Journal of the Korean Physical Society, Vol. 54, No. 6, June 2009, pp.
2274$2281
Zoom Lens Design for a Slim Mobile Camera Using Liquid Lens
Sung-Chan
Park
and Jun
Park
Department of Physics, Dankook University, Cheonan 330-714
(Received 6 April 2009) In this study, we have designed a two-group inner-focus zoom lens system using a prism and a liquid lens and we evaluated that zoom system. By xing the rst group, the inner-focus zoom system has the same total track at all zoom positions and the advantage of mechanical size, which is useful for a slim mobile camera. To acquire a slim and compact zoom system, we directly inserted a right-angle prism of 5.5 mm in thickness in front of the rst lens group. Specially, the second group of the zoom system is placed into a liquid lens whose power is varied to maintain the image plane at di erent magni cations. This variable power can be acquired by changing the curvature between the oil and the water in a passive device. In conclusion, the focal length range of this zoom lens is 4.3 mm to 12.90 mm and the aperture is F/3 at a wide position and F/5 at a tele position. This design procedure resulted in a slim zoom lens system useful for a mobile camera.
PACS numbers: 42.15.Eq, 42.15.Fr Keywords: Mobile camera, Liquid lens, Electro-wetting, Zoom lens design, Aberrations
I. INTRODUCTION
The liquid lens is explained using an electro-wetting phenomenon [1{4]. The contact angle of a liquid droplet is normally determined from the force balance between the electrostatic force and the surface tension. This phenomenon, known as electro-capillary behavior, was rst formalized by Lippmann and is quanti ed by the expression known as Lippmann's equation [5]: @
(1) ( SL )T ; P = ; @V where
SL is the interfacial tension between a solid and a liquid, V is the external voltage, is the surface charge density in Coulomb/m2 and the derivative is taken at constant temperature and pressure. Consequently, the change in the interfacial tension at a given potential is closely related to the surface charge density. The variation in the contact angle due to the external voltage is shown in Figure 1. Although the contact angle change caused by the external voltage is known empirically, it is not easy to acquire the exact contact angle. At voltage above the threshold voltage, although the external voltage is increased continuously, the contact angle no longer changes [1]. This phenomenon, known as contact angle saturation, is quanti ed by the following expression known as Young's equation, which is the equilibrium condition for a three-phase contact line [1{4]:
LV cos =
SV
SL ; (2)
where
SV ;
SL and
LV are the interfacial tensions of the solid/gas, solid/liquid and liquid/gas interfaces, respectively. Eq. (2) predicts the contact angle for any given three-phase. The capacitance (C) of the system is mainly determined by the capacitance of the insulating layer and has a constant value [4]. From Eq. (1), we obtain 1
SL (V ) =
SL CV 2 2 (3)
where
SL is the interfacial tension of the solid/liquid interface at zero voltage. If Eq. (2) and Eq. (3) are combined, the Lippmann-Young equation is given as [1{ 4] cos (V ) = cos 0 + 1 CV 2 2
LV (4)
-2274-
This expression is currently used as the fundamental equation for the electro-wetting e ect. Since the electrowetting phenomenon is reversible and has a fast response due to low electric power, it is useful for vari-focus lenes and actuators for focusing. Figure 2 shows the change in the focal length of a liquid lens at di erent voltages. The external voltage changes the curvature of the liquid lens, so various focal lengths can be obtained by using a liquid lens. Because of these properties of a liquid lens, in this research, we used a liquid lens as a variator and a compensator in a compact zoom-lens design.
Zoom Lens Design for a Slim Mobile Camera Using Liquid Lens { Sung-Chan Park and Jun Park
-2275-
Fig. 3. Layout of the two-group inner-focus zoom system. (a) Position 1 (Wide Position), (b) Position 2 (Middle Position) and (c) Position 3 (Tele Position).
Fig. 1. Variation of the contact angle and its saturation by electro-wetting. (a) Contact angle 0 at no external voltage, (b) Contact angle at external voltage V and (c) Contact angle s at saturation.
the zoom system is composed of a xed front lens group and a second lens group for zooming and compensation. Their powers are denoted by k1 and k2 respectively. The rst group is always xed. While the second group moves to the object side to have a longer focal length, the power of the second group should vary to keep the image position stationary. This variable power of the second group can be realized because of the liquid lens, whose curvature is changed by the electro-wetting phenomenon. Among the various zoom types, the inner-focus system has a structural advantage [6]. Since the rst group is always xed, the total track length of the zoom system has the same value at all zoom positions. This con guration is useful to have a slim zoom system for a mobile camera. From Figure 3, the four zoom equations can be formulated as follows:
Kwide = k1 + k2 k1 K2 z1w ; Ktele = k1 + k2 k1 k2 z1t ;
(5) (6)
Fig. 2. Various focal lengths of the liquid lens due to the electro-wetting e ect. (a) short focal length and (b) long focal length.
II. ANALYSIS OF A TWO-GROUP INNER-FOCUS ZOOM SYSTEM
where Kwide and Ktele are the optical powers at wide and tele positions, respectively. Since the axial ray has zero height at the image plane, hiw and hit should be zero:
hiw = [k1 ; z1w ; k2 ; z2w ] = 0; hit = [k1 ; z1t ; k2 ; z2t ] = 0;
(7) (8)
The zoom lens for a camera has continuously changing magni cations or focal lengths at various zoom positions. These are obtained by moving the variator for zooming and the compensator for stable focusing. The layout of the two-group inner-focus zoom system is shown in Figure 3. From the object to the image side,
H where z1w is the distance between H1 and H2 at wide H position and z1t is distance between H1 and H2 at tele position. From the Eqs. (7) and (8), the back focal distances (z2w ; z2t ) at the wide and the tele positions are given by
z2w =
(z1w )k1 + 1 Kwide
(9)
-2276-
Journal of the Korean Physical Society, Vol. 54, No. 6, June 2009
(z1t )k1 + 1 (10) Ktele After rearranging the four equations, Eqs. (5)$(8), we can get expressions for the unknown quantities k1 , k2 , z1w and z1t [6,7]:
z2t = k1 = k2 = Kwide Ktele (z2t z2w ) (Ktele z2t ) (Kwide z2w ) Kwide Ktele ; Kwide z2w Ktele z2t
(11) (12) (13)
z1w = z1t =
1 Kwide z2w ; k1
1 Ktele z2t ; (14) k1 These four parameters(k1 ; k2 ; z1w ; z1t ) are the initial design parameters required for the zoom lens. If the back focal distance (z2w ; z2t ) and the focal lengths at wide (= fw ) and tele (= ft ) positions are given, the four initial design parameters (k1 ; k2 ; z1w ; z1t ) can be obtained from Eqs. (11) to (14), sequentially.
III. INITIAL ZOOM LENS DESIGN USING LENS MODULES
Fig. 4. Optimized zoom system consisting of two lens modules.
For the initial design of two-group inner-focus zoom system, we select the proper back focal distances and the total powers at the wide and the tele positions by using paraxial studies, as follows:
z2w = 7 mm; z2t = 11 mm; Kwide = 1=fw = 0:23 mm1 ; Ktele = 1=ft = 0:077 mm1 ; T = z1 + z2 $ 12 mm; =
(15)
where T is the overall length of the zoom system, which determines the size of the system. By inputting these initial data into Eqs. (11) $ (14), we can get the proper values for the powers of each group and the zooming locus:
k1 = 0:093 mm1 ; k2 = 0:2 mm1 ; z1w = 5 mm; z1t = 1 mm:
(16)
The next step is set up for the lens module zoom system by inputting these data into each lens module. Each group of the zoom system is generally composed of several lens elements. When the higher-order aberrations are neglected, the lens system can be speci ed by its rst-order quantities and the third-order aberrations. In other words, if we assign the rst-order quantities and the third-order aberrations of the lens modules to the real lens system, then both lenses are equivalent to each other
within the limited aperture and eld [8,9]. Hence, each group of the zoom system can be replaced by a thicklens module by specifying its focal length (FLM ), front focal length (FFM ), back focal length (BFM ), magni cation (MGM ), entrance pupil position (EPM ), entrance pupil diameter (EDM ), eld angle ( ) and aberrations at limited aperture and eld [10{12]. We have set up the zoom lens system with two thicklens modules, for which initial rst-order inputs are given, from Eqs. (15) and (16), to work as a zoom system. This is based on the retro-focus system, for which the second module has positive power, but the rst module has negative power. The air distances between each module should be con rmed for the mounting space. Since the lens modules do not re
ect higher-order aberrations, it is desirable to reduce the aperture and the eld size of the system so that the lower-order aberrations are dominant. We have taken a zoom system with a half image size of 1 mm and f-numbers of F/5 at position 1 to F/7 at position 3. The distances between modules are constrained to be longer than 0.2 mm over all zoom positions. We selected the overall length to be as short as possible for a compact system. In order to get an optimum zoom system, we optimized the lens module prescriptions so that the speci c constraints were satis ed. The design variables are the focal lengths, the front and the back focal lengths, the conjugate points, the spacings and the aberrations of each module. Figure 4 shows the initial design of the zoom system obtained from this process. The focal lengths are from 4.3 to 12.9 mm. Table 1 shows the zooming locus for each position and Table 2 shows the data for each module. In Table 1, d1i is the air space between the modules at the i-th zoom position. In Table 2, the values of W040 , W131 , W222 , W220 and W311 denote the third-order wave aberrations calculated at the edge of the eld and at the exit pupil in units of waves at the d-line. Therefore, they correspond to the wave aberration coecients for spherical aberration, coma, astigmatism, Petzval curves and distortion,
Zoom Lens Design for a Slim Mobile Camera Using Liquid Lens { Sung-Chan Park and Jun Park Table 1. First-order properties and zooming locus of the zoom system consisting of two lens modules (z1i : distances between the principal planes of the rst and the second groups; d1i : air-spaces of these groups).
efl bfl ffl d1i z1i z2i
-2277-
Position 1 4.300 1.804 2.203 4.216 5.066 6.938
Position 2 6.450 3.185 0.190 2.834 3.684 7.937
Position 3 12.900 5.820 5.294 0.200 1.050 11.050
Fig. 5. Schematic diagram of the real lens system with ve elements.
Table 2. Design data (in mm) for the lens modules in the optimized lens module zoom system. module I FLM FFM BFM MGM EDM Field ( ) Thickness W040 W131 W222 W220 W311 module II at wide 5.878 0.744 0.294 0 1 1 3.913 0.0280 0.0059 0.0008 0.0026 0.0002 module II at middle 7.178 2.425 0.658 0 1 1 3.913 0.0252 0.0106 0.0004 0.0020 0.0001 module II at tele 9.529 4.298 4.384 0 1 1 3.913 0.0251 0.0142 0.0015 0.0014 0.0004
bfl = f [k1 ; d1 =n1 ; k2 ; ; k9 ; d9 =n9 ];
(18)
12.046
0 1 1 1.968 0.0092 0.0024 0.0031 0.0004 0.0006
F ront focal length : ffl = f [d1 =n1 ; k2 ; d2 =n2 ; ; d9 =n9 ; k10 ]; (19) Magnification : M = 1=[d0 =n0 ; k1 ; d1 =n1 ; ; d9 =n9 ; k10 ]: (20)
12.896
11.015
respectively. These wave aberrations are measured as departures from the reference sphere at a given aperture and eld. They are easily calculated using the Code-V macro [11].
IV. REAL LENS DESIGN FOR GROUPS
The real lens group is composed of several lens elements, which should be equivalent to the lens module given in Table 2. The schematic diagram of this lens system is depicted in Figure 5. The aperture stop lies on the rst surface and the chief ray makes an angle with the optical axis at the stop. The focal length (f ), the front focal length (ffl), the back focal length (bfl) and the magni cation (M ) at a given conjugate of this system are expressed in terms of Gaussian brackets as [12{15]
F ocal length : f = 1=[k1 ; d1 =n1 ; k2 ; ; d9 =n9 ; k10 ]; Back focal length :
In the above equations, kj (j = 1, 2, , 10) is the optical power of each surface, dj (j = 0, 1, 2, , 10) is the distance between surfaces and uj (j = 0, 1, 2, , 10) is the convergence angle of the ray from the axial object point, as shown in Figure 5. Therefore, the optical power kj is given by cj (nj nj 1 ), where cj and nj are the curvature and the refractive index of the j-th surface. The refractive indices in the object (n0 ) and the image (n10 ) spaces are assumed to be unity and the square brackets denote the Gaussian brackets. For the real lens system to be equivalent to the lens modules to within the limited aperture and eld, all the rst-order quantities and all the third-order aberrations of the real lens should be equal to those of the lens module. Analytic and numerical approaches using extensive computer calculations to obtain the real lens data have been reported [12, 15, 16]. However, it is very complicated to handle all the rst-order quantities and thirdorder wave aberrations at the same time. First of all, it is hard work to handle analytically all the third-order wave aberrations of the liquid lens having variable curvatures. In this research, therefore, the rst-order quantities given in Eqs. (17)$(20) were used to make the real lenses equivalent to the lens modules, except the thirdorder aberrations. Accordingly, we can set the equivalent conditions as
F LM = 1=[k1 ; d1 =n1 ; k2 ; ; d9 =n9 ; k10 ]; BFM = f [k1 ; d1 =n1 ; k2 ; ; k9 ; d9 =n9 ]; F FM = f [d1 =n1 ; k2 ; d2 =n2 ; ; d9 =n9 ; k10 ];
(21) (22) (23)
(17)
MGM = 1=[d0 ; k1 ; d1 =n1 ; ; d9 =n9 ; k10 ]: (24)
-2278-
Journal of the Korean Physical Society, Vol. 54, No. 6, June 2009
Fig. 7. Power changes of the liquid lens part in the second group at three positions. Fig. 6. The power variation (in diopter) of a liquid lens due to external voltage.
If Eqs. (21)$(24) are satis ed simultaneously, the real lens is equivalent to the lens module, within paraxial optics. In this paper, an optimization design method is proposed to design a real lens group equivalent to the module of each group. The design variables of the real lenses are changed to obtain a lens system in which the four rst-order quantities are matched to those of the lens modules. Thus, the constraints are composed of the four rst-order quantities of each lens module given in Table 2. Therefore, the real lens that satis es the constraints is equivalent to the lens module within paraxial optics. At the stage of initial zoom system design, the groups must be as compact as possible to improve the portability of the camera. Therefore, each group must be designed with a few elements. In the initial zoom system design using lens modules, the rst group is always xed for zooming. This makes this group free from independent color correction so that it must be as compact as possible to have a slim camera. Therefore, it is desirable to design the rst group as a single element with a negative power. In the singlelens case, there are four design variables, i.e., c1 , c2 , d0 and d1 . Therefore, the four constraints given by Eqs. (21)$(24) can be satis ed by specifying the lens design variables by using the optimization design method in Code-V. After a few iterations, the real lens of the rst group is obtained. The design data with small chromatic aberrations are selected and evaluated. There are aberrations that are not corrected, but the agreement for the rst-order properties is complete. The second group in Table 2 has variable focal lengths due to the zoom positions. They are 5.878 mm at wide position and 9.529 mm at tele position. Compared to the rst group, these strong powers reduce the amount of displacement of this group to have a higher zoom ratio for zooming. Also, this group must make the image plane stationary for compensation and balance the aberrations generated by the rst group. A liquid lens is
used to vary the power of the second group at each zoom position. Therefore, a liquid lens and a few elements, including a doublet, are at least needed to have a lens system equivalent to the lens modules. This con guration is useful to correct the chromatic aberrations and the coma. To obtain the powers of this group at three positions, as given in Table 2, we optimized the design variables including the liquid lens. Figure 6 denotes the maximum power change for a current liquid lens due to an external voltage. Actually, contact angle saturation limits the magnitude of the power change of the liquid lens, as shown in Figure 6 [1,2]. At the initial design stage, therefore, we should constrain the power range of the liquid lens so that we limited its power to range from +13 m1 to 18 m1 , in diopter. Figure 7 shows the powers of the designed liquid lens at three positions. From this process, the real lens of the second group is obtained. This lens group and lens module II exhibit the expected properties; i.e., there is little aberration, but the agreements for the rst-order quantities and power changes at each position are good. Therefore, this group is equivalent to lens module II within paraxial optics.
V. OPTIMIZED DESIGN FOR THE SLIM ZOOM SYSTEM
The groups separately designed in the previous section, due to the zooming locus of Table 1, are then combined to establish an actual zoom system. If a zoom system equivalent to the lens module zoom system is to be achieved, the airspaces (dji ) between groups should be set according to the zooming locus of Table 1 at each position. This procedure results in a zoom system equivalent to the lens module zoom system within paraxial optics. From the evaluation for these zoom systems, the agreement for the rst-order quantities between both systems is complete. However, there are color and residual aberrations that are not corrected in the designs for the rst and the second group designs.
Zoom Lens Design for a Slim Mobile Camera Using Liquid Lens { Sung-Chan Park and Jun Park
-2279-
Table 3. Design data of an aberration-balanced zoom system (in mm). No. Radius Thickness Glass Object In nity In nity 1 In nity 2.7500 FDS90 2 In nity 2.7500 FDS90 3 In nity 0.2799 A A 4 10.570 0.6000 BAFN11 A 5 14.315 *4.573 6 In nity 0.2000 7 2.3956 0.7005 FD10 8 3.9519 0.7226 TAC4 9 39.0187 0.2359 A 10 8.316 0.6000 FDS90 A 11 9.736 0.6345 A 12 5.582 0.6000 NPSK57 13 In nity 0.6000 Water(nd = 1.35, vd = 59) L 14 4.194 0.6300 Oil(nd = 1.45, vd = 44) 15 In nity 0.6000 FDS90 16 In nity 0.5146 A 17 4.578 0.6669 E48R A 18 3.322 *0.200 19 In nity 0.3000 BSC7 20 In nity 1.2806 Image In nity 0.0194 L A (* air-spaces of moving groups, : liquid lens, : aspheric surfaces). Table 4. First-order properties and zooming locus of an aberration-balanced zoom system (in mm).
efl bfl ffl d5i d18i R14 (liquid surface)
Fig. 8. Layout of an optical zoom system with a liquid lens and a right-angle prism.
A plastic block is inserted to fold the ray path. It is located in front of the rst lens, not inside the lens system. Its thickness is just 5.5 mm, which con guration realizes a much slimmer zoom system than any other one [17]. Since the axial rays are parallel to the optical axis, there are no aberrations induced by the prism block. In the initial design, we reduced the aperture and the eld size so that the f-number was too large and the image size was too small. If current speci cations for a zoom camera are to be met, the aperture and the eld size should be increased. The f-numbers are extended to F/3 at position 1 and to F/5 at position 3. The half image size should be 2.2 mm for a 1/4-inch CCD. In an extended aperture and eld system, however, higherorder aberrations that are not corrected in the previous design become signi cant. In order to improve the overall performance of the zoom system with an extended aperture and eld, we balance the aberrations of the starting data by using a lens design program. In this process, the rst-order layouts are xed. To correct for the residual aberrations, we use aspheric lenses. The aspheric surface has many design parameters, so aberrations can be well corrected [18,19]. The rst group lens is a singlet and relays the ray refracted from the prism to the second group. Because the stop is placed in front of the second group, the chief ray height in the rst group is very high at the wide position. Therefore, the rst group generates many o -axis aberrations. Both surfaces of this lens are aspherized to balance these aberrations. The second group is designed into a cemented lens and a liquid lens including singlet. To correct for the spherical aberration, it will be generally e ective to use an aspheric lens in this group. In addition, the liquid lens is used to change the power of
Position 1 4.300 1.300 3.026 4.573 0.200 4.194
Position 2 6.450 1.300 0.806 3.267 1.486
Position 3 12.900 1.300 14.276 0.200 4.542
4.182
3.050
this group at each zoom position. Other lenses balance the residual aberrations. Finally, a zoom system having good performance is obtained, as shown in Figure 8 and the design data are listed in Table 3. Table 4 shows the rst-order speci cations and zooming locus of this zoom system. Figure 9 presents the distortion and Figure 10 shows the modulation transfer function (MTF) characteristics of the system at two extreme positions. Aberrations are significantly reduced and the MTF at 200 lp/mm is more than
-2280-
Journal of the Korean Physical Society, Vol. 54, No. 6, June 2009
zoom positions. In this system, the ratio of relative illuminations is more than 66 % over all positions, as shown in Figure 8. The chief ray angles of incidence (AOI) into the image plane are shown in Figure 8. The variation of the AOI from a wide to a narrow eld is less than 5.5 degrees. That is an extremely small value, so a stable image quality for zooming can be realized. The depth of the zoom lens is less than 5.5 mm, which results in a slim and compact mobile zoom camera. Consequently, this zoom system has enough performance to satisfy current requirements for a mobile zoom camera.
VI. CONCLUSION
Fig. 9. Distortion plot of an aberration-balanced zoom system. (a) wide position and (b) tele position.
For compact zoom lens design, we set up an inner-focus two-group zoom system consisting of two lens modules with initial data obtained from paraxial study. The optimum initial design with a zoom ratio of 3X was derived by assigning the rst-order quantities and aberrations to each module along with speci c constraints. From an optimization design procedure, a good design for the real lens of each group was quickly obtained by matching the four rst-order quantities of the modules and speci c constraints. The separately designed groups were combined to establish an actual zoom system. A prism of 5.5 mm in thickness was inserted to fold the ray path, which con guration resulted in a slim zoom system useful for the mobile zoom camera. Through balancing of the higher-order aberrations in the extended aperture and eld, we improved the performance of the zoom system further. A compact system with a zoom ratio of 3X , whose aperture was F/3 at wide eld and F/5 at narrow eld and which had an image size of 1/4-inch on a CCD, was obtained. The zoom system developed in this work performs reasonably as a slim zoom camera with 3 to 5-mega pixels.
REFERENCES
Fig. 10. MTF characteristics of an aberration-balanced zoom system. (a) wide position and (b) tele position.
30 % at all zoom positions within 0.7 elds. The relative illuminations are calculated at marginal elds for all
[1] H. J. Verheijen and M. W. Prins, Reversible Electrowetting and Trapping of Charge (Philips Research Laboratories, Eindhoven, 1999). [2] A. I. Drygiannakis, A. G. Papathanasiou and A. G. Boudouvis, JCIS 326, 451 (2008). [3] C. Gabay, B. Berge, G. Dovillaire and S. Bucourt, Proc. SPIE 4767, 159 (2002). [4] W. J. Welters and L. G. Fokkink, Langmuir 14, 1535 (1998). [5] A. D. McNaught and A. Wilkinson, Compendium of Chemical Terminology, 2nd ed. (IUPAC, Zurich, 1997). [6] K. Tanaka, Appl. Opt. 22, 2174 (1983). [7] M. Herzberger, J. Opt. Soc. Am. 33, 651 (1943). [8] T. G. Kuper and M. P. Rimmer, Proc. SPIE 892, 140 (1988). [9] S. C. Park, Y. J. Jo, B. T. You and S. H. Lee, J. Korean Phys. Soc. 50, 1243 (2007).
Zoom Lens Design for a Slim Mobile Camera Using Liquid Lens { Sung-Chan Park and Jun Park [10] W. T. Welford, Aberration of Optical Systems (Adams Hilger Ltd., Bristol, 1986). [11] CODE V Reference Manual, Version 9.70 (Optical Research Associations, Pasadena, Calif. 2007). [12] S. C. Park and R. R. Shannon, Opt. Eng. 35, 1668 (1996). [13] M. Herzberger, Modern Geometrical Optics (Interscience, New York, 1958). [14] K. Tanaka, Paraxial Theory in Optical Design in Terms of Gaussian Brackets, in Process in Optics XXIII, edited
-2281-
by E. Wolf (North-Holland, Amsterdam, 1986), p. 63. [15] S. C. Park and Y. J. Jo, J. Korean Phys. Soc. 52, 1048 (2008). [16] S. C. Park and J. U. Lee, J. Korean Phys. Soc. 32, 815 (1998). [17] M. Sueyoshi, U. S. Patent 7110186 (2006). [18] G. I. Kweon, J. Opt. Soc. Korea 12, 79 (2008). [19] J. H. Lee, T. S. Jang, H. S. Yang and S. W. Rhee, J. Opt. Soc. Korea 12, 262 (2008).
