CYCLOPS2 Commissioning - July/Sep 2012

Tue, 4 Sep 2012 00:25:28 +1000

The latest CYCLOPS Report Aug 2012 is available as a PDF file - the image below is the front-page summary. Here are the main points

Summary of what’s good
• CYCLOPS2 bundle has been fed through its fibre run. CYCLOPS1 remains potentially available, but all future CYCLOPS runs should be changed to CYCLOPS2.
• CYCLOPS2 IFU and CURE guider are closely enough to confocal that no further work on this front required (within 0.1mm of top end travel)
• CYCLOPS2 Slit placement and rotation has been finalised, and is acceptable as is focus wrt to the UCLES slit entrance.
• Deployable calibration pick-off mirror is reliable all over the sky.
• GUIDER binning and temperature control works.
• Have verified we can acquire and guide on V=14.0 stars.
• FITS headers (with two exceptions) from CURE are being correctly created.
• Relative fibre throughputs measured using the Moon as a flat screen (with the telescope spiralled out across the surface during the observations to smooth out craters etc). Did this 5 times and get the same relative throughputs with a scatter between observations of a few percent. Then comparing these to the relative throughputs from July (using dome flats) I get the same answers to within a few percent.

Summary of what’s not good
• CYCLOPS2 pneumatic mode scrambler does not work. It is never going to produce the level of fibre agitation required – it has just too much damping built into it, and acts more as a fibre “massager” than a fibre “agitator”.
• A modified version of the CYCLOPS1 fibre agitator (likely needing a different eccentric motor or other arrangement to generate more shaking) can deliver significant modal noise suppression, and should be used for CYCLOPS2 use in future.
• The CURE lamp system produces a ThAr ‘spot’ at the f/8 focal plane that is 20” across. The CuR and FeAr spots are slightly larger and the Quartz spot is much larger.
• The CURE lamp stage is unable to reliably position its lamp spot on the f/8 focal plane – over the course of the run the spot was seen to move about by ~10-20” (ie by a whole spot size). The stage moving the lamp unit is either losing steps or suffering backlash, or just not positioning correctly. In any case, its positioning accuracy and reliability is not accurate enough to meet CURE requirements.
• Additionally, even once at a position the CURE lamp stage wobbles by enough to cause lamp illumination of CURE to move by +-1.5” as telescope moves around sky. This demonstrably comes from visible slop in the lamp box on its stage, and so should be able to be reduced by at least a factor of 10 with some suitably improved engineering (like a brake that actually clamps the stage down when it arrives in its position).
• CURE MIRROR keyword is always IN
• CUREGUI does not provide any feedback on status of CURE MIRROR, or LAMP position or whether lamp is on other than via the radio button controls. So when saaps sends drama commands direct to the CURECALIBLAMPS task, we get no feedback on status.
• There is no hardware support for knowing what position the calibration flip mirror is in, and what position the lamp slide is in.
• GUIDER frame averaging does not work – we need this for guiding on bright stars.
• Rearrange filters in CURE Guider filter wheels that position 0 (the default that the camera is in when it starts up) has the most commonly used filter (V), rather than the least commonly used filter (U)!

WHAT NEEDS TO HAPPEN
• CURE lamp stage needs to be tested as a single unit, built into CURE with all its cables connected, and demonstrated to reliably position the lamp spot on the CYCLOPS2 beamsplitter. This means these tests and fixes need to be done at Site (i.e. not just testing the stage by itself in Epping).
★ It needs to reliably reposition
★ It needs to stay in position.
• The accuracy with which the stage positions the lamps goes from good at the quartz lamp end to awful at the ThAr lamp end. So one obvious change to make is to swap the FeAr and ThAr lamps around. This also has the advantage of putting the two lamps used by CYCLOPS closest together so less time is spent waiting for the stage to move.
• GUIDER Frame average needs to be made to work (as it works for all the other apogee guiders)
• Hardware support for mirror position and lamp position needs to be added.
• A CYCLOPS2 fibre agitator (ie an updated version of the CYCLOPS1 fibre agitator) is needed. Additionally, it should be mounted on the wall outside the pre-slit area, not hanging from a clamp in the preslit area.

Also of potential interest is the following movie which shows some modal noise suppression results from an interim vibrator, and from just shaking the fibre bundle by hand. Quality is not great, since it was taken with a phone, but you can see the modal noise, and the degree to which different approaches smooth it out.

CYCLOPS Commissioning - Feb 2011

Mon, 28 Feb 2011 16:25:53 +1100

CYCLOPS' final on-sky commissioning run took place in Feb 2011. Thanks to a yet more hard work by AAO and AAT staff the flexure problem seems to have now been resolved. While we could not take data on sky (due to torrential rain), back-illumination tests showed the flexure between the guide camera and the fibre head to be unmeasurably small (ie. < 0.05" on sky).

This achieves our final goal, and the CYCLOPS LIEF grant has been wound up. CYCLOPS is now freely available to all AAT observers!

CYCLOPS Commissioning - Sep 2010

Wed, 13 Oct 2010 15:57:16 +1100

CYCLOPS' second on-sky commissioning run took place on Sep 24-27, 2010. Thanks to a lot of hard work by AAO staff (especially Luke Gers and Stan Miziarski) the optical bench is now well aligned with the telescope. Unfortunately, there are some new and problematic flexure issues that need to be addressed in the guide CD mounting. But they'll be dealt with soon. A report can be found here.

CYCLOPS Commissioning - June 2010

Sat, 26 Jun 2010 11:47:17 +1000

CYCLOPS' first on-sky commissioning run took place on June 18-22, 2010. A report can be found here.

Pre-Commissioning Tests

Thu, 27 May 2010 11:54:11 +1000

Pre-commissioning tests were carried out at the AAT during the daytime over the period 23-28 May, 2010. A report on what was found can be obtained here.

CYCLOPS News

CYCLOPS2 Commissioning - July/Sep 2012

CYCLOPS Commissioning - Feb 2011

CYCLOPS Commissioning - Sep 2010

CYCLOPS Commissioning - June 2010

Pre-Commissioning Tests