Date: Thu, 17 May 2001 10:12:25 -0700
From: Russell E Owen
Subject: Re: Pointing with the spectrograph (was 2.5-m ObsLog 20010515
At 9:35 AM -0700 2001-05-17, Russell E Owen wrote:
>At 8:38 PM -0400 2001-05-16, jeg@astro.princeton.edu wrote:
>>I think it very likely indeed that we will want an instrument block
>>per cartridge. I do not know what the nominal tolerances are, but there
>>are a LOT of parts between the kinematics and the plate, and I would
>>be very surprised if the plate is located better than 10 mils or so
>>with respect to the kinematics. That is 250 microns, about 4 arcseconds,
>>or the radius of the guide fibers, and a big chunk of the 10 arcsecond
>>pointing accuracy.
>
>We may want an instrument block for each cartridge, and we certainly
>can do it, but I suggest we try without, first. I believe French
>went to some trouble to align the clamps on the cartridges so they
>would all be the same.
I'm afraid I did not explain why I'd like to start with one
cartridge. The issue is that it takes an hour or so of on-sky time to
obtain the data needed for an instrument block. And that's for an
imaging instrument with a reasonable field of view -- the cartridges
may be worse. So it will be costly to obtain 9 instrument blocks.
Once obtained, we would set up each cartridge as a separate
instrument (or possibly as a separate view for the base instrument
"spectro" -- either would do fine), and ask SOP to do a
cartridge-specific SET INSTRUMENT command for each cartridge. This
part quite reasonable, though it does have some costs, including a
bit of new code in SOP and introducing another possible source of
error into the system (darn, it picked the wrong cartridge).
Hence I believe we should try one instrument block first. If that
works well enough that stars are almost always found on all
cartridges, great. If not, then the extra trouble to obtain
individual instrument blocks will probably pay off in increased
observing efficiency.
This also gives us a chance to learn how to measure an instrument
block on one cartridge first. Then, with any luck, we'll learn some
tricks that will make it easier to measure the others.
-- Russell