[Aavso-photometry] How can I beat the FLIP?

[Brad Walter]

I got two pretty good runs of HD37605 on Christmas eve. Seeing was poor
before midnight and the best I could get was .008 mag error for the pre
meridian flip run although measurement consistency improved during the
run. After the flip sigma was down to .004, which isn't too bad for a 10
cm scope. However, when I look at the magnitudes before an after the
meridian flip, the means of the measurements of the two runs are more
than a three sigma confidence interval apart. The only thing I can
attribute this to is change in location of the stars on the CCD. Even
though I spent an hour trying to get the before and after flip FOVs
exactly lined up, they were still several pixels off. I suspect the
small star size of the wide field refractor I am using (FWHM around 2
pixels) makes the situation much worse. 
 
Does anyone have suggestions to reduce this systematic error? I am at 30
sec. integrations now with sharp focus. I thought I would try to
de-focus as much as I can and still stay within 60 sec exposure
duration. Has anyone experimented with dithering? It would hurt the
consistency between individual integrations but since I am making my
measurements on an image resulting from the average of 10 integrations,
it seems as though dithering would have an averaging effect that could
help.
 
Brad Walter
Director of Product Marketing
 
Active Power
2128 West Braker Lane
Austin, TX 78758
Phone: (512) 744 9414
Fax:     (512) 836 4511
bwalter at activepower.com