Change page style: 



Canopus is the Adaptive Optics bench of GeMS . Below we described the characteristics of the main component of Canopus.

Content on this page requires a newer version of Adobe Flash Player.

Get Adobe Flash player

Calibration Sources

Canopus is equipped with a set of calibration sources. These sources are used to perform all the day-time calibrations, and to exercise the system in a control environment. The set of calibration sources:
  • 5 Laser Guide Sources: These are 0.8" sources (500um in the f/16 entrance focal plane). They can be moved following the optical axis to simulate sources from 90km to 160km
  • 6 Natural Guide Sources: These are 0.6" sources(300um in the f/16 entrance focal plane). They are conjugated to infinite, and they are used to simulate NGS seen through typical seeing conditions, and they are used for the calibrations of the NGSWFS
  • 24 Natural Guide Sources: These are diffraction limited (10um in the f/16 entrance focal plane). They are conjugated to infinite and they are used for Non Common Path Abberation (NCPA) calibrations and performance evaluation

Deformable Mirrors

GeMS uses 3 Deformable Mirrors (DMs) conjugated respectively at 0km, 4.5km and 9km. The characteristics of the DMs are summarized in the table below:

  DM0 DM4.5 DM9
Altitude (km) 0 4.5 9
Total actuators 293 416 208
Valid actuators 240 324 120
Slave actuators 53 92 88
Actuator coupling 33% 33% 33%
Pitch 5mm 5mm 10mm
Magnification 7.9m -> 80mm
Layout notes In the layout plots above, black crosses, red numbers show valid actuators. Blue crosses, blue numbers show slave actuators. The dotted lines mark the various beams (for WFS 0 through 4). For the altitude DMs layout (DM4.5 and 9), the solid line marks the outer edge of the WFS beams, that is, it encloses the sensed part of the DM (dotted and solid lines computed for a LGS altitude of 90km, i.e. at zenith).
These DMs are based on a piezo-stack technology (developed by CILAS) the DM electronic has been made by XX

Tip-Tilt Mirror

The Tip/Tilt Mirror Assembly is composed of a custom Tip/Tilt stage, a Silicon Carbide lightweighted mirror, a kinematic mounting interface for the Tip/Tilt Stage, and the assembly aluminum mount. Our TT mirror is located just downstream from DM0, and has a stroke of +/- 1.4 arcsec, with a bandwidth of 400Hz. It's a Physik Instrument device.

Science Beam Splitter Assembly

The science dichroic let the NIR through to the science instrument and reflects everything downward of 1 micron into the WFSs environment. The mechanism has 2 positions: the second dichroic has a cutoff at 850nm, to allow science at z band.

Science Atmospheric Dispersion Compensator

At large zenith angle, a compensation for atmospheric refraction within the science instrument bandpass (for broadband filters) will be needed. This can be done via the science Atmospheric Dispersion Compensator (ADC). The Science ADC can be IN or OUT the optical path. If it is not really required, it is recommended to leave it OUT, as it affects the transmission.

Output Focal

Canopus modifies the entrance f/16 beam to a f/32.5 beam

Laser notch filter

It is a dichroic that reflects the 589nm light into the LGSWFS and let the rest of the visible light through into the TTWFS.


Canopus is using five Laser Guide Star Wave-Front Sensors (LGSWFS) each one looking at one LGS. Characteristics of these LGSWFS are described in the table below:


Those are EEV39 80x80. They are controlled by a SDSU controller (San Diego State University). Each subapertures use quadcell, with 1.38arcsec/pixels (thus 2.76arcsec field of view).


Because the LGS range varies from 90km (at zenith) to 180km (at zenith angle 60 degrees), plus the natural variation of the sodium layer altitude, we need the LGS WFS to accomodate this variations. This is taken care of by the LGS WFS zoom. There are many active elements in the LGS WFS which job is to:
  1. compensate for the range induced defocus, and doing so
  2. keep the wavefront aberrations small, and
  3. keep the deformable mirror / lenslet registration error small.
This is done by a combination of Look Up Tables (LUT) and active loop (see the SFS loop).


This device includes 3 TTWFS, each one a quadcell detector using Avalanche Photodiodes (APDs). Each one can be moved into position using X and Y linear stages. To facilitate dithering and insure the relative positioning of each probe w.r.t the other ones, probe #1 and #2 are actually located on top of probe #3. Thus, to dither, we need to move only probe #3 (#1 and #2 will move with it). Probe 3 can span the whole 2 arcmin field. Probe 1 is limited to the top-part, and probe 2 to the bottom part. See this page for all the details about probe positioning, and observing strategy.


The Atmospheric refraction across the visible (400-1000nm) is quite large. Thus the star images formed on the apex of the TTWFS pyramid (quadcell) will be somewhat elongated when working at large zenith angles. The use of an ADC will compensate for this refraction-induced elongation and will increase the SNR on the TT WFS.

TTWFS (Quadcells)

Each of the TT WFS is in fact a quadcell. It is made of a 4 faces-pyramid, which apex is located in the infinity focal plane. Each face reflects the light into a small doublet, that focuses it onto a fiber. The light is transported through the fiber to the APD itself, located outside of the optical table. There are 3 TTWFS, each with 4 APDs, thus a total of 12 APDs. The APD does photon counting, and passes the counts to the RTC, at an adjustable rate (multiple of the LGS WFS sampling period, i.e. 800, 400, 200Hz if the LGS WFS works at 800Hz).

Slow Focus Sensor (SFS)

Because the Sodium layer altitude can vary slowly across a night, the science image would slowly drift out of focus, if only the LGS WFS focus information were used. To prevent such a drift, we need to sense the focus error on a nearby guide star. We have choosen to split the light from the TT probe#3 (about 70% for the TT WFS and 30% for the Slow Focus Sensor). The SFS provides a focus error every few seconds, which is sent to the LGS WFS zoom mechanism. Here is how the thing works:
  1. Let's assume we start with everything in focus
  2. The Sodium layer drifts by 200meters up
  3. This is seen by the LGS WFS
  4. The system (DMs) compensate immediately for this defocus (normal close loop, using default control matrix)
  5. The NGS seen by the SFS will go out of focus (Bbecause this only corresponds to a sodium layer characteristics, and not a true defocus in the system, the NGS did not undergo any real defocus. However, it sees the defocus induced by the DM).
  6. The SFS will report a defocus
  7. From this defocus, the system will drive the Zoom corrector
  8. Because of the zoom corrector motion, the LGS WFS will see a defocus and will correct it
  9. Assuming a proper calibration and sign, this will eventually converge and both the SFS and LGS WFS will be happy and with zero average focus.