- 2016B Programs and Schedule Announced
- Gemini Home
- Telescopes and Sites
- Science Visitors at Gemini
- Observing With Gemini
- Retired Instruments
- Status and Availability
- Nod and Shuffle
- Camera Properties
- Detector Array
- Flatfield Features (GMOS-N)
- Observing Strategies
- Spectroscopy Overview
- Long-slit Spectroscopy
- Multi-Object Spectroscopy
- Integral Field Spectroscopy
- ITC, Sensitivity and Overheads
- Guiding Options
- Observation Preparation
- Data Format and Reduction
- Visiting Instrument Policy
- Visiting Instrument Telescope Interfaces
- DSSI Speckle Camera
- TEXES (North)
- Integration Time Calculators
- Magnitudes and Fluxes
- Near-IR Resources
- Mid-IR Resources
- Observing Condition Constraints
- Performance Monitoring
- SV/Demo Science
- Future Instrumentation & Current Development
- Queue and Schedules
- Data and Results
- Gemini Research Staff
Change page style:
GMOS-S Array (Hamamatsu)
The upgraded GMOS-S detector array consists of three ~ 2048x4176 Hamamatsu chips arranged in a row. Two of the detectors (CCDr and CCDg) have an enhanced red response, these CCDs are referred to by the ITC as "Hamamatsu Red". The right-most CCD (CCDb, the blue end of spectral dispersion) in the focal plane array has improved blue response in addition to red response very similar to the Hamamatsu Red CCDs. This third CCD is referred to by the ITC as "Hamamatsu Blue". The orientation of the CCDs has not changed and continue to support the Nod and Shuffle observing mode. The plot below gives the anticipated QE comparison to the current E2V CCDs in GMOS-S. These QE plots are taken from general Hamamatsu information and lab measurements done in Hilo. See the Status and Availability webpage for more details.
QE Comparison for the GMOS-S CCDs upgrade.
The table below summarizes some of the Hamamatsu detector/controller characteristics (updated October 2015)
|Pixel format||6266x4176 pixels (mosaiced)|
|Array layout||Three 2048x4176 chips in a row with 61pix=0.915mm gaps|
|Pixel size||15 microns square; 0.080 arcsec/pixel|
|Spectral Response||approx 0.36 to 1.03 microns [ Hamamatsu Red data / Hamamatsu Blue data / plot ]|
|Bias level||Bias Image|
|Flat field response||not yet available|
|Readout time||1x1: 83sec(slow)/35sec(fast) 2x2: 24sec(slow)/11sec(fast)|
|Chip ref no.||BI5-36-4k-2||BI11-33-4k-1||BI12-34-4k-1|
|Dark current||~ 3e-/hr/pix||~ 3e-/hr/pix||~ 3e-/hr/pix|
|Full Well||~120,000 e-||~122,000 e-||~115,000 e-|
|Fringing at 900nm||<2%||<2%||<2%|
Readnoise and Gain Values (*)
The table below gives gain/read-noise values for the new GMOS Hamamatsu CCDs operating with the SDSU controller. The values are averaged over all 12 amps.
|rate||level||Gain (e-/ADU)||noise (e- rms)|
The Slow Read / High Gain mode will not be offered for the Hamamatsu CCDs as it has been deemed to be of little scientific use. Slow Read / Low Gain is the primary mode for science use. Fast Read / Low Gain may be of use, for example, with acquisition observations or for time resolved observations. Fast Read / High Gain is expected to be used primarily for very bright targets.
(*) These values were updated after the installation of new video boards, and are valid for data taken from September 1, 2015 onwards. For data taken between July 2014 and August 2015, please look at the previous values.
The linearity is better than 0.5% up to 60k ADU counts in these detectors, as measured from data taken during July-September 2014.
Cosmic hit rate
The cosmic ray hit rate is higher on the Hamamatsu CCDs - about 2.5 times more pixels are affected by cormic rays, as compared to the E2Vs.
Saturation banding in binned data (before september 2015)
The issue described below has been fixed as of August 31, 2015 and is currently no longer present.
This effect is a lowering in counts (with respect to the bias level) that happens when one or more pixels saturate, affecting the whole amplifier width. Examples are shown below, for imaging (left) and spectreoscopy (right). In the imaging example the "band" can be seen as it spans over the width of the corresponding amplifier. The counts within the "band" are lower (by a few hundreds) than the bias level (which is ~3000 ADU). For spectroscopy it looks more dramatic since it is a saturated arc line therefore the whole amplifier is 'lowered'. It does nor appear in 1x1 binning though.
The cause is understood (it is originated in the controller electronics, it is not a problem of the detector itself) and a fix implementation is currently (as of December 2014) under study; will probably be performed during 2015A (TBC)