Revised December 10, 1999
Each wheel of the triple wheel assembly consists of
a metal belt driven wheel with a dual closed loop encoder for position
control. The motor and drive system uses a Galil 50-1000 motor to drive
a 50 to 1 harmonic drive. Additional drive reduction exists due to the
drive wheel to filter wheel diameter ratio. The ratio between the two wheels
is 32.9 to 1 (drive to driven). The second loop of the servo system is
encoded by a 5 micron Renishaw optical encoder. For the filter wheel dia
of 790 mm the Renishaw encoder provides 2.6 arc sec of angular resolution.
The primary servo loop which uses the 4000 counts per revolution Galil
motor encoder has a resolution of .2 arc sec per motor encoder count.
The actuators are designed as three independent cartridge
assemblies that mount in the lower tubes of the main frame. The line of
force of each pair of collimator support struts passes through the node
in the main frame when focus is at the nominal setting. When the actuators
move from nominal focus the small moments transferred to the main frame
have no noticeable effect. Each assembly is mounted to the inside of the
tube with six push-pull screws to allow alignment of the actuators to the
optical axis of the instrument. A complete cartridge assembly is composed
of a THK High Precision ball screw (6-mm lead) slide driven by a Galil
servomotor (4000 count/ revolution encoder) through a 100:1 gearbox (figure
8 and 9). Included in the assembly is a 1/10-micron Renishaw linear encoder
that forms the second branch of the dual servo loop. The assembly also
includes primary as well as secondary limit switches and mechanical hard
stops. Due to the large reduction through the gearbox and lead screw, six
motor encoder counts equals one Renishaw count. This servo system is capable
of repeatable positioning to 0.04 of an arc second of collimator tilt over
the entire 50-mm range of focus.
The pre-dispersion prism static support uses three
tantalum pucks that are bonded to the glass as a connection point for the
six struts that support the prism. The support struts use machined cross
flexures to insure that the 5 degrees of freedom other that the axial direction
of the strut are sufficiently soft to insure that the structure is statically
determinate. The cross flexures also provided the necessary range of motion
for alignment of the prism. The prism is surrounded by a non-contacting
safety cage that catches it if one or all of the bonded tantalum pucks
The ESI grating is supported in a front defined cell.
Adjustable hard stops in the side- wall of the cell provide the defining
points for translation of the grating. The grating is held against the
defining points (front and side) by spring plungers which act at points
on the opposite side of the grating blank. The grating and cell attaches
to the Optical Sub Structure with a three point kinematic mount. Each of
the kinematics is adjustable to provide tip and tilt alignment of the grating
to the spectrograph. Once the kinematic mounts are adjusted for alignment
the entire grating assembly can be removed and replaced manually without
realigning the grating.
Post dispersion Prism:
The post dispersion prism is support by a determinate
space-frame structure that is identical to the support for the pre-dispersion
prism. The space-frame structure is mounted to a steel plate that rides
on two THK linear slides. The stage drive consists of a lead screw, gearbox
(10:1) and Galil motor. The position of the prism is defined by the reference
fiducial and the motor encoder. An electro-mechanical brake is used to
hold the prism in place. This brake prevents the ball screw and gearbox
from back-driving when the slides are in the vertical direction. The brake
must be energized to release it. This provides safe operation in case of
sudden power failure. Each motor encoder count corresponds to .8 microns
of linear motion.
The camera shutter uses two blades to insure even exposure
times to all pixels on the CCD. Exposure linearity is approx. 1 percent
across the CCD. The shutter has two titanium blades that are moved by separate
Bimba air cylinders. Shutter logic control is hard wired in the control
box to determine which blade to move when a command to change state is
given. Typical blade opening or closing duration is approximately 140 milliseconds.
The CCD dewar holds approx. 4 liters of liquid nitrogen
to cool the 2k x 4k devices to û120 C. Hold time for the half-full
vessel is approx. 20 hours. The CCD is supported in the dewar housing by
a hexi-pod structure. This structure provides a very rigid support and
thermal stand-off for the CCD. The flexure of the cryogenic container is
isolated from the CCD by a flexible cooper braid that connects the CCD
to the cold finger. The vacuum housing with cryo container is kinematically
mounted to the back flange of the camera. The kinematic mount has 3 adjustable
screws to tip, tilt and piston the CCD for alignment to the focal plane.
The dewar can be removed from the camera and reattached to the kinematic
mount without disturbing the dewar alignment. The alignment of the spectra
on the CCD is also adjustable by means of a cam which rotates the dewar
about the optical axis. Like the piston and tip/tilt adjustment, this is
a one time only adjustment that is preserved if the dewar is removed.
The imaging mirror is a cell-less optic that is support
via 3 kinematic balls that are attached to the mirror by bonded invar pads.
The mirror is held loosely is a transport frame by springs which provide
the pre-load on the kinematics when the balls engage the v-grooves. The
transport frame is attached to a THK linear slide and a cam follower which
are propelled by a ballscrew, gearbox, and motor combination. The mirror
uses the same drive system and brake as the other two translation assemblies
in ESI (Post-dispersion Prism and Low Dispersion prism). The gearbox ratio
and screw pitch is the same for all three assemblies. Like the Post-Dispersion
Prism the stage encoding is provided via the motor encoder. Safety clips
are included to catch the mirror in the event a glue joint should fail.
Low Dispersion Mirror:
The imaging mirror is identical in size and design
to the imaging mirror. The drive mechanism and kinematics are identical.
The guider system for ESI uses a 200 mm Canon lens
and a Photometrics PXL camera with a 1kx1k (24 micron pixels) detector
to produce a 4x4 arc minute FOV. The guider FOV includes a 1.2X4 arc minute
portion of the instrument FOV with the remainder of the guider field being
located on a fixed mirror adjacent to the slit masks. Focus of the canon
lens is controlled remotely by a motor and timing belt driving the manual
focus ring on the camera. The guider system also has an eight-position
filter that can accommodate 50 mm square filters. The holders and filter
mechanism is identical to the filters systems used in HIRES.
The ESI calibration system consists of two separate
systems. The first is selectable projector for Neon, Argon, Mercury-Neon
and Quartz continuum source. The second is a projection system for the
Low Pressure Copper argon Line lamp. Each system consists of a lamp, an
optical illumination system feeding a flexible light conduit and a projector
The ESI instrument hatch slides open and closed on
4 linear bearings that ride on solid hardened shafts. The hatch is moved
by a double acting Bimba air cylinder which is controlled by the same skinner
4-way valves used on the camera shutter. The inside of the hatch is coated
with a specular paint to reflect the cal lamps into the slit. The hatch
is part of the Upper Structure of the spectrograph. This attaches directly
to the bearing and does not transfer moments into the science structure
of the instrument.
All electronics for ESI are contained in one of 3 cooled lockers. Each locker contains a radiator equipped with two cooling fans to re-circulate chilled air to cool the electronics. Two boxes contain the Galil servo electronics that control the motion stages. The third electronics locker houses the CCD crate and the electronics for the PXL guide camera.
The boxes are part of the exterior structure of the
instrument and form part of the insulated shell of ESI. The mass of the
boxes is transferred to the bearing by a structure that is completely independent
of the structure that supports the optics.
Cable Wrap System
The cable wrap system presents the electrical, optical and fluid lines to the rotating portion of the instrument by containing them in a single tube that wraps around a track on the instrument. This tube is pulled into a tubular shell by a constant force spring pulling on a wheel. This shell attaches to the cassegrain module and extends out close enough to the telescope mirror cell to clear the telescope structure. A total rotation travel of 540 degrees is obtained by turning ¾ turn either side of the attachment point on the instrument. The total travel is minimized by reducing the path length around the instrument with a triangular shaped spool and by rotating either side of center instead of wrapping 1 » turns in one direction on the spool.
The tube contains 2 power cables, two » in cooling hoses, one 3/8 in air hose and one zipper tubing containing 4 fiber optic and 5 coax cables.