• Journal of Semiconductors
  • Vol. 42, Issue 6, 062301 (2021)
Bowen Liu1 and Jiangtao Xu2
Author Affiliations
  • 1Tianjin Key Laboratory of Imaging and Sensing Microelectronic Technology, School of Microelectronics, Tianjin University, Tianjin 300072, China
  • 2Tianjin Key Laboratory of Imaging and Sensing Microelectronic Technology, School of Microelectronics, Tianjin University, Tianjin 300072, China
  • show less
    DOI: 10.1088/1674-4926/42/6/062301 Cite this Article
    Bowen Liu, Jiangtao Xu. Modeling the photon counting and photoelectron counting characteristics of quanta image sensors[J]. Journal of Semiconductors, 2021, 42(6): 062301 Copy Citation Text show less
    The QIS conceptual illustration. An 8 × 8 × 4 spatial-temporal data cube of jots in the QIS (left) is reconstructed to a 2 × 2 data plane of pixels in the output image (right). Each data of pixels is equal to the sum of a 4 × 4 × 4 data sub-cube of jots.
    Fig. 1. The QIS conceptual illustration. An 8 × 8 × 4 spatial-temporal data cube of jots in the QIS (left) is reconstructed to a 2 × 2 data plane of pixels in the output image (right). Each data of pixels is equal to the sum of a 4 × 4 × 4 data sub-cube of jots.
    The signal chain model of quanta image sensors.
    Fig. 2. The signal chain model of quanta image sensors.
    (Color online) The PDF P[URO] as a function of URO on the condition that μe = 2 e- and un = 0.2 e- r.m.s.. N0, N1, N2, and N3 are four quantization levels corresponding to 0, 1, 2, and 3 ADU for a 2-bit QIS. The dashed lines are three quantization boundaries corresponding to 0.5, 1.5, and 2.5 e- for a 2-bit QIS.
    Fig. 3. (Color online) The PDF P[URO] as a function of URO on the condition that μe = 2 e- and un = 0.2 e- r.m.s.. N0, N1, N2, and N3 are four quantization levels corresponding to 0, 1, 2, and 3 ADU for a 2-bit QIS. The dashed lines are three quantization boundaries corresponding to 0.5, 1.5, and 2.5 e- for a 2-bit QIS.
    (Color online) The ideal D-logH response curves for 1-bit to 5-bit QISs in solid lines. And the ideal linear response curves for 1-bit to 5-bit CISs in dashed lines.
    Fig. 4. (Color online) The ideal D-logH response curves for 1-bit to 5-bit QISs in solid lines. And the ideal linear response curves for 1-bit to 5-bit CISs in dashed lines.
    (Color online) The realistic response curves for the 3-bit QIS. The different conditions of curve 1 to 5 are listed in Table 2.
    Fig. 5. (Color online) The realistic response curves for the 3-bit QIS. The different conditions of curve 1 to 5 are listed in Table 2.
    (Color online) The photon counting signal error rate SERph as a function of the mean value of incident photons μph for ideal 1-bit to 5-bit QISs.
    Fig. 6. (Color online) The photon counting signal error rate SERph as a function of the mean value of incident photons μph for ideal 1-bit to 5-bit QISs.
    (Color online) The photon counting signal error rate SERph as a function of the mean value of incident photons μph for the 3-bit QIS. The five conditions of curve 1 to 5 are listed in Table 2.
    Fig. 7. (Color online) The photon counting signal error rate SERph as a function of the mean value of incident photons μph for the 3-bit QIS. The five conditions of curve 1 to 5 are listed in Table 2.
    (Color online) The photoelectron counting signal error rate SERphe as a function of the mean value of incident photons μph for the 3-bit QIS. The five conditions of curve 1 to 5 are listed in Table 2.
    Fig. 8. (Color online) The photoelectron counting signal error rate SERphe as a function of the mean value of incident photons μph for the 3-bit QIS. The five conditions of curve 1 to 5 are listed in Table 2.
    The conceptual illustration of the Airy disk and jot array. The Airy disk diameter DA = 3.8 μm, the jot area Ajot = 1 μm2 (left) and Ajot = 0.25 μm2 (right).
    Fig. 9. The conceptual illustration of the Airy disk and jot array. The Airy disk diameter DA = 3.8 μm, the jot area Ajot = 1 μm2 (left) and Ajot = 0.25 μm2 (right).
    (Color online) The photon counting signal error rate SERph as a function of the mean value of incident photons μph for 1-bit to 5-bit QISs based on the parameters listed in Table 3.
    Fig. 10. (Color online) The photon counting signal error rate SERph as a function of the mean value of incident photons μph for 1-bit to 5-bit QISs based on the parameters listed in Table 3.
    (Color online) The photoelectron counting signal error rate SERphe as a function of the mean value of incident photons μph for 1-bit to 5-bit QISs based on the parameters listed in Table 3.
    Fig. 11. (Color online) The photoelectron counting signal error rate SERphe as a function of the mean value of incident photons μph for 1-bit to 5-bit QISs based on the parameters listed in Table 3.
    (Color online) The relationship between jot area Ajot, the mean value of incident photons μph, and integration time τ under different illuminance level Ilux for the 3-bit QIS in Fig. 11. (a) Ilux = 0.1 lux. (b) Ilux = 1 lux. (c) Ilux = 10 lux. (d) Ilux = 100 lux. (e) Ilux = 1000 lux. (f) Ilux = 10 000 lux.
    Fig. 12. (Color online) The relationship between jot area Ajot, the mean value of incident photons μph, and integration time τ under different illuminance level Ilux for the 3-bit QIS in Fig. 11. (a) Ilux = 0.1 lux. (b) Ilux = 1 lux. (c) Ilux = 10 lux. (d) Ilux = 100 lux. (e) Ilux = 1000 lux. (f) Ilux = 10 000 lux.
    SymbolParameterUnit
    kphNumber of photonsphoton, p
    kpheNumber of photoelectronselectron, e-
    kdNumber of dark signal electronse-
    keNumber of total signal electronse-
    VCGVoltage-referred jot outputmicrovolt, μV
    UCGElectron-referred jot outpute-
    VROVoltage-referred readout circuit outputμV
    UROElectron-referred readout circuit outpute-
    DNADC output digital numbersADU
    QEQuantum efficiency of the jote-/p
    CGConversion gain of the jotμV/e-
    vnVoltage-referred read noise of the readout circuitμV r.m.s.
    unElectron-referred read noise of the readout circuite- r.m.s.
    vthVoltage-referred quantizer threshold of the ADCμV
    uthElectron-referred quantizer threshold of the ADCe-
    Table 1. The parameters of the signal chain model in Fig. 2.
    NumberQE (e-/p)μd (e-) un (e- r.m.s.)
    Curve 1100
    Curve 20.800
    Curve 310.010
    Curve 4100.3
    Curve 50.80.010.3
    Table 2. Different conditions for realistic QISs in Figs. 5, 7, and 8.
    ParameterValue
    Jot size1.1 × 1.1 μm2
    Quantum efficiency79% at 550 nm
    Dark current0.16 e-/s/jot
    Read noise0.21 e- r.m.s.
    Table 3. The correlation parameters of the QIS chip in Ref. [16].
    Bowen Liu, Jiangtao Xu. Modeling the photon counting and photoelectron counting characteristics of quanta image sensors[J]. Journal of Semiconductors, 2021, 42(6): 062301
    Download Citation