One of the issues regularly discussed results from the widely divergent designs of LED light sources, including the misapplication of 'bulbs' designed for tail-light use (where 'end' light output can be more important than side) or, not uncommonly, manufacturers not understanding headlamp design requirements and producing 'bulbs' of dubious practical use. The latter may be done in an effort to obtain greater overall light output from a high number of small runout LED sources - more is better, right?
Well, no, not necessarily; and so that you can make a more informed choice with regard to headlamp light sources I thought it worth briefly discussing headlamp design fundamentals, with particular regard to the early 20th century. For the purposes of this discussion I'm concentrating on the typically Model T headlamp with 'fluted' lens, although the short explanation below will give some background to other headlamp design.
The Model T fluted lens headlamp is an optical system that combines two properties to create the desired light beam cast. These are Catoptrics (reflected light through the use of a mirror), and Dioptrics (the refraction of light through the use of a lens), together this makes the Model T headlamp a Catadioptric system. For those interested in ancient history it's likely that Catoptrics preceded Dioptrics by several hundred years, but it's not until the 19th century that the two really combined to make an effective system. The French were early experimenters here, but in my view it was only around the time that the Stevenson's began their remarkable period of lighthouse construction that things developed sufficiently to set the stage for comparatively high-performance and easy to use T headlamps in the 1910's.
With a T headlamp you have a source of light, originally intended to be from a small V-shaped tungsten filament, that is applied at a specific place within a curved reflector that, when adjusted correctly, will project a relatively round focused beam outward. This utilises the catoptric principle, it provides a reasonably bright beam, but it can be dazzling to oncoming traffic, and will usually not give a laterally broad enough light coverage for the driver.
Enter the fluted lens. When applied to a working lamp the vertical flutes act as lenses and will widen the original catoptric focused beam sufficiently to improve the lateral spread without losing focus or increasing the beam height. Later lens still can often have a combination of differently formed lenses in the glass so as to provide an even more controlled beam output. The Ford Model A is such an example, as is the ubiquitous 7" sealed beam headlamp of later years - by comparison the Model T headlamp lens is very simple thing - but not ineffective all the same.
As well as the optic system itself light source design has developed over time. The original V-shaped tungsten filament bulb was a 6V 30-odd 'candlepower' device that required a reasonable amount of current to operate satisfactorily. They were specified (see SAE Handbook, Vol1) as having a light centre of 1 1/4" +/- 3/64", as measured from the centre of the filament field to the bulb side of the lock pins in the base (of the socket). The specification goes on to say that the longitudinal axis of the lamp filament at the focal point shall not deviate more than 3/64" from the axis of the lamp base.
These dimension limits were required to standardise bulbs such that they would provide a suitable output at the correct focal point of a typical paraboloidal reflector then in use. The resultant beam should have little spurious light spill (ie. it should be as near 'round' as possible), and be sufficiently 'powerful' with the then somewhat limited light source.
In later years I've noticed that some tungsten filament bulbs in use have not had the requisite V-shape, meaning a dimensionally different light source than originally specified, and more recently people have utilised a variety of LED bulbs, some of which have an extremely broad light source, often both along and orthogonal to the bulb axis. Some may also have a thin optically impervious plane, either side of which is mounted a (light emitting) diode or two. It should be clear to the reader that both the 'incorrect' tungsten bulb, and many LED 'bulbs' will lead to an imperfect Catoptric derived beam, that in turn can affect the output of the system as a whole when the glass lens is added.
How imperfect, and whether it is of consequence will depend upon several factors. Primarily, from an oncoming driver's perspective, the beam should not spill upwards strongly enough to dazzle them, and for the driver of the source vehicle they need a beam that has good breadth and cast, and is sufficiently powerful enough to provide good illumination of the road surface and surrounds. Given that many LED light sources will have significantly greater light output than a traditional 32CP tungsten bulb it perhaps matters less from a beam intensity perspective that their light is often cast from greater area, however this can affect the 'perfectness' of the beam and through rogue light rays from the reflector and refraction from the lens may spill sufficient light to bother the oncoming driver. Thus in choosing a LED lamp there are several things to note:
- Try to obtain a unit that has as narrow a light source as possible - ideally it should output ~1.25" from the pins, and be of limited width.
- When trialling a LED lamp it's best to focus with the lens off. This goes for tungsten-filament bulbs as well.
- LED's can be very 'white' or 'cool' (i.e. they have a high temperature). It's best perhaps to select bulbs with lower temperatures, around 2800/2900K is a good number. That said many modern headlamps operate successfully at higher temperatures, however drivers may complain they're too 'bright', or 'they don't look right'.
- Although it's difficult to provide a direct correlation, there are simply too many variables, bear in mind that the original 'candlepower' measurement was a somewhat crude measure and is no longer used. Nowadays LED's are often rated in lumens and, without wanting to get into pages of discussion, you could crudely say that a 400-lumen LED lamp will have similar light output to say a 32CP tungsten filament bulb (L = CP * 12.57). This gives a guideline at least.
Finally, this post is just intended to go over some rough points and provide basis for informed and informal discussion. It's certainly not exhaustive, nor is it a scientific article - particularly given I've left a lot out and tried to avoid too much esoteric 'stuff' (and some of it is opinion) - and indeed parts may even be contentious, but hopefully it'll be of interest to some. As a matter of further interest and useful record I also include below some relevant pages from the Bureau of Standards Motor-Vehicle Headlighting circular No. 276 of 1925 - no doubt the fluted lens displayed will be recognisable!
Luke.