NIRC 2 and OSIRIS Interfaces

Message sent from R. Dekany to J. Larkin, and K. Matthews:

Hi, James, Keith,

In considering a point design for NGAO, we've been wondering what hard constraints, if any, should be imposed on delivered F/# and pupil location. One of our working Level 1 requirements is that NGAO work with OSIRIS and NIRC2, but it's not clear, at this early date, that this strictly means presenting these instruments the same delivered F/# and pupil location. A more global systems engineering approach may weigh modifications to your instruments (by 2011 they may need/wish some overhaulin'), multiple relay output F/#'s, or other approaches.

To help us understand what's easy and what's not, can you comment for your respective instruments on 1) the impact/difficulty of handing faster working F/#'s (say, down to F/9?), and 2) the impact/difficult of handing a telecentric pupil location (e.g. pupil at infinity).

Similarly, how OSIRIS and NIRC2 might deal with different-than-current field curvature could also impose constraints on our initial survey of NGAO options. Can you comment on what could/couldn't be done to deal with different delivered field curvature? (If my terminology is to call the native Keck field curvature 'negative', then NGAO is likely to prefer a more negative FC than delivered by the current AO relay, though not guarenteed.)

• Response from Keith Matthews:

Rich,

Any of the changes would require new collimator optics. The change in scale would require different masks in the focal plane. Right now these are flat. The would be some what more difficult if there were larger field curvature, but not impossible. Theoretically the changes could be make without taking the instrument out of service. That is in about a month or less of down time. This would allow operation in imaging mode. For spectroscopic mode, there might have to be a subsequent focus adjustment that would take another thermal cycle.The true cost considering design, manufacturing, testing, installation etc. I would guess at between $500K to $1.0M. That might sound high, but I think that it is realistic if everything is considered.

Keith

-- RichardDekany - 17 Apr 2006