[Gb-ccb] some lab test results

[Brian Mason]

Thanks Rich-- I'm glad someone is being critical. I was quite careful 
about units on both axes and in fact the ccb configuration for the scan in 
question was such that the integration time was 2 ms resulting in a 
maximum nyquist-sampled frequency of 250 Hz (which is exactly what my 
plots go to). You can verify the config in the ccb fits file header and 
the time stamp deltas in the FITS file are consistent with the 2ms 
integration time as well.  I'm a little puzzled.  It is possible that the 
10 MHz wasn't plugged in, or its signal generator was not at precisely the 
right frequency, so the CCB clock was running a little slow: I don't 
believe I checked this part of the setup when doing the hot/colds, I just 
verified the receiver connections and checked that we had reasonable data 
in the demo client.

Since we were going to repeat this anyway with the better cables & on the 
telescope, we'll see what happens with our nie 60 Hz calibrator signal 
then...

  Brian

--------------------------------------------------------------------
Brian Mason                           | office: +1(304)456-2338
Associate Scientist                   | fax:    +1(304)456-2229
National Radio Astronomy Observatory  | mail:   PO Box 2 
bmason at gb.nrao.edu                    |         Green Bank, WV 24944
http://www.gb.nrao.edu/~bmason/       |
--------------------------------------------------------------------

On Wed, 26 Oct 2005, Richard Lacasse wrote:

> Brian,
>
> One suspicious thing that I noted in your plots is that your 60 Hz harmonics 
> (60, 120 and 180) are all systematically low.  There must be a scaling factor 
> off somewhere?
>
> Rich
>
> Brian Mason wrote:
>
>> 
>> hi all- I've been looking at the ccb lab data collected last week and still 
>> have more to do, but here is a progress report.
>> 
>> 1) Tsys & Tcal measurements.  I collected data with hot and cold loads, 
>> with no cals, one cal, and the other cal firing. Results are in the 
>> attached ccbTsys pdf file (Tsys and Tcal values) as well as analys19oct, 
>> which shows a more detailed consideration of non-linearity.  The png file 
>> shows a plot of the non-saturated channels (10 and 11 saturate against a 
>> hot load) Trx values vs frequency. At a level of 10 or 15% I am not 
>> confident in the overall absolute calibration of Trx because our cold load 
>> is a little sketchy, and my room temperature is just a nominal 300 K. We'll 
>> do a better measurement on the telescope (at least record the real ambient 
>> temperature).
>> 
>> More importantly there is dramatic gain compression in some of the 
>> channels.  I measure this by comparing the cal diode on - cal diode off 
>> signal vs a hot load and a cold load respectively-- the cal signal in 
>> counts is in one case, against a hot load, only 50% of the cal signal 
>> against the cold load (this is one of the channels with -20.7 dBm into the 
>> detector). The raw total power counts against either hot or cold load are 
>> stable over the duration of the measurements to ~1% so overall gain drifts 
>> aren't a big factor here.
>> 
>> Without a model or measurements of how the linearity varies for the system 
>> as a whole vs input power level it's hard to correct for 10+% 
>> nonlinearities or to have confidence in the Trx or Tcal results.
>> 
>> In spite of the sometimes large gain compression I am reluctant to 
>> recommend we consider hardware changes at present even were such changes 
>> straightforward.  We have a good baseline of tests with the current system; 
>> and we don't have a firm assessment of where we are with respect to the 
>> post-detector noise floor yet.  Data with the full CCB on the telescope 
>> (and possibly with the receiver off, much later) may be helpful to form an 
>> overall picture of what's going on so we can come up with a reasonable 
>> solution should that be needed. It is not very surprising to me that over 
>> such a huge RF band it is a challenge to balance things equally as well as 
>> meet our stringent noise requirements!
>> 
>> 2) power spectra: I collected 100 sec against a cold load with no cal diode 
>> firing and derived tp and differenced power spectra. They are at:
>> 
>> http://wiki.gb.nrao.edu/bin/view/Projects/CcbNineteenOctSpectra
>> 
>> There is visible 60 Hz and other stuff, variable from channel to channel 
>> and usually present in the differenced data.  Considered in quadrature to 
>> the other noise, the line contamination typically introduces << 1% to the 
>> total noise RMS in the differenced data although in a couple of cases it's 
>> up to 2 or 3%. There are some low-frequency zooms on that page as well to 
>> look for microphonics.
>> 
>> These tests were with the box not properly sealed up, and the old/bad 
>> cables, so should be taken with a grain of salt. I will repeat the test on 
>> the telescope with the final system.
>> 
>>  Brian
>> --------------------------------------------------------------------
>> Brian Mason                           | office: +1(304)456-2338
>> Associate Scientist                   | fax:    +1(304)456-2229
>> National Radio Astronomy Observatory  | mail:   PO Box 2 bmason at gb.nrao.edu 
>> |         Green Bank, WV 24944
>> http://www.gb.nrao.edu/~bmason/       |
>> --------------------------------------------------------------------
>> 
>> ------------------------------------------------------------------------
>> 
>> 
>> ------------------------------------------------------------------------
>> 
>> _______________________________________________
>> gb-ccb mailing list
>> gb-ccb at listmgr.cv.nrao.edu
>> http://listmgr.cv.nrao.edu/mailman/listinfo/gb-ccb
>
>
> _______________________________________________
> gb-ccb mailing list
> gb-ccb at listmgr.cv.nrao.edu
> http://listmgr.cv.nrao.edu/mailman/listinfo/gb-ccb
>