightly journey being practically zero, it became unbearable once cars started to become more commonplace, especially in conurbations. Consequently the number of accidents in such areas due to dazzled drivers became a serious issue.

According to my research, it was yet again America who took the initiative and thus the first ever automotive lighting regulation was adopted in the state of Massachusetts on 27/10/1915. This regulation provided that wherever there was not sufficient light on the road to make all substantial objects visible for a distance of at least 150 feet, the lamps which a motor vehicle was required to display, should throw sufficient light ahead to make clearly visible any such object within the specified distance. They provided further that any light thrown ahead or sidewise should be so directed, that no dazzling rays should at any time be more than 3.5 ft above the ground, 50 ft or more ahead of the vehicle, and that such light should be sufficient to show any substantial object 10 ft on each side, and 10 ft ahead of the vehicle. Needless to say that there are descriptions of the difficulty to enforce this regulation, as you where. However, to comply, motorists attempted to diffuse the glare by attaching pieces of frosted glass or other “glare eliminating devices” (read snake oil) available by mail order businesses to their headlamps. But it became pretty apparent that as time went on, the regulations were not very effective.
Depending on the sources, Baltimore, Maryland passed an ordinance which limited glare but did not control visibility either 1916 or 1917.

I once saw illustrations of the baffling array of different bulb sizes and filament shapes being used to meet these new regulations as well as the ones that were anticipated at the time, in a 1917 catalog. These bulbs were made by different manufacturers and were neither standardised, nor interchangeable. What they had in common, though, were simple transverse filaments, coiled-coiled spiral axial designs, and center supported V-shaped coils. The size of these filaments alone prevented reasonable control of the beam pattern and the output of the bulb designs varied from 135 to 584 lumens. However, lighting engineers back then were increasingly concerned with glare and visibility.

A contemporary survey (Clark, 1916) concluded that passenger car headlamps were located approximately 107 cm above the ground and the approximate location of the driver’s eyepoint was 137 cm above the ground. In comparison, the corresponding values for today's passenger cars are 62 cm and 111 cm respectively. What this means is that the headlights in the pioneering days were much closer to eye level, which made the avoidance of glare and dazzling a lot more difficult. This technical challenge led to numerous solution attempts that appear overly complex and downright bizarre today. Headlamps began to be festooned with all kinds of gadgets and “glare preventing” gizmos to still not really satisfy the concerns being raised by drivers and being written about in the technical literature. One common method, however, was to switch a resistor into the headlamp circuit to reduce the current to the headlamp bulb filament. Some headlamps were built with reflectors tiltable from the driver’s seat. Another method used a vacuum connection from the inlet manifold. A bellows, on the top of the headlamp reflector, was connected to the vacuum line and held the reflector in the correct position until the driver turned a valve closing the vacuum line, then the bellows expanded tilting the headlamp reflector downward. Also, some installations used electrickery tilting the reflector.

As you can imagine, these devices simply did not work and the reason for that wasn't even restricted to them malfunctioning, which of course happened frequently. But to an even greater degree, headlamp bulbs did not have an accurate location of the filament with respect to the base, which made any attempt to direct the beam accurately futile. The positioning of the bulb within the headlamp assembly also left much to be desired. Each of these issues contributed to glare.

A method for minimizing glare from filament mislocation is only described as late as 1921. The method described is to use two designated areas on the lens, so the filament image will only converge to one specified point when the filament is located at the focal point of the parabola. If the images don’t converge, an adjustment needs to be made. This same method could also be used to aim the beam pattern down to reduce glare. This general idea was discussed in the NHTSA (National Highway Traffic Safety Administration) Regulatory Negotiation meetings in Washington, D.C. in 1995-1996 for visual/optical headlamp aim. No one at the event realised then that it was a 1921 idea.

But I digress. Back in the days, along with technical product developments, standards and regulations were also being written with gay abandon. In 1918, the first joint IES (Illuminating Engineering Society, not related to the Illuminati) and SAE (Society of Automotive Engineers) specification relating to the optical performance of headlamps was developed. Vehicles only had one beam, much closer to a main beam than what is now considered a dipped beam. Four test points were determined for measuring the beam pattern. The specification was written for the combined light output from both headlamps on a vehicle, to my knowledge the first instance of a vehicle based specification. The complete specification was as follows:

Measurements shall be made at the following positions at a distance of 100 ft ahead of the headlamps:

Position 1: Directly ahead and at a height not less than one half the distance of the center of the headlamps above the level surface. The indication of the foot candle meter shall be not less than 0.48 ft-c (5.2 lux) for a motor vehicle and not less than 0.24ft-c (2.6 lux) for a motorcycle.

Position 2: Seven feet to the right of Position 1 at any point not above the level of the headlamps.
The indication of the foot candle meter shall be not less than 0.12ft-c (1.3 lux) for a motor vehicle and not less than 0.06 ft-c (0.6 lux) for a motorcycle.

Position 3: Directly in front, 5 ft above the level surface. The indication of the foot candle meter shall be not more than 0.24 ft-c (2.6 lux).

Position 4: Five feet above the level surface and 7 feet to the left of the axis of the vehicle. The indication of the foot candle meter shall be not more than 0.08ft-c (0.9 lux).

Note: In order to allow for any possible inaccuracies of a test of this character, a tolerance of 20 per cent may be allowed on the above values.

The following diagram shows the test point locations and intensity values for this beam pattern. A headlamp mounting height of 90 cm was used to establish the locations of Positions 1 and 2.

Fig1.png (54.69 KiB) Viewed 5706 times

Perhaps the first properly documented study of headlamp performance was done in 1920 (published by Devine, 1921). Thirty headlamps were obtained and photometered in a test laboratory developed for the state of - you guessed it - Massachusetts. These headlamps were then used on road tests to compare the laboratory photometry results with on road performance. These headlamps used 1 3/8 inch focal length parabolic reflectors and were 9 inches and 9 1/4 inches in diameter. The glare limits listed above were found to be reasonable, while it was determined that additional light was necessary forseeing objects on the road.

Why, you ask rightfully, have I been hitherto so America centric? Well, up to this point, the technical evolution of the headlamps were similar enough in Europe to not warrant the effort to outline them separately. Also, as usual, the Americans have it much better documented and archived, thus researching it there is a lot easier and they consistently were one little step ahead of the Europeans.

It wasn't until later in the 1920s that Europe and America went entirely separate paths in automotive headlight technology, thus making my further research a lot more work.

[Junkman]

4. 1921-1930

Another early study was conducted by engineers Magdsick and Falge at GE Nela Park in Cleveland in 1921. Observers determined the amount of light in various parts of the beam pattern that “would satisfy the driver from the standpoints of safety and afford a reasonable degree of convenience in operating the car”. Considering this was a study within a commercial corporation, it was surprisingly thorough and comprehensive, conducted on city and country roads, and in a variety of weather conditions. The study also included comments about the legislative activities in several states. The results led to a revision of the IES-SAE photometric requirements for headlamps, in 1922. More test points were added and the “seeing light” values below the horizontal were increased. Again, the numbers refer to the light intensity in candela (cd):

Fig2.png (17.6 KiB) Viewed 5698 times

During this decade, inspection stations were established in various U.S. states, with the requirement that drivers of vehicles obtain a certificate indicating the headlamps of their cars were properly aimed had sufficient light output, and did not cause excessive glare. There were discussions at this time to specify a minimum light value above the horizontal line, to adequately illuminate “low hanging trees and other obstructions from above”. Indeed, a minimum light value above the horizontal line was included in the 1922 revision (see above diagram). These early developments regarding vehicle headlamps were still very similar for the U.S. and Europe. Bulbs, lighting components, and beam patterns were essentially the same.

However, in the 1920s two different approaches to automotive lighting began to appear. Many people believe this must be attributed to the more rapid increase in the vehicle density in the U.S. than in Europe. Those Fords, you know.
But, and now it becomes really interesting, the real reason was that there are fundamental differences in the use of the automobile here and there. The Americans adapted to the mobility provided by the automobile because it was necessitated by cities expanding and people beginning to move to Suburbia, while the industries also developed outside of the town centres. Hence Americans drove their cars often on the open roads at night, while Europeans retained the focus on the city centres and daytime driving, a terrible tradition that plagues European cities to this day. It is generally acknowledged (Nelson, 1954; deBoer, 1955; Meese, 1972; Olson, 1977) that these factors are the ones that contributed to the two different approaches, which can be summarised as follows:

U.S. philosophy:
The primary emphasis is to develop as much light as possible to maximize seeing ahead of the vehicle, while considering the other driver and trying to do something to minimize glare is only the secondary emphasis.

European philosophy:
The primary emphasis is to do whatever is necessary to minimize glare in the other driver’s eyes, while trying to make sure there is enough available light to drive is only the secondary emphasis.

Rules and specifications prepared by the IES were submitted to the American Engineering Standards Committee for adoption as a tentative U.S. standard in 1921. These same rules and specifications were then submitted for consideration to the International Commission on Illumination (Commission Internationale de l’Eclairage, CIE) for a 1921 meeting. At that meeting the following resolution was adopted:

The International Commission on Illumination records its strong opinion that regulations respecting limitations of the light from automobile headlamps should be framed with reference to international agreements so that the regulations may be uniform for all countries.

Is this the first attempt ever at global automotive standards harmonization? If yes, then it was about headlamps and you may now begin to realise why they are so significant in the eye of automotive historians.

Also established at that meeting was a committee to study the problem of automobile headlighting and to report at the next session of the Commission. There are groups still working on this today, believe it, or not.

In the U.S., the IES was working on improving the details of their first rules. It was decided to only consider 21 candlepower bulbs in headlamp sizes of 8 5/32 inch, 8 1/2 inch, 9 inch, and 9 1/2 inch diameter. Before 1924, all headlamps generated a single beam pattern, best described as a spot beam resembling today’s main beam.

Then, in 1924, Philips in Europe developed the first two filament headlight bulb ("Bilux"), followed shortly thereafter by a similar design developed by Guide Lamp in America ("Duplo"). The two filament automotive bulbs enabled a main and dipped beam pattern to be created in one and the same headlamp assembly. In these bulbs, the main beam filament was placed on the focal point and the low-beam filament was mounted above and to the left. Note that I deliberately didn't say "or to the right for use in countries that drive on the left". This will be explained later.

Anyway, the two filament bulbs allowed the low beam pattern to move down and to the right (side shifting). This arrangement of the filaments continued until the mid 1950s in U.S headlamps. In Europe the two filament Bilux bulbs with their internal shield became the commonly accepted standard, yielding an asymmetrical flat topped beam pattern. This type of beam pattern lasted until shortly after WWII.

[Hooli]

This is actually a lot more interesting than I was expecting, thanks for posting it all.

[Junkman]

5. 1931-1940

Since the UK and other countries driving on the left could not adopt the Bilux system for obvious reasons, different solutions had to be employed. One approach to somewhat control glare had a movable reflector assembly to be used when another vehicle approached. This assembly was tilted downwards and to the nearside by means of an electromagnet operated by a switch under the control of the driver. The switch also caused the offside headlamp to be turned off!

Meanwhile in the U.S. several different combinations of beam patterns were developed to try to provide improved lighting with some measure of glare control. Three filament bulbs were developed to create four beam patterns:

a regular main beam, a regular dipped beam, a meeting beam to use when approaching other vehicles, and a super beam for straight roads in the country with no opposing traffic for miles on end. These headlamps were used on some cars in 1932 and 1933, such as Packards. Three beam headlamps had a “traffic beam” aimed high on the right and low on the left, in addition to the regular main and dipped beams. Thus, an American automobile might have been fitted with either a single beam, two beams, three beams, four beams, or side shifting beams.

By the mid 1930s many states had lighting regulations in place to control glare. Creative inventors had developed diffusers, intensity reducers, masks, etc., in an attempt to deal with glare as well as statutory regulations. As you can imagine, sharlatanry also fell on fertile grounds. Headlamps came in all kinds of different shapes and sizes, and the mechanisms for aiming these headlamps were not at all reliable. Some involved bending of sheet metal brackets, while others called for moving the bulb position inside a headlamp. All of this only caused state officials to enact even more legislation. Lighting engineers, automotive manufacturers, and state officials were all trying to do something to provide better lighting with lower glare. During these discussions the SAE prepared a revised beam pattern in 1933). This pattern is shown here (Intensity in candela):

Fig3.png (18.9 KiB) Viewed 5695 times

In 1937 France legislated that headlamps emit selective yellow light. The French gave the following reasons for requiring this:

improvements in visibility under conditions of glare, and, even more important, reductions in fatigue under glare conditions. The yellow headlamp color was obtained by several different methods: a glass bulb with a different chemical composition, coatings on the glass bulb, a yellow headlamp lens, or a yellow filter.
Because the yellow glass for bulbs reduced the emitted light intensity by about 18%, it is often claimed this may have contributed to the perception of reduced glare from yellow headlamps, however, this is not the case. However, many tests were run in different countries, all failing to show any significant advantage of yellow headlamps. However, within France, the mandatory use of yellow headlamps continued until the early 1990s.
This warrants a separate chapter, which will be delivered later.

Back in the U.S, by the late Thirties the dissatisfaction with the endless headlamp legislations and the chaos thusly created had reached such a level that something new was needed. The solution came in from of the glass sealed beam headlamp, which became the answer to most of the above problems. The sealed beam consisted of two pieces of glass, a lens with optics, a parabolic aluminized reflector, and tungsten filaments supported by metal lead wires. The two glass pieces were fused (melted) together, and the assembly was filled with an inert gas to prolong the life of the tungsten filaments. Within two model years, 1940 and 1941, all U.S. car production switched to one common headlamp size, 7 inch round, with two headlamps per vehicle. There was no legal alternative to this until 1958. You can slag the yanks as much as you like, but when they standardise something, it's standardised.

However, at the time of its inception, the all glass sealed beam was arguably the world's most advanced control of an automotive lighting product sold to the general public at that time. Because of patent restrictions, some “sealed beam headlamps” were made with a metal reflector, a soldered in bulb, and a glass lens permanently fastened to it. These metal backed “sealed beams” were finally phased out by the mid fifties, when the entire U.S. headlamp industry began making only all glass sealed beams. They are thus highly collectible today.

Also, in 1940, the maximum intensity per vehicle was raised from 50,000 cd to 75,000 cd. The SAE revised the dipped beam specification accordingly by decreasing the allowed glare light and increasing the minimum required values below the horizontal line:

Fig4.png (20.01 KiB) Viewed 5695 times

[Junkman]

6. 1941-1950

There weren't any significant technical advances in headlamp design during this decade, since a selected few clinical sociopaths had other priorities in its first half, making sure people being too poor and too few to drive cars a lot in its second. However, once the hostilities had ended, there were many American cars on European roads, of course equipped with sealed beam headlamps. This had an oddly positive side effect. It accelerated the evaluations and comparisons of U.S. and European headlamp beam patterns throughout the early Fifties.

[Junkman]

7. 1951-1960

At the beginning of this decade, all U.S. headlamps were as follows: 7 inch round sealed beams with the main beam filament on focus and the dipped beam filament off focus, visual aim, and more consistent lighting than previously obtainable.

European headlamps had replaceable Bilux bulbs with an internal bulb shield on the dipped beam filament, a meanwhile symmetric dipped beam pattern to make them useable in the UK and other countries driving on the left, visual aim, and continued emphasis on the accuracy of the filament location. This commonly agreed standard remained virtually unchanged until well into the 1970s.

Consequently headlamp beam patterns were vastly different in the U.S. and Europe. Lighting engineers in collaboration with lighting and vehicle manufacturers organized a series of tests to yet again try and develop a common beam pattern. These tests were the origin of the Groupe de Travail Bruxelles (GTB), in 1952, an international group of experts from light source, lighting device, and vehicle manufacturers. But even with the driving evaluations and the considerable research that was done, it was not possible to reach a compromise on one common beam pattern, again caused by the different priorities for minimizing glare versus maximizing seeing light. On the positive side, a common aiming method (visual aim) in both Europe and the U.S. was agreed on. This, however, was rather short lived, as you will soon find out.

In 1954 the Americans increased the wattage of their sealed beams from 40 to 50 for the main beam filament and from 30 to 40 for dipped beam one, while at the same time tilting the reflector axis down, which resulted in a change of the way the optical prescription was generated. The dipped beam filament was now placed on focus and an external shield or “fog cap” was installed to block the direct forward light.

The SAE Lighting Committee and other industry groups held many tests with participants, also international ones, visually aiming U.S. beam pattern headlamps. The results indicated that visual aiming the existing U.S. beam patterns was not very accurate. Mechanical aim removed the subjectivity of visually aiming the beam. Consequently the agreed on GTB method of visually aiming the headlights was abandoned and mechanical aiming became the only legal method to adjust headlamps in the U.S. until 1997. Therefore, in 1956 “aim tips” or “aim pads” were added to the front of the glass lenses of the sealed beams to enable this mechanical aiming.

In 1957 the 5 3/4 inch round four headlamp system was developed in the U.S., so for the first time since 1940 America had an alternative to the two round 7 inch sealed beams. The new system became legal in all states in 1958 and consisted of two round main beam lamps and two round dipped beam lamps, the latter being fitted with a secondary filament to provide some fill light on main beam. Thus, the lens optics compromise necessary for a combined dipped and main beam light was eliminated, because each headlamp had a single primary function.

[fried onions]

Thank you. My studies are enlightening.

[Datsuncog]

Very much enjoying this.

It's the kind of stuff I used to buy classic motoring magazines to find out about - but nowadays I rarely seem to encounter informative, in-depth articles written by articulate people with a deep knowledge of their subject.

Instead the big-meejah boys just seem to churn out hundreds of sooper-HD Photoshopped images accompanied by a scant few lines of text each month. Guess that's what the punters seem to want these days while browsing the mag stands in Tesco, because I know the journalists themselves are capable of much more.

Guess that's why I'm here, and not flicking through Practical Classics...

[8BAK465]

Looking forward the the Yellow Headlghts bit

I saw a modern Eurobox the other day sporting Yellow Light wayy after the 93 compulsary ending but I guess it was just Pineapple Scene

Incidentally I remember during the 80's everybody buying the yellow clip on's for going to France but would take them off the second they rolled off the boat like they were going to be flogged in public if they didnt

Whats the law here regarding yellow headlights ?

Your Renault would look the part with them

[Junkman]

8. 1961 - 1970

In 1965 SAE J579a was written to make some further refinements to the U.S. dipped beams.

Fig5.png (19.72 KiB) Viewed 5665 times

Also, in 1966, the U.S. government passed the Motor Vehicle Safety Act. The U.S. National Highway Traffic Safety Administration (NHTSA) was subsequently created in 1968 to improve the safety of road transportation by focusing on the design of vehicles. Prior to this, lighting was regulated solely by the individual states, although they all adopted the guidelines set by the SAE. Now, within two years, many existing SAE lighting standards were written, in whole or in part, into the Federal Motor Vehicle Safety Standards (FMVSS).

The evaporation of tungsten filaments while the light source is switched on was already discovered way back in 1912. This evaporation process will concentrate at some weak spot, scratch, or mark on the tungsten surface. Eventually the filament becomes thin enough to burn out. It was also discovered, that when filling a light bulb with certain halogen gases, the gas molecules combine with the evaporating tungsten molecules to form a compound that will circulate back to the filament, where the tungsten will dissociate and redeposit itself on the hot filament. As a result, this allows the filament to shine brighter and hotter without shortening its life. However, it will also burn out after a while.
That automotive applications of halogen headlamp bulbs did not occur until the mid Sixties, despite they were used in industrial and household applications much earlier than that, is downright unbelievable. In the end, for once Europe stopped procrastinating quicker than the U.S. to take advantage of this improved light source, by introducing single filament H1, H2, and H3 halogen filled bulbs, first for use in auxiliary lights, then also in headlamps. In the late Sixties, the double filament H4 bulb for specific use in headlamps was released. Although the dipped filament in the H4 bulbs is centric, H4 bulbs are exclusively used in headlamps with optical beam bending for asymmetric dipped beams, necessitating specific headlights for left or right hand traffic on the same make and model of car.