QE measurement
, made at -100C.
QE measurement
, compared to other EEV CCDs and Loral/Lesser 2k3eb.
Note that the QE measurements were made when the output was non-linear
by +/- 1.5%. No correction to the QE data has been attempted.
334nm flatfield
,cuts are +/- 10% of median level. Strange diamond pattern.
550nm flatfield
,cuts are +/- 3% of median level. Uniformity is nice.
1060nm flatfield
,cuts are +/- 10% of median level. Sensitivity is approximately proportional
to thickness.
Vignetting at corners is caused by the test-stand.
830nm monochromatic flatfield
,cuts are +/- 20% of median level. Fringing peak-to-peak amplitude is
about 40%. Fringes and 1060nm flat seems to indicate that the chip is
thinner in the middle region than at top/bottom. From counting fringes,
a thickness difference of 2.1um peak-to peak is estimated.
Note foam-like structure, also seen in 1060nm flat.
Made at high temperature (-34C), making
bad columns prominent.
The files below contains two graphs each.
The upper one is gain vs. ADU, measured from noise statistics.
This method has proven to have poor precision.
Below, the counts per second for a variety of exposure times are
shown. Corrected for shutter delay and light source drift, the
accuracy of this method is far better.
Linearity
, channel A, low gain, RON = 4.8e-.
Response is linear to within +/- 0.15% over the entire dynamic range of
the ADC: 130000e-. Conversion factor has later been increased a little to
increase full well.
Linearity
, channel A, high gain, RON = 3.7e-.
Some deviating points are seen, which may be due to
testing equipment instabilities.
Excluding these, the response is linear to within +/- 0.15%
up to ADC saturation.
Linearity
, channel B, low gain, RON = 5.3e-.
Response is linear to within +/- 0.15% over the entire dynamic range of
the ADC: 150000e-. Slightly other voltages are used at commissioning,
making the response flatter.
Linearity
, channel B, high gain, RON = 3.9e-.
Some deviating points are seen, which may be due to
testing equipment instabilities.
Excluding these, the response is linear to within +/- 0.15%
up to ADC saturation.
The RON in these measurements is higher than expected mostly due to electronic interference noise.
No precise measurement of CTE has been made, but cosmics does not display any tails in darks, indicating a good CTE.
MTF-graph from the first, crude reductions of the data. The combined MTF of Silicon and pixel geometry is shown. Measurements were made at 670nm and 830nm. A better reduction is underway, but already this confirms the EEV chip to perform much better than the Loral/Lesser 2k3eb devices.
Last updated September 22, 1998 by
norup@astro.ku.dk