<html>
<head>
<meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
</head>
<body bgcolor="#FFFFFF" text="#000000">
<div class="moz-cite-prefix">Hi Richard,<br>
<br>
I took another look at your scope pictures. The scale is 10ns/div
or 2ns per sub-division. Take 2 of these small divisions (4ns) and
slide them by 2ns: I easily see a difference of 30% in the peak
value that the ADC would register.<br>
<br>
You can also change the scope settings so that the sampling on the
scope is 250MSPS, adjust the trigger level and set to envelope.
You should see the 30% effect.<br>
<br>
Best regards,<br>
Fernando<br>
<br>
<br>
On 2/13/2014 11:11 AM, Fernando J Barbosa wrote:<br>
</div>
<blockquote cite="mid:52FCEEA5.5050209@jlab.org" type="cite">
<meta content="text/html; charset=UTF-8" http-equiv="Content-Type">
<div class="moz-cite-prefix">Hi Richard,<br>
<br>
You may also want to check your software and get a raw data dump
to confirm.<br>
<br>
Best regards,<br>
Fernando<br>
<br>
<br>
<br>
On 2/13/2014 10:22 AM, Fernando J Barbosa wrote:<br>
</div>
<blockquote cite="mid:52FCE351.1060102@jlab.org" type="cite">
<meta content="text/html; charset=UTF-8"
http-equiv="Content-Type">
<div class="moz-cite-prefix">Hi Richard,<br>
<br>
The fADC250 has a linearity of ~1% full scale and also a
ch-to-ch gain variation of 1%. We measured these before and
during production with pulses and levels. There is no
processing at all in raw data mode. Your signals are so close
in amplitude and so far away from saturation that I doubt
there is any non-linearity effects, other the edge speed and
sampling as I mentioned earlier. You can easily check this
with a ramp signal or a precision variable attenuator.<br>
<br>
I would check with another scope channel and another ADC
channel to be sure there are no inconsistencies.<br>
<br>
Best regards,<br>
Fernando<br>
<br>
<br>
On 2/13/2014 1:01 AM, Richard Jones wrote:<br>
</div>
<blockquote
cite="mid:CABfxa3Rk2qiS+Br4v6C2Q992jn6eAMgBA73Ug-xrD=cPhjG10g@mail.gmail.com"
type="cite">
<div dir="ltr">Fernando,
<div><br>
</div>
<div>We cannot use the scope for these tests because we can
only see a few channels at a time. We would very much
like to understand the DAQ so we can move on and start
fiber QA studies. The decay time of these signals is so
slow that the rise time will not affect the max pulse
height by more than about 10%. Running the signals shown
on the scope through a 125MHz low-pass filter reduces the
pulse height somewhat, but does not affect the ratios.
There are variations based on the exact delay through
each channel and the FADC clock, but these effects are +/-
10% because of the slow decay time. As you state, we are
comparing ratios of 40% on the scope and 15% on the ADC. </div>
<div><br>
</div>
<div>We do not show the pictures on the wiki, but if we swap
the inputs on the FADC, the results do not change by more
than a couple percent. This shows that it cannot be gain
variations between the different channels on the FADC, but
must be some kind of signal processing that is doing very
different things to the peak heights depending on what
preamp output generated them. Is there possibly some
nonlinear processing of the raw FADC values that is
happening in the module before we see them? As Alex
points out, the pulse shapes of channels 1 and 3 are
virtually identical, but the ratio of their amplitudes
disagrees: 6% on the FADC vs 20% on the scope.</div>
<div><br>
</div>
<div>-Richard J.</div>
</div>
<div class="gmail_extra"><br>
<br>
<div class="gmail_quote">On Thu, Feb 13, 2014 at 12:16 AM,
Fernando J. Barbosa <<a moz-do-not-send="true"
href="mailto:barbosa@jlab.org">barbosa@jlab.org</a>>
wrote:<br>
<blockquote class="gmail_quote">Hi Richard,<br>
<br>
From the pictures, I get for each set the following peak
amplitudes:<br>
1) 64mV, 278 channels=4.3 ch/mV<br>
2) 46mV, 248 ch=5.4 ch/mV<br>
3) 76 mV, 295 ch=3.9 ch/mV<br>
<br>
Clearly not linear where a 40% change on the scope
corresponds to 16% on the ADC. This is because we are
looking at peak values and the pulse rise time is less
than 4ns (ADC sampling). If you are triggering the ADC
from the laser pulser, try adding cables in 1ns
increments to see how the ADC amplitude changes. The
pulse risetime is just too fast for the ADC to sample
properly (must be > 4ns). Because the pulse duration
is much longer than the pulse risetime, charge should be
OK as it is to be used in the experiment. The scope is
the better tool for your tests of the fibers unless you
put a low pass filter to increase the pulse risetime.<br>
<br>
Best regards,<br>
Fernando<br>
<div>
<div class="h5"><br>
<br>
<br>
<br>
----- Original Message -----<br>
From: Alexander Somov <<a moz-do-not-send="true"
href="mailto:somov@jlab.org">somov@jlab.org</a>><br>
To: Richard Jones <<a moz-do-not-send="true"
href="mailto:richard.t.jones@uconn.edu">richard.t.jones@uconn.edu</a>><br>
Cc: Hall D beam working group <<a
moz-do-not-send="true"
href="mailto:halld-tagger@jlab.org">halld-tagger@jlab.org</a>><br>
Sent: Wed, 12 Feb 2014 22:02:13 -0500 (EST)<br>
Subject: Re: [Halld-tagger] first results on fiber
light yields<br>
<br>
<br>
Hello Richard,<br>
<br>
Actually fadc spectra for 1) and 3) agree reasonably
well<br>
<br>
Why 2) is about 20% wider (was it a different
readout channel)?<br>
<br>
(there could be some sampling effects due to the
different pulse<br>
shape around the peak. I would also compare amps
without the<br>
splitter).<br>
<br>
Cheers,<br>
Alex<br>
<br>
<br>
<br>
<br>
<br>
On Wed, 12 Feb 2014, Richard Jones wrote:<br>
<br>
> Dear colleagues,<br>
><br>
> We are working to obtain light yields for the
first article fiber bundle<br>
> that was produced for the tagger microscope
some time back. We have run<br>
> into a problem trying to interpret what we see
coming from the FADC250<br>
> module that is installed in the data
acquisition crate we are using for<br>
> fiber QA tests. Please see the plots on the
wiki page linked below that<br>
> illustrate the problem we are seeing.<br>
><br>
> <a moz-do-not-send="true"
href="https://halldweb1.jlab.org/wiki/index.php/Tagger_Microscope#Light_yield_tests">https://halldweb1.jlab.org/wiki/index.php/Tagger_Microscope#Light_yield_tests</a><br>
><br>
> When we look at the signals on the scope, we
see one set of amplitude<br>
> ratios that clearly reveal the differences in
light yield resulting from<br>
> different lengths of the fibers that are
covered in reflective paint. The<br>
> results from the FADC are much less
distinguished, and disagree in the<br>
> ratios of pulse heights. Comments or
suggestions as to what we are doing<br>
> wrong would be appreciated.<br>
><br>
> -Richard Jones<br>
><br>
</div>
</div>
_______________________________________________<br>
Halld-tagger mailing list<br>
<a moz-do-not-send="true"
href="mailto:Halld-tagger@jlab.org">Halld-tagger@jlab.org</a><br>
<a moz-do-not-send="true"
href="https://mailman.jlab.org/mailman/listinfo/halld-tagger">https://mailman.jlab.org/mailman/listinfo/halld-tagger</a><br>
<br>
</blockquote>
</div>
<br>
</div>
</blockquote>
<br>
</blockquote>
<br>
</blockquote>
<br>
</body>
</html>