[0001] The invention relates to an electrodeless low-pressure mercury vapour discharge lamp
with a discharge vessel which encloses a discharge space provided with a filling of
mercury and rare gas in a gastight manner, which discharge vessel comprises a light-transmitting
enveloping portion and in addition a recessed portion in which a coil for generating
a high-frequency magnetic field is enclosed, and which discharge vessel is provided
with a luminescent layer at least on a portion of a surface facing towards the discharge
space.
[0002] Such a lamp is known from EP 0.162.504 B1. The lamp is operated in that the coil
is connected to a high-frequency electric supply source. The magnetic field generated
by the coil induces an electric discharge in the discharge space. The coil in addition
generates a comparatively strong electric field in the discharge space as a result
of potential differences across this coil. The electric field strength may be very
great especially near the recessed portion of the discharge vessel. In addition, comparatively
high temperatures prevail in the wall of the discharge vessel. The temperature of
the recessed portion may even assume values above 200° C.
[0003] Under these circumstances, luminescent materials present in the luminescent layer
may react with particles from the discharge space which collide with these materials.
Depending on the application of the lamp, this may give rise to disadvantages.
[0004] Usually, several kinds of luminescent materials are present in the luminescent layer,
and the luminous efficacies of these materials are affected by the reactions with
the particles to different degrees. The result of this is that the colour point of
the light generated by the luminescent layer shows a shift during lamp life. This
is no disadvantage when a single lamp is used because this process takes place gradually.
Even a comparatively great difference with the colour point at the beginning of lamp
life is not observable in that case. Clear differences in colour point, however, are
observable in applications where known lamps of mutually differing ages are positioned
in one and the same space. If one of the lamps is defective, it is necessary to replace
not only the defective lamp, but also the other lamps in said space in order to avoid
colour point differences under these circumstances, which is expensive. It was found
that differences between colour points are observable when at least one of the colour
points is present outside the Von Kries transformed MacAdam ellipse of another colour
point (see: New Insights in Chromaticity and Tolerance Areas of Fluorescent Lamps,
J.J. Opstelten and G. Rinzema, Journal of the IES, Winter 1987, pp. 117-127). This
is the case if there is no ellipse half the size of the transformed MacAdam ellipse
which comprises all colour points.
[0005] It was found to be favourable to operate the lamp of the kind mentioned in the opening
paragraph by means of a pulsatory supply. The lumen output of the lamp may be adjusted
between, for example, 10% and 100% of its rated lumen output in this mode of operation
in that the ratio of the pulse duration to the time interval between the pulses is
varied. In this mode of operation, however, comparatively high voltages are required
for re-igniting the lamp at the start of each pulse. Electric fields will then occur,
especially near the recessed portion, which are even stronger than those during nominal
operation. It was found that mercury is bound to material in the luminescent layer
under these circumstances, which mercury is no longer available for lamp operation.
A comparatively large quantity of mercury is necessary if a sufficiently long lamp
life is to be guaranteed in spite of this. This is bad for the environment in the
case of inexpert disposal at the end of lamp life.
[0006] The invention has for its object to provide a measure in an electrodeless low-pressure
mercury vapour discharge lamp of the kind described in the opening paragraph which
counteracts the interaction between the luminescent layer and particles from the discharge
space.
[0007] The electrodeless low-pressure mercury vapour discharge lamp according to the invention
is for this purpose characterized in that at least a portion of the luminescent layer
bears a protective layer of aluminium oxide particles with a coating weight of 10
to 500 µg/cm². It was surprisingly found that this measure counteracts the interaction
between the luminescent layer and particles from the discharge space in spite of the
comparatively strong electric fields and the comparatively high temperatures prevalent
in the lamp.
[0008] It is noted that US 4,639,637 discloses a low-pressure mercury vapour discharge lamp
with a tubular discharge vessel which is provided with a luminescent layer coated
with an aluminium oxide layer at a surface facing towards the discharge space. In
contrast to the lamp according to the invention, however, the lamp is ignited and
the discharge is maintained by a pair of electrodes arranged in the discharge vessel.
The electric fields occurring near the luminescent layer are weak in such lamps. The
temperatures prevalent in the discharge vessel wall are comparatively low.
[0009] A first attractive embodiment of the lamp according to the invention is characterized
in that the surface of the recessed portion facing towards the discharge space is
provided with a luminescent layer which bears a protective layer having a coating
weight of 100 to 500 µg/cm². This embodiment of the lamp according to the invention
has the advantage that only little mercury is bound by the luminescent layer also
during operation by means of a pulsatory supply. A luminescent layer on the enveloping
portion of the discharge vessel is, for example, absent. The lamp may then be used
as a UV radiator.
[0010] A second attractive embodiment is characterized in that the surface of the enveloping
portion facing towards the discharge space is provided with a luminescent layer bearing
a protective layer having a coating weight of 10 to 50 µg/cm². Lamps according to
this embodiment of the invention have the advantage that the colour point shift is
so small that no difference is perceivable between the colour points of lamps of the
same embodiment but differing in age. The recessed portion may be provided, for example,
with a reflecting layer. In a modification, the recessed portion is also provided
with a luminescent layer. It was found in this modification of the second embodiment
that the colour point shift is also very small without a protective layer on the luminescent
layer of the recessed portion. It is favourable, however, if the luminescent layer
on the recessed portion bears a protective layer corresponding to that of the first
embodiment in this modification. The lamp is then also suitable for pulsatory operation.
[0011] The protective layer may be readily provided in the form of a suspension of aluminium
oxide powder, after which the layer is sintered,
i.e. is heated for some time in order to drive out auxiliary substances such as binders
from the layer. The suspension is applied to the surface, for example, by spraying.
For providing the protective layer on the recessed portion, it may suffice to immerse
the recessed portion in the suspension. Alternatively, the layer may be provided,
for example, through electrostatic coating.
[0012] It is found in practice that lamp characteristics of electrodeless lamps, such as
the lumen output or the change therein during lamp life, depend on the production
location of the lamp.
[0013] It was surprisingly found with the lamp according to the invention that the change
in lumen output depends much less on the production location than with the lamp not
according to the invention. The measure according to the invention thus improves the
reproducibility of said lamp characteristic in the case of manufacture in different
locations.
[0014] Embodiments of the electrodeless low-pressure mercury vapour discharge lamp according
to the invention are explained in more detail with reference to the drawings, in which:
Figs. 1 and 2 show a first and a second embodiment of the lamp, respectively, partly
in side elevation and partly in longitudinal sectional view. A supply device is also
shown diagrammatically.
Figs. 3 and 4 show the colour point shift of lamps according to the second embodiment
and of lamps not according to this embodiment, respectively.
[0015] The embodiment of the electrodeless low-pressure mercury vapour discharge lamp according
to the invention shown in Fig. 1 is provided with a pearshaped, gastight discharge
vessel 1 which encloses a discharge space 2. The discharge vessel 1 has a light-transmitting
enveloping portion 3A and a tubular, recessed portion 3B. The discharge vessel 1 further
has a flanged portion 3C which connects the recessed portion 3B to the enveloping
portion 3A. The discharge space 2 is provided with a filling of mercury and a rare
gas, here argon. A coil 10 for generating a high-frequency magnetic field is accommodated
in the recessed portion 3B. The coil 10 has a length of 25 mm and is provided with
a first and a second winding 11, 12, each of 15 turns around a core 13 of soft-magnetic
material. In an alternative embodiment, the coil may have, for example, an air core,
or a core of ceramic material. The first winding 11 is connected to a high-frequency
supply source 20 by means of current supply conductors 14, 15 in order to operate
the lamp. The supply source 20 here has a frequency of 2.65 MHz. The second winding
12 has an end which is connected to current supply conductor 15 and also has a further,
free end. The supply source 20 is connected to poles P and N of the mains and is earthed
at pole M. The recessed portion 3B of the discharge vessel 1 is provided with a luminescent
layer 5B at a surface 4 facing towards the discharge space 2, this layer comprising
red-luminescing yttrium oxide activated by trivalent europium (YOX) and green-luminescing
cerium-magnesium aluminate activated by trivalent terbium (CAT). The luminescent layer
5B has a coating weight of 8 mg/cm². A surface 4 of the enveloping portion 3A facing
towards the discharge space 2 is provided with a luminescent layer 5A which comprises
blue-luminescing barium-magnesium aluminate activated by bivalent europium (BAM) in
addition to the luminescent materials YOX and CAT. This luminescent layer 5A has a
coating weight of 3.5 mg/cm².
[0016] The luminescent layer 5b on the surface 4 of the recessed portion 3B facing towards
the discharge space 2 bears a protective layer 6B of aluminium oxide particles. The
protective layer 6B was obtained in this case in that the recessed portion 3B was
immersed in a suspension of Alon-C of the Degussa company, after which the layer remaining
on the recessed portion 3B was dried and then sintered. The coating weight depends
on the concentration of the aluminium oxide powder present in the suspension. Aluminium
oxide powder of the Alon-C type comprises aluminium oxide particles with a size of
approximately 0.01 to 0.04 µm, and has a specific area of approximately 100 m²/g.
[0017] Lamps corresponding to the embodiment shown in Fig. 1 and having a coating weight
of 170 µg/cm², of 250 µg/cm², and of 300 µgcm² were manufactured, two of each Kind.
The coating weights of the lamps accordingly lay between the limits of 100 and 500
µg/cm². Two reference lamps were also manufactured in which a protective layer of
aluminium oxide particles on the luminescent layer of the recessed portion was absent.
The six lamps according to the invention and the two reference lamps were subjected
to an endurance test of approximately 170 hours in order to assess the influence of
strong electric fields of the kind which may arise during re-ignition. Each lamp was
provided with a coil whose windings were interrupted for this purpose. The windings
jointly extend over a length of 30.5 mm around the coil core. The interruption in
the coil creates two coil parts each of 7.5 turns, approximately 5 mm spaced apart.
A voltage of 700 V was applied to this coil during the endurance test. Since the coil
is interrupted, it does not generate a magnetic field, so that no arc discharge is
generated in the discharge vessel. The lamp accordingly remains permanently in a condition
corresponding to that during re-ignition in reduced operation during the endurance
test. In reduced operation of the lamp, the comparatively high re-ignition voltages
occur during a fraction of each cycle only. This is because, on the one hand, a voltage
is applied to the coil during a portion of each cycle only in this mode of operation,
and on the other hand because the voltage across the coil drops quickly after lamp
re-ignition. It is assumed that the above endurance test of 170 hours is comparable
to approximately 5000 hours of pulsatory operation with a lamp power which is reduced
to 15%.
[0018] A strong blackening of the luminescent layer between the coil parts was observed
in the reference lamps after this endurance test. In the lamps according to the invention
having a coating weight of 170 and 250 µg/cm², the luminescent layer was much less
strongly discolored. A discoloration was even substantially absent in the lamps according
to the invention having a coating weight of 300 µg/cm².
[0019] An electrodeless lamp was manufactured for comparison whose recessed portion was
provided with a luminescent layer coated with a silicon oxide layer and one whose
luminescent layer on the recessed portion was coated with an yttrium oxide layer.
The lamps were also subjected to an endurance test. After no more than 16 hours of
operation at a voltage of 600 V, the lamp with the yttrium oxide layer exhibited a
pale brown band between the coil parts. The section of the recessed portion lying
between the coil parts in the lamp having the silicon oxide layer remained unchanged
in colour. The recessed portion, however, did show a grey discoloration on either
side thereof.
[0020] It was also investigated, after the endurance test had been completed, to what extent
mercury was bound to the luminescent material in the lamps according to the invention
and in the reference lamps. For that purpose, the section of the recessed portion
situated between the coil parts was subjected to a wet chemical analysis.
[0021] The average quantity of mercury bound to the wall in the reference lamps (REF) and
the respective average quantities in the lamps having a coating weight of 170, 250
and 300 µg/cm² are listed in the following Table.
coating weight (µg/cm²) |
mHg (µg) |
REF |
62 |
170 |
33 |
250 |
30 |
300 |
18 |
[0022] It is evident from the Table that the quantity of bound mercury in the lamps according
to the invention with a protective layer having a coating weight of 170 and 250 µg/cm²
is approximately half that of the reference lamps. The quantity of bound mercury is
even about one third in the lamp according to the invention whose protective layer
has a coating weight of 300 µg/cm².
[0023] The drift in the lumen output between 1 and 100 hours of operation was measured for
10 lamps according to the invention whose luminescent layers on the recessed portions
were coated with a protective layer of aluminium oxide particles, and for 10 lamps
not according to the invention. The lamps not according to the invention correspond
to those according to the invention except for the absence of a protective layer.
Five of the lamps according to the invention were manufactured in a first production
location (A), and five in a second production location (B). Similarly, five of the
lamps not according to the invention were manufactured in the first production location
(A) and five in the second production location (B).
[0024] The drift in the lumen output in the period from 1 to 100 hours of operation is shown
in the following Table.
|
Lamp not according to the invention |
Lamp according to the invention |
A |
98.2 % |
98.3 % |
B |
93.7% |
95.7 % |
difference |
4.5 % |
2.6 % |
[0025] Manufacture in different production locations leads to a difference in the lumen
output drift of 4.5% in the lamps not according to the invention. This difference
is no more than 2.6% in the lamps according to the invention.
[0026] In Fig. 2, parts corresponding to those of Fig. 1 have reference numerals which are
100 higher. In the embodiment of the electrodeless low-pressure mercury vapour discharge
lamp according to the invention shown here, the surface 104 of the enveloping portion
103A facing the discharge space 102 is provided with a luminescent layer 105A which
bears a protective layer 106A with a coating weight of 28 µg/cm². The coating weight
accordingly lies within the limits of 10 and 50 µg/cm².
[0027] In the embodiment shown, the flanged portion 103C has neither a luminescent nor a
protective layer. In a modification, this portion 103C also has a luminescent layer,
possibly be provided with a protective layer.
[0028] Six lamps were manufactured corresponding to the embodiment of the invention described
with reference to Fig. 2.
In addition, five reference lamps were manufactured without a protective layer of
aluminium oxide particles on the luminescent layer of the enveloping portion.
[0029] The above 11 lamps were subjected to an endurance test. The lamps were switched off
for half an hour after 2.5 hours of operation each time during this. The x- and y-coordinates
of the colour point were measured for the said lamps both at the start and after 2000
hours of operation. Fig. 3 shows the colour points at the beginning (open dots) and
after completion of the endurance test (closed dots) for the six lamps according to
the invention. Fig. 4 shows the colour points of the five lamps which do not have
a protective layer of aluminium oxide particles on the enveloping portion. In these
Figures, the colour points at the beginning of lamp life are indicated with open dots
and those at the end of lamp life with closed dots. Only four colour points are visible
in Fig. 4 because two colour points had equal coordinates both at the beginning and
at the end of the endurance test.
[0030] It is evident from Fig. 3 that the colour points of the lamps measured at the start
and after completion of the endurance test lie jointly within an ellipse having half
the size of a Von Kries transformed MacAdam ellipse. The average deviation of the
colour coordinates at the end of the endurance test compared with those at the beginning
is 0.001 and 0.003, respectively.
[0031] Fig. 4 shows that the colour points of lamps not having a protective layer on the
enveloping portion differ considerably after completion of the endurance test from
those at the beginning of the endurance test. The average deviation of the x- and
y-coordinates is 0.017 and 0.010, respectively.
[0032] It is clear from the above that lamps corresponding to the embodiment of the invention
described with reference to Fig. 2 show only a very small drift in their colour point
compared with lamps not corresponding to said embodiment. The average initial value
of the colour point (x = 375, y = 383) differs slightly from the average initial value
of the colour point (x = 381, y = 379) of the lamp not corresponding to said embodiment.
This difference in the initial value of the colour point may easily be eliminated
by means of a minor change in the composition of the luminescent layer.