SiPMs can be either a P-on-N or an N-on-P diode structure depending on the intended operating wavelength but in all cases they are operated Reverse Biased, that is to say the cathode is on the positive potential and the anode is on the negative one. With this in mind it can be seen that an SiPM can be biased with either potential provided the potential difference between anode and cathode has the correct polarity.
Broadcom's NUV-MT SiPM are P-on-N (the part numbering system AFBR-S4NxxPxxx4M references this for those interested in such things!) and the diagram below shows a possible readout circuit schematic:-
Where the application demands the use of an array of SiPMs to create a larger active area than the ~36mm² of the AFBR-S4N66P014M "6mm" SiPM customers can design their own nxn arrays using the 4mm or 6mm surface mount components or one of Broadcom's range of overmolded chip-on-PCB arrays:-
| AFBR-S4N44P044M | AFBR-S4N44P164M | AFBR-S4N66P024M |
| 2x2 array with 4mm pitch | 4x4 array with 4mm pitch | 1x2 array with 6mm pitch |
 |
 |
 |
It should be noted that these three arrays are designed with 4-side-tiling in mind and can be used to facilitate far larger configurations.
Broadcom's arrays are designed to allow access to the signals from each SiPM in the array and can therefore be read out using multiple single-channel circuits as outlined above - the same would apply to customer's own designs.
However, some applications require a single signal output from the array - for example accurate energy information or trigger signals. In these cases Broadcom do not recommend direct parallel ganging since this leads to summing of the individual SiPMs terminal capacitances and a slow signal with increased noise. The recommended approaches are:-
Signal summing via software: In the event an accurate acquisition time stamp is available the individual SiPM signals can be binned and signal summing of the digitised signal performed.
Analogue summing via hybrid-ganging: The circuit biases multiple SiPMs via a common electrode (cathode in schematic below). A high-pass filter is used to decouple the anode signal to the cathode of the neighbouring channel allowing the signal of all firing SiPMs to be extracted via a decoupling capacitor on the cathode side. This signal is then fed into a preamplification stage.
This approach is highly effective - the plot below shows the Gamma-Ray spectrum of three radioactive point sources (152Eu, 22Na & 137Cs) using a hybrid-ganging method to readout one signal from a 2x2 arrangement of Broadcom AFBR-S4K33P6447L (8x8 array, 3mm SiPM) giving a total of 256 SiPMs. A 51 x 51 x 30 mm3 LaBr3:Ce scintillator was used and Resolution of 3.5% at 662keV achieved.
For customers seeking an off-the-shelf readout solution for SiPM arrays - and also APD arrays or Multi-Anode Photomultiplier Tubes (MAPMTs) AP Technologies works with Vertilon Corporation who have a range of multi-channel charge integrating and photon counting Data Acquisition Systems (DAQs) together with Sensor Interface Boards (SIBs) for many standard pixelated arrays. Vertilon also offer OEM versions of their electronics suitable for integration into customer instrumentation.
Broadcom have produced a number of helpful application notes including Working with Broadcom SiPMs which discusses the issues highlighted in this article. This and other application notes can be accessed below
 |
Broadcom's SiPM Family overview |
 |
Product Brief, NUV-MT SiPM Range |
|
A Brief Introduction to SiPMs |
|
Working with Broadcom SiPMs |
|
SiPM Dynamic Range, Linearity and Saturation |
|
NUV-MT Performance Correlation |
|
NUV-MT Single-Photon Measurements |
|
SiPM Characteristics for PMT Users |
For more detail please visit the Support pages on the right hand side of this page - including Vertilon DAQs and interface boards.
