Patent Grant
7323820
The invention is based on a metal halide lamp having a ceramic discharge vessel, the inner contour of which is convex in form with rounded ends, wherein the discharge vessel contains a fill which comprises starting gas, preferably as noble gas, mercury and metal halides, the metal halides comprising two groups, namely the first group made up of the emitters and the second group made up of the wetters. These are in particular high-pressure discharge lamps with ceramic discharge vessel for a neutral-white luminous color.
U.S. Pat. No. 6,218,789 has already disclosed a metal halide lamp. In that document, a halide of Yb is used to generate molecular radiation. The discharge vessel consists of quartz glass.
U.S. Pat. No. 6,483,241 has disclosed a mercury-free metal halide lamp which uses Mg iodide as fill in a ceramic discharge vessel.
It is an object of the present invention to reduce the color scatter in metal halide lamps with a convex geometry of the discharge vessel, in particular with fills used for neutral-white luminous colors.
This object is achieved by the following features: the second group at least comprises one of the halides of Mg and Yb, with the proportion of these constituents of the second group amounting to at least 15 mol %, with the option for Ca halide to be an additional constituent of the second group, in which case the proportion of the entire second group amounts to at most 55 mol % of the metal halides.
Particularly advantageous configurations are given in the dependent claims.
The color scatter of metal halide lamps has long been the focus of attempts to improve quality. This problem in itself appeared to have already been solved, since a corresponding fill composition is known for a cylindrical geometry of the discharge vessel. In this case, certain ratios for the surface area also have to be taken into account.
Surprisingly, however, it has emerged that these established approaches aimed at finding a solution fail if, instead of a cylindrical geometry, the more isothermal convex geometry is used. This is to be understood as meaning a discharge vessel with rounded corners, which has either a straight center part or an elliptically shaped volume. The rounding may be circular, elliptical or of some other shape. This problem is particularly pronounced when using fills for a neutral-white luminous color, i.e. for a color temperature from approximately 4000 to 4900 K.
According to the invention, therefore, the inner contour is convex in form with rounded ends, while the discharge vessel contains a fill which comprises starting gas, preferably as noble gas, mercury and metal halides, the metal halides comprising two groups, namely the first group made up of the emitters and the second group made up of the wetters, and wherein the second group at least comprises one of the halides of Mg and Yb, with the proportion of these constituents of the second group amounting to at least 15 mol %, with the option for Ca halide to be an additional constituent of the second group, in which case the proportion of the entire second group amounts to at most 55 mol % of the metal halides.
It is particularly preferable to add halide of Yb, in particular in a proportion of from 10 to 60 mol %, preferably 15 to 45 mol %. In particular, a fraction of the Yb, preferably up to 50%, may be replaced by halides of Mg. A suitable halogen in this context is preferably iodine, but bromine may also be suitable, in particular as a fraction which replaces iodine, preferably up to 30%.
Operation may be implemented at electronic ballasts or conventional ballasts.
Metal halide lamps with convex ceramic burners, in particular to set a neutral-white luminous color (NDL, typically 4000 to 4900 K), require a relatively high proportion of RE iodide in the metal halide melt. RE here stands for rare earths. The term burner means discharge vessel.
Therefore, over the illumination time and service life of the lamp, there is an increase in the restarting peak voltage UI and the crest factor (UIs/UIrms), which can lead to critical lamp conditions and premature failure through extinction of the lamp.
In the case of cylindrical discharge vessels, this problem is normally remedied by the addition of CaI2, which is known per se. However, it has emerged that the wetting properties of the metal halide melt changes significantly beyond typical CaI2 concentrations of at least 20 mol %, in particular 25 mol %, since in the operating state the wetting angle of the melt on the lamp components is increased.
In the case of lamps with high power densities, the altered fill wetting results in a relatively high individual scatter of the desired color temperature as a result of fluctuating extent of the fill wetting on the inner wall of the discharge vessel. In this context, the power density p is to be understood as meaning the power P of the lamp in W per unit area S in mm2, differentiated between the inner and outer power density pin=P/Sin and pout=P/Sout (where S respectively denotes the surface area on the inside (in) and outside (out) of the discharge vessel) and typical surface area ratios between the inner and outer surfaces eo_back in the electrode back space (eo_back: =total space or burner extent in the interior and exterior behind the electrode tip, including the capillary with regard to the neck region) to the total surface area of the discharge vessel (S inter_deo/Si_tot; So, back_deo/Si_tot), as is the case with convex lamps with hemispherical end shapes.
Typical ratios for both shapes are explained in Table 1 below:
On account of the different surface area ratios, which are substantially responsible for dividing the power transported by radiation transport and heat conduction between the inner wall and from the outer wall of the discharge vessel to the environment, a very homogenous temperature distribution is formed with convex discharge vessels.
For example, the ratio of outer surface area to inner surface area is typically from 1.6 to 2.0 when using a cylindrical geometry (in Table 1 it is 1.8), whereas when using a convex geometry this ratio is typically from 1.0 to 1.35 only (in Table 1 it is 1.16). The difference between comparable power stages is typically 50% (in Table 1 it is 55%). Furthermore, the ratio of the inner surface area which lies behind the electrode tips to the inner surface area which lies between the electrodes is 0.95 with a cylindrical geometry but just 0.7 in the case of a convex geometry, i.e. is 35% greater with a cylindrical geometry. The ratio of outer surface area Sback which lies behind the electrodes to the outer surface area Seo which lies between the electrodes is 1.78 in the case of a cylindrical geometry but only 0.77 in the case of a convex geometry, i.e. is 131% greater in the case of a cylindrical geometry.
The result of this is that under certain circumstances, if a defined wetting angle of the metal halide fill is exceeded, a boosted distribution of the fill into the interior of the burner occurs. This leads to increased individual scatter of the color temperature and consequently to a corresponding scatter in the electrical characteristic variables.
The individual scatter of the color temperature is now reduced by an altered fill composition, so as to produce a defined degree of fill wetting on the inner wall of the discharge vessel in the electrode back space. At the same time, the electrical lamp data, such as restarting peak and crest factor, are as a result comparable to fills with a high CaI2 fraction (low activity of the RE iodides).
A typical target value for the color temperature is, for example, 4000 to 4400 K. The novel fill reduces both the scatter in the color temperature, with only a slight deviation from the Planckian locus in the CIE diagram and with a low crest factor.
An acceptable variation range δ after 100 h illumination time for the color temperature Tn and the crest factor Cr is
δTn≦±75K, Cr=UIs/UIrms<1.9.
The addition of CaI2 in metal halide melts to set NDL color temperatures is typically 40-50 mol %, thereby reducing the activity of the trivalent RE iodides, which over the illumination time and service life leads to a reduction in the reaction rates of the RE halides with the lamp constituents and therefore limits the formation of free iodine. This in turn restricts the increase in the restarting peak voltage and the crest factor.
To achieve a comparable effect with convex discharge vessels, according to the invention metal halide additives are substituted for proportions of CaI2 in the fill, without altering the main fraction of the RE halide concentration and therefore the chemical activity thereof. The result is a change in the degree of wetting, which produces a defined fill distribution in the electrode back space of the lamps, with a low individual scatter occurring when setting the color temperature.
It has been found that the divalent metal halide components of Yb and if appropriate Mg, preferably MgI2 and YbI2, are suitable for completely or partially, but at least to an extent of 20 mol % in the total fill and therefore 20/45 of the molar CaI2 quantity, performing the role of the CaI2.
It is preferable to use a quantity of 20-25 mol % YbI2 while maintaining 20 to 25 mol % CaI2 or to simultaneously use halides of Mg and Yb, in such a way that, in particular when using the iodides MgI2 and YbI2, the total quantity of YbI2+MgI2 forms a proportion of at least 20 mol %, in particular 20 to 35 mol %, of the metal halides, and together with CaI2 form a total proportion of 40-50 mol % of the total fill of metal halides.
In the text which follows, the invention is to be explained in more detail on the basis of a number of exemplary embodiments. In the drawing:
The discharge vessel 10 may in particular be internally spherical or elliptical or may deviate from the spherical geometry by virtue of having a short cylindrical center piece between the half-shells of the sphere. In particular, it has the dimensions shown in
Specifically, in this example, Lcyl=1 mm, L=15 mm and R=7 mm.
The ratio of the external radius to the internal radius is Ra/Ri=7.8/7=1.11. The ratio of internal radius/cylinder length is Ri/Lcyl=7/1=7. The electrode gap is 9.2 mm.
Electrodes 3 project into the discharge vessel. The electrode gap EA and the length L of the discharge vessel are in a ratio EA/L=9.2/15=0.61.
An ignitable gas selected from the group consisting of the noble gases is located in the discharge vessel at a cold fill pressure of 300 mbar. The discharge vessel also contains mercury and a mixture of metal halides consisting of the following molar compositions (mol %) in accordance with Table 2 below:
The power consumed is in a range from 140 to 150 W. If the ratio of the power to the external surface area of the discharge vessel is considered, the wall loading is typically in a ratio of from 17.2 to 18.45 W/cm2.
If the ratio of the power to the internal surface area of the discharge vessel is considered, the wall loading is typically in a ratio of from 21.2 to 22.75 W/cm2.
The color temperature for these lamps is in each case approximately 4200 K.
The exemplary embodiments reveal a considerable reduction in the scatter in the color temperature and the crest factor. Evaluation after an illumination time of 100 hours gives the following result:
This table in each case shows the mean value and the standard deviation for the crest factor Cr and the color temperature Tn.
The lowest scatter in the color temperature combined, at the same time, with an acceptable crest factor is found in exemplary embodiment 2, in which 66% of the CaI2 molar fraction is substituted by YbI2 (totalling 31.2 mol % in the overall mixture).
A similar behavior with regard to the reduction in the scatter of the color temperature can be achieved if CaI2 is partially substituted by MgI2. The effectiveness of the admixture in reducing the scatter in the color temperature results from the reduction in the wetting angle of the molten metal halide melt on the aluminum oxide ceramic. The effectiveness in reducing the scatter in the color temperature, both with MgI2 and with YbI2, becomes significant once at least 15 mol % has been added, preferably 20-35 mol %, in the metal halide melt. The proportion should not exceed 55 mol %.
This is linked to the replacement of the CaI2, which improves the read fraction and may typically be present up to within the range from approx. 40-45 mol % as a constituent of MH fills for a color temperature of 4000 K.
The CaI2 may be replaced completely or partially by the substances YbI2 or MgI2, individually or together, preferably in a proportion of approx. 50-70% of the Ca iodide. This means that optimum conditions are achieved in fills with typical contents of 15-25 mol % formed from at least one of the metal halides DyI3, HoI3, TmI3, and that the proportions of the group of the wetters made up of MgI2 and YbI2 should be in the range from 15 to 55 mol %, optionally including CaI2, preferably in the range from 15-35 mol %, in the overall mixture.
It can be seen that in the case of the convex discharge vessel, there is a smooth relationship between the integrated inner surface area (i) and outer surface area (a), and the two are closely related. In the case of the cylindrical discharge vessel, the relationship involves sudden jumps, cannot always be differentiated and the relationship alters. In particular, the inner surface area may temporarily even be greater than the external surface area.
| Number | Date | Country | Kind |
|---|---|---|---|
| 10 2005 013 003 | Mar 2005 | DE | national |
| Number | Name | Date | Kind |
|---|---|---|---|
| 4243906 | Wilson | Jan 1981 | A |
| 5936351 | Lang | Aug 1999 | A |
| 6218789 | Takahashi et al. | Apr 2001 | B1 |
| 6483241 | Stockwald | Nov 2002 | B1 |
| 6501220 | Lambrechts et al. | Dec 2002 | B1 |
| 6646379 | Nohara et al. | Nov 2003 | B1 |
| 7245081 | Ashida et al. | Jul 2007 | B2 |
| 20030015949 | Higashi et al. | Jan 2003 | A1 |
| 20030020408 | Higashi et al. | Jan 2003 | A1 |
| 20050073257 | Takeuchi et al. | Apr 2005 | A1 |
| 20060028113 | Arndt et al. | Feb 2006 | A1 |
| 20060049765 | Ota et al. | Mar 2006 | A1 |
| 20060164016 | Rintamaki et al. | Jul 2006 | A1 |
| 20060238127 | Van Esveld et al. | Oct 2006 | A1 |
| Number | Date | Country | |
|---|---|---|---|
| 20060208643 A1 | Sep 2006 | US |
