dark

[a.walker x295]

At CTIO we have just commissioned our first Loral
3K CCD (3072 x 1024) that has been thinned, packaged and
AR coated by Mike Lesser at U of Arizona.  We UV flood
the CCD using a Zinc lamp, and read it through amplifier
C.  Clock waveforms and all voltages are "normal" Loral.

We find that at 143K the dark rate is approx 0.4 electrons
per hour, which is the lowest I can recall measuring.
CTE is still excellent at this temperature, as is QE.
We ran TI CCDs at this temperature, and EEV's almost
this cold, and the dark was always a few electrons per
hour.

I would be interested in knowing whether there are other
CCDs that achieve this dark rate, and in particular
for Lorals -- maybe it is just a function of this batch
of silicon...  The scientific importance of this is
mostly for low S/N, low background work (eg Echelle
spectroscopy, Fabry-Perot) where exposures are long and
the dark noise contribution ranks with read noise as
the most important noise sources.  Unfortunately with
the Loral 3K we are always read noise dominated for this sort
of work (7 electrons rms read noise).

Alistair Walker