<html>
<head>
<meta http-equiv="Content-Type" content="text/html; charset=utf-8" />
<style type="text/css">
<!--
html{color:#555555;}body{line-height:1.5;font-family:'Trebuchet MS','Helvetica Neue',Arial,Helvetica,sans-serif;font-size:87.5%;}h1{font-size:1.6em;}h2.field-label{display:inline-block;font-size:1em;padding-right:5px;min-width:10em;margin:0.3em;}.problem_report{line-height:1.5;max-width:60em;}fieldset.problem_report.resolved
legend{background-image:url(data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABAAAAAQCAYAAAAf8/9hAAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAy0lEQVQ4jWP8//8/AyWAiZACd3f3/xYWFrht+f//P1a84t3e/0obff4rbfT5D1GGXR0LuoEr3+/7X3W4n2gvwA0gVSOKAcqbfPGGpImJCU45JgYGBoa7fpsZ22wLSbadgYGBgRE9GrF55Vf2BYbHjx8zYjWB0ljAcAGGExkZ/0MtwuoCggmJEBh4AzBS4pMnT/7fuXOH4dKlSwwnT56EiwcGBv43MDBgMDExYdDX12eQkZGBhAlyiC5YsOA/AwMDUXjLli3/iYoFQgAA+pSxZrXofD0AAAAASUVORK5CYII=);background-repeat:no-repeat;padding-left:18px;}fieldset.problem_report.needs_attention
legend{background-image:url(data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAABAAAAAQCAYAAAAf8/9hAAAABmJLR0QA/wD/AP+gvaeTAAAA9ElEQVR42sWTvUoDQRSFv9wMKWxSBVmzdhZJIwTWv9pyLWxTpbE1kBeJPoLxBazzBgGFKNqlHXAhsITUw1y7sMpmjER0YJrDPWcO556BPzhjYPJjlhfT82LUpxcK6Lo5U0JMgcu56tXy7BQajeBDpkAcAEdz1W6uyrLdYieK2DMmKCArcqczpH/ddc0msy+OkyQJC4h3N0ynx+Q5ALtUNs5q5U+8e+R+VPFi4kjk3NZqd++qUK+TZVnYwSfAOyvejeLXt/2qVG/dYoG1dqseBNco27bs/wXKWhIDB8AhcFLAH4Bn4Al4AUqT7RVC++6mv/JVPwDi3VGzomYvyAAAAABJRU5ErkJggg==);background-repeat:no-repeat;padding-left:18px;}.problem_report div.field-items{display:inline-block;}div.date-vitals p{font-size:87.5%;}a{text-decoration:none;}.Readme a:link,.Readme a:visited,.Readme
a:active{color:red;}
-->
</style>
</head>
<body id="mimemail-body" class="elog-logentry-notify">
<div id="center">
<div id="main">
<style>
<!--/*--><![CDATA[/* ><!--*/
div.field-vitals{
margin: 0.5em 0;
}
div.field-vitals .field-type-taxonomy-term-reference {
margin: 0.1em 0;
}
article.comment {
padding-left: 10px;
}
article.comment.odd {
background-color: #EEEEEE;
}
article.comment.even {
background-color: #DDDDDD;
}
div.node-content.logentry table{
width: auto;
border-collapse: collapse;
border-spacing: 0;
border-width: 1px;
}
div.node-content.logentry th{
border: inherit;
}
div.node-content.logentry blockquote{
background-color: #FFFFFF;
}
div.node-content.logentry caption{
font-size: 1em;
font-weight: normal;
}
table.field-vitals{
margin-top: 1em;
margin-bottom: 1em;
font-size: 87.5%;
}
table.field-vitals th{
vertical-align: middle;
text-align: left;
width: 15%;
padding: 0.1em;
}
table.field-vitals td{
vertical-align: middle;
text-align: left;
width: auto;
padding: 0.1em;
}
table.field-vitals td li {
margin-left: 0;
list-style-type: none;
list-style-image: none;
}
table.downtime {
width: 30em;
margin-bottom: 1em;
border: 1px black dotted;
}
table.downtime th {
text-align: center;
}
table.downtime td {
text-align: center;
}
tr.caption th {
border-bottom: none;
}
table.downtime tfoot{
background-color:#EEEEEE;
}
div.field-name-body{
margin: 1em 0;
font-size: 110%;
}
div.date-vitals p{
margin: .1em 0;
}
article div.ctools-collapsible-container{
margin-left: -5px;
clear: both;
}
#comment-form{
margin-left: 5px;
border: graytext outset medium;
-moz-border-radius: 15px;
border-radius: 15px;
padding: 1em;
}
div.comments-form-box {
margin-top: 2em;
margin-bottom: 5em;
}
h3.comment-title {
/* display: none; */
}
p.author-datetime{
font-weight: bold;
}
/*--><!]]>*/
</style><article id="node-518548" class="node node-logentry article ia-n clearfix" role="article"><header class="node-header"><h1 class="node-title">
<a href="https://logbooks.jlab.org/entry/3445710" rel="bookmark">Performance testing of the FDC F1 TDC crates</a>
</h1>
</header><div class="date-vitals">
<p class="author-datetime">
Lognumber <a href="https://logbooks.jlab.org/entry/3445710" class="lognumber" data-lognumber="3445710">3445710</a>. Submitted by <a href="https://logbooks.jlab.org/user/somov">somov</a> on <time datetime="2016-12-11T20:59:13-0500" pubdate="pubdate">Sun, 12/11/2016 - 20:59</time>. </p>
<table class="field-vitals"><tr><th>Logbooks: </th><td><a href="https://logbooks.jlab.org/book/hdlog">HDLOG</a> <a href="https://logbooks.jlab.org/book/hdtrigger">HDTRIGGER</a></td></tr><tr><th>Entry Makers: </th><td>Alex, Lubomir</td></tr></table></div>
<div class="logentry node-content">
<p>We checked performance of the new F1 tdc firmware<br />
provided by Ed using 4 FDC crates. In general,<br />
the F1 performance was found to be 'robust' for<br />
large trigger rates and hit multiplicities.</p>
<p>Setup:<br />
-----</p>
<p>Crates: rocfdc 1,4,13,14<br />
each crate has 12 low-resolution F1 TDC modules<br />
(except for rocfdc 1, where two more tdcs, not used<br />
by the FDC are installed)</p>
<p>Trigger: TS FP random trigger<br />
trigger holdoff parameters consistent with fadc 125</p>
<p>Hit multiplicity: changed by lowering LV thresholds</p>
<p>Block level = 40, buffer level = 4, readout window = 1.4 usec </p>
<p>(we also checked performance using different block levels)</p>
<p>Data were read out from TI and TDC modules, but was NOT<br />
transmitted from the crate.</p>
<p>(the maximum transmission rate of a gigabit Ethernet link<br />
was measured separately using an individual FDC crate and<br />
a setup of 4 FDC crates and random trigger. The rate is<br />
about 115 Mbps, see Fig. 1)</p>
<p>Fig 2. Top<br />
----------</p>
<p>Data rate R (Mbyte/sec), read out from TI/TDC modules<br />
(in rocfdc 13 ) as a function of the trigger rate<br />
for various hit multiplicities.</p>
<p>( R = Input trigger Rate * Livetime * Size of 40 blocks / 40 ) </p>
<p>Fig 2. Bottom</p>
<p>Corresponding live times as a function of the trigger rate.</p>
<p>Comments:<br />
---------</p>
<p>There are two factors limiting the performance:<br />
VME data transfer rate and the F1 TDC processing time.</p>
<p>For very large hit multiplicities, the VME transmission<br />
limitations occur at a small trigger rate, when the<br />
processing time is negligible, see blue curve in the<br />
top plot of Fig. 2.</p>
<p>The maximum VME data rate was estimated to be about<br />
180 Mbyte/sec</p>
<p>At higher rates and modest hit multiplicities, the<br />
processing time becomes the limiting factor, see<br />
other curves on the same plot.</p>
<p>The green curve satisfies the GlueX high-rate performance<br />
(expected L1 rate < 100 kHz, 1 Gbit/s Eth.)</p>
<p>Note, the processing time depends strongly on the number<br />
of hits in the particular F1 TDC module. The hit<br />
multiplicity corresponding to the green curve (rocfdc 13)<br />
is dominated by two F1 TDC modules, see Fig. 3. Each<br />
module has 3 (out of six) 'noisy' chips producing the<br />
maximum number of hits (64 hits/chip).</p>
<p>The crate performance is expected to be better if hits<br />
are more uniformly distributed over F1 TDC modules in<br />
the crate.</p>
<p>Note, the new firmware provides better handling of burst<br />
of triggers: data for some triggers can be lost (the<br />
corresponding flag is reported).</p>
<p>We'll keep on testing the performance.</p>
</div>
<div class="attachment-box">
<div class="image-wrapper"><img view_mode="elog_email" class="image-style-plentybig" src="cid:3bb0d9fd7fea5ad27b9cdd71538aa37a@logbooks.jlab.org" width="960" height="763" alt="" title="Fig. 1 " /></div> <div class="image-wrapper"><img view_mode="elog_email" class="image-style-plentybig" src="cid:60aa6357621a482b7f9b4772395d074e@logbooks.jlab.org" width="796" height="772" alt="" title="Fig. 2" /></div> <div class="image-wrapper"><img view_mode="elog_email" class="image-style-plentybig" src="cid:86f67532ac8c08f701bf8fd801b7b6c5@logbooks.jlab.org" width="798" height="374" alt="" title="Fig. 3" /></div> </div>
</article> </div>
</div>
</body>
</html>