NRES Status Update

NRES Overview

• Fiber-fed, Littrow design with double-pass prism cross-disperser
• Accepts starlight from one or two 1m telescopes
• Resolution (R) ~ 53,000
• Almost full spectral coverage 380 nm - 870 nm in one shot
• Temperature and pressure stabilized for 3 m/s instrinsic stability
• Continuous wavelength calibration with ThAr
• Funded in part by NSF grants from MRI and ATI programs

We plan to deploy 4 NRES spectrographs to 4 sites: (Chile, Texas, South Africa, Israel) in 2017.  In Chile and South Africa, each spectrograph will be coupled to 2 X 1m telesocpes.  The Texas and Israel sites have only 1 X 1m telescope each.  By the end of 2017, we thus expect to have 6 X 1m spectroscopically-capable telescopes, 4 in the southern hemisphere and 2 in the north.

We have parts for 2 more NRES spectrographs, but we have not yet worked out a schedule for deployment or a decision on sites.

Update as of June 1, 2017

Commissioning of NRES-1

The first NRES instrument (NRES-1) is complete, installed in its environmental chamber and attached to two 1m telescopes at CTIO (Chile).

The spectrograph performance parameters are close to the design targets.

Spectral dispersion and the dimensionless resolution R (the ratio of wavelength to the FWHM measured for ThAr emission lines) vary across each echelle order, but not from order to order; this is a consequence of the large incidence angle of light on the grating. The measured linewidths are about 4.15 pixels, almost independent of wavelength.  The resolution R is about 44000 at the redmost end of each order, 55000 at the center, and 72000 at the blue end.

To estimate throughput, we measured the number of detected electrons per 4.15-pixel resolution element for two moderate-brightness Sun-like stars: HD76151 (V=6.0, Teff=5500) and HD115383 (V=5.22, Teff=6050).  Scaled to V=10.0 and a 10-minute integration time, these measurements give:

These numbers suggest better performance than our TESS target of S/N = 30 for V=11 at an integration time of 1 hour.  Throughput will improve when the telescopes' optics are recoated.

We have obtained spectra of stars from 0.4 < V < 11.5 with exposure times 5 sec < Texp < 1 hr.

Environmental chamber servos have been tuned remotely since we left the site.  Present temperature stability is better than 5 mK on timescales of 1 day.

Problems encountered:

1. When we unpacked the main CCD camera, we found that the coating on the outer surface of the dewar window had begun to peel from the glass. We have never seen this before, and the cause of the damage is undetermined.
2. Alignment of the exposure meter optics drifted after the environmental chamber was sealed, causing a loss of sensitivity. We need to open the chamber so we can realign these optics. We are lab-testing a permanent mechanical fix for this tendency to drift. As of May 30, this problem appears to be solved.
3. About 10 days after we left the site, a circuit in the CCD controller clock board failed, causing us to be unable to read out the CCD. This problem was traced to a power supply failure. The supply has been replaced, and as of 16 May, the CCD is back on line. We are still working on various readout and minor noise issues.
4. Transitioning to full site software operation is proving challenging and time- consuming.

Assembly of NRES-2

NRES-2 is assembled and performs well in the lab.  It has been moved into its environmental chamber (igloo).  It will be packed for shipment to Texas in mid-June.

Development of Data Reduction Pipeline

The data reduction pipeline corrects raw CCD images for bias, dark current, and stray light, extracts 1-dimensional spectra covering 67 echelle orders and 2 or 3 fibers, identifies about 800
ThAr emission lines from the calibration fiber and uses them to estimate a high-accuracy wavelength solution, and computes a radial velocity estimate. Results are written to a compressed tar file which is stored in a data archive. A searchable database of metadata will allow quick location and retrieval of data for any selected objects or observing times.

Work is proceeding on a queue system that will run the pipeline on an LCO server, so that the NRES pipeline will be invoked automatically for each incoming data file, performing near-real-time reduction of NRES data. We plan to make the pipeline available in three successive versions.

(v1 - due Aug 1) Performs automatic calibrations, extracts to 1-dimensional (per order) spectra, estimates wavelength solution from ThAr spectrum. Performs a rudimentary radial-velocity analysis, meant (at first) largely as a testbed for later improvements. Produces diagnostic plots and a tarball of intermediate and final data products.

(v2) Provide user access to and analysis of data from the exposure meter and from the autoguiders; also from the radial velocity analysis.

(v3) Perform an automated classification of the target star in terms of the parameters {Teff, log(g), log(Z), vsini}, and add information about this analysis and its result to the output tarball.

Schedule for remaining installations