Silicon Photomultipliers are now very well-established in key markets such as medical imaging, where they are the detector-of-choice for combined MR/PET and MR/CT systems as well as "homeland security" portable spectroscopic radiation detection and dosimetry instruments. However, there are many other potential applications and so new users are learning the basics of SiPMs and how to get the best from the technology.
With this in mind we have compiled a list of Frequently Asked Questions and provide links from these to application notes and support pages on our own website where topics are discussed in more detail:-
- What is an SiPM?
- What are SiPMs suitable for?
- At which wavelengths are SiPMs sensitive?
- Can SiPMs detect X-rays and gamma-rays?
- Are SiPMs easy to damage or destroy?
- What is the active area of Broadcom SiPMs?
- What do breakdown voltage and overvoltage mean?
- What is the maximum voltage that can be applied to Broadcom SiPMs?
- Which bias polarity do you recommend?
- What is Photon Detection Efficiency (PDE)?
- Does the Photon Detection Efficiency (PDE) change with temperature?
- How do you calculate the responsivity of SiPMs?
- What is crosstalk in SiPMs?
- What is afterpulsing in SiPMs?
- What is the recovery/recharge time of an SiPM?
- How can I evaluate Broadcom SiPMs?
- What do I need to operate an SiPM?
- Can I buy SiPMs in interposer PCBs?
- Can you supply the footprints for your SiPMs?
- What does an SiPM pulse look like?
- What is the dynamic range of an SiPM?
- Does the SiPM gain change with temperature?
- Can I operate an SiPM in ambient light?
- Do you have tips for testing your SiPMs?
- What are the most common circuit schematics to read out an SiPM?
- How do I digitise an SiPM signal?
- What is the fill factor of Broadcom SiPMs and SiPM arrays?
- How can I buy Broadcom SiPMs?
- I would like to learn more about Broadcom SiPMs. Where can I get more information?
- Support page: What is an SiPM?
A silicon photomultiplier (SiPM) is a semiconductor device that can detect light (photons) and convert it into a current signal. Its basic building blocks are single photon avalanche diodes (SPADs), also referred to as Geiger-mode avalanche photodiodes (GM-APD). GM-APDs are biased with a reverse voltage that is high enough to create a Geiger discharge when a photon is absorbed in the active (high field) region of the diode. The minimum voltage at which the Geiger breakdown occurs is referred to as breakdown voltage (VBD). The Geiger discharge provides a high multiplication factor (gain) and a well defined charge output, resulting in excellent single-photon resolution. To stop the discharge current (and the self-sustainable current), quenching is necessary. In an analog SiPM, this is conventionally realised by using a quenching resistor (Rq) in series to the SPAD. Once an avalanche occurs in the SPAD, the voltage drop across Rq effectively reduces the voltage across the diode below its breakdown voltage. Subsequently, the SPAD recharges over Rq and is ready to detect photons again.
A SPAD with its dedicated Rq is often referred to as a microcell. An SiPM is an array with several ten to several ten thousand microcells connected in parallel. The microcell density of an SiPM depends on the SPAD pitch, which is typically between 10 μm to 100 μm. The resulting size of the SPAD (partially) determines many electro-optical performance parameters of the SiPM, such as the photon detection efficiency (PDE) and the gain.
SiPMs are best suited to detect light (photons) from the near UV over the visible to the near infrared electromagnetic spectrum.
Higher-energy photons, such as X-rays and gamma-rays, can be detected using scintillators. Scintillators convert high-energy photons into a number of UV or visible photons that are proportional to the energy of the incoming X-ray or gamma-ray.
SiPMs are applied in a large variety of applications in fields such as the following:
- Medical imaging (PET, gamma cameras)
- Radiation detectors and dosimeters
- Microscopy (for example, confocal laser scanning microscopy, two-photon microscopy, fluorescence microscopy)
- Astrophysics
- High-energy physics (HEP)
- X-ray imaging
- Biophysics (for example, fluorescence detection, cytometry)
- Time-of-Flight (ToF) and LiDAR
At which wavelengths are SiPM sensitive?
- Support page: SiPM Photon Detection Efficiency
The sensitivity of Broadcom® SiPMs ranges from 250 nm to 980 nm, with the exact PDE spectrum determined by the technology. Therefore, SiPMs are best suited to detect light (photons) from the near UV over the visible to the near infrared electromagnetic spectrum.
Broadcom’s SiPM portfolio offers dedicated solutions for the detection of UV/blue/green (NUV-MT SiPM) light and for yellow/red/IR (NIR SiPM) light.
The PDE spectrum of Broadcom’s NUV-MT SiPM ranges from 250 nm to 900 nm.
The PDE spectrum of Broadcom’s NIR SiPM ranges from 500 nm to 980 nm.
Besides the peak sensitivity and spectral sensitivity, both technologies differ in the other key parameters, such as noise characteristics and the dynamic range.
Can SiPMs detect X-rays and gamma-rays?
Although SiPMs can directly detect X-rays, the efficiency is very low and signal generation is dominantly based on scattering.
However, SiPMs can be used with the majority of scintillators, and, therefore, they can be used to detect X-rays and gamma rays with high efficiency.
We have broad experience in the field or X-ray and gamma-ray detection and can support your application-specific needs.
Are SiPMs easy to damage or destroy?
SiPMs are very robust devices both mechanically and optically. When the recommended current limit is obeyed, SiPMs are virtually indestructible. Accidental illumination, daylight illumination, or the wrong bias polarity is usually not critical and does
not damage SiPMs. Furthermore, SiPMs can be used immediately after being exposed to bright light and need no time to recover.
What is the active area of Broadcom SiPMs?
- Product brief: AFBR-S4N Series – Broadcom High-Sensitivity NUV Silicon Photomultipliers
- Product brief: AFBR-S4 SiPM Family – Silicon Photomultipliers
Depending on the technology, Broadcom offers various SiPM form factors and channel areas.
Blue-optimised SiPMs
NUV-MT SiPMs are available in three active area variants:
- AFBR-S4NP014M
- AFBR-S4N44P014M
- AFBR-S4N66P014M
The 4x4 mm² SiPMs are also available as SiPM arrays in the following arrangements:
- 2 x 2 AFBR-S4N44P044M
- 4 x 4 AFBR-S4N44P164M
The 6x6 mm² SiPMs are available in a 2 x 1 array configuration (AFBR-S4N66P024M).
All Broadcom NUV-MT SiPMs use a SPAD technology with a 40-μm pitch, providing an optimal balance between efficiency and linearity.
Red/NIR optimised SiPMs
Broadcom’s high-sensitivity NIR SiPMs (AFBR-S4P11P012R) are available in a 2 x 1 dual-channel package. The channels are integrated in a monolithic array of 1 x 0.5 mm2 active area per channel. The package design allows connecting the two individual cathodes on a customer PCB to obtain a 1x1 mm2 SiPM.
Broadcom’s fast NIR technology is designed for d-ToF and LiDAR applications and is available in the following configurations:
- 1mm², 1 x 2 monolithic line array with a 0.5 x 1.0 mm² channel area (AFBR-S4P0102L3R)
- 1 x 24 monolithic line array with a 1.1 x 0.233 mm² channel area (AFBR-S4P0124P3TA)
Broadcom NIR SiPMs provide ultra-high SPAD densities with a pitch of 12.5 μm, resulting in excellent dynamic range.
What do breakdown voltage and overvoltage mean?
- Support page: Breakdown Voltage
- Product brief: AFBR-S4XX – Brief Introduction to Silicon Photomultipliers
- Application note: AFBR-S4XX – Working with Broadcom SiPMs
SiPMs are arrays of single photon avalanche diodes (SPADs) in a parallel circuit.
Breakdown voltage refers to the minimum voltage that is required to cause a Geiger breakdown in the SPADs; in other words, the minimum voltage that must be applied to an SiPM to operate it properly.
Overvoltage is the excess voltage above the breakdown voltage. Most SiPM parameters are a function of the applied overvoltage.
The overall bias voltage supplied to an SiPM is VBias = Vbreakdown + VOV.
What is the maximum voltage that can be applied to SiPMs? Related support page Breakdown Voltage
The maximum voltage that can be applied to SiPMs depends on the technology.
- For Broadcom’s NUV-MT SiPMs, the maximum overvoltage is 16V, resulting in a maximum bias voltage of 48.5V at 25°C.
- Broadcom’s NIR SiPMs can be operated at up to 12V overvoltage.
For details, refer to the data sheet of your device.
Which bias polarity do you recommend?
- Support page: SiPM Biasing and Readout
- Application note: AFBR-S4XX – Working with Broadcom SiPMs
Depending on the connection scheme and therefore the used polarity of the SiPM bias, the signal can be obtained from the anode or the cathode. However, Broadcom recommends using the following convention for readout
- Broadcom SiPMs with AFBR-S4N and AFBR-S4K serial numbers (p-on-n substrate), such as NUV-MT SiPMs, prefer the anode for signal extraction and therefore a positive bias via the cathode.
- Broadcom SiPMs with AFBR-S4P serial numbers (n-on-p substrate), that is NIR SiPMs, prefer the cathode for signal extraction and a negative bias via the anode.
What is Photon Detection Efficiency (PDE)?
- Support page: Photo Detection Efficiency
- Application note: AFBR-S4XX – Brief Introduction to Silicon Photomultipliers
Photon detection efficiency (PDE) is the SiPM parameter that refers to sensitivity and is the ratio of detected (primary) photoelectrons and the number of incident photons. An SiPM’s PDE is given as follows:
PDE = QE × FF × ε
With QE the quantum efficiency, FF the geometric fill factor given by the ratio between the sum of the SPADs’ active area and the total SiPM area, and ε the probability of triggering an avalanche.
Broadcom SiPMs provide best-in-class PDE:
- 63% at 420 nm for NUV-MT SiPMs
- 37% at 750 nm for high-sensitivity NIR SiPMs
Does the Photon Detection Efficiency (PDE) change with temperature?
Support pages: Photo Detection Efficiency and Temperature Effects
Yes, there is a measurable influence of the SiPM temperature on its PDE. The exact temperature behavior depends on the underlying technology.
The temperature coefficient of the PDE depends on the wavelengths and applied overvoltage and tends to increase for longer wavelengths. More details on the NUV-MT temperature behavior of the PDE are given in Schmailzl et al. 2022 JINST 17 P12009.
The plot shows the measured PDE at three wavelengths (365 nm, 463 nm, and 633 nm) at 9V overvoltage in the temperature range between –30°C and 60°C. Overall, the temperature coefficient of the PDE is a small value and in the order of 10-3/°C.
How do you calculate the responsivity of SiPMs?
- Application note: AFBR-S4XX – Brief Introduction to Silicon Photomultipliers
- Application note: AFBR-S4NxxPyy4M – SiPM Characteristics for PMT Users
The photon detection efficiency (PDE) is a sensitivity parameter usually used for SiPMs and can be directly compared with the QE of photodiodes, APDs, and PMTs.
However, often, the radiant sensitivity is provided as a parameter of the sensitivity of a photodetector. The radiant sensitivity (S [A/W]) of an SiPM can be calculated from the PDE, gain, and correlated noise components (PCT: crosstalk probability and PAP: afterpulsing probability) using the following equation:
With e being the electron charge, λ being the wavelength of light, h being the Planck’s constant, and c being the speed of light in a vacuum.
Note that for a comparison of photosensors with and without gain (photodiodes, APDs, PMTs, SiPMs), QE and PDE are more useful parameters than the radiant sensitivity S. This is due to the fact that the gain and noise parameters are included in the preceding equation.
For example, a sensor with very poor PDE can show a huge radiant sensitivity when the gain or correlated noise is very high.
This, however, would not translate into an acceptable signal-to-noise ratio.
The gain of the sensors is one for photodiodes, tens to hundreds for APDs, and millions for SiPMs and PMTs.
- Application note: AFBR-S4XX – Brief Introduction to Silicon Photomultiplier
Crosstalk (CT) in SiPMs is a form of correlated noise. During the avalanche process in a SPAD, optical photons (from recombination) can be emitted. These photons can propagate through the silicon or the protective layer of the SiPM package and trigger a neighboring cell. As a consequence, the pulse amplitude (and charge) is increased by the contribution of additionally triggered SPADs.
What is afterpulsing in SiPMs?
- Application note: AFBR-S4XX – Brief Introduction to Silicon Photomultipliers
Afterpulsing (AP) in SiPMs is a form of correlated noise. Charge carriers of an avalanche can be trapped in the silicon and released after some time. As the SPAD is already (partly) recovered, a second delayed pulse can be observed next to an
initial one.
What is the recovery/recharge time of an SiPM?
- Application note: AFBR-S4XX – Brief Introduction to Silicon Photomultipliers
The recharge of a SPAD follows an exponential behavior proportional to e–t/T with being the recharge time constant. After the recharge time, the signal is back to 1/e = 37%, and after (approximately) five time constants, the signal is back below 1% of the peak amplitude.
The recharge time constant is determined by the capacitance of the SPAD and the quenching resistor value. Smaller SPADs tend to have shorter recharge times and therefore provide a higher rate capability.
- For Broadcom NUV-MT SiPMs, Tau (τ) is 55 ns.
- For Broadcom NIR30 SiPMs, Tau (τ) is 15 ns.
How can I evaluate Broadcom SiPMs?
- Application note: AFBR-S4E001 – Evaluation Kit for the AFBR-S4NxxPyy4M SiPM Family
Our AFBR-S4E001 NUV SiPM evaluation kit includes an SiPM of all three available active area variants (2x2 mm², 4x4 mm², and 6x6 mm²). In addition, it ships with two types of readout PCBs:
- A DC readout without amplifier for direct access of the SiPM pulse
- An amplifier PCB with an amplified energy channel and a pole-zero filtered timing output
What do I need to operate an SiPM?
SiPMs are simple to test and evaluate. The AFBR-S4E001 SiPM evaluation kit comprises an amplified energy and pole-zero filtered timing output. For applications with high dynamic range requirements and bright signals, it is shipped with a DC readout PCB that allows reading out the SiPM directly without amplifier or bandwidth limitation.
A voltage supply (~ 50V) and an oscilloscope are all the equipment you need to test the SiPM with its evaluation kit. If you also want to use the amplifier PCB, you will need a dual-source voltage supply to (5V) to bias the amplifiers.
NOTE: Cables are not shipped with the evaluation kit, you must provide these.
Can I buy SiPMs in interposer PCBs?
Broadcom does not sell single SiPMs on an interposer PCB. The Broadcom AFBR-S4E001 evaluation kit includes three SiPMs on an interposer PCB to be used with the dedicated amplifier and/or readout PCB.
If you need additional SiPMs on an interposer PCB to be used with your evaluation kit, contact AP Technologies and we will be happy to work with you to facilitate this.
Can you supply the footprints for your SiPMs?
Yes, contact AP Technologies for this information.
What does an SiPM pulse look like?
- Application note: AFBR-S4NXX – SiPM Waveform and Bandwidth Considerations
The rising edge of an SiPM signal is caused by the Geiger discharge and is determined by the diode resistance in a conducting state.
The falling edge of the signal is determined by the recharge of the SPAD (and parasitics) and is characterized by two exponential decays (a fast and a slow component).
The following figure shows the response to a single photon for an SiPM with Broadcom’s NUV-MT technology with three different active areas (without amplifier).
Furthermore, the effective pulse measured at the output of the SiPM also depends on the detected light pulse and is a convolution of the SiPM pulse shape with the light pulse.
What is the dynamic range of an SiPM?
- Support page: SiPM Linearity & Saturation
- Application note: AFBR-S4NxxPyy – SiPM Dynamic Range, Linearity, and Saturation
The dynamic range of an SiPM addresses the maximum number of photons that can be detected by the SiPM and depends on the properties of the incident light pulse.
- For short light pulses (approximately < recharge time constant), the dynamic range is determined by the number of SPADs and the PDE.
- For light pulses longer than the recharge time constant and CW light, SPADs can recharge and be triggered by further photons, which effectively increases the dynamic range.
Does the SiPM gain change with temperature?
- Support page: SiPM Temperature Effects
The gain of an SiPM does not change with temperature if the bias voltage is adapted such that the temperature coefficient of the breakdown voltage is accounted for (meaning that the SiPM is operated at a constant overvoltage).
However, customers may see a reduced signal amplitude/charge pulse response at high temperatures due to saturation effects. This effect can be prominent in NIR SiPMs, which are qualified at high temperatures and show higher DCR compared to NUV-MT series SiPMs.
At high temperatures (for example above 80°C) a non-negligible number of SPADs can be triggered by dark counts. This is observed by a DC current offset of the signal baseline. In numerous applications where high temperatures (and often also bright ambient light conditions) must be handled, AC coupling is used to suppress this DC baseline component. The high DCR at high temperatures effectively reduces the dynamic range by occupying SPADs. These SPADs cannot contribute to the overall SiPM signal when a signal light pulse is detected by the SiPM. The reduced number of contributing SPADs results in a reduced signal amplitude/charge output.
Broadcom NIR SiPMs use a 12.5-μm SPAD pitch, providing excellent dynamic range and a target to minimise saturation effects under bright ambient light and/or high-temperature conditions.
Can I operate an SiPM in ambient light?
An SiPM is a highly sensitive photon detector and thus will saturate when exposed to bright ambient light. However, exposing an SiPM to ambient light will not damage the SiPM if a suitable current compliance limit is applied.
If SiPMs are required to be operated under bright ambient light, narrow-band pass filters can be used to suppress the ambient light of wavelengths different from the signal wavelength. Follow these tips to operate an SiPM in ambient light:
- Always use a current compliance limit (of a few mA; the exact value depends on the type of SiPM).
- Use a narrow-optical-band-pass filter matched to the signal wavelength to suppress a broad spectrum of ambient light.
- AC coupling (for example, using a high-pass filter or balun transformer) can suppress a DC offset caused by frequent SiPM triggers by ambient light.
Note that ambient light will constantly trigger SPADs on the SiPM, which effectively reduces the available dynamic range of the detector. In an AC readout scheme, the increasing saturation with increasing ambient light brightness may appear to reduce the signal amplitude (given a constant signal power). This signal amplitude reduction is not due to gain changes; rather, it is caused by a reduced dynamic range (similar to the effect caused by high DCR at high temperatures, Q 1.23).
Do you have tips for testing your SiPMs?
- Application note: AFBR-S4XX – Working with Broadcom SiPMs
- Application note: AFBR-S4NxxPyy4M – Single-Photon Measurements
- Application note: AFBR-S4NxxPyy4M – NUV-MT SiPM Performance Correlation
SiPMs are easy to use and to test. The following are a few tips for first-time users:
- Current compliance: Make sure to apply a current compliance limit to the SiPM to prevent high currents to the device when involuntarily applying wrong bias polarities of exposing the SiPM to ambient light. To start with, you can use 3 mA as the current compliance limit.
- Consistency checks: It is good practice to ensure proper measurement conditions:
- Firm cable connections
- Dark conditions
- Correct oscilloscope settings
Before starting a detailed test, a brief setup consistency check can be performed, such as the following:
- Does the SiPM behave like a diode (diode test on multimeter)?
- Is the SiPM dark current reading on an SMU (if used) reasonable?
- Can dark pulses be observed on an oscilloscope?
Operational point: SiPMs are complex sensors with most of their performance parameters depending on the applied overvoltage. To find the best operation point in your application, we strongly recommend a bias sweep over the full specified overvoltage range.
What are the most common circuit schematics to read out an SiPM?
- Support page: Biasing and Readout
The ideal way to read out an SiPM strongly depends on the application and its requirements in the signal. The SiPM can be DC or AC coupled; impedance matching and filtering may be necessary; and signal amplitudes can be increased using
amplifiers or even reduced using attenuators.
In most applications, the SiPM signal is amplified and filtered/shaped. A very common readout design therefore comprises a transimpedance amplifier (TIA). TIAs can be realised based on a wide range of operational amplifiers. The choice of operational amplifier should be based on the performance requirements in the end application.
The SiPM signal can be coupled to the TIA input DC or AC, and each variant has its own advantages and complications.
If only moderate gain (few tens of V/A) and/or low bandwidth is required, a single-stage transimpedance amplifier can be used:
Schematic of a single-stage transimpedance amplifier circuit
If the required gain at a given bandwidth cannot be achieved by a single-stage TIA, a second amplification stage can be implemented. For the second stage, a TIA design can be used, but other options, such as voltage amplifier designs, are also
suitable options.
Schematic of 2nd-stage transimpedance and voltage amplifier circuits
As previously discussed, Broadcom's AFBR-S4E001 evaluation kit includes a high performance two-stage amplifier design. The SiPM is biased via its cathode with buffer capacitors used in the bias line. The SiPM signal is first fed into a current divider (R6 and R8), which effectively reduces the gain of the first TIA stage. With a TIA gain of 200 (given by the feedback resistor R5) and the current divider, the gain of the first stage is 82 V/A. The pre-amplified voltage signal is subsequently split into two channels. The “energy” signal (full DC signal via OUT1) is fed into a voltage follower (unity gain), whereas “timing” (pole-zero filtered via OUT2) is fed into a second TIA stage.
Both outputs are impedance matched to 50Ω.
Simplified Readout Scheme of the Broadcom AFBR-S4E001
How do I digitise an SiPM signal?
- Support page: Biasing and Readout
The SiPM current signal is typically digitised by either of the following:
- Charge integration or amplitude sensing
- Time-over-threshold measurements if the energy must be assessed
For timing measurements, the typical time at which the signal exceeds a predefined threshold (for example, versus a trigger or second detector) is used.
What is the fill factor of Broadcom SiPMs and SiPM arrays?
The fill factor within the active area is included in the PDE.
The achievable fill factor of SiPM arrays, meaning the ratio between the active area of the SiPM channel and the overall covered area, depends on the used product. Generally, SiPMs with larger active areas also result in a higher fill factor if single-channel devices are considered. However, the fill factor can be increased by using Broadcom SiPM arrays. Highest fill factors are achieved with the Broadcom 16-channel NUV-MT array (AFBR-S4N44P164M).
The following table provides an overview of the device fill factor for Broadcom NUV-MT SiPM products.
| AFBR- S4N22P014M |
AFBR- S4N44P014M |
AFBR- S4N44P044M |
AFBR- S4N44P164M |
AFBR- S4N66P014M |
AFBR- S4N66P024M |
|
| Active Area of SiPM (mm) |
2.00 x 2.00 | 3.72 x 3.62 | 3.72 x 3.62 | 3.72 x 3.62 | 6.00 x 6.00 | 6.00 x 6.00 |
| # of channels | 1 | 1 | 4 | 16 | 1 | 2 |
| Active Area of device (mm²) |
4.00 | 13.47 | 53.88 | 215.52 | 36.00 | 72.00 |
| Package Footprint (mm) |
2.48 x 2.71 | 4.31 x 4.18 | 8.26 x 8.26 | 16.00 x 16.00 | 6.71 x 6.48 | 6.54 x 13.54 |
| Package Area (mm²) |
6.72 | 18.02 | 68.23 | 256.00 | 43.48 | 88.55 |
| Fill Factor (%) | 59.5 | 74.77 | 78.97 | 84.19 | 82.80 | 81.31 |
Broadcom NIR SiPM arrays are based on a different technology and are "monolithic arrays"; that is the channels are implemented in a single silicon chip. This allows a fill factor of close to 100% for the single array (i.e. without package).
AP Technologies is not just here for technical support, you can contact us for a formal quotation on any quantity of Broadcom SiPMs - from low quantities for evaluation and development purposes through to full production.
I would like to learn more about Broadcom SiPMs. Where can I get more information?
The AP Technologies website contains numerous support pages with associated application notes from Broadcom and you are always welcome to contact us using the website contact page, phone (01225 780400) or by email at
