Compute responsivity, NEP, specific detectivity D*, SNR, shot noise, and thermal noise in real time. Visualize noise contributions and SNR vs optical power curves.
Responsivity:
$$R = \frac{\eta e}{h\nu}= \frac{\eta e \lambda}{hc}\quad \text{[A/W]}$$Shot noise:
$$i^2_{shot}= 2e(I_{photo}+ I_d)\cdot B$$Johnson (thermal) noise:
$$i^2_{thermal}= \frac{4k_B T B}{R_L}$$SNR:
$$\mathrm{SNR}= \frac{(R\cdot P)^2}{i^2_{shot}+i^2_{thermal}}$$Specific detectivity:
$$D^* = \frac{\sqrt{A \cdot B}}{\mathrm{NEP}}\quad \text{[cm}\cdot\sqrt{\text{Hz}}\text{/W]}$$The core conversion efficiency of a photodetector is defined by its Responsivity. It links the quantum efficiency (η), a fundamental property of the material, to the wavelength (λ) of the incident light.
$$R = \frac{\eta e}{h\nu}= \frac{\eta e \lambda}{hc}\quad \text{[A/W]}$$Where: $R$ = Responsivity (A/W), $\eta$ = Quantum Efficiency, $e$ = Electron charge, $h$ = Planck's constant, $\nu$ = Optical frequency, $\lambda$ = Wavelength, $c$ = Speed of light.
The total noise current determines the minimum detectable signal. It is the root-sum-square of three independent noise sources: Shot Noise from the signal and dark current, Johnson (Thermal) Noise from the load resistor, and a constant noise floor.
$$i^2_{total}= i^2_{shot}+ i^2_{johnson}+ i^2_{other}$$ $$i^2_{shot}= 2e(I_{photo}+ I_d)\cdot B, \quad i^2_{johnson}= \frac{4k_B T}{R_L}\cdot B$$Where: $I_{photo}=R \cdot P$ = Photocurrent, $I_d$ = Dark Current, $B$ = Electrical Bandwidth, $k_B$ = Boltzmann constant, $T$ = Temperature, $R_L$ = Load Resistance. The Signal-to-Noise Ratio is then $SNR = I_{photo}/ i_{total}$.
Fiber Optic Communications: Maximizing SNR at the receiver is critical for data integrity. Engineers use this exact noise model to choose between PIN photodiodes and Avalanche Photodiodes (APDs), balancing responsivity against excess noise. The bandwidth (B) parameter directly relates to the data rate.
LiDAR and Laser Rangefinding: These systems detect extremely weak reflected pulses. A low Noise-Equivalent Power (NEP) is paramount. Designers optimize the active area (A) and cool the detector to minimize dark current, as you can explore in the simulator, to achieve longer detection ranges.
Scientific Spectroscopy: In instruments like spectrophotometers or astronomical spectrometers, detecting faint spectral lines requires minimizing all noise sources. High load resistance (R_L) and cooling are standard practices to measure precise light intensities across wavelengths.
Biomedical Imaging: Techniques like fluorescence microscopy or pulse oximetry rely on detecting low-light-level signals from biological tissue. Understanding the trade-off between detector area (A), bandwidth (B), and noise is essential for achieving clear, high-contrast images without damaging samples with excessive light.
There are several key points you should be especially mindful of when starting to use this simulator. First, the misconception that "setting the quantum efficiency η to 100% is always optimal." While a higher η does increase responsivity, in real sensors η varies significantly with wavelength. For example, the η of a silicon photodiode is high in the visible to near-infrared range (~1000nm) but drops sharply at the telecom wavelength of 1550nm. By moving the wavelength slider in this tool and changing η, you can get a feel for this trade-off.
Next, the relationship between bandwidth B and noise. Increasing the bandwidth (e.g., from 1MHz to 10MHz) worsens the NEP (makes the value larger) in the calculation, right? This is because "you are now picking up noise from a wider frequency range," not because the sensor's performance instantly degrades. A wide bandwidth is essential for handling high-speed signals, so "setting a narrow bandwidth just because you looked at the NEP" is not a universal solution.
Finally, the gap between simulation and actual measurement. This tool is based on an ideal model. Real devices have additional factors not included here, such as 1/f noise (flicker noise), noise added by the amplifier itself, and variations in dark current due to temperature. Even if the tool indicates "thermal noise is dominant," it's not uncommon for the noise from the amplification stage to become the bottleneck in practice. Use the simulation results as "a starting point for design and for understanding trends."