Design and Testing of a Wireless Demonstrator for Large

DPF2013-221
October 24, 2017
Design and Testing of a Wireless Demonstrator for Large
Instrumentation Systems
arXiv:1310.1819v1 [physics.ins-det] 7 Oct 2013
H. Sahoo1 , P. De Lurgio, Z. Djurcic, G. Drake, A. Kreps and M.
Oberling
High Energy Physics Division
Argonne National Laboratory, Argonne, IL-60439, USA
R. Hashemian and T. Pearson
Electrical Engineering Department
Northern Illinois University, Dekalb, IL-60115, USA
In this proceeding, we report the development of a wireless demonstrator intended to readout instrumentation systems having thousands
of channels. A data acquisition system was designed and tested based
on compliant implementation of 802.11n based hardware and protocols.
This project is for large detectors containing photomultiplier tubes. Both
free-space optical and radio frequency techniques were tested for wireless
power transfer. The front-end circuitry, including a high-voltage power
supply was powered wirelessly, thus creating an all-wireless detector readout. The system was successfully tested as a single detector module, which
was powered wirelessly and transmitted data wirelessly. The performance
of the prototype system and how a large scale implementation of the system might be realized are described in this proceeding.
PRESENTED AT
DPF 2013
The Meeting of the American Physical Society
Division of Particles and Fields
Santa Cruz, California, August 13–17, 2013
1
Presented by Himansu Sahoo at the DPF 2013 Meeting of the American Physical Society
Division of Particles and Fields, Santa Cruz, California, August 13-17, 2013.
1
Motivation for Wireless DAQ
In several areas of scientific research, the size and complexity of detectors have become
exceedingly large. For example, detectors used in Nuclear and Elementary Particle
Physics can have dimensions of the order of 10-100 meters and contain thousands
to millions of readout channels. This create a significant challenge to power the
electronics as well as to transfer data. Traditional approach of using electrical cables
have become complicated and expensive at a larger scale. Specially, cabling is not
practical for detectors in remote location or in hostile environment. To overcome these
problems, we developed an alternative approach that uses wireless techniques to power
and transmit data. A stand-alone photomultiplier tube (PMT) base detector was
designed and tested in free space, that operates from wireless power and then transfers
data wirelessly. We carried out this case study for large detectors containing PMTs,
which are commonly used in high energy physics research. The primary purpose is
to ascertain the feasibility and practicality of such devices as single detector modules
that can be configured as arrays in a large detector. This approach has potential to
eliminate the need of expensive and massive cable plants, to simplify the process of
installation and repair, and to reduce the detector dead mass.
2
Design Considerations
We have explored different technologies, such as free-space optical and radio frequency (RF) to wirelessly transmit data and power. But, we selected the technology
that is inexpensive and off-the-self that can be easily implemented while meeting the
performance goals of our R&D project.
For wireless data transmission, optical links support higher data rates (1 Gbit/s [1])
than those use RF. However, RF transmission does not require line-of-sight, and an
individual receiver (i.e. access point) can communicate with many front-ends. Since
both of these advantages provide significant simplification and cost reduction, RF
data transmission was chosen for this project. We focused on wireless local area network (WLAN) technologies with 802.11n variant, because it offers the highest data
throughput and has sufficient range for data transmission. For instance, a single
steam 802.11n link has a total data rate of approximately 65 Mbit/s [2]. While this
is sufficient for our single prototype front-end, it provides greater challenge for large
detectors in transferring data from thousands of readout channels over a limited and
common frequency spectrum. We addressed this problem by filling the available frequency space with many access points, as described in Ref. [3]. One of the frequency
range in 802.11n is centered around 5.5 GHz with an overall bandwidth of approximately 1.2 GHz (4.9−6.1 GHz). Single stream 802.11n access points can have an
individual operating bandwidth of 20 MHz. This enabled us to populate up to 48
1
access points with the usable 1.2 GHz frequency spectrum, each communicating to
nearly 64 PMT wireless read-outs or front-ends (a total of 3072 front-ends). This
design provided an overall data throughput of 1.68 Gbit/s.
For wireless power transmission, we tested both optical and RF power transfer
methods. The optical power demonstrator utilizes a high power light-emitting diode
(LED) that is collimated into an 8 inch diameter beam and is received by a photovoltaic (PV) panel, as shown in Fig. 1. The LED is an OSRAM SFH 4751 with 3.5
W optical output, operated at a maximum DC current of 1 A. The LED wavelength
is 940 nm which matches the peak efficiency of the Delsolar 156 × 156 mm2 photovoltaic cell used in our 312 × 280 mm2 PV panel array. This test system met our
requirements of receiving 250 mW of power at 5 meters.
Power received (mW)
600
550
500
450
400
350
300
250
200
2
3
4
5
6
7
Distance (m)
Figure 1: Apparatus for optical power transmission. Left figure shows the tube
containing LED and lens, and the photovoltaic receiver at the far end. Right figure
shows the power received (mW) by the photovoltaic panel as a function of distance
(m) from the optical source.
The RF power demonstrator uses high-gain directional microwave antennas. The
setup consisted of a function generator driving a 14 dBi gain Yagi antenna at 915 MHz
with an output power of 10 dBm, which was received by a 11 dBi gain patch antenna,
as shown in Fig. 2. The power received was measured in free space to minimize
scattering from surrounding objects. As the transmission distance increased, the
power received fell rapidly; the power loss being 20 dBm at 5 meters. Thus, a 25 W
source will be required to receive our targeted 250 mW power.
For this feasibility study, we chose the optical method to implement in our prototype. It provided a DC source at the receiver end, which is relatively easy to utilize.
In order to transmit RF power, it has to be converted into a DC supply at the receiver end, which is commercially available for only a 100 mW power [4]. However,
RF power could be a better choice for a large detector because one source can power
2
Power loss (dB)
0
-5
-10
-15
-20
-25
-30
1
2
3 4 5 6
Distance (m)
7
8
Figure 2: Apparatus for RF power transmission. Left figure shows the RF transmitter (14 dBi gain Yagi antenna) in the foreground, with the RF receiver (11 dBi gain
patch antenna) at the far end. Right figure shows the power loss (dB) as a function
of distance (m) from the source.
many front-ends, thereby reducing the complexity and cost of the system.
3
Wireless Data Acquisition System
The wireless prototype system is comprised of four boards: a power board, which
receives wireless power and generates different voltages needed by the system; a digital
board, which processes the data and does wireless data transmission; a front-end
board, which does shaping and digitization of the PMT signal; and a high voltage
board, which generates high voltage for the PMT. The high voltage board uses a
standard Cockroft-Walton (CW) switching circuit to boost the 24 V input voltage
up to 2000 V, which is needed for the 10-stage tube PMT. We chose a large 10 inch
diameter PMT, a Hamamatsu R7081HQE [5], which has dark noise rate of ∼10 kHz.
All the boards are physically arranged inside a tube as shown in Fig. 3. One end of
the tube was fitted into the base of the PMT and the other end was connected to a
photovoltaic panel. Details of the wireless demonstrator are described in Ref. [3].
The prototype readout system consisted of a Scientific Linux computer and a Cisco
E3000 [6] running DD-WRT firmware as an access point. The front-end transmitted
data wirelessly once per second as a single UDP packet using 802.11n in the 5 GHz
band. A readout program running on the server received and stored the incoming
UDP packets. For each asynchronous PMT trigger, we stored the pulse height (2
bytes) and time stamp (4 bytes) information. We have collected data with the prototype system operated from wireless power (using the optical source and received by
the photovoltaic panel) and with wireless data readout. The performance achieved
3
Figure 3: The printed circuit boards used in the wireless prototype system (left) and
configuration of the prototype readout system (right).
along with target specifications are summarized in Table 1. The system is capable of
sending greater than 10k events/s. To test the data acquisition capability, a sodium
iodide crystal was attached to the PMT and tested with 241 Am and 137 Cs sources.
The 137 Cs data yields a 17% energy resolution with sodium iodide. The response from
the sources are shown in Fig. 4.
Table 1: Summary of the initial goals and achieved performance for the wireless data
acquisition system.
Specification
Target
Performance
Total power consumption (@ 10 kHz)
250 mW
386 mW
Digital
120 mW
216 mW
Front-end
30 mW
39 mW
HV
80 mW
131 mW
Maximum event rate
10 kHz
80 kHz
Data transfer rate
35 Mb/s
11 Mb/s
Bit Error Rate
< 1×10−12 Dropped Packets
4
Summary
We successfully built and tested a wireless demonstrator, which was implemented in
a photomultiplier tube base and received power and transmitted data wirelessly. The
4
3
×10
60000
160
140
120
40000
Events
Events
50000
30000
20000
100
80
60
40
10000
20
0
0
200
400
600
800
1000
1200
0
0
1400
ADC value
500
1000 1500 2000 2500 3000 3500 4000
ADC value
Figure 4: Response from 60 keV 241 Am source (left) and from 661.7 keV 137 Cs source
(right). PMT HV in on and system is powered wirelessly and wireless data readout.
system transmitted data at a rate of 11 Mbit/s, which can support up to 16 frontends per wireless channel. While the power consumption was slightly greater than our
target, it was still low enough to allow the system to operate from our optical power
system. In the longer-term, we intend to implement RF power transfer to facilitate
the simplification by using one transmitter to power many receivers. Additionally,
we will also investigate the use of a custom ASIC for lower power operation of the
front-end and Cockroft-Walton control circuitry for a larger system.
ACKNOWLEDGMENTS
We acknowledge the support of Laboratory Directed Research & Development funding
from Argonne National Laboratory to carry out this project. I thank the conference
organizers for their invitation to present this work at DPF conference.
References
[1] P. De Lurgio, W. Fernando, B. Salvachua, R. Stanek, D. Underwood, D. Lopez,
New Optical Link Technologies for HEP Experiments, in Proc. DPF-2011 Conference, Providence RI, August 8-13, 2011.
[2] cBOWL221a Product Brief, cBProduct-0811-01 (1.6), Aug. 2011, ConnectBlue
AB, Norra Vallgatan 64 3V, SE-211 22 Malmö, Sweden.
[3] P. De Lurgio, Z. Djurcic, G. Drake, R. Hashemian, A. Kreps, M. Oberling,
T. Pearson and H. Sahoo, A Prototype of Wireless Power and Data Acquisition
System for Large Detectors, arXiv:1310.1098.
5
[4] P1110 915 MHz RF PowerharvesterTM Receiver Datasheet, Rev A, Apr. 2010,
Powercast Corporation, 566 Alpha Drive, Pitsburgh, PA 15238, United States.
[5] R7081HQE Datasheet, Nov. 12. 2003, Hamamatsu Photonics K.K., 325-6,
Sunayama-cho, Naka-ku, Hamamatsu City, Shizuoka Pref., 430-8587, Japan.
[6] E3000 Datasheet, 10031010NC-JL, Oct. 2010, Cisco Systems, Inc., 170 West
Tasman Dr., San Jose, CA 95134, United States.
6
Download PDF
Similar pages