Time-of-flight pixel with charge storage
By introducing a charge storage unit into the time-of-flight pixel, charge deflection and storage are achieved by voltage modulation of the photogate and transfer gate, solving the problem of large space requirements, realizing efficient demodulation and storage, supporting correlation double sampling under global exposure, and improving sensor performance.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- PMDTECHNOLOGIES
- Filing Date
- 2021-11-02
- Publication Date
- 2026-07-03
AI Technical Summary
Existing time-of-flight sensors have large space requirements for time-of-flight pixels, making it difficult to achieve efficient demodulation and storage of photogenerated carriers within a limited space.
A charge storage unit is introduced into the time-of-flight pixel. Charge deflection and storage are formed by voltage modulation of the photogate and transfer gate. Combined with the readout diode, charge demodulation and storage are realized, reducing space requirements.
It significantly reduces the spatial requirements of time-of-flight pixels, while achieving efficient demodulation and storage of photogenerated carriers, supporting correlation double sampling under global exposure, and improving sensor performance.
Smart Images
Figure CN116601771B_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to a time-of-flight pixel having a charge storage section below a modulation gate and a time-of-flight sensor including the corresponding pixel.
[0002] The time-of-flight sensor according to the invention is particularly useful in time-of-flight camera systems, wherein the time-of-flight camera system obtains distance from the phase shift of emitted and received radiation. It is suitable as a time-of-flight or 3D camera, particularly a PMD camera having a photomixing detector (PMD) (e.g., the PMD described in DE 197 04 496A1). Summary of the Invention
[0003] The purpose of this invention is to reduce the space requirement of time-of-flight pixels in time-of-flight sensors.
[0004] The time-of-flight pixel according to the invention achieves this objective and enables demodulation and storage of photogenerated carriers under a single photogate to be executed. This significantly reduces space requirements. Attached Figure Description
[0005] The attached diagram schematically illustrates:
[0006] Figure 1 This is a top view of the pixel structure according to the present invention;
[0007] Figure 2 It shows a cross-sectional view of a pixel structure with a potential distribution;
[0008] Figure 3 This illustrates another embodiment of the structure according to the invention;
[0009] Figure 4 It shows an embodiment for front lighting;
[0010] Figure 5 It illustrates an embodiment for backlighting;
[0011] Figure 6 It shows an embodiment with back illumination having an optical isolator;
[0012] Figure 7 This illustrates an embodiment of back illumination with a scattering element; and
[0013] Figure 8 The example shown is a backlighting system with an optical isolator and a scattering element. Detailed Implementation
[0014] like Figure 1As shown, according to the invention, photogenerated charges in silicon are deflected and collected thereby by modulation voltages on photogates GA and GB, and by the formation of known carrier swings in the directions of storage regions MA and MB, respectively. Storage regions MA and MB are configured such that as long as the corresponding transfer gates TXA or TXB have a low potential, the carriers are permanently (i.e., independent of the voltage applied to photogates GA and GB) retained in storage regions MA and MB. When transfer gates TXA and TXB have a high potential and the corresponding photogates have a low potential, the charges collected in the storage regions discharge toward the readout diodes DA and DB.
[0015] In addition, the drain gate GD with the attached drain diode DD allows for concentrated, limited suppression below the storage regions MA and MB.
[0016] On a semiconductor substrate with appropriate p-type or n-type doping, two photogates GA and GB are formed in a centrally located photosensitive region FAB within a time-of-flight pixel. Below the photogates GA and GB and defining the photosensitive region FAB, memory regions MA and MB are disposed in the semiconductor substrate, wherein the photogates GA and GB only partially cover the memory regions MA and MB.
[0017] Adjacent to the storage regions MA and MB, corresponding transfer gates TXA and TXB are arranged, which in turn are adjacent to the readout diodes DA and DB. In a direction perpendicular to the photogates GA and GB, the transfer gates TXA and TXB, and the diodes DA and DB, a drain gate (GD) is arranged, which connects the mixer region or photosensitive region FAB of the photogates GA and GB to the drain diode DD (discarded node) in a switchable manner.
[0018] The storage regions MA and MB are preferably partially fabricated below the photogates GA and GB by means of enhanced n-type injection. These injections result in a local increase in the electrostatic potentials below the photogates GA and GB. These regions can be used as storage regions MA and MB for intermediate storage in demodulating optoelectronics.
[0019] Therefore, photogates GA and GB are used to achieve a dual function: demodulating photogenerated carriers on the one hand, and storing these carriers in the middle on the other, in order to realize the function based on the principle of correlation double sampling using simultaneous global electron aperture (global shutter). This structure can be used for front illumination FSI and back illumination BSI of pixels.
[0020] Figure 2The structural and potential cross-sections through the time-of-flight sensor or time-of-flight pixel according to the invention are shown when a typical applied voltage is applied. In addition to the structural cross-sections, potential cross-sections for the following modes are also shown: a) concentration, b) holding, and c) readout.
[0021] a) Centralized mode:
[0022] During the concentration phase, transfer gates TXA and TXB remain at low potentials, thus separating storage regions MA and MB from readout diodes DA and DB. Drain gate GD is also at a low potential, separating the drain contact from the mixer region or photosensitive region FAB. Photogates GA and GB form a well-known carrier swing. Correspondingly demodulated photogenerated carriers enter storage regions MA and MB, respectively. During further concentration phases, the carriers stored there cannot escape in the direction of readout diodes DA and DB, nor in the direction of photosensitive region FAB. Therefore, the storage capacity of storage regions MA and MB is independent of the voltage applied to photogates GA and GB. Thus, during the concentration phase, there is a continuous concentration of carriers in storage regions MA and MB.
[0023] b) Maintaining the mode:
[0024] In hold mode, the two photogates GA and GB are held at low potentials, and the transfer gates TXA and TXB separate the storage regions MA and MB from the readout diodes DA and DB. To prevent further concentration of charge carriers in the storage regions, the drain gate GD is now switched to a high potential, resulting in a direct connection between the mixer regions GA and GB and the drain diode DD. The photogenerated charge carriers now discharge in the direction of the drain diode DD, meaning they are no longer available for further concentration in the storage regions MA and MB.
[0025] c) Readout mode:
[0026] The readout mode is used to transfer carriers concentrated below the storage regions MA and MB along the directions of the readout diodes DA and DB. For this purpose, the photogates GA and GB are held at a low potential. The drain gate GD is also held at a high potential and will discharge all photogenerated carriers towards the drain diode DD during the readout mode.
[0027] Now, the transfer gates TXA and TXB are switched to high potential, causing a potential gradient to form in the direction of the readout diodes DA and DB. This causes the charge stored in the memory regions MA and MB to flow to the readout diodes DA and DB. With proper timing, the readout can be performed as a noise-reduced correlated double-sampled CDS.
[0028] Figure 3An implementation of a time-of-flight pixel comprising only a single diode node DA (one tap) is shown. In this embodiment, one of the two readout channels, A or B, is omitted. In this example, the pixel comprises a photogate GA and a storage node for channel A, while channel B is formed by a drain gate GD and a drain diode DD. Therefore, the modulation signal is provided at the photogate GA for channel A, while the complementary modulation signal is provided at the drain gate GD. Photogenerated carriers are thus concentrated in the storage region MA or discharged via the drain diode DD.
[0029] In principle, designs with more than two diode nodes can also be envisioned.
[0030] Figure 4 The image shows illumination (front illumination, FSI) of the time-of-flight pixel according to the invention from the front. Figure 5 The illumination from the rear is shown (rear illumination, BSI).
[0031] Figure 6 An embodiment is shown that includes storage areas MA and MB with optical isolation for back-illuminated BSI, wherein a trench or buried reflector ISO is used for optical isolation.
[0032] Figure 7 A variation of the scattering element SR, including a scattering element on the back side of a time-of-flight pixel, is shown. In this embodiment, one or more scattering elements SR, such as trench or pyramid structures, can be configured to increase quantum efficiency.
[0033] like Figure 8 As shown, the scattering element SR can be combined with the optical isolation element ISO to optically isolate the storage areas MA and MB.
[0034] List of reference numerals
[0035] GA, GB photogates A and B
[0036] MA and MB storage areas A and B
[0037] TXA and TXB transfer gates A and B
[0038] DA, DB readout diodes A and B
[0039] GD drain gate
[0040] DD drain diode
Claims
1. A time-of-flight pixel, comprising: At least one modulation gate (GA, GB) has a photosensitive region (FAB) and a memory region (MA, MB). The memory regions (MA, MB) have n-type localized enhanced doping below the modulation gates (GA, GB) and define the photosensitive regions (FAB) of the modulation gates (GA, GB). At least one transfer gate (TXA, TXB) is adjacent to the memory region (MA, MB) of the modulation gate (GA, GB); At least one readout diode (DA, DB) is located after the transfer gate (TXA, TXB); At least one drain gate (DG) is adjacent to one side of the photosensitive region (FAB) of the modulation gate (GA, GB); and At least one drain diode (DD) is located after the drain gate (DG).
2. The time-of-flight pixel according to claim 1, wherein, It is configured for back illumination and includes a scattering element (SE) for scattering light incident on the back of the time-of-flight pixel.
3. The time-of-flight pixel according to claim 2, wherein, The scattering element (SE) is formed as a groove or pyramid-shaped structure.
4. The time-of-flight pixel according to any one of the preceding claims, wherein, The storage areas (MA, MB) are protected from direct light incidence from the back by an optical isolation structure (ISO).
5. A time-of-flight sensor comprising time-of-flight pixels according to any one of claims 1 to 4.
6. A time-of-flight camera or time-of-flight camera system, comprising the time-of-flight sensor according to claim 5.