# INTRODUCTION ecently a lot of effort has been concentrated to develop standard 3D vision imagers due to the drastic increase in demand of 3D imaging system. The 2D-imaging system can evaluate only the intensity projection of a scene, there is no information about the depth of the 3D objects. Range-imaging sensors acquire threedimensional (3D) maps from a scene and can be used in a variety of applications such as bio medical appliances, surveillance system; several applications in automobiles, robomechatronics single point measurement etc. 3D image isextracting information from the geometric estimation of third co-ordinate of a scene. In this work, we present a finger typed field assisted electro-optical demodulator fabricated in custom technology. After reviewing related research work in Section 2, the device architecture with its working principle and ISE-TCAD simulation are introduced in Section 3. In Section 4 the electro-optical characteristics of the device is reported. Finally the paper is concluded in Section 5. A number of applications that can detect the time or phase information of reflected light for 3D imaging are available in the literature. Depth information can be determined by correlating the incoming modulated light signal from the scene with a reference signal synchronous with the modulation signal of the light source [ ]. In time-of-flight optical ranging, the phase information is used to plot the distance map of the observed scene, thus enabling the reconstruction of the shape and position of the observed objects [2]. TOF technique provides the best performance in terms of acquisition speed, reliability, overall cost of the system and is most suited to integrate electronic circuitry with more functionality. Several studies on image capturing techniques using specialized pixels coupled with active illumination have reported to produce images with information even at a low intensity level [3 4].TOF based 3D imagers so far reported in different literatures depending on the type of photo detector used in the pixels. The time or phase information in addition to signal intensity is based on a variety of lightsensitivedevices such as: p-i-n photodiodes, linear or Geiger mode avalanche photodiodes and photomultipliertubes. Several works [5 -8] reported a standard photodiode coupled with complex readout circuitry using indirect time-of-flight. The key element of 3D range camera of photo demodulators have been implemented with different types of technologies such as: Charged Couple Device (CCD), Complementary Metal Oxide Semiconductor (CMOS) and CMOS/CCD hybrid approach. The photo generated charge is mixed on two or more photo-gates thus achieving an intrinsic demodulation effect [9 -12]. The advantage is the read-out channel simplicity which results in a small pixel size.The disadvantages are the lower sensitivity due to the presence of the "photo-gate", the lack of immunity to the ambient light and the cost of the nonstandardtechnology.A Photonic Mixer Device (PMD) is an interesting solution for dynamic 3D-vision that is reported in [ 3]. An alternative demodulating detector structure, the Current Assisted Photonic Demodulator (CAPD) has been reported in [14].The function of detection and demodulation in a single device uses a modulated electric field that infiltrates deeper into the substrate to enhance the charge separation and collection mechanism.A linear Current Assisted Photonic Mixing Device fabricated on high resistivity silicon has been described in [15]. # III. DEVICE ARCHITECTURE AND WORKING PRINCIPLE A finger-typed electrode based electro-optical demodulator consists of multiple strips. The cross sectional view and the layout of the device is shown in Figure 1 (a) and (b) respectively. The above Figure1(a) shows an active pixel that contains a finger typed photonic mixing demodulator that is integrated with a read out circuit and a row select transistor for each collecting electrode.This device has four electrodes, two of them known as collecting electrodes and connected to V ce1 and V ce2 and rest of them are known as modulating electrodes and connected toV me1 and V me2 These modulated electrodes are connected to the device substrate. This device consists of nine collecting electrodes. Among these electrodes, the seven central regions (from region 2 to region 8 as shown in Figure .01) consist of a p+ type detection junction and two n+ type substrate contacts. The rest of collection regions contain one collecting junction and one substrate contact, like the region 1 and 9 (Figure .1) [16]. All of the substrate contacts and the collection junctions are connected as shown in the device cross sectional diagram. The collecting electrode V ce1 and the modulating electrode V me1 are connected to the 2, 4, 6 and 8 regions according to the detection junctions and substrate contacts. On the other hand, the collecting electrode V ce2 and the modulating electrode V me2 are V me2 -V me1 is applied at the modulated electrode V me2 . An electric field formed inside the substrate of the device guides the photo-generated charge carriers towards the detection electrode V ce2 . At 780 nm light incident on the device surface the hole current density of this device shown in Figure 2. his simulated photograph shows the region 1and 2 according to the cross sectional view of the device shown in Figure 1(a). Most of the generated holes move toward the collecting electrode V ce2 , guided by the electric field with a voltage difference applied between two modulating electrodes i.e. V me2 > V me1 . # IV. PERFORMANCE CHARACTERISTICS OF THE DEVICE The customize Fabricated structure is characterized electrically and optically. The demodulation contrast of the device and the effect of frequency and modulation voltage on it are assessed. Phase measurements are carried out to evaluate linearity of the device. connected to the region 1, 3, 5, 7 and 9 accordingly. This device is also enclosed with an n+ bulk-contact, shared along the array and placed at a minimum distance of about 20?m from the pixel boundary. A p+ ring is surrounded by n+ bulk-contact at a distance of about 20?m for better isolation of each device. The distance between the neighboring modulating electrodes is 20 ?m and the total area of this device is 0.4mm × 0.4mm.Firstly, the device simulation software ISE-TCAD is used to investigate the operational behaviour of the device. In this device a potential difference is applied between the modulating electrodes to direct the signal charges towards the two detection regions. Now we can calculate the demodulation contrast by using the following equation. The device shows a maximum DC demodulation contrast larger than 90%, thus indicating that, this device is potentially enabling a very efficient mixing process. # b) Dynamic Characteristics In order to measure average current at the collecting electrodes and the dynamic demodulation contrast we have conducted and experiment. The schematic representation of this experiment set up is shown in Figure 5. characterizations Two sinusoidal waves are generated by using a function generator. One of the two sinusoidal waves is used to modulate a laser emitter and illuminate the device. The other is connected to the input of a differential amplifier. The differential amplifier outputs with 180 0 phase shift are connected to the modulating electrodes V me1 and V me2 of the device. The electric field formed in the substrate average current through the collecting electrodes V ce1 and V ce2 is read out with a Semiconductor Parameter Analyzer. For this measurement, the sinusoidal signal for laser emitter and two modulating signals are needed to use with an appropriate synchronization. At different modulation frequencies, the average current is measured under a 650nm red laser with 90% modulation depth used to illuminate the test device.The capability to separate and transfer the charges of a sensor to the corresponding output node can be expressed as a demodulation contrast. For data acquisition a LABVIEW software program was developed for the interface with PC and the experimental set-up. The dynamic demodulation contrast is the most important performance indicator for this device. The demodulation contrast depends on both the amplitude of the modulation voltages and frequencies. The dynamic demodulation contrast can be defined as: Where I max and I min are the photo-generated currents flowing at collecting electrodes V ce1 and V ce2 . The demodulation contrast for the finger typed device as a function of the modulation voltage amplitude at eight different frequencies from 100Hz to 30MHz is shown in Figure : 6. By increasing the modulation voltage it should be possible to increase the majority current that cause the drift of the minority carriers, namely holes. When the modulation voltage is applied to the modulating electrodes, the photo generated holes arrive at the collecting electrode of the device. If applying more voltages, the electric field penetrates deeper in the substrate so that more holes reach detection node resulting in a higher demodulation contrast. Due to the larger voltage applying at the modulating electrodes the power consumption is increased. So the amplitude of modulation voltage should be carefully chosen. By increasing of the modulating frequency the decrease of the demodulation contrast can be described with respect to diffusion time. The photo-generated charges in the deeper of the substrate need more time to reach the active region where the demodulating electric field is present and thus reduces the demodulation contrast. The phase linearity measurements performed between the applied phase and measured phase of the device. In these measurements a variable phase delay V is applied between the laser input to illuminate the device and two modulation electrodes of the device. The value of Vcan be recovered acquiring four amplitude measurements with four different phase shifts 11 , 12, 21 and 22 applied to the modulated laser signal considered as -180°, -90°, 0° and +90° respectively [17]. The phase shift can be calculated with the equation (ii) given below: At three different frequencies-3MHz, 1 MHz and 100 kHz the C-V response of the device is shown in Figure : 9. Due to a larger depletion width the higher reverse bias produces a lower capacitance. At lower frequency, the capacitance is larger than at higher frequency. Because of their finite charging and discharging time the deep-level impurities in the space charge region make the capacitance to be frequency dependent [18,19]. # CONCLUSION This paper has described the characterization of a finger typed electro-optical demodulator fabricated in a custom technology on high resistivity silicon substrates. A 400 m × 400 m structure with finger typed electrodes has been considered and tested in terms of electrical and electro-optical performance. The maximum phase linearity error between the applied phase and the measured phase is 4.09% for square wave. In particular, the DC and dynamic demodulation performance of the multiple strip devices has been investigated. The measured dynamic demodulation contrast is more than 20% at 20 MHz modulation frequency. This customize device corresponds to understand field assisted photo mixing demodulator in term of optimizing the performance to make them in complementary metal-oxide-semiconductor technology. 1![Figure 1 : (a) Cross sectional view of multiple strip CAPD and (b) Device layout](image-2.png "Figure 1 :") 2![Figure 2 : Simulated hole current density under illumination hen a potential difference VV me2 -V me1 is applied at the modulated electrode V me2 . An electric field formed inside the substrate of the device guides the photo-generated charge carriers towards the detection electrode V ce2 .At 780 nm light incident on the device surface the hole current density of this device shown in Figure2.his simulated photograph shows the region 1and 2 according to the cross sectional view of the device shown in Figure1(a). Most of the generated holes move toward the collecting electrode V ce2 , guided by the electric field with a voltage difference applied between two modulating electrodes i.e. V me2 > V me1 .](image-3.png "Figure 2 :") ![Typed Electrode Based Electro-Optical Demodulator Fabricated on High Resistivity Silicon © 2013 Global Journals Inc. (US)a) Dc CharacteristicsAn experimental characterization was carried The DC characterization set up of the test device is shown in above Figure3.The device is enlightened with a wide spectrum lamp.](image-4.png "Finger") 3![Figure 3 : Experimental setup for DC characterizationsIn this measurement the required voltage at different electrodes can be supplied with a voltage source. A semiconductor parameter analyzer is used to read out the detection current from the collecting electrodes. Table:1 shows the detection current from two collecting electrodes I ce1 and I ce2 at different modulation voltages.](image-5.png "Figure 3 :") 4![Figure 4 : Demodulation efficiency vs. Modulation Voltage](image-6.png "Figure 4 :") 5![Figure 5 : Experimental setup for Dynamiccharacterizations Two sinusoidal waves are generated by using a function generator. One of the two sinusoidal waves is used to modulate a laser emitter and illuminate the device. The other is connected to the input of a differential amplifier. The differential amplifier outputs with 180 0 phase shift are connected to the modulating electrodes V me1 and V me2 of the device. The electric field formed in the substrate average current through the collecting electrodes V ce1 and V ce2 is read out with a Semiconductor Parameter Analyzer. For this measurement, the sinusoidal signal for laser emitter and two modulating signals are needed to use with an appropriate synchronization. At different modulation frequencies, the average current is measured under a 650nm red laser with 90% modulation depth used to illuminate the test device.The capability to separate and transfer the charges of a sensor to the corresponding output node can be expressed as a demodulation contrast. For data acquisition a LABVIEW software program was developed for the interface with PC and the experimental set-up.The dynamic demodulation contrast is the most important performance indicator for this device. The demodulation contrast depends on both the amplitude of the modulation voltages and frequencies.](image-7.png "Figure 5 :") 1V me2 (V)I ce1 (?A)I ce2 (?A)0.0000.44220.2800.2000.93219.7900.4001.89518.8200.6003.32417.4800.8005.00615.8401.0006.78314.1501.2008.28212.7401.4009.39211.7401.60010.25010.9701.80010.93010.380 F © 2013 Global Journals Inc. (US) © 2013 Global Journals Inc. (US) F Finger Typed Electrode Based Electro-Optical Demodulator Fabricated on High Resistivity Silicon © 2013 Global Journals Inc. (US) ## ACKNOWLEDGEMENTS This work has been funded by "Netcarity" European Integrated Project (www.netcarity.org) and supported by the Fondazione Bruno Kessler (FBK), Trento, Italy. * Acquisition of three dimensional image data by a scanning laser range finder IMoring THeikkinem RMyllyla AKilpela Opt.Eng 28 1989 * Handbook of Computer Vision and Applications Sensors and Imaging BJaehne HHaussecker PGeissler San Diego Academic Press 1999 1 * Time-Domain Correlation Image Sensor: First CMOS Realization of Demodulator Pixels Array SAndo AKimachi IEEE CCD/AIS Workshop Karuizawa, Japan 1999 * An Image Sensor with an In-Pixel Demodulation Function for Detecting the Intensity of a Modulated Light Signal JOhta KYamamoto THirai KKagawa MNunoshita IEEE Trans. ED 50 1 January 2003 * A CMOS Photosensor Array for 3D Imaging Using Pulsed Laser RJeremias WBrockherde GDoemens BHosticka LList PMengel Proceedings of ISSCC'01 ISSCC'01Sanfransisco (CA) USA Feb, 2001 16 * A 4x64 Pixel CMOS Image Sensor for 3D Measurement Applications OMSchrey PElkhalili 'Mengel MPetermann WBrockherde BJHosticka IEEE Journal of Solid State Circuits 39 7 July, 2004 * A 16x16-Pixel Range-Finding CMOS Image Sensor DavidStoppa LuigiViarani AndreaSimoni LorenzoGonzo GianmariapedrettiMattiamalfatti Proceedings of IEEE European Solid-State Circuits Conference-ESSCIRC'04 IEEE European Solid-State Circuits Conference-ESSCIRC'04 September, 2004 * A 50×30 pixel CMOS sensor for TOFbased real time 3D imaging" proceedings of the IEEE workshop on CCD and advanced Image sensors DStoppa LViarani ASimoni LGonzo MMalfatti GPedretti 2005 * Solid-State Time-of-Flight Camera RLange PSeitz IEEE Journal of Quantum Electronics 37 3 March 2001 * CCD-Based Range-Finding Sensor RMiyagawa TKanade IEEE Trans. ED 44 10 1997 * An all solid-state optical range camera for 3D real-time imaging with sub-centimeter depthresolution(SwissRanger) TOggier MLehmann RKaufmann MSchweizer MRichter PMetzler GLang FLustenberger NBlanc Proceeding of SPIE 5249 2003 * Mesa Imaging website * Dynamic 3D Vision RudolfSchwarte Proceedings of International Symposium on Electron Devices for Microwave and Optoelectronic Applications International Symposium on Electron Devices for Microwave and Optoelectronic Applications 15-16 November, 2001 * Novel Standard Detector using majority for guiding Photo-Generated Electrons towards Detecting Junctions WVan Nieuwenhove MVandertempel Kuijik Proceedings Symposium IEEE/LEOS Benelux Chapter Symposium IEEE/LEOS Benelux Chapter 2005 * LPancheri DStoppa NMassari MMalfatti CPiemonte G.-FDallabetta Current Assisted Photonic Mixing Devices Fabricated on High Resistivity Silicon" Sensor Conference Page 2008 * Multiple Strip Photo Mixing Demodulator for 3D Imaging Implemented on High Resistivity Silicon Gian-Franco DallaQuazi Delwar Hossain LucioBetta DavidPancheri Sagar KumarStoppa Dhar Proceedings of 7th International Conference on Electrical and Computer Engineering 7th International Conference on Electrical and Computer Engineering * Timeof Flight Range Imaging with a Custom Solid State Image Sensor RLange PSeitz ABiber RSchwarte EOS/SPIE International Symposium Munich, Germany June 14-18 1999 * DKSchroder Semiconductor material and device characterization New York Wiley 1998 2nd ed * Micro probe analysis CFRobinson 1973