NRES performance

The Network of Robotic Echelle Spectrographs

Site Dome(s) Telescope(s) Camera Installation First Science Obs Current status
LSC (Chile) C 1m009 fl09 March 2017 2017-09-22 available for science obs
ELP (Texas) A 1m002 fl17 September 2017 2017-11-08 available for science obs; exposure meter not working
CPT (South Africa) B 1m013 fl13 November 2017 2018-08-20 available for science obs; exposure meter not working
TLV (Israel) A Wise fl18 July 2018 2019-03-22 available for science obs; exposure meter not working


The targets for commissioning observations are typically isolated stars with V magnitudes between 3 and 8. Fewer observations are made of targets in the range 8 < V < 12. Based on the spectra from those targets, the signal-to-noise per resolution element is estimated from the order of the spectrum that includes the Mg b lines (5167 - 5184 Å). The results are scaled to an exposure time of 60s, and then a model is fit to the points with the highest quality spectra (i.e. successful acquisition and clear sky conditions).

The models for the LSC-NRES, CPT-NRES, and ELP-NRES units are shown in the graph below. Although the scheduler regards the NRES units as identical, the commissioning observations have demonstrated that their performance somewhat differs. Those differences are predominantly driven by the state of the feeding telescope, not the spectrograph itself. We recommend to set the exposure time based on the performance of the lowest throughput NRES system. 

nres throughput model


Beyond the exposure time, NRES observations are subject to typical overheads. The overheads we've included in our database are the following:

The total, 667s, will be added to the exposure time of the request. You must make sure that your time window can accommodate the sum of exposure time plus overheads.

  • slew & settle time: 180s
  • time to switch between calibration mode and "sky observing" mode: 20s
  • acquisition camera exposure time: 10s
  • acquisition image processing time: 400s
  • post-exposure readout time: 57s

Known Limitations:

  1. Target Acquisition: Successful target acquisition typically takes 3 minutes or less. However, in some cases, acquisition can take up to 10 minutes. While the NRES team is working to improve the acquisition efficiency, the overhead time is set to 10 minutes to increase the likelihood that observing requests complete. Also, because the target acquisition code sometimes fails to center the star on the fiber, some exposures have lower than expected signal-to-noise. The acquisition strategy is continually being improved.

  2. System throughput: Initial data indicate that all three deployed spectrographs have a comparable system throughput. Differences in mirror reflectivity and collimation status can partially account for the differences between the NRES units.

  3. Exposure Meters: The exposure meters in all NRES units are misaligned and not functional. The cause of the misalignment and possible mitigation efforts are under investigation. 
  4. RV accuracy: The RV precision currently achievable is of order of 50 m/s to 100 m/s. The limitation in the data processing, and with special processing, a precision of order of 10 m/s has been demonstrated. The design goal of 3 m/s RV precision appears achievable with the spectrograph hardware.

Shared Risk

We recognize that the instrument performance is not yet at the level of its ultimate capability; we expect to continue to deliver significant improvements. We welcome your assistance in that effort. We encourage you to send us your criticism and your suggestions for improvements: Send email to and put "NRES" in the subject line.