APRIL 2020 LIGHTING JOURNAL28

THE

ROAD AHEAD

THE

ROAD AHEAD

Vehicle headlamps were once rated in units of candle power, which as an apt description for my first car, a 1969 Morris Minor.

The designer of the Morris Minor (also of the ubiquitous Mini) was Sir Alec Issigonis, who once opined that, in the inter-ests of safety and staying alert behind the wheel, he should make his cars as uncom-fortable to drive as possible; to the extent of fitting a spike to the steering column to impale the drive in the event of a collision.

I’m not sure how serious he was about this particular ‘passive’ safety feature, but the dim illumination provided by my Morris’ headlamps led in no small part to the obser-vance of speed limits, at least at night.

Today, vehicle lighting technology has advanced out of all recognition, with my cur-rent car luxuriating in the excess of ‘active pixel LED headlamps with laser assisted main beams’. Short of the light output from a thermonuclear explosion (or two ‘peta-watts’, or 2 x 10^12 candelas per sq m if you’re interested), you won’t find a brighter source of artificial illumination, with night almost being turned into day.

In this article, I intend to review the development of vehicle headlamp tech-nology (from candles to LEDs and lasers) and explain the metrics used to express the output of lamps and the methods used to measure them.

FROM CANDLES TO TUNGSTENThe very earliest vehicle lighting was based on candles or burning oil (for example acet-ylene), but light output was less of a con-cern because at that time vehicle speeds were strictly enforced by a man walking in front of the car carrying a red flag. Through-out most of the 20th century vehicle head-lamps were based on incandescent lamps.

The invention of the incandescent lamp is popularly credited to Thomas Edison in 1879. Incandescence (as I imagine most ILP members will well know) is the emission of (visible) light as a result of resistive heating of a metal (filament of tungsten). The more current that flows, the greater the heating effect, the greater the incandescence.

With an efficacy of about 10 lumens per Watt (electrical to optical efficiency), incandescent lighting will therefore never set the world alight. The lamp was posi-tioned at the locus of a reflector (usually parabolic) so as to collimate the beam, while the desired beam distribution was achieved using prismatic optics moulded into the headlamp lens. Figure 1 overleaf shows a schematic of a lens-optic headlamp.

COMPLEX REFLECTOR HEADLAMPS (1983) The 1980s saw the introduction of com-plex reflector optics in headlamp design, and the gradual shift away from simple,

www.theilp.org.uk

APRIL 2020 LIGHTING JOURNAL 29

Automotive lighting

The arrival of LED has transformed car and automotive lighting, providing safer, less stressful illumination for drivers while avoiding dazzle, glare and discomfort for other road users. But robust testing

remains of paramount importance within the industry

By Robert Yeo

The arrival of LED has transformed car and automotive lighting, providing safer, less stressful illumination for drivers while avoiding dazzle, glare and discomfort for other road users. But robust testing

remains of paramount importance within the industry

parabolic reflectors. The optics required to shape the headlamp beam were designed into the reflector, rather than into the lens itself.

Complex reflector optic headlamps pro-vided an improvement in light collection from the incandescent lamp and better control of the beam shape. The 1983 Austin Maestro is credited with having the first production complex reflector headlamps. Figure 2 shows a schematic of a reflector headlamp.

PROJECTOR HEADLAMPS (1986) The 1980s also saw the development of pro-jector headlamps. This design places the lamp at the focus of an ellipsoidal reflector and uses a condenser lens to collimate the beam.

This type of headlamp offered the vehicle designer greater flexibility because of the reduced cross-sectional area of the head-lamp, albeit with greatly increased depth.

The 1986 BMW 7-Series (E32) is cited as being the first volume production car with polyellipsoidal projector low beam head-lamps. Figure 3 shows a schematic of a pro-jector headlamp.

HALOGEN LAMPS Separate from developments in headlamp optical design, the simple tungsten incan-descent lamp was being replaced by the quartz tungsten halogen (QTH) lamp. First developed in 1955 by Elmer Fridrich and Emmet Wiley at the General Electric Com-pany, the so-called ‘halogen’ lamp improved upon the original tungsten lamp by burning brighter and extending the life of the fila-ment. A halogen gas (typically iodine or bromine) is placed within the quartz (fused silica) vacuum envelope of the lamp.

As the tungsten filament evaporates, rather than condensing on the glass enve-lope of the lamp and darkening it (as occurs with a standard incandescent lamp), it reacts with the halogen gas to form a halide, which doesn’t deposit on the glass. Instead, the tungsten halide dissolves back to

tungsten when in close proximity to the fil-ament, returning the metal back from whence it came.

This process is known as the halogen cycle and it has the major benefits of keep-ing the glass of the lamp clear and extending the life of the filament. The luminous effi-cacy of quartz tungsten halogen lamps var-ies in the range from about 12 to 24 lumens per Watt.

COMPACT FLUORESCENT LAMPS (CFLS) As an aside, spiral compact fluorescent lamps (CFLs) are unsuitable for vehicle lighting. It is more difficult to efficiently collect and collimate light from an extended (larger) source area, and the sheer size of the emitting area of CFLs alone precludes them from automotive lighting applications.

The spiral format CFL was developed by Edward Hammer at General Electric in 1976 in response to the 1973 oil crisis.

HIGH INTENSITY DISCHARGE (HID) LAMPS, AKA ‘XENON’ (1992) During the 1990s, high intensity discharge (HID) lamps started to become popular, referred to as ‘xenon’ in the vernacular.

The very first dipped-beam HID head-lamps were developed by Hella and Bosch and launched on the E32 BMW 7-Series in 1992. HID lamps of course produce light as a result of an electric arc that is struck in a metal halide gas containing xenon; the addition of xenon reducing the warm-up time compared to an argon-only gas mix-ture that would otherwise take several min-utes to reach full output.

The spectrum of light emitted from an HID lamp contains a distinct blue peak, which compares with relatively little blue content from an incandescent lamp. Thus, the correlated colour temperature (CCT) of an HID lamp is much higher (4000K to 5000K versus 2800K to 3200K for tung-sten or tungsten halogen).

The presence of a blue peak from the HID

lamp and the use of a condenser lens in a projector headlamp can result in a charac-teristic blue glint (colour fringe) at certain angles of view, caused by chromatic aberra-tion from the lens.

A byproduct of the higher CCT and blue-rich output of HID lamps is their higher perceived brightness as a result of the semi-dark-adapted state of the human vision sys-tem. The scotopic response of the eye favours blue light over longer wavelengths. HID lamps typically exceed a luminous effi-cacy of 100 lumens per Watt, an order of magnitude higher than an incandescent lamp.

THE ARRIVAL OF LED (2006) The new millennia saw the start of the next automotive lighting revolution: the light emitting diode (or LED).

The first series production car with LED-based dipped-beam headlamps was the 2006 Lexus LS (courtesy of Koito), with the 2007 Audi R8 V10 featuring the first all-LED headlamp (manufactured by Automo-tive Lighting, the joint venture between Magneti-Marelli and Bosch).

An LED of course produces light by elec-troluminescence, a phenomenon discov-ered in 1907. The recombination of elec-trons and electron holes in the semiconductor produces photons of light, with the bandgap of the semiconductor determining the wavelength (colour) of the light emitted.

Much as elsewhere within the lighting world, LEDs provide a unique set of advan-tages over all other lighting technologies in automotive applications.

They are directional, compact, efficient (>100 lumens per Watt) and long lasting (if driven and thermally managed correctly) and present the automotive lighting designer with new opportunities for inno-vative design concepts based on their com-pact form factor.

An LED itself does not produce white light on its own. White light LEDs are in fact

p Figure 1. A schematic of lens-optic headlamp (all three original images by Duk)

p Figure 2. A schematic of reflector headlamp

p Figure 3. A schematic of projector headlamp

APRIL 2020 LIGHTING JOURNAL

www.theilp.org.uk

30

Automotive lighting

www.theilp.org.uk

APRIL 2020 LIGHTING JOURNAL 31

As part of the ‘The Exchange Public

Realm’ project in Aylesbury, GIFAS

in-ground units and a Charles

Endirect Ltd feeder pillar were

specified by the consultant. The

purpose of the in-ground units is

to supply single phase and 3-phase

electrical supplies for eventing

purposes as and when required.

When not in use the in-ground

units remain in the closed position

keeping The Exchange free of street

clutter.

For this project a Charles Endirect

electrical feeder pillar was also

required. Initially the housing

measured 2m high, but the

customer required something much

lower, and the file design shown

on the photo below measures just

1m high and 5m wide.

Ingenuity at work

This demonstrates the flexibility in

our design to suit the customers

requirements. The purpose of

the feeder pillar is to control and

supply electrical power to the street

lighting in this area, as well as the

in-ground units.

The housings are made of stainless

steel, and the safe opening and

closing of the lid is achieved using

gas pistons, meaning that the unit

is a maintenance free product.

GIFAS in-ground unitProject: The Exchange, Public Realm, Aylesbury

CharlesEndirect.com

+44 (0)1963 828 400 • info@CharlesEndirect.com

CEL_Lighting journal advertorial_Full page.indd 1CEL_Lighting journal advertorial_Full page.indd 1 12/03/2020 16:4912/03/2020 16:49

of course blue LEDs with a phosphor coat-ing through which the blue light is trans-mitted. The phosphor absorbs some of the blue light and fluorescence converts this to light of longer wavelengths.

The combination of blue LED light with green to red phosphor converted light cre-ates the white light output that we desire. The high brightness indium gallium nitride (InGaN) blue LED was invented by Shuji Nakamura at Nichia in 1995.

MATRIX AND PIXEL LEDS (2013) Bringing us almost to the present day, ‘matrix’ LED or ‘pixel’ LED headlamps employ an array of LEDs.

A forward-facing camera determines the presence of oncoming vehicles, or a vehicle travelling in the same direction in front, and dims individual LEDs in the array so as to avoid dazzling the other traffic. The headlamp beam is adjusted in real time so as to illuminate the road around or to the side of the other traffic, thereby maximising the illuminance for the driver of the vehicle.

Matrix or pixel LED headlamps operate on full main beam outside of urban envi-ronments and above urban speed limits, with the headlamps automatically dim-ming in bands or zones to avoid blinding other traffic.

Matrix LED headlamps were first intro-duced on the 2013 Audi A8, which featured 25 LEDs that could be individually switched on, switched off or dimmed. As an example of the current state-of-the-art, consider the 2018 model year Range Rover (L405) and Range Rover Sport (L494) model ranges. These were introduced with matrix LED and pixel LED headlamp technology.

The Land Rover matrix LED-type fea-tures 52 LEDs splitting the beam into verti-cal strips. The Land Rover pixel LED-type incorporates 142 LEDs, which allow the main beam pattern to be split vertically and horizontally. This provides for more pre-cise control of the beam with improved

illumination while at the same time mini-mising dazzle to other vehicles.

LASER ASSISTED HIGH BEAM (2014) Supplementing some high-end LED head-lights is laser-assisted high beam. This operates on a similar basis to how white light is generated from an LED. In this case, an intense blue laser ‘pumps’ a remote phosphor, which converts some of the blue light to longer green and red wavelengths.

As with white LEDs, the residual blue laser light is combined with the fluorescent green and red to create white light. Again as with LEDs, the laser light is also directional, except even more so.

The near-collimated white laser beam has a much further reach than standard LED beams, so is activated outside of urban environments above a threshold speed (typically 50mph) so as to illuminate the road further ahead of the vehicle (claims of useful illumination out to 500m are made). The first production vehicle equipped with laser-assisted high beams was the BMW i8 in 2014.

DIGITAL MATRIX DMD LED (2019) The very latest innovation in headlamp technology was announced in November last year by Audi.

The Audi E-Tron Sportback features ‘Digital Matrix LED’ headlamps that employ three LEDs and a DMD (Digital Micromirror Device) chip with one million ‘micromirrors’, each of which can be tilted up to 5,000 times per second.

The DMS chip is similar in concept to the DLP (Digital Light Processing) tech-nology developed by Texas Instruments and used in projectors. Each of the micro-mirrors on the 2D mirror array either allows the light from the LEDs to pass through to the headlamp lenses (and so illuminate the road), or to deflect into a beam dump, which absorbs the light. The resultant ‘high definition’ headlamp illu-mination pattern improves upon the beam steering and selective illumination

capabilities of matrix and pixel LED headlamps.

One of the more novel aspects to the Audi DMD technology is the ability to project a ‘carpet’ of light in front of the vehicle, so indicating to the driver which lane they should be in and their position within the lane. Selective illumination of hazards (for example pedestrians) is also highlighted as a benefit of DMD lighting, as is the ability to perform car-to-car communications to warn of hazards.

TESTING VEHICLE HEADLAMPS In order to comply with UN ECE regula-tions (the World Forum for Harmonization of Vehicle Regulations) for vehicle type approval, the illumination performance of a headlamp must be tested.

There are two basic methods used when measuring the photometric performance of vehicle headlamps. The first is to project the headlamp on to a screen and measure the illuminance (in lux) as a function of position within the beam.

The alternative option is to mount the headlamp on to a motorised, two-axis goni-ometer and measure the luminous inten-sity (in candelas) as a function of angle as the headlamp rotates and tilts with respect to a fixed light meter. I’ll refer to these methods as the ‘Projection Test Method’ and the ‘Goniometric Test Method’ respec-tively. Both require that the headlamp is measured at a distance of 25m for regula-tory compliance testing.

Why measure at 25m? The short answer is that this is the measurement distance specified in applicable ECE regulations. However, it is helpful to understand the physics of why 25m is chosen and not some other distance.

In the photometry of light sources, we refer to near-field and far-field measure-ments. In the photometric far-field, the light beam can be considered to be ‘fully formed’. The source itself behaves as a point light source, the luminous intensity (candela value) of the beam remains

APRIL 2020 LIGHTING JOURNAL

www.theilp.org.uk

32

Automotive lighting

Automotive lighting

www.theilp.org.uk

APRIL 2020 LIGHTING JOURNAL 33

COLLABORATE

INNOVATE

PREPARE

DELIVER

COLLABORATE

INNOVATE

PREPARE

DELIVER

COLLABORATE

INNOVATE

PREPARE

DELIVER

WORKINGWITH YOU

WORKINGWITH YOU

COLLABORATE

INNOVATE

PREPARE

DELIVER

WORKING WITH YOU

29177 V-Max advert v4.indd 1 07/10/2019 1:43 pm

Automotive lighting

constant and the illuminance (in lux) decreases in proportion to the inverse square of the distance travelled.

In the photometric near-field, the beam is not fully formed, and the size of the light source and photodetector input aperture both influence the value of illuminance and luminous intensity recorded. In addition, the beam illuminance does not follow the inverse squared relationship and the lumi-nous intensity is not a constant.

It may be perfectly reasonable to want to measure illuminance in the near-field in some instances (for example, you may wish to know the lux level up close to the lamp). But the problem with near-field measurements is that the readings are to some extent dependent upon the actual size of the photodetector and the intensity value will be lower than the true far-field reading. In other words, a near-field illu-minance (lux) measurement cannot be used to calculate the far-field luminous intensity (candela) value as the inverse squared relationship does not hold.

For simple, diffused wide-angle light sources the photometric far-field is typi-cally at a distance of five times the luminous aperture on the lamp. This would be the case for automobile side and marker lamps, with the ECE regulations requiring their measurement at 3m. For narrow-angle sources and those with complex beam structure (such as vehicle headlamps), the far-field distance is much greater. To ensure that headlamps are correctly measured in the far-field, ECE and other regulations stipulate a measurement distance of 25m.

While in reality the true far-field distance for a given headlamp may be less than 25m, this distance is specified to ensure that all conformity measurements are made in the far-field irrespective of the design of the headlamp optics.

GONIOPHOTOMETRIC TEST METHOD Goniophotometers intended for testing vehicle headlamps need to be very precise, with the beam sampled every 0.1°. To avoid undesirably long measurement times, high-end automotive goniometers commonly measure ‘on-the-fly’, meaning they don’t stop and measure at each angle.

Instead, the photometer records the inten-sity as the headlamp is in motion. As a conse-quence, goniometers designed for testing automotive lighting tend to be quite expen-sive. If you are prepared to live with a longer measurement time, or perform lower resolu-tion scans for indicative testing, ‘stop-and-go’-type goniometers are a lower cost option.

In either case, the photometer must be placed at a distance of 25m from the head-lamp mounted on the goniometer platform. The reason for this is to ensure that the

headlamp beam is sampled in the photomet-ric far-field, in which the beam can be con-sidered to be ‘fully formed’.

The complex beam shape of headlamps would yield significant measurement errors if sampled closer to the source, in the near-field. Generally, the goniometer control soft-ware will perform a scan, and then automati-cally analyse the beam shape and intensity values so as to determine whether the head-lamp satisfies the requirements of the rele-vant ECE regulation.

PROJECTION TEST METHOD Provided that you have a sufficiently large dark room as your lighting laboratory, the projected illuminance using imaging pho-tometer method of testing holds many advantages.

This technique employs an imaging pho-tometer to record a 2D illuminance (lux) pattern of the headlamp beam as it is pro-jected on to a vertical surface (wall or screen) at a distance of 25m.

The lux versus XY position data-set can be mathematically transformed into an inten-sity versus angle data-set, often performed automatically though instrument software. The projection method is the method cited in regulations.

The simplest approach is to manually position an illuminance photometer (lux meter) on the projection wall facing the headlamp and record the beam illuminance as a function of position. However, this is a very time-consuming and laborious process and requires manual computation of the ECE test point compliance matrix. Manual illuminance measurement of headlamp beams using lux meters is not considered practical because of the time required to sample all of the required beam spots.

The more recent development is to replace the simple lux meter with an imaging photometer. The imaging photometer is used to take a photometrically-calibrated, high-dynamic-range digital picture of the beam pattern and automatically align the required test point matrix to the headlamp HV point (datum).

The ECE test point compliance matrix will then be automatically generated based upon the best possible alignment of the test points within the beam that the software can find. Compared with goniometric measure-ments, the projection test method using an imaging photometer has the following attributes:

• It is faster, measurements taking sec-onds to perform• It is simpler; the headlamp can remain mounted on the vehicle (assuming that the laboratory is large enough that the car can be driven into it)

• It is less expensive, with the typical cost of an imaging photometer being a fraction of the cost of an automotive goniophotometer• The imaging photometer can also be deployed for other kinds of testing, for exam-ple analysing the lit area uniformity of side or marker lamps or of DRLs (daytime run-ning lights)

On the downside, you do need a 25m-long dark room to perform projected beam tests. However, indicative tests at shorter working distances (for example 10m) have been shown to correlate closely with 25m measurements.

In other words, a headlamp that passes the ECE test points at 10m is very likely also to pass at 25m, and vice versa. Thus, even if space is limited, projection headlamp test-ing will still serve as a useful developmental and benchmarking tool for vehicle manufac-turers and for those manufacturers of vehi-cle headlights.

IN CONCLUSION Headlamp technology continues to develop, with the very latest matrix LED, laser-as-sisted high beams and digital matrix LED designs providing safer, less stressful illumi-nation for the driver, while at the same time avoiding dazzle, glare and discomfort to other road users.

Yet the basic metrology of headlamps remains the same, regulations requiring that one measure the luminous intensity of the headlamp as a function of angle (or, equally, one can measure illuminance as a function of XY position within the beam).

There are two basic approaches to the testing of vehicle headlamps, one using a goniophotometer, the other being to project (shine) the headlamp beam on to a vertical surface or screen and record the 2D illumi-nance pattern using an imaging photometer.

The goniometric technique is probably the one that most engineers will be more familiar with, the headlamp being mounted on a two-axis motorised rotary stage (the ‘goniometer’) which is viewed by a fixed light meter (the ‘photometer’) that records the intensity for each angle of view as the lamp rotates.

The more recent, and significantly lower cost, approach is to employ an imaging pho-tometer to record the beam pattern as it illu-minates the screen. This projection method of testing is not only cheaper, it is faster, but you do need a large dark room to project the headlamp down (headlamps being tested at 25m, as per regulations).

APRIL 2020 LIGHTING JOURNAL

www.theilp.org.uk

34

Robert Yeo is a qualified physicist and co-founder and director of photonics equipment distributor and photometry trainer Pro-Lite Technology

Automotive lighting

www.theilp.org.uk

APRIL 2020 LIGHTING JOURNAL 35