WFS Home

Science
  WFS Overview
  Science Programs
  Survey Fields
  Publications

Data Access
  Archive
  Relase
  Cookbook
  DQC interface

Image Gallery

Technical
  WFC obs notes
  WFC CCD
  WFC foibles
  WFC filters
  WFC pipeline
  Observing logs
  Photometry
    Colour Equations
    Zero Points
  Flats+Biases
  Defringing
  Astrometry

What's New

Help
  
Search Site
  F.A.Q.
  Web Changes
  Getting Help
  About

WFC foibles

During the April 1998 commissioning phase and whilst developing the pipeline processing software for the WFS, several interesting "features" specific to the WFC were noticed and logged. Since these may be of more general interest than solely for WFS applications we have collected them together in the following list. If anyone out there finds any other unusual WFC instrumental signatures please contact either Nic Walton, naw@ing.iac.es , or Mike Irwin, mike@ast.cam.ac.uk , so that we can investigate and update this problem page as appropriate.


The following polynomials approximate to the LUTs used in the semester 1998B pipeline reduction and convert the bias-corrected CCD data onto the CCD#1 system, note that -ve values relative to bias level are left the same.

August 1998
CCD#2 true = 0.987086*value - 1.06221E-06*value**2 + 5.19301E-13*value**3
CCD#3 true = 0.993867*value + 4.15398E-07*value**2 - 1.02134E-12*value**3
CCD#4 true = 0.987744*value - 1.15224E-06*value**2 + 2.58758E-12*value**3

September 1998
CCD#2 true = 1.000010*value - 1.75546E-06*value**2 + 1.13966E-11*value**3
CCD#3 true = 0.998063*value + 2.51351E-07*value**2 - 8.71941E-13*value**3
CCD#4 true = 0.994517*value - 1.49265E-06*value**2 + 8.10674E-12*value**3

October 1998
CCD#2 true = 0.991703*value - 1.40928E-06*value**2 + 5.84305E-12*value**3
CCD#3 true = 0.993855*value + 5.36676E-07*value**2 - 5.50707E-12*value**3
CCD#4 true = 0.995565*value - 1.10012E-06*value**2 + 5.77076E-12*value**3

November 1998 - August 1999 inclusive, the non-linearity is identical within the errors of measuring it, to that of October 1998.

The new DAS based on SDSU controllers was commissioned at the end of August 1999. This uses new controller electronics and ADCs. Since we suspected most of the non-linearity originated in the electronics/ADCs it was not surprising to find that new transforms are needed. We currently only have non-linearity measure for September and October 1999 which give the following coefficients:

CCD#1 true = 1.0*value - 2.0E-06*value**2 + 8.0E-12*value**3
CCD#2 true = 1.0*value - 0.5E-07*value**2 - 4.0E-12*value**3
CCD#3 true = 1.0*value - 6.0E-07*value**2
CCD#4 true = 1.0*value - 1.5E-07*value**2 - 2.0E-12*value**3

Note that none are now linear and that since the relative "slope" is arbitrary until after flatfield renormalisation we have simplified the transforms and set the first coefficient to unity for all devices. For the September and October 1999 data these transforms are good to between 0.1-0.2% in the range 0-50K counts.

Subsequent measurements in March 2000 showed that the non-linearity of CCD#1 changed sometime between the WFS run in December 1999 and the first WFS run in March 2000. The new values for CCD#1 are

CCD#1 true = 1.0*value - 2.5E-06*value**2 + 1.2E-11*value**3

The other CCDs remained as before. Subsequent linearity checks in June 2000 showed the same set of coefficients found in March 2000.

These coefficients remained unchanged until a significant hardware upgrade during summer 2003. Measurements in August 2003 showed the new coefficients to be:

CCD#1 true = 1.0*value - 7.0E-07*value**2 + 2.0E-12*value**3
CCD#2 true = 1.0*value - 1.5E-07*value**2
CCD#3 true = 1.0*value
CCD#4 true = 1.0*value - 2.5E-07*value**2 + 4.0E-12*value**3

But just as thought you were on top of this, subsequent measurements in March 2004 showed they had changed again - d'oh! We have no idea when, or if this simply reflects alternative operational choices.

CCD#1 true = 1.0*value - 2.5E-06*value**2 + 1.5E-11*value**3
CCD#2 true = 1.0*value + 1.9E-07*value**2 - 4.0E-12*value**3
CCD#3 true = 1.0*value - 4.0E-07*value**2
CCD#4 true = 1.0*value + 0.7E-07*value**2 - 2.0E-12*value**3

After yet more heroic non-linearity measurement taken during November 2004, using default readout options etc., we find we are back with the August 2003 coefficients - don't call us !

CCD#1 true = 1.0*value - 7.0E-07*value**2 + 2.0E-12*value**3
CCD#2 true = 1.0*value - 1.5E-07*value**2
CCD#3 true = 1.0*value
CCD#4 true = 1.0*value - 2.5E-07*value**2 + 4.0E-12*value**3