-
Notifications
You must be signed in to change notification settings - Fork 0
/
introduction.tex
114 lines (104 loc) · 5.35 KB
/
introduction.tex
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
\section{Introduction}
The main purpose of the LBNE DAQ system is to read the raw data from the
Front End Boards (FEB), which are mounted on the Anode Plane Arrays (APA) inside
the cryostat, to build events from the different parts
of the detector and to pass these events on to long term storage.
The Level 3 requirements for this system include\cite{DAQ_REQ}:
\begin{itemize}
\item{LArFD-L3-DAQ-3: The DAQ shall be capable of receiving raw data from a freely running readout from all detector systems.}
\item{LArFD-L3-DAQ-7: The DAQ shall be designed to collect data continuously}
\item{LArFD-L3-DAQ-8: The DAQ shall perform prompt processing of data}
\end{itemize}
The DAQ-7 requirement is relevant mainly to non-beam physics.
As such, it was left out of the requirements at the time of CD-1, which
assumed a surface-located Far Detector (FD) with 3 meter-water-equivalent (MWE)
overburden.
Nonetheless, continuous readout remains a valuable goal that would be desirable
to have in the final LBNE DAQ system and is planned for in the 35t prototype.
Table \ref{tab:rates}, which is taken from Jon Urheim's CD-1 presentation
\cite{DAQ_CD1}
lists
the rates expected in a single Anode Plane Array (APA) for a surface-located
FD.
Assuming that a zero-suppression
algorithm is running in the cold electronics,
the rate is dominated by cosmic ray muons.
Note, however, that common-mode noise,
which may be appreciable,
is not included in this table.
\begin{table}[h]
\begin{tabular}{|p{1.5in}|c|c|c|}
\hline
Process&Rate (kHz/APA)& Samples (per APA) & Avg. Data Rate (Mbps)\\
\hline
Generic 1.4 ms interval (not zero-suppressed)&
0.70&$7.3 \times 10^6$& 61,000\\
\hline
Cosmic Ray Muons & $\approx$ 1 & $2.5 \times 10^4$ & ~300 \\
\hline
Radioactivity: &&& \\
U/Th ($\gamma$'s) & $\approx$ 1&40& 0.48\\
$^{39}$Ar/$^{85}$Kr ($\beta$'s)&40&24&12\\
\hline
Electronics Noise (not common mode) & $\approx 1$ & 15 & 0.2\\
\hline
\end{tabular}
\caption{\label{tab:rates} Expected rates for one APA in a surface-located
FD}
\end{table}
The first (and only) opportunity to test many features of the LBNE APA
design will be the 35 Ton prototype (35t) to be constructed at FNAL.
The APA(s) of this device are likely to be somewhat smaller than
those of the full LBNE FD and there will be very little overburden.
The rates listed above are thus only a rough estimate of those to be
expected in the 35t.
Nonetheless, we adopt them for the following discussions.
The key electronics module that needs to be provided for the back-end DAQ is
one that is capable of reading the data streams from each of the
APA's and concentrating the data down to a smaller number of high-bandwidth
data streams that are then passed to an event-building network.
This is a commonly needed function in modern HEP experiments that has frequently
been addressed with custom modules built explicitly for a single experiment.
This may require significant development time, and the modules produced quickly
become obsolete, as available networking technology progresses.
Furthermore, the modules are not typically produced with re-usability
in mind.
These two factors limit the desirability of reusing such modules in
subsequent experiments.
The SLAC Research Electronics Department (RED) via the DOE generic detector R \& D effort has developed a solution to these
problems of obsolescence and lack-of-reusability
by producing a hardware system, together with firmware and software, that can
be easily adapted for use in different experiments.
The development costs are then leveraged over multiple experiments, allowing
each of them to benefit from the latest networking hardware, at a significant
reduction in development costs.
This ``DAQ toolkit'' uses the modern Advanced Telecommunications Architecture
(ATCA) for its physical structure.
The toolkit is not an ``off the shelf'', DAQ system.
Rather, it provides tools that can be used by individual experiments to
build modern DAQ systems with significantly lower development costs than
would be needed to build fully custom systems.
The key element of the toolkit is the Reconfigurable Cluster Element (RCE).
The RCE is the third generation of development (Gen 3) and is
based on a Virtex Zynq ``System on a Chip''.
A single board combining several of these RCE's can handle very high bandwidths
measured in the 100's of Gigabits/second.
This system has been adopted in several HEP experiments already, and will likely
be adopted by more in the future.
The REG is continuing to develop and support new generations
of the toolkit to take advantage of new networking equipment as it becomes available.
We are proposing to make use of this toolkit in the DAQ systems to be produced
for the LBNE 35 ton prototype and the full Far Detector.
The bandwidth available in the current generation of the toolkit far
exceeds that of the baseline system based on the Nova Data Concentrator
Module (DCM).
The increased flexibility afforded by this extra bandwidth may be highly valuable to
ensuring LBNE success.
Furthermore, leveraging the work already done by the REG as well as benefiting from
their support in the future, will provide many benefits to LBNE and may reduce
the development costs.
We have begun exploring the viability of the SLAC DAQ toolkit in the LBNE
context with the help of
an SLAC FY13 LDRD grant that we successfully applied for.
This grant is supporting initial prototyping work, which thus comes at no cost
to LBNE.