• Introduction
• From dsimulator to detector plane
• CCD issues
• 2023-02-26: All CCDs (B amplifiers) show nominal bias noise and response to illumination
• 2023-01-06: CCD5 high bias noise with gradient but nominal throughput
• 2020-04-23: All CCDs (B amplifiers) show nominal bias noise and response to illumination
• 2020-09-15: CCD5 high bias noise but nominal throughput
• 2020-02-14: CCD5 bad response to illumination

Introduction

This page describes issues that are currently affecting the DEIMOS science detector. The page will be updated regularly with the latest status.

Each CCD on the detector mosaic can be read using two amplifiers. In spectroscopy mode, we use the B amplifiers (AMPMODE = SINGLE:B), because A amplifiers are overall noisier. In direct imaging mode, we normally use both amplifiers (AMPMODE = DUAL:A+B) to read the detector, because the readout time is twice as fast than with one amplifier for mask alignments. It is not possible to read the detector using both amplifiers in spectroscopy mode.)

February 26, 2023, the DEIMOS detector mosaic presents the NO ISSUES when read out with B amplifiers. The following issues with A amplifiers are still present, but the do not affect science operations (see CCD status for details):

• CCD5A is not sensitive to illumination. Therefore, DEIMOS spectra should always be taken with the B amplifiers
• CCD3A bias frames have a horizontal noise pattern. This does not affect spectroscopic observations. It does not affect mask alignment images either. It only affects science observations taken in direct imaging mode. This issue can be mitigated by taking direct imaging science data with the B amplifiers instead of both amplifiers.
• CCD6A bias has all pixels at 1024 counts.
• CCD8A has a lower gain than the rest of the CCD mosaic. This issue does not affect either spectroscopy or direct imaging observations, but it prevents us from using A amplifiers as an alternative when there are issues with B amplifiers. Additionally, CCD8A bias frames have a horizontal noise pattern.

January 6, 2023, the DEIMOS detector mosaic presents the following issues (see CCD issues for details):

• CCD5 spectra taken with the B amplifier (CCD5B) has a bias noise about 3 times higher than nominal and a gradient along the detector rows. Its response to illumination is nominal. Therefore the spectra of targets placed on CCD1 will have a readnoise component higher than the spectra of targets placed on other CCDs.
• CCD5A is not sensitive to illumination. Therefore, DEIMOS spectra should always be taken with the B amplifiers
• CCD3A bias frames have a horizontal noise pattern. This does not affect spectroscopic observations. It does not affect mask alignment images either. It only affects science observations taken in direct imaging mode. This issue can be mitigated by taking direct imaging science data with the B amplifiers instead of both amplifiers.
• CCD6A bias has all pixels at 1024 counts.
• CCD8A has a lower gain than the rest of the CCD mosaic. This issue does not affect either spectroscopy or direct imaging observations, but it prevents us from using A amplifiers as an alternative when there are issues with B amplifiers. Additionally, CCD8A bias frames have a horizontal noise pattern.

April 23, 2022, the DEIMOS detector mosaic presents the NO ISSUES when read out with B amplifiers. The following issues with A amplifiers are still present, but the do not affect science operations (see CCD status for details):

• CCD5A is not sensitive to illumination. Therefore, DEIMOS spectra should always be taken with the B amplifiers
• CCD3A bias frames have a horizontal noise pattern. This does not affect spectroscopic observations. It does not affect mask alignment images either. It only affects science observations taken in direct imaging mode. This issue can be mitigated by taking direct imaging science data with the B amplifiers instead of both amplifiers.
• CCD6A bias has all pixels at 1024 counts.
• CCD8A has a lower gain than the rest of the CCD mosaic. This issue does not affect either spectroscopy or direct imaging observations, but it prevents us from using A amplifiers as an alternative when there are issues with B amplifiers. Additionally, CCD8A bias frames have a horizontal noise pattern.

September 15, 2020, the DEIMOS detector mosaic presents the following issues (see CCD issues for details):

• CCD5 spectra taken with the B amplifier (CCD5B) has a bias noise about 4 times higher than nominal. Its response to illumination is nominal. Therefore the spectra of targets placed on CCD1 will have a readnoise component higher than the spectra of targets placed on other CCDs. The same issue affects to CCD5 when read with the A amplifier.
• CCD5A is not sensitive to illumination. Therefore, DEIMOS spectra should always be taken with the B amplifiers
• CCD3A bias frames have a horizontal noise pattern. This does not affect spectroscopic observations. It does not affect mask alignment images either. It only affects science observations taken in direct imaging mode. This issue can be mitigated by taking direct imaging science data with the B amplifiers instead of both amplifiers.
• CCD6A bias has all pixels at 1024 counts.
• CCD8A has a lower gain than the rest of the CCD mosaic. This issue does not affect either spectroscopy or direct imaging observations, but it prevents us from using A amplifiers as an alternative when there are issues with B amplifiers. Additionally, CCD8A bias frames have a horizontal noise pattern.

February 14, 2020, the DEIMOS detector mosaic presents the following issues (see CCD issues for details):

• CCD5 spectra taken with the B amplifier (CCD5B) look like a bias when illuminated with photons. This isssue has been occuring intermittently for over a year. It affects the spectra of slits placed on CCD1. The same issue affects to CCD5 when read with the A amplifier.
• CCD3A bias frames have a horizontal noise pattern. This does not affect spectroscopic observations. It does not affect mask alignment images either. It only affects science observations taken in direct imaging mode. This issue can be mitigated by taking direct imaging science data with the B amplifiers instead of both amplifiers.
• CCD8A has a lower gain than the rest of the CCD mosaic. This issue does not affect either spectroscopy or direct imaging observations, but it prevents us from using A amplifiers as an alternative when there are issues with B amplifiers.

From dsimulator to detector plane

This section describes the relative orientation between the dsimulator user interface, mask design blueprints and detector plane. This is intended to serve as a reference to determine which areas on the dsimulator interface should be avoided to prevent the spectra of the highest priority targets from being affected by any given issue on the detector mosaic. Plesae, click on the images below for an expanded view.

1. Mask location on the detector plane

The next figure shows the location and orientation of the instrument focal plane with respect to the detector as it is normally seen on a modern DS9 display tool. The image was opened using DS9 with the option to match the coordinate system to the WCS. This is not a sky WCS but a detector WCS that contains the relative location between amplifiers on a DEIMOS mosaic image. This is also how the image looks like when opened with the DEIMOS real-time DS9 GUI. The image was taken with one of the internal calibration lamps and no mask inserted in the optical path. Note that CCD5 looks like a bias due to the issue described reported on February 14, 2020.

The next image was also illuminated with one of the internal calibration lamps, but a mask was inserted in the optical path (please, expand the image for a better visualization of the slits and alignment boxes).

Next figure shows an internal spectral flat taken with the same slitmask that was used to take the mask image shown in the previous figure.

2. Blueprint and dsimulator user interface

The following figure shows the blueprint corresponding to the mask shown in the previous section. Note that the blueprint is rotated by 180 degrees with respect to the mask image displayed with DS9. CCD1 is on the right-hand side in the blueprint, instead of the left-hand side.

The following image shows the dsimulator user interface for a different mask than the one shown in the image above. It is important to note that the disimulator orientation matches the blueprint orientation. The slitmask diagram shown in the dsimulatior is rotated by 180 deg with respect to the image displayed with DS9.

CCD issues

2023-02-26: All CCDs (B amplifiers) show nominal bias noise and response to illumination

The following figure shows a flat and a bias frame taken with the B amplifiers on March 14, 2023. The bias noise and response of illumination of all CCDs is nominal.

2023-01-06: CCD5B high bias noise with gradient but nominal throughtput

The following figure shows a flat and a bias frame taken with the B amplifiers on January 6, 2023.

2022-04-23: All CCDs (B amplifiers) show nominal bias noise and response to illumination

The following figure shows a flat and a bias frame taken with the B amplifiers on April 23, 2022. The bias noise and response of illumination of all CCDs is nominal.

2020-09-15: CCD5B high bias noise but nominal throughtput

The following figure shows a flat and a bias frame taken with the B amplifiers on October 17, 2020. The CCD5B bias is noisier the rest of the CCDs. The flat frame displayed on the left-hand side of the figure is nominal for all CCDs, including CCD5B. Note also the horizontal bands on the CCD8B bias.

The following figure shows a flat and a bias frame taken with the A amplifiers on October 17, 2020. The CCD5A bias is nominal, but CCD5A does not respond to illumination. Note also the horizontal bands on the CCD3A bias. The CCD6A bias has all pixel values at 1024 counts. The CCD8A flat counts are about half the flat counts on the neighbor CCDs.

2020-02-14: CCD5 bad response to illumination

The following figure shows a flat and a bias frame taken on February 14, 2020. The bias frame on the right-hand side of the figure looks normal. The flat frame displayed on the left-hand side of the figure does not contain any signal on CCD5. Therefore, spectra from slits placed on the the CCD below CCD5, i.e. CCD1, will be missing the red side of the spectrum.