Photon Detection Efficiency Calculator (SiPM)
Photon detection efficiency calculator is a handy tool that allows for a quick conversion between responsivity and Photon Detection Efficiency (PDE) in Silicon Photomultipliers (SiPMs).
Photon detection efficiency40.8%the PRODUCT of three independent factors, so it is always lower than the smallest of them. A device with 80% quantum efficiency can still detect under half the photons that reach it
Quantum efficiency80%the chance a photon creates a carrier at all
Fill factor60%the fraction of the surface that is active. The gaps between microcells are dead area, and this is usually the biggest single loss
Avalanche trigger probability85%the chance a carrier actually starts a detectable avalanche — raised by increasing the overvoltage, at the cost of more dark counts and crosstalk
Photons detected, ignoring saturation408
Microcells actually firing385.73N(1 − e^(−fired/N)) — each microcell can fire only ONCE per pulse, so a second photon hitting the same cell is lost
Saturation loss5.459%significant — the response is noticeably non-linear at this light level
Photons per microcell0.11333keep this well below 1 for a linear response
Why PDE is not quantum efficiencytwo extra factorsa photodiode's efficiency is essentially just QE. A silicon photomultiplier must also land the photon on active silicon and then win the avalanche lottery, so quoting QE alone overstates a SiPM badly
The formula
PDE = QE × fill factor × avalanche probability
Three factors multiply
A silicon photomultiplier detects a photon only if all three things happen: the photon converts to a carrier (quantum efficiency), it lands on active silicon rather than in the dead space between microcells (fill factor), and the carrier triggers a detectable avalanche (trigger probability). Because they multiply, the overall efficiency is always below the smallest of the three — a device with 80% quantum efficiency may detect well under half the light that reaches it.
Microcells saturate
Each microcell fires once per pulse and then needs time to recover, so a second photon hitting the same cell is simply lost. The response follows N(1 − e^(−k/N)) and becomes noticeably non-linear once the number of detected photons approaches a tenth of the cell count. Keeping that ratio low is the central constraint in using these devices quantitatively.