<!DOCTYPE html PUBLIC "-//W3C//DTD HTML 4.01 Transitional//EN">
<html>
<head>
<meta http-equiv="Content-Type" content="text/html;charset=ISO-8859-1">
<title></title>
</head>
<body text="#000000" bgcolor="#ffffff">
Martin Shepherd wrote:<br>
<blockquote type="cite"
cite="midPine.LNX.4.63.0510021956550.29275@goblin.caltech.edu"><br>
Using a breakout cable that JD made for me on Friday, over the weekend
<br>
I managed to use one of the GPIO-card's miscellaneous analog outputs
<br>
to drive known DC voltages into the differential inputs of the
<br>
CCB. This should make the tests that involve inserting known DC
<br>
voltages at the CCB inputs, much less tedious and more versatile than
<br>
setting these voltages by hand, as we had previously planned.
<br>
<br>
Using this new feature, I now have a test program that generates a
<br>
staircase ramp of known voltages at one of the CCB inputs, and at each
<br>
step in this staircase, measures the mean and standard deviation of a
<br>
frame of dump-mode samples. At the moment, I'm just printing these
<br>
numbers to the screen, in tabular form, but my next task will be to
<br>
include code to compute a weighted least squares fit of the numbers to
<br>
a straight line, in order to work out the ADC gain, zero-offset and
<br>
goodness of fit. I'll also try plotting the generated values in
<br>
gnuplot.
<br>
<br>
Note that in order to be able to take over control of the CCB from the
<br>
manager (which was left connected over the weekend), I reconfigured
<br>
the CCB server not to accept connection requests from computers in the
<br>
192.33.116.* IP address range. So I imagine that there will be lots of
<br>
error messages in the manager's log regarding failed connection
<br>
attempts. I have reenabled access to this block of addresses now.
<br>
<br>
Martin
<br>
_______________________________________________
<br>
gb-ccb mailing list
<br>
<a class="moz-txt-link-abbreviated" href="mailto:gb-ccb@listmgr.cv.nrao.edu">gb-ccb@listmgr.cv.nrao.edu</a>
<br>
<a class="moz-txt-link-freetext" href="http://listmgr.cv.nrao.edu/mailman/listinfo/gb-ccb">http://listmgr.cv.nrao.edu/mailman/listinfo/gb-ccb</a>
<br>
<br>
</blockquote>
Martin ( & All),<br>
<br>
This seems like a reasonable (and clever) test; however, please be
advised<br>
that, at this time, the modified detector amplifier board we're using
to convert<br>
single-ended signals to their differential counterparts has NOT been
carefully<br>
calibrated against any sort of standard. Actually, (unless someone
else has<br>
"tweaked" its calibration in my absence), it has only received a very
crude<br>
adjustment which was only intended to verify its functionality. Given
this, I'd<br>
expect the proposed test to yield results which should exhibit a
reasonable<br>
degree of precision; but, most likely, very poor accuracy...<br>
<br>
Randy<br>
</body>
</html>