Krzysztof ZAREMBA

1

, Andrzej PAWLAK

2

Bialystok Technical University (1), Central Institute for Labour Protection - National Research Institute (2)

Parameters of model luminaire with high power LED diodes

Abstract. Specially manufactured reflectors with limited height were applied in the designed luminaire model. Thanks to that approach, a reduced
luminous flux distribution was achieved along with more appropriate luminous intensity curve, with simultaneous high efficiency reaching 92%.
Application of 30 white and 5 red diodes in the same luminaire provided the resulting optical radiation with the colour temperature of 3702K and
colour rendering index of 81.

Streszczenie. W modelowej oprawie zastosowano specjalnie zaprojektowane i wykonane odbłyśniki o niewielkiej wysokości. Dzięki temu
osiągnięto ograniczony rozsył strumienia świetlnego, krzywą światłości przystosowaną do równomiernego oświetlenia powierzchni roboczej i
jednocześnie wysoką sprawność oprawy wynoszącą aż 92%. Jednoczesne zastosowanie 30 diod barwy białej i 5 diod barwy czerwonej
spowodowało uzyskanie światła o temperaturze barwowej 3702K i wskaźniku oddawania barw 81. (Parametry modelowej oprawy z diodami LED
o dużej mocy)

Keywords: high power LED diode, general lighting, colour rendering index, colour temperature.
Słowa kluczowe: diody LED o dużej mocy, oświetlenie ogólne, wskaźnik oddawania barw, temperatura barwowa.

1. Introduction

In the case of high power LED diodes, it is possible to

assume that their luminous intensity distribution curves are
very similar with the cosine luminous intensity distribution
curve of the Lambertian surface. They should not be
however applied directly in general lighting luminaires [1].
Obtaining the constant illuminance value on the working
surface, which is directly connected with the highest
possible degree of the lighting uniformity, requires
application of a luminaire with the special luminous intensity
curve [2]. Such luminous intensity distribution can not be
obtained in a typical general lighting reflector luminaire with
the

α

ob

cut-off angle ranging from 55° to 65°. A uniform

lighting luminaire must be constructed in such a way, that
the central part of the luminous flux emitted by the diode is
dissipated by an additional reflector module or a lens. That
additionally complicates the construction of such a luminaire
and increases its final price.

Fig. 1. Geometric layout of the designed luminaire with alternated
luminous intensity curve

In accordance with the standard EN 12464-1:2004 –
Lighting of Indoor Work Places, the uniformity of target area
lighting (calculated as a quotient of the minimum and
average value of the illuminance observed on the given
target area) for continuous work should be estimated at
least 0.7. That is exactly the reason why we decided to
examine the lighting uniformity for luminaires, where the
central part of the luminous intensity distribution curve,
characterized with the

α

ox

angle (Fig. 1), is identical with the

luminous intensity distribution curve of the light source and
then has a constant value of the illuminance. Under the
assumption that the reflection coefficient for the reflector
module was defined at 0.82, the minimal normative lighting

uniformity ratio, defined at 0.7 is achieved for the

α

ob

cut-off

angles lower or equal to 60,4°. Thus, it is possible to design
a single-reflector luminaire module, providing lighting
uniformity ratio better than 0.7. A luminaire with the direct

α

ob

cut-off angle estimated at 60° can theoretically provide

the lighting uniformity ratio of 0.713, while the said ratio
increases to 0,852 with the

α

ob

cut-off angle of 55°.

However, predicting a practical decrease in the estimated
lighting uniformity ratio, the luminaire with the

α

ob

cut-off

angle of 55° was selected for construction. It is
characterized with the

α

ox

angle estimated at 37,6° and the

luminous intensity I

ox

= 252,2 cd/klm.

2. Construction of the luminaire model

The shape of the luminaire reflector was estimated for

the I

o

α

luminous intensity distribution curve using the flux

method (Fig. 2) [2]. The used method for estimation of the
shape of the reflector module does not compensate for the
real size of the light source. Thus, the applicability analysis
for the derived profiles was conducted in systems with the
finite size of the light source. Its results confirm the
applicability of reflectors, designed for the light source with
negligible size, in the lighting systems with the real light
source of finite size. Visible alternations in the luminous
intensity distribution curves are noticeable only when the
size proportion between the light source and the reflector
module reaches the ratio 1:5.

Fig. 2. Calculated shape for the rotationally symmetrical reflector
with the direct irradiation angle

α

ob

equal to 55°

Construction of the model of the examined luminaire,

equipped with the LED diodes type LXHL-LW3C with the
diameter of 5.6 mm, featured a reflector module with the
output diameter of 60 mm (Fig. 2). The LXHL-LW3C diode,
emitting white light, is characterized with the standard
operating current of 700 mA with the operating voltage of
3.7 V, thus dissipating 2.6 W. Under such operating
conditions, the diode emits a luminous flux of 66 lm (with the
minimum value of 60 lm). The diode can also be powered

PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 83 NR 5/2007

50

with the current of 1000 mA and voltage of 3.9 V,
dissipating 3.9 W and emitting a luminous flux of 80 lm. The
LXHL-LW3C diode, just like most of the currently
manufactured devices of his type, emits the daylight type of
visible radiation. Such diodes produce significant amount of
blue and green - yellow light, thus their colour temperature
is quite high, since it is typically estimated at 5500 K. The
manufacturer allows for a very wide colour temperature
distribution, ranging between 4500 K and 10000 K.
Additionally, the manufacturer does not also provide any
estimation of the colour rendering index, stating only that it
is good. Based on the catalogue – provided spectral
distribution for the LXHL-LW3C diode, its colour
temperature was estimated at 6177 K while the overall
colour rendering index was established at 67. The colour
temperature of 6177 K is too high for the majority of
applications, however, taking into the account the
admissible wide colour temperature distribution provided for
by the manufacturer, it becomes obvious that the target
luminaire should be equipped with a mechanism allowing to
adjust the colour temperature at will [3]. The impact of
applying red LXHL-LD3C diode in order to alter the colour
temperature of white diodes, was examined in detail. Fig. 3
present the relation between the colour temperature T

c

and

colour rendering index (CRI) for the resulting mixture of
optical radiation from a single red diode and a varied portion
n

W

of radiation originating from white diodes. Increase in the

number n

W

of white diodes is closely correlated with the

increase in the colour temperature and colour rendering
index of the resulting light. The designed luminaire
contained thus a single, red LXHL-LD3C diode for every 5
white emitting LXHL-LW3C diodes. The optical radiation
emitted by such mixture of diodes is characterized by the
overall colour temperature of 2709 K and the colour
rendering index estimated at 67, which is equal to the ones
of the white LED diodes alone. Selection of such proportion
of diode types was also advantageous in terms of power
budget, since most of the available power supply sources
are characterized with the maximum output voltage of 24 V,
which corresponds to 6 diodes connected in series in a
single branch.

Fig. 3. Calculated colour temperature T

c

and colour rendering index

CRI for radiation mix originating from a single red and n

W

white

diodes

Utilization of such diode proportion can also produce the

output white light with the colour temperature higher than
2709 K and colour rendering index reaching 79 providing

that a regulation circuit is employed in the luminaire,
allowing to decrease the relative quantity of the luminous
flux

Φ

Rr

originating from red diodes. (Fig. 4).

Fig. 4. Calculated colour temperature T

c

and colour rendering index

CRI for radiation mix originating from a single red diode with the
relative quantity of luminous flux

Φ

Rr

and 5 white diodes

a

b

Photo 1. Overall construction of the model luminaire with 36 LED
diodes: (a) – with no attached reflectors, (b) – with attached
reflectors

The target model luminaire was to contain 30 LXHL-

LW3C type LED white light diodes and 6 LXHL-LD3C type
LED red light diodes (Photo 1 – red diodes are located on
the diagonal of the luminaire). Each diode was mounted on
a separate, electrically insulated radiator (Photo 1.a). In the
result of the conducted analysis using the Philips
Xitanium™ 80W/3.15-24V/3150mA power supply system,
assumptions concerning the system power source were
modified. Instead of original two parallel power supplies,
only one was applied. It was decided that at the research

PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 83 NR 5/2007

51

stage, assuring stable and safe operating conditions for the
diodes is much more important than obtaining their
maximum lighting parameter ratings. Closer analysis of the
power supply system indicated that under failure conditions
(e.g. disconnecting one of the parallel branches), the whole
current provided by the power supply would flow through the
remaining branches, meaning that with a 3 branch design,
most of the diodes would be damaged. Thus, a single power
supply system was connected to 6 parallel diode branches.
It must be noted, that while the average current value in
branches was estimated at 0,537 A, the maximum observed
difference was measured at 0,09 A, providing the relative
value of approximately 20%.

2. Conclusions
The measured luminous intensity curves for 36 LED
diodes and the model luminaire were depicted in Fig. 5. The
constructed luminaire is characterized by low luminance
value and limited luminous flux distribution. The obtained
luminous intensity distribution curve is characterized by
significantly lower maximum luminous intensity value when
compared with the theoretically estimated one, which
originates from the manufacturing technology for model
reflectors, the surface of which has slightly diffusive
properties. Despite this fact, the model luminaire achieves
significantly improved lighting uniformity when compared
with the application of individual diodes alone. The luminous
flux of all 36 diodes was measured at 1954 lm (on average,
54,3 lm/diode). The model luminaire has very high
efficiency, reaching 92%, which is 15÷20% higher when
compared with other luminaires with similar parameters but
different light sources. The designed luminaire is relatively
low, which is a design target on its own. It could be further
reduced providing that lower profile radiators are used. The
measured colour temperature of the designed luminaire was
estimated at 3702K, which is higher than the theoretically
expected value. The measured colour rendering index was
estimated at 81. It is therefore conclusive that the designed

model luminaire with high power LED diodes may be
successfully applied for general lighting purposes.

Fig. 5. Luminous intensity curve: 36 LED diodes (black dotted

line), luminaire model (black solid line) and calculated theoretically
(grey solid line)

REFERENCES

[1] U c h i d a Y . ; T a g u c h i T . : Lighting theory and luminous

characteristics of white light-emitting diodes, Optical
Engineering. Dec. 2005; 44(12): 124003-1-9

[2] Z a r e m b a K . : A Synthetic Method of Designing Rotational

Reflectors, 13

th

European Simulation Multiconference 1999,

Modelling and Simulation: A Tool for the Next Millenium,
ESM’99, Warsaw, June 1-4 1999, Poland, Volume II, p.307-309

[3] B r o w n D . ; N i c o l D . ; F e r g u s o n I . : Investigation of the

spectral properties of LED-based MR16 bulbs for general
illumination, Optical Engineering. Nov. 2005; 44(11): 111310-1-
4

Autors: Krzysztof Zaremba, Ph.D. (E.Eng), Białystok Technical
University, Chair of Optical Radiation, ul. Wiejska 45D, 15-351
Białystok, Poland; phone [phone], [e-mail];
Andrzej Pawlak, M.Sc. (E.Eng.), Central Institute for Labour
Protection - National Research Institute, ul. Czerniakowska 16, 00-
701 Warszawa, Poland; phone [phone], fax [phone]-
3695, [e-mail];

PRZEGLĄD ELEKTROTECHNICZNY, ISSN 0033-2097, R. 83 NR 5/2007

52