Photocircuits in active pixel sensors

 

The pixel circuit (photocircuit) in active pixel sensors are often based on the charge integration method (See section 7.3.8). The photodetector structure used for APS can also be based on any of the photodetector structures described in section 7.2. The read-out circuit used in APS is often a source follower based circuit described in section 7.9.1. In fact there is very little amount of new concepts in APS.

Figure 7.24 illustrates two of the common photocircuits used in APS. In the photogate-based circuit initially the node X is reset and charge starts integrating in the the potential well under the photogate created by applying a large voltage to the gate of the photogate device. After the integration cycle the charge is transferred to node X and read out. In the photodiode-based circuit after the reset operation charge is continuously integrated on the node X until the next reset. Notice that there is no Sample&Hold stage in either of these photocircuits, and in reality while the output of some of the photocircuits are being read out other photocircuits are still integrating charge. In simple imagers this does not impose any problems, as the read-out time is much shorter than the integration time.

The advantages and disadvantages of both structures are as follows.

• Qnatum efficiency
  The quantum efficiency (QE) of a photogate is about two times less than the photodiode, because the gate usually blocks a large amount of the incoming light.
• Fixed pattern noise
  Fixed pattern noise (FPN) in photogate is more than photodiodes, as there are some random surface recombination processes, which are more effective in a photogate than in a photodiode.
• Simplicity
  A photogate requires a biasing voltage for the gate. Also a transfer gate, TX, is required for transferring the charge from the potential well to the read-out node.
• Correlated double sampling
  The reset and charge integration nodes are isolated in a photogate-based APS. Therefore, it is possible that after resetting the node X, first the reset value of node X be read out and stored on a capacitor, and then after transferring the accumulated charge to node X its value be read. By subtracting the two values a large amount of FPN can be cancelled.

  In the illustrated photodiode-based circuit the integration and reset nodes are the same. Therefore, it is not possible to perform the correlated double sampling operation in the same frame period. In order to add this facility to the photodiode-based circuit, a transfer gate similar to the photogate-based circuit, can be added to this circuit.

  
Figure 7.24: a) APS photocircuit using a photogate. b) Using a photodiode.