Photon Force PF32 User manual

PF32 USER MANUAL
v1.4.9

Page 2 of 16
W A R R A N T Y A N D D I S C L A I M E R S
Use of this product and the associated software implies acceptance of the Photon Force terms of
use.
Bare electronic components are susceptible to electrostatic discharge (ESD). Failure to
employ good practice in handling the hardware, or to observe and comply with the
warnings and handling precautions stated in this guide will void product warranty.
This product may not be used in military, aerospace, medical or other safety critical
applications without the express written permission of Photon Force Ltd. Any such use is
undertaken entirely at the customer’s own risk.
Specifications are subject to change without notice.

Page 3 of 16
T A B L E O F C O N T E N T S
Warranty and disclaimers ................................................................................................................... 2
Warnings ............................................................................................................................................. 4
Package Contents ................................................................................................................................ 5
Introduction ........................................................................................................................................ 6
Features at a glance: ....................................................................................................................... 6
PF32 General characteristics ............................................................................................................... 7
PF32 Modes of Operation ................................................................................................................... 9
TCSPC mode .................................................................................................................................... 9
Photon Counting mode ................................................................................................................. 12
PF32 electrical connections .............................................................................................................. 13
Technical Specifications .................................................................................................................... 14
Mechanical Drawing ......................................................................................................................... 15

Page 4 of 16
W A R N I N G S
Store the camera at temperatures between 0 and 40°C
Avoid humid environments (70% and above)
Keep out of direct sunlight and away from heaters
Avoid vibrating surfaces when mounting the PF32 system
Avoid areas that contain dust/particulates

Page 5 of 16
P A C K A G E C O N T E N T S
Upon receiving your PF32 camera, please check the contents:
PF32 Camera
5V DC power supply
USB3 cable

Page 6 of 16
I N T R O D U C T I O N
The PF32 system consists of 1024 single-photon avalanche diodes (SPADs) arranged in a 32 × 32
array. Each pixel has its own photon counting and timing electronics, making it a powerful and
unique single-photon sensitive time-resolved imager. It can be used in time-correlated single-
photon counting (TCSPC) mode with a photon timing accuracy of 55 ps, or it can be used in photon
counting mode as a single-photon sensitive high frame rate camera. In TCSPC mode, the sensor
can be operated in conjunction with a wide variety of light sources, such as pulsed lasers and
LEDs, to perform time-resolved measurements. To accommodate the differing properties of these
devices, a number of clocking/synchronisation options can be provided.
The camera has a CS-mount thread enabling the use of a wide variety of lenses, and with the
addition of adapters this can be extended to almost any lens the user wishes to employ. It is easily
connected to a computer through a USB 3.0 connection, and requires a simple 5V supply for the
entire system.
F E A T U R E S A T A G L A N C E :
 32×32 pixel TCSPC imaging array.
o Fully digital photon counting and time-stamping (no analogue readout noise).
o In-pixel dual mode electronics:
 55ps resolution 10-bit Time to Digital Converter (TDC) time-stamping
(1,023 time bins).
 7-bit photon counting (count up to 127 photons per pixel per frame)
o Pipelined operation: simultaneous data acquisition and readout.
 Flexible readout timing allows an increased frame rate for a subset of pixels or reduced
number of counter/TDC bits (LSB first).
 Row and column enable/disable settings to define region of interest.
 Optional calibration mode to lock TDC resolution to laser or reference frequency.
 External laser synchronisation signal and reference clock inputs to provide TDC stop
signal and calibration loop reference frequency.
 Individual power and bias supplies for noise-sensitive features.

Page 7 of 16
P F 3 2 G E N E R A L C H A R A C T E R I S T I C S
Within each pixel of the PF32 sensor, the detection of single photons is performed by the SPAD
active area. This active area is surrounded by the necessary electronics to bias and quench the
SPAD, as well to time and count the detected photons. In a traditional TCSPC setup, each detector
would need its own external electronics to time-tag the photon’s arrival, whereas the PF32 is
intended to provide a compact and highly parallel means of data acquisition for a variety of
applications such as fluorescence lifetime imaging microscopy (FLIM) and 3D active imaging.
The active area is a silicon p-n junction which is reverse-biased beyond its breakdown voltage in
the so-called “Geiger-mode”. In this regime, the absorption of a single photon gives rise to a
detectable current pulse due to impact ionisation events within the semiconductor lattice enabled
by the acceleration of the initial photo-generated carriers in the high electric field. After the onset
of an avalanche, the SPAD is disarmed by reducing the bias to a value below the breakdown
voltage. The SPAD is then re-armed after a short period known as the “dead time”. The dead time
(sometimes called the “hold-off” time) is needed to ensure that any carriers trapped within the
semiconductor structure following an avalanche event are released; if the dead time is too short
and the SPAD is re-armed, any remaining trapped carriers may trigger another avalanche,
referred to as an “afterpulse”. Afterpulsing effectively increases the dark count rate (DCR) of the
SPAD, thus decreasing the signal to noise ratio (SNR). For this reason, the dead time is set to a
sufficiently long period to negate the deleterious effects of afterpulsing.
The DCR itself is a function of temperature and bias voltage; increasing either, or both, of these
parameters also increases the DCR. The final contribution to the DCR is through cross-talk,
whereby during the avalanche process some photons are emitted during the large flow of high-
energy carriers. Due to the large distance between active areas, the contribution of cross-talk is
negligible.
The photon-detection efficiency is also a function of bias voltage as shown in the figure below.
The final important figure of merit is the jitter, or instrument response function (IRF), which
relates to the uncertainty in absolute timing of the stochastic avalanche process in response to
the detection of a photon. The IRF of the SPADs within the PF32 array is ~150 ps.
Photon detection probability vs. wavelength of a PF32 pixel.
300 400 500 600 700 800 900 1000 1100
0
5
10
15
20
25
30
0.2 V excess bias
0.6 V excess bias
1.0 V excess bias
Photon Detection Probability (%)
Wavelength (nm)

Page 8 of 16
Further reading is available here:
 Paper describing the SPAD device: J. A. Richardson, L. A. Grant and R. K. Henderson; “Low
Dark Count Single-Photon Avalanche Diode Structure Compatible With Standard
Nanometer Scale CMOS Technology”, Photonics Technology Letters, IEEE, Jul. 2009, vol. 21,
no. 14, pp. 1020-1022.
 Paper describing the PF32 device: J. Richardson, R. Walker, L. Grant, D. Stoppa, F. Borghetti,
E. Charbon, M. Gersbach and R. K. Henderson; “A 32×32 50ps resolution 10 bit time to digital
converter array in 130nm CMOS for time correlated imaging”, IEEE Custom Integrated
Circuits Conference (CICC), San Jose, USA, September 2009, pp. 77-80.

Page 9 of 16
P F 3 2 M O D E S O F O P E R A T I O N
T C S P C M O D E
Often, TCSPC is thought of as a fast stopwatch – the start being the emission of the pulsed light
source and the stop being the detection of a single photon. However, like many TCSPC systems,
the PF32 performs reverse start-stop measurements: the detection of a single photon starts the
time-to-digital converter (TDC) and the next synchronisation pulse stops this process. This is
explained in the following diagram.
Using a basic laser-ranging setup as an example, the reverse start-stop principle can be explained.
A pulsed laser beam passes through a diffuser and illuminates 2 surfaces, A and B. Photons are
scattered from these surfaces and some will be collected by the PF32. Below the ranging setup we
show the timing of the processes involved (for simplicity, we consider a half round trip). The laser
period is tl, and at some time later, photons are detected by the PF32. Considering forward start-
stop, the time stamp given to a photon from surface A (tAf) has a lower value than that of a photon
from surface B (tBf). In reverse start-stop, the detection of a photon starts the timing process and
the timing information is given relative to the subsequent laser sync pulse. This results in the time
stamp for surface A (tAr) being of a greater value than that of surface B (tBr).
Reverse start-stop measurements are advantageous since they ensure that the counting
electronics are only active when a photon has been detected; in forward start-stop, the
synchronisation signal starts the timing and the detection of a photon stops the timing. This
means that the timing electronics are in constant use and need to be reset every synchronisation
period, potentially increasing readout dead time, power consumption, and heat dissipated
through the device.
The PF32 can accept a synchronisation signal via the Sync SMA connector (3.3V max as set out in
the synchronization document) in laser-is-master mode. Alternatively, the FPGA within the
camera can be used to provide the synchronisation signal to the TDC and also output this to the
light source via the TRIG SMA connection (0 to 3.3V).

Page 10 of 16
The image below demonstrates the basic setup in the laser-is-master mode:
Here, the laser driver is the master clock for the TCSPC measurement: it provides the timing for
both the laser to fire and for the stop signal for the camera through the sync input SMA of the
PF32.
If the laser being used does not have a synchronisation output, a beam-splitter can be used to take
a small portion of the outgoing power and direct it onto a fast photodiode. The output of the
photodiode can then be used as the sync pulse as shown below.
If the laser driver does not have a synchronization output, the master clock for the TCSPC
measurement can be provided by taking part of the outgoing light and directing it onto a fast
photodiode.
If, however, you’d like to use the internal clock from the PF32 to trigger your laser, the following
configuration can be used.

Page 11 of 16
If the pulsed light source accepts a trigger, the PF32 can be used as the master clock via the TRIG
output (3.3V)
The following diagram shows the intended method of operation of the PF32's TCSPC mode as
detailed in the figure caption.
Timing diagram showing intended TCSPC use: In this example, (a) a laser firing at 20MHz is
directed at a sample that absorbs the laser light and re-emits at a longer wavelength due to
photoluminescence with a temporal decay as shown in (c). Here we examine the timing of a single
PF32 pixel where the TDC is in the ready state (e) until a photon is detected from the sample (d).
The TDC starts to run and is stopped by the following laser sync pulse (b). This time stamp (f) is
then awaiting transfer to memory until the start of the next frame, during which time the TDC
cannot register any other photon arrivals. The frame clock (g), therefore, dictates the highest
count rate achievable by any single pixel within the PF32 array since each TDC can only generate
one timestamp per frame time.
While there are benefits of reverse start-stop TCSPC measurements, some care does need to be
taken in order to extract the best performance from the system. Since the TDC provides 1,023
time bins (10 bit) with ~55ps resolution, the maximum unambiguous measurement is
time
…
(b) Laser S
ync
(g) Frame Clock
TDC
running
<50ns
Later photons
ignored
(c) Sample Emission
(d) Detected Photons
(a) Laser E
mission
Timestamp
generated on
sync rising
edge
<50ns laser period
(>20MHz rep rate)
Photon
triggers
TDC
Timestamp
read out on
frame clock
pulse
(e) TDC Operation READY
READY TIMESTAMP N AWAITING TRANSFER TO MEMORY RESET
(f) Pixel Memory TIMESTAMP N TIMESTAMP N-1
…
37.5ns

Page 12 of 16
approximately 56.3ns. If longer intervals are measured, the time-stamp value will ‘wrap around’,
such that a 57.3ns interval is indistinguishable from 1ns. However, it should be noted that the
measurement uncertainty (jitter) accumulates over time; the more times the TDC wraps around,
the higher the jitter degradation. Care should therefore be taken to provide a maximum STOP
waveform period of under 56.3ns. This can be accomplished by delaying the synchronisation
pulse with an active delay generator, or passively with lengths of coax cable.
If no photon is detected within the frame time, a count is added to the zero time bin.
As with all TCSPC systems, care should be taken to avoid the effects of pulse pileup, which will
distort the photon-counting histogram. It is essential that the photon counting rate per pixel is
less than 10% of the sync frequency in order to avoid pulse pileup effects.
In both modes, an in-pixel memory stores the data captured during each frame time. This data is
then read out during the following frame, while the pixel acquires new data.
P H O T O N C O U N T I N G M O D E
In photon counting mode, the TDC is configured to simply count the number of photons received
during each frame time. No sync is required for the measurement process, making photon
counting mode a useful tool for initial setup/alignment of the system before taking TCSPC data.
In this mode, each pixel can count multiple photons per frame as shown below.
In photon counting mode, the frame clock (a) dictates the maximum frame rate (a single pixel has
been shown for clarity). At the start of each frame there is a short deadtime (shown in red) of
~50ns before which detected photons will not be counted, after which each pixel can count multiple
photons within a frame (7 bit counter, therefore 127 photons per frame). After the detection of a
photon (b), the counter increments (c) and the SPAD is reset. During the reset process the SPAD
is off for ~50ns. This process is repeated until the end of the frame when the next frame clock
pulse initiates the transfer of the data to the readout stage (d). It should be noted that the 7 bit
counter will reset to zero after 127 detection events. A sufficiently short frame time should be
chosen to ensure that wrap around does not occur. The camera’s ‘frames to add’ setting allows
multiple frames to be summed digitally to achieve a larger maximum count value if needed.

Page 13 of 16
P F 3 2 E L E C T R I C A L C O N N E C T I O N S
As shown in the figure above, at the top of the camera housing, there are 5 input and 2 output
SMA connectors, a USB 3.0 connector, and a 5 V power supply terminal. The table below describes
the SMA connector functions:
Name Voltage Range (V) Description
FRM 0 to 5 Frame sync input for use with scanning systems
LINE 0 to 5 Line sync input for use with scanning systems
PIXEL 0 to 5 Pixel sync input for use with scanning systems
BLK 0 to 5 Blanking input for use with scanning systems
SYNC 0 to 3.3 External laser sync input signal
TRIG 0 to 3.3 External laser source trigger signal (output)
SHUT 0 to 3.3 Shutter output signal

Page 14 of 16
T E C H N I C A L S P E C I F I C A T I O N S
Sensor Dimensions
Array size 32 × 32 pixels
1.6 × 1.6mm
SPAD active area 6.95µm
⌀
Pixel pitch 50µm
Optical fill factor 1.5%
Optical / Electrical Performance
Photodetection efficiency Peak 28% at 500nm
Dark noise: <100Hz for 80% of pixels
Afterpulsing <0.02%
Optical/Electrical Crosstalk: None
Timing jitter: <200ps FWHM
Photon Counting Mode
Photon counting: 7 bit in-pixel
16 bit in firmware
Maximum photon counting rate per pixel: 50MHz
Time Correlated Mode
Temporal bin size: 55ps
Temporal range: 55ps - 57ns
TDC resolution: 10 bit
Maximum laser sync frequency (input or
output):
100MHz
Laser sync input amplitude NIM / 3.3V / 5V
Laser sync output amplitude 3.3V
Readout & Control
Maximum sensor to firmware frame rate,
10-bit data
0.5Mfps - 8Mfps [1]
Raw data streaming frame rate to PC 300kfps (USB3)
Inter-frame dead time <50ns
X/Y Scanner synchronisation input signals Pixel, line and frame clock
Exposure synchronisation signals Blanking (3.3V / 5V
input)
Shutter (3.3V output)

Page 15 of 16
M E C H A N I C A L D R A W I N G
Dimensions in mm and [inches]

Photon Force Ltd.
34 Melville Street,
Edinburgh
Scotland, UK,
EH3 7HA