in its environmental
chamber in the lab at the
University of Chicago.
Credit: Andreas Seifahrt
for the MAROON-X team.
Efficiency curves for the
from the fiber exit to
the focal plane. The
measured efficiency for
the as-built spectrograph
(arrows) are lower limits
only due to limitations in
resolving the blaze peak
of each echelle order in
setup. A model based
on the measured
throughput of all
is sho wn as the blue
line. For comparison we
show the theoretical
predictions from an
optical model with
minimum and best-effort
specs for the throughput
and efficiency of all
The delivered optics
exceed the best case
expectations over most
of the bandpass of
the blue arm and lend
confidence in achieving
similar efficiencies in the
red arm, currently under
construction at KiwiStar
African Large Telescope. Table 1 provides a
summary of MAROON-X’s properties. We in-
tend to bring MAROON-X to Gemini North as
a visiting instrument beginning in 2019.
In January 2017, KiwiStar Optics delivered the
core spectrograph and the blue wavelength
arm to the University of Chicago. This has
been installed in a chamber with temperature
control to better than 20 milliKelvin (Figure 1).
The spectrograph is currently undergoing
an intensive test and calibration campaign.
All the expected characteristics of the spec-
trograph (e.g., resolution, scattered light,
and efficiency) have been confirmed with
The spectrograph’s efficiency in the blue
arm is particularly impressive, with peak
throughputs from the exit of the fiber feed
to the focal plane of over 60% (Figure 2). Ini-
tial testing is being done with only the blue
wavelength arm implemented and a smaller,
off-the-shelf 2k x 2k e2v detector in place of
the final 4k x 4k custom STA science detector
systems that will be used on the telescope.
Orders for the red arm and the final detec-
tor systems have been placed and delivery
is expected for mid-2018. One of the detec-
tors is a thick, deep-depletion CCD that of-
fers quantum efficiencies of over 90% out to
900 nm to fully exploit the high throughput
of the instrument and to suppress fringing
which would otherwise limit the achievable
radial velocity precision.
The primary wavelength calibrator for the
instrument is a stabilized Fabry-Perot etalon,
traced to the hyperfine transition of rubid-
ium. This device delivers a comb-like spec-
trum of about 500 bright and unresolved
lines per spectral order with frequencies
traceable to a few cm/s (Stürmer et al., 2017).
In addition, an automated solar telescope
delivers solar light to the spectrograph, to
test and improve the data reduction and ra-
dial velocity analysis pipeline delivered with
the instrument (Figure 3).
First tests with the etalon calibrator demon-
strated that even over the limited spectral
coverage of the smaller and less stable lab
detector system, the science and calibration
fibers track each other to better than 20 cm/s
over timescales of minutes to days (Figure 4).
The high line density and exquisite stability
of the etalon allows for unprecedented sta-
bility vetting and calibration at a level other-
wise offered only by a much more complex
and expensive laser frequency comb.