Comparing Photometric Measurements made with
a Kodak KAF-0400 and KAF-0400-ABG CCD
CCD's with anti-blooming gates can suppress the tendency for bright point sources, such as stars, to "bleed" along columns. Kodak offers anti-blooming versions of such CCD's as the KAF-0400, which provide more than 300x protection from charge overflow. Anti-blooming control works by draining off excess charge from the pixels, which would saturate and spill over to other chips in the same column. Modern anti-blooming control as implemented in the Kodak CCD's uses an electronic drain between the columns of the CCD. This Anti-Blooming Gate ("ABG") is called a lateral drain, and it acts like a rain gutter to slough away overflow charge and thus prevent bleeding from occurring, at least up to the capacity of the drain.
While anti-blooming protection may be desirable for some types of CCD imaging, and it permits images to be taken under extremely adverse conditions of high dynamic range, ABG control has the deleterious effect of reducing the quantum efficiency of the CCD chip. For the Kodak CCD, the physical presence of the drain circuitry reduces the collecting area of the CCD to 70% the area of the regular, non-ABG version of the chip. There appear to be further electronic considerations as well, since response measurements of the Kodak KAF-0400 ABG chip indicate that the overall quantum efficiency is approximately 1/2 that of its non ABG sibling. However, a partial compensation for the loss in quantum efficiency appears to be that the ABG chip has about 1/2 the dark current of the regular chip. How these factors affect photometric measurements made with these CCD's is the subject of this Tech Note. Below, we use numerical models to compare the two versions of the KAF-0400 CCD under somewhat "typical" conditions.
Model Results
Signal to Noise ratio models were constructed for an observation made with a Kodak KAF-0400 CCD and the anti-blooming gate ("ABG") version of the KAF-0400 CCD. From these models, the photometric uncertainty, or magnitude error, was calculated and plotted for each chip. The magnitude measurement errors are plotted against CCD temperature, as the temperature controls the dark current, which in turn adds noise to the measurement. The instrumental setup and conditions used by the models are chosen to represent a typical "good" photometric setup by a small college observatory or an advanced amateur astronomer. The models measure a reasonably faint star of V=19.0 against a moderately dark sky of V=20.5. The exposure time was chosen to be 1200 seconds, which may be obtained as 1 single exposure or as a set of shorter exposures (e.g., 5 240-second exposures) and combined in software. The model results are shown below; model details are given in the figure caption.
In the figure below, the lower the error, the better the photometric precision. Thus the lower curve wins. It is clearly shown that the "regular" KAF-0400 gives more than 40% lower magnitude error at all temperatures. Models for a brighter star would give a similar result in that the ABG results are always worse by at least 40% at a given magnitude. It is also important to note that cooling the CCD below about -15 to -20C gives little gain in photometric precision.
|
