Depth Sensing Pixel Shared Readout Port Design

The patent combines multiple readout functions into a single shared readout port that sequentially handles both in-phase and out-of-phase charges from depth sensing pixels. This merging approach maintains signal processing capability while reducing device size and manufacturing complexity by eliminating the need for separate readout circuits for each charge type.

The patent implements periodic action by sequentially reading out in-phase charges during a first time period and out-of-phase charges during a second time period through the shared readout port. This time-division multiplexing approach allows a single readout port to handle multiple signal types without requiring simultaneous processing paths, thereby reducing device complexity while maintaining measurement precision.

The patent segments the sensor into distinct regions: a first region containing image sensing pixels and a second region containing depth sensing pixels. This spatial segmentation allows each pixel type to be optimized independently while maintaining their composite functionality, reducing the need for precise alignment between overlapping structures and thereby lowering fabrication precision requirements.

The patent transitions from a vertical stacking approach (positioning depth sensing pixels behind image sensing pixels along the light path) to a planar arrangement where depth sensing pixels are positioned in a second region adjacent to or around the first region containing image sensing pixels. This dimensional change eliminates the need for precise vertical alignment through multiple layers, significantly reducing fabrication precision requirements while maintaining composite pixel functionality.

The patent merges the readout paths for in-phase and out-of-phase charges into a single shared readout port. This ensures that both charge types undergo identical signal processing operations through the same circuitry, eliminating mismatches in gain, offset, and noise characteristics that would occur with separate readout ports, while maintaining signal processing accuracy through sequential processing.

The patent uses correlated double sampling where the out-of-phase charge readout serves as a reference copy that mirrors the signal processing path of the in-phase charge readout. By processing both charge types through the same shared readout port in sequence, the system creates a matched pair of measurements that can be differential processed to eliminate systematic errors, ensuring signal processing consistency while maintaining accuracy.