IO

Updated 29th September 2011

Introduction

IO (Infrared-Optical) is a replacemet for the RATCam and SupIRCam cameras, with the eventual aims of providing wider fields of view, improved image quality and the ability to simultaneously image in both the optical and IR bands. The instrument consists of four fundamental components:

• IO:O - an optical imaging camera with a 10x10 arcmin field of view.
• IO:I - a near-infrared (JH) imaging camera with a 6x6 arcmin field of view.
• IO:THOR - a fast readout optical camera with a 4x4 arcmin field of view
• IO:BEAM - a beam stearing system incorporating dichroic mirrors and tip-tilt stages

The instrument is being built and comissioned in stages. At present IO:O and IO:THOR are fitted to the telecsope and a simplified IO:BEAM system in place.

Current Status

Updated 28th February 2012

• NEW: Fringe frames are available to download below. These are preliminary so please let us know how you get on with them.
• The current implementation of IO:O suffers two problems (low blue QE and high scattered light) that depending on your science targets and observing conditions may adversely affect your data. A plan is in place to address these problems, however in the meantime users should consider on a case by case basis whether or not the instrument is best suited to their programmes.
• Using the default Aluminium fold mirrors, overall throughput for RATCam and IO:O are the same at r-band. At longer wavelengths IO:O throughput is higher and for shorter wavelengths RATCam wins. The cause of this has been traced to the detector itself. We do not recommend using IO:O in u-band at all. For the Bessell B filter we also suggest using RATCam unless the extra field of view is absolutely vital. You should also be aware that g band and shorter data passed through our pipeline will still have some residual features from the flat fielding. At these wavelengths the QE contains large (up-to 50 per-cent) variations on scales of a fews tens to one hundred pixels. These variations are very much larger than encountered in traditional flat-fielding and leave few-percent residuals in the reduced image. Procurement of a new array is underway in order to address these QE problems which we aim to install before the end of the Semester.
• At all wavelengths, but particularly noticeable through the redder, broadband filters, scattered light onto the detector from elements of the telescope structure inside the primary mirror hole causes large scale linear structures to be superimposed on the images. These structures are a function of the Cassegrain axis rotator angle. The resulting image appearance looks similar to bad flat fielding, however this is an additive rather than multiplicative effect. The effect is obviously worse around full moon when the sky is brighter and you should consider only observing with the moon below the horizon . It should be noted that this problem does not affect the flat fields that were used to reduce the data, as these are constructed from a mean at a large number of different rotator angles, which serves to wash out the effect. For observers with point or small sources, the implication is that they should use a local rather than global sky solution for sky subtraction for doing their photometry. For large sources that fill the field of view there is no easy solution apart from building up a large sample of data and implementing some sort of iterative subtraction of a smoothed sky. A new baffle structure will be fitted to the telescope in early May to fix this problem.
• There is currently an option to use an over-coated gold feed mirror instead of the Aluminium mirror. This gives roughly 7% better throughput in the I and Z filters.
• Only 2x2 binning (0.303arcsec/pix) is being offered. Full frame 2x2 binned readout time is 17sec so for short integrations of bright sources, RATCam may remain more efficient if you do not require the enlarged field of view. No windowed modes are currently available, although we aim to deploy by mid-Semester.
• In short exposure frames a low level residual horizontal stripe structure in the bias level is visible. This is well below the general read-noise of the array and should not generally be a concern.
• Finally we note that IO:THOR is available for use with a single OG515 filter defining the blue response and the CCD cutoff the red response. No dichroic is fitted, so simultaneous imaging with IO:O is not currently supported.

IO:O

Specifications

Updated 20th Feb 2012

IO:O is equipped with a large diameter iris shutter. This results in a different exposure time between the centre and edge of the field of 80ms. For resonable photometric accuracy exposure times should therefore be greater than 10-20 seconds.

*Note the quantum efficiency figures are as supplied by the manufacturer. In practice using the current detector we find much lower figures blueward of about 5000 Angstroms.

Filters

IO:O has a single, 12 position wheel. Sloan, Bessell B and V and a set of H-alpha filters are available:

Standards

Eventially regular (every 2-3 hours) photometric standards will be obtained in the broad band filters. The standard fields to be used have not yet been determined and this programme is not currerently executing. Users requiring standards should ensure they are included in their own observation requestes.

Sensitivity

The exposure time calculator was updated with sensitivity estimates based on engineering data. Sensitivity was expected to be similar to RATCam apart from in the u-band (where IO:O should have around 3 times greater sensitivity) and the z-band (where IO:O should be around 2 times more sensitive). Using the current detector these sensitivities are not obtained.

Pipeline

A data redution pipleine similar to that for RATCam will be available

Fringe Frames

Preliminary Fringe Frames obtained 9th Feb 2012:

• Z band z-fringe.fits.gz
• I band i-fringe.fits.gz

Observing Overheads

These will be similar to RATCam in the long term. Using the current detector however the readout time is 17 seconds.

IO:I

Specifications

IO:THOR

Specifications

Filters

IO:THOR has a single fixed OG515 filter. However the incoming light will have passed through either the "Red" or "Blue" dichroics in the IO:BEAM system, giving two possible passbands ("blue" ~5150 - 6700 Angstroms) and ("red" ~6700 - 9000 Angstroms) where the long wavelength cutoff of the "red" option is defined by the CCD QE cutoff.

Standards

Regular (every 2-3 hours) photometric standards will be obtained in the broad band filters simultaneus with IO:O. The standard fields to be used have not yet been determined.

Sensitivity

Sensitivity should be similar to RATCam.

Pipeline

A data redution pipleine will be available.

Readout Speed

Depending on window size, readout speeds vary between around 0.5 seconds and 0.007 seconds. Fast timing experiments and lucky imaging are theirfore potential applications of THOR. In addition it will eventually serve as a generator of tip-tilt correction signals for the beam being fed to IO:O.

Observing Overheads

If you propose to use THOR, you must consult with the LT team prior to proposal submission in order to determine overheads etc.

IO:BEAM

Specifications

The IO:BEAM system contains a number of mechanisms designed to flexibly route the optical beam to the IO:O, IO:I and IO:THOR instruments in a number of different configurations. The final system will comprise

• A deployable optical/IR dichroic mirror that can be placed in the beam feeding IO:I, allowing optical light to pass through to IO:O and IO:THOR
• A deployable tip-tilt mirror that can be placed in the beam feeding IO:I and used for tip-tilt correction and beam-switching.
• A tip-tilt mirror permanently in the beam feeding IO:O and IO:THOR to be used for tip-tilt correction
• A three position mechanism allowing IO:O to be fed by either a mirror or short or long pass dichroics. When the dichroics are used, the beam will also pass through to IO:THOR.

The current system comprises the final two elements only.

A custom Phase2GUI panel will be provided to control these elements as part of your observing sequence. A screenshot of the prototype is shown below: